• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Medical Devices

  • Print
  • Share
  • E-mail

Transcript for Public Workshop - Magnetic Resonance Imaging Safety, October 26, 2011

UNITED STATES OF AMERICA

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

+ + +

CENTER FOR DEVICES AND RADIOLOGICAL HEALTH

+ + +

MRI SAFETY PUBLIC WORKSHOP

+ + +

October 26, 2011

8:30 a.m.

FDA White Oak Campus

10903 New Hampshire Avenue

The Great Room (Room 1503)

White Oak Conference Center, Building 31

Silver Spring, Maryland 20903

FDA: SIMON CHOI, Ph.D., M.P.H.

Senior Science Health Advisor, Office of Center Director, CDRH

TERRY O. WOODS, Ph.D.

Office of Science and Engineering Labs, CDRH

MITCHELL SHEIN, M.S.

Chief, Pacemaker, Defibrillation and Leads Branch, Office of Device Evaluation, CDRH

OWEN FARIS, Ph.D.

Deputy Director, Division of Cardiovascular Devices, CDRH

SUNDER RAJAN, Ph.D.

Office of Science and Engineering Labs, CDRH

WOLFGANG KAINZ, Ph.D.

Office of Science and Engineering Labs, CDRH

JANA DELFINO, Ph.D.

Office of In Vitro Diagnostic Devices

ERICA JEFFERSON

Press Contact, Office of Public Affairs


SESSION 4 : SCANNING PATIENTS WITH IMPLANTED MEDICAL DEVICES

DEVASHISH SHRIVASTAVA, Ph.D., University of Minnesota

TOBIAS GILK, M.Arch, RAD-Planning & Mednovus SAFESCAN

CHARLIE McINTYRE, Chickasaw Nation Medical Center

JEFF BURROWS, M.S.E.E., Medtronic Cardiac Rhythm Disease Management

SANDY WIXON, Medtronic Cardiac Rhythm Disease Management

ASHOK SARASWAT, M.S., B.Ed., RT(R)(MR), MRI Education Program Director,

Ohio State University Medical Center

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

NIELS KUSTER, M.S., Ph.D., Director, Information Technologies in Society IT'IS Foundation

FRANK G. SHELLOCK, Ph.D., FACC, FACSM, University of Southern California

JOHN WELTER, Medtronic

ROSS VENOOK, Ph.D., Senior R&D Engineer Neuromodulation, Boston Scientific Corporation

JIM OLSEN, B.S.M.E., Sr. Manager, MRI Technology, Medtronic Neuromodulation

DANIEL (JOE) SCHAEFER, Ph.D., GE Healthcare

JOHAN VAN DEN BRINK, Ph.D., Philips Healthcare MRI/MITA

JOSEPH OCH, American Association of Physicists & Medicine

ROUNDTABLE 2 : SCANNING PATIENTS WITH IMPLANTED MEDICAL DEVICES

MITCHELL SHEIN, M.S., Moderator, Chief, Pacemaker, Defibrillation and Leads Branch

OWEN FARIS, Ph.D., Division of Cardiovascular Devices, FDA

JOEL FELMLEE, Ph.D., Mayo Foundation

GEORG FRESE, Siemens AG

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

ROEE LAZEBNIK, M.D., Ph.D., GE Healthcare

JIM OLSEN, B.S.M.E., Sr. Manager, MRI Technology, Medtronic Neuromodulation

KAREN SMITH, M.Sc., RTMR, ACR, RTR, Vancouver Coastal Health

PAUL RUGGIERI, M.D., Cleveland Clinic

FRANK G. SHELLOCK, Ph.D., FACC, FACSM, University of Southern California

CHARLES STANLEY, RT(R)(CT)(MR), CIIP, President, Section for Magnetic Resonance Technologists (SMRT)

ROSS VENOOK, Ph.D., Boston Scientific Neuromodulation

TERRY O. WOODS, Ph.D., , Office of Science Engineering Labs, FDA


SESSION 5 : ACCESSORY MEDICAL DEVICES IN THE MR ENVIRONMENT

TOBIAS GILK, M.Arch, RAD-Planning & Mednovus SAFESCAN

JOSEPH OCH, American Association of Physicists & Medicine

ASHOK SARASWAT, M.S., B.Ed., RT(R)(MR), MRI Education Program Director,

Ohio State University Medical Center

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

ROUNDTABLE 3 : ACCESSORY DEVICES IN THE MR ENVIRONMENT

SUNDER RAJAN, Ph.D., Moderator, Office of Science and Engineering Labs

HEIDI EDMONSON, Ph.D., Medical (MRI) Physicist, Mayo Clinic Rochester

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

SONJA ROBB-BELVILLE, MSRS, RT(R)(MR), The University of Texas MD

Anderson Cancer Center

ANNE MARIE SAWYER, B.S., RT(R)(MR), FSMRT, Stanford University School

of Medicine

DANIEL (JOE) SCHAEFER, Ph.D., GE Healthcare

ANNA SRB, Kopp Development Inc.

STEVEN WEDAN, B.S., M.S., Imricor Medical Systems

TERRY O. WOODS, Ph.D., Office of Science Engineering Labs, FDA

WOLFGANG KAINZ, Ph.D., FDA

OTHER PARTICIPANTS :

MICHAEL STECKNER, Toshiba

VERA KIMBRELL, SMRT

JOHN KELLY

BRIAN SHULLS, Cook Medical

MARK RANSICK, Ethicon Endo-Surgery

BRAD STAGER, Biomet

MARY VAPORIUM, Sinai Hospital of Baltimore

KEITH KOPP, President, Kopp Development

FELICE SUN, Neuropes


INDEX

PAGE

STRUCTURE OF THE DAY AND LOGISTICS - Simon Choi, Ph.D., M.P.H. 230

DAY 2 INTRODUCTION: IMPLANTED AND ACCESSORY MEDICAL 230

DEVICES IN THE MR ENVIRONMENT - Terry O. Woods, Ph.D.

SESSION 4 : PUBLIC COMMENT ON SCANNING PATIENTS WITH 230

IMPLANTED MEDICAL DEVICES, INCLUDING DEVICES LABELED AS

"MR CONDITIONAL" - Mitchell Shein, M.S., Moderator

Devashish Shrivastava, Ph.D. 230 Tobias Gilk, M.Arch 230

Charlie McIntyre 230

Jeff Burrows, M.S.E.E. 230

Sandy Wixon 230

Ashok Saraswat, M.S., B.Ed., RT(R)(MR) 230

Emanuel Kanal, M.D., FACR, FISMRM 230

Niels Kuster, M.S., Ph.D. 230

Frank G. Shellock, Ph.D., FACC, FACSM 230

John Welter 230

Ross Venook, Ph.D. 230

Jim Olsen, B.S.M.E. 230

Daniel (Joe) Schaffer, Ph.D. 230

Johan van den Brink, Ph.D. 230

ROUNDTABLE 2 : SCANNING PATIENTS WITH IMPLANTED MEDICAL 230

DEVICES - Mitchell Shein, M.S., Moderator

SESSION 5 : PUBLIC COMMENT ON USE OF ACCESSORY MEDICAL 230

DEVICES IN THE MR ENVIRONMENT - Sunder Rajan, Ph.D., Moderator

Tobias Gilk, M.Arch 230

Joseph Och 230

Ashok Saraswat, M.S., B.Ed., RT(R)(MR) 230

Emanuel Kanal, M.D., FACR, FISMRM 230

ROUNDTABLE 3 : ACCESSORY MEDICAL DEVICES IN THE 230

MR ENVIRONMENT - Sunder Rajan, Ph.D., Moderator

ADJOURNMENT 230


M E E T I N G

(8:30 a.m.)

DR. CHOI: Good morning again. If you can start taking your seats, we'll begin shortly.

Good morning and welcome to the second day of the MRI Safety Public Meeting. I'm Simon Choi, the Office of the Center Director, Center for Devices and Radiological Health.

We look forward to another day of lively scientific discussion on the topics. The format for this public meeting will be to hear presentations from the public on the following topic areas: implants and accessory medical devices. During each public comment session, each speaker will be limited for six minutes. Following each open public session, we'll have a roundtable discussion among the experts in the area.

Written comments may be submitted to the public docket for this meeting, which may be accessed online at www.regulations.gov, and the docket number is FDA-2011-N, as in Nancy, 0640. If you run out of time, you may refer attendees to your written comments that you submit to the docket.

Members of the press or media should contact Erica Jefferson from the Office of Public Affairs. Erica, if you could stand up? Thank you.

This meeting is being recorded for transcription purposes. Each presenter and roundtable participant should identify themselves before speaking.

Same as yesterday, we would like to keep this a scientific discussion and not a promotion of a particular service or product.

At this time, I would like to invite Terry Woods up to the podium.

DR. WOODS: Good morning, everybody. It's now my turn to wonder if the screen's going to stay up or down this morning.

First, before we get into today, I wanted to start with some data that David Grainger, who worked for the MHRA in the U.K., that's the U.K.'s FDA equivalent, he sent me some data. He was on the phone yesterday and said that he would be happy to share this with us. This is their incident data from '94 through 2011 with scans per year and incidents reported per year. And he said that, in the U.K., the incident reporting, he feels like it's pretty comprehensive. So, again, just kind of food for thought.

This one is interesting because it shows incident per million scans per year, and in 2002, the U.K. released their MR safety guidance document, the MHRA. If you haven't seen it, it's a really nice document. So they released it in 2002, and it looks like their number of incidents per million scans started going down and they updated it again in 2007.

This, he kind of constructed on the fly with some data from a couple of the graphs he saw yesterday, the U.S. data. So, again, I can't make any comments about these numbers, except for the fact that, again, this is the incidents reported per year in the U.K., and this was his interpretation of the data in the U.S.

This, that you can take a quick look at, are the types of incidents that they had. Again, I guess, from one of those graphs, it looks like we should maybe be talking more with our friends in the U.K. to see what they're actually doing to decrease the number of accidents or incidents in MRI. Again, just a little food for thought that, but it's not really relevant to today.

Simon, if we could go into the other presentation?

So I'm going to introduce today, what we're talking about, implants and other devices. I'll start with general concerns, the FDA goals, a little bit about standards and labeling, and then what we're up to today.

A lot of this, we heard a lot of yesterday and don't have to spend a lot of time on, but MR safety concerns, there are projectile concerns produced by the static field and the spatial gradients. There are artifact concerns. The thing about artifacts is, sometimes you do want them if you're doing image-guided surgery and you see where your biopsy needle is. Other times, you don't want them, and many of you in the audience know a lot more than I do about things that you can do once you know an implant is safe to improve the image that you get.

Safety concerns are also produced by the pulsed gradient magnetic fields, dB/dt, by the RF fields and by the combined electromagnetic fields. RF heating is probably one of the trickiest concerns. There has been a lot of work done into appropriate testing and modeling to determine RF heating, and again, we have had severe injuries and burns caused by RF heating in various devices. Here's a couple of active implants.

So to talk a little bit about specific concerns for specific types of devices, you can divide the world of devices into devices that you intend to go into the scanner and devices that you don't intend to go into the scanner. There are both passive and electrically active devices in both of those places. Given the type of device and the place that you want to put it, there's different specific concerns. For everything, you have to worry about force and torque. For passive devices, you also need to worry about RF heating, image artifact, and for long implants, electrical stimulation of the heart or neurological tissue.

Electrically active implants has been the subject of a tremendous amount of work by ISO/IEC Joint Working Group over the last, I guess, almost five years. It doesn't seem like that long, but it has been. You can see that they've done the best job that they can to define all the additional concerns that are caused that you have to worry about when you have an electrically active implant that you want to put into a scanner.

Because there is a tremendous variation in active devices, if you have an active device that you want to come up with MR conditional labeling for, we encourage people to work interactively with the FDA as you define your test plan in the first place so that we're all on board with the testing that's going to be done before you actually complete it.

Let's see. I guess that's more of the same, again. I think this is supposed to say not intended to enter, things that aren't implants, but are intended to go into the scanner. Okay. So this would be things like EEG leads or pulse oximeters. Again, you probably need to at least consider all of the factors that you have to consider for an active implant, but again, work interactively with FDA so that we can come up with the best set of tests for you to do before you actually do your testing.

For devices that aren't intended to enter the scanner, for passive devices, again, really you have to worry about projectile accidents, force, and torque. For electrically active devices, again, they're susceptible to all the effects for things that need to go into the bore of the scanner, except RF heating. So you may not have to actively test for all of these things, but you need to consider them as you come up with your test plan.

In addition, there's some electrical equipment that you can only operate within a certain distance of the MRI scanner, either because the equipment itself will become a projectile hazard or perhaps there's some kind of electrical interference between the equipment and the scanner that would either affect the operation of the scanner or that would affect the operation of your equipment. For that kind of equipment, we need you to take great care in deciding how you can physically fasten down this equipment so that it doesn't go beyond the place that it's supposed to go.

So, what's FDA's goals in all this MRI testing and labeling? We want to ensure that MR safety is safely demonstrated based on the best available science, and we want to assure that that information is communicated to people that need it, the medical people at the MRI sites through device labeling, and we want to do that in an efficient, least burdensome way.

What do we ask sponsors to do for devices? We want you to use a risk-based approach to address potential hazards that are caused by the interaction of your device and the MRI scanner's electric and magnetic fields. The hazards are those that I've listed before, and they depend on the nature of the medical device and what you want to do with it in the MRI scanner and where you want to put it. Again, as I kind of outlined before, the testing analysis you have to do very substantially depending on what kind of device it is and what you really want to do with it.

So then we ask you to label your device for safety based on the testing analyses that you did, and whenever possible, we encourage the use of consensus standards.

The first standard that I'm going to mention today is the standard that contains the safety terms and icons. It's ASTM F2503 that was published in 2005. And if you notice, the scope of the standard is Marking Medical Devices and Other Items because the intent of the group that worked on this standard was that you could come up with a set of terms and icons that could be used for anything, not just medical devices. You can use it on your buckets and fire extinguishers and other things that you might need in your MRI suite. The goal in developing this terminology was to help prevent accidents.

So these are the terms that hopefully everybody has seen together with their associated icons. The icons come in black and white and color versions, and again, the terms are MR safe, MR conditional, and MR unsafe.

An MR safe item is one that poses no known hazards in all MR environments. To paraphrase Manny, the only way it's going to hurt you is if somebody picks it up and throws it at you in the MRI suite. MR safe things are things that are nonconducting, nonmagnetic, and nonmetallic. So if anything includes any metal at all, it can't be MR safe.

An MR conditional item is one that's been demonstrated to pose no known hazards in a specific MR environment with specific conditions of use. So you need to list the conditions that need to be satisfied. That would be things like the field strength, the spatial gradient, some statement about the RF fields or SAR. It turns out that most devices are MR conditional. So MR conditional is not a detrimental label to put on something. All it means is you need to look a little bit further to see where you can actually safely use this device in the MR environment.

MR unsafe is something that's known to pose hazards in the MR environment, you know, ferromagnetic scissors. There's a lot of things that are MR unsafe.

Then there are other standards for implants and other medical devices that many of you have worked with us to develop. They're ASTM standards for measuring displacement force, image artifact, RF heating, and torque. Then there's the brand new ISO technical specification, 10974, that deals with active implants.

As far as labeling goes, the FDA has recognized the terms and the icons in ASTM F2503, and we ask people to use them. There's a guidance document that scope is passive implants that talks about testing and labeling. The information in that guidance is useful for other devices in addition to passive implants.

This is some information about MR conditional and MR unsafe implants. Again, the purpose of doing this is to give MR facilities information that will help you make an informed decision about whether to scan a patient with an implant. A critical factor in this is making sure that the labeling information that you use to make the decision about scanning a patient is the information for that specific manufacturer model of device. You can end up with serious problems if you say this guy's got a pacemaker; here's some pacemaker MR labeling. If it works for that one, it must work for this one. That's absolutely not true. You need to make absolutely certain that you've got the information for the specific manufacturer and model number that the patient has.

One thing that I think is probably not widely used, but one thing that patients can do, is we ask the implant manufacturer to recommend that a patient register the MR conditions under which their device can be scanned with MedicAlert. The language is awkward because, again, FDA regulations deal with device manufacturers, not with patients, so we can't directly go to the patients, but we ask that the manufacturers recommend that the patients register their information with MedicAlert or some other equivalent organization.

Hopefully you have all seen labeling for a specific device in a format like this. This is, I guess, the standard format that we ask for generally where you say that nonclinical testing has demonstrated that this device is MR conditional and can safely be scanned under the following conditions, listing at least the static field, the spatial gradient, and some statement about SAR, preferably with normal operating mode or first level control mode.

We also ask you to have some kind of statement about image artifact, and again, this is just to give the people at the sites information that will let them know if it's worth doing this scan or not. If the artifact is going to obscure the region that they're interested in, then there's no point in doing the scan.

So what we're going to do today is look at -- we have two different sessions: one for scanning patients with implants. What we want to hear from you is how you guys make a decision about whether to scan someone or not, what information do you need, what information do you actually have, and where do you get that information.

We would also like to hear that about accessory medical devices in the MR environment. By that, we mean electronic equipment that maybe needs to be outside of the bore of the scanner that you need to use to do whatever procedures that you're doing. This also would be -- I guess we're seeing more and more requests for things that go into the scanner with the patient that are not actually implanted. So this would be accessory, often electrically active devices and pulse oximeters, various kinds of electrodes and various kinds of monitors of different physiological conditions. How do you make a decision about bringing these things into the MR environment, and again, what information do you need, what information do you actually get, and if you have a device that's restricted to a certain area in the MR environment, how do you make sure it stays there?

For those of you who aren't speaking or who have something else you want to share with us, again, as Simon said, if you go to this website, this is the docket number that you have to send your comment to. I apologize, I don't know how long the docket is open precisely, but we should be able to find that out and give you the day that it actually closes before the end of today.

With that, I guess we'll hand things over to Mitchell, who's going to moderate the next session of public comments.

DR. SHEIN: Good morning, everyone. I'm Mitchell Shein. I'm the Branch Chief for the Pacemaker, Defibrillation and Leads Branch. It's an honor to be able to join in with this group. In addition to my day job with Pacemakers and Defibrillators, I've been active with the joint working groups developing the technical papers to address this very topic.

We've got a robust discussion ahead of us yet this morning, a lot of people to speak, so what I will, as Terry has mentioned, restrict everybody to six minutes, try to keep things moving, and then we will take a short break and then come back for a roundtable discussion on the topic. Certainly, during the roundtable, I would encourage people in the audience to come up and get the mics.

I apologize for not being here yesterday, but did we have somebody collecting questions in the audience, Terry? Do we have cards to collect questions for the roundtable discussion? Okay. If you're shy and you prefer not to come to the mic and you want to slip me a note as to questions you would like to ask, I would be happy to share those with -- otherwise, please feel free to get up.

With that, why don't we move into the first discussion? Dr. Shrivastava, it's going to be a name I'm not going to be well with. If you would come on up and join us, that would be great. That's you. I'm sorry. Didn't see you standing by.

DR. SHRIVASTAVA: Thank you. This is a scientific discussion. We wanted to study radiofrequency heating near deep brain stimulation lead electrodes during MRI activity with the goal that maybe there are ways by which we can minimize heating the electrodes by changing clinical practice a little bit, and also, if we can develop computational ways of modeling and measuring and imaging heating near the electrodes.

In short, the goal of this presentation was to improve RF safety for imaging patients implanted with DBS devices by (1) investigating the role of the extracranial DBS lead placement on the heating near the lead electrode, this is the clinical part, and then the studying the feasibility of endogenous contrast-based MR thermometry on imaging heating to verify heating later.

The motivation came from the fact that clinically harmful RF heating can be produced near conductive medical devices, and there are no validated engineering tools exist to determine, which is predict an image, RF heating, that is temperature, near conductive medical devices in vivo. Safety depends on determining its spatial and temporal distribution of temperatures in vivo and not local and global SAR.

So for this particular study, we took three excised porcine heads, and we used Siemens 3T transceiver head coil, scanner reported SAR for these experiments was 3.164 W/kg. Sequence was TSE for 641 seconds. We used fluoroptic thermometry to measure heating at and near DBS electrodes, and we used GRE-based PRF thermometry and thermal modeling to determine heating at the electrodes. We used tissue because it is our experience that heating is significantly different in tissue versus in a phantom.

This is the typical example. This is a DBS lead. You can see the artifact. This 1 millimeter-wide lead was shown as 1 centimeter-wide lead for the thermometry sequence that we were using, and therefore, it was impossible to image any temperatures right next to the electrodes. But then we wanted to see how we can use thermal modeling and MR thermometry together to predict peak temperatures near the -- and I'll show you the results on that.

This is the typical heating response that we measured. We looped extracranial DBS lead on the top, on the side, on the back of the porcine head, and you can see that on the electrodes, there are two -- right at the electrode. This was a distal -- first electrode. This was the second electrode from distal, and this was 6 millimeters away because artifact range was 5 millimeters. You see that no matter how you loop these extracranial DBS leads, the heating was significant. However, if you take this lead and put it parallel to the RF coil so our clinical situation would be behind the neck and parallel to the coil, the heating was significantly minimized.

Then, because we realized that you cannot get any signal from, or the sequence that we were using, near the DBS leads, we did thermometry around it, and then we used our thermal modeling to predict the peak temperature. So what we said, okay, we can measure 5 millimeter away from this lead, and we can input the information, our thermal modeling, say, okay, so if, in 10 minutes, this is the temperature that we are measuring, what is the peak temperature, what should be the peak temperature that is needed to get this temperature? You'll see that the model temperature and the measured peak temperature were very close, and that's how we validated the feasibility of this hybrid modeling using MR thermometry and thermaling together. This here is the measured image temperature in MRI, and this is the simulated temperature in the MRI near the DBS lead. You can see the comparison there.

This is the challenges of developing MR thermometry near this DBS lead electrode, and that is, yes, we measured and imaged temperatures there, but that was very high. That example that I showed you, the peak temperature near the electrode was roughly 16 degrees, and that's why the temperature 5 millimeters away was close to 3 or 4 degrees that we measured. The challenges here is we develop our constraints for our MR thermometry.

So we said okay, if I want 1 degree temperature rise, which is the current limit for MR thermometry, the accuracy of MR thermometry is close to 1 degree. If we can identify an image 1 degree C, 6 millimeters away, what would be the -- sorry. This is not that one. This is the next one. Sorry. This one here is the -- so FDA center says we cannot have more than 1 degree temperature change in vivo, so this is the slide, what we'll need to detect if 6 millimeter away if, at the peak, we only have 1 degree C temperature.

So if the peak temperature change is only 1 degree C, the temperature that we will need to detect in a perfused tissue is close to 0.2, and that is the constraint that we have in MR thermometry. If this is what we want to limit ourself to, we will have to have MR thermometry, which can detect better than 0.2 degrees C accuracy in perfused tissue.

This here is another look at the same thing. If we are detecting only 1 degree C, 6 millimeters away, this would be the peak temperature and this is the scan time. So if you have a five-minute sequence, roughly, the peak temperature near the electrodes in the perfused tissue will be roughly 4 degrees, which will be 41 degrees.

In summary, RF heating is coil-dependent and was minimized when the extracranial DBS lead was placed parallel to the coil axial direction, in this particular case. Looping of the extracranial DBS lead produced harmful heating in all the cases. We show that the hybrid modeling is an attractive option for determining heating in vivo, and MR thermometry sub-degree C accuracy is needed to keep the temperatures below 1 degree C threshold. Thank you.

DR. SHEIN: Thank you.

Is Mr. Gilk available? Thank you.

MR. GILK: Thank you very much. I'm the same person I was yesterday with the same involvements I had yesterday.

Today, I would like to speak with you this morning about a number of issues related to the safety of imaging patients with implants. One of the things with respect to statements or guidance that I would like to offer to the FDA is that the safety of imaging MR patients with implants begins long before the patient arrives at the MR center. It actually begins with the testing and labeling procedures, both labeling for the MR device and testing labeling for the MR implant.

Currently, we image about 10 percent of the U.S. population on an annualized basis in MR. Ten percent of the population is roughly equivalent to everyone in the United States who self-identifies as being of Irish ancestry. So if we can apply the notion of a percentage of the population to an ancestral group or an ethnic group, I think maybe that might reinforce further the importance of proper identification of implants and devices.

With respect to such a large proportion of the U.S. population receiving MR exams on an annualized basis, it is inconceivable to me that it is permissible to introduce new products on the market, new implants and devices without minimum statements with respect to MR imaging and safety, even if that statement is to say this particular device, implanted device has not been tested for MR safety and there is no information available on it. Even that, as a minimum, is preferable to the position we find ourselves in today.

So if we identify prospectively the safety, or identify the fact that we have not identified prospectively the safety of an implanted device, then at the time of implantation, we need to make sure that the specific information about what has been placed inside the patient, be it active, be it passive, be it a "temporary implant" like hemostatic clips for GI procedures, that information needs to be entered into the patient's medical record and the information needs to be directly provided to the patient.

If the information is not entered in the medical record or provided to the patient, in my opinion, the procedure is not complete. If the documentation associated with a medical procedure is not complete, should we actually be reimbursing for these procedures? Should it not be a condition of reimbursement that the paperwork gets done? If it doesn't, in my opinion, makes perfect sense to withhold reimbursements for facilities that fail to provide this critical information in the patient's medical record and directly to the patient themselves.

Now, if we start with the manufacturer, we continue to implantation, now we come to the point of care where we're actually imaging the patient. If we do not have appropriate product labeling, both for the implant and for the MR because, ultimately, this is a question about the interaction of those two devices, if we don't have full and accurate information for both, that makes the risk assessment for imaging any specific device patient difficult and often impossible.

Even with complete and accurate information from one half of that equation, if we don't have the information from the other side, and in this particular instance, I'm referring to information from the MR system manufacturer in particular, for magnetic spatial gradient and the spatial gradient maps to determine the relative safety of imaging a particular patient with a particular implant in a particular anatomical location, then, again, it confounds the risk assessment by the physician.

I find it very curious that the existing disparity between implant testing and labeling and MR device testing and labeling make it such that clinical providers who have the exact same magnet that was used to demonstrate the safety of a particular implant and device, the information with which they're provided suggests to them that it is unsafe to scan a particular implant or device on the exact same magnet. I think this is a problem that needs to be addressed.

There is a lot more information, more technical information than six minutes affords us the opportunity to get into. This is an article by myself, Dr. Shellock, Dr. Kanal that is in AJR, the January of this year, one of the issues. Anybody who is interested in this, this is available for free, the full article online.

So with respect to point of care issues, one of the critical things that I would like to ask the FDA for is for any newly approved MRI device products, that the FDA require full magnetic spatial gradient and force product mounts. There are vendors who have released this information and they can do it in a variety of different formats, but I think it's essential that we simply get, not only the maximum value for a particular scanner, but that we get the information about the magnetic spatial gradient and the force product in the patient accessible volume.

So steps, require MR safety statements for all newly improved implants and devices, require the complete implant information to be entered into the patient's medical record and the patient provided with that information, require magnetic spatial gradient force product maps for newly approved MR system products and require regular MR safety training for MR personnel.

With that, I would like to thank you very much.

DR. SHEIN: Thank you.

Dr. McIntyre.

DR. McINTYRE: I'm Charlie McIntyre from the Chickasaw Nation in Ada, Oklahoma. The scanning of MR patients at my facility is especially difficult when it comes to conditional implants. As Toby brought out, my spatial gradient strength are patient accessible, so we are in the target range of those higher forces. The term conditional is still not very well understood, at least radiologists, technologists, and surgeons that I've talked about. Now, I hope to help alleviate some of that at the next Oklahoma Society of Rad Tech meeting. I'll try to present that and continue to reinforce the new terminology.

What I run into a lot is, when we get these conditional implants, I get a confused ordering physician, I get a confused and upset surgeon that put the implant in, I get an upset patient that had a successful MRI in the past and I have to explain that to them, that it's for their own good. Sometimes, they accept that. Sometimes, they storm off upset and mad.

The end result is, we all want to be safe. We all want to make sure, and unfortunately, when it comes to my particular problem with the spatial gradient, the wording that was brought out, by definition, by the terminology, that's all we have to go by as technologists, and the radiologists rely a lot on this.

My facility is not under any pressure to push patients through if there's any question, so if a condition is not met, we don't do the patient. We do not chance it, even if urged on by the provider. There are just too many unknowns. I would hate to see on one of the commercials, you know how you had the NSF, warning, warning, did you get MRI contrast. I would hate to see that, warning, warning, was your implant scanned on a scanner that did not meet the conditions?

Another thing: if the patient had a coincidental MI shortly after a heart attack, are they going to go back and investigate that magnet, like, see what changed about this patient? Oh yeah, they had an MRI. Then they may investigate further. Oh, was that MRI -- did it meet all the conditions? So those are just little, may be paranoia, but it's things that we have to look at it because they will find it.

In conclusion, I would recommend that, again, testing be uniform. Testing would include all types of magnets, be it vertical, horizontal, all orientation, possibly tested on an extra high field -- if it's possible. I don't know. Soup up the RF power. Make it higher than normal so that there's no question about the implant or device that we may encounter.

Thank you for your time.

DR. SHEIN: Thank you.

Mr. Burrows.

MR. BURROWS: Morning. I am Jeff Burrows. I work at Medtronic. In February of this year, FDA approved our first MR conditional pacing system, and I would like to just share a little bit of feedback that we've gotten from users specifically related to the MRI environment and patients getting their first MRI.

So the four items I would like to go through briefly this morning: number one, online training, patient monitoring, patient ID cards, and general contraindication from MRI manufacturers. You'll see some common themes from what we've been discussing yesterday and today.

First, training requirements. Interesting from the notes yesterday in our discussion about whether FDA can require training and just how much, but as part of our approval, we were required to implement training programs for both cardiology and radiology, and this is the verbiage that exists in our labeling as it stands today. We were reluctant to take on the training requirements because our feeling at the time, well, no other manufacturers had been required to do that and so this was something new and different for us.

I would have to say, after six months experience in the field, we've gotten very positive feedback from both cardiology and radiology on the training, the availability of the training. They're taking it very seriously. So our feedback or our feeling at this point is that we support MR conditions training for life-sustaining devices. There are some additional requirements related to scanning a pacing system that are above and beyond just the simple SAR management, magnet management, things like programming the device before and afterwards. So we're in full support of doing this.

The next item is monitoring. We've gotten quite a bit of feedback on this. We have, in our labeling, as you see here, the patient must be monitored, and this includes visual and verbal, ECG and pulse ox. This is an example provided by one of our physician advisors, Dr. Rod Gimbel, of a recent patient that he scanned. You can see here, both the pulse ox and the ECG. This is representative of much of the feedback that we've gotten that they're questioning the utility of the ECG in the MRI environment because, even when you have a high quality ECG, you can look on any one of those three traces for three different field sequences and see it would be very difficult to pick out the patient's heart rate. Very clear on the pulse ox.

Dr. Gimbel's comment was, if I had to take one means of monitoring with me to the desert island, it would be pulse ox, that ECG may be useful on some occasions, but it's really not the one that you can really rely on.

So the feedback we've gotten, number one, the MR technologist is really only responsible for verbal and visual. So the questions we've had are, well, who needs to monitor that patient, is it from cardiology, is it from radiology. That's really something that the clinic has to decide for themselves in their particular environment. We've had questions related to ECG. If the ECG is required, do I have to be able to interpret the ECG, and then what happens during the scan when there is so much noise that I really can't see what's going on? Our proposal would be that ECG really should be optional as we go forward, and to be specific, that the monitoring is for heart rate and not for ECG details.

Next item is patient ID cards. We've heard a couple comments on patient ID cards. This is two examples of a patient ID card. Because the system requires a complete implant, the pacemaker and two MR conditional leads, you'll see on the left, you see the patient ID card verbiage for when a patient has that. It does say that they have a complete system, but it doesn't say that this patient is ready to be scanned, and that was after much discussion with FDA in our final labeling that we decided to make sure that we said it's a complete system, but nothing more because we really do need to check further. On the right-hand side, you have a case, a very unusual case, but this would be a case where, if somebody had an MR conditional pacing system, but not MR conditional leads, it does very clearly state that they are not available for a scan.

We've gotten quite a bit of feedback. Requests to put an MR conditional symbol on the patient ID card itself. Could we indicate that this patient can get a scan, and then an MR unsafe symbol for the incomplete system? What we really feel at this point is that the ID cards really are sufficient as is, that the MR conditional symbol could potentially be misinterpreted as this patient is ready to go into the scanner. The medical record check is really always going to be required. Then we also refer to the manufacturers, our website, which has the up-to-date conditions on the website, so you should always check that as well.

Last item I have today is a conflict that we've seen. We've had at least one case where we had a call where a patient was refused a scan with our system, and that's because of the conflict that exists today between some of the MRI manufacturers' labeling, as far as allowing implants, and then MR conditional labeling.

So what we have is we have the MR conditional labeling for our system, an MRI machine contraindication for all implanted devices. The MR technologists called us. We said we have this labeling, the patient can be scanned. Then called the manufacturer's representative. The manufacturer's representative said it's contraindicated in our labeling, we can't recommend a scan, so the patient was refused a scan on that day.

Now, most of the MRI manufacturers, I think this is a situation that will relieve itself in the near future. Some of the manufacturers have already updated their labeling and others are in the process. So I think, in the near future, like I said, this will take care of itself.

DR. SHEIN: Mr. Wixon.

MR. WIXON: Good morning. I'm Sandy Wixon, and I'm an employee of Medtronic's Cardiac Rhythm Disease Management Division, and I've been involved in MRI technology and MRI safety on our products for the last several years.

What I would like to talk about this morning is kind of an extension of what Mr. Burrows just discussed, some of the things that we've run across since introducing our initial system to the market. With that, I would like to proceed.

We've already heard some discussion about spatial gradient, and we, I think, will continue to hear that throughout the day today. It's a spatial gradient and creates confusion between what goes into the device manufacturer's labeling, what the MRI scanner manufacturers put out, and how the MRI technologists or radiology community interpret that. So the spatial gradient determines the force that's going to be exerted on anything that goes into the magnet room.

The test method ASTM 2052 tells you to measure the deflection angle, so you hang your device from a string, you measure the deflection angle with a protractor, and from that, you determine the force. And what you're supposed to do is report the location in X, Y, and Z where that maximum force occurred and also the magnitude. For those that aren't familiar with it, that upside down triangle is shorthand for spatial gradient.

The FDA requires that spatial gradient be in the MR conditional device labeling, and in addition to that, scanner manufacturers provide maps of both the magnitude of B 0 and the magnitude of spatial gradients as a function of position around the magnet. Sometimes, however, those maximums can be behind the shroud, so there isn't consistency from manufacturer to manufacturer as to what's reported in terms of that spatial gradient.

So the issue is clinical confusion when somebody goes to scan. The device manufacturer looks at and tests and measures the spatial gradient of 4.5 T/m. The scanner manufacturer might report a much higher number, but it's not necessarily in a patient accessible area, and as a result, the clinic, the concern that there's such a disparity between the two numbers, could refuse to scan the patient when, in fact, as indicated earlier, it's the same magnet that was used to do the safety testing.

So the proposed solutions, there's actually two proposals that I'm making. One is, in the labeling, you only specify what the maximum spatial gradient was that you tested at. It could very well be that the force applied to your device is a tenth or a quarter of what the requirement is. That doesn't end up in the labeling. So one proposal is to be able to extrapolate the maximum force that was measured at that spatial gradient to indicate what the spatial gradient maximum would be for that device. Then the second piece of the proposal is for the scanner manufacturers to very, very clearly report the spatial gradient only in patient accessible areas.

The other topic I want to cover this, and I know this is going to be a topic for roundtable this afternoon or later today, is the marking of accessories for a pacing system. On our first system, the Revo system, when we conducted the clinical and we introduced it to the market, there were no requirements for accessory marking. Our second conditional system, when we started the clinical study, the FDA asked us to put MR unsafe markings on the programmer, the patient home monitor, patient self check.

To date, since we have released to the market, we have really not received any solid user feedback on the need for marking the accessory equipment. In addition, it never came up in the clinical study from any of the clinical centers. So Medtronic's proposal is MR unsafe marking is unnecessary. MR unsafe is typically assumed if there is no marking on it.

I want to be clear that Medtronic is very, very concerned about patient safety in MR environment, and we certainly don't want any of our equipment to be associated with a missile effect, but there are many things like a home monitor, an activator, or things like that that are not needed for the person to receive an MRI scan, and that would be no different than a cell phone or a radio or any other thing that a patient might bring into the MRI suite. So we feel that marking those as unsafe would be unnecessary.

Thank you.

DR. SHEIN: Mr. Saraswat.

MR. SARASWAT: Good morning again. We had a good session yesterday, and I'm looking for another exciting day today. After Dr. Woods gave that presentation, she gave a nice summary, so that was a refreshing review for all of us.

I would like to touch on a few things, and I want to share one of our case studies that we did at the Hofstra University Medical Center. We know there are a lot of devices and passive/active and all those things, and the technologist is in the middle most of the times. We debate that, what is scope of the technologist, and I do believe it's part of the technologist's scope to be in the middle of this. Those who don't want to deal with these issues, we believe in further training.

Yesterday, there was some discussion on training, what are the limits, what FDA can do or cannot do, and that was inspiring to me, but when we expect that the device safety and all those discussions, by the same token, there should be a recommendation that these people need to be properly trained and these devices need to be used by skilled people. So, again, the recommendation goes a long way. Somebody mentioned manuals. I know manuals have some redundancies and some funny things, but I do find them very useful. One has to, you know. There's so many sources of training, and training is crucial.

I don't want to use the manufacturer's implant, but I may have to use it. At OSU, we had a very successful case study when the Medtronic label system came. What we found was useful, that this manufacturer had an accident website, good program for technologist education, physician education, including cardiologists and the radiologists. We took the online test. I found the patient card information was very clearly visible.

I believe, in Ohio, we were the first one to go ahead and try it because I come from an academic medical center. The good news is that I was part of it. I was charged with this study, so I met with the cardiologists, and it was a good, successful meeting. Then I met with the pacer clinic, another successful meeting. This was done with the presence of Medtronic representatives. Then we did the technologists training, and we also did the radiologists training, so there was a lot of training that went on, but in the end, everybody knew we were on the same page.

When we did the first patient, I believe it was early April, there were a lot of experts available and a lot of help from experts in the field. We had the cardiologist, pacer clinic person, nurse available, of course, the technologist. All the specialists were there, and we monitored. We even put EKG, even though EKG is not reliable, pulse ox, all these things were done, and we had a good result, and we continue to do that. But my question is that we can afford that at the OSU Medical Center, but I still don't believe that the smaller centers have the same resources available. As we are doing more and more patients, we are feeling that we can do that, and I can see it can pan out down the line.

So, again, this was a good case study for us, and the lesson we learned is that technologists need to ask for the implant cards. Not all manufacturers, as I said yesterday, they have good information, useful information on the card. Mr. Gilk said that some of the manufacturers do not have reliable information. By that, I mean, I'm not insulting them. I'm just saying, the information, I can't use that. It's not meaningful to me. So it should be simple, comprehensible, and to-the-point information.

So there's a lot on the table, and I think this is a great example, and I believe the FDA is requiring now, I may be wrong on that, with the new applications, they require some training. That's just wonderful if it is. I'm not 100 percent on that, but the manufacturers need to provide that training and coming from an academic institution, anybody can -- there are multiple sources of the training, so we just need to figure out which training is useful and which is not useful. That's an ongoing process.

So with that, again, I want to just kind of summarize that we do not want to -- it's much easier to say we're not going to scan this patient, but with my experience in dealing with Dr. Kanal and Frank Shellock and a lot of people, you don't want to deny that care just by saying patient has implant, we're not going to scan this patient. It happens a lot of places. So the question is that who are we not going to scan. To me, the question is how can we do the benefit versus risk analysis with the help of the physician and go ahead and provide the care to that patient without putting the patient at risk.

So, again, as implants are becoming a norm, the MR community is aware of it, and we need to successfully scan these patients by monitoring and providing the care they deserve. So thank you so much.

DR. SHEIN: Dr. Kanal.

DR. KANAL: Morning. My name is Dr. Emmanuel Kanal. I'm the Chair of the American College of Radiology Magnetic Residence Safety Committee and the Director of Magnetic Resident Services at the University of Pittsburgh Medical Center.

In the issue of implanted devices, let's start first with what the problems, at least the identifiable problems might be, and let's go for potential cures for each one. Amongst the problems that I can think of are poor or uncertain history as to the presence of a pacing system in the first place in that patient. Another potential problem is the lack of ready MR safety-related information even when you do identify it.

So now you have a patient in front of you that you do know that there is an implant, and now you want to know can I scan that implant. Now, you're having the difficulty of how do you find that out, even if you know exactly what the make, model, and type is, where do you go for that information and what is the reliability of that information. Are disclaimers available, or is this something that's going to be not only binding medically, but legally?

Number three, a lack of understanding how to apply the information that might be available to safely scan or to make the decision not to scan that patient. And the lack of understanding may be because the implant itself may be poorly labeled and/or because the magnetic resonance unit may be poorly labeled.

Let's go through each of these. The first one, I believe, is perhaps the most difficult one for us to solve in the short-term, and that is, if there's a history as to whether this patient does or does not have an implanted device that is unreliable or unknown, what do you do? The American College of Radiology MR Safe Practice Guidelines has standards for unconscious patients, patients that are incommunicable or what have you, so we know that there are things that you can do.

(a) They are relatively burdensome; and (b) They're going to expose the patient to ionizing radiation if you have somebody that you need an emergent MR on and you need to find out if they have a certain implant on them or in them or in their brain or what have you. This takes time and this takes, in certain cases, irradiation, but at least let's standardize our approach for screening these patients.

Number two, far more important, this would be the eventual future's goal for us is to have an electronic health record which is accessible. Several solutions for this electronic health record exist or potential solutions. One of them can be that the patient carries the card, but I think everyone here who is a physician or a technologist recognize the unreliability of expecting to have that card, let alone for emergent cases or patients that just don't happen to carry them with them, unconscious patients.

Another possible solution, therefore, would be to have a central repository of some sort. Possibilities might be essential database for each patient. Another possibility that I would like to at least throw out here is for the manufacturer. As long as the patient permits it, have the manufacturer of that device maintain a central repository that is HIPAA compliant and that is secured, that will provide access to, Dr. Kanal is a patient of ours, a patient who has one of our devices implanted, and it was planted on such and such date, this make, this model, and beneath that needs to be updated about it. That manufacturer has all the information about what was implanted or information that might change about that implant. That's just some food for thought as to how we might centralize that data.

Another problem that we've dealt with that we've discussed was the, even once you identify that there is something there, we don't know information about that implant. So one of the suggestions that you've heard now several times, and I think is going to be a common theme, is that we would recommend that the FDA mandate MR safety testing, and more importantly, reporting of safety testing results for all devices that are designed and sold, distributed, marketed for implantation, or being attached to a body, an insulin pump, or a pacing system, or what have you. It's meant to be implanted or attached to a human; there has to be some MR safety testing information, even if that information, as you heard already from several, simply is we have not formally tested.

More than the physicians, more than the radiologist, every technologist who deals with this area recognizes the tremendous amount of time that is spent or wasted trying to get information about an implant and then you find out that no one's ever formally tested it. Even knowing that it was not tested at least helps us towards a risk/benefit decision.

Again, it's been about a decade now that we've already discussed with the FDA, and we haven't had any progress at all. There are certain clamps or clips, for example, the endoscopic vascular clips that we know are powerfully, strongly ferromagnetic. They're used every single day. There is no mandate that we report that these were even implanted. And as Dr. Gilk said, if we should mandate it, they would be on the medical record. The patient needs to be advised. The physician who requested that procedure needs to be advised that these are not MR safe, these are directly MR unsafe, and the patient needs to be aware that these have to pass. We don't know when they passed. We have to irradiate all these patients to clear it. This is still untouched in the past, approximately, decade. Nothing has changed.

The solution for, what's the information source that you go to. Right now, everybody's familiar with Dr. Shellock and other sources of information that say we've tested and this is what we've tested and this is what we've compiled from available testing. These compilation lists are great, as long as I can rely on them, but if the information changes, as you know, all of these sites' statements that disavow any legal responsibility for the information being disseminated, well, that may be great for the site, but it's not great for me because, in the courtroom, I will be alone. People say you're alone when you're born and you're alone when you die. Please take it from me. You're alone in the courtroom. The only response that I can see is, who should be held legally responsible for the information being provided that I'm going to make a risk/benefit decision on that patient? Clearly, the manufacturer of that device should be held responsible for that information and maintaining and updating it. And that should be done with every manufacturer of every implant or attached device. It should be mandated that they run a website that will give me all the current information about that site.

Finally, the last problem is implant labeling not understood by the operator, mandate MR safety education for the device operators, and standardize the device labeling that was to be used so that the end user -- the labeling will be written, as Ashok said, it should be written so the end user will understand it.

We also need to standardize the MR scanner device labeling itself. This is the MR unit now. That labeling has to be standardized. The FDA, for reasons unknown to me, mandates that they report the maximum spatial gradient. I actually have no idea why that is mandated. Why aren't we asking to report the maximum force product? Why are we introducing new terms that the end user doesn't need?

To the best of my knowledge, if I'm wrong, I really would like to hear in the roundtable discussion the FDA explain this. It may be that I'm missing something very basic, but it seems to me that what we're looking at is, number one, information for the patient and healthcare practitioners only. I do not believe it is something that we need to know because the service rep is going to be exposed. That's not why we're asking for this information. We want to know, is this device going to be affected by a certain force and, if so, what is that force and where will that be, at maximum?

So why don't we mandate that we have the correct information? Maximum spatial gradient reported today, as you know, is meaningless if it's not three dimensionally clearly identified. If we can't reach that area when we're in there, then what difference does it make what that is? Number three, that isn't the area in which safety testing has been done on devices. Since we can't reach that area anyway, if it's --

DR. SHEIN: Manny, could you bring this to a close? We're running out.

DR. KANAL: Thank you. Let's summarize. We also mentioned that there should be a low power or fixed operating mode which will make it easier to scan implants.

So, in summary, central repository of patient implant data, standardize the prescreening of magnetic residence of these patients, and mandatory MR safety testing and reporting even if you're just reporting that no testing was done, web-based safety device information, education center for all MR devices, standardized labeling, standardize the MR scanner labeling, and standardize fixed or implant operating modes.

Thank you very much.

DR. SHEIN: Dr. Kuster.

DR. KUSTER: Good morning. My name is Niels Kuster. I'm from the IT'IS Foundation and ETH Zurich. I was Chair of the SAP Group, defining the RF requirements of the implant testing in the new Joint Working Group, IEC standards.

So I'm going to present a study we did to test the shortcomings and the strengths of these guidelines. The study was conducted by the Medical Device Research Program, which is a pool in which we can conduct industry research. The plea I'm going to make here is that in the interest of the industry, in the interest of regulators and foremost in the interest of the patient, it's important that we do more such research because the study shows not everything has been solved because it's a new process. This process includes a substantial in silico evaluations, meaning computer simulation, which is new, I think, for the industry, as well as for the regulators.

The background, we heard it. We have continuously increased usage of MR scans. Patients are excluded. The ASTM standard that we have today for -- the work done in the working group showed that there are severe shortcomings of this standard. So it might endanger and underestimate the patient risk, and so this Joint Working Group I was talking developed these guidelines, these procedures.

The concept of these guidelines are based on first, generating an RF model of implants, by irradiation, experiments, or numerical simulations. This model is then taken into validation tests. It must be sure that these validation tests are sufficient. We have this model of the implant. Then we can go into in silico, into simulations, and perform the energy deposition at the lead by the presence of the implant in the MR. Then, based on that, the risk assessment is conducting either on temperature or on in vivo experiments, for example, testing thresholds.

This is a very simple, generic implant for deep brain stimulators which have been defined by this group. We went through the process here. There is a system in which the implant has been validated. The validation shows that the numeric model is good. That was here, simple because the model was designed so it can be easily simulated in full way.

So this shows you the system setup has been defined in contrast to the ASTM phantom, that we have very well defined excitation of the implant, of the structure so we know exactly what we're exciting on. This is the structure, so it moves a little bit slower. I apologize for that. Then, the energy is deposited here at the tip, so this is the concept and such function that the implant is tested first.

Maybe we can move on. So then this implant is taken. Then the excitations are evaluated, as Manuel Murbach showed yesterday, by using the virtual population, the different anatomies and then, exciting these different anatomies with the lead pulse shown here through the birdcages, different kind of birdcages. Then, for the lead pass, for this general lead pass, different lead trajectories are evaluated as shown on the right side, and then I have another movie which shows you a full 3-D simulation. Here, the head with this generic implant introduced. This is done for the different anatomies, and the energy, as you see, is then picked up by the lead structure and deposited at the lead tip.

So you see here, the energy is picked up, and then here, this is the temperature pattern which then is deposit the energy and can cause these injuries. We don't want to worry that such injuries will occur in the future.

So, in conclusion, MR safe implants can be comprehensive tested with a combination of experiment and in silico evaluation, that both are needed. If the model is correct, can only be validated by experimental tests. Experimental tests are always needed. Coverage of patient population for the different MR technology can only be obtained in silico, but need always -- basic tools, as shown here, are now available, but we need more for suitable for such a simple structure, but it's too demanding for complicated structure.

So we needed better tools and we need better procedures because I have seen that sometimes, companies tend to use nonphysical-based models and match them -- to the test configuration which then, overlapped with the validation, which is dangerous. Such a procedure is very difficult to detect the problems of that procedure by the regulators. So we need procedure better defined so we have sufficiently independent and sufficient number of validation tests. We need to assess how many models are needed, how many MR coils are needed. So much better defined procedures are required here, need to be able to define such that we can rely on the test and Dr. Kanal is not alone in the courtroom in future lawsuits.

So in the interest of the industry, in the interest of the regulators, because the process we have now that we have a trial error, so each industry has developed their own process going to FDA. FDA may detect or may not detect shortcomings, and we suppose that is far too long, too costly. So we need more industry research, in the interest, foremost, of the patient.

Thank you for your attention.

DR. SHEIN: Dr. Shellock.

DR. SHELLOCK: Morning. Frank Shellock, University of Southern California. I'm going to skip through a bunch of these slides in the interest of time and kind of cut to the chase. Of course, I'm going to talk about the MRI labeling issues for implants and devices.

One of the things I just wanted to bring up, this document, for one reason or another, 1997 draft document still exists as a draft still on the FDA's website, produces a lot of confusion to implant and device manufacturers who look at that and think it has some relevance today. In fact, it doesn't. This is the pertinent document that should be there and referred to.

Moving on with regard to the terminology, I think we're going to talk a lot about this today because there continues to be a lot of confusion, despite the fact that we made an attempt to basically walk everybody through the current labeling information and how it's applied to implants and devices, and that was published in radiology a couple years ago. You all know the terminology: MR safe, MR conditional, MR unsafe. Then you saw this template and you look at this template, and it does look very concise, very straightforward, shouldn't produce too many problems or issues for those of us that have to look at it, particularly the MRI technologists, with the possible exception of that term, the spatial gradient information. Of course, we'll talk more about that later.

Again, this is what the template is, as it's been proposed and presented by the FDA. A lot of issues, though, related to that because, as you'll soon see, there are many deviations from that template, and the labeling is just way too complicated, I feel. So the issue at hand is poorly worded, difficult to understand, and inconsistent FDA approved labeling exists. The issue is, the MRI healthcare workers, MRI technologists, and radiologists, if they don't understand the labeling, they either may make a mistake or they may not scan the patient, or there could be a considerable delay between the time they get the patient in hand and when they actually perform the MRI exam because of the confusion that exists.

So one of the things, as you already saw, there's an article that's, again, another editorial where we attempted to walk people through the information, the terminology, the highest spatial gradient, and how it's applied to implants and devices. Despite that article, there's still ongoing confusion, and the main issue is where we're actually setting up this test for the translation retraction assessment versus the specifications that the MRI users get from the manufacturers.

When you see that, it usually tends to cause even more confusion because there are two different places where the measures are made. The values don't easily relate to one another. I would be an advocate of abandoning the term high spatial gradients and probably get back to something a lot more simple. I don't know if force product is the way to go, but keep in mind though, at the present time, there are probably a few thousand implants than the labeling, the highest spatial gradient information is it's 720 G/cm.

This is information, actually, from an FDA reviewer that I've seen over and over again, where it says when a manufacturer device presents something to the FDA, they get comments back, typically and oftentimes, an FDA reviewer may say because the MRI scanners with higher maximum spatial gradients are becoming available, I don't know where these MRI systems are coming from. I don't know about them, but what is happening is the FDA is seeing these specs that are higher that are presented by the manufacturers and think that that information should be applied.

Anyway, so they make this comment, and your device has relatively low magnetically induced deflection force in the spatial gradient used for your testing, the FDA encourages you to perform additional analysis to determine if there was a larger spatial gradient for which your device can safely be scanned. Seems like a relatively benign comment, but in fact, all that does is produce a lot more confusion. When they deviate from something that the MR system users that healthcare professionals see, there's typically confusion and issues remain. They're not taking into consideration a recent editorial and the information on the high spatial gradient value, as it's been specified.

So now let's look at a couple of FDA approved labels that exist for some implants that have been recently assigned FDA approved labeling for MRI.

Here's one. This is a very simple stent. The spatial gradient value is reported at 2500 G/cm or less, which is a far nexus of what you may typically see for an implant, that 3T. The other thing that I found interesting is that stent heating was derived by using measured, nonclinical, in vitro temperature rises and a GE Excite 3T scanner and in a GE 1.5T coil in combination, and I don't know what that refers to, the coil, what type of coil, so on and so forth. Anyway, if we're not exact with our labeling, we're going to see some confusion.

Now, here's another example. This is another stent, and here is spatial gradient information, less than 100 T/m, and it has gradient. I'm not even sure what this means, this static magnetic field gradient product, less than 200T, T2/m values extrapolated. So when did we start extrapolating values, and again, a term that's going to produce a lot of confusion. It may be meaningful to a physicist, may be meaningful to some people, but it's inappropriate to have in labeling for the average MRI healthcare worker.

Then look at all this information. I mean your eyes glaze over trying to just get to the point of what all this other information pertains to, and it pertains to some of the heating scenarios and the heating aspects that are related to those, and it goes on and on and on. It's difficult to sift through to get to the right information. You know, are you going to scan a patient with that device or not?

Another good example, poorly worded, difficult to understand labeling. We've got the fighting chance of helping out the patient, at least we have 700 G/cm or less normal operating mode and what have you, but then there's all this additional, I think, superfluous information for the RF heating, a 3T and a 1.5. And there's this statement here about, it is recommended that the patients register the conditions under which the implant to be scanned safely with the MedicAlert Foundation or an equivalent organization.

So very confusing information, and I'm not a big fan of assigning -- you know, now you're asking the patient to interpret this information and present it to MedicAlert for one reason or another and register the information. So with regard to that matter, and this statement is another situation that we see a lot of times in comments back to manufacturers to put this in their labeling. It's not a good idea or a good recommendation.

Again, we cannot rely on the patients. I mean most of the time, the MR healthcare workers can't even sift through all that information and understand it, and now the patient is supposed to do that; they have to understand, interpret, and present the MRI conditions, which may be confusing to all of us. And there really is no known precedent for doing this, you know, and I don't know what the point or the motivation for having such a statement and information applied in the labeling. So that's all I have to say on that.

I thank you for your attention, and hopefully we will get some improvement with regard to how the issues of implants and device labeling is addressed, particularly with regard to the role of the FDA. And I would love to see in-house standards developed for MRI labeling and disseminate it to all the FDA reviewers so that everybody's on the same page, and perhaps you should create an expert panel with experienced reviewers to provide consistent reviews and comments to help out the device manufacturers and, of course, ultimately help out the MRI community who has to deal with these issues and problems on a day-to-day basis.

Thank you very much for your attention.

DR. SHEIN: Mr. Welter.

MR. WELTER: Hello. My name is John Welter. I'm from Medtronic. This is actually a joint presentation by several implantable medical device manufacturers that are part of the Joint Working Group that you've heard about several times today. In an effort to not repeat ourselves, we've actually developed a joint presentation and will each present our sections separately.

As is evident by your attendance here today, we are all interested in increasing patient safety, as well as access to MRIs. Existing MRI safety regulations limit the exposure fields based on human physiological responses, and each manufacturer can interpret those limits and use proprietary algorithms that adhere to those. This has been safe for un-implanted patients in the past and will continue to be in the future. However, this does result in uncertainty within the fields and makes device testing and regulatory approvals for implanted devices very challenging.

Our proposed solution is to develop industry-wide scanner modes that limit these electromagnetic fields directly and have well-defined values. The Joint Working Group and MT 40 have done work collaboratively, and the results are promising and should lead to good things in the future. The Active Implantable Medical Device Committee requests input from this group in implementing this and ensuring its success.

The motivation for this work is to increase access to MRIs for patients with implantable medical devices. The patient population with implanted devices is increasing. This includes both passive and active devices, and they range in size from small to large and in complexity from a simple metal object to something that includes electronics and can be affected by these RF fields in their internal circuitry. Then, also, MRIs are on the rise as well and have become the gold standard for many diagnostic scenarios. The majority of implants include, in their pathway in MRI, which is required during their implant process. Then scanning patients with these MRI conditional devices is complex, and we believe that using these fixed modes will help alleviate this issue.

So when discussing the physiological limits, the existing limits focus on the physiological responses, such as the specific absorption rate and the peripheral nerve stimulation threshold, and these can vary widely within a given patient as well as across the manufacturers and the way that they actually implement adhering to these limits. This plot here shows the allowable B 1, which is essentially the applied RF field as it varies when you landmark a patient at the head, which is shown in the front, and then towards the extremities towards the right of the plot. Then there's three lines there. Those are for three different models, Fats, Hugo, and Suzy, and it shows that for different human body compositions, as well as sizes, you can result in different limits as well.

Then the need for the fixed modes is that these limits are implemented very differently across the manufacturers, and this results in a large amount of uncertainty when you try to test or label for scanning with an MRI. The challenges that are presented to the regulators, as well as the device manufacturers, are that we must first determine each manufacturer's implementation and then, secondly, there's nothing that prevents an MRI manufacturer from changing their implementation while still adhering to the physiological limits.

Then this may force the AIMD committee to evaluate devices at extreme limits that are not clinically relevant and potentially reduce access to MRI exams or create additional conditions in the labeling, making that more complex. We believe that fixed modes add clarity to these levels, at which the devices are to be tested and labeled for.

At this point, I'll turn it over to Ross Venook from Boston Scientific to complete the presentation.

DR. VENOOK: Thanks, John.

I want to thank the FDA for holding this workshop and everybody for attending and being interested in this topic. As John said, my name is Ross Venook. I work for Boston Scientific Neuromodulation. I studied MR safety in graduate school and am continuing with this with Boston for the last five years, a long-term employment.

So John motivated why we're interested in fixing certain parameter values in MRI systems with the way that the MRI systems operate. I'm going to talk a little bit about which parameters we think are important and why, and then talk about the progress that we've made in establishing which values and which limits should be used for establishing safety for patients.

What you see here is chart that has the four primary values, and these are physical field values. Notice the units and micro-tesla and tesla per second that are really important. These are the critical values for the device manufacturers to assess and to understand the patient safety profile for patients with our devices. All of the implant manufacturers who have been involved in this work believe that these are the four values.

Importantly, the ones in the middle, the average RF power and the peak gradient switching fields, those are related to, although not necessarily easily related to the existing MR safety standards that are based upon physiologic limit. But the first on the list and the last on the list are not covered by those units, and so one of chief aspects of establishing fixed parameters is to enable us to cover these other important aspects for patients who have implantable devices that aren't covered for patients who don't have implantable devices. Why is this; just a quick motivation. The peak RF power, the B 1 peak, as we call it and it's related to the maximum induced voltages that you can have going into the circuitry and in the complex device. It's really important to be able to limit that. Right now, in MR systems, it's limited only by the maximum capability of the RF amplifiers and the MR system.

The last on the list, the average gradient switching field, this also doesn't have direct physiological impact on the patient, but the patients who have implantable devices, particular those that have, rather than the long antenna type wires that create lead heating, this is the flat surfaces and the cans that contain the electronics. And device heating scales, as approximately radius to the fourth power, and so as devices get larger and for larger devices, this is really an important consideration.

So why are we fixing the parameter values or why do we think that's the right solution here? The real key is to define the environment. In order to assess whether a device can be safe and under which conditions it can be safe, what you really need is to understand what the applied exposure fields are. That's the key, not the SAR values, but the applied B 1 field and micro-tesla. The purpose is to put the AIMD manufacturers, the MR manufacturers, and the FDA all on the same page about, these are the conditions, these are the fields that a patient is subjected to so that we can understand which conditions to use when testing our devices.

Once we have that, we think that this is going to be really powerful in increasing MRI access for our patients because devices will be able to be labeled more effectively, more easily, and more consistently. As has been described by several folks, there are a lot of issues when this issue gets to the MR tech, having a set of fixed parameter values that's consistent across the MR industry while labeling of devices that works in different scanners in the same way, to be consistent and to really enable MRI access safely.

What was the forum for this work? I won't go too far into depth, but the Joint Working Group that's been described previously had a subgroup that separated and said we're going to talk about fixing parameter values, what should those values be, how do we define these. This is a couple years ago. I'm extremely excited to report that we have made good progress there. We have a proposal that describes and fully defines the four parameter values that I mentioned earlier, as well as recommends limits that are appropriate.

The proposals that were put forward include a really solid proposal for a single mode at 1.5T that we think will work for everyone involved, as well as a roadmap that describes how to take this concept forward to 3T, and to lower values that are inherently safer in terms of exposure, and also to higher values that allow patients with devices that can handle it to take the benefits of MR at higher levels.

I want to acknowledge at this point all of the people who have worked on this, particularly the partners on the MR manufacturers' side. There's been a significant amount of work here just to get to this point, and I also want to further acknowledge the work that it's going to take in the future to take it from where it is today, which is potentially being implemented in a standard and giving people direction on what to do. Taking it the next step through implementation to where it's really getting to patients is going to take a lot more work probably from everyone in this room, and I hope you're as excited about it as I am.

So without further ado, I thought I'd share this or we thought we'd share this. This is the proposal that was made. These values are, I would call them near final, or we haven't fully decided on all these values yet. This is something that's still in progress, but this gives a perspective on what we're thinking in terms of this initial first fixed parameter value option. The important thing here, for those of you who aren't as familiar with all the technical details, is that what this would allow is essentially normal operating mode scanning for patients who have implantable devices, but in a way that the limits to the exposure are fixed and consistent across scanners and we think that it will provide a lot of value.

Again, we're excited about this first step, we, the active implantable device makers, and we're really looking forward to continuing working with all the different stakeholders as we do this. And as we hopefully can extend this proposal to work at 3T and to address patients and devices that need more restricted conditions or can have more flexibility.

At this point, Jim Olsen, a colleague from Medtronics, will make some concluding remarks.

MR. OLSEN: Hi. I'm Jim Olsen from Medtronic Neuromodulation, and first, I would like to thank the FDA for holding this workshop and inviting me to talk.

John and Ross have already covered most of the fundamentals here, so I just want to make a few concluding remarks. As already stated, current MR systems limit their maximum RF and gradient exposure based on physiologic response instead of the radiated levels. Each manufacturer can do this in significantly different ways, and there can be some very large differences in the maximum radiated power that's emitted by some of the scanners. This is an issue for testing.

We realize the incidents of burns and nerve stimulation on patients without implants has been incredibly low for many, many years. For patients without implants, this strategy has worked really well, and we acknowledge that, but for patients with implants, that this becomes a problem. As we all know, the number of patients with implants is steadily increasing. So implanted devices focus energy and, therefore, increase this risk of burns and nerve stimulation.

So we believe that when patients implanted medical devices, it becomes increasingly important for all parties, including the radiologists, the FDA, for everybody to understand what the radiated power levels are and the fixed parameter modes will accomplish this. So as Ross already said, the proposed mode for 1.5T limits the power levels to near normal mode, and what we're really doing is just eliminating some of the extreme cases that are possible on some of the scanners. This new mode would enable the vast majority of scans that are currently run in normal operating mode and, thereby, we believe that image quality will not be affected.

So the Joint Working Group which includes the medical device and active implantable medical device manufacturers and the MR manufacturers are close to finalizing the details for this first fixed mode. We believe that this is going to be a big step forward for scanning patients with devices, and we really want to thank the MR manufacturers for all their cooperation and support of this very important advancement that will enhance the safety of scanning patients with active devices. Thank you.

DR. SHEIN: Dr. Schaefer.

DR. SCHAEFER: Hi. I'm Joe Schaefer, GE Healthcare. I'm not going to go into a great deal of detail over some of the background since it's been extensively covered here, but just to mention that, of course, our concerns are heating from the RF and gradients, forces and torques from the magnetic field, and making sure that the implant function is not compromised.

One other thing that hasn't been mentioned much is sometimes there's confusion about whether a time varying gradient is the same as a spatial magnetic gradient, so we need to make sure that that's dealt with.

So let me see if we've covered most of this. Typically, the MR conditional limits will include things about the gradient slew rate per axis, the SAR level for induced heating, and then, the static field, both because of the SAR level, which may be different because the electrical length may be different, causing heating to change, and also for the magnetic forces and torque.

The forces that are produced in the magnetic field are related to the product of the magnetic field locally and the gradient of that magnetic field. This is generally termed the force product and the IEC safety standard requires a report of location and value of the maximum field, the maximum gradient, and the maximum force product. It's important to note that there's no translational force in a uniform magnetic field, but there may be torque that could be considerable, and that forces increases roughly with the square of the B field. Another thing, if we get into accessories later, is that -- or implants that are active, the cores may saturate and fail in the static field.

So the proposed safety improvements we would like to mention here, which I subtitled "Do Not Confuse the Operator," is all connected with trying to keep it simple for the operator because we're asking them to interpret a great deal more or fairly complex information to decide the safety of these implants and what have you.

So it seems to me that we need to standardize forms and the information on the forms, including the things like the units, the presentation of the data, and this sort of thing. And possibly IEC or MITA or ASTM or some other group may be able to do that with some cooperation from the FDA. We need to make clear static fields are not time varying fields, and we should consider using the force product instead to replace the spatial static fields because that's what the force is related to, and in addition, it's good at all field strengths. You wouldn't necessarily have to measure it in another system where the force of gravity would be the same as it would for the force product.

Let's see. We should extrapolate the static spatial gradients to the force of gravity on the implant. I don't think that's always done, and that would be a useful thing to know. If there are any maps for spatial gradients or force products, right now, there's some that are contour plots. There are some that are concentric cylinders. The maximum value on those, I think all that should be standardized so, again, an operator can look at it and quickly understand what's going on, and once they have a little experience, maybe we'll get past that problem.

So, in conclusion, it's imperative that the operator be provided this standardized information to interpret with no unit conversions some kind of consistent data layout. Understand this is between both what's provided by the implant manufacturers and the MR manufacturers so it's easy to look at it and understand it. And that we should replace spatial gradients with a force product, that major data should be extrapolated through what field would result and the forces equal to gravity, and that the field maps, if there are any, should be standardized in terms of units and type and that we may need some standard organization to be involved to undertake this effort, possibly ASTM or MITA or IEC. And I kind of ran through it, but that's all I got.

DR. SHEIN: And to bring us home, Dr. Van den Brink.

DR. VAN DEN BRINK: Thank you.

It seems that we all agree, more or less, at least on the problems that we are facing, and we are using the same words. I'm Johan Van den Brink from Philips, and we discussed and organized our presentations together with MITA.

As Joe showed you already, you can see that there's several pieces of information that play a role in MR conditional labeling, and we've been seeing a lot of these this morning. Often, it's really difficult to judge labeling information, as excellently explained and demonstrated by Frank Shellock, I think.

There are other problems related to this. For example, that MR manufacturers report magnetic fringe field in T/m while the FDA labeling asks for G/cm, and applying the factor that's in between there is not that easy. So apart from that, we do see a strong increase due to short and wide bore designs in static field gradients, and I think the tests will show this, are not really appropriate and correct. Furthermore, on the labeling, you'll find the main field strength limitations on the SAR and heating testing, and I'll say a few words on this because what's safe at 3T may be perfectly unsafe at 1.5T.

So we think that there is a need for consistent, concise, and complete labeling, and actually, at the end of the day, this may require relabeling for existing implants, even though there have been a thousand or more than a thousand labels already.

So coming to the static field gradients and effective forces, you need to carefully check the data in the MR conditional labeling information for an implanted device, relative to MR scanner information. MR scanner information originally wasn't intended for that use. It was meant to assist accessories getting to the scanner. So that's why we showed a maximum value at the bore location as it is, as has been shown this morning. Apart from that, of course, as Joe also mentioned, there is the confusion between strength fields and spatial encoding gradients, which are the switch gradients. So, yesterday, it's a lot of information to be dealt with.

I want to come to some proposals to include the ASTM standards for static field gradients measurements. What you see here is a typical drawing from the standards on the left, and on the right, a photograph from this editorial in Radiology. Well, you wonder, why is this device on the edge of the bore, and why is it kept by this -- on the sides.

Actually, the reason for that is -- and I'll move on to the next slide -- that the force that's supplied to the device isn't going into the bore, but is going into the magnets. What you see is the red arrow there. It's in the Zed-X plane, and actually, the intent of measurement is that the magnetic force is perpendicular to the force of gravity, actually. And these tests said that it's not. So in a way, this whole setup here, coming with the 720 G/cm statements, is probably -- well, should be improved, let's keep it like this.

I think it's easy to come with a proposal for this, so we have to ensure that the magnetic force is perpendicular to gravity and that it must be at the point where the device ends at the end of the measurement procedure. Then you measure the deflection angle and apply some basic mathematics, and you calculate the static field gradient to tell you that, which corresponds to one times gravity, and use that for labeling. That really would help a lot.

Then, of course, you can ask, well, you're all scanning patients with these devices, so why is it considered safe currently? Well, most of these devices have very low deflection angles, so there is no force, and actually, it works, apparently. But some of these devices may have a high force, for sure.

On the other side, the devices are mostly used near the bore center line, and as Joe said, some of us provide onion shell or bore liner, graphs like this. You look into the bore of the MR system, and we provide the force that the static field gradient values both in T/m and G/cm to avoid confusion for the different dimensions. So if you have a device which is labeled at 400 G/cm and it stays in your aorta, which is in the central -- more or less, you can be pretty sure that that's going to work.

Then coming to the main field and SAR issues, I want to make the point here that what is safe at cylindrical 1.5T may be quite unsafe at lower fields. And, also, what's unsafe at 1.5T may be safe at 3T. So in the labeling that's provided, the templates that we saw, it says 3T, and it states something about SAR. We are not fully happy with that because we would like to see that there is explicit labeling for 1.5T and SAR and heating, and 3T in SAR and heating in this template.

So there is one example here which is also interesting. I think it's close to a good example for labeling for one of our products from Philips. What you see here, it's pretty complicated still. There is one caveat here. In the testing information we have, it shows that the temperature rises after 15 minutes by 0.4 degrees in a specific location, which is specified in the labeling, but the labeling does not specify effects that is 15 minutes -- labeling information here. Today already, where the time is explicitly specified, I think that's extremely helpful to enforce that as well.

So, finally, our recommendations are to reconsider requirements for testing and implant labeling especially for the static field gradients or force products, ensure completeness and accuracy of the information on implant labeling and not by the many -- that you see from Philips examples, but -- ensure consistency between MR vendor and implant device vendor information. We all agree on that, I think, after this morning's presentations, but also, the expectation of the information requires education, which probably needs CME accreditation. At the end of the day, the final judgment for all of this, of course, is for medical staff. Thank you.

DR. SHEIN: Thank you very much to all of our public commenters this morning. We're actually a few minutes ahead of time, so I'm going to take the prerogative of the chair and give us an extra five minutes at break. We will be reconvening promptly at 10:45 back in the room to begin a roundtable discussion. Please formulate any questions you might like to in the break, but we'll see you back here in 20 minutes. Thank you.

(Off the record.)

(On the record.)

DR. SHEIN: Would like to get started again. I would like to invite the roundtable participants for this session to please come up to the podium. We've carefully arranged names in front of the seats, so if you could find yours and settle in, we'd appreciate it. Thank you.

(Pause.)

DR. SHEIN: And welcome back. This is the second roundtable session for the workshop involving scanning of patients with implantable medical devices, including those that are labeled MR conditional. We had some good robust discussion this morning, a few common themes. Before we get into a larger discussion now, I would like to give everybody up here at the rostrum the opportunity to introduce themselves. If we could start with my colleague at the far end and work our way back, that would be great. Terry?

DR. WOODS: Hi. I'm Terry Woods. I'm in the Office of Science and Engineering Labs, and I've been added to this roundtable as the person who should be the target for the spatial gradient discussions. I was involved in developing the standards and in writing questions, for instance, suggesting that you can do an analysis to come up with a larger spatial gradient that you can put in your device labeling. So let me have it. We're really interested in hearing what everybody has to say and, again, trying to come up with the best labeling that we can for these devices so that people can be scanned safely.

DR. FARIS: I'm Owen Faris. I'm the Deputy Director in the Division of Cardiovascular Devices at FDA, and I've also been something of a team leader in FDA's efforts to develop MR conditional labels for active cardiovascular implants.

MR. OLSEN: Jim Olsen. I work at Medtronic Neuromodulation, and I've been working on developing MR safe systems for six or seven years.

DR. VENOOK: My name is Ross Venook. I work at Boston Scientific Neuromodulation, and I've also been in MR safety for about 10 years.

DR. SHEIN: Frank, you need to push the button to talk.

DR. SHELLOCK: Why do you give me these high tech things on the Panel?

DR. SHEIN: I'll get to that in a second.

DR. SHELLOCK: Frank Shellock, University of Southern California. I've personally tested, in the last 27 years, about 3500 plus implants and devices. And thank you for the FDA for letting me be on the Panel.

DR. RUGGIERI: My name is Paul Ruggieri. I'm the Administrative Director of MR at the Cleveland Clinic Health System, head of Neuroradiology.

MS. SMITH: My name is Karen Smith. I would like to thank the FDA for allowing me to be on the Panel. I've been an MRI technologist since 1987, and I've been very involved in the Canadian Association of Medical Radiation Technologists and currently serve as a regional practice lead in British Columbia for MRI.

DR. LAZEBNIK: My name is Roee Lazebnik. It's my pleasure to be here on the Panel today. Thank you very much. I am the Chief Medical Officer for GE Healthcare MRI.

DR. KANAL: Manny Kanal, University of Pittsburgh Medical Center. I am the Chair of the American College of Radiology Magnetic Resonance Safety Committee and the Director of Magnetic Resonance at the University of Pittsburgh Department of Radiology. And I have personally tested Frank Shellock.

DR. SHELLOCK: MR unsafe, by the way.

MR. FRESE: Georg Frese, Siemens AG, Erlangen, Germany. I'm also sitting here as Chair of MT 40. You've heard the name MT 40 a couple of times. MT is standing for Maintenance Team, and the team is responsible for the MR safety standard, IEC 60601-2-33.

DR. FELMLEE: My name is Joel Felmlee. I'm a medical physics physicist from Mayo Clinic and Foundation.

MR. STANLEY: Hi. Charles Stanley, Stanford University School of Medicine, and I'm here as the President of the Section for Magnetic Resonance Techs.

DR. SHEIN: Welcome to everybody again. I'm Mitchell Shein. I'm the Branch Chief for the Pacemaker, Defibrillation, and Leads Branch in the Office of Device Evaluation.

We started out with two quick tests for the folks up at rostrum this morning, first with being able to identify your seat and follow where your name assignment was. The second one, most of you passed and benefited from Frank's mistake, and you can learn from others. I will remind you that you do need to press the button on your microphone to talk. This is a recorded session, and it's important that we get that so that it all comes through.

And with that, there are three microphones in the audience. I want to invite those in the audience to come up.

Before we get to that, one of the things that I heard in the presentations this morning was the need for full and accurate labeling. I was wondering, to start things off, if I could get a little discussion from the Panel, perhaps, about what would constitute full labeling? I heard discussion about having full information, but I think you can blizzard somebody with information and hide what's important in that.

So both from an MR system label perspective, as well as from an implantable device perspective, what would constitute the minimum appropriate information to allow a technologist, at the end of the day, to understand whether the patient with a device could be scanned? So I'll open up to the floor. Why don't we open it up to the technologists in the middle there? Dr. Kanal?

DR. KANAL: The idea of minimum labeling actually has two answers in my opinion, and that would be a wonderful thing to see, if we can expand on that here. The first one would be scientific, meaning we would like the scientist who is MR safety aware be able to make a determination as to the conditions under which the device was tested for a static field, for spatial gradient field and translational and torque related forces and the conditions of testing under time varying magnetic fields, both RF, as well as gradients.

All of that makes sense to be able to determine the safety of scanning, but in the end line, there's a second complete labeling requirement, and that one is the radiologist and the technologist. At some level, everything I just said is meaningless for way in excess of 99 percent of the radiologists that are out there. What we need is a quantum state; I can or I cannot scan this patient if I press this mode or this operating environment, this can be safely scanned or this can't, and that's part of labeling.

DR. SHEIN: Karen, if I could ask you, at the end of the day, you have a patient, maybe, maybe does not have their patient ID card with it, which may or may not include language as to MR susceptibility. They may or may not, if they have an active implantable, have a copy of their most recent programming session showing that the device was put into an MR mode, if that's available. And then you have an MR technologist in the room. I've had several MRs myself, and there's not always a radiologist in the room. It's a technologist and me.

So what do you need to see or what would help facilitate for you, as a technologist doing these scans, what information do you need to have at your disposal? What would make it easier on you?

MS. SMITH: Karen Smith. I agree with Dr. Kanal, what Dr. Kanal just said. I mean at the end of the day, what we need to have is some kind of a card saying whether it's safe or not for me to scan that patient on what type of mode I'm going to be scanning and at 1.5T or 3T. I would hope, by the time I've received the patient, that would already be cleared. Often, we do that in our prescreening so that when the patient has an implant, a lot of times, we're chasing them ahead of time so we're not doing it the last minute on the time when we would be putting him into the scanner. But when that happens, sometimes it comes down to risk versus the benefit with the radiologist involved or sometimes we have to cancel or delay the scan.

MR. STANLEY: Charles Stanley. Well, I think some type of standardized template that all the device manufacturers had to use would be quite helpful where it would be much easier to go to that particular line that had that yes or no question or answer or that information that you need. I mean, as someone who's dug around, started in Google, found the company's website, found the legal affairs part on the website, then tried to dig through the legal language to the one scientific paragraph buried in the middle that may have talked about whether the device was safe or not, it makes it very difficult.

The way that Frank has his website set up, it's easy. You can go. There's a template. You know where to go. You can go look at one, two, three, four, or five, and I just think that device manufacturers be required to provide it kind of in a template format, and whatever that central repository is, that there be a central repository of that information as well, would be quite helpful. Because, again, at the end of the day, it comes down to what Karen just said. It's the question, and if there's a question, it's going to go to the radiologist. The first question is, is there a question. If we can find the information as a technologist and it's deemed to be safe, then it's a sign off by the radiologist at most. But, again, that difficult information is incredibly difficult to find sometimes.

DR. SHEIN: Frank.

DR. SHELLOCK: Let me push the button first.

DR. SHEIN: Thank you.

DR. SHELLOCK: With regard to the patient ID cards, great idea. Always see recent comments coming from the FDA to have a patient ID card included as part of the information for implants and devices. The enormous issue is that the patient ID card has space limitations. So what happens is either important points are missed or left out on the patient ID card. There's no standardization of the information that's on the ID card and/or a lot of times the patient just doesn't remember to bring it with him or her to the MRI procedure.

So patient ID cards can only go so far. We need standardization with regard to the information. It needs to be concise and straightforward, and it needs to be understandable by the MRI healthcare workers.

Since a lot of the implants and devices have extensive information associated with them, as you saw for some of those stents, there are huge deviations from what the original intent was for labeling with regard to a very straightforward template, and now we see reams of paper applied as far as labeling information for something that's relatively simplistic, like a passive implant, like a stent or a prosthetic heart valve. We need to get back to the basics, get the straightforward information there and get the information to the people that need it so they can make decisions and do the MRI procedures in a timely manner and get standardized information on those patient ID cards.

One of the options that I've seen some of the manufacturers use, since they can't put all the information on a patient ID card, for example, for the MR conditional cardiac pacemakers, that there's that website and/or customer care information there so they can at least call the customer care individual and/or go to the website and get the information they need. That's, I think, a viable option.

As long as the customer care individuals have been trained to understand what the questions are and give the proper response, because I have seen instances where customer care information are then now making the matter even worse by providing more confusing information or not even understanding what the labeling questions and issues are. Then it just makes for problems all the way around.

DR. SHEIN: Thank you. If I could hear from the implant manufacturer, maybe some thoughts on what they would like to see in terms of labeling or some of the constraints that you see in the label.

DR. VENOOK: So this is Ross from Boston Scientific. One of the concerns that we think about a lot is that, particularly let's say with regard to patient ID cards, is that patient ID cards capture a moment in time. When we think about the MR safety for our patients, they get implanted with the device, and then they walk out the door and different things happen to them. For our devices, they go and they get spinal fusion surgery after they have our devices, or a lead breaks and they have a revision, and they move to Minneapolis and then another doctor does this other thing.

So one of our concerns is how do you address that? Not to say that patient ID cards aren't a really important aspect and shouldn't have the sorts of features that people have been talking about. I think that they should, but there are limitations there, and we're trying to understand how to make it work appropriately so that the MR tech is able to look at the patient and understand, getting to Charles' question, is this a patient we need to be asking questions about or is this a patient we can do scans? And if it's a patient who has a more complex device, probably the answer is you need to maybe ask some more questions, you need to dig a little deeper than the card that they got when they were implanted eight years ago. That, I think, is one of the concerns and the things that we're thinking about in terms of the effectiveness at Boston Scientific.

DR. SHEIN: I think a lot of discussion that I'm hearing thus far on this topic is addressing the implant where a patient aspects. Is it safe to assume that the tech knows all that he needs to know about he MR equipment that's set up in the room, the scanner he's about to use, or is there labeling, perhaps, that needs to be addressed there as well? If I could maybe get some of the MR manufacturers to address that.

DR. LAZEBNIK: Yeah. Roee Lazebnik, GE Healthcare. If I could reframe the issue a little bit, at least from my perspective, it's not so much about full disclosure versus minimal requirements. I believe the role of the manufacturer is to provide a responsible amount of labeling that allows a clinician to make an informed decision about the scan. Ultimately, the decision to scan is, of course, not determined by the label. It's dependent on the clinician's decision in context of the labeling.

I think there is certainly a tradeoff. As we've heard in the discussion this morning, we've been providing too much and confusing information and providing too little information. That's why I think the comments that were made by my colleague, Joe Schaefer, are very important in that standardization of the terminology, of the information that is needed is important across the entire industry so that a technologist or a radiologist will not be confused by the information provided through the labeling, that is, that the information on the label is well understood in the process of making that informed decision.

DR. SHEIN: Thank you.

Questions from the audience this morning?

DR. VENOOK: While they're working on the mics, this is Ross from Boston Scientific. I might just add to Roee's comment, that one of the hopes for the fixed parameter option that we've been developing is that there will be this standardization, and we'll allow a device manufacturer to be in a position such if a label doesn't require the MR tech to know the insides and the outsides of the MR scanner, that sort of thing is dealt with by the people who understand the technology and that there's a simplification across all that.

DR. SHEIN: In the middle first.

MR. STECKNER: Michael Steckner, Toshiba. Dr. Shein, I just want to actually make a statement, less of a question. There was comments during your session this morning with regards to the spatial field gradients and confusion whether the number provided by vendors are covers on or covers off. I just canvassed the five major vendors that are in the room today, four of them report covers on, one reports covers off. In the case where the covers are off of the numbers, the consequence for clinical use is not relevant. The difference is sufficiently small. So I hope that does clarify it, and I believe there were some people on the Panel who might want to add some further comments to that.

DR. SHEIN: Thoughts from the Panel?

MR. FRESE: Michael, thank you for putting this on the table. I would like to go a little back in the history of the MR safety standard and spatial gradient. The safety requirement of the MR safety standard is fulfilled by every MR system in the world; otherwise, it would not be possible to market such a system in Europe. NCIC standard is also recognized by the FDA.

Requirements for maximum spatial gradient came in the second edition. That was about 2001. We also have provided information about some maximum B 1, the maximum dB/dt and also the maximum static field. It's not necessarily 1.5T. You might find a place where the static field is even higher, up to 2T. This information was provided in order to enable the user to get a feeling, where do I find the maximum attraction force, where can I test an auxiliary device. It was not meant to make an assessment on implants.

In the meantime, we've learned these data are used to check implants or requirements if an implant can be put into that specific MR system. Therefore, with the third edition, we have revised this complete chapter and provided much more information, much more detailed information on different location and so on, in close connection with the implant manufacturers and the Joint Working Group to improve the situation.

This third edition is now going to be implemented, but it might take half a year, a year, or even a little bit longer. There, the maximum spatial gradient is plotted for the patient accessible area. That makes it clear, and hopefully it gets a little bit of confusion out of the debate. Furthermore, we are already working on the first amendment of the third edition, and that's why MT 40 is here in Washington this week. One of the major topic is -- you've heard already about that specific experimental option where we go and achieve maximum output values, which helps the implant manufacturer to test and to validate the implants. We have heard about that in the previous presentations.

DR. SHEIN: Further thoughts? Frank?

DR. SHELLOCK: That's great that there are manufacturers, you know, we have cover on and cover off information. Nobody knows that, and that information isn't presented to the MR users when those maps are provided to them. I think the basic question is do we truly need highest spatial gradient or passive or can we go back to the basics of utilizing 3T or less, 1.5T or less. Skip that spatial gradient information.

I'm trying to think, and I know we went over this when we were going from the short bore to the long bore systems and knew that there were substantial differences in the high spatial gradient. And I think even I indicated that there may be a situation where something on a short bore versus a long bore system may pose a problem with regard to the magnetic field interactions translational attraction.

Thinking back on all the different implants now that have been tested over the 25 plus years by myself and others and recorded in literature, I can't think of an example where something is going to be hazardous as a result of different spatial gradient on a short bore versus a long bore, 1.5 or 3T system. I wonder, since Manny tested me and I know he's tested a lot of implants, what his thoughts are on that.

DR. KANAL: Manny Kanal. The question is, to what we're testing for, I think it's important that we take one step back and make sure everyone understands, I am not talking down to you. I am simply explaining the system.

The only reason we want to know is because we have an end goal that has been identified for us, for better or for worse, which is G, which is, as Dr. Schaefer was mentioning, gravity. We want to know, is it 45 degrees or not? Does it mean we're going to cancel the study if it's 47 degrees, but we want to know, does this deflect 45 degrees? If it deflects less than 45 degrees, for whatever reason, the MR community has accepted that that's not going to be a concern for us, a significant concern whether we scan the patient or not. The only thing I need to know is can it deflect 45 degrees if I bring this patient or healthcare worker into the magnet? I don't want to know about if the magnet's shroud is off. I just want to know can it? In order to answer that question, I simply need to know one thing; where in that room will I need to test?

Now, the question that Frank asked is a very important one. If it's at 1.5T long bore, how much more does the actual force go up if you make it into a short bore? Well, the question is more complicated than just long and short bore. Is there shielding? Is there not shielding? Is it active? Is it passive? How tight is the shielding? Did they add shielding to the site after they purchased the scanner, which would be unusual, but I guess, possible?

So all I need to know is where in the room is the maximum possible force, and will this device in this patient, at that location, deflect 45 degrees or more? We're not talking gradients, temporal gradients now. It's the simplest question to solve. How can we be sitting here worrying about whether it's dB/dx or whether it's field strength? Maximum force product, where is it? Is it more than 45, and that part of our question is --

DR. SHEIN: Dr. Woods.

DR. WOODS: I guess, to partially answer Frank's question, for many devices, for many implants, it's true. You could say that they're safe as far as force and torque go for 3T or less or 1.5T or less systems, but what we were faced with is trying to come up with a set of parameters that we could use for all devices.

Again, with the parameters we came up with, we wanted to try to come up with a consistent list of things that people at sites could see all the time, and that's why we included the spatial gradient for everything because, again, while there are many devices which you could probably label as safe, as far as force and torque goes for less than 3 or less than 1.5T, there are also quite a few that are not, for which you do need to include the spatial gradient information.

Also, we wanted the information for the implant to reflect that the testing and evaluation the implant had undergone, not knowing what kind of new scanners the vendors were going to come up with that may have substantially different spatial gradients or field strengths than the ones that exists when the testing was done. And so that's kind of why we came up with a list of things that we do ask for in the labeling.

MR. STANLEY: I'm going to provide a nontechnical answer to this. Did that hurt? We've been asking this question on the LISTSERV. We've been asking individually and collectively to vendors for years and so, to kind of go back to something we discussed yesterday, does the FDA need to be involved in helping us solve some of these issues? Well, yes. We finally got an answer in this room today, that four or five vendors provided information that countless people in this room have spent countless hours trying to track down from the vendors, and we got it today.

So, from a nontechnical standpoint, again, that's why we need someone else to be involved to help us. It isn't even a question of whether we wanted to ask the question or needed to ask the question. The MR community did ask the question, and we couldn't get an answer from all the vendors at any one time until today.

DR. SHEIN: On the left of the room.

MS. KIMBRELL: Hi. Vera Kimbrell, SMRT, MRI technologist. I just want to say and perhaps ask for comment. As an MRI technologist, this is what I need.

I need to be able to match the implant parameters with the vendor specifications. I don't really care what those specifications are. Maybe it's not even important that I completely understand them, but they need to be in the same language. If the vendors have specifications that say gauss per tesla per second, then the implant needs to say the same thing. Don't make me do math. I didn't do well in school in math. I need to quickly assess a patient. I need to understand if, in my scanner, and I need to know what those parameters are in my scanner, and I need to match the implant to that.

As to where I get that information, it's probably naïve for us to ask patients to carry implant cards. It's sad that it's naïve. I would relate this a little bit to drugs. Most patients know what they're allergic to. They don't, however, know what's implanted in their body, and that's a little sad. Perhaps education, both from implant manufacturers and from referring physicians and education directed to patients is needed in this field. It really is. But from you, from the manufacturers of equipment, I need to easily Google, right?

The international community can Google. I can type of Medtronics, whatever device, and then I need to easily be able to get to the implant information if I can't get it from MRISafety.com or whatever. But that information needs to be available, and I need to understand it. It needs to say words that I get. I need to look at that information, and then I need to look at my vendor information, and I need to be able to quickly figure this out. I don't need to go through four layers. I don't need to delay care for that patient that needs this scan not an instant. I need to understand what that is. That's all.

DR. SHEIN: Thank you.

MS. KIMBRELL: Thank you.

DR. SHEIN: Comments

DR. SHELLOCK: In getting back to what I was saying before about the inconsistencies in the labeling information, and a lot of that has been the result of, I think, FDA reviewers. And as I showed you examples, that there were just phenomenally high values for the spatial gradient information, and I assume that information was extrapolated and not a typo. It could have been a typo, or there was other information relative to the magnetic field that, where did that ever come from; huge deviations from what was originally put forth as a simple template and what was supposed to be used as standardized information.

We need more consistency with regard to what happens once the information from the implant and device manufacturers gets presented to the FDA and undergoes review and then, subsequently, gets put into the instructions for use.

DR. WOODS: Terry Woods. Actually, that 2500 G/cm number was supported by testing and analysis done by the sponsor, and it's a very reasonable value for that device.

DR. SHELLOCK: But was that a calculated, an extrapolated value, or was it something based on incorporating what the deflection angle was or what have you? And if that's not put in parentheses, if I was to see that value on an implanted device, and this actually happened with a technologist who, that's how I got hold of that instructions for use, they didn't understand why it was so high. They assumed it was a typographical error.

So if something is going to be extrapolated, let us know that that's extrapolated and/or let us know that that's an option to do that because it winds up causing a lot of problems and confusion when you, for a couple thousand implants, you see 720 G/cm for a 3T scanner and then, all of the sudden, now you see 2500.

DR. SHEIN: Again, I think I'm hearing some consistency or theme coming out that the desire for a consistent, clear label with common information so that it's easily translatable for the technologist in the room with the patient. I worry if we over-spec or over-provide information, particularly on a real estate, expensive place such as an ID card, too much information that might not provide the information we need.

I'm going to go back to the left mic again, please?

DR. FRICKE: Stanley Fricke from Children's National Medical Center. I want to bring up an issue and then ask a question.

Before coming to this meeting, I briefly flipped through our two-volume, two-ream set of internal documentation on implanted devices, go or no go, 1.5T or not, 3T or not, and I came across, rapidly, one curious example. I don't know if you can see it, but I'll just hold it up.

This is a picture of a pump implanted in a patient, and it shows, basically, a coronal x-ray of the patient with a pump. It's two major axes in a, more or less, coronal orientation. The statement is, this is safe in this orientation and then written on hand on the side, if you see it flipped 90 degrees so it looks like this and there's a little rectangle drawn, it's not safe. The motor can permanently stall. When we're at this point, we've gotten past the ridiculous. I mean the patient could move from laying on their back to laying on their side. There's a lot of things going on here. Admittedly, this is a little bit of an extreme case, but it's not something that we don't see on a day-to-day basis.

A lot of the discussion I've heard today and yesterday is about the labeling and, more importantly, interpreting the labeling. I don't know if anybody has a virus scan software on their computer. I do. It's updated on a continuous basis. To me, to have a software that takes the parameters of your scanner and the parameters of the manufacturer's label and gives you a go, no go call with intelligent directions seems to be merited here. It could even be on the manufacturer's scanner or on your scheduling system.

If it were on the scanner, and we've seen this already, you can perform a very low, I'll call it power scan for the moment, that can detect metal, detect the position of metal. That information, with this, can render all kinds of -- this is not to take away from the technologist. It's to take the magic out of the box. These are things that, the scanner could go get the information, the most current information on that device and bring it in to play as well. So comments on that.

MR. FRESE: Well, I think a cooperation between MT 40 and the Joint Working Group in creating the fixed parameter option and also talking about labeling is a step in this direction to harmonize this somehow. Unfortunately, it's a fixed parameter option, addresses only B 1 of RF and gradient issues, not necessarily spatial gradient issues on something like that, like force and torque. I think we are moving in this direction, and I think that was also subject of the presentations before.

DR. VENOOK: Again, from an MR vendor perspective, I think it would be difficult, if not impossible, to give you a so-called, as you said, go, no go answer on a particular implant for a variety of reasons. One, the interaction between the MR vendor and the implant vendors is not necessarily continuous or thorough enough to understand everything that's available out there. I think the best we can do, again, is to provide you with responsible labeling so that you can make the decision in context of the MR device's labeling and the devices labeling.

I think it's also important to consider things like whether there's any suspicion that the device is operating within specifications or not, which is, again, impossible to determine on any given day. So I think the labeling and the standardization of the labeling is the key here. We have to arrive at a mode where the information that's provided by the MR vendor is useful and interpretable by the vast, vast majority of the MR community in order to make that informed decision.

DR. SHEIN: Dr. Kanal?

DR. KANAL: Actually, I'm only going to comment on the first thing you brought up. I am not familiar with the specific device that you held up, but I was actually a bit amused that your conclusion was that we've gone to the ridiculous.

As an MR physicist, you'll understand that I tested an oxygen delivery system, a tank stem regulator that was marked as MR safe and meant to say conditional, and actually, under one orientation, it did, in fact, deliver -- if it says 6 liters per minute, it was delivering 6 liters. If I turned it perpendicularly, it was markedly impaired because of the ferromagnetic spring that was positionally dependent in its function and, therefore, it was markedly impairing the amount it was delivering, was less than 1 liter per minute.

So, frankly, we're not just worrying about, is something going to fly in and hit our patient, is it going to torque or is it going to translate and cause damage to the patient. One of the aspects of implanted devices or, as we'll talk about in the next session, accessory devices, is the safe function of the device to continue to function as it was intended, and position actually is not ridiculous. It is a consideration that we can't ignore.

DR. FRICKE: Dr. Kanal, I agree with you 100 percent. My comment stems from the fact that, if I get an x-ray like this, I think that there is a large number of people that will say well, if it's in this orientation, I can scan the patient, and they might miss the fact that if you place the patient laterally, this is probably no longer a valid assumption. So the level of knowledge one has to have for this transcends which direction.

DR. KANAL: I think that --

DR. FRICKE: That's all I'm saying. Do you have to have a physicist there for every scan in this case? Maybe that's a good idea.

DR. KANAL: I'm not sure I would say physicist, but I do believe that this supports the ACR's recommendation that an MR safety officer needs to be named before you open your door.

DR. SHEIN: Dr. Shellock.

DR. SHELLOCK: I am familiar, that you're referring to. It's a fusion pump, and I would disagree with you that it's ridiculous to have that information, and what it does afford you is the ability to scan a patient with an electronically activated device by following highly specific conditions. And the situation is, if it gets oriented to the Z-axis of the magnetic field, you could potentially have the issues with regard to the motor stalling.

So you have the option to scan that patient or not, as long as you follow the highly specific conditions, and this is what we see for electronically activated devices. It's not peculiar. We know that that's what we have to do for the MR conditional pacemaker, DBS systems, and so on and so forth. I think, going to the extent of having a fixed parameter mode, hopefully that will be embraced by everybody, but it's not going to solve the problem for all the thousands of implants that exist, and also it really isn't necessary for a passive implant, which, the vast majority of the implants that exist are passive implants that we have to deal with on a regular basis.

What we would like is to see three or four lines to allow us to quickly ascertain as to whether or not it's going to be acceptable to scan the patient or not. We don't need a fixed parameter mode to do that. I think it's necessary for electronically activated devices, but it likely is going to be impractical to roll that out to the MRI community because now you're telling them, the device manufacturers and everybody in the MR community, to upgrade their software and understand fixed parameter mode. It's going to be a struggle to get there, but hopefully we'll get there at some point.

DR. FRICKE: I agree. I wasn't saying that the amount of information is ridiculous. I'm saying the imposing interpretation of that, and that's why I'm suggesting, at some point, a connection between your scanner and what the device manufacturer's information is so that the interpretation, whether it's a lack of mathematical abilities or physics knowledge or something so that it can help interpret it. It's not just a static piece of information or a customer service line that may or may not have the appropriate level of information or disseminate too much information that can't be interpreted. The idea is to scan the patient and safely. Interpretation based on the differences in data and the differences in scanners presents a real problem and challenge every time that, in part, could be solved through software today.

DR. SHEIN: Thank you very much. I think that, from my own experience in the pacing area, I go to the Heart Rhythm Society meetings on an annual basis, and it's not infrequent that there's a presentation on, I've done 100 patients, I've done 50 patients, I've done 200 patient scans, and I've never had a problem. What I've learned in my interaction through the Joint Working Group, that this situation is complex, and the more we learn about it, the more we learn we really don't know.

I would caution against ad hoc testing that, I did one last week, therefore, if I do another one, then I'm going to end up with the same result. The likelihood of problems arising may be small, but they're real. The implications for patients in that small population to whom those problems might occur are very real, and we need to make sure to avoid that. What I would encourage is for more cooperation, and I'm seeing a lot of it in the Joint Working Group between the MR community and the implant manufacturing community because I think, there, when they share the technical information or descriptions of the devices and the modeling that was alluded to by Niels earlier, that's where the true strength in solving some of these issues are. To go ahead and do limited scanning testing and publish on that is not a way to proliferate the safety of MR scanning of a class of devices.

I'll take a question in the middle of the room.

DR. CRUES: Thank you. John Crues from Radnet. I would just like to emphasize that, outside of an academic center where most of the people in this room come from, this is a very difficult area. The default mode in most of the centers in Southern California is simply now not to scan anyone with any implanted device because nobody wants to take the risk. And I don't think there are very many centers outside of academic centers where either the technologist or the radiologist is able to interpret even the current labeling that we have.

As things get more complicated and with at least four aspects of the physics that are important for safety, I think it's going to be imperative as we go forward, as was suggested a minute ago, I think you really need to -- the tech needs to go into the scanner, put in the information concerning the device that that person may have in the body before they get into Zone 4. If there's a warning, it should come up on the scanning device, coming from the internet, from some website that has all that information on it. Once they get into Level 4, the MR machine itself should know what limitations there are for that individual device and require manual override to do any pulse sequences, which would be considered unsafe with that particular device.

Without having that kind of interface, you're not going to have the right people at the center to make the kind of decisions you're talking about. It's becoming too complicated. So I think experts in the industry need to have this available to the scanners so that we can have the best possible advice to the technologist at the time of scanning for each individual advice, if that's possible.

DR. SHEIN: Okay. Before I --

DR. CRUES: I was wondering if that's possible.

DR. SHEIN:  -- open it to discussion, I would like to complicate that a little bit further. There are currently, in MR systems, outright contraindications for scanning in these devices, certainly, with the approval systems such as a Medtronic recent pacemaker that has been shown to be MR conditional. Frank, an outright indication would seem, at least to me, in all cases, to be obsolete. How do we transition from those devices from when they were labeled and approved to move forward to get them to be more contemporary and up to date with the devices that are available to today? I'll open that for discussion.

DR. LAZEBNIK: Roee Lazebnik. One of the issues is exactly what you mentioned, is that, obviously, from a vendor perspective, it's extremely important for us to be 100 percent compliant with our labeling. And we do recognize that as technology involved, particularly, for instance on the implant side, it's important for us to continually re-evaluate their labeling and act accordingly.

The pace for that is, unfortunately, not necessarily real time. That is, it is very difficult to change labeling, for instance, on a weekly basis in response to new devices appearing on the market, new technologies. Again, I don't think it's realistic to expect the MR device to be completely aware of every implant out there, of every specification for every implant and to continually compare that with the specific MR system, again, and still be compliant with the currently published labeling.

Having said that, certainly, it's important to make every effort to stay up to date in terms of labeling and to be very cognizant of the evolution of the technology, and I believe the proper forum for that are groups like MT 40 and adherence to other technical standards from the IEC, from ASTM that continue to examine the issues involved. I believe that's the best industry can do, given the ground rules under which we operate.

DR. SHEIN: So from what I'm hearing you say, that you would need help from the agency, perhaps, in ways to facilitate more frequent, perhaps, updates to label? I mean I'm stepping outside of the box a little bit here because we don't do this in pacemakers either, but would it help if there were periodic required updates in labeling for you to do periodic searches of the literature to see what's out there, to come up with updates. Perhaps, one of my colleagues at the end of the table?

DR. FARIS: Well, I think that, while I think it would be very impractical to expect the MR manufacturers to specifically update for new devices that become MR conditional, it might be practical for, instead of, for example, the blanket contraindications that pacemakers might use, maybe some caveat to that, that because it says unless a device is MR conditional and responds to the condition of use as set and defined by the parameters of that scan.

DR. SHEIN: Dr. Kanal.

DR. KANAL: I just want to go back to the comment that was made by Dr. Crues. I think that, after it's all said and done, one of the most important things for the FDA in this session is to hear what he said because I think it's not a matter of whether Manny agrees. Nobody cares if Manny agrees. He's selling you a de facto observation. It's a true statement when I teach about MRI safety, one of the quotes I like to say is that it's very easy to say no. It takes knowledge to be able to say yes, and if the knowledge isn't there, the default is no.

So I'm not sure this really falls under the purview of the FDA because the FDA is not supposed to be there ensuring access to MRI. You're there for safety and efficacy, but the fact remains that it's become so complicated or we've learned enough to recognize how complicated it is, that it requires significant knowledge in some cases to approve the scanning of the device, of a patient with the device. So it might be that you might want to entertain two types of sites, sites that can scan patients with the devices or certain types of devices, active or otherwise, and sites that are approved to scan patients without implants if, in fact, it requires that much expertise.

DR. LAZEBNIK: Roee Lazebnik. Just to address your question directly of whether it would be useful to have, for instance, mandated, continuous relabeling, I believe, actually, the answer is to continue to work towards standards that the entire industry implements because it's not really a question of just continuous evaluation of the labeling, that the labeling can become confusing if it doesn't follow well agreed upon standards. Again, terminology being similar across all vendors, both on the MR side and on the implant side and making sure that those standards are both comprehensive, but also not overwhelming and palpable. So I think that's where we approach to let groups like MT 40 work on what is appropriate and what will actually benefit the patient from a safety point of view in context of all the data that is available.

DR. RUGGIERI: It's also important to consider the other side of the picture and remember that the implants were tested at one point in time, and the machines have developed, gradients are different. If you're trying to compare testing that was done years ago on a specific implant, we're looking at a system that doesn't necessarily apply. From a tech standpoint, he or she sees that the spatial gradients don't come close to what exists on the machine. They just look for the fixed value and consequently turn the patient away or the person who is taking care of the scheduling.

With regard to John's comment a little bit earlier, it was particularly the outpatient world or practice world, that if those decisions actually take place before the scheduling rather than when the patient shows up at the scanner, and it may be that the safety data precludes the patient from being scanned on that particular scanner. Does that mean he or she drives across town to get his scan at a different facility affiliated with that site? Does it mean that they go to the local university setting who may have a larger array of machines that would permit him or her to be scanned?

So those decisions should probably, from a scheduling standpoint, from patient satisfaction standpoint, from patient safety standpoint, those decisions should probably take place up front rather than after the time that the patient's arrived, and particular made aware of before the patient shows up in Zone 4.

DR. VENOOK: Ross Venook from Boston Scientific. I think from a device manufacturer perspective, I appreciate the question and I agree with everything that was said here by my colleagues. I think that the paradigm is well established in medicine that if you're a site that can't handle the patient that you're currently treating, you find a site that can treat that patient. I think that there is a goal for our company and I think for our colleagues to have devices where a patient can just show up and go in and nobody has to know anything about it. That's a long-term vision.

Where we are today is that there's a lot of complexity, different devices, different patients, all sorts of things come into the decision making. So certainly, for the near term, I think it's appropriate to consider, and I would strongly support any MRI site that thinks that they're not capable of making a decision about whether a patient can be safely scanned or not, to not scan the patient. I mean I think that's a prudent decision, and I think that if you want to develop that expertise and, on the device side, we're going to make it as simple as possible, but when it comes to patient safety, you need to know your limits. And the fact is that this is a very complex issue, and we're trying to simplify it as much as possible, but we're all in a position where we're going to need to be making tough decisions, and we need to know our limits when we're trying to make those decisions.

DR. SHEIN: Yeah. We'll take a question at the left, please.

MR. KELLY: John Kelly. I am an MRI technologist for many years. I've worked at both outpatient sites and also in the hospitals. My concern, and hoping that from this workshop will be developed, a concise information center. Over the many years that I've been scanning, things have changed, the ethnicity that I've scanned, communication, language barrier.

So when implants are placed in a patient and they may not speak the current language, that's a problem, so the idea of having some sort of central depository of implants and where a technologist can go and find this. Now, being that there's three shifts, days, evenings, and nights, and then we have weekends, hopefully, from this workshop, we can come up with some sort of solution that, for the technologists, it would be easy to find these things. Thank you.

DR. SHEIN: In the middle.

MS. SAWYER: Anne Sawyer from Stanford University, MRI technologist. Just a couple of questions. Hopefully they'll be quick answers.

At this meeting, this is regards to the spatial magnetic gradient, and I'm always all in favor of things becoming more specific if they help you, if they don't confuse the current situation any further, but I'm also wondering, when these ideas come up and are evaluated and determined -- and this is with regards to, I hear a lot about working groups and alliances and this and that. I guess my first question always is, is there an MR technologist that sits on any of these groups? Is there an MR technologist that the FDA runs any of this by?

DR. SHEIN: I can speak to that question. At the Panel meeting for the Medtronic device, we were specific, and we actually included a couple people from the MR technology and people who ran MR systems and institutions, but that was absolutely important to us and it's a critical question.

MS. SAWYER: So my only suggestion, and I'll go out on a limb and Charles can throw something at me, but I would offer the section for MR technologists at the ISMRM team to be your tool. If you need things reviewed, send them our way. We'll be glad to help you.

DR. WOODS: That would be great. I know we have, from ASTM, at various times, submitted things to the website, and we really appreciate and would value the input of the technologists and what we were doing within the ASTM group and, I suspect, in the other groups as well.

Frank will tell you, I'm a tough reviewer, so I think we work hard on these things. Did you have --

DR. SHELLOCK: Frank Shellock. I think that's a great idea, and I fully understand why the technologists were involved with the MR condition pacemaker, but it's really not just for that. It's for every single implant.

The suggestion that the FDA developed an in-house expert group that looks at all the labeling aspects for implants and devices and run it past a technologist and get input and say do you understand this information. Then we're not going to see these inconsistencies and these very lengthy and virtually un-understandable instructions for use for passive implants, the vast majority of the implants and devices that exist.

So we don't need to just do that for electronically activated devices. Absolutely, we need technologists' input, but the MRI technologist needs to be involved in all labeling aspects that get run past the FDA because I believe that you have MR working group members that may not fully appreciate the level of understanding that the technologist may be able to provide as far as input, and it should go a long way in solving some of the issues that currently exist with the present labels that have FDA approval for MRI information.

MR. STANLEY: I'm Charles Stanley, and after hearing something that Frank is saying, I would even go one step further to that and say not just a technologist as chosen out of the air, but to actually use an organization like the SMRT because, as we've heard even from some of the people in the audience, you know there's a wide variety of clinical settings. So to just pull up technologists that's worked in an academic medical center might not be completely representative of the workflow issues that you're trying to work through.

So I would say the SMRT offering our services to you, we also are the keepers of the LISTSERV which many of the people in here are on. We have access to ask these questions to the list and to take advantage of the numerous technologists that stay active and involved with the SMRT.

MS. SAWYER: My second question is, now that we have all these very smart people and the safety experts in the room, has anyone ever been able to confirm a burn due to the time varying gradient magnetic fields?

DR. KANAL: Manny Kanal. I believe the question was, and you actually -- I found that interesting that your question was that there are many smart people in the room, as well as MR safety experts. Very appropriate.

The question, I believe, was are there burns that have been reported from gradient DBT as the cause, and my answer to that, from personal experience, is none.

MS. SAWYER: Thank you.

DR. KANAL: It doesn't mean that there have been none. That means I am not aware of any.

DR. SHEIN: Dr. Kuster, on the left.

DR. KUSTER: Niels Kuster at the IT'IS Foundation of ETH Zurich. I have an issue I don't understand. I'm aware there are devices labeled as MR conditional safe, which are specified with the requirement, the 1 watt or .1 W/kg. At the same time, we know that none of the machines out there can control these parameters. So especially from somebody outside of legal system of United States, it's difficult to understand that the regulator can approve something which the radiologist cannot control. So my question to the FDA is why this is possible. Does it make sense from the implant manufacturer? What is the motivation, and from the radiologist, what they do with this information?

MR. OLSEN: This is Jim Olsen from Medtronic. I'll take a shot at that question. That testing for deep brain stimulators, a submission was done approximately 10 years ago using what, at that time, were the industry accepted standard methods of ASTM phantoms and doing testing in scanners. The original labeling was for .4 W/kg and then, three or four years later, we learned that some scanners, as we've talked about earlier today, run at significantly higher power than others for a given reported SAR.

So Medtronic reduced the power level on their labeling from .4 to .1 W/kg, and there were a number of other extremely conservative measures taken in that testing, like keeping the temperatures below 2 degrees C and running the leads at the very edge of the phantom.

So I think Medtronic's position on that labeling is that we have substantial safety margin as demonstrated by the numerous literature reports of scanning at much higher power levels than at our labeling. So that labeling, at the time, was using the methods that were the best that were available, and we believe that there's substantial safety margin in that labeling, and it did go from .4 to .1 W/kg to try to address some of the variation in scanner power levels.

DR. KUSTER: I believe that my point was that machine cannot control .1. Even the calibration is occurring in much higher levels.

DR. RUGGIERI: I would argue that it is possible to look at -- you can achieve SAR levels on that level using an SVGR or MPRA (ph.), whatever you would like to call it, because you're using a nonselective RF pulse. You don't have 180-degree inversion pulses. Those data would acquired with the transmit receive head coil, which uses considerably more efficient RF pulses than would be the case with the body coil transmit. So that combination of factors allows you to get down to .1.

Beyond that, you can, in fact, control that if, for example, you want a -- acquisition, you can reduce echo train length, for example, which would reduce the SAR considerably in contrast to what would come out, what you would get otherwise. So you do, in fact, have control over that, extends the exams time certainly, considerably. If you have the option of using transmit receive head coil, you can certainly reduce that considerably, I think, if you're doing a head study, but you don't have too many options, for example, when you're doing body imaging with one potential exception of the way the extremity coils can, you know, transmit receive head coil.

DR. SHEIN: Terry.

DR. WOODS: Terry Woods. And also, to echo what Jim said, this testing was done quite some time ago, and we've learned a lot about RF testing since then, but at the time FDA's position was the company had demonstrated that the device should be safe with that power level, and so we felt it was more useful to label the conditions for which they had demonstrated it to be safe and then let that decision be up to the sites, whether they thought they could achieve that SAR level with the scanners that they had.

DR. FARIS: Just to echo Terry's point, I think that we've heard a lot of discussion today about how the labeling is relatively complicated, not necessarily consistent across devices. I think we need to take that information and take those thoughts and take that back and discuss that. I think it's important to just make the point that we're trying to walk a fine line between putting out a very simple message that could be understood by all, but also recognizing that there are going to be times where a more sophisticated interpretation was needed and we want to make sure the information is there to allow that interpretation to make a very difficult decision. And also, to allow information that reflects understanding that the technology will evolve and that newer devices may have those abilities, that it may not be present right now and that that device labeling should be relevant then as well.

DR. SHEIN: Thank you.

In the middle.

DR. SHULLS: Brian Shulls (ph.), Cook Medical. I have question about the RF labeling, specific to passive implants. The labeling, we've approached it very conservatively, and there's testing information that's provided. We list the temperature rise in this phantom, we list the specific MR system that we're using to do the testing. We provide the telemetry or the whole body SAR telemetry value. My question is do the MR technologists know how to interpret this, and is all this information needed?

DR. SHEIN: I guess, getting back to the topic of what would go into the label, what the technologist needs to see. I know that we at FDA are trying to standardize the implants, as well as with MR systems, including information about field, dB/dt, and SAR. Why aren't those parameters sufficient for technologists at the end of the day to know what they need to do? If we could address both of those issues under the same discussion, it would be helpful. So I'll turn to the table.

DR. RUGGIERI: I'm going to bridge that. I think SAR is very straightforward. There are different scan modes that the manufacturers provide, but recognize that their safety modes that exist for manufacturers vary from one to the next. Siemens is different than Philips, and Philips is different than GE, et cetera, so you have to be aware of that. Field strength's easy. The big dilemma is the spatial gradients, and that has become particularly difficult to interpret.

DR. SHELLOCK: Frank Shellock. I think you're getting to a very important point. There for the longest time in the labeling, it's stated the MR system reported SAR level, then the temperature information. Then, at some point, there was the decision to go and include information. Well, it's either going to be the normal operating mode and/or the first level of operation. Interestingly enough, when I polled well over 100 MRI technologists, only 3 understood what the normal operating mode was. So that's a problem right there.

And so the question is, if we're going to select parameters relative to the RF aspects of an MRI procedure, what are those going to be, and if it's normal operating mode, fine, let's be consistent or first level of operation, but let's stop there. Why do we need to put paragraph after paragraph of information? Calorimetry. Who cares what the calorimetry information is. I mean, yeah, physicists do, whoever's doing the testing that's information that we have, but why put all these combinations of temperature rises relative to the body coil and so on and so forth.

I mean you have all this additional information that's just not needed on the labeling to make a decision about whether or not it's going to be acceptable to scan a patient, and you certainly are not going to be able to put that on a patient ID card. All it does is it drives the poor implanted device manufacturers crazy because now they have to then lengthen their instructions for use for the MRI section and get it translated into worldwide languages.

We need to have some consistency with regard to what we're presenting, make it short and sweet, cut to the points that are pertinent to ensure safe use of MR in the patient with the implant. This is for a comment relative to passive implant in particular.

DR. FARIS: If I could respond to that, and then, actually, I would like to pose a couple of questions to the group as well. I think the thinking has traditionally been at FDA, that we want a simple message at the beginning of that label that says these are the conditions of use, but then, for situations in which, again, maybe one of the conditions isn't met or is slightly not met, we want to provide a little bit more detail so that a more sophisticated interpretation could be taken. I'm hearing here that maybe it more of a problem and providing less utility than we had thought.

My first question would be, should the label be much simpler and just be, these are the conditions of use. This is your bullet one, this is your bullet two, this is your bullet three, and these are the only important conditions that would need to be.

My second question is, I admit this is a little bit of a surprise to me, all of the concern around spatial gradients and the interpretation. I think this is a message we need to take from this meeting. I think it's very useful for us to hear this out in the world. And so my question to you is, let's say we test a device where we think a conservative spatial gradient testing showed 1 degree, 2 degrees, 3 degrees of deflection, nearly none, such that we think that even with evolving technology, it's very unlikely there will ever be a situation where spatial gradient was an issue for that device. Should that simply not be a condition of use? Should it not be discussed at all?

DR. KANAL: Manny Kanal.

DR. SHELLOCK: Yeah. Manny first and then me.

DR. KANAL: Sorry? I think that the idea of not including the information, I am put off by that. I think having it there and saying, putting a green, it's past, I don't care how you want to put it, but to not present it, I don't like the idea.

Also, to keep the label simple is extremely important, but if there's further information that could be used to help with either an off label application or what have you, there has to be a way for us to get to that, and I'm not sure if that should be part of the simple label, but at least we keep going back to I need a website. There's no way around this. No matter what we say today, I don't want to hear about cards. You're not going to put the information on a card. It's not going to fit, and we're going to leave it home. I need a website. The only person that can update that information is the manufacturers who create that device.

So at some level, I don't want the information to be not present. I want the information that we need to be readily accessible. And the reason I keep saying it that way is because you've got to go back to the very first comment made, I'm afraid I forgot the name of the young lady, the technologist who spoke first, and Dr. Crues, because that's critical. Because the end user is not always Frank Shellock or Manny Kanal or somebody else that thinks that we're God's gift to mankind. The end user is the technologist. The end user is going to be somebody who just wants to get the patient done safely.

I will absolutely guarantee you that in excess of 99 percent of the radiologists that will be held responsible for the decision making on that patient don't understand your label today. And when that person wants to say should I scan, and the technologist, poor young guy or girl is going to turn around and talk to this doc and say come one, doc, you're the expert, yes or no, what are they going to base that on? So there has to be a second level of information exchanged that is targeted to that person. If it can't be reached, I can accept that, and then they're going to have to say no, go somewhere else where somebody who has the expertise can interpret it.

DR. RUGGIERI: Paul Ruggieri. I would clearly support that and clearly keep the labeling as simple as possible for the technologist, but there should be some data available that you can go look up in a system like ours where we've got over 60 machines that -- safety issues on. I constantly get problems presented from various sources, private practice and academics. If you're going to make a risk/benefit analysis, can a patient be scanned, if so, what are the issues, how do we modify the scan? With a simple go, no go, I can't make those decisions whatsoever. So that data should be available at some place, but yes, keep it as simple as possible for the masses, if you will.

DR. SHELLOCK: So basically, do everything?

DR. RUGGIERI: The general labeling should be simple, as Dr. Kanal mentioned, but have some site, some reference site that we can go to and look up that more detailed data and then make a risk/benefit analysis.

DR. SHELLOCK: What I would suggest is you have the basic information at the top, the bullet points, the things that 99 percent of the people are going to really be interested in, and then say additional information to follow, which provides other information for conditions and make sure that one is not inconsistent with the other.

Because that's what I've seen in a lot of instances, that you though you understood the first couple lines, and then you read everything below -- the 12 paragraphs below, and it's like, well, now I really don't know if I can scan the patient or not. Keep it consistent with regard to the message that's being presented and make sure that one section is for most people that really want to understand is it going to be safe to scan the patient or not. Then, for those of you that want to get into the physics and want some options before scanning the patient, haul that out and say other conditions are as follows or other options with regard to the MRI labeling information are as follows so people don't confuse the top part of the labeling with the bottom part and it's very straightforward with no unnecessary interpretation required.

MR. STANLEY: Charles Stanley. To follow on those two points, that is what we do today. We do use the internet. We do try to find that information. The problem is, it's incredibly inefficient. We go to Google. We go to the MRI LISTERV. We e-mail Manny directly. We e-mail Frank directly, get answers at all times of the day and night, and it's incredibly inefficient.

Then the other problem is, is because you go to so many different sources, sometimes the information's conflicting. If you go and Google a device, you may be getting a PDF that's three years old, you may be getting the most current information. To leave that in the technologist's hands to figure that out, especially when there's nonstandardization of the language and the format makes it incredibly difficult, so yeah, the point is, is that's what's happening right now. It's just we're doing it in an incredibly inefficient and dangerous way.

DR. RUGGIERI: To reinforce that, you can go to one manufacturer's website, go to three different pages, and find three different criteria for the same implant.

DR. SHELLOCK: Exactly.

DR. LAZEBNIK: Roee Lazebnik. To add to those comments, I know there's definitely, certainly a need and almost a temptation to put everything into a very simple to understand label, but I do not believe that a label, no matter how well phrased, is going to ever mitigate all the complexity associated with some of these ultimately clinically based decisions. Again, the role of the manufacturers here is to facilitate an informed decision, and that means creating a responsible label.

Again, to echo what was said yesterday, there's a tremendous amount of highly important information in the other documentation associated with the device, and that needs to be considered, along with the label. There were comments, for instance, that perhaps there's not sufficient awareness of what is the normal mode of an MR scanner mean. It's tremendously important to understand what the different modes of operation for an MR scanner are and are not and what parameters change before you scan the patient.

That's something that, certainly, from the vendor side, we advocate quite highly, and that's why we provide a tremendous amount of documentation. At the end of the day, MRI scanners, the capabilities only increase generation to generation. That means that the complexity of the device intrinsically increases. We do our best to simplify things like workflow, automate as many things as possible, but that will never mitigate the complexity intrinsic of the technology.

DR. SHEIN: I would like to magnify on that a little bit also. From my perspective as a regulator, what I want to make sure is in the content of the label, be it the implant, be it the MR system, there's enough information for the clinician, at the end of the day, to be able to make an informed decision on a risk basis on whether a patient should or should not be scanned. There are going to be a number of patients.

We don't regulate medical practice here at FDA. What we are looking to do is make sure people are very well informed so they can use these devices as safely and effectively as possible. As mentioned earlier, a slide earlier today, I forget who had it up, MR has become the gold standard for scanning some patients, for imaging some patients in whom the risks associated with scanning patients who have implants needs to be taken into account, but it still may be the right thing to do on a risk/benefit analysis.

I think we need to keep that in mind, that it's the patient at the end of the day, and as technologists, I wouldn't want see people blindly turning people away because you have an implant and you have to come back. We need to make a good, balanced decision as to whether the risks associated with scanning a patient with that implant is appropriate in light of the information you will get from that. Is there another modality that can be used for imaging in that patient? If not, it may be very appropriate to go forward just the same.

So with that, I'm going to turn to the left of the floor.

MR. GILK: Thank you very much. Tobias Gilk. I wanted to offer an observation about sort of the general content of the conversation and then pose a question to the Panel.

The observation being that it seems as though the issue associated with labeling and testing for implants and devices really comes down to effective and efficient coordination between the implanted device and the MR device and making sure that the information provided is consistent and equivalent from things such as units, down to the specific information about the relative safety of scanning patients with implants.

I would encourage any consideration from modification of current testing and labeling conditions to anticipate future technology. The FDA has already approved MR scanning of up to 8T for human use and research purposes, and we all know that the research magnets of today are the clinical magnets of tomorrow. We find ourselves today in a situation where we cannot compare the relative safety profile of an implant or a device very effectively between 1.5T or 1.2T vertical field or 3T. Any changes to this testing and labeling strategy should anticipate newer magnets, stronger magnets, higher spatial gradients, and we should do our best to come up with universally applicable standards that are coordinated between the implant device and the MR manufacturer.

My question, however, is related to the creation of information about implant and device safety. Dr. Kanal sort of touched on this just a moment ago. I'm curious to hear the Panelist's individual impressions, but also, the FDA position with respect to the availability, accessibility, and accuracy of information for technologists, for radiologists, for patients, and is there a need in some sort of official central repository. The current resources and conglomerations of manufacturer data that currently exists necessarily, legally disavow responsibility for that content. Or is now the time to look at some sort of sanctioned database, repository for this kind of information, either coordinating existing sources, collections of information, or looking at new strategies? I'm interested in both Panelists' individual thoughts on this and also the FDA's sort of position overall with regard to creation, accessibility, implant device, MR safety information.

DR. FARIS: I would say that that's one of the major take-homes for me from this meeting is just hearing that that sounds like a need out there. I don't know that I can take away an FDA position on doing that, but I certainly can say that FDA has heard that message, and I think we'll be discussing it.

MR. GILK: Let me refine the question a little bit. Today, as it stands right now, what is the FDA's expectation with respect to -- take away the idea of central repository for all the information. What is the FDA's expectation for an individual implant or device and the accessibility and contemporary nature of the information? What is the obligation for the manufacturer, if there is one? What is the FDA position with respect to information availability for implant safety for an MR scan?

DR. FARIS: So right now, from an implant perspective, that's done strictly through the labeling. FDA approved that labeling, and that labeling is typically available on that manufacturer's website. If the manufacturer has new information that they want to update on the website, that has to be reflected in the labeling, so they come to FDA to go through that process. So I completely understand that that can be a challenge from radiologist or a technologist perspective to try to navigate exactly where to find that label for that device. I think that it would be useful to have a better way of doing that.

DR. SHEIN: I think I would add to that, Owen, that the label and the content of the label are driven by the manufacturer of the device and are typically driven internally there by what marketing wants to be able to say about the device, if they want to speak to that. I think what I've heard this morning is we need to get back and revisit, as an Agency, perhaps, whether there is a minimum amount of information that we will expect and ask of manufacturers to provide regarding these devices, and I think that's a good take-home for us.

MR. GILK: I would encourage also to take a look at, the implanted device is persistent in the patient, even if the company that manufactured it isn't persistent. That doesn't -- the fact that the company went out of business or got bought out or changed their name or God knows what does not change the safety concerns for that patient, although, now we rely strictly on the labeling information provided with that device at the time of approval as the only source of FDA monitored information.

DR. SHEIN: That's certainly very important, and that's a great point. When devices become obsolete or the people who manufacture them become obsolete and they move on to other endeavors -- certainly, my career here at the Agency, I've watched significant contraction in the pacing community on the number of manufacturers that are out there. It's been sliced in about half -- What do you do with those patients?

What I would caution though, is what's available on the last day, you can't extrapolate information from 1.5 T to 3T, for example, in the case of defibrillators.

By the same token, you also can't extrapolate that in the same type of scanning system, that one device that performs well, if this is close enough -- there are enough device idiosyncrasies in pacemakers and defibrillators. I apologize for referring there, but it's what I know best, and it's where I live, but there are idiosyncrasies about the electronic designs of those devices that have a specific circumstance and implications for how they'll interact with the field. To speak to some of those things, once those manufacturers are out or to print something that might be interpretable by a technologist at a later date is going to be very, very hard to do. But I think that the overall perspective of us going back and considering some basic information to address MR systems would be good.

Frank, you've been wanting to comment.

DR. SHELLOCK: Yes. I think, at a minimum, the FDA should, not so much advise, but require the implant device manufacturers to -- we all know that when the device gets opened up out of the package, the IFU gets thrown away, so nobody ever sees that, so it's useless. Put it on their website and companies like Medtronic and Cook Medical and others have isolated websites, specific homepages that have their MRI information for all their implants and devices. That's a great start.

Rather than advising implant device manufacturers to have the patient list the information with MedicAlert, which I think we all know that's just not -- it's a ridiculous recommendation as far as I'm concerned. Advise them to have it on their website and make it readily available to the MRI healthcare professionals that need that information. The IFU is never in the hand of the MR healthcare workers. They never get to see it, but they could have a chance to see it if it's posted as a PDF file on the manufacturer's website. So I think that should be something to be carefully considered by the FDA.

One other comment with regard to the labeling information, and I agree with my colleagues here in some of the things that were stated, but we need to make a distinction between passive implants and active implants. So yes, we need to see all that information, the laundry list of information for the active implants. That's necessary to ensure safety for the patients, but it's really not necessary for the vast majority of implants and devices. Let's just see a couple of bullet points that get the point across in a safe and efficient manner with regard to the MRI labeling.

DR. SHEIN: Dr. Kanal.

DR. KANAL: Manny Kanal. The point that Frank just made is exactly what I wanted to speak to and to just add that the FDA could be of substantial benefit to us. I don't know if you can do this, but it's a wish list. It might be possible for the FDA to help us in the accuracy of the data that makes it to the labeling.

Now, I don't mean to suggest that the FDA is not ensuring the accuracy, of course, but in the opposite extreme, you're not. We have stents out there that have zero degree deflection angles and have in their written recommendations four to six-week waiting periods before you can scan them. This confuses the industry. The technologists don't understand why that's there, for good reason. There's no reason for it to be there and yet somehow, it makes it -- somebody said it. All the FDA typically does is either approve or not approve of the wording. I would recommend that the FDA consider requiring any statement made by the manufacturer to be scientifically defended and not just let it through. It may not be doable, but it's a suggestion.

DR. SHEIN: In the center of the room, please?

DR. RANSICK: This is Mark Ransick. I work for Ethicon Endo-Surgery. I would like to just kind of bring a different perspective actually from a passive device manufacturer. Actually, what I came up here to do is reinforce Vera's comments in terms of the labeling. It would be wonderful to basically have kind of a -- I won't say simple, but a standardized type of labeling so that I don't get the calls from the call center, one of the technologists that say do you report this. How do we dial it in from our end? So that would be wonderful.

Also, as another example from -- got an e-mail directed from Charlie McIntyre, the same type of issues in terms of what we label. We had quite extensive testing done to the new FDA guidance, and when we reported the data at first, I actually didn't publish it, but we had Frank review it, and thank God because what I was going to actually place in the writing was even more nonsensical than -- it was pretty darn nonsensical where I was interpreting the test data because I'm not a technologist from the end user standpoint, but I'm the one that's going to be actually writing or preparing and working with regulatory to publish this stuff.

Mitchell, you made a comment that I would just like to also respond in terms of marketing versus who actually puts the stuff in the labeling. I can speak from our standpoint, our company, that what we have put in our IFUs or packaging inserts was directly related to our regulatory requirements. So it was actually driven from regulatory based on the FDA guidance that came out, and also the ISO regulations that came out around the same time.

DR. SHEIN: Certainly, there are requirements through guidance and expectations around that, and certainly, there's the propagation through the years of the content that's always been there, but I think overall, that when a new claim is made, for example, MR conditional, that those are things that are done at the discretion of the manufacturer, and I don't think we drive all of those things, but certainly, there is some that we do drive.

Thoughts from the table?

DR. VENOOK: As a device manufacturer representative, I want to address a couple of the comments that were made, including to agree with the comment here, which is the statements that we make are very strongly driven by regulatory and our primary concerns are providing accurate information that's useful for making risk based assessments and risk/benefit decisions by appropriate personnel.

So sometimes, that results in a label that isn't easily readable to people who don't get that bill. Sometimes it results in labels that we all look at and say, well, why was that there and I think that's regrettable, but we're all working on it. The truth is that it's really difficult to reach every audience with a given label.

Something we struggle with, we're working on, and I think that we're taking away or I'm taking away from this discussion and the comments is that there are a variety of audiences, and this is something that isn't a complete surprise to us at Boston Scientific, that there are a variety of audiences for MR related labeling that are distinct from the audiences for which we typically write labels. It's important to understand that, and it's important to distinguish between the different audiences for different pieces of labeling. Sometimes it's a challenge to do, particularly because a label is a label and a website is a website.

It's not that there are distinct capabilities of these different technologies, but I think I want to acknowledge the comments and say that Boston Scientific is working towards the types of things that people are looking for, and I also want to commend Medtronic for what I think they've done a really good job with, which is their MRI safety website. Those who haven't seen it, you should probably go there and check it out. I think that getting specific comments on what they've done right and things that they could do better will help them and will help the rest of the industry to create the kinds of accessibility of the information that people are requesting.

MR. OLSEN: I kind of appreciate what Frank was saying about what should be on a website because we struggle into 14 different languages we have to put on an IFU. Now, you add this little comment in here that it's not that long, but now you've got to add all these extra pages. You can actually raise the sterilization dose required now because now you've got larger mass. Is there someplace that we can -- your comment is an excellent comment, something readily assessable with the technical data, basically a statement saying, under these conditions, you can do it, and then we can have further comments or more technical data depending upon the audience that needs it greater.

DR. SHEIN: Before we move on, I want to redirect the discussion for a few minutes because we've had a lot of discussion already this morning around what makes a patient eligible for scanning and what information needs to be made available with clinicians who are going to subject that patient to scanning. I want to get back to the patients themselves and the physicians themselves. We had a few slides earlier this morning regarding monitoring, what appropriate monitoring of patients to make sure that should the untoward happen, even with our best of intents, something go on during a scan, what needs to be in place and what's the minimal monitoring that needs to be in place to patients?

DR. FELMLEE: Joel Felmlee here, Mayo Clinic. We monitor a lot of patients with their devices. We typically use pulse ox, and we'll add ECG, and we do that because some patients, the pulse oximeter doesn't work. They're either too sick or their hands are cold or something like that, and in some of those cases, the ECG actually works well. There's a few cases where it doesn't, of course. I think you have to double up and use them both, and we also add the scanners, pulse plethysmograph just to be an additional backup for a heart rate monitor. It's not designed for that, per se, and there's a label on the scanner for it. We use it that way. So we have too PPG, pulse ox, and an ECG.

MR. STANLEY: Just please keep in mind that scope of practice for technologists often doesn't allow them to be the monitors of those monitors, but they are put in those situations quite often.

DR. KANAL: It's important to realize that this is exactly what we went through in the review of the Medtronic devices for monitoring. It's important for the FDA to recognize that it is inappropriate to attempt to pursue a single answer to that question. What is required for monitoring is exquisitely patient-dependent, not just because of the implant, but because of the patient. If I have a person who's got COPD and I need to make sure he's not retaining carbon dioxide, I might have a different requirement then if what I'm looking for is a child that I'm putting through anesthesia or sedation.

So the monitoring, although it may be in some way related to the implant, almost all the time I would assume it would be determined by the patient and the need of the study and the duration of the study and any potential intervention for anesthesia or sedation, not necessarily driven by the implant.

DR. SHEIN: Great point. Other thoughts?

DR. SHELLOCK: Well, to my knowledge, there's only one implant that really requires or necessitates that you begin a consideration for monitoring, and that's currently the Revo pacemaker, MR conditional pacemaker, so the vast majority. Everything else, again, trying to think through all the implants and devices that exist, it's really only one situation where monitoring is needed, and I agree with Manny, I hate to say it, but I do, with regard to the issues relevant to monitoring and what should be selected and implemented on the patient. As long as you've got the right people involved with the monitoring procedures that know and understand not just how to use the monitors, but also to respond in the event of an emergency, have the proper equipment to manage those patients.

DR. EDMONSON: Heidi Edmonson, Mayo Clinic. Also just to add to the question posed earlier about using a scanner that's not able to limit the SAR to .1 W/kg, we also have a physicist on site monitoring the scanning parameters to make sure that we're maintaining safe scanning conditions for the particular implant.

DR. SHEIN: Okay. In the middle. I'm sorry. We'll go to the left.

DR. STAGER: I'm Brad Stager with Biomet, and I've got several comments about things I've heard earlier this morning. I wanted to try to add a little different perspective as far as a passive implant manufacturer. For example, Charlie, this morning, you asked about why we don't test to a souped-up version of the parameters. And simple answer to that is that we are giving you parameters under which we've determined devices can safely or patients can be safely scanned. So although we may test for worst case, we provide you with information to the highest degree under which those patients can be scanned.

As far as labeling things across the board with either tested or not tested, FDA does allow us to say that they have not been tested, but BSI, for example, in the U.K. will only allow that labeling for a certain period of time. If we say it's not been tested, we have to come back and say okay, we're going to test it in the next six months. I'm not sure if they'll even allow us to put that labeling out there for any more than a year. I'm not sure what kind of time period they have.

With regards to patient ID cards, I actually feel your pain because I field a lot of these questions from MRI techs. I'm the point of contact. They call me when customer service says we have an MRI question and between knees, hips, shoulders, wrists, elbows, sports med, trauma, any of those different sites, we actually have different sites on our website for that. One thing I'm going to push for is a central point of MRI language.

If it's not on there, and assuming we get to that point, if it's not on there, you have to assume that it has not been tested because we put forth a lot of effort, a lot of money to have these devices tested, and if they have been, we're going to want them out there for you to have that information, and also so that you don't have to wait an hour or a day or whatever it may be to determine if you can safely scan someone. That would also be, anything that's out there would be the most current because we maintain, and assuredly, if we're going to this effort and expense, we're going to want the most current information out there.

Also, because I am the person that determines what goes into that IFU as far as MRI compatibility, and I know you hate that statement, but I'm going to put something in there, develop wording to the effect that it is highly recommended that at least the system identification, meaning in our brand name, is included in the patient information. I got a call last week, and they said well, we've got this trauma screw, didn't know what system it was associated with. They didn't even know that it was a trauma screw. In the end, all the tech said was I just need to know that it's not made of surgical steel. Yes, I know it's not. It's titanium. They went on and did the scan even though it hadn't been tested.

I'm also a bit baffled about the statements concerning a standardized version of the labeling because, as we have new submissions, the FDA hammers us hard on, well, you need to state certain parameters in there. We do express everything, and typically between 1500 and 2500 gauss, as far as the spatial gradient, there are many other aspects that they hammer us on, as far as expressed in whole body average SAR, which now I understand why that's so critical because typically our testing is expressed in a local SAR, which is obviously worthless to somebody that's in the scan center because that's not how your system reports it. So it's been very enlightening as far as being here and learning what everybody needs.

I guess, in my final, I have a question as far as whether or not there have been any known reported incidents of burns occurring as a result of orthopedic implants.

MR. FRESE: We have got some reports on burns due to RF loops, and I would count four or five cases when they had both sides, artificial hips. That's just an assumption. I don't know if that is related, but I could mention that the core intensities changed dramatically due to the conductive implant compared to a situation with a patient without hip implants.

DR. KANAL: Manny Kanal. I was just going to say that the case of a third degree burn in medial thigh is symmetric third degree burns, medial thighs in a patient with bilateral artificial hips, also reports of burns associated with external fixator devices.

DR. SHEIN: Question in the middle.

DR. SHRIVASTAVA: Devashish, Center for Magnetic Resonance Research, University of Minnesota. I have two comments. One is regarding the whole body SAR, and this is what Niels said earlier.

Our experience is that we do not understand what full body SAR is. We can monitor very accurately how much power is going to the RF coils, that are going into the RF coils, but we cannot quantify how much of that power is going from the RF coil to the body. That is where the fudge factor is, and that is what people estimate to compute their SARs. So there needs to be some kind of a standard or some way of computing this, validated with experiments so that everybody knows that when a scanner says 3 W/kg, it is 3 W/kg.

On the second one, somebody said about temperature time curves on the website. It is my opinion that this kind of information should be available on the internet because, in the mammalian brain, I'm talking about now, brain implants, the repeated or single insult of temperature versus time between the range of 1 to 3 degrees, the effect of those is not known.

So if that kind of information is provided, at least in future, we will have some hope that we can look at those curves and see how because we are not going to have just one scan. You can provide information saying we raised 2 degrees C in 10 minutes, but we are going to have five, six scans in a single session. We don't know how our body's going to respond to those repeated scans and those repeated results. Those things need to be understood, they need to be researched, but that kind of information would be necessary to be on the website for future use. Thank you.

DR. SHEIN: Dr. Kanal?

DR. KANAL: Relative to the first point you brought up, I think I would like to bring up a point that, I'm not sure it was supposed to be brought up at this meeting, but since you've gone there, there are devices available that are radiofrequency shields. They can deflect and absorb the RF energy that's being transmitted, and we can wrap the limb that has that device or what have you that I'm worried about heating up or I can wrap the cables in these.

The problem of using those in MR scanners today seems to be that it drives the deflected power detected by the system, it drives it crazy. It thinks that there's a problem, and it shuts the system down. So it would be truly ideal, from a patient safety point of view, if the manufacturers could work in conjunction with the people that create these materials, that they're actually the same materials used by the military to deflect and absorb radar and they coat entire battleships with it. We can use that in the MR scanner dramatically to decrease RF power deposition in specific locations in the patients.

It won't do anything for ferromagnetic considerations or for gradient safety, but for RF power, it's tremendous, and we've done a whole mess of testing on this. It would be wonderful to see the manufacturers working in conjunction with them to allow that to be used safely in the MR environment.

MS. KIMBRELL: Hi. Vera Kimbrell, SMRT. Just a question and comment to the vendors. We don't necessarily, as MR technologists, need shutoff valves for implants and devices. What we really need are an easy way to look at our parameters so that we set up a scan, we know exactly what we're delivering to a patient so that if we choose to go over the speed limit, we know that we're doing that, that we can stay within the speed limit. But while I don't think that the vendor's in a position to decide whether implants can be scanned in their scanners or not, I don't think that's your job at all.

I would like the operating software to give me a readout for each scan that I set up to tell me what I'm delivering, how much RF I'm delivering, the gradient strength I'm delivering. Give me an idea of what the dose that I'm delivering to the patient is so that I understand what that means.

DR. EDMONSON: Heidi Edmonson, Mayo Clinic. I really appreciated your comments earlier about the risk/benefit analysis revolving around the patient. I think that maybe we need to remind ourselves, as one step further, that as potentially a risk/risk analysis, that by denying that patient their MRI scan, that we're potentially putting that patient into more invasive procedures that may carry their own risks or risk of infection or continued pathology or pain or something like that and dependent on pain medication for extended periods of time and the potential risks for that. I would just like to maybe --

MR. STANLEY: There's also the litigation aspect as well. If you deny somebody something that was the standard of care because your facility doesn't scan those implants or that kind of thing and so, yeah, that's been discussed before. There are litigation issues as well.

DR. VENOOK: Ross Venook, Boston Scientific. I think this is a really good point, and it brings up something that I was thinking about other comments that were made before, which is that what we're striving to do is give access to a patient to MRI, and sometimes that means access in ways that are less than perfect.

So the challenge that I think we face as device manufacturers is this balance between providing the information that allows us to make really tough choices -- pardon me -- that allows physicians to make really tough choices versus providing information that's confusing or information that's not relevant or just saying, you know, this needs to be simple and so the answer is no. I think that it's an important comment, and I think Manny before said well, it's easy to just keep people out of scanners, and we're trying to walk that balance and we appreciate your comments.

DR. SHEIN: The right.

(Pause.)

DR. SHEIN: Thank you. In the center.

DR. LAUNDER: Yeah. Jason Launders from ECRI's chief. Just to follow up on that, really, point out that you can go to the FDA website and find any information on pharmaceuticals or the labeling is a fairly straightforward find. Would it be impossible, difficult, whatever, to do the same for implants?

DR. WOODS: Terry Woods. I was at that public meeting where people discussed the possibility of an electronic labeling depository for devices. The big problem is, for drugs, drug labeling is fairly standardized from one drug to the other. The dilemma that we have for devices is that devices just scan the spectrum from a comparatively simple, nonimplant to an MRI scanner and coming up with some kind of a standard format for the electronic label and trying to decide what's going to be in that label is a big part of the discussion that's going on.

I encourage all manufacturers out there and also people who would be interested in this information to find out about that effort and give us your input because, again, we're trying to figure out how do you do this for devices because it's much more difficult for devices than it is for drugs.

DR. SHEIN: So with that, let me provide a couple observations based on the hour and two-quarters that we've been having this discussion.

I've heard definitely a lot of discussion around content of minimal labeling, a call for a consistent, clear presentation of information. I heard a request for FDA to require implant manufacturers to keep websites of information regarding the products up to date on the web, most of all, for the ability of techs to match specifications for implants to those to the scanner so that they know and they can more easily determine whether patients were appropriate for scanning.

I've heard repeated calls for a leverage in the use of IT to update information for all parties regarding installed systems and devices that are approved. I've heard a discussion, maybe not fully, but it's an intriguing idea regarding a concept of a site of excellence, if you will, regarding particular sites certified or trained for the scanning of implant patients. I've heard a call for FDA to base all of the MR content and implant manuals to the data, backed and supported and validated. And last of all, from the patient specific side, I've heard that monitoring is not as simple as just doing pulse oximetry or ECG, but it's something that may be specific to that patient, and we need to be mindful of that as we move forward.

With that, I'll remind folks that we will be having lunch at the window outside as we did yesterday. I want to thank everybody's time at the rostrum. I want to thank those who asked questions and those who sat patiently throughout. With that, I'll close this session. Thank you very much.

(Whereupon, at 12:40 p.m. a lunch recess was taken.)

A F T E R N O O N S E S S I O N

(1:30 p.m.)

DR. RAJAN: I think we're ready to start the next session. There appears to be a reluctance on the part of our audience to come back. Give it another minute. Shall we give it another minute for people to come in?

(Pause.)

DR. RAJAN: Okay. I think we should get started so we can all go back home on time and catch our flights.

Welcome back to the last session of views of accessory medical devices in the MR environment and the impact on patient safety, MR safety. We'll start off. First speaker in the session is Toby sitting in the front.

MR. GILK: Good afternoon, everyone. Toby Gilk again. Same person, same credentials.

This afternoon, I would like to address the topics of accessory medical devices, and with respect to comparison of direct injury, direct reported injury product of accessory medical devices as compared with direct injury related to implanted medical devices. It appears as though the accessory medical devices are involved in a greater number of the reported accidents. Why is this? To a large extent, it's a product of the attractive force. Simon, can you pop up the video?

This video illustrates the attractive force exerted on probably a 1.5, 2 pound wrench in the attractive pull of an MR. That's 500 pounds of attractive force on a 1 pound, 2 pound crescent wrench, illustrating the extent to which we can and cannot analogize between gravitational force and the force that can be exerted on an object from gravitation versus what we see in terms of magnetic force attraction is pretty remarkable. So the forces, even on relatively small objects are pretty significant.

If we could jump back to the slide? There are five steps, five protections that we could implement that would dramatically curtail the potential for these types of accidents. Number one, to label portable accessory devices in the MR suites, and not only label them with the MR safe, MR conditional, MR unsafe, but for the MR conditional products to actually identify on the item, on the object, what the safe conditions for use are.

In a single magnet facility, this probably isn't terribly critical, but if you're in a multi-magnet hospital where on one side you have a 1.5 and on another side you've got a 3T, the critical definition of what is an effective limit for safety may be really relevant. And that pump, that patient transport, whatever it is may be safe in one environment, but not in another. And if we don't actually put that information on the device where we have enough real estate to do so creates a substantial hurdle in terms of the MR departmental staff and the technologists' ability to identify on safe conditions for use.

If we utilize the ACR 4-zone principle, if we tie access to the suite, access to the controlled access portions of the MR suite to effective screening, not only for individuals, for patients, but also for staff and any equipment that's brought into the suite, and we apply screening protocols comparable to what we do for the patient to any new equipment that's brought in; that too helps to mitigate these concerns.

For projectile risks, these portable equipment, portable objects can assist in being identified through the use of ferromagnetic protection systems as well. As has been mentioned previously, there are medical devices, ventilators, anesthesia machines, pumps that have specific limitations in terms of static magnetic field or spatial gradient magnetic field, beyond which they should not be exposed in order to operate as intended.

Some manufacturers have now begun to require that the site be marked with, hypothetically, the 450 gauss line. Since we have no perceptual cues, the hair on the back of our neck doesn't stand up when we get to 449, there is nothing in the environment that tells us where that is if we don't identify and prospectively mark that. Draeger, I think, is a manufacturer that has an MRI anesthesia machine that, as a condition of purchase, they expect the host facility to mark critical field values within their suite if that's not already done. I would like to see more of that and overt support for point of care protections, point of care identification safety threshold.

Number five, and this has been repeated in many contexts related to many MR safety aspects, provide regular, repeated, ongoing MR safety training for the MR departmental personnel to help keep them informed about the relative risks and effective preventions for reducing accidents involving accessory medical devices.

With that, I would like to thank you very much.

DR. RAJAN: Well, that was a good introduction. I think we're on to the next speaker.

MR. OCH: Thank you.

My name is Joseph Och, and I'm here representing the American Association of Physicists and Medicine, the AAPM. I am actually a member of the MRI Subcommittee on Magnetic Resonance Imaging. We're very pleased to be able to provide information to the FDA and want to express our appreciation for arranging such a public workshop. I think the ability to hear this information and disseminate it is just really valuable.

We have actually, I'm sad to say, over the two days that we've been here, I don't really have new information to add, just adding to what has already been going over. We agree that the rate of accidents and incidents has been increasing and needs to be addressed, and that the manufacturers of implanted devices have recently increased the number of MRI conditional devices, and that this has required more observation and scrutiny by the field.

Our recommendation is that we need a standardization of protocols, safety standards, and we do accept and endorse the ACR Guidance Document for safe MR practices. We're citing here the 2007 version, but I know 2011 is just around the corner, and we're eagerly anticipating that. We believe that the vast majority of these accidents and incidents could be mitigated with fundamental point of care protections and that there must be appropriate initial and continuing safety education for all of those who enter MR environments. We are in favor of careful and consistent site access restrictions.

I think all of these things have been pretty much gone over in the past couple of days. As such, I can just say that, in addition, I would like to add a couple of comments that are not necessarily those of the AAPM. As I said, I'm a medical physicist and I do have some personal instances. The FDA had asked about the whole idea of MR conditional and MR conditional labeling and does it confuse, is it sufficient. I just want to run a couple of things past people.

One is, I got a call from one of our units about a month ago asking whether they could do a certain patient on their magnet. I, of course, looked it up, and it was a spatial gradient question. I found that sure, that was no problem. A few days later, I checked in with them to see how everything went, and they said it was fine, except that we were going to do it on the 1.5 magnet that was a closed magnet that you said was okay, but just to be on the safe side, we did it on our 1T open magnet. I hear some people laughing. If you don't get the joke, then that explains why that labeling may be a little confusing, okay?

Another, just a fresh incident that happened just last week, and I'm just using this to give some examples. Last week, I had a call from one of our sites. They wanted to do a patient with an implant, but it was conditional, and I said okay, what are the conditions, and I brought out all my data sheets. They said we can't find any conditions. We have the documentation, but there's no conditions. I said well, is there like tesla or gradients or does it -- and they said no. So I ran down to the site and I looked at it and true enough, it's MR conditional. There were no conditions indicated, so they were ready to turn this patient away.

I've heard some talk in the past couple of days about turning patients away and how dangerous that could be. Being a physicist, I could, of course, come up with a high tech solution. I called the manufacturer. That was my high tech solution, okay? I found, indeed, that the condition was that all external wiring and connectors should be removed from the device, and that was the condition. No other indications, and actually, that was actually in their literature, but nobody indicated that that was the condition. So, again, there's confusion there.

Over the past couple of days, we've had a couple of discussions also about risk/benefit, and when you're doing one of these scans, the radiologist must take into account the risk/benefit. Well, a couple of ideas there. One of them is the fact that when a device is presented, that presents a question, whether you should scan or not. One of the things you commonly hear is do you have previous experience with the device, imaging or anecdotal. If you're in a large enough institution, I guarantee that your radiologist will have a colleague who, when they were at Southeastern Tech State, they imaged this device all the time and nothing ever happened. Of course, if you look into it, it's maybe not the same scanner or the same field strength or even the same device. Okay. So that's one danger there.

The other danger that I see is when you talk about the fact that, after these devices have been in a patient, they become epithelized, that is, that after a certain amount of time, it's safe to scan them, and that time, of course, is three months or 10 days or six weeks or whatever the last person that the radiologist talked to says. If we're going to have radiologists and technologists in a position that they have to make that sort of a decision on whether the device is securely in place or there is prior experience, I think there needs to be more guidance given to these practitioners so that they can make a more informed decision.

Thank you.

DR. RAJAN: Next speaker is Ashok.

MR. SARASWAT: Thank you. Again, I'm the same person. I haven't changed in two days, but actually, I have changed a lot. I have learned a lot through all the discussion and interaction. As the previous speaker mentioned, when you talk to different people, how the message gets distorted, changed, that's human nature. I heard Dr. Massengill talk about yesterday he had a heading, "Nobody's perfect."

The message, whatever is given, somehow, it gets changed. It happens all the time. The prime example is one of the famous news media got the hint that we are having this talk here, FDA, and there was a little news on I believe it's CNN or something. I don't know. They're saying that somehow, the beauticians have more training than the -- I don't think they even know what the title is. I think they used something as technician maybe, I don't know. So they're saying that the news is, the sensation is that beauticians have more training than MRI technologists, but we know better, that we are living in a challenging age.

I truly agree with you that when you talk to different sources, things get changed. That's why we need standardization, training, and simplification, but also deal with the complex issues. We're proud to be in this middle of this earlier, and yes, most of the accidents are preventative. We can prevent them.

Without taking much time, the MRI technologist role is to be gatekeeper. We are on duty, and a lot of speakers alluded to that fact that we are the ones who make sure that nothing goes past from Zone 3 to Zone 4. So we take time and, so to speak, police that, and it's important. Even though there's a lot of room for improvement, I, as an educator and administrator, when I see the MRI doors open, I don't like it, but I don't get mad. I just educate them. I repeat often, and soon they follow it.

So we can prevent a lot of things. Yes, I support the ACR's concept of four zones. Places that are grandfathered, they cannot have this one, two, three, four zone. At least they can separate that Zone 3 and 4. Everybody has Zone 1 and 2, so it's not that hard, but if they can't, there are ways to kind of modify that a little bit. Again, I am a big supporter.

Also, with my experience, I started doing MRI in 1985, then I had different roles, but I still scan. I can say from my own experience that yeah, we do not report all the accidents that are due to translational force or attractions, and you all have seen what the oxygen tanks and -- well, that's implant, but oxygen tank has done.

There's a long list of accessory devices, starting from your oxygen tank, patient comfort devices, functional MR equivalent, emergency crash carts, they are MR compatible, or fire extinguishers, injectors, ventilators, all those complex systems, something as simple as cleaning supplies, biopsy devices, and yes, the intraoperative suites. We have a 1.5T intraoperative suites. It's not a .2T or .5. So all the surgical equipment needs to be looked at. Things are getting complex, but that's okay. If you're equipped, you can handle that.

There are a lot of challenges that we need to handle, but the good news is, again, that we can manage these things, and it's doable, and again, it is 100 percent preventable.

Ferromagnetic detectors, as I said, they're optional, and I can see a use for those, but I can't tell you what is good, but for me, it makes sense that it's a tool that can be used and I have nothing against smart devices. In future, we are getting more and more smart devices. Automation, innovation, all these things are okay. There are some good things and bad things about it, so let's just see what the future holds. Dr. Manny Kanal, I mention him in almost every talk because what he just said, that the new MRI technique is going to be totally different. So we've got to be proactive and keep changing as the needs are, and I think, again, it's a good dialogue, good discussion, and I want to thank you, the FDA, for this opportunity to talk about the issues. Thank you.

DR. RAJAN: Dr. Kanal is our next speaker. Calling Dr. Kanal.

DR. KANAL: Thank you. Apologies. Clinical care never stops. Emanuel Kanal again, University of Pittsburgh Medical Center.

For accessory devices, there are really three areas, some of which we've covered, so it should be relatively brief. The potential problems for the accessory devices in the magnetic resonance environment, again, number one, unclear as to how to safely scan the patients in the presence of the device. That means that the device itself may be poorly labeled, or it may be that, again, the MR scanner may have labeling that contributes to my confusion as to how I can safely scan the patient with this accessory medical device or if I can safely scan the patient.

Number two, possible modification of the device. Now, this is a big issue. A device may be fine, but since I have purchased it and placed it in the MR scanner, personal experience, it goes to anesthesiology, it gets sent back to me, and now, the same cart no longer is acceptable for the MR environment and has to be rechecked.

I'm sorry. Number three, I may have something such as an oxygen delivery system, a tank, a stem, a regulator. I may have an oxygen delivery system that is perfectly fine, but as recently, in a site that I was asked to check, as I go in to look at it, the wrench that's used to tighten the regulator is left on a rubber band, literally, dangling over the top of the regulator. The wrench is powerfully ferromagnetic.

So how do we address these, poor labeling of the device? Nothing to discuss. We've already gone through this ad nauseam through the entire morning. Standardize the device labeling and again, just to emphasize, target the clinical end user and the language and the communication skills that we're using are towards the individuals with whom we're attempting to communicate. In this case, that will be the technologist, it will be the radiologist.

For poor labeling of the device, also make certain that the accessory medical device was tested in the worst-case scenario, and when it is being used, don't just check that it's not going to fly in. Don't just check the functionality of this oxygen delivery system or what have you, but don't forget also to check the MR scanner to be certain that the MR scanner doesn't interfere with the function of the device and that the device does not significantly interfere with the function of the magnetic resonance scanner as part of the integral part of the safety labeling of that device.

Poor labeling of the MR imaging unit or the MR site itself, again, all I can do is recommend over and over that you use the ACR MR safe practice guidelines and standardize your first template based on that. That means use the four-zone design and make certain that, as the document itself recommends, any ferromagnetic or potentially metallic object being brought into Zone 4 should be labeled.

Medtronic brought up in one of their discussions before that they didn't feel it was necessary to label their accessory devices with MR unsafe. I have no problems with that unless someone tries to bring it into the MR scanner. Anything that goes into Zone 4, regardless of what it may be, if it has the potential to be metallic or ferromagnetic, possibly, it needs to be clearly labeled as safe, unsafe, or conditional and mandate that all MR manufacturers, once the installation is complete, I still think it would be tremendously helpful for all sites if we could mark off, at the very least, the three-dimensional 5 gauss line, certainly on that floor, on that site. But if it's going to extend into the floor above, the floor below, that that also be clearly identified.

Lastly, the recommendation of maximum spatial gradient was already discussed, and again, we've discussed why it's limited. I would recommend that we consider replacing that with the maximum force product, or better yet, don't even use that. If we're communicating with a radiologist, if you're communicating with a technologist, say the greatest translational force, the greatest tugging force that will be experienced by a device will be at this location in the space in your room. And it is at that location that this device was tested.

The possible modification of the device, that's a very significant issue. Once the device has been labeled, if it's removed and especially if the device has multiple objects on or in it that can be changed or can be moved or removed, then it is imperative that the site recognize it's under my watch, it's under my responsibility. So when they bring it back in and they've put in batteries or removed or exchanged or now there's an oxygen tank that's no longer the same one that it left with, once the device has left my sight, the label MR conditional or MR safe or whatever is still on the side of the cart, to the tray, the device, but the contents may have changed.

So we don't allow it to leave our site. If it does leave, part of the written protocols of our institution is that anything that leaves with a label, when it comes back, has to get reassessed and rechecked. In our site, at this point, it's almost always checked by the technologist for ferromagnetic content or properties.

Lastly, the loose ferromagnetic accessory devices, obviously, anything that's going to be not just the accessory medical device, but anything attached to it needs to be either removed or adequately tethered so that it can't leave that accessory medical device which is being labeled, certainly, before it can be brought into the Zone 4 regions or the magnet itself.

That brings us to the summary. What are the potential problems and solutions for the accessory medical devices? Worst-case scenario, testing of the device. Check for functionality, not just of the device, but also, the MR scanner. We're not just looking for projectile effects. Standardize the ACR MR safe practice guidelines, not just for the 4-zone diagram, but also, for ensuring proper labeling of anything brought into Zone 4, visualization, clear visualization of the 5 gauss line, how simple that would be to do.

By the way, and you all know this from human nature, if you have a nice, red mark on the floor, people will say what's that? People will tend to avoid going beyond it. It's just human nature. It can only help. Consider standardizing on reporting the maximum force product or where is my translational force in this room with this magnet going to be greatest, and make certain that testing is performed in those areas or reported for those areas. Finally, remove or secure anything that's going to go in with an accessory medical device so that there are no loose things that are there, and anything that is loose that has left your control, when it comes back, needs to be verified, checked or rechecked. Thank you very much.

DR. RAJAN: Thank you.

So I guess we got done a little bit early again and we are scheduled for a break. Maybe we can start in about 10 minutes or so for the roundtable so we can go home early? Thank you.

(Off the record.)

(On the record.)

DR. RAJAN: Okay. Good afternoon. I think we are ready to start the next session, which is the roundtable session, so I invite the roundtable or the straight table Panel members to come take their seats.

(Pause.)

DR. RAJAN: Well, while we are getting started, we were wondering about the answer to a couple of questions. We should have perhaps asked this question yesterday when we still had a full house. How many of you have, at your facilities, a marking for 5 gauss line? That's a small number. How many of you have any kind of marking? About the same. Okay.

The second question is, I should say, third question is how many of you have experienced accessory devices that you started out thinking was safe or reasonably safe to be tied up somewhere and then has decided to become magnetized? Okay. This is a significant problem.

Okay. With that, I would like to open this roundtable for discussion. I invite the audience to come up to the microphone if you have any questions, especially provocative questions, even if it makes the FDA look bad.

MS. VAPORIUM: Hi. Sure. I'm Mary Vaporium (ph.) from Sinai Hospital of Baltimore. I've been in MRI since it was FDA approved.

DR. RAJAN: I'm sorry. Could you hold on just one second?

MS. VAPORIUM: Too loud?

DR. RAJAN: Before you start, I should have asked our members to just go through the table and introduce yourself one more time real quickly.

DR. WOODS: Terry Woods, FDA.

DR. KAINZ: Wolfgang Kainz, FDA.

MR. WEDAN: Steven Wedan, Imricor Medical Systems.

MS. SRB: Anna Srb, Kopp Development.

DR. SCHAEFER: Joe Schaefer, GE Healthcare.

MS. SAWYER: Anne Marie Sawyer, Stanford University.

MS. ROBB-BELVILLE: Sonja Robb-Belville, UT MD Anderson Cancer Center.

DR. KANAL: Manny Kanal, ACR University of Pittsburgh Medical Center.

DR. EDMONSON: Heidi Edmonson, Mayo Clinic.

DR. RAJAN: Thank you.

Okay. Please?

MS. VAPORIUM: Hi. In the early days of MRI, in referring to Dr. Shellock's book for conditional ratings -- well, back then, it wasn't called conditional. It was called compatible. In my own mind, I had a strong sense of what Compatible 1, 2, and 3 meant. Well, now we have Conditional 1 through 8, and I rarely see a three when I look up an implant, but I would like to know why it's expanded to eight conditional ratings and can they be -- so can I, in my own mind, know if something's a Conditional 4, I have a strong sense of the precautions I would need to take versus a Conditional 7 or a Conditional 1. Is there a move to maybe simplify those conditional ratings so that, if I look at a previous history, see an implant that has been researched and it says it's a Conditional 3, then I know what to do with that device. I should know what my precautions should be.

DR. RAJAN: Okay. I think we can start with Terry on this one or Wolfgang?

DR. KAINZ: Yeah. Can I ask you first, where you have and seen this and where you have these one, two, three, four, five, six, seven conditions?

MS. VAPORIUM: That's in Dr. Shellock's book.

DR. KAINZ: Okay.

MS. VAPORIUM: Conditional ratings for implants.

DR. KAINZ: Yeah. What we have, we are not recognizing a Conditional 1, 2, 3, and so on. So we only have MR Safe, MR Conditional, and MR Unsafe. Actually, I also do not know what an MR Conditional 3 is, and I have never seen a definition from that and that's all I know about it. Maybe someone else --

MS. VAPORIUM: Well, I wish Frank were here.

DR. KAINZ:  -- can answer this question.

DR. WOODS: This, as I understand it --

DR. RAJAN: Maybe Terry can make a general comment on the whole concept again, of conditional and compatible.

DR. WOODS: Really, I have the same comment as Wolfgang. I have no idea what the 1 through 8 is. We ask people to label things as either MR Safe, MR Conditional, and list the conditions for which it can be scanned or MR unsafe.

DR. RAJAN: Okay. Dr. Kanal?

DR. KANAL: Frank, are you here now? All kidding aside, I think that's a question you really should address to Dr. Shellock. The numbers and labels that you're using are specific to Dr. Shellock. There's no standardization that's agreed on that or anything of that nature. It's his definitions, and the only one you can ask that question of is him. We, too, don't use that. I don't know what those numbers are, and we don't reference them ever, so you have to ask him.

MS. SAWYER: I just wanted to ask, are you a technologist?

MS. VAPORIUM: Yes.

MS. SAWYER: Are you on the Tech Server?

MS. VAPORIUM: Tech Server?

MS. SAWYER: Yeah.

MS. VAPORIUM: No.

MS. SAWYER: I'll just have to say, you don't have to be a member for the Section MR technologists, but you can go to their website, and you can sign up to be on a worldwide Tech Server, and this would be a perfect question that would go under discussion, and also Dr. Shellock's on that, so he could address it as well.

MS. VAPORIUM: Okay.

MS. SAWYER: I would highly encourage any of the technologists in the room that if you're not on the Tech Server, it's just a real value. There's safety information, Dr. Kanal, Dr. Shellock, and other are on it. There's protocol exchange. There's staffing information, just all sorts of wonderful things, and sometimes the e-mails are like, the boys talk a little too much, and then you just go delete, delete, delete, but then there is really valuable information on there. So get on it.

MS. VAPORIUM: Okay. Will do.

MS. SAWYER: Go to the website.

MS. VAPORIUM: Will do. Thank you so much.

DR. SHULLS: Brian Shulls from Cook Medical. I just wanted to make a comment about Shellock's ratings. As a medical device manufacturer, we have no input to those, and when you call us, we don't know what they mean. They confuse the people that are trying to asses it, answer the questions, and so I really advise looking at the labeling of our device versus looking at the ratings on the website that Shellock has.

DR. RAJAN: Thank you. This is generally speaking. Books can get out of date, and websites can also get out of date, so one always has to be careful interpreting some of the stuff you read out there.

DR. MEYER: Sorry. I'm Jim Meyer from Children's Hospital, Philadelphia. In defense of Dr. Shellock, on his website, he defines them and explains what each one of the conditional measures are. It is there. It's nicely described, and it kind of does give you a feeling, if you are going to use it, and I personally have found MRISafety.com has been very helpful. Again, he's not here. I feel like he does a good job of explaining how those different -- he even has Unsafe 1, Unsafe 2, et cetera.

DR. KANAL: I do think that it's important to recognize one of the strongest calls you've heard from yesterday until now, inclusive, is for standardization. I actually would love to have this continued with Frank here, and he's not, so we probably should not, but I would suggest standardization is the direction I believe we're moving to, and that question is the result of a lack of standardization.

DR. KAINZ: I would ask is there need that we need -- do the technologists see a need that we need subgroups in MR conditional with this? I don't see it actually because the conditions are clearly labeled and the technologist can compare them, what's on the labeling to the scanner, and I personally don't see a need to subgroup them. I am curious if there are any other opinions on that.

DR. RAJAN: Yeah. Let's open it up to the Panel here. I think somebody else --

MS. ROBB-BELVILLE: Yes. Sonja Robb-Belville. There is a need for, again, standardization. We've heard this as a common theme throughout these last two days, and so yes, if we could classify devices as MR conditional, but we need a template that goes through the different conditions that may be present with every scanner, with every situation that would have us make sure that we're following those conditions.

DR. RAJAN: Thank you. This is very interesting that there is one book or one author who seems to have published a lot in this area and whose work is apparently read quite a bit, but still should be careful, it's not standardized in the way in which we do things in a transparent way, in a way in which we can all adopt it and agree and then adopt it. So we should be careful when we try to interpret what's coming out of one book. Thank you.

Dr. DELFINO: If I could just follow up with that question? The FDA has a labeling guidance document that yes, at the moment, is written only for passive implanted medical devices but is applied effectively for anything that we think may go in the MR environment. In that guidance document, it's listed as some information about the field strength, some information about the gradient field, and some information about the SAR level or the RF heating. That, to me, is an attempt at least at standardization, and I would very much like to hear from the technologists and from the radiologists who are making decisions about what is missing in this standardization.

MS. SAWYER: I just wanted to say -- Anne Sawyer -- about going back to Dr. Shellock and also the standardization. To me, they sort of go hand in hand. I really like what Frank's done because devices aren't just conditional or not. There are a lot of gray areas with regards to field strength and SAR and body part and coil. So I don't think you can just say it's either black or white.

I think we have to be, at least in my opinion, more intelligent about how we make these decisions and have all the information and all the specific details, but still clearly laid out to us so we don't turn patients away and so we don't cause undue harm. I think Frank, I really wish that he was here -- I just texted him. I'm like, come back, because I think he's done a really good thing. He's put a lot of thought into this, and also he really understands the challenges of the technologists and the decisions that they have to make. So I think that's important to have that sort of breakdown.

DR. KANAL: Not to mention that it's all in one location. One of the things that technologists have been screaming that they seem to want, that the radiologists want, is to know exactly where to go for the information I'm seeking. If it's not there, I'll go to anybody who compiled it for me, whether it's guaranteed, whether it's legally binding, whether it's up to date, I'm just going to go to where I think I can find it.

MS. SAWYER: The other thing, I should say, on top of that, you have to be careful because I know there are facilities that will go to Dr. Shellock's website, and they'll make a decision based on one piece of information. I think there's a lot of homework you need to do. My radiologist will kid me and he'll say what will Frank say. I'm like here's what Frank says, but I want to know what you say because you're the one that's going to make the decision.

So I think there's a lot of different places that we get information, information we get on the Tech Server, from Dr. Kanal, and so on and so forth. Then, when you to the websites of the manufacturers and the medical devices, my wish is, when you hit that top page, you see MR technologist tab so I can go there and get the information immediately and so that I don't have to dig. What I'm saying is you've got to go to a lot of different places to make sure you've really done your homework to make sure that you've got all the information so you make an intelligent decision for your facility and for the patient.

DR. DELFINO: So there are two separate issues there. One is accessibility of the information, and the other one is the content of the actual information. I understand very much that they're separate issues. I don't want to minimize the first one, but I would like to focus on the second one, in that the specific question is what information do you, the technologist, or you, the radiologist, need in order to make that informed decision.

DR. KANAL: That's very straightforward. We covered that the last session. We need to know anything that might be restricted in the conditional labeling. If it tells me a dB/dt restriction, I need to know it. If it's a field strength restriction, if it's an orientation restriction, if it's an RF restriction, whatever the restrictions are, I need to know it, and I need to know that I can stay within them in order to scan this patient safely.

MS. ROBB-BELVILLE: And as Vera requested, in a common language.

DR. RAJAN: Okay. Just to follow up, what is in the multi-level conditional grading that's in Dr. Shellock's book? Could somebody just give a one minute intro to that? Also, some suggestions on how, if you all think that is something that's really worthwhile, your comments as well and how we might try to adopt it, standardize it, what could be the next steps because I see a risk of having the FDA have a conditional labeling, calling something conditional and having somebody else having a different set of meaning for a multi-level conditional grading. I see this being a potential problem. I would like a little bit more discussion about that. Who would be able to give me a two minute intro to the Shellock conditionals?

UNIDENTIFIED SPEAKER: Can I explain the Shellock conditions a little bit? I'm an MRI technologist up in New York. What he did was, instead of naming a condition after every single implant that he has listed in his book, he just grouped them into eight groups. So if you have implants that you have to wait six to eight weeks, he just says look up Condition 6 or whatever it is. I don't know the exact ones. If you have a condition where you can only go to this point, scan for this long, look up this condition. So he just grouped it together to make it easier, shorter for his book. That's the way I took it.

DR. KANAL: I can also add that, for the first version of that book, we were co-authors, and one of the areas in which we do disagree is whether those conditions should be subdivided. The purpose was to try to make it easier for the technologist. That, I know for a fact. The very fact that the questions exist suggests that it's no longer making it easier for the technologist, even if it ever succeeded in doing so in the past. But I think that the real key is that, even if it makes it easier, if it adds to the complexity because the rest of the system doesn't accept that differentiation, then either the FDA needs to change to that standard or that standard needs to change to the FDA's. I'm not even sure it makes a big difference, but one thing I am pretty confident about is that multiple standards are not going to do us a service.

DR. RAJAN: Joe?

DR. SCHAEFER: I've got one somewhat orthogonal comment, and that is, we've been discussing this possibility of this fixed parameter option under the IEC modes, and it's conceivable that the devices would be approved only for operation in the fixed parameter option. So there should probably be some provision, or basically, this is a heads up since it's not a done deal yet, but it may be by the end of the week. I just want to mention that.

DR. EDMONSON: Heidi Edmonson. I might just jump in here and mention that I recently had a conversation with a technologist that works for a mobile company, and while she's in the mobile unit, the only internet connection is going straight to the scanner. So her only MRI safety resource is that Shellock book. There's no Googling anything to check it. That's her one and only resource while she's working.

DR. RAJAN: Okay. Well, maybe we should go on to the next question.

MR. GILK: Tobias Gilk. I was curious to ask the FDA and if any of the other Panelists want to chime in on this as well. There are lots of accessory medical devices out there that are marketed as nonmagnetic or MRI friendly or warm and fuzzy, lots of ambiguous terms that describe the magnetic safety, the projectile safety aspects of all sorts of accessory medical devices.

What does the FDA think or feel or how do you respond to these ambiguous markings that, as far as I'm aware, have no substantial definition behind them, no criteria, and it just becomes a fluff marketing term but conveys the sense that there are specific parameters that may or may not be true?

DR. WOODS: Terry Woods. If it's a medical device, the purpose of our Office of Compliance or one of their purposes is to investigate false or misleading claims, and I would encourage people who think that there may be false claims to contact the Office of Compliance. If it's not a medical device, unfortunately, we don't have any control over that. You could, perhaps, try the CPSC, the Consumer Product Safety Commission, or someone else. Naturally, we don't like the idea of people labeling things that are not accurate, and again, talk to the Office of Compliance if it's a medical device.

MR. GILK: Could you help me to better understand, when, on the provider side, we look at everything that's not inside the patient. Everything that comes into that room in support of patient care as an accessory to the MR, can you give me an idea of what types of items that might be brought into the MR environment that are FDA regulated as opposed to those that aren't? Is the MR fire extinguisher or the wheelchair or the step stool?

DR. WOODS: Well, again, we don't regulate fire extinguishers, but we do regulate wheelchairs, I guess, medical device is something that --

DR. RAJAN: Class I exempt wheelchairs?

DR. WOODS: Well, you're right, yeah, but a wheelchair is a medical device.

DR. RAJAN: Yeah.

DR. WOODS: It's something that -- I'm fumbling here on this definition, but a medical device is something that's used to treat a medical condition. So things like buckets and pillows and most blankets are not medical devices.

DR. RAJAN: Dr. Kanal?

DR. KANAL: I think this is insolvable. As an illustration for how complex it is and that you won't get what you're looking for, the oxygen tank is actually regulated in the color of the tank, the cylinder itself, it's regulated by the Department of Transportation. The stem is controlled by the Compressed Gas Association. If you take the stem, the regulator, and the tank and put them together, now it's meant to deliver oxygen, and that's under the domain of the FDA. So if you take it apart, it's no longer the FDA's.

So I'm not sure you're going to have an answer to the office chair or the bucket or the fire extinguisher specifically bought for the MR environment. That's maybe what we bought it for, but it's not a medical care delivery device.

DR. RAJAN: This reminds me of an incident a long time ago when I was standing with a technologist across the patient bed, and I noticed that, in slow motion, the little brass cap on the top of this pencil just unhinged itself from the pencil and flew inside. This was a standard brass, the yellow pencil. So it is a tough problem to catch all these sorts of things, what accessories. Okay. We've got one more comment.

MS. SRB: Anna Srb, Kopp Development. I think I understand was, what Toby was asking is the danger of ferromagnetic detectors because we often, when we going on site to install our detectors, the MR care providers will tell us, well, your detector is -- by my nonmagnetic wheelchair or nonmagnetic gurney, which is not correct labeling because it's indeed magnetic because if magnet sticks to it, it's magnetic.

So, therefore, what his question was, if I understood correctly, is how do we avoid this confusion of the false labeling nonmagnetic. Also, before we labeling, you think it should be tested, and the testing should be also standard for those devices as well. Somebody mentioned -- work yesterday with respect to testing the ferromagnetic detectors, but the conditions in which those ferromagnetic detectors were tested were not clinical. They were not tested in the environment where they intend to be used. So I just wonder how the gurneys or wheelchairs are tested as well. Are they tested in somebody's conference room, living room, or is it tested in the MRI environment, and how is it being tested? And, therefore, how important that false labeling is.

DR. RAJAN: Well, that is, indeed a good question. I don't believe I've ever seen performance data to test extended ferromagnetism. Maybe somebody else from the Panel could address that, that that is, indeed, an issue.

DR. KAINZ: Well, I mean, if somebody advertises a device or an accessory as nonmagnetic and they submit an application, they have to say what kind of metal it is in the summations, but it's entirely conceivable that there is some change in time or it's not a binary situation.

DR. RAJAN: Okay. Next question.

MR. STECKNER: Michael Steckner, Toshiba. Simplicity versus complexity, I think, especially in the last session, less so on this one. We were hearing various opinions across a range of items of, let's make this more simple, let's standardize, complaints about complexity, it needs to be simplified or cases of a certain physician saying please give me more information so if I want to do something off label, I'm appropriately informed. I think these are reasonable things.

A lot of the things that are being discussed, we're working with some leading edge or bleeding edge technology, and there's risk and reward. Physician makes the determination. The patient could bleed. We're trying to use the technologies to help. There's initially a lot of complex information that needs to be digested and understood, and over time, with experience, there's an understanding of what can be simplified, can be made easier to operate, and ways to simplify, and there's some people who are concerned about that progress, as we've heard. Others who welcome it, depending on what customer you speak to.

So the question becomes, is what are the guidance that can be given to the vendors to help us decide when to try something more simple or more complex. I mean some of the things are already in action, but some things go very slowly. For example, the Joint Working Group has been in existence for basically five years to produce a document that's not even a standard. It's a technical specification. It's a level below a standard because it's so complex and it's so new; that's our way to grow into the situation ,and that standard will be published shortly, and then we're going to be continuing to work on it, and it's going to be perhaps another six years before that turns into a standard. So these things take time to get out so that people can benefit from these things, but it also takes time to learn from that and make the simplifications.

So the question to the Panelists is what's the guidance? When do we make things more simple? When do we release more information so that appropriate decisions can be made for unanticipated off label uses? Thank you.

DR. RAJAN: That's a tough one. Okay. Wolfgang.

DR. KAINZ: Let me try this in order to add something, which then eventually gets to a question to the MRI manufacturers. So we face the following problem that there are medical devices, accessories, or implants which simply cannot be labeled at the normal operating mode, particularly for heating. So we have now, as a regulator, the choice to either say to the manufacturer, okay, we do not grant you any labeling because you cannot scan it 2 W/kg safely. Everything below 2 W/kg, you cannot dial in.

As I have heard in these two days, MR technologists have a lot of problems to understand the information. So we face the problem, what are we doing? Do we allow labeling with 1 W/kg for an accessory or if it's 0.5 W/kg? And we have done this, so this labeling is out there. My question now to the MRI manufacturer and to what you said, Michael, simplify it and what additional information would be needed? How can we solve this problem? Because it cannot be solved from a regulatory point of view. It can be, but only by not allowing labels below 2 W/kg, and it can be done so by manufacturers of medical devices, accessories, and implants. It can be only solved by the MRI manufacturers to provide some simple, easy to understand way that medical devices can be scanned below the normal operating mode minutes.

DR. SCHAEFER: This is Joe Schaefer. One of the motivations for the fixed parameter options, as you know, is they have another level where we control fields that are generated by the scanner to a certain level that perhaps more implants might be able to use. That's the only way we can deal with it. If there's a device that falls below a level that we can guarantee we designed our system to and tested it to operate, if we don't have that situation, we can't possibly consider it to be safe.

So we just have to say, anything that's below some level -- it may not be 2 W/kg, it may be whatever this FBO turns out to be, would be the level that we could operate to. If .1 W/kg is safe and .2 W/kg is not and we're not designing our system to be that accurate, especially at the low end, then it would not be so.

DR. KAINZ: I was more talking about 1 W/kg than 0.5. This is labeling which we have seen needed for some medical devices, external pump fixators and other devices. So I am not talking about the low end. An example, Medtronic brought about the newest stimulator, which has 0.1 W/kg. So I am talking more about 1 watt, half operating mode limit.

MR. STECKNER: That's a useful example. I mean, you can have various vendors who are struggling to get first technologies out, early adopters. We've heard all of these phrases. They come out with different numbers. They're pushing the best technology they can to have different numbers. Then, the technologists need to train to understand what this mumbo jumbo of numbers actually means and how to interpret, like we've been discussing the last few days.

Then the bar gets raised, the technology improves, we get together, we discuss, we need to have standards on this. Standards eventually come out. The vendors come out with the controls on their new products. The old ones don't have it, necessarily, depending on how things are handled. It takes time for that all to disseminate out and then learn from that and then simplify on the second round and hopefully make it a more of a push button operation.

It's just the way things go, and I just want everybody here to understand that that's the march of technology, and hearing the invaluable feedback from the last couple of days does help identify some of the trends and some of the needs.

DR. KANAL: If I understood your question correctly, Dr. Steckner, apparently what you're saying is when do we simplify and when do we provide a dictionary's worth of information. If I may try to respond to that based on an analogy that we've used earlier this week, nothing to do with the manufacturer of the MR system now, let's go to a stent.

A perfect example, a stent, and it's been tested, and at 1.5T it's been found to be safe. It's acceptable, MR conditional 1.5T, and then the other numbers are given as well, and there's quite a nice safety standard and plenty of buffer. Then how many e-mails do I have to get that will say it's not been tested yet at 3T, Dr. Kanal, what should I do? So I don't believe your question is necessarily restrictive. You're saying should we be simple or should we provide all the information. As a physician, my answer is always yes, do both.

In the official label, keep it simple. This is approved to take that stent at 1.5T at this dB/dt, at this SAR, it's approved. Now, did it deflect 44 degrees? Did it deflect 1 degree? Did it not deflect? So if you have somebody who has the potential to benefit from what you would call an encyclopedia of an additional knowledge, as long as you can make it in a way that's readily accessible to those few that would want it, absolutely, you should provide that information, whatever the legal framework of making certain that it's there without imposing any legal exposure, but if you could provide the information, what a tremendous benefit to all of us if Dr. Smith can go online and say it deflected 2 degrees. Oh give me a break. I'm not worried about heating because it's a pellet. Go ahead, put him in.

It's tremendously helpful, but I don't think we should be cluttering up the basic approval with all the encyclopedia, and I think that applies to your question about the manufacturing of the system itself.

MR. STECKNER: To a certain extent, yeah, but I'm also thinking of what you had said earlier when I said four vendors report the spatial gradients covers on and one off, and your comment, I believe, was something like, I don't care, let's just make it simple, let's come to an agreement. And that's something for standardization. There's a need there, and the vendors have attempted to respond, each individually in different ways. Reporting covers off is a very conservative way and safe way of giving the number because it's the worst possible case. Then, on another portion of the conversation, for example, we were talking about, I think, labeling for something, and you had, if I remember correctly, indicated it would be nice to have more information because, for your off label usage, that extra information could be valuable. So it's how do you balance those things?

DR. KANAL: On the first statement you made, if I may address that and correct it, I don't believe anyone said who cares. It was the exact opposite, and I think you actually have heard that from Charles Stanley. He said that we've been asking for I don't know how long, and I think Frank really led that cause for the longest time, and we just didn't have access to that information. We very, very much did care.

Even if it wasn't standardized, we just wanted to know and had no central repository where we could identify the answer to that question, so no decisions could be made accordingly. It was so frustrating, as you saw, we published an article saying what we don't know. As Charles brought out, we had to come here today to hear the answer to that question. You know how long it takes to get an article researched, written, submitted, revised, approved, published, and then a few months later, we get the answer here.

So just recognize, from our side of the table, we have nowhere to turn to but you guys, and there's a lot of you in you guys. I can't just say Mike, can you answer the question for me. If I'm Dr. Joe Smith, it's not a reasonable present situation, but the other side of things, I completely agree with what you said. If there is a place we could go to, keep it simple. If there's a place that the interested and the initiated can go to for further information, I don't even see why we would dream of withholding it.

DR. EDMONSON: Okay. When I go to, say, one of the manufacturers' websites for their MRI safety-related information, I see a sheet of information for the patient and I see a sheet of information for the healthcare provider. Would it be possible to have a sheet of information for the MRI expert? Because even if they don't want to provide any more information, it would be useful to know that maybe they just did the bare minimum and that's really all that's been tested.

DR. SCHAEFER: I just have one quick correction. The IEC standard that's out currently came out in 2002, and it required all vendors to report the maximum force product, the maximum gradient, and maximum B in their locations, and I think everybody's been doing that for all these years. The maps and stuff were not required, so I just want to correct that.

DR. RAJAN: This is where the Philips person should jump up and say simpler settings.

Okay. Let's go to the next question.

MR. WEDAN: So just to comment on the pencil, I've been working very hard night and day to get that MR conditional label on those brass tips at the end of the pencil.

Part of the problem with the MR conditional label today is if you go to any vendor that sells accessory products, and when I say accessory products, those products will range from a plastic garbage can to things that are intimately connected to the patient and provide healthcare information to the patient, the MR conditional label is going on everything.

Recently, I've been fortunate enough to have received several new magnets, and we're deploying furniture objects all the time. A plastic garbage can has an MR conditional label on it. Now, this may not be the purview of the FDA in any way, shape, or form, but somewhere, like Homeland Security, the issue is we have a failure to communicate. This might be an EPA issue, but it certainly is a Zone 4 issue, and anything that goes into Zone 4 ultimately may or may not affect a patient or a worker adversely.

So when there's a conditional label, and we've heard this in different cases, I think it should be completely unreasonable to allow the use of a conditional label without a statement of what the conditions are. That should be some kind of law somewhere. It's just unreasonable to deploy a label with no conditions. We can come up with what we think the conditions are in some cases, but not in all cases.

So turning this around in terms of standards, the labels have standards, they are well defined, they've been redefined, and the label itself in terms of, for this equipment has changed. But the question is, the vendor selling us these products, telling us they're MR safe, at what tesla, all the words. If it's conditional, how, when, and where. If it's not conditional and it's MR safe, it needs to be labeled appropriately as well. If it's not conditional but it can actually kill someone, is it MR hazardous? And that needs to be evaluated.

Again, this may not be the implanting device, where if we use it wrong, it can kill someone, but it can be used. Where is that line? It's not just the technologist that's involved. It's in the people that plan the architectural composition of these centers, Zone 3 and 4. If I check the guidelines, it says no loose metal objects in Zone 3, not Zone 4. So conditional needs to mean something.

DR. RAJAN: Thank you. Any comments from the Panel on that? Okay.

DR. KANAL: There was one other questions that I was going to bring up to the FDA for this Panel, so we may as well bring it up now. Is there a definition of MR unsafe for the FDA? Hear me out.

We have devices. I'm going to just use it as a metaphor. We have a device that I would like to get approved, and it's an accessory medical device, as this is discussing. It's an EKG machine or something, I don't know, and it is 100 percent demonstrated to be safe as long as it is kept somewhere outside of the 5 gauss line because anything above that, holy cow, it's going to explode and cause World War III.

So do we call that MR unsafe, or do you call it MR conditional as long as it's exposed to no magnetic fields? Where is your cutoff that you won't accept a -- and the reason I ask, because there's some devices that were labeled MR conditional, and they fly into the machine and hit the technologist and cause --

DR. RAJAN: There's a lot of precedence for that. Maybe I'll let Terry take a crack at that.

DR. WOODS: Well, for that, we have cleared accessory equipment that was labeled MR conditional within a certain distance or outside a certain gauss line. The thing that we required people to do then is give us some means for physically restraining the device, you know, bolt it to the wall with a strap that's not going to break so that it can't go closer than it's supposed to to the scanner.

DR. RAJAN: Lockable wheels don't count.

DR. KANAL: I recognize that, and that is the present state of the art. The confusion is a crowbar is MR conditional, and at what level does the FDA mandate MR unsafe?

Remember, for example, from the American College of Radiology guidelines, we were exquisitely careful to follow one mandate that is universal in medicine, which is don't tell me how to practice medicine. So we were extremely careful not to design into those guidelines, you cannot bring it. You can do whatever you want. You just can't claim you didn't know, so you have to label MR unsafe to a piece of equipment that can hurt that patient, but you can bring it anywhere you want to. Just, it's your responsibility, Manny, if you're bringing it in. So where does the FDA say to this manufacturer, I don't accept an MR conditional labeling on this 72-pound piece of equipment? It's not MR conditional in my opinion. I now want this labeled MR unsafe.

DR. WOODS: I guess, in general, it depends on, this is something we look at on a case-by-case basis, and so we don't have a hard and fast definition. It depends on the rationale for why they need this equipment in this certain space, and if we're confident that it really will be restrained the way that they say it will be restrained, so there's not a black and white.

DR. RAJAN: If I may add, there are many examples of devices that can be operated in the console area where now we do require that's MR unsafe because that stuff either will not work or is too risky to take inside the scanner room. So that fits in, sort of, the group that you were talking about.

DR. KANAL: I would like to suggest -- it doesn't have to be debated now. I'm just making a suggestion for your consideration. I would like to suggest that you markedly expand and formally define what you would consider MR unsafe. There is nothing wrong with a manufacturer of a device, an accessory medical device receiving a label of MR unsafe. The only thing that could happen if they receive that label is that I will be held that much more responsible before I decide to bring that into the room.

Then the device manufacturer, some of you are here. We would all like to hear from you. You're certainly encouraged to respond. No, no, no. We want MR conditional. I would like to hear the reasons because the only thing that could happen by calling it MR unsafe is it draws attention to the fact, this is so clearly unsafe that we're going to give it a red sticker. So you want to use it, you really better know what you're doing. If it requires tethering that can't break, please forgive me, I would call that unsafe in the MR room. I may have to tether it, but one thing's for certain, it's unsafe, and I would love to see the FDA consider expanding that definition as well.

DR. RAJAN: Okay. Should we move on to the next question?

MR. KOPP: Keith Kopp of Kopp Development. We're a manufacturer of ferromagnetic detectors. This is actually related to this last which question, which, Terry, you partially answered. We run into a number of situations where our detector detects ferromagnetic material, conditional ferromagnetic, such as a ventilator, patient monitor, et cetera. So we are asked regularly of what's the best way to know that we're not getting it too close to the magnet? You're talking about tethering. A lot of sites have a 5 gauss or multiple rings.

I would like to really get two answers. Is there anything the FDA recommends for conditional items that are traditionally used that are on wheels, for example, like ventilators, traditionally used inside the magnet room to keep them at the MR conditional distance, which is a gauss line, generally? And the second question is, the technologists and that, I've heard a number of sites say yeah, tethering is great, but it's in the way. We have a small room. We have three pieces of equipment that have to have different tether lengths, and they get caught into each other, so on. So they're saying tethering is another safety issue that you're bringing up. So I would like to get a consensus of what people feel about this.

DR. RAJAN: Turn to the technologists on the Panel?

MS. ROBB-BELVILLE: Sonja Robb-Belville. One thing that I would like to point out is that a lot of sites don't have 5 gauss lines marked. As an educator over the last 11 years, one of the favorite things for me to do is to see how everyone's suites are set up. I would say probably 25 percent might have that 5 gauss line marked. Then, yes, you are, as a technologist, dealing with the size of the room, the restrictions of the room. We do have an interventional suite in MD Anderson. It is a very large suite, so it does provide the opportunity for that equipment to be tethered and moved relatively easily, but it is still something that we struggle with.

MS. SAWYER: Anne Sawyer from Stanford. We don't have the 5 gauss marked or 3T generally because the 5 gauss is contained within the magnet room. Now, there's 7T. We have 121 gauss in the control room, which keeps me awake at night, but I have a wall at the end of the door at the 5 gauss line for that magnet.

With regards to tethering, I'm not in favor of it at all. I don't like it all. I won't allow it in my magnet suites. Speaking of, I have to fight with some of my faculty because somebody else is going to say oh, I need to move this, and then they're going to unhook it, and so I'm just not in favor of it at all.

Also, the other thing that I worry about, and I maybe have limited knowledge of this, but I'm always concerned that it might become magnetic over time if it's in the magnetic field, depending upon what it's made from, but I don't like tethering at all. If it needs to be tied on a leash, it shouldn't be in the magnet room.

Having said that, I know everybody has limited space, but I also intermittently am asked to review designs for MR facilities, and I have always said, and people always say well, I don't have enough money to do that. Well, if you don't have enough money to do it, then you shouldn't be in that business because something bad is going to happen. You're planning for failure.

MR. KOPP: Well, actually, the question was not so much about the 5 gauss line, because that's generally near the walls and that's things that will be outside. I'm talking, take a specific case like a ventilator which has to be in the magnet room, but most of the ventilators either are ferromagnetic enough to be sucked to the magnet or they stop functioning properly when you go above, say, their 250 gauss rating. How do you handle this kind of situation, and how do you accommodate the things that can be greater than 5 gauss but are not limited in the static field?

MS. ROBB-BELVILLE: Well, I'll speak to an example where I went to check on interns at a clinical site and literally walked in minutes later after a -- it's not a ventilator. The nurses were calling it an MR compatible IV pump that was not fully secured to the IV pole. When you look at the labeling for the device, it was to be mounted on a countertop or a wall or something of that nature, not on an IV pole. So the result of that is that a child now has stitches on her face.

MS. SAWYER: I just want to say, I don't work in a hospital. I work in a research facility, but we do have MR conditional ventilators that we use on anesthesia cases in my facility, and we know how and where to use them, but they don't need to be tethered because they're safe or they're conditional at a certain distance, but we use them all time. Again, it involves knowledgeable MR technologist and also techs as well. Am I wrong, but is there not MR conditional or MR safe ventilators that can be used in a clinical situation with patients? Yeah. So what are you saying? That they didn't want to pay the money to get the MR safe? I'm sure we see that.

MR. KOPP: Well, yeah, that happens a lot too, but there are even, at least from my knowledge of it, the ones I've seen that are meant for the MR environment always have a restriction of the maximum magnetic field that it can be exposed to. There may be some that don't know. Some of the manufacturers actually build in an alarm that, when you bring it in closer than that, the alarm goes off. So people are doing it different ways, but there's a whole range of this equipment. Some of the new patient monitors, for example, can literally be brought up to the bore, but that's, from what I've seen, kind of the exception, and that's not the population of equipment that's in a lot of the sites that we have come in contact with.

MS. SAWYER: I mean generally, at my facility, we have a standard rule. Length of the patient table all the way around the magnet, which may be, but unless you know exactly that that monitor or whatever you're using has been tested and is fully safe, and when we purchase these items, these things are reviewed carefully with the vendor to make sure that we know exactly what we're doing and how they're going to be used with that particular magnet.

DR. KANAL: Manny Kanal. Once again, I think we are creating the problem that you are asking about. For whatever reason, the industry has standardized and the entire industry has standardized that if it deflects less than 45 degrees, we will consider that acceptable, and we even give that a label, MR conditional. If it's not a threat at all, we even go so far as to call it MR safe.

I have no clue how we have deteriorated, in my opinion, into a situation where, if it's above 45 degrees, we change it from MR conditional to MR conditional. It doesn't make any sense. My proposal is that anything that deflects anywhere in that room, up to and including maximum force product location, more than 45 degrees, is MR unsafe. You're still at leisure to bring it in if you feel responsible and you're going to do it and you're going to practice tethering. No one's telling anyone how to practice medicine. We're telling you, this is a potential projectile, that's all, and isn't that the real reason for the label in the first place?

The problem is -- please forgive me -- is that we, the official agencies, the FDA, the entire industry has said this is MR conditional. This 72-pound piece of equipment is MR conditional as long as we keep it the heck out of the room and it doesn't make sense.

MS. ROBB-BELVILLE: What we forgot to tell people is that it's conditional, not compatible as well.

DR. WOODS (I guess this is not Dr. Woods speaking): I guess, in our intraoperative suites, we proactively, at installation, used colored floor tiles to demonstrate where the 5 gauss mark was, but there were very specific requirements from the infection control folks that we could not paint a line or put tape down on that area. In some of our other suites that are not used for intraoperative procedures, we have marked some tape temporarily when we know that we are handling MR conditional or potentially unsafe devices in the MRI scanning room.

DR. RAJAN: Okay. Thank you. Let's move to the next question.

UNIDENTIFIED SPEAKER: If I'm listening to, all day, I'm getting the impression that devices are tested against variable parameters, that the test conditions are or test procedures are sometimes adventurous, that the labeling is often confusing, and the parameter specified in the labeling can hardly controlled by the radiologist with the machine. So there's no possibility to dial in the parameters in the machine to control it.

So from an outside, this is rather nonsatisfactory. Luckily, we also heard that the manufacturers are implementing this FPO, that the standards are moving into the direction of defining sound and comprehensive test procedures to test against this well-defined parameter sets. But the radiologist can dial in, click on FPO or not FPO, and then everything is defined if the labels say it's compatible and safe and the FBO conditions. My first question is, is my impression correct of the situation today, and second, if yes, does FDA consider to require that the devices which are labeled today in these various conditions be retested against new requirements?

DR. RAJAN: Wow. Joe, I believe you're the person to answer the first question since you like FPO so much.

DR. SCHAEFER: Repeat it a little bit. I think I got lost there somewhere in it.

DR. RAJAN: The question, as I understand it, is -- if you could repeat it yourself.

UNIDENTIFIED SPEAKER: So the impression I have is that the parameter sets, devices are tested today are quite variable from test houses to test houses and may not reflect the parameter space today in the scanner world, that the test procedures are sometimes very adventurous, not technically thorough, and that the labeling is confusing and the parameters specified in labeling are not able to be controlled by the radiologist using standard software.

DR. SCHAEFER: Okay. Yeah. This is Joe Schaefer. I have to start out by saying I haven't actually done the test, but this morning, I believe, Johan presented a paper showing there's some issues with the procedure that's used, or there's some ways we could extrapolate it a little bit better, which is one aspect of it. There seems to be widespread agreement among some of us that going to this idea of force product would be a superior idea. I guess, on the other hand, to my knowledge, there aren't a lot of reports of injuries in the last few years from the number of scans that have been done with people with implants. Must say that the limits of force and all that are quite conservative, and maybe that's why there's not been an issue.

DR. RAJAN: I think he was also referring to the RF heating, so from the FPO perspective, whether your assumption was correct, that it's going to be an improvement in terms of picking these settings from a machine so you know what you're setting it.

UNIDENTIFIED SPEAKER: (Off microphone.)

DR. KANAL: So you're not talking about accessory medical devices at all. You're talking about what's inside the bore, being scanned?

DR. RAJAN: I think he went back to implant. I think a general consensus was your assumption is correct if FPO will improve the situation in terms of testing, but maybe we can address the second part of your question, which is, what would FDA do if FPO becomes a reality? How would FDA handle previous devices that have been cleared with different kinds of testing? I think that I can turn that over to Wolfgang and Terry, but something tells me that maybe we don't know the answer to this question.

DR. KAINZ: This is Wolfgang Kainz from FDA. So as far as I know, we will not be able to force manufacturers which have existing labeling, based on the test methods and on test conditions, which was done previously and now it changed to other parameters like the B 1 field instead of SAR. We have no restrictions, no authority to require them to do retesting, but we certainly are willing or more than willing to work with them if the companies decide to come to FDA, and we are open to work with them, to do the best, to adapt their, what you called adventurous testing to a better testing and to adapt it to these new parameters.

As I said before, I think it's a problem. You mentioned the issue that the parameters cannot be dialed in, and I think this is -- I had it before, the question to the MRI manufacturers, what are we doing about SAR levels less than the 2 W/kg, and in the interest of the patient, we have accepted such labeling and labeling is out there. I think we need to now work all together to resolve this problem. As I have just heard from Joe, have a fixed parameter mode as it will come hopefully as soon as possible with one levels close to the normal operating mode, but they're done, hopefully, as quickly as possible to have all the levels less than the 2 W/kg at 50 percent or maybe at 25 percent mode with fixed parameters.

DR. WOODS: Terry Woods. One thing I would like to add is that even when fixed parameter options becomes a reality, the current labeling is not going to be obsolete, and the manufacturers have demonstrated that their devices are safe to be scanned under a particular set of conditions, and as long as you continue to scan and satisfy those conditions, then it should be okay. So, again, fixed parameter mode doesn't mean everyone has to redo their testing. There is additional testing and analysis that they could do to extend the claims that they have, but it doesn't invalidate what's currently in the approved labeling.

DR. RAJAN: Okay. I think Johan was first or were you? Go ahead.

DR. VAN DEN BRINK: Johan van den Brink speaking, Philips. A question related to accessories and regulating medical practice, which has a relation. What I find as an interesting question from myself is that, again, we contraindicate a use of devices that have not received a compatibility statement from -- MR manufacturers in the close facility of the MR system. Obviously, many systems are brought into our rooms, brought into your rooms, basically in close proximity of our systems that likely do not have such a compatibility statement from our side.

What I would like to learn a bit more before leaving this room is how does FDA perceive the use of all kinds of equipments in the facility of the MR system that may be labeled as MR Conditional, probably, but has not receive such compatibility statements. We are contraindicating that use.

DR. WOODS: Terry Woods. I guess this is really no different from the accessory equipment than for all the implants that are contraindicated currently, so I don't see a real difference there. I guess what we've seen in interventional MR and a lot of new procedures, the medical community is telling us that they need this equipment in the vicinity of the scanners, and I guess the position that we've had from the third party equipment is you come and try to demonstrate to us an MR environment in which you can safely use your device and we will consider that evidence. And if we agree with it, work with you to come up with the labeling.

DR. VAN DEN BRINK: Well, the reason why I would perceive it as a different situation is that these devices are used on a more regular basis in a wider variety of patients and may have serious EMC impacts on the functioning of our device. From my perspective, which has more impact on the labeling of our devices and potential safety and efficacy it requires? So I really think that the problem is regulations should be adhered to very strictly.

DR. WOODS: So I guess, how would you respond if, I don't know, a ventilator manufacturer came to you, then what's the process for coming up with a certification where you say that the equipment is safe to be used with your system?

DR. VAN DEN BRINK: We offer a verification and validation procedure with them as a service, and you can get a compatibility statement.

DR. WOODS: I guess do the other scanner vendors do the same thing?

(No answer.)

DR. RAJAN: I think we should move on to the next question please.

UNIDENTIFIED SPEAKER: Actually, I was interested in an answer to that question, but silence is golden sometimes.

DR. RAJAN: I think the question was directed to Johan. Terry, could you repeat the question?

DR. WOODS: Yeah. I would just be curious to know what the other MR vendors do, if they have such a similar process for certifying additional equipment to be used for their scanners.

DR. VAN DEN BRINK: Well, Siemens has such a process, and equipment which is, well, like regulated and so on should have this compatibility statement. Of course, some, there is no need for this kind of compatibility statement. For example, if you want to take a chair in the controlled access area, therefore, there is no need for permission.

DR. RAJAN: Yeah. I think there is a spectrum of devices, and certainly, I know there are many devices where we do see cooperative agreements; when they see devices that use the machine's function, there is collaboration or notice from the OEMs for the use of the device. But there are also other devices where it's like an infusion pump or something that's sort of outside and it's not really that critical to get. It will work in all MR systems.

Okay. Joe?

DR. SCHAEFER: Well, I'll make a quick statement here. Joe Schaefer. I think there's some differences between the way things are done in Europe with this than in the US. Typically, unless we vendor some device, we don't really test it or go into the compatibility work with the vendor, at least to date. That's always subject to change. I think there may be some differences there.

MR. FRESE: Georg Frese from Siemens. Well, I think this could explain why we get so many complaints about interfacing problems from the States, but not from Europe. An observation.

UNIDENTIFIED SPEAKER: I just have one thing to say. Joe, you use that word compatibility; don't use that word, please.

DR. VAN DEN BRINK: Okay.

MR. KOPP: This is a perfect segue --

DR. VAN DEN BRINK: Johan van den Brink. The word compatibility comes up because there are other standards that we are using, talk about EMC compatibility. It's going to cause another piece of confusion. I appreciate what you're saying, but it's there. I'm sorry.

MR. FRESE: Maybe I know the difference, why from Europe and America -- the Medical Device Directive has a certain article if you want to combine medical devices. Then you make a declaration according to Article 12. Internally, we are calling it Article 12 declaration, which means MR compatibility statement.

DR. RAJAN: Yeah. I should also point out that electromagnetic compatibility is also a pretty important part of FDA review process for all these types of devices, so it's not overlooked. Okay.

MR. KOPP: Well, I think, as a segue to one of my questions, I think part of this, if we looked at it in detail, we would find that UL and IEEE standards may be overlooked in the way we do business in the United States because there is combining of accessories concept that we have here. It's just that inspectors today are entirely unable to identify the physiological monitor that's in my induction bay, that it might be going into my MR room to be connected to a patient during a scan. Make that connection and say hey, we haven't certified this.

That being said, I want to get back to the projectile question. I think the statement of about 75 pounds is very true today. There are a lot of things that are going into the room, but not everything that's a projectile necessarily hurts and not everything that hurts is a projectile. So part of the problem defining hazard; my son can shoot me with his Nerf gun, pretty much, I'm still alive, standing here to tell you about it, but there are transformers that have to go in the scan room. I don't like them. They're our patient comfort systems that include motors that need to be services that go into the room, and then there's the medical devices themselves, ventilators, physiological monitors, every kind of device that you can imagine.

To me, the line has to be, if that device is going to kill someone or has killed someone, the MRI hazard label needs to be in place; otherwise, what does it mean? What does the circle or the red line through it mean? It says can harm or kill; otherwise, everything's conditional. We don't need that other sticker. So we just need two and we simplify the process. If it's killed someone, or it very clearly can kill someone, we need to consider this. For the implanted devices, it's hard because, used wrongly, they can kill someone, but for the projectiles, these are two different and distinct categories at some point. A projectile flying across the room can kill anyone in the room at any time.

DR. KANAL: Manny Kanal. For what it may be worth for this audience, the incidence of which I am aware, people have asked me what is the smallest ferromagnetic object that you've seen that has caused injury to a patient. It's hard to define injuries, so I would just like to use common sense. The smallest that has caused what I would consider true injury that I'm aware of personally is not a 72-pound monitoring device, but is nail clippers, as I have mentioned before.

A hairpin in another patient was overlooked, and as they went to the magnet, it flew into the nasal passages, and that patient required surgical extraction. I'm not sure that's injury, but it's certainly intervention.

I completely concur with the statement that it doesn't need to be a 72-pound device. My concern is that we approach the labels of MR Safe, MR Conditional, and MR Unsafe as if we're doing it from a regulatory point of view, but at least to the best of my knowledge, that's not the intent. The intent is educational. The intent is to advise the site this is MR Safe. This is MR Unsafe. This is MR Conditional. For all three of those, there might be conditions of use that are acceptable. It may be that the acceptable way to use it is just leave it out of the room. It's still a condition of use that is still acceptable if you judge it to be, but we can't continue with the present situation where everything is called conditional as long as someone sets some kind of a standard for which they think it's going to be all right.

There are so many things that we left to these two days to discuss, very few of them are exquisitely concrete and so unbelievably readily accomplishable. If it deflects more than 45, it's unsafe. There's nothing to discuss. It doesn't require a reviewer to have to make a decision. It means that you just communicated to the end user, this item is unsafe. That one is safe. You want to use it safely, keep it far enough away, and if they want, the manufacturer of that device can tell you, by the way, in order for it to not -- think of what we're talking about, deflect 45 degrees -- let alone if you put it on wheels. In order for this to not find itself becoming deadly, you have to keep it beyond 200 gauss, 300, whatever it is, fine.

You can add that label. It's not mandatory. I already know it's MR unsafe, and that certainly puts this site on notice for what you really want them to understand, but calling everything conditional just because someone made it safe under a certain weird condition is doing a disservice and, in fact, it is negating the very label and unsafe in the first place. So that's something I think you can accomplish.

DR. RAJAN: Now, there is a minor counter argument to your argument, if I may. That is, if you have to consider a world where everything is either safe or unsafe, you'll really have a difficult time to figure out which of these boxes you can carry and keep by the side of the wall. You wouldn't be able to figure that out because, clearly, you'll be forced to eliminate all the devices, unless by trial and error, if you're going to take the chance. At least now, if something is conditional, you might have the courage to carry it to the 2 gauss line and check it out.

UNIDENTIFIED SPEAKER: No. I disagree with that. I disagree. That's where your physicist comes into play. That's where your safety officer comes into play. If it's indicated and it needs to go into the MR room, it's going to go in there, whether it's unsafe or not, but for the person that doesn't know what it takes, that unsafe label means I have no idea, I'm not going to touch it, I'm not going to bring it in there. For the people that know what they're doing, those are the only people that should dictate that process. They're going to know what to do.

DR. KANAL: Manny Kanal. Channel 4 brought their color TV camera into an MR scan room; they were advised not to. They said I'll keep it far enough away. Now, I don't think there's anybody in this room that's going to look at that camera, and their first reaction, this thing that he puts over his shoulder and he walks around like a weightlifter, there's nobody in this room that's going to look at that and say that's MR Conditional.

The reason I'm saying that is because we have devices that weigh twice that much that are labeled MR Conditional because we, in the FDA, have allowed ourselves to look at the intended use and cloudy up our thinking. It's not safe. You can restrict it so that the patient doesn't get hurt and I need this and don't tell me how to practice medicine, Manny, I'm bringing this pump in. Absolutely, we encourage, that's appropriate patient care, but it doesn't change physics, and it doesn't change facts, and it shouldn't be allowed to change the labeling.

DR. WOODS: Terry Woods.

DR. RAJAN: I think he had a comment before you.

DR. WOODS: Okay. Go ahead.

MR. WEDAN: This is Steve Wedan from Imricor. I think, actually, one of the benefits of having the MR conditional is for intent. So we make a system, we make an interventional system where components stay inside and components stay outside the magnet room. Those components which are intended to go in the magnet room we label as or we tend to label as MR Conditional. Those things that would stay outside and must stay outside, we put MR unsafe on.

That gives the site some guidance as to how this device should be used, and those things that do go inside, some are perfectly safe. Some are conditional, not necessarily because of projectile hazards, but because of functionality and so forth. But they're devices that are intended to be in the room and need to be in the room for interventional system to work. So I think there is some benefit to having conditional, but not just safe or unsafe.

DR. RAJAN: Okay. Terry was next.

DR. WOODS: Yes. Terry Woods. I guess, Manny, I can see your point. The difficulty is that deflection angle depends on the force product, and so unsafe where then? Again, it's going to deflect different amounts in different scan rooms, so it might have a 45-degree deflection in one room, but not in another. That's the difficulty we'd end up with if we chose to do something like that.

DR. KANAL: But we defined the standard as the greatest it could possibly reach, and we've done that for implants because the patient is designed, no pun here, to go into the bore, whereas this isn't designed to go into the bore. The entire concept of four zones is a recognition that, human nature being what it is, you're in the room. All of us immediately recognize, there's no longer a physical constraint. That's a problem. We need to build one in.

Recognizing that you need to build in a physical constraint means that the intent -- with tremendous respect. It's just a difference of opinion. The idea that it's intended to go into the room, okay, no one's disagreeing, but the end user doesn't care about that. They care about, is it safe or not. I intended it to be safely used. It just killed someone. The intent isn't what's being labeled. It's the safety that we're labeling, and the intent can be written in to be conditions, and if it can't be used ever, then even as an unsafe, there should be another category. This will blow up if it's brought into the scanner, but since we don't have that, we've created three categories.

Unsafe simply means we know -- that's exactly what the gentleman said. We know this is going to cause harm. This does have the ability to kill or maim or cause significant bodily injury. How do we allow ourselves to get that clouded? It can't be more concrete, but we've said since I want it in the room, let's make labeling adjust to my intent and desire. But please remember what somebody has so beautifully labeled Joe Six-Pack Radiologist. The average Joe Six-Pack Radiologist is going to do a barium enema after this, and then he's going to read a CT scan, and he's got to cover nuclear medicine this afternoon, and somebody's bringing two mammos for him or her to read while they're sitting there, asking him a question about, can I bring this in the room.

They don't have the luxury of spending most of their career on MR safety. If you call it conditional, he's going to assume it's conditional. If it would have been unsafe, they would have told me. That's the concern I have, that communication is not accurate for the end user that we're intending to communicate with.

MS. ROBB-BELVILLE: Sonja Robb-Belville. I get that, Dr. Kanal, you're calling for a change in the labeling definitions, and yes, a consistency under what the current model is. I want to come back to that infusion pump that was used improperly. It was not labeled whatsoever. It was here is a device. We call this MR compatible, right? It was marketed that way.

Now, certainly, had they read the manual, which we've all established that people don't necessarily read the manual in these last few days. Had they read the manual, they would have known the conditions that it could be used safely in that system room. However, they did not because they were using that bad C word, compatible, believing that wherever they brought it in the room, it would be fine.

What my plea is, is that whatever we decide as the standard, that they are actually on those devices so that they are in our face, so that the nurses that may or may not have been trained have that immediately available. If they are looking at a pump that is labeled as MR Conditional versus MR compatible, that immediately triggers well, what conditions would this be safe under versus it's MR compatible and I can take it in there and be fine.

MS. SAWYER: It's Anne Sawyer. I just wanted to play the devil's advocate here because I don't think I entirely understand what Dr. Kanal is saying, but let me ask this. If they are labeled unsafe, then all these other facilities that don't have a Dr. Kanal, then they just won't take it in the magnet room at all?

DR. KANAL: I'm taking that as a question. Will the facilities that don't have an MR safety officer or a physicist or someone that they can turn to to understand the safety considerations, they won't take it in.

Number one, if somebody will take the advisory, the educational label as regulatory, then that is, in fact, a problem, so we would have to clarify that what these are -- in other words, if they followed the ACR Guidance Document, they would understand this is not meant to tell you what you can and cannot do. It specifically states, you can bring anything you want in the room. All you have to do is label it as to whether it has the potential to be safe or unsafe so that anyone, even the nurse anesthetist or someone that's not typically in this environment, when they look up and they see something, if it has a red sticker on it, it will have a certain meaning to them. If it has a yellow, it may have an entirely different meaning.

So it's not meant to tell you how to practice. You're saying people may misinterpret that and, of course, that's an education issue we would have to address.

UNIDENTIFIED SPEAKER: And there is one more issue. You said for the sites that don't have an MR safety officer. I'm sorry, but when you have a hazardous object, you need to know it, and you need to go find somebody that can inform you of the exact process and how to use that because this should not be taken lightly, and it probably, you're right, should not be used there. So I don't understand the question in the sense that either you do have the knowledge or you don't have the knowledge to use it. So if you do, this issue is reduced to protocol. Otherwise, you need to get somebody and the devices need to reflect that.

MS. SAWYER: What I was thinking in my mind, and again, this is still a training problem because I don't think people fully understand the word Conditional yet, but when they finally do understand Conditional and they see that, I am hoping that they'll stop and they'll go find out, what are the conditions under which this item can go in the magnet room or go in the magnet and be used. That's my hope.

DR. RAJAN: Okay.

UNIDENTIFIED SPEAKER: Can I ask Manny for a clarification. I'm sorry. I'm still not quite following your --

DR. RAJAN: Are you going to ask a question of the same point of discussion?

UNIDENTIFIED SPEAKER: Yes, it is. Okay. Thanks, Sunder. MR Unsafe, MR Conditional, you're concerned about the blur there. Now, consider the pacemakers and your concern that MR vendors have historically contraindicated, and by the way, that just changed. MR vendors looking at these devices, the older ones, and by design, they weren't necessarily designed to be MR Conditional, so we contraindicated, but you're concerned about that. So I'm trying to understand what --

DR. KANAL: That's very straightforward to address. We have identified three areas at least. We've identified the static field. We've identified the time varying magnetic fields that are radiofrequency, and we've identified the gradients. Of those three, at least one, we've identified what we would consider as an extreme cutoff.

I'm not even sure, historically. The article was why Dr. Knew (ph.), if I remember correctly, in 1985 in Radiology where he first described the deflection angle tested 45 degrees, but it caught on like wildfire and everyone accepted 45 degrees as essentially, 1G will be our cutoff. It doesn't mean that you're going to die if you put somebody in there if it went in with a 45 or 46 degrees. It means there's something you need to be aware of.

Conceptually, we've accepted a very, very longstanding 30-year or 25-year acceptance of a nonmoving standard, for some reason, of 45 degrees. So what we've said is that, what this single aspect -- I don't want to hear about RF, I don't want to hear about gradient. The single component referred to as projectile effect because it's not torque, the single component called projectile effect, we're going to need a, is it a problem or is it not a problem. We've used 45 degrees as our cutoff.

So all I'm suggesting is, by saying that something is 45 degrees, it doesn't mean it's going to kill you. It means you might have an injury, whatever the official wording is, serious harm or injury or death or whatever the actual outcome is. For example, if I put a paperclip in there, am I suggesting we need to get another terminology because the mass is small enough that it probably won't? No, no, no. We've accepted a number.

We recognize the seriousness of the number, and from an educational point of view, we're telling them, this is safe, meaning, don't think about it. This is unsafe, meaning don't even think about it. Know for a fact, this can hurt you. Then we've created conditions under which they can safely use certain things for RF, safely be used, and quite frankly, the safely be used really was -- the conditional was really defined for RF and gradients, but that's for another time. No one ever suggested we designed a sequence, a name of MR Conditional because it's only slightly attracted to magnetic field. It's like being slightly pregnant. At some point, you say you are or you aren't a potential life or limb hazard.

The conditional label grew out of a need to address radiofrequency issues. It grew out of a need to address dB/dt issues, but since we threw in the static field, we suddenly said, and don't worry about the static field. We'll always find a condition under which it's safe, such as don't expose it to one. That's why the more quantum state of the static field or the static field gradient has suddenly found itself changed into a conditional spectrum.

It's okay six feet, but it's not okay five feet. That's never been the intent of MR conditional. That is the root cause analysis cause of so many of these accidents, where things are in the room. They think it's fine. They didn't tether sufficiently, they didn't tether at all, someone wasn't aware, someone else pushed it. If we're here to prevent that from happening, root cause analysis is the way to see why they happened and to change it. Calling it conditional is part of the problem.

MS. SAWYER: Thank you. Now I get it, and I didn't know that, and I bet there's a lot of people that didn't know that either about Conditional.

DR. KANAL: Well, the person who suggested the names to the FDA, MR Safe, MR Unsafe, and MR Conditional is addressing you now.

MS. SAWYER: Thank you.

DR. KAINZ: Are you asking for a redefinition of these three classes we have? That would mean then, are you advocating for an MR Unsafe with conditions, or am I misunderstanding? That's how it sounded to me.

DR. KANAL: I think we need to stick with the definitions we already have, which is MR Safe --

DR. RAJAN: Time out. Excuse me. Who would have thought that we would find a debatable point at the end of the day, but I invite all of you to take this up with Dr. Kanal after the session, and let's see if we can charge him down a little bit.

I just wanted to emphasize that all this labeling aside, the FDA does not regulate clinical practice, so the labeling part sort of came independently of the -- even though we had some thought about it, no intention of regulating clinical practice here.

Okay. So I go on to the next question. Somebody was standing there with the next question. No? No more?

Okay. In which case, if somebody has a follow-up comment for Dr. Kanal? No?

UNIDENTIFIED SPEAKER: I think we need an MR missile icon.

DR. RAJAN: Okay. Well, we've got one more question. All right.

MS. SUN: This is Felice Sun (ph.) from Neuropes (ph.). I have a question for the Panel for accessory devices, as well as for implanted devices from the discussion I just heard. There is a definition of MR Unsafe, which means it could possibly be unsafe in the scanner room. My question is, I think Dr. Kainz actually brought this up kind of as a side note, but I think it's a real question.

Can something be labeled MR Unsafe and yet consistently have a list of conditions under which it can be used appropriately in the scanner? Is it a contradiction to have both, or is it actually informative for the user to have both, say it's MR Unsafe but say, by the way, you can use it under these conditions in this scanner?

DR. KANAL: Well, in universities, at least across the country, MR Uunsafe equipment is constantly brought into the room and labeled MR Unsafe, but we have, in a research setting in a transplant center, this is a one of a kind infusion pump. I have to have it on this patient. It happens all the time. The purpose of the label, once again, is to warn everyone around that piece of equipment that it is a potentially unsafe piece of equipment and should not be taken lightly. The unfortunate extension has been beyond that, as you can see, but of course, MR unsafe, we can use and do so routinely as long as we take the responsibility for doing so.

UNIDENTIFIED SPEAKER: But by the definition, no. The definition is that if it's MR Unsafe, it is unsafe to bring it into the room, period.

MS. SUN: I think I heard two different things then. Is it a contradiction to have something that is labeled MR Unsafe and at the same time provide information that presents cases when it can be used in the scanner room safely?

DR. KANAL: It is absolutely not a contraindication under the ACR Guidance Document from the first day it was written in 2002. That has always been provided for.

MS. SUN: Then I would also like to direct the question to FDA. In terms of labeling, is it appropriate to have labeling that says MRI Unsafe and at the same time provide conditions under which it can be used safely?

DR. KAINZ: This is Wolfgang Kainz from the FDA. No, we don't allow this currently. If it is MR Unsafe, and you said at the beginning, it's unsafe, I think, I'm not absolutely sure about a standard technician, but it says it is known to cause harm. So if it's labeled MR Unsafe, then it is unsafe, and we certainly do not accept an MR Unsafe labeling. That's what I just asked Manny if it should be there. If this is asked for, MR Unsafe label with conditions. I don't know how we would do that. Currently, no, it's not the case.

DR. RAJAN: Currently, you don't get that testing data on your labeling. If it's unsafe, it just says it's unsafe, so it's a little bit of a problem.

DR. KAINZ: But as a solution to what you asked Manny, it could be that an additional warning. Is there, this is MR Conditional, but has the potential to kill or has the potential to harm a patient. That would be -- I don't know. I'm just throwing this out, but certainly, I don't think it's a good idea to have MR Unsafe with conditions. Well, I am not sure if you're asking that.

DR. KANAL: There is a tremendous dichotomy that you're looking at. The dichotomy is clinical practice on one side and how the FDA approaches it on the other. The clinical practice is that this isn't safe. There's nothing to discuss. We know that, but I'm going to use it anyway because I have to. As a risk/benefit analysis, we do that all the time. Every drug, just about, I ever give has the potential to be unsafe. These are risk/benefit assessments. That's what we deal with on every single decision we make for our patients.

As an unsafe, warned item, I now know that I have to treat that with a certain respect, and I can't change the fact that it is, in this environment, unsafe. This is semantics. This is the same mistake the FDA made with compatible versus safety. It meant something to you because you approached the words and defined them, but the end users didn't use them the same way, and we're repeating the same exact mistake.

Now, this time, I know it's a repeat because I did create those words and put them in an e-mail to you guys and suggested we use them and it eventually went through ASTM. In that suggestion, it specifically said, since there is confusion in the industry as to what compatible and safety means, we have to clarify it to what a natural language speaking, American or English speaking patient or healthcare worker would assume that the words mean.

So an unsafe product, which I think you'll find in your own definition, is one that is shown to be harmful. No one's going to look at that TV camera and say it's conditional. We know it's harmful in this environment. So we can now recreate the conditions, and we do all the time, and people with TV cameras go in there at certain places and certain distances. It's not what it's designed for. It is what it's designed for. That's not the issue. The same thing can apply.

Here, the FDA says it's the other way around and it's appropriate. I understand where you're coming from. We define the terms, but from then on, the term rules, not the end user. So now, oh no, the system has MR Unsafe and so, by definition, we won't let you have conditions of safe usage. That's your decision. It's got nothing to do with the educational mission and objective of the label itself.

DR. KAINZ: What solution do you suggest? What are you proposing?

DR. KANAL: The solution is to go back to exactly what we were -- not to go back, as to maintain exactly what your own definitions today state. If it's shown to be harmful, not a question, but it is, it's labeled as unsafe. You, internally, have not permitted that request of saying, by the way, if this thing goes beyond the so-and-so gauss line, it's going to pull in automatically. It'll deflect 90 degree. You haven't permitted that to happen. I don't know where that came. I'm not sure what the origin of that incompatibility is internal.

DR. RAJAN: Okay. I think we have time for one last question.

MR. OCH: Well, this is just a comment on that. This is Joe Och from the AAPM. It seems to me, from what I'm hearing here, that the confusion is between unsafe and forbidden. What I think I'm hearing is that things, as a projectile, not as an implant, not something that's going, but as a projectile, things either are going to have a possibility of flying into the magnet or not. If they don't, don't worry about them. If they do, they're unsafe and you ought to know about that.

It's not a question of saying they can't go in level form, zone form. They can go in, as long as you know about this. I mean, from radiation safety issues, medical physicists, the radiation safety officer goes through this all the time. It's this is forbidden, except that you can do this and this and this. It's a question of alerting somebody that an object has to be treated specially. How do I treat it specially? Let me tell you, here's how you treat it specially. That's what I'm getting.

UNIDENTIFIED SPEAKER: Well, I'm just looking at the definition to some extent. MR Unsafe is defined by the ASTM international. Is it any item that is known threat or possesses a hazard in all MR environments?

An example of an MR Unsafe item would be a pair of ferromagnetic scissors or any item constructed of ferrous materials. The MR Unsafe icon consists alone.

DR. KANAL: To respond to your question again, just to reiterate, what am I asking for? Stick with the definitions we agreed on. Don't superimpose your own; FDA's arbitrary. If it's unsafe, let's make that regulatorily unsafe. No one said that. The device is unsafe. Don't tell me how to practice medicine. The device isn't safe. That's never changed.

DR. RAJAN: Okay.

DR. KAINZ: Not to give any conditions to it, then you are left alone under what conditions and how far you can get it to the scanner.

DR. KANAL: I see no reason why you need to cut us off. If that information is there, why does the FDA refuse to -- the very response that you gave to that question, if it's MR Unsafe, to date, the FDA has chosen to say, we will not accept conditions of use. That's an internal decision. No one else out there is forcing you or mandating that you say that. You need to readdress, why did you choose to do that?

DR. RAJAN: Okay. Looks like we've got some good arguments from one side, so with that, I would like to close this very provocative session. Maybe we need to call a panel to sort this out, but thank you very much for your participation and asking questions. Thank you.

(Off microphone comments.)

DR. RAJAN: Folks, we've just got a couple of minutes. Simon wants to close the session here.

DR. CHOI: Diana, could you forward to the last slide, please? This concludes the public meeting. Thanks to all the presenters, roundtable participants, and attendees.

On this slide, you'll see the webpage, regulations.gov, the actual FDA docket deadline to submit comments is November 22nd, 2011. Thank you again. The transcripts, the webcasts, and the slides will be available approximately six to eight weeks from this meeting. Thank you so much.

(Whereupon, at 5:00 p.m., the meeting was adjourned.)

C E R T I F I C A T E

This is to certify that the attached proceedings in the matter of:

MRI SAFETY PUBLIC WORKSHOP

October 26, 2011

Silver Spring, Maryland

were held as herein appears, and that this is the original transcription thereof for the files of the Food and Drug Administration, Center for Devices and Radiological Health, Medical Devices Advisory Committee.

____________________________

SUSSY MOREHOUSE

Official Reporter