• 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

Thermal Aspects of Radio Frequency Exposure: Transcript for January 11

U.S. FOOD AND DRUG ADMINISTRATION

MOBILE MANUFACTURERS FORUM

GSM ASSOCIATION

WORKSHOP ON THERMAL ASPECTS OF RADIOFREQUENCY EXPOSURE

January 11, 2010

Gaithersburg Hilton
620 Perry Parkway
Gaithersburg, Maryland

“This transcript has not been edited or corrected, but appears as received from the commercial transcribing services. Accordingly the Food and Drug Administration makes no representation as to its accuracy”

TABLE OF CONTENTS

Session 1: Scope of the Workshop: Joe Morrissey

The Needs of the FDA: Abiy Desta

The Needs of the FCC: Ed Mantiply

IEEE C95.1 Brief Overview: C.K. Chou

ICNIRP Brief Overview: Bernard Veyret

Report Presentation: Review of Thermal Damage to Various Tissues

Mark Dewhirst

Marv Ziskin

Session 2: Whole Body Heating

Introductory Presentation: Chris Gordon

Session 3: Immune System

Thermal Influences on the Immune System: Elizabeth Repasky

Session 4: RF Energy Deposition and Temperature

Robert McIntosh

P R O C E E D I N G S (8:30 a.m.)

Agenda Item: Session 1: Scope of the Workshop

DR. MORRISSEY: We are going to go ahead and get started here. Thank you all for coming to the workshop today. Hopefully this will be a very valuable workshop. We will be able to make some progress toward the issue of radiofrequency exposure.

We certainly have a great turnout today. I think we have some more chairs coming. Some very quick announcements.

(Logistical announcements)

We have many of the invited experts, and we will make introductions here in a minute. We do have great participation from across different stakeholder groups, and I wanted to point that out. We have several from academia as well as the government. We have a lot of government agencies not only in the U.S., but internationally, represented here at this meeting, speaking to the importance of this issue, and we will get to that. We have the German Radiation Protection Program. We have Health Protection Agency from the U.K., Health Canada, the Health Council of The Netherlands, the Australian Radiation Protection Program, Japanese EMS Information Center. We also have a lot of industry participation.

As many of you know, this workshop was born out of the issue of wireless communication with the huge growth of wireless communication. I guess it is now four million mobile phone prescriptions now worldwide, which is more than half of the global population. So we have a huge interest, and I'll go over this just really briefly.

But we also have some interest from the medical device manufacturers. They responded to the Federal Register announcement. I thought this was very interesting. I guess what we can do right now is, Joe, if I can start with you, just for the invited experts around the room, can I have you say your name, your affiliation, and maybe just a very brief sentence with regards to your expertise and how it fits into the thermal workshop.

DR. ELDER: Good morning. My name is Joe Elder. I am essentially retired. I worked for many years with the U.S. Environmental Protection Agency, and then worked for the past about eight years with Motorola.

My background is on the biological effects of radiofrequency radiation, and I have also helped with the IEEE ISO standard setting activities.

DR. CHOU: C. K. Chou. I have been working in this field since 1971, on biological effects of many applications. I worked in academia for 27 years. The last almost 12 years I worked for Motorola. I am currently the Chairman of IEEE ISIS TC-95.

DR. DESTA: Good morning. My name is Abiy Desta. I am with the U.S. Food and Drug Administration. My primary responsibility there is to coordinate scientific issues that affect multiple devices, medical devices and radiological health.

DR. MANTIPLY: Hi. My name is Ed Mantiply. I am with the Federal Communications Commission. I started work with the Environmental Protection Agency in 1976, and I came to FCC in 2000. Most of my work has been exposure assessment and enforcement of the exposure levels.

DR. VEYRET: Good morning. My name is Bernard Veyret. I am from the University of Bordeaux. I have a team working on bioeffects for the last 25 years, and I belong to the ICNIRP.

DR. GORDON: Hi. I am Chris Gordon from the U.S. EPA. I used to do a lot of work with microwaves on temperature regulation. I have been out of it, and now I am back into it again.

DR. DONALDSON: Hello. I am Gavin Donaldson. I am from the University College, London. I started off as a physiologist doing hot and cold human exposures, and then moved into the field of epidemiology to do with heat waves and that sort of work.

DR. KYATKIN: Good morning. My name is Eugene Kyatkin. I am working in the National Institute of Drug Abuse. I am physiologist, and I am mainly working with recording of temperature, specifically brain temperature in experimental animals during different situations, emotional situations, behavior and addictive drugs.

DR. DANDRAYA: Good morning. My name is John D'Andrea. I am with the Naval Research Unit in San Antonio. We are a medical research organization, traditionally looking at questions of RF safety. We also include combat casualty care and dental research now. I am not an invited expert. I just simply took an empty seat.

DR. ZERIACS: I'm enjoying the empty seat as well. I am John Zeriacs, also with the Naval Medical Research Unit in San Antonio. I am head of the Directed Energy Bioeffects Group.

PARTICIPANT: I'm not an invited expert, either, but I work for Nokia. I am responsible for safety issues of Nokia.

DR. MILLER: Good morning. My name is Morton Miller. I am a professor in the Department of Obstetrics and Gynecology at the University of Rochester. My area of interest is hyperthermia induced birth defects. I'll have a short 15 minutes of fame at one o'clock today, and a one-hour discussion, too.

DR. BERGERON: I am Mike Bergeron. I am with the National Institute for Athletic Health and Performance. My background is probably very different than most everyone else here, but I am probably one of the few here that have looked at thermal stress in children and adolescents.

DR. SHARMA: Hi. My name is Hari Sharma. I am coming from Uppsala University. I have as my interest to understand brain function changes in hyperthermia. I am doing this work since 1977. One of our early researchers suggested that this is leakage of blood-brain barrier, and that hyperthermia can produce brain dysfunction. So we have interests together with also Dr. Kyathkin, because I have a physiology background. We want to now study how heat induced hyperthermia can be altered by drugs and diseases, for example hypertension, diabetes and even exposure to nanoparticles. Thank you.

DR. DEWEY: I am Bill Dewey. I am an emeritus professor from UCSF, retired, in other words. I have been involved in hyperthermia research at the cellular molecular level since about 1970. I will review a lot of this work. Basically I am a biophysicist, trying to understand mechanisms at the cellular level.

But I would just like to make one comment about an expert. As far as I know, an expert is a drip under pressure.

DR. MC DANNOLD: Hi. My name is Nathan McDannold. I am a physicist. I am at Brigham and Women's Hospital, and we work on therapeutic uses of ultrasound, including thermal therapy and MRI based temperature mapping of thermal changes.

DR. HOOPES: I am Jack Hoopes. I am a professor at Dartmouth College, trained as a radiation biologist and veterinary pathologist. I have been working on various aspects of hyperthermia, primarily to treat cancer, for most of my career. Currently we are using iron oxide nanoparticles to treat cancer. A lot of the same controversy we are seeing with RF as to whether it is truly heat or some other effect.

DR. WANG: My name is Christina Wang. I come from the University of California in Los Angeles. I am a professor of medicine. I'm not an expert really, but we did a lot of work on heat to see how it affects thermogenesis, in the hope that we can find the pathway to find the target for developing male contraception. So that is the area of my work.

DR. FOSTER: My name is Ken Foster. I am from the government and I am here to help you. I'm sorry, that's a joke. I am a professor of bioengineering at the University of Pennsylvania. I have been involved in RF safety issues for a number of years, and most recently have done some thermal modeling for heating in tissue.

DR. SODERBERG: My name is Per Soderberg. I am a professor of thermology at Uppsala University in Sweden, where I am heading a research group that is working on optic radiation effects to the eye. I am also heading the optic radiation subcommittee in ICNIRP.

DR. BEACHY: Good morning. My name is Sara Beachy, and I am a postdoctoral fellow in the Department of Genetics at the National Cancer Institute. I am interested in the vascular effects of hyperthermia.

DR. REPASKY: I am Betsy Repasky from Roswell Park in Buffalo, New York. I am immunologist, and I am very interested in the beneficial effects of thermal stress on immune cells and signalling pathways that may enhance natural immune response to tumors, hopefully taking advantage of some of the natural thermal sensitivity of the immune system as it has evolved to respond to fever.

I come into this field mostly from a group of the Society for Thermal Medicine with Mark Dewhirst here to my right. We are very interested in various applications of thermal energy in a positive application for cancer.

DR. DEWHIRST: Non-cancer, too. Mark Dewhirst, professor in radiation oncology at Duke University. There was a workshop run by the World Health Organization about ten years ago that I chaired. It was the first attempt to try to put some kind of quantitative assessment of the effect of thermal energy on tissue damage and also on neurocognitive behavior and other kinds of end points. It was a very productive workshop, and this is a follow-up to that. I think we are going to have some very interesting discussions, and I hope for further refinement that would help with standard setting and so forth by the FDA and the regulatory agencies.

DR. ZISKIN: Good morning. I'm Marv Ziskin. I am a professor of radiology and medical physics at Temple University Medical School in Philadelphia. I have been involved in ultrasound safety for my entire career. For the past 19 years I have also gotten involved in the safety of electromagnetic fields. I am the current co-chair of the IEEE committee SC-4, which is involved with setting standards in RF radiation. I am here today to learn a lot and also to talk about some of my experiences.

DR. MORRISSEY: Thanks. We have some wonderful experts here. Like Mark said, the purpose of this workshop again is to try to gain as much insight as we can on the thermal thresholds of different tissues in support of these radiofrequency exposure standards, which are becoming more and more -- catching the public eye, becoming more and more important.

The format of the workshop. It will largely be discussion based. This first session in the morning is going to help lay out the framework or point us in the right direction, that is, trying to have discussion to better clarify what we know and where we need to go to better determine the thermal thresholds or the temperature thresholds of different tissues as it relates to exposure from radiofrequency devices, whether these be wireless communication devices, medical imaging devices, et cetera. I think the scope of the workshop is fairly broad in that way.

There will be two moderators, myself and Ken Foster, the one from the government over there that is here to help you. We will try to help steer and direct the discussions, but the meat of the workshop is certainly going to be dependent on good discussion by the participants. So I want to urge everyone to be very actively involved in the discussion and help us to address the goal questions.

We are going to have two sessions that are going to be largely discussion based. We are also going to hear a very interesting talk, I'm sure, by Dr. Repasky about the immune system, and we are going to hear some discussion about SAR versus temperature later on today. But again, the bulk of the session will be discussion based.

Because we are trying to put so many sessions in a short time, it is necessary for us to have some breakout sessions. I had e-mailed an agenda to the group several weeks ago. We will have a breakout session today at 11 o'clock. After this breakout session, everybody will return here to lunch. It is a bag lunch, but we can continue the discussion. Make sure you fill out a form for the lunch sandwiches, because if you don't fill out a form you are not going to have a sandwich.

We will have another discussion session on pregnant women and the fetus, as well as on children. The other breakout room, the Rockville Potomac, is on the other end of the hotel. It is a bit of a walk, but I'm sure we can all manage it. I will lead the first breakout session to the Rockville Potomac.

I have not broken out the experts in terms of which breakout session you should go to, because I don't know where your interests and your ability to contribute lie. But I will say that I would like the experts to try to divide up fairly evenly. If you see -- we have 23, 24 experts; if you see 20 experts here and only three in the other room, I'm going to ask the experts on your own to help us divide this, so we can have good discussions in both sessions.

We did make one small adjustment. The session on the testes is going to be this morning in the Rockville Potomac Room. We are going to have the session on behavior tomorrow. This is to accommodate some flight schedules.

The goal questions that we are going to try to stick to. Again, the purpose of this workshop is to help define temperature thresholds in support of radiofrequency exposure and exposure standards, what are the most appropriate health end points. For tissues like the brain, this could be incredibly complex. The brain as an organ could be divided up into several sub-organs. Is this a behavioral end point or a pathologic end point, a physiologic end point? The discussion will help to bear that out.

What are the most appropriate time periods for acute and chronic exposure? This is very important with regards to standards and the government's needs for compliance. So while these may seem like very difficult questions to get your arms around, they are important parameters for us to try to get as close as we can to settling on what appropriate acute and chronic exposure time periods should be.

We will talk about established time-temperature thresholds, and also what we hope to come out of this is, if you are paying for research out of your own pocket, out of your own 401K, what cost effective research could one come up with that could help us get closer to better definition of time-temperature thresholds. We can support and contrast this with the wonderful report that Mark Dewhirst and his colleagues helped to write, that I also circulated prior to the workshop.

I'm not going to get into the physical interaction of electromagnetic energy with dielectric median tissue, other than, the workshop will be focused on thermal effects. Clearly, if anybody has read the media, the newspapers, there are questions about potential non-thermal mechanisms that are out of the scope of this workshop. We are going to be looking at thermal thresholds in support of the standards.

I forgot. In Session 1 we are going to hear from the FDA and the FTC with regards of their needs to better clarify what they would like to see coming out of this workshop. We are also going to hear from the two relevant RF exposure standards committees. Dr. C. K. Chou and Dr. Bernard Veyret will tell you very briefly about the current standards, which are recognized by different government agencies for defining limits for exposure to radiofrequency energy, perhaps how they were developed, how the whole body and local limits are developed, and where we would like to go.

The direct application for human health, what are we doing here, what is the goal of this workshop and where are we actually applying this to real human health questions. Clearly there is a huge proliferation of mobile phones, and everybody here is well aware of that, four million subscriptions, I believe, currently, I think it is projected six million by 2013. Each of these is a radiofrequency transmitting device. So exposure levels are of concern in the general public, and these standards would like to evolve to become as solid as they can, robust as they can, as science based as they can, to define the correct limits for human exposure to radiofrequency energy.

There has been some interest, and I'm not going to go into these in detail, but some interest in the U.S. Congress. We have had some Congressional hearings on the health effects. But I would also like to mention that the scope of the workshop can be much broader. We are all very aware of the body imaging, which can be in the terahertz range. That is another non-ionizing source of electromagnetic energy. There are wireless medical devices. Medtronics has wireless links for control of pacemakers, radiofrequency ablation. I know there are some medical device manufacturers here. And medical imaging, the RF component of MRI does jump RF energy, which is absorbed by heat. Ultrasound, the scanning ultrasound, getting into the higher megahertz range, also has energy that can be deposited as heat. So all of these are within the scope of this workshop, trying to identify these RF exposure limits

Don't forget to fill out your lunch form. We will be meeting in the lobby at 6:30 tonight. If you would like to go to dinner, make sure that you sign up.

Next on our list is Ably Desta.

DR. DESTA: Ken, I do work for the government and I am here to help you.

I think Joe has done a wonderful job describing the scope of the workshop, so my talk is really going to be short. I would like to thank MMF and GSMA for co-sponsoring the workshop with the FDA, Joe Morrissey for all the hard work he has done, and all the invited experts here.

Under the radiation control provision of the Food Drug and Cosmetic Act, FDA is responsible for assuring the safety of radiation emitting products, whether they be cell phones, or other electronic products that emit radiation, we are responsible for assuring the safety. One of the ways we assure the safety of the products we regulate is to make sure they comply with safety standards.

RF safety standards have been around for a very long time. The whole body exposure limit that has been used by both ICNIRP and IEEE has been based on biological end points that have been scientifically accepted as being valid. The partial body exposure that is being used by IEEE at least and probably ICNIRP. is one that, in terms of identifying exactly what hazards it is supposed to protect against, has been a little bit harder to follow. The rationale seems to be based on a number of things that are not necessarily biologically based.

From the FDA perspective, we would like to see safety limits that are clearly tied to biological effects that are not necessarily destructive. I go back to the whole body exposure, which is work stoppage. If that is the case, then I think the partial body exposure limits should not be something as severe as cell necrosis, but something that could be identified as being significant and yet not catastrophic. We are hoping that this workshop will come up with real examples and real numbers that could be used to guide the standard setting process.

I look forward to a very productive workshop, and I look forward to moving forward from this point so that in the very near future we actually have a standard that is thermally based that we can all point to. Thank you very much.

DR. MORRISSEY: (Off Mic.)

DR. MANTIPLY: I am somewhat humbled to be among the experts in thermal physiology in this academic community. I have worked for a long time in the area, and would like to give you briefly as much as anything personal feelings about this from an FCC perspective.

FCC is about exposure limits. I spend relatively little time thinking about the exposure limits and lots of time doing two things. One is interpreting and implementing the exposure limits, enforcing them, determining how the compliance is going to occur in real world situations, making measurements in the field, determining what type of systems and standards we are going to use when we determine compliance of devices. The other half of my time, I try to answer Congressional inquiries and calls and concerns about this field in general.

The exposure limits are really key to all of this. I wanted to talk also about some of the process, so you get a sense of where this ends up, and also about potential occupational over exposure. It is possible to look at the exposure limits from the bottom up. If they are above me and I think they should be lower, I want to be more conservative.

But there is another community out there. There is short wave broadcasters, there is RF broadcasters and such, and people who work in very high fields, who are not really convinced that the standards are not too conservative, that they are being over protected, and they would like to be able to take responsibility for their own exposure more, instead of having to comply with the limits that we have.

Limits are legal requirements for a large industry, and it has a big impact. It also affects small businesses. There are people who have to comply and relatively small broadcasters and such that are nonprofit, that have to deal with these issues.

They should be responsive to public health concerns. They apply to potentially continuous exposure of children, and people are very concerned about that. They should be specific and fully defined. They should be well founded on science, and they should be easily explained. Obviously these things tend to be in conflict with each other, so we struggle with this always.

They should have a minimum impact on business, especially for small businesses. We would like to see manuals used and this kind of thing so that small business is treated more like a consumer as much as possible.

I was just thinking about this overall thing and reading about the end points and the thermal studies. I wanted to go down the list here of how this all seems to work from my perspective. You have some adverse effect. This is a very general way of thinking about this. But you need a mechanism, you need some physical interaction. In this case, we are talking about thermal, but we need to know this in order to do anything quantitative in order to extrapolate from an effect to a limit that applies to lots of different kinds of sources, different modulations, wide frequency range. You have to understand how that is occurring in a way that allows you to present a theory behind it. So thermal is a very complex theory for this and works quite well in this process of creating standards.

There is a safety factor, statistical uncertainty about how well you know where the effect thresholds are and how safe you want to be, and if you want to try to take in account other mechanisms that are unknown. Sometimes the safety factor gets into that.

There are basic restrictions in the language of ICNIRP standards, if it is an internal field or a specific distortion rate, more specifically for thermal effects. For low frequencies these are commonly electrostimulation and low frequencies or induced current density or internal electric field strengths at low frequencies. So the reference level, what you can actually measure in the field. When you go out with a meter and you make a measurement, you are not putting it in a person's body, you are not normally clamping probes on people, so most of the time you need to measure what is there, the unperturbed field in an environment.

Evaluation is the general term I use for either measurement or computation. In other words, how do we determine whether we are compliant with the reference level or a basic restriction. Some of the computations are simple calculations we do, and we also do more sophisticated computational models. We have a lot of measurement equipment that has evolved a lot over the years, and I have been mostly involved in that aspect.

Finally, we have exemption. In other words, if we can avoid the impact of these standards as much as we possibly can in terms of difficult compliance measurements and computations, we would like to do that. If you can say you have a certain source power and certain separation distance, you can be very confident that you are compliant with the exposure limits, we should do that.

This is basically my last slide. There are cases out there where people do get overexposed. It is not very uncommon. Then they will claim injury at some later date because of their exposure. These are some questions that I think this forum might be able to contribute to more than most meetings. It is really about the limits versus injury. Typically I am talking about a tower climber, a broadcasting antenna. The antenna can have ten or 20 kilowatts going into it. He can be seriously exposed over the exposure limit. What do I tell him? What kind of expectations do we have for him? In other words, for a whole body exposure, there are some variations in terms of the exposure because if the metal tower is present it may be very close to the antenna. Is there going to be localized internal thermal damage or just whole body over heating? In other words, is the blood flow really going to dominate so that the whole body is uniformly overheated, so that the result we would expect would be heatstroke, or will there be localized heating of internal organs that might cause some other kind of damages that would be important and possibly show up later.

In other words, this last part is, would you expect somebody to go away from an overexposure like this and not think anything of it, maybe has a headache, maybe felt hot, but not really be injured, not have to go to the hospital, and then later develop some illness that he would want to claim workers compensation for and that kind of thing. So some input in that n this meeting I think would be helpful.

That is basically all I have. I would like to offer any -- answer any questions or have any discussion for a short period. I've only got ten minutes, and I have probably already used up all my time. But if there is anything burning, we can talk about it during breaks, and we are going to have lots of discussion.

DR. MORRISSEY: Let us hold the questions until the break. The first session we are just directing the discussions, and we will have a lot of discussion.

DR. MANTIPLY: I just wanted to offer.

DR. MORRISSEY: That's fine. Thanks, Ed. Next up we are going to learn about the existing standards for radiofrequency exposure and where we are with these standards. The first presentation will be by Dr. C. K. Chou of the IEEE ISIS Society.

DR. CHOU: Good morning. If we do anything, we must have a purpose why we are doing this. Joe described a little bit earlier. I will try to point out how this connects to the IEEE standards.

I will just give you a brief history. IEEE has been dealing with safety of RF exposure since the first standard came out in 1966. At that time it was based on a simple thermal model. The chairman of the committee was Herman Schwann. It was based on very simple heating. If you go up 100 milliwatt it causes thermal effect. So that is simple. You divide by 10, 10 milliwatt per square centimeter across frequency range is a stable frequency range. That is the flat frequency response.

Bill Guy took over the standards in 1974, made a little change here, but the major change was in 1982, incorporating dosimetry. That was based on the most sensitive biological effect. That was the time on the behavioral disruption that was mentioned earlier. The standard in '82 and '91, in both standards in IEEE we don't want to talk about thermal effect. We just want to protect against any known effects, and don't want to say this is due to thermal.

When we did the revision in 2006, in this particular one we agreed with the ICNIRP. They said in 1998 to protect against known health effects, for ICNIRP they looked at all thermal and non-thermal effects, so in our IEEE standard we also said the same thing. For low frequencies protect against known adverse effects, electrostimulation. So for RF it is protecting against heating effects. So this is the background.

In 1999, we started getting serious about revision of the 1991 version. In 1999 at that time we still tried to relate the relationship of the other adverse effect to any dosimetry or quantity measured in the lab, and what are the different parameters. At that time it is current density, total current, contact area, how to relate the behavior response to SAR, how to relate physiological function to SAR. Now we want to see the temperature increase to SAR. It is one of the items among the list of things that was discussed at that time.

I have many of these records all from the minutes. IEEE is an open consensus, open meeting to anyone interested in this committee, so all these minutes are all posted on the IEEE website. Anybody can look at old minutes, can find all this. I describe some of the highlights from many of the discussions.

This is the source of this discussion. This is by Ric Tell. He was the chairman of the risk assessment working group among our SC-4. He came up with many questions, what are exposure assessment, dosimetry questions forming a basis for a technical rationale for this revision.

In question 23 he says, temperature based standard for both average and local tissue would be more directly related to potential injury from RF field. That is the known adverse effect we know today. Does the present scientific basis provide scientific support for such a standard. This is the basis for the two workshops, the one chaired by Mark in the early time, and this is the follow-up one.

So among all these people, this shows some examples. This is Vitas Anderson. He is saying the ultimate intent is to protect against excessive temperature rise sustained over sufficient time. Ken Foster here, also talked about exposure designed to protect against excessive local temperature rise. May Swicord, used to work for FDA, also Motorola, based on local and partial body, you can theoretically determine field limits in terms of power density or field strengths, below which it is impossible to elevate the temperature to a critical level, how do we define all those correlations.

Dimbylow from the U.K., NRPB at the time, heating is the basis for restriction on exposure, the primary limiting fact, the quantity should be temperature rise. SAR is a secondary indicator of temperature rise. Also at that time he correctly predicted the average should be over an average mass of ten grams. That has been proven later by Hirata, here in the audience, and also Rob McIntosh from Australia, that indeed a ten gram average correlated with temperatures much better than the one gram average.

In San Antonio, May Swicord made a motion to develop a thermal basis for the localized exposure standard, and we tried to formulate some kind of action. But the trouble is, the time was getting too long. Vitas Anderson from Australia asked to change the temperature base standard by that time. We said, no, we cannot do that, we have to finish the existing revision, and that means anything to change the temperature standard must be to the next revision.

So that is what we are here for, is to come up with something, more scientifically correct information for the next revision.

In 2006 we published this new standard. I searched for the name for SAR. I found 332 mentions in there, and I searched for temperature, there is almost 600 mentioned in the current version of the standard. Ron Petersen was the ICES chairman at that time, in 2006. That was before the publication of the standard. I had an interview, and Ron Petersen said there was considerable discussion in the SC-4 committee about the rationale for the next revision, whether it should be SAR or temperature based. Most people felt that the temperature based standard at the higher frequencies would make more sense.

So that was published in 2006. In the 2007 IEEE meeting, there was a motion for the subcommittee to pursue the investigation of relationship between localized tissue temperature increase and peak SAR for the next revision. This question has been unanimously approved by the committee.

So in terms of that, we thank all the physiology experts who come to this workshop. We welcome you to continue to help our committee for the continuous effort. I invite you to join us as an IEEE ICES member.

Thank you very much.

DR. MORRISSEY: The second to last speaker is Bernard Veyret.

DR. VEYRET: Thank you, Joe. This is the logo of ICNIRP. This is a picture of the Commission as it is today. This is from our last meeting in Salzburg, Austria two months ago. Everyone is here, including our dear Chairman, Paolo Vicchia, our scientific secretary is there. So these people are all smiling. They are not a black hole, they are real people, and you can talk to them, including to us during the meeting.

This is the ICNIRP structure. We are a 14-member Commission. At the moment we have one vacancy, so we have only 13 members. We have four standing committees, epidemiology, biology, physics and optics, to make it short. Each of these standing committees has one or two members of the Commission plus about five extra members, and then 30 consulting members who are not part of the standing committees, so altogether it is about 70 people.

What is most important right now is to talk about the timetable for the revision of the guidelines, because the outcome of this meeting will help us do that, that is, to stay in tune with the science, and with the needs for the guidelines.

As you know, the latest addition of the guidelines for ICNIRP was in 1998. Then there was a gap. There was of course a constant update of the science. We just published the book this year, because it is in press now, so it should be 2010. But you can download it from the website of ICNIRP. We had a statement on RF which I will comment on briefly later.

Then what is in italics, including at the bottom, is the future. You have the WHO research recommendations, which should come out in within the next two months, which will help run the strategy for the science, but also it will be used for the public guidelines setting. Then we have the IARC cancer assessment, which should come soon now, within a year or so, at least the meeting to design the classification. Then the EHC WHO on RF which should come about a year later. It is already -- the people are working on it already. Then the ICNIRP guidelines will come later. We need all of these to be able to publish a new set of guidelines.

Now, this past year we had a statement on RF to reaffirm basically the guidelines, the '98 guidelines. This is the title of the statement, which you can also download from the website. I took a few sentences from the statement, including the fact that there is no evidence of any adverse effect below basic restrictions, so there is no need for an urgent revision. In fact, you might argue that this is true if you don't find any evidence of adverse effects below the critical effects rather than a guideline, but that is something we can discuss later.

Then we acknowledged that the relationship between the basic restriction and the reference levels in part of the spectrum, which is around two gigahertz to make it simple, is wrong. It is wrong because it was done in a crude way in the past, and now you can get much better calculation that tell you that there is something wrong. So it is up to 40 percent off.

Then the epidemiology is obviously awaiting the final outcome of Interphone, so we can only say that this is the conclusion at the moment. So we are missing the world picture for the epidemiology, as you all know.

We were asked to comment briefly on the basis of whole body and local exposure limit. And of course I don't have the time to go into details, but it seems to me that there is no mention of new findings that would lead to a modification of the whole body SAR basic restriction level, unless we reconsider everything. We had already some comments about this this morning, but there will be more comments. So this is the basis for the guidelines for both IEEE and ICNIRP for whole body exposure, but everything can be reconsidered.

For local exposure, this is not true. As you all know, the rationale for the local exposure limits in the '98 guidelines was a bit fuzzy, let's say. That is true also for IEEE in a way, but not in the latest version, but in the previous one. So that means that we need to do some work on this, and we are here for that purpose.

The future directions for ICNIRP is obviously an update of the scientific database, which is always ongoing. A good proof of this is the publication of the Blue Book.

The SC biology to which I belong is starting the process of writing the draft of the new guidelines, simply in terms of the overview panorama of the science. No conclusion, but at least to start to write this.

As I said, the outcome of this workshop will help ICNIRP rethink about the use of SAR, and the rationale for setting the local exposure limits. When I say the use of SAR, I mean versus temperature. We have been discussing this, but we have no conclusion at all at the moment. For me it is unlikely, that is my personal view, that we move to temperature in the next revision, but we are open to any good idea and good move for the health and science.

Thank you very much.

DR. MORRISSEY: Thank you, Bernard, very good. So hopefully the group now has an idea of the purpose of the workshop and what it feeds into, the different standards that are developed to control RF exposure in humans.

Now we are going to hear from Mark Dewhirst. He is going to help us even more to frame the scope of the workshop. He has developed a report, which I passed along to many of you, which does go over quite a bit of this information.

Agenda Item: Report Presentation: Review of Thermal Damage to Various Tissues

DR. DEWHIRST: Thank you, Joe. Before I start into this, I want to say that we have been in discussions about coalescing some of the parts of what will be presented in this workshop into a special issue of the International Journal of Hyperthermia. We did this before, after the World Health Organization workshop. It has been an extremely useful tool for people to go back and look at, so I am looking forward to putting together an issue, and Joe will be the guest editor for that.

We published the first paper in 2003 from the World Health Organization conference, in which Jack Hoopes played a major role in the first paper. We have updated this. We were asked to do this by looking at the literature since 2002 up until now, and see what we can find in the literature and see if it changed anything from our prior report or if there was new information.

I want to just point out that Ben Viglianti was extremely important in helping me with this. William Wetsel is a neurocognitive expert at Duke who wrote a separate report on neurocognitive effects. Most of you or many of you saw it; we have it distributed. I will touch a little bit on his report at the end.

I also want to just acknowledge these four students of mine, because if they hadn't helped me with this, I wouldn't be standing here today. This ended up being an enormous task, much bigger than I could ever have thought it would be. Also, Daryl Hockman, who is a neurobiologist at Duke, who helped us with some of the CNS effect interpretation.

I am going to give you a brief summary of some highlights from the prior paper, because I think it is important to set the stage. Then I'll talk about the new review and some highlights particularly of brain and testes, and I'll touch a little bit on some other stuff. Throughout this, I am going to talk about challenges and some recommendations for future research.

We know basic thermal biology tell us that hyperthermia kills cells. This is just data from Roizin-Towle. We are looking at the rate of cell killing as a function of temperature here. You can determine what the slope of each of these curves is as a function of temperature, and what you get is what is called an Arrhenius plot. So this is looking at one over the slope of those cell killing curves versus temperature in Celsius or in Kelvin down here. If you want to derive an Arrhenius relationship, you need to use Kelvin.

But the point is that these are well defined relationships. And because there is a well defined relationship, you can relate the rate of cell killing at any temperature to a different temperature through this Arrhenius relationship.

Bill Dewey had proposed this formulation which he will talk about more later, and I am not going to go into detail. But it stands for cumulative equivalent minutes in 43. You can do that by converting any time-temperature history by this equation here to an equivalent minutes in 43. It gives us a standard by which we can compare everything. So it gives us a standard by which we can compare everything. So everything I am going to talk about in terms of tissue tolerances is going to be in this unit here, cumulative equivalent minutes in 43.

The advantage is that you can convert any time-temperature combination to this standard, and because of that, if you know what an iso effect is, whatever the damage is, you can determine that threshold for that. You can also determine isoeffect lines which will allow you to determine time-temperature combinations that will keep you below what the threshold will be. So that is an advantage. I'll show you some more information about that in a second.

The challenges are that these plot slopes could vary by various kinds of end points. It turns out that the slope really doesn't vary all that much, but the absolute sensitivity of tissues does vary a lot, and we will go into that. But you have to have a lot of data. You have to have at least one time-temperature combination for any kind of threshold in order to get this number.

This is data from Henrieck and Moritz, 1945, circa 1945, still probably some of the best data on human skin tolerance that there is. This is the threshold for pain down here, this is temperature here versus isoeffect time. As you increase the temperature, the time to sense pain becomes shorter, that is what that is telling you there. Also notice that there is a big difference between the threshold for pain and the threshold for damage, so big difference here. So pain is a good moderator for wanting to know whether or not you are going to have serious damage to skin.

But there is not much difference between what we might call moderate damage, a blister and full necrosis. They almost lay on top of each other on this curve. So once you get above this, you don't have much room for error.

The other point I want to make is that these curves are parallel. That is extremely important, because it validates the idea of using the Arrhenius relationship to do these conversions.

This is a comparison of time-temperature thresholds across tissue types for just the mouse. Again, note that there are variations in sensitivity of these different tissues to heat, but all the curves are parallel, again validating the idea of using the Arrhenius relationship for any kind of tissue that you want to look at.

It turns out that these slopes that we see in mice are very similar to what we see in man. They are slightly different. I'll just point this out here. In mouse there is what is called a break point, where the slope in the Arrhenius plot changes. In mice it is around 43. The R value below break point is about .25. That means for every degree that you go below 43, you have to increase the heating time by a factor of four to get the same isoeffect. The R value above the break point is about .5. That says if you go from 43 degrees for 60 minutes, it is the same as 44 degrees for 30. So you go by a factor of two above the break point.

We did this fit for human cells, and we found that the slopes are slightly different. But because most of the data that is out there is on so many different species, we decided to stick with the mouse. So all the calculations used the break point for mouse. The reason we do that is because it gives you a bit more conservative estimate on what the thresholds would be than if we used the human. The only data we had on man was skin, so we just decided for the sake of safety we would stick with the mouse.

Let me then go on to this new review that we did. These are the search criteria that we used, trying to look for papers. We also did recursive searches against papers that cited our 2003 review in the Hyperthermia journal. After that, we found 463 papers that had been published since 2002 with some combination of these key words.

We have 117 papers that are included in the review. This is a breakdown by tissue and whether or not we are talking about local or whole body heating. Why only 177 out of nearly 500 papers? Here is the problem. Nearly a third of these papers had lack of adequate thermal data. The temperature is either not measured at all or not measured adequately. Often health was not measured at the site of damage assessment. This kills us, because if we don't have the measured temperature at the place where the assessment is being made, then we can't make any sense out of it. So there were a lot of papers that we had to throw out because of that.

There was a large fraction of papers on laser ablation. These doses that they get with laser ablation were very, very high, and really didn't help us in terms of determining thresholds.

There were some papers on modeling which didn't have data; there were some reviews. There were experiments done on excised tissues, and we decided to throw those out, because there is no way of knowing how viable those excised tissues are. It is difficult to assess, so we threw those out except for some brain slice data which I'll show you in a little bit, which were carefully done. We did not do any excised tissue evaluation.

Accurate calculation of the CEM 43 requires full thermal history. Oftentimes that is not given to us. Sometimes what will happen is, they will say, we did 43 degrees for 60 minutes. We don't really know what that means. We have to assume that they did 43 for 60 minutes. We don't know anything about the heat-up, we don't know anything about the details of what they did, and that makes it difficult. So it is better if you have full thermal data.

Why is it better?

Here is my question. Here are two temperature profiles. Here is one that goes up to 46 for a few minutes, comes back down. It is lower than the other one for the remainder of the heat. Here is one that is flat and comes back down. Which one has the highest CEM 43? Take a vote. Option number one, option number two. Any guesses? Anyway, it turns out profile two is more damaging than profile one, even though you might be fooled by the fact that this one looks not as bad. The reason is because that short amount of time that it is at a high temperature, remembering that the rate of cell kill doubles for every degree that you go above 43.

That is why it is important to have the full thermal history. In ICES oftentimes we don't have it. When you look at this literature, it is just not there.

Here is a paper just published. I just did a Medline yesterday or the day before. Here is one on MRI. With MRI you can get very good thermal data. It could help us a lot to establish some of these standards.

But here is a typical example of a problem. We are looking at thermal ablation in the rat brain. Here is a hot spot, here is the temperature profile for that. Very nice data. The temperature profile is right smack dab in the middle of the ablation zone. Useful data, but the only temperature is in the center of the ablation zone. Where we need the data is at the margin and outside the ablation zone. If we want to set thresholds, we need to know what the temperature is here, what it is out here. If you are just looking in the ablation zone and you tell me you killed the tissue, great, but that doesn't tell me how close can I get to the ablation temperature without causing ablation. That is what we really want to know.

This is just a summary of some of the thermal sensitivities that we had in the last paper. Basically they are divided into different species here. We had ranked different organs, testis and brain being extremely sensitive. Here is CEM 43 here, whereas things like prostate, muscle, fat, skin are pretty heat resistant, and then a bunch of other organs in the middle.

An important point to make here, which I think is important when you think about thermal tissue sensitivity versus something like radiation. We would typically say that brain is a fairly radio resistant organ, because it doesn't have any proliferating cells in it. So brain on the radiation scale would be pretty radiation resistant, whereas testis, which has a high proliferating fraction, a lot of stem cells, is radiation sensitive.

Here you can see brain and testis are on the same order of magnitude in terms of sensitivity with heat. So you can't use the rules that we learned about radiation sensitivity when we talk about organ sensitivity. It is a different ball game.

In the report that we distributed around, we have expanded that figure that I just showed you. There is a table here that is again updated from our prior paper.

A couple of points I want to make here. We grouped various organs by ranges of CEM 43. Here is from zero to 20. Here is tissue type and whether or not there is acute or chronic damage detected. Then we have it broken down by whether or not there is histologic evidence or dysfunctional evidence or whether there is just gross assessment of tissue at the time it was taken out. So there are these different kinds of end points here.

I want to point out that anything in red here is something that is new compared to what we had in the last report.

There are very few assessments made of chronic consequences of thermal damage. This was brought up earlier by the gentleman from the FCC. I think we do need to know more about what the chronic consequences of thermal damage are. There is just not a lot of data out there. People look at acute effects, and then they stop and that's it.

The other thing that they don't do is, they don't look at the effects of repeated exposure. This was brought up also. There is very little information out there about that, and we will touch about that later, and the need for that kind of information.

We looked at both local and whole body heating in this review. In the prior review we didn't do that, we stuck with local heating only. But this time we included whole body heating. In fact, here is some data from Dr. Sharma, who is here at this meeting, thank you.

What we are looking at here is the fault increase in blood-brain barrier damage as a function of CEM 43 for whole body heating of rats. One point to make here is, different parts of the brain are more sensitive to change in blood-brain barrier. You can see there are some fairly important parts of the brain here, the thalamus and the hippocampus, that are fairly sensitive to this, whereas other parts of the brain like the cerebrum are less. So that is important to know. As was mentioned earlier, different parts of the brain behave differently with respect to thermal exposure.

I would point out that when you talk about total body heating, there is different physiology going on there. You can't cross correlate what happens with local heating with what happens whole body. Jack Hoopes awhile back looked at thresholds for blood-brain barrier damage in the dog, and found that for local heating that is between ten and 20 CEM. So that is a factor of ten higher than what you get with whole body heating and maybe ten or 20 fold higher. So something is different with whole body, and I don't think we necessarily know what it is, and I think it is something we need to try to sort out. But clearly the brain is much more sensitive to whole body heating than it is to local.

Sub-regions of the brain vary in thermal sensitivity. The age and time of the assessment are important. I emphasize the need for multiple time point measurements after a thermal exposure, not just one.

Here we are looking at cell death in adult rats here and also rat pups. You can see that the rat pup is much more sensitive to cell killing by heat than the adult. But even with the adult in this molecular area, you get about a twofold increase in cell death after heating. This is at a CEM 43 of 5.9 minutes, so it is fairly low, but still a substantial amount of cell death occurring there.

We did look at some functional assays that were done, particularly this hippocampal neuronal excitability. It has implications about seizures. This was done by looking at a certain part of the brain, the hippocampus, which has a dentate gyrus pyramidal cells, and do a stimulating electrode at different places and then do a recording electrode.

We look at this thing called P1 spike intensity. The major point I wanted to make here is that, you notice that with heating, you do get a reduction. This is percent change and it is going down, so that means you have a reduction in the P1 spike intensity, which has to do with the inhibitory neurons, their ability to dampen an excitatory response. So hyperthermia causes the inhibitory neurons to dysfunction, and it increases the likelihood that there would be some kind of an epileptic or seizure event.

The other point to make here is that immature rats are a lot more sensitive to that. The black line here is immature rats compared to more mature rats.

So it indicates that these inhibitory neurons are sensitive to heat, and the young is much more sensitive to this than mature. In the report I show that they have to do with changes in neurotransmitter, particularly GABA, which is maybe involved in this.

There was a lot of data on brain. Much of this report deals with brain. The other major organ where there was a lot of new data was testis. This is looking at decrease in weight of the testis for mouse, rat and down here, human. The major point I wanted to make here is that any data you get in the mouse completely overestimates what would happen in a human. Not that the rat is the same as the human, but in terms of the dose range, human seems to be similar to rat.

I think the mouse is probably not the right species to use for this, if you want to look at sensitivity of the testis to heat. Probably rat or monkey or human is a better choice.

The effect of heat on the testis is dependent on time after exposure, again indicating that it is really important to look at multiple times. This is looking at the fold increase in germ cell killing and for different species and different thermal exposures. But I want you to notice here this MMM, those are monkey data, and here is human.

The one human paper we have came from China. I can't believe they actually did this. They got the volunteers to sit in a hot bath at 43 degrees for 180 minutes, and then they measured testis size and did biopsies on them afterwards. Only the Chinese, I guess. But it is very good data. The only thing I would say about it is, it didn't measure the temperature of the testis. I think that is a drawback of this study. They measured the temperature of the water bath, but we don't know what the temperature of the testis was. I think it is probably lower than what this indicates.

I think a threshold for sperm damage is still not determined. We talk about stem cell damage. Monkey and human data are similar. So I think probably the best thing to do would be to do a study in monkeys to understand what is the threshold for germ cell damage in testis, which is the critical question, and it is unanswered right now. We don't have it. I think the tools are available. It just needs to be done.

If you look at sperm function, we do get recovery over time, but there is a rather decided decrement. Here is the human data. You can see a very decided decrement in number of sperm, sperm viability, sperm motility, and normal morphology after heating. But it does recover. It takes awhile, 100 days after this thermal exposure for sperm to recover back to baseline.

Just a note on skin. That has to do with thresholds for pain. I said pain is your best friend when it comes to thermal damage to skin. The point is that this threshold for pain is dependent on prior thermal exposure. So if you have an initial thermal dose of 112 CEM 43, and then there is an increase in the threshold for detection of pain if you have a prior thermal exposure. So this gets to the issue of repeated heat exposure and what that does to our ability to sense temperature change and also the tissue's ability to respond to that.

The laser ablation community has also spent a lot of time on the issue of thermal dose. Instead of using the Dewey formulation, they have used a different thing they call the damage index. It is derived from the Arrhenius relationship. It assumes that damage occurs with first order kinetics.

This is the equation for this rate of damage. You can see it is related back to the Arrhenius type analysis here. These are the different abbreviations here. C is the percent undamaged tissue, K is the reaction constant, A is the frequency factor. Ea is the activation energy in joules per mole. This is the key factor that they try to look at. Then R and T.

You can take various data here. This is data on thermal ablation from a variety of different fathers. You can look at the natural log of A, that is the frequency factor, and Ea. There is a log-linear relationship between them that is awfully good. Looks really good, doesn't it? We go, wow, we can use this for what we want to do, looking at thresholds for thermal damage rather than the Dewey formula. Maybe this would be a different way to do it that might be useful.

We tried this. The requirement is that you need to establish a range of time-temperature combinations to see the same isoeffect. There aren't that data like that out there in the literature. The Henrieck and Moritz data are probably some of the best, but it is hard to find this kind of information.

I think for it to be valuable, you need to have a predictable relationship between Ea and severity of damage. What we found out was that there really wasn't such a thing. So this is looking at Ea in skin, various end points for skin, just to point out a few highlights for you.

That is 50 percent necrosis of skin. That is a complete epidermal necrosis, and that is a full blister. So a full blister being less serious than full epidermal necrosis, and this is in between the two, it is obviously not a relationship that is meaningful here, that would be useful to relate the severity of one type of damage to another. That is a mild vascular reaction which is on the same order of magnitude as these others.

Future directions for this. I think it is important to assess temperature at the site above and below the threshold for damage. I can't emphasize this strongly enough. It is so frustrating to look at this literature and be so close to having an answer, and then have to throw a paper out because it is just not there.

It is important to utilize standardized isoeffects, and to identify various levels of severity effects. This gets to the issue that was briefly mentioned. That is, you would like to have a threshold that is lower than killing a cell. I agree. There is hardly anything in the literature about that.

So I think there are things that could be done. For example, you could look at -- we have got the neurophysiology, which for the brain would be great, but other things could be done. We could do genomic analyses on tissue. We could do all kinds of things to try to look at sublethal events, but somebody has to do the work and publish it, and so far there isn't much like that out there. There is insufficient data between 40 and 300 CEM for many organs and tissues. We either have stuff like whole body heating which is really low, or we have thermal ablation, which is really high, and there is not much in the middle. So we need a lot in the middle.

Assessment time after exposure is critical. As I said, there is few data on chronic effects, virtually no data on repeated exposures.

With regard to repeated exposures, we have to think about thermal tolerance. This was in the prior paper that we published. If you look here, you can look at the time to maximum resistance, that is, thermal tolerance resistance versus heating time at 43 and a half. The severity of the initial exposure affects how long it takes to get the maximum thermal tolerance. We are talking about hours here. So you get thermal tolerance, and it lasts quite awhile. It peaks out at somewhere between 20 and 40 hours after exposure, and it can last a long time.

This is looking at the time of thermal tolerance decay. This is heating time at 43 and a half, and it takes a long time, four or five days, for it to go away. So when you are talking about chronic exposure now, you are talking about bringing in the issue of thermal tolerance and how does that affect tissue response. The bottom line is, it is going to make the tissue less sensitive, but we know virtually nothing about this when it comes to normal tissue. There needs to be a lot more work done to really understand that.

Just a brief highlight on the report that Bill Wetsel put together. He has a fairly large section in that report on the TRPV receptors. TRPV receptors were just discovered around the time the first report came out, so there wasn't any discussion about it at all that I recall in the initial papers. These are the receptors that sense temperature change, so they are extremely important.

This one here is really interesting, because it is a negative regulator of fat generation. Activation of this channel is necessary to prevent maturation of pre-adipocytes to adipocytes. If you could activate this thing, it would ultimately reduce the number and size of fat cells, and could be used as a weight loss drug. Capsaicin, which is an extract from hot peppers, is known to activate this channel, giving rise to efforts to develop agents that could activate the channels as a means of weight control.

So this receptor is extremely interesting because it has such cliotropic effects, one being control of fat cell growth. But it is one of a family of such receptors. It has a sensor for both temperature and low pH, but it also responds to inflammatory lipid metabolites. Capsaicin is another of what are called vanilloids.

The knockout mouse for this receptor has a low baseline body temperature, so it communicates directly with nerves, goes back to the central nervous system and regulates the thalamus with respect to temperature control. So if you have a knockout mouse with this receptor missing, you have a lower baseline body temperature.

They are subject to regulation by other stimuli such as inflammatory pain, phosphorylation. These can alter the thresholds for detecting temperature change, which may be important. As I said, the channel activation can activate or influence hippocampal excitation and increase the propensity to febrile seize, all this coming from this one receptor, and it is only one of several. So I think it is an important area of investigation. I know Betsy's work is interested in it, other people in the room probably. I think there is a lot of science that can go forward with these receptors, trying to understand how they affect body tolerance. It could be that there are polymorphisms in this receptor that affect individual responses to heat, but as far as I know, no one has looked at that.

Bill is an acknowledged expert in neurocognitive behavior, but has never done anything in hyperthermia. So I think it was good for him to do this, to look at it completely unbiased, have no preconceived notions about the field, and came to his own conclusions.

He discussed the effects of neonatal and perinatal hyperthermia exposure later in life. He updated the literature related to thermal exposure and cognitive function in humans. I want to make a point here. There was inadequate statistical design, he thought, in several of his studies, and that heat stress can decrease motor function, which could influence cognitive tests. That was something that I wasn't aware of before, but maybe the rest of you that have done this stuff know that already. But I thought it was interesting.

He emphasized the need to include a battery of different neurocognitive tests for this kind of research. Several of the papers were, he thought, deficient in this regard, making conclusions difficult to reach.

That is the end. So I'll turn it over to Marv.

DR. MORRISSEY: Next we are going to have Marv Ziskin.

DR. ZISKIN: Good morning, everyone. My name is on here as the presenter of this talk, but really as far as the report goes the mass of credit should go to Joe, who is responsible for most of the actual report.

Joe has done a fantastic job of organizing this workshop. The instructions from Joe to me was to assume that everyone has read the report in detail, and just give a very quick review, and particularly make it very short.

So in summary, I would like to present that the bottom line thresholds for causing developmental abnormalities is greater than one and a half degree C for over an hour would be required, from two to two and a half degree C for a half to one hour would be required for producing abnormalities. Then when we get to high temperatures, greater than four degrees, in the order of a few minutes, ten to 15 minutes would be appropriate.

That essentially would be my talk. However, I would like to say a few other things before stopping.

When we come to thermal bioeffects, there are two areas in which cells are most sensitive to. Rapidly dividing cells are the key thing, rapidly dividing, particularly in the embryo and fetus where we have rapidly dividing cells, we have importance.

From the point of view of medical concerns, the impact of these things, cells can die. You have several thousand brain neurons dying every day in non-replacement. We think that for the most part, on a daily basis there is not much effect. Skin cells are dying in multiple thousands, and red blood cells and so on. Most of the time they are reproducible, and from a medical point of view are not terribly significant, but you hit a few cells in the embryo, you can have a major effect. A single cell in fact can have a major lasting damage in an embryo, so that non-reversible things will have very important significant medical consequences.

In terms of hyperthermia for the embryo and fetus, it is known to create congenital abnormalities of teratogenicity effects in every single laboratory animal that has been studied. Here is a list of birds, hamsters, mice, rats and non-human primates, and it is presumed to be also true for humans.

I would like to give a little bit of my personal experience in this when working in the laboratory of Marsh Edwards in Sydney, Australia, whom I give credit for having the largest database in maternal and fetal effects. This is a charming young lady who is typical of the 29 guinea pigs who volunteered for this study. The next slide shows the specially designed cage to allow the pregnant abdomen to be exposed by ultrasound, which was my major interest at the time.

Here she is placed into this water tank. There is the transducer producing the sound. It is aimed particularly at her pregnant abdomen. The actual beam was very well tailored to reproduce the type of temperature elevation history that was occurring in the guinea pigs. What Marsh Edwards would do is, he would put them into an incubation oven with a hot air type of circulation. Over many years he developed a very large database on the effects.

So in my studies that I did, I showed that my results from heating with ultrasound were consistent with those that he had done with the hot air incubator. What are the effects? The most important effect, the most sensitive effect, the first effect that you would see is, at the right time of gestation there is a small brain weight that we would see.

This is a normal brain for a guinea pig neonate. This is one that has been affected by the hyperthermia in the uterus. If you look at the two, it is the same shape and size and so on. I mean, size is much smaller, that is the big difference. There is a lack of number of neurons that went into this deformation here.

The lack of cells has an effect. It is no challenge to decide which of the little piggies did not receive the hyperthermia in this case.

Here is another example. You are seeing a newborn, the lower half of a newborn guinea pig. Here is the leg which is normal. But look at this. This is an example of a clubfoot or talipes equinovarus. This is a typical type of defect that heating the fetus by several degrees over a period of time can result.

This is not the worst that you see. There is a more generalized problem called arthrograposis multiforme which affects many joints, many muscles and many bones. As an example, this is a normal tibia and fibula in a newborn guinea pig, but look at this. The fibula is already absent. Here is the tibia, which is very much distorted, smaller and twisted in this example.

This is looking at the side of a guinea pig stillborn. This is the head region up here, this is the rear end at this end. If you look at the surface of this rear structure, you see that this is normal looking flesh. But look what is happening at the forelimbs. All of the skin, the muscle, has turned to a fatty waste, atrophy and so on. It is non-functional, and the animal of course died here.

The defect that is causing this is located in the spine, in the cervical region which innervates this region. If you look at the cross section of the cervical vertebrae, this is what it is.

For those who are familiar, looking at the cross section of the spinal cord, there is a very small central canal that runs through the spinal cord, and the rest of it is filled with cells. However, here we have this great big evacuated region over here, where the cells never divided and they never filled in this region. The lack of these neurons which innervates the limbs is what caused that atrophy and the defect.

The reason for the lack of the cells that would grow is because of the active cells that produce the multiplication that the ependymal cells have been terminated in their cycles. Here is what that looks like. This is one hour after exposure to the hyperthermia. You see a cell that is in mitosis. This is relatively a normal appearing one, but look at this cell over here. The chromosomes are all clumped up. They are not normal at all. They are swollen, they are sticky, they don't separate very well. Here is another one that has been affected.

Then if you look at it a half hour later in these ependymal cells, you can see where the chromosomes have formed an agglutinated mass. They are totally non-viable and they die. When an embryonic cell here in the brain dies, all of the progeny that would have come from this is gone; you don't have this produce. So this has major impact. If they were replaced, this wouldn't be a medical problem, but this certainly is.

Now, this shows that the same type of problem exists in humans, too. This arthrograposis multiforme does occur, it is not that common, but it does happen in humans also.

I mentioned about, it is important when during gestation the exposure occurs. This is a graph of the incidence of small brain size as a function of time during the exposure. During the active development of brain in the third and fourth weeks of gestation in the guinea pig, we have the most production. If you go before that or various places beyond that, you don't see it very much, so that is very important during the gestation period.

Working with Mort Miller, I had gathered all of the world's literature on congenital abnormalities produced by hyperthermia. What you see here is a graph of the temperature at which an abnormal effect occurred, and this is the log of duration of exposure. These points are the very lowest that occurred for any particular time or for temperature. This is the effects.

There were some experimenters who looked at the same abnormality, but at different times or different temperatures. I connected the points on those similar things. These three points were for the same abnormality. This for example was the small brain size.

What was very interesting up here is that all these lines were parallel. That is no accident. It turns out that is very consistent with the Arrhenius theory of activation energy defects. But what was so fascinating here was that the Arrhenius theory is based upon a single chemical reaction changes in which you would get this type of effect. But here is a very complicated fetus, and the same law is applicable to this, too. I drew this dotted line, which happens to be where the CEM 43 equals one, where that would occur. It turned out for this data, that boundary was a very good boundary in separating out anywhere where you would have an effect or you would not have an effect.

I was also asked to compare whole body exposure things from SAR, what sorts of temperatures would you anticipate that these SAR values would produce, and how they might relate to fetal abnormalities.

If we look at about 15 watts per kilogram, it would produce a temperature rise something on the order of four degrees C, and if extended over a certain period of time would lead to abnormalities. IF we look at four watts per kilogram, this is a common one that we speak about as a threshold for an effect, we would have a one degree rise. Although there might be some behavioral changes, there is no tissue damage that occurs at that level.

If we reduce down to a one and a half watt per kilogram, we would have a temperature elevation of .4 degrees, and that could be used as a convenient conservative safe level for actual tissue damage. If we look at .4 watts per kilogram, which happens to be more or less the universal RF safety standard for this thing, we see that that would give us something in the order of about a tenth of a degree temperature elevation.

It is also from a biomedical point of view interesting to see, in normal pregnancy -- well, even without the pregnancy, there is a normal diurnal variation that your body temperature rises and falls each day, on the order of plus or minus a half a degree. So these things are very conservative, compared to our normal range.

Anyway, that is all the comments that I wanted to make at this time. I want to thank you for your attention.

DR. MORRISSEY: Thank you, Marv. Amazingly, we are ahead of schedule. I did not anticipate that. But we will take advantage of it, because I don't think we will afford ourselves the same luxury during the discussion periods.

Thank you very much, all the speakers. That gives us our direction and our marching orders. Why don't we have a break right now? I'll load Chris Gordon's slides. We will take a short break, 15 minutes, and we will come and listen to Chris give us a talk, and then we will go into our breakout sessions.

(Brief recess.)

Agenda Item: Session 2: Whole Body Heating

DR. MORRISSEY: Our next speaker is Chris Gordon from the EPA. He is going to talk about human thermoregulation. With my announcement, that is another hint to make your way back to your seats and we will get started. DR. GORDON: Thank you very much. We will get started here. I used to do research in microwaves and temperature regulation with EPA back in the '80s. Then EPA phased out their program, so I have been out of it for about 20 years. But I have always been doing temperature regulation. So Joe was kind enough to ask me to come back and prepare a talk on human thermoregulation and microwaves and so forth, so I'm glad to be here.

The outline of my talk. I am going to talk about fundamentals of thermoregulation and sensitivity versus capacity of the thermoregulatory system, and then human thermoregulatory responses, and then finally if we have time, I will contrast RF exposure with exercise and drug induced hyperthermia.

Thermoregulation is the ability to regulate a stable core temperature independent of changes in ambient temperature. It is the classic definition. The unique aspects of thermoregulation relates to a lot of what we talk about here and some of the talks alluded to. It is a unique autonomic regulatory system that relies on behavior and conscious awareness of the environment. In other words, you can walk around for years and years with high blood pressure and never know it because you just don't know it until it is too late. You might not feel good, but you don't know that you have high blood pressure, but you always know if you are too hot or too cold. We live and die with the Weather Channel and the weather report. We always want to know what our thermal comfort is almost all the time. Yet it is an autonomic system. It is regulated day and night with and without conscious awareness.

Let me just talk about the stability of core temperature using a comparative physiological perspective, comparing a 30-gram mouse and a seven million gram African elephant. Their mean core temperatures are for the mouse and the elephant almost the same, which is pretty amazing when you think about the differences in body mass.

But this is the time course of core temperature of an elephant over five days. They measure this temperature by feeding the elephant a data logger inside a cantaloupe. Then whoever drew the shortest straw five days later had to go find the data logger. What you see is the elephant has a circadian rhythm of temperature day and night that is very stable. There are probably a few little artifacts here. So the elephant has an average core temperature over five days of about 36.2 degrees.

The mouse has about the same average temperature, but the mouse gets there a different way. This is two days of core temperature of a mouse monitored by telemetry, and the mouse has a different strategy to thermoregulate; it is up and down and up and down and up and down day and night. This is a mouse at 25 degrees Celsius, active and asleep and so forth, and he is eating.

So with that to introduce you to the stability of core temperature, when we talk about the fundamental of thermoregulation, it is the body heat balance equation. This describes the relation between the heat storage in the body, and it is dependent on the metabolic rate minus the avenues of heat loss. So heat storage when it is zero over a long period of time, then the animal is normothermic. When metabolic heat production exceeds heat losses, then the heat storage is greater than zero and you are hyperthermic, and when heat loss by these avenues exceeds heat production, then your S is less than zero and you are hypothermic.

It is a little more complicated than that, though. In this diagram here from the 1950s, there is the thermal core and the thermal shell. Here are isotherms in a human body at 20 degrees ambient temperature. There is the thermal core that we are all familiar with, it is at 37 degrees. Then there are these heterothermic areas where temperature decreases out to the periphery as such. At 35 degrees, under heat stress, you see that the thermal core now expands into the arms and the legs. This is significant when we start talking about the thermal effects of RF.

A simple equation then to express the heat balance in the core, the heat balance of the core is the metabolism minus the conductive and convective heat transfer between the core and the shell minus the heat that you would lose when you expire water by respiration. Then if we throw in the energy that is absorbed from RF radiation, then the heat balance of the core is metabolism plus your RF heat minus the conductive convective heat exchange.

Then the shell now is slightly different. We can express the heat balance of the shell as the absorbed heat from RF plus or minus the heat that is exchanged between the core minus the heat that is lost by evaporation, convection, conduction and radiation, that is, from your skin outward.

The basic thermoregulatory pathway I am depicting here. You have the body with internal thermal receptors in the brain and in the core, and then you have cutaneous warm and cold receptors which feed up through the spinal cord, the brain stem and into the hypothalamic area, which I will talk about more later.

Under heat stress conditions down here, you activate the mechanisms to dissipate heat. That is, prefer a cool environment, which is a higher level cerebral process. You eventually activate the pituitary gland. You sweat, you respire more, and skin blood vessels dilate to dissipate heat. With cold stress, the pathways are activated to retain heat and increase heat production, so you select a warmer temperature. The pituitary activates the thyroid axis, that is, thyroid simulated hormone. Skin blood vessels construct, piloerection and hair follicles, the adrenal medulla releases adrenaline, and then the skeletal muscles activate, and you shiver. So that is the basic breakdown.

To get everything into perspective, I want to talk about the basic thermoneutral profile of a homeotherm. What we have is several things going on. There is a thermoneutral zone of ambient temperatures, where our basal metabolic heat production shown here in the blue is at basal levels. Our core temperature is also nicely controlled. Then we thermoregulate by varying skin blood flow within the thermoneutral zone. So we maintain metabolism at the lowest level, doesn't cost us much, and then we can change skin blood flow to modify heat loss.

So when you start to cool, metabolic heat production starts to go up. This core temperature stays the same, and skin blood flow is very minimized until you get to a certain point. If you get too cold, your body starts to recognize that there is a danger of frostbite, and starts to increase skin blood flow to protect the tissues. Then evaporation will increase. As you go above the thermoneutral zone, then evaporation starts to rise and you sweat, and core temperature regulation is not very effective above the thermoneutral zone. Skin blood flow maxes out at a certain point. You can't do much else with skin blood flow because the ambient temperature is too warm and you can't dissipate much heat.

Taking that diagram now and relating to RF exposure, I get the same lines right here, skin blood flow metabolism, this classic profile here. Here is evaporation and core temperature. If you are exposed to RF in the cold, your options are to maintain heat balance, you can lower metabolism and you can also increase skin blood flow. This has been shown in various experimental animal studies.

Then if you move the RF exposure up to the thermoneutral zone, the option to lower metabolism is no longer there because of your basal levels. So your only options are to increase evaporative water loss, sweating, and to increase skin blood flow within this zone right here. Then when the RF exposure occurs at above the thermoneutral zone, shown right here, now your options are much less, because your skin blood flow can really not do much more. Your only option is to sweat. If you can't sweat, then you are going to be in deep trouble real quick.

Some basic characteristics of the human thermoregulation and the capacity in the cold. It is pretty amazing that our overall skin blood flow varies from about 150 to 2,000 mls per minute, depending on ambient temperature. Below 25 degrees, a semi-nude man is maximally vasoconstricted. We double metabolic rate by shivering after exposure for one hour at five degrees.

Here is some old data, but it is still pretty reliable, of skin blood flow in the hand of human subjects, and here is skin blood flow on the Y axis. In the top panel we have got ambient temperature, and it is a little more variable, but you see above ambient temperature of about 25 degrees, skin blood flow in the hand increases. Then if you express in terms of skin temperature it is lot more reliable because the thermal receptors in the skin are driving that response, so you reduce the variability. You can see that relationship between skin temperature and blood flow, just to show you how it all comes together.

In the heat, our basic thermoregulatory characteristics are, sweating is activated during rest at an air temperature of about 32 to 34 degrees. We have about two and a half million sweat glands. The opening of our sweat glands is about 90 square centimeters. A working human that is in pretty good physical condition can evaporate sweat at about two liters per hour for more than five hours, totally a heat dissipation of 1400 watts.

Here is some old data from classic thermal physiologist Adolph back in 1947. This shows here air temperature versus evaporative water loss which I have expressed in watts. You see in real terms what I was trying to depict in that diagram earlier, how sweating is minimal, and then it increases through the thermoneutral zone. Then with working we have a very well developed capacity to increase sweating when we work.

Now I thought I would talk about some basic neural mechanisms of thermoregulation and help everyone put this in perspective. We have internal thermal sensitive neurons in our hypothalamic area, preoptic area, as well as other parts of the central nervous system. These are driven by peripheral thermal stimulation. As the last speaker alluded to, the TRPV receptors in our skin that are activated drive these pathways.

I have got a diagram here to show you a little more about how it all works. In the skin we have warm sensors and cold sensors. These fibers eventually pass through the spinal cord into the hypothalamic area. So we have a neuron here that is sensitive to heating. What we have up here is the temperature of the brain, 35 to 39 degrees, versus impulses per second. This neuron here shows a response like this, and then the cold neuron down here increases activity with cooling. So these neurons have been identified in a variety of mammalian species as well as reptiles and fish, believe it or not. So we have receptors, hypothalamus to the outputs, so these drive thermal receptors for heat dissipation, and then these drive the thermal receptors for heat production.

Also, I am going to show here that an integrative neuron drives skin blood flow. So you have an integrative neuron here that increases activity due to this reciprocal inhibition of this pathway, where it is activated by cold and suppressed by warming. Your skin blood flow is driven by your sympathetic pathways, and the more those pathways fire, as I show here, the more your sympathetic neurons fire, then that is going to construct these blood vessels.

We have a dashed arrow showing that at normal thermal, everything is in balance. Then if you raise the temperature of the brain to 39 degrees, then you start to heat up this neuron, and you are driving sweating. This all makes sense, prefer cool temperatures, but you also inhibit this integrating neuron. So you slow down the activity here and you vasodilate. Then if you cool the brain, then we just drive the opposite. So now the cold pathway is being activated. This is being suppressed because you have this reciprocal inhibition, and you are stimulating this integrative neuron, which causes constriction of the blood vessels. So it is an oversimplification, but I think it works for right now.

Set point and thermoregulation is what I am going to talk about next. Thermoregulation is nice to explain in terms of thermoregulatory control. It still remains the fundamental concept with most people. There is always disagreement. It can also apply to the local regulation of tissue temperature.

Like I said, the anterior hypothalamus and preoptic area is the site of the integration of those pathways which I just showed in that diagram earlier. We have a typical servo-loop where you have a set point, a set point temperature. It is compared to the feedback and the error signal generated drives controlling elements which elicits control over a system that is affected by a load of some type in the classic engineering sense. Then if you apply these same things at temperature, you have your set point temperature, you have the thermal effectors that I was showing earlier. The controlled system is basically body temperature. Then the load is either heat and cold stress or RF exposure. Then the feedback is the thermal receptors in the skin and the core.

A little diagram here to show the set point and how it all works. I work a lot with mice and rats, but I could have a diagram of a person here.

You have an animal in here, and he is in a gradient. So he can go to the cold end or the warm end. That is the thermal preference. We all know how sensitive our thermal preference is, and behavioral regulation is very sensitive. The idea is that there is a set point temperature and the core temperature regulates around this set point. Heat dissipation brings the temperature down and then heat gain conserving mechanisms brings it back up. We basically oscillate around a set point. The set point of course can change.

Showing here is with a fever. If you present the animal with a pyrogen that raises the sudden set point, then the animal feels very cold, because now the set point is suddenly raised above its body temperature and it goes to the warm end of the gradient, and then as it reaches a steady state it eventually comes back to normal. As when we get a fever we feel very cold and when we seek a warm temperature, put on blankets and so forth.

The case we studied here with RF internal and external heat sources, the set point stays the same. If you present the animal with an external heat challenge or an RF heat challenge, basically the animal feels very hot and seeks out cold temperatures, and then regulates behaviorally, if it can. Behavior is a very inexpensive way to regulate, whereas metabolism, sweating, those are a lot more costly. So if we can use behavior, we will.

This is from some studies I did back in the early '80s. I used a temperature gradient for mice and hamsters. The temperature gradient was made of a wave guide that passed 2450 megahertz RF through it. It was heated at one end and cooled at the other. The animal could move about in a shuttle box. This is about 30 feet long. You could track and animal with little photo cells in this gradient.

This is some old data that I thought I would throw up here, force you through it. Here is a mouse, and he goes into the gradient and he stays in there for an hour, and then eventually the animal is at a preferred temperature of about 31 degrees. Mice prefer nice warm temperatures. Then you turn on the microwave exposure and the animal suddenly starts moving around and slowly starts moving to the coldest end of the gradient down to about 18 degrees Celsius. Then you turn off the power and suddenly the animal realizes that its body temperature is quite warm, and it recovers and comes back to the warm end of the gradient.

In terms of this group here today, I think the two questions that people want to talk about with thermoregulation RF exposure is sensitivity and capacity. In other words, sensitivity is what is the magnitude of the error signal required to activate heat dissipating behavioral and autonomic responses, and the capacity is how effective are heat dissipating thermal effectors to maintain homeostasis.

Factors that affect sensitivity are the frequency of RF, the ambient temperature, age, rest versus exercise, preexisting conditions, genetic background and thermal adaptation.

Here is some data that is just recently coming out in the literature. This is not RF, but this is just to show how genetic background and thermal adaptation could affect our sensitivity to RF exposure. This is the sensitivity of people living in Japan in a temperate climate or Malaysia in a tropical climate. They measure their sensitivity in different parts of the skin to a heating element.

Number one at the forehead is very sensitive. The skin on the calf is least sensitive. You see that someone raised in a temperate climate has much more sensitivity to heating than people raised in a tropical climate. I'm sure you have probably met people through the years that were born and raised in a tropical area, and they come to this country and they are miserable in a climate like we have right now. Then in the summertime they may be wearing a sweatshirt or T-shirts and so forth. So thermal adaptation and where you were born and raised and your genetic background will have a definite impact on your sensitivity to warm and cold temperatures.

Factors that could affect the capacity to dissipate RF energy would be age and health of the subject, ambient temperature, relative humidity, your previous acclimatization to a warm or cold climate, and again genetic background.

To understand the thermal sensitivity of the CNS, receptors evolved in lower vertebrates like I said earlier to respond to changes in body temperature. Homeotherms like us respond more to changes in skin and shell temperature, because our body temperature stays much more stable than the lower vertebrates. But the central receptors are nonetheless critical in exercise and in fever and in RF exposure.

Years ago, 30, 40 years ago, they used thermal probes in the brain to characterize the behavior of these central thermal receptors. I want to show you a diagram here and how it relates to the RF exposure I am going to talk about in a minute.

Not that anyone has ever lowered a thermo into a human brain, but it has been done in animals, from mice to goats to pigs. You lower a tiny steel tube into the preoptic area and you perfuse water through this thing, and you can heat and cool the brain stem selectively and monitor the thermoregulatory responses.

This is an overview of what was found in these studies. If you start to raise the hypothalamic temperature, you can elicit sweating, and in a thermoneutral room sweating is elicited at 37 degrees. Then if you lower the temperature of the room, then this threshold shifts upwards to higher and higher temperatures, so you don't activate sweating in a cold room until you get out to two degrees above this threshold at 37 degrees. That is because thermoregulatory responses are classically defined in terms of an interaction between brain temperature and skin temperature, so the response is a combination of a proportionality constant multiplied by the difference between brain temperature and the set point temperature in skin. So you can see how this might all work, and in the data I will show you in a minute, you will see how it does work.

The thermal receptor responses to RF radiation is what I wanted to talk about next. The study I am going to summarize here is from Eleanor Adair, who did a lot of these human studies, some very nice work in the late '90s into the early part of the 2000's. I have tried to capture some of her work and present it here.

I have had to copy the graphs right out of her papers. Let me just walk you through this real quickly, and see. What we have here is the esophageal temperature, this line right here, then these are various skin temperatures that are plotted. Then down on this panel is sweating, measured with a sweat capsule, then this is metabolic rate right in here, which doesn't really change. This is skin blood flow in the back and in the chest and in the arm, done with laser Doppler probes.

In this example, this is the mean responses of seven volunteers to 2450 megahertz at 70 milliwatts per centimeter squared, at an ambient temperature of 28 degrees, which is right at the thermoneutral zone. So you would not expect metabolic rate to change much in the thermoneutral zone.

The whole body SAR is about one watt per kilogram. The subjects had their back to the antenna. There is an immediate rise in the temperature of the back, and very little change in the esophageal temperature. The esophageal temperature is probably the best measure of the core temperature, although not a very comfortable way of measuring it.

But you see here also that the skin temperature of the back starts to rise really quickly, and then the sweating eventually starts to become activated in the chest and the back as well.

The next graph, she did several different frequencies. This is very nice here. This is the mean response of seven people to 100 megahertz, which is near resonance for an adult human. This now is eight milliwatts per centimeter squared at an ambient temperature of 31 degrees, and the average SAR is about .54 watts per kilogram.

The interesting thing about this exposure is, the changes in core temperature, esophageal temperature, are fairly modest. What you will see is that the temperature of the ankle rises abruptly. I am pretty sure that is a hot spot for focusing of the energy, but maybe someone can allude to this later. The sweating eventually starts to kick in, but it is not really robust. You see skin blood flow start to kick in a little bit. Just keep in mind, this is a half a watt per kilogram SAR.

Here is a summary. This is what I was alluding to, to the hypothalamic probes to heat and cool the hypothalamus. Here is the sweating response for seven subjects at 2450 megahertz at 31 degrees, at 28 degrees and at 24 degrees. The nice thing about thermoregulation is that it usually always makes sense. Yes, you are at a warmer temperature, so you are going to start sweating at lower SARs as compared to a cooler temperature, where at 24 degrees there is hardly any sweating, and at 28 degrees there is a little bit of sweating. This is like those parallel curves I was showing in that earlier slide. This is sweating on the back.

Here is another summary of the same kind of thing at 220 megahertz, again at three different ambient temperatures. This is at 31 degrees, so the sweating is greater. This is power density. This is change in sweating on the chest and the back. You see that at the warmer temperatures, the system activates quicker, which it should do to dissipate the heat in both cases.

But behavior is the key here. We use our behavior all the time to tell us if the environment is okay or if it is aversive. The RF field I have always thought is fascinating, because some frequencies of RF are deposited primarily on the skin, but sometimes it is deposited deep in the body. I have always wondered, when it is deposited deep in the body, can your system tell you that there may be a problem. So like I say, it is our first warning of an impending thermal stress of heat induced damage.

From the studies of Adair, and this is a summary here, the subjects are all being exposed for about 50 minutes. Then she used a classic scale of thermal comfort, so the subjects could report a numerical score. Zero means it is uncomfortable, I am slightly uncomfortable, that is a one, and so on up to number four, a scale of four would be it is intolerable.

Now let me just go slowly here. This is 2450 megahertz. This is power density, but I put the SAR of one watt per kilogram right here. Then I have three different curves, 24 degrees ambient, 28 and 31. So you see that you get up to about a half watt per kilogram, and the reporting of thermal discomfort of about 1.5.

I think it is interesting that this is not affected at all by ambient temperature. At 31 degrees they don't report any more thermal discomfort than they would at 24 degrees. But the autonomic response is definitely different. Now at 220 megahertz we are getting closer to 2450, is deposited more peripherally in an adult human, and now 220 megahertz, the curves are quite a bit different. But here is 24, 28, 31 degrees; they are pretty much on top of each other. You can also see that the thermal discomfort at the highest SAR is about the same, give or take, for these two conditions.

Then 100 megawatts down at the bottom, I hope you can see this. Notice first that the scale is different here. The highest level of thermal discomfort is only .6 at the highest SAR of only a half watt per kilogram, whereas to put a perspective, .6 would be down here. So while this frequency elicits autonomic responses, the ambient temperature effects are not all that exciting. In fact, at 31 degrees down here in the red, the thermal discomfort is apparently less than at the cooler temperature. I can't tell you whether that is significant or not, but there is no remarkable effects of ambient temperature on the thermal behavior based on thermal comfort in these subjects, which I think is pretty darn interesting.

So the capacity to thermoregulate during exposure is, body size is critical when the dose of RF is normalized to body weight. This is some work that I had done with interspecies extrapolation years ago. This is before we had all this nice human thermoregulatory work that is published in the literature now.

The work I did, I found that if you plot body weight in logarithmic scales versus the SAR and calculate the threshold SAR to elevate core temperature, you see this nice inverse relationship where mouse is up here at about 30 watts per kilogram, then golden hamster, rat, rabbit. That is at ambient temperature of 20 degrees. Then at 30 degrees these thresholds, not surprisingly they drop. So the threshold SAR at a raised core temperature decreases dramatically.

Then I have also plotted the work of Adair for 2003, because she did calculate a threshold SAR to raise body temperature in humans to 31 degrees. This trend comes out here, but it is flattening out.

Threshold is a statistical calculation. What I like to do is look at the SAR for a one degree rise in core temperature. It is a little more reliable. You get a nice linear relationship between body size and the SAR to raise core temperature at 20 degrees here and 30 degrees; of course this all shifts. We don't have human data, because the SARs were never increased enough to raise core temperature by one degree in those human studies.

Now I want to finish up here comparing the physiological responses to RF heat versus exercise and drug induced hyperthermia. These are all unique circumstances, I feel.

Here is human studies where they looked at the thermal response to MDMA or Ecstasy, which is a drug that is abused quite a bit in nightclubs and so forth. People die from Ecstasy because they become hyperthermic. They come to the emergency rooms with body temperatures of 42 degrees Celsius, if they overdo it.

In this clinical study they looked at -- the volunteers take a radio transmitter pill so they can measure their core temperature, and then they measure their oxygen consumption. At 30 degrees their core temperature compared to controls goes up to about half a degree.

Look at oxygen consumption. They get the Ecstasy, and then their oxygen consumption starts to increase. Let's just convert this to SAR, 100 mls of oxygen per minute is equivalent to about 33 and a half watts of heat. Down here what I have converted is small print. This is power now in terms of metabolism. So what we are talking about is a half a watt per kilogram led to about a half a degree rise in core temperature under this condition.

Now, exercise. This is a nice study from quite awhile ago. I copied this out of the original data here, so it is a little fuzzy. Humans were exercising on an odometer, so they are working at 40 watts or 90 watts and then recover. What I want to show you here is that this is their metabolism that goes straight up. That is equivalent to about 2.4 watts per kilogram of additional heat, in addition to basal metabolic rate. Then when they exercise at 90 watts up here, that is about five and a half watts per kilogram.

So they also measured their body temperatures when they are exercising. In this paper they were very interested in trying to measure evaporative and radiated heat loss and that sort of thing. The point here I want to show is that this is the oral temperature and the tympanic temperature at 30 degrees.

What you see is a very modest rise. During the same period, two and a half watts per kilogram, their oral temperature only goes up just a few tenths of a degree. Tympanic drops. Tympanic is a little shaky. That is dependent on blood flow to the area, to the external jugular and so forth. Oral if it is done right is pretty accurate. When they exercised at 5.6 watts, 5.5, the core temperature starts to go up quite a bit more. It never really gets up all that far.

Now I am at the point to summarize all this, and the various thermal effects of these treatments. It is pretty much all at 30 degrees Celsius, right in the middle of the thermoneutral zone for humans.

What we have then is the treatment SAR, and then the change in core temperature. For the drug, the MDMA, a half a watt per kilogram led to about .6 rise in core temperature. Exercise at 2.5 watts per kilogram, that is only about a tenth of a degree rise in temperature. That is at 30 degrees. Then 5.4 watts per kilogram, about a half a degree.

With RF, 2450, one watt per kilogram, whole body, about a .15 degree rise in core temperature after 50 minutes of exposure. At 100 megahertz, which is resonance, half a watt per kilogram SAR, about the same, about a .1 degree rise in core temperature.

There is not a lot of equivalency between -- I can't imagine ever subjecting someone to RF of five watt per kilogram, especially at resonance. We can talk about that later.

But anyway, these things are difficult to compare, but it is nice to put them out here so everyone can get a feel for how the adult human body, these are all fairly healthy adults as well, how they respond to these different types of SARs from a drug, from exercise and from RF.

I think that is it.

DR. MORRISSEY: Thank you, Chris, that is a great talk. At this point we will have a few questions if anybody does have a burning question.

DR. REPASKY: (Off mic.)

DR. GORDON: I actually skipped that part. It goes up because the animal's tissues are getting warmer and it is breathing a little faster, and it is trying to escape. The lower level is much easier to talk about, it is much easier to predict mathematically.

DR. REPASKY: (Off mic.)

DR. GORDON: How do we compare that? That is interesting. With fever, of course, the set point is going up. It is a totally different thermoregulatory mechanism, because you are trying to raise heat production. I didn't really look for the data, but I'm sure there is some old data to measure metabolism, there must be, with fever and to compare that.

DR. REPASKY: (Off mic.)

DR. GORDON: Exactly. Again, it is regulated versus forced, right.

DR. MILLER: Let me ask you to restate the question, because back here we couldn't hear what the question was.

DR. GORDON: She was asking about fever and the metabolic load, the metabolic load you encounter with fever, and comparing that with these other situations of hyperthermia. I didn't have the answer.

DR. MILLER: I understood the question more or less from your answer, but it would have helped the audience to have the question first.

I have a different issue with your presentation. It was a fine presentation. The third trimester fetus during pregnancy becomes a half a degree warmer than the mother. It is a very substantial thermal dose. It lasts for roughly 13 weeks. Any elevation above that temperature, the fetus is caught in a 37 degree water bath, and it is very difficult for the placenta to dissipate that heat. In fact, it isn't quite capable of doing it.

So there is a different concern here, or an elevated concern, for the pregnant woman and elevated temperatures per se, particularly in the third trimester, where the fetus is already elevated by a half a degree. I think that ought to be addressed.

DR. GORDON: That is a good point. Yes, the fetus has nowhere to go; it is bathed in --

DR. MILLER: It is stuck.

DR. MORRISSEY: Chris, we have a session devoted to that. We are going to get into that in more detail. For right now, before we break out into the discussion groups, just questions that are directly pertinent to the -- although that was a pertinent question, but we have another session to expound on that.

PARTICIPANT: I have a question about your last slide, where you show frequency versus SAR.

DR. GORDON: The summary slide?

PARTICIPANT: Yes. Since it is a whole body rise of temperature, what can we deduce from this? We know that at 2.5, the absorption is mainly reversible. What we describe is the whole body SAR, so we mask somewhere the local increase of temperature. When you say 2.5 giga, one watt per kilogram in 0.15 degrees rise of temperature, and when you have 100 megahertz with half watt per kilogram, you increase 0.1. But at 2.5 giga, the energy deposition is mainly -- to have the core rise of temperature you need to have a high rise of temperature of the skin, for instance.

DR. GORDON: I'm not sure if I understand the question. I think there are issues with showing the core versus the shell. I think that is something that we don't understand with RF deposition. The body has the ability to move heat through the shell, so you might not see a rise in the core temperature very much, even though there is still a heat load.

The heat has got to go somewhere. For example, the responses of 100 megahertz aren't very robust, but they are still getting the heat, it has still got to go somewhere.

DR. FOSTER: A long time ago, there was a controversy about thermally based standards. It was pointed out even by you that if you have someone under high ambient temperatures with a high relative humidity, even a small amount of radio frequency energy might push them over the edge. That may have been the point that you raised.

Do you have any comments on that? Should that be taken into account in exposure limits?

DR. GORDON: Yes, I think it is very critical. If it is in a warm humid environment, your system has nowhere to go. Sweating is ineffective, vasodilation is ineffective because you have maxed out. So any little bit of our effort, let's say we are talking about the chronic exposures, maybe that would put you over the top, whatever that might mean.

DR. FOSTER: Is that something that needs to be taken into account in any thermal based limit?

DR. GORDON: Yes, definitely.

DR. MORRISSEY: If there are no more questions, we are going to break out into our breakout groups now. Chris, that was a great talk, and very interesting. Let's give Chris a round of applause.

We are going to have -- the breakout on the cardiovascular system will be here. We will take a minute or two to shift over. The rest of us, we will march over to the Rockville Potomac Room. I would like about half of us to go to the other room so we balance this out and have a good discussion in both areas. For the participants as well, about half of you will follow over to the Rockville Potomac Room and listen to the session on testes. Here we will have a session on cardiovascular.

(The meeting recessed at 11:10 a.m. to resume in breakout sessions.)

 

 

 

 

 

A F T E R N O O N S E S S I O N (3:15 p.m.)

Agenda Item: Session 3: Immune System

DR. REPASKY: I come to this meeting as a member of the Society for Thermal Medicine, and several other individuals here including Gerard van Rhoon and Mark Dewhirst and some others. Our field has been propelled in enthusiasm over the last decade, because finally after a very slow start based on difficulty in trying to heat tumors and the physics of trying to understand how to best heat up tumors, a lot of major problems have been solved in terms of temperature measurement, and non-invasive thermometry is on the horizon now, and it is getting better and better in terms of learning how to use heat, and in particular using heat to maximize the effects of radiation and chemotherapy for cancer patients.

What I have here is a list of not phase I and phase II trials, but phase III trials, which have shown significant positive survival benefits or benefits not only in overall survival, but time to recurrence and a variety of other issues. There is data here now that is really starting to give everybody a sense of momentum for trying to figure out what is going on with heating tumors during cancer therapy.

I also want to indicate that it is not published yet, but a new phase III trial has been completed. Ralph Issels and Gerard van Rhoon is a part of that as well. This will be the largest sarcoma trial of any therapy that will be published very soon as the data is complete, showing a significant survival benefit. Particularly while the protocol here is not what we used in the States in terms of what we use in hyperthermia and radiation for advanced sarcoma, the local control that was demonstrated in this trial was really impressive to everyone.

Why does hyperthermia enhance radiation in chemotherapy? There is a lot of hypotheses. From my point of view, I think there is good evidence, we don't know yet, but we are accumulating evidence that in part it might have to do with a potentiation of a long term immune response against the tumor, which helps in provoking inflammation. The heat may be providing some type of stimulus, co-stimulation, a danger signal, which in turn can help stimulate a long term anti-tumor immune response.

The work that we do in my lab focuses on a variety of different end points by which the immune response is activated to control tumor growth, including changes in vasculature and vasculature profusion of tumors, which heat does very well, all the way over to release of soluble factors such as cytokines like TNF-alpha, IL-1, IL-6, proliferation of effective cells, and then changes in the tumor itself that when heated makes them better targets for NK cells and CD8 cells.

I will be talking just about one small proportion of this work today. What I wanted to indicate is that the range of temperature that we are using in my lab is below 40; 39.5 is the maximal temperature that we use, in part because it was the simplest temperature to use in a mouse, and that is where we see the best effects in terms of immune responses.

We get our rationale from some of the figures that you saw earlier in the beautiful introduction to some of the work I am talking about from Chris Gordon's talk this morning, where he elegantly talked about how thermoregulation is so fundamental to biology. I am very fascinated. This is the same figure he showed. I put it in color this morning.

The gradients that normally exist in humans are not appreciated well. Even in the hyperthermia field, when we dump heat into a tumor, you can see that depending on where this tumor occurs in the body, the extent of change of temperature in that tumor is going to be very different. But basically it is all about the brain. The brain needs to be at 37 degrees. You need a heat sink, a significant heat sink, to move heat, in case you are exercising and getting up and moving. That is why the rest of the body is significantly lower, particularly out here in the hands and the feet.

What is interesting about fever is not so much that there is an increase in temperature. By the way, the increase in temperature out here in the shell is enormous. If the basal temperature is 30 and you are going to 40 degrees, that is a very big shift compared to the change of temperature in the core. So I think that the magnitude of the change that we are talking about in fever can be quite significant.

But in addition to the increase in temperature, I think what is the most interesting aspect for me is the disappearance of these gradients. That brings up another point that was mentioned this morning.

These gradients are maintained actively by thermoregulation neurons and receptors for temperature that differ in their concentration and sensitivity to different temperatures throughout the body. So if you have 37 degree blood coming from the core out to your skin, which is normally 28 degrees, it feels very hot, even though it is only core temperature, because the set points for these all differ, and that is what keeps temperature different in different parts of the body.

That entire gradation is lost when you have a fever, and it is maintained that way for hours, sometimes as long as six to 12 hours, before body temperature falls again and these gradients are re-established.

But even under normal circumstances, even when you don't have a fever, most of the organs in the body stay put. So they are bathed in an environmental temperature which is relatively constant. But the immune system, particularly lymphocytes, have to migrate through these gradients day and night, and then experience a pretty significant thermal shift, as does the rest of the body.

So it brings up a lot of interesting questions as to what these gradients do to cells, how do lymphocytes even know they are in different parts of the body, is it a temperature dependent recognition that tells a lymphocyte whether it is in the skin or in the core.

These gradients are enormous. I also have pictures from older literature that I didn't change, but this is a very important picture from my point of view. This is telling us something important about how does the liver heat to tumors.

What you are looking at here is hand blood flow, out here in the tip of the finger. Within seconds, you are looking at time in seconds when heat is delivered to the trunk of a human. Here is the increase in temperature out here in the hand within milliseconds of a change in temperature in the core. So that heat is gotten rid of very, very quickly, and that is why we need these significant heat sinks.

So this to me seem to be a very sensitive way of telling a clinician whether or not a tumor has been heated that is in the core, simply by looking at heat release at the fingertips. It would simplify some of our problems in knowing tumor vasculature or the temperature.

This is an even more impressive experiment. Here there is an infusion of warm water or saline in the left ear, and you see the temperature go up. What you are looking at here is left finger heat elimination. Again, within seconds of a slight change of temperature going into the brain you have a very rapid release of heat out at the periphery. So it really is about maintaining brain temperature for reasons which are not entirely clear to me. The rest of the body can be much cooler, but it serves as a very rapid receptacle for the release of heat whenever there is even a small change in brain temperature or core temperature.

I like to collect pictures from nature in terms of identifying for the classes I teach up in Buffalo, about why temperature matters, why should we care about it. This is the one I picked for my lecture today.

I don't know if anyone owns Siamese cats; I don't. I found this to be a pretty fascinating story. When they are born they have no color points at all, they are all one color, because during gestation they have been exposed to the mother's temperature, which is relatively warm. But after they grow up and these gradients are established pretty quickly, they have a mutation for pigment. Whenever anything makes them warmer, such as gestation or fever or becoming overweight, they get white again. They lose that pigmentation. So these gradients of temperature tell me something about how very mild changes in temperature are able to influence gene transcription and protein production, and it is reversible. Very subtle things change, like their weight or whether they have a fever, is certainly enough to change a significant protein production. This is affected by the level of the gene.

Fever is interesting, because we still don't understand why it happens. But it happens not only in humans and mammals, but throughout nature, including the cold-blooded animals. What they do, as we heard about this morning as well, they have to move to a hot spot. They feel very, very cold as a result of infection. But nevertheless the overall end point of body temperature elevation is something as distant from us as fish is very similar, in terms of the overall shift in temperature as things like mouse and dogs, which, like we heard about the guinea pig, dogs also run a little hot, but human, mouse and pigeons are also hot as a bird.

Febrile temperature is relatively constant throughout nature, whether it is generated from hypothalamic, shivering, metabolic changes as in humans, or whether or not you move to a hot spot. This is a very powerful response to nature.

Does it matter? It does. These experiments are not easy to do in animals. We keep our mice in all IACUC approved institutions at relatively cool temperatures. As we heard from Dr. Gordon this morning, if you let them choose where they want to stay, mice choose invariably 28 to 32 degrees Centigrade. All IACUC institutes make us keep our mice at a maximum of 22 to 23 for some good reasons which we can talk about later.

But if you try to measure fever in an animal that is mildly hyperthermic, you don't see it, because their temperature is below the normal thermic temperature. But in any case, there are ways to do these experiments. These investigators in Jeff Hasday's lab house these animals at two different ambient temperatures, allowing one group to get a fever following a bacterial infection to the gut, and another group to have normal thermic or slightly lower temperatures. There is a significant difference in survival with no treatment, simply by allowing animals to warm themselves up.

Out in the ectotherms, this response is even greater. This literature comes from the '60s and the '70s. A large amount of it was done by Matt Kruger. This data is so fascinating, because these animals really do need to seek out very hot temperatures as soon as they feel sick. This data is remarkable; 34 degrees Centigrade is hardly cold, but the survival of lizards given this experimental infection is very poor if they are allowed to be at 34 degrees, compared to 40 degrees. If you allow them to choose where they want to go instead of just sticking them in a temperature, they all choose 42 degrees. That is where you get the best survival.

Does this matter to humans? I think it does. Everybody in this room has had this experience. This is my daughter. When I came home from work she had been let out of school early because she was sick. I don't generally allow the dog or the cat on our beds, but she said, I am sick, I feel freezing, and would I let them stay there because they keep her warmer.

I took a picture of her, because I thought this is what I need to remember, that humans don't get fevers only by this natural internal increase in temperature. We also inherit a very significant behavioral response. I think most of us in the room have experienced the symptom of feeling cold. You think, I am getting sick, you want to go sit in a warm room if you have one in your lab, or go home and cover up with a bunch of blankets.

That turns out to be about 40 to 50 percent of the supportive energy you need to help increase temperature in fever conditions, even in humans. So when people talk about creating a fever and looking to see what happens, it is very complicated. How do you add the heat to create a fever? It is hard to ask mice whether they feel hot or cold, and you have to use a combination of behavioral events plus biochemical injections of either something like LPS or bacteria, and hope that the mouse is warm enough to sustain these large amounts of energy that it takes to raise body temperature.

How do we do it? What are the questions we are asking? We have chosen the most simple question, which is simply to take lymphocytes and look ta them at various temperatures to see how they differ, and whether or not those differences affect their ability to enhance the killing from infections.

I think there is no doubt that an increased body temperature is necessary for improved survival, through a large number of studies. But whether it is immunologically driven has not been very clearly established. There are a lot of different end points you can look at if you are looking at immunological events. We have looked at many of these.

What I am going to focus on now is data on CD4 cells. Most of this has been replicated for CD8 cells, which is the cytotoxic arm of the T cell response.

First of all, I would like to let you know that the plasma membrane fluidity itself is not just sensitive to heat shock temperatures. This has been studied before. What this experiment does is, it takes a look at temperatures between 30 to 42 degrees. What you are looking at here is anisotropy using DPH. So we are looking at the energy and the movement of these molecules as a direct function of membrane fluidity. Membrane fluidity here is increasing as the points go down.

So as you increase temperature, membrane fluidity increases over this physiological range of temperatures. This is looking at fresh CD4 cells from people or mice, and a Jurkat T cell line, which gives us the larger number of cells we need for biochemical studies.

Another interesting thing that was discovered from these experiments is that untreated cells have a given fluidity. Anti-CD3 and anti-CD28 are antibodies which are used to mimic the T cell receptor ligation and co-stimulation that we use to stimulate how T cells see an antigen presenting cell. It is known from the literature that ligation of these molecules does increase fluidity, so that made sense.

Anti-CD3 alone, that is a triggering of the T cell receptor or signal one, has no effect on fluidity or on activation, but anti-CD28, which is a very important co-stimulation signal two that is needed for most effective activation, did increase fluidity, and so did heat. So heat and co-stimulation through a known molecule both provided a similar amount of an increase in fluidity.

What could that mean? When you look at what goes on in the plasma membrane of T cells that are resting, and we are using as a readout the production of a cytokine, IL-2, very important, very potent cytokine that gets all of the rest of the cells with the immune response going, you don't get any IL-2 in resting cells unless they are activated, and that is important.

But also what you see in resting cells is something known as lipid rafts. These are particular membrane domains that differ because of the amount of cholesterol and other saturated and unsaturated lipids, and the presence of important signaling molecules, including the T cell receptor complex, which is here identified by CD3, and some other signaling molecules.

What happens upon normal activation is that membrane domains cluster. You find these much larger rafts that are thought to present a platform for the T cells to interact now with an antigen presenting cell, which would approach here bearing antigens for the T cell receptor.

The reason why rafts may form -- and by the way, this is not a fluid region, this is a much more highly organized region -- is to cluster signaling events which tell the cell, go ahead and make a potent cytokine. Antigen is here, so activate and make cytokines. Under normal circumstances you see significant clustering of lipid rafts upon activation.

Just to summarize very quickly, lipid rafts are distinguished based on the types of increase phospholipids, single lipid cholesterol. We know that they are important signaling molecules that are also clustered there. Luckily, there is a reagent out there, cholera toxin reagent that you can buy that has a fluorescein or some type of marker attached to it that attaches very specifically to GM-1, which is a component of lipid rafts.

What I am going to show you now is cholera toxin binding to GM-1 as a marker for what happens to lipid rafts when cells are activated or heated.

We know that fluidity changes in bulk fluidity, but now looking at individual membrane components and lipid rafts, what do we see? Just to tell you again what happens normally, you have these domains that are separated, that upon binding of the T cell through the T cell receptor, here in our experiments will be marked by an anti-CD3 antibody, you can see where the MHC and peptide is coming in from the antigen presenting cell.

Another important part of the story is the role of signal two or CD28, that usually binds its receptor on the antigen presenting cell, or B7.

When those events happen by reasons which we don't really understand, a major reorganization of lipids occurs. You form these visible by eye clusters that you can see by GM-1 staining. When that happens through signaling, IL-2 is produced. If CD28 is missing, and there are nice knockout animals for CD28, or if you don't add the antibody or signal two, IL-2 is not produced.

These are some of our earliest data on temperature. What we are looking at here is CD8 cells or CD4 cells, and we looked at some different temperatures, 33 to 39.5, and then just simply look at them. This is what resting T cells are thought to look like, this is what the literature says they look like. This is what heated lymphocytes look like. It is a very obvious change in the clustering of lipid rafts in the membrane.

Again, the effect is seen at 39.5 and also at 38.5, but we didn't look at 38 or anything lower. It is not 100 percent of the cells. Some cells of course have aggregates to begin with because a low level of activation is always going on, but the effect is very obvious when you look at these cells by fluorescence.

I think the previous slide was CD8. I wanted to show you both here. This is CD4, this is another experiment at 37, diffuse staining of phospholipids, 39.5, clear aggregation, very similar to what happens if that cell has been activated.

You don't have to count these by eye. Because of our purchase of an image stream flow cytometry at Roswell Park, we are able to take a picture using flow cytometry. So you can look at millions of events, put your cursor on any one of these dots and a picture of that cell comes up. So it allows you to quantify and get past the bias that has always been true of fluorescence in pictures. You take one picture, no one knows whether that is true of the whole field or just that cell.

Here you can see that one photograph corresponding to every single one of the counts we evaluated has a diffuse or an aggregated phenotype. So it is data like this that is really convincing us that this is a pretty global and perhaps non-specific effect now. Without adding antigen, nothing happens except this reorganization. I'll come back to that point in just a moment.

It is temporary. It takes about two hours to see a significant effect, but it peaks at six. So there is no point in heating longer than six hours. I don't know whether two or four or six differ in terms of function, but certainly if you heat longer there is no benefit.

What is interesting about that duration of time is that that is pretty similar to what a fever is, if you were to extrapolate this back to the heating exposures that take place in most animals. It dissipates very quickly. If you take cells after they have been heated for six hours and put them back at 37 degrees, those aggregates are lost. So within two hours you are back to the control level of aggregation versus diffuse staining in these populations.

Now, what about looking at immune function? We know heat alone does something to change the appearance of plasma membranes of lymphocytes. So what I want to do now is to show you a biochemical analysis of the lipid rafts.

We use GM-1. We make density gradients of the lymphocytes, either heated or not, and separate them out based on their density. Generally in the literature and in our lab, the raft fractions usually are anywhere from two to four. Again, GM-1 can be nicely marked as a raft marker using cholera toxin.

The non-raft fractions are more of a diffuse group, but we try to stay as far as we can out here. We use markers like CD71 which never go to the raft or CD45 to tell us for sure we are in the non-raft regions of the domain.

What you are looking at here is resting cells, no activating signal like antigen. What you are looking at is a fold change. It is a mild but reproducible increase in several molecules, including Lck and LAT and CD3 and CD28. Zap70 has disappeared, but it doesn't change in this analysis. Just like we see by eye that there is clustering of lipids, there is also a measurable increase in the concentration of some proteins that we would call raft associated in response to temperature alone.

This is the important part of the story. Now we are looking at IL-2. I told you that under normal circumstances you need signal one and signal two in order to produce IL-2 from these cells. So with comparing 37 and 39, and we are now looking at only CD3, we should expect no production of IL-2, and we don't unless you really use huge amounts of signal one, and then the green finally starts to produce IL-2. But this is extremely high from an immunological point of view. But down here at zero, with neither heating or not heating, we don't get IL-2.

I think I think is important, because if you went out and laid on the beach or if you exercised in the fitness room, you don't want lymphocytes making IL-2. So I am happy by seeing that there is this level of control. You need also for the cell to think that there is an antigen present as mimicked by CD3 antibodies.

This is a cell line, Jurkat. These are human peripheral CD4 positive cells where the effect is even more enormous. In other words, our own cells are more sensitive to this heating effect than the cell line which has been around for about 20 years, but still, the cell line definitely shows a significantly increased production of IL-2 without signal one. But it is even more evident using fresh peripheral blood from humans.

This is ELLI spot assay. Now we are using just different ratios of signal one and signal two. Under every circumstance there is more IL-2 produced even with suboptimal amounts of stimulation with both CD4 cells and both types of cells.

What about the IL-2 gene? The answer is yes. The mild thermal stress enhances IL-2 message production. You can see that here on the gels. We were also able to do RT PCR. You can show that with signal one alone, with no signal two, there is no message produced saying IL-2. However, if you use heat plus signal one you get IL-2 produced. At 37 degrees which is the green bar, it is very suboptimal because there is no signal two provided. You haven't added any anti-CD28. So the real question we have is, does this temperature shift substitute for the need for signal two in the body.

We are not really sure yet. This is very new data. But what you are looking at here is a biochemical analysis of control cells without any treatment at 37 degrees. You get so much of the protein inside of the lipid raft. This is 37 degrees, in which we have added both signals, a complete immunological signal, signal one and signal two. This is the fold increase that occurs under optimal activation conditions using these antibodies. You can see that if we only have CD3 alone, no signal two, that we get a comparable amount of these proteins in the lipid raft, suggesting that this temperature shift from 37 to 39 which again is very small in terms of where we could go in the body, we could look at shifts as great as 30 to 39, because this would be relevant, most of your immune system is way out in the periphery. Until they become activated, when they migrate to a draining lymph node or someplace deeper in the body, also experiencing a temperature shift.

The most conclusive experiment with the use of the CD28 knockout animal, and this is just a characterization that we don't have in CD28 in these cells. Now what we are doing is tasking CD3 alone and to our great happiness -- remember, there is no CD28 in the animal, so adding heat gave us a great response of IL-2. Anti-CD28 alone, heated or not heated, had no effect, again good because we have never been able to demonstrate a heat effect alone on IL-2 production. There is measurable changes in the physical positioning of molecules in the membrane, but it doesn't go on to release IL-2 non-specifically.

Then here is a very important experiment. We added both antibodies, expecting that CD28 would have nothing to bind to. Yet without heat you still get no effect, even though we have added the anti-CD28 antibody. There is still absolutely no effect on IL-2 production over control levels, and we get a pretty substantial IL-2 produced in this animal model.

I am going to skip to this slide because I am running short of time now. What we are doing right now is looking at 2D gels and doing mass spec in heated and non-heated cells, where we can look at proteins that are either expressed in a higher level after heating or proteins that are expressed after a lower level after heating.

This is work that Sara Beachy who is here, established while she was a graduate student in my lab. These are the spots that we cut out for our first analysis of protein in the lymphocytes which increased after heating.

These are some of the data we are getting. To our happiness, most of these turn out to be cytoskeletal proteins, which people know are brought to the membrane with real activation of lymphocytes. But remember, we are only using heat, we are not adding any other antibodies.

We were also interested in the fact that we are finding several heat shock proteins, because they are induced even under these mild conditions in the cell, which do migrate to the membrane. Then the one that interests us the most right now is this mitochondria protein. It is a stomatin-like protein. The reason why we are interested in this particular protein -- remember, this is the effects of heat -- is that this particular protein -- people know that mitochondria do migrate near the surface of the cell upon lymphocyte activation. It is thought that they provide some of the great energy that lymphocytes need to experience the proliferation that they need to go through to really deal with the invader. So we are very excited by the fact that temperature alone helps at least to bring some of these same proteins to the membrane that a co-stimulation like CD28 is needed at 37 degrees.

So our overall conclusion here is that the energy provided by increasing the temperature of these cells may help supplant the energy that CD28 is using or providing as a co-stimulatory molecule.

In conclusion, I would like to come back to a couple of points that were made this morning by Dr. Gordon. I have been using increased temperature. We have been talking a lot today about fever, exercise and just adding heat. I really can't agree more with what has already been stated, that these events are extraordinarily different in terms of their effect on an intact animal in terms of a whole body sense.

I use this slide, which is from Dr. Gordon's work, and it is published in a book that I brought here with me. But I love this picture, because I think it says a lot in a short space about why these things matter in terms of how we are going to study heating.

What we are looking at here is a comparison of what happens to mice in this case under normal thermic conditions. Here is mimicked a cold spot and a warm spot. Mice like it about 30 degrees. Everything is fine, nothing is bothering them. Their set point in their brain is the same as the temperature stimulus. You have some in blood flow and you have some metabolism, and everything is in balance. They like it a little bit warm.

Now what we are looking at here is what I do. What we do in our experiments is, we put animals in a warm box or we pick their cells out of the animal and put them in a warm box. Now the analogy is, the brain set point is forced to go up because we are changing it.

Excuse me, I skipped one. This is forced hyperthermia here. We are forcing the animal to get hot, but its brain set point hasn't been changed, so that is shown here. Now the animal acts as if it is really hot, and it is struggling to find a cool spot. You can see this behavioral response. It will move toward the coolest spot you provide. We have a thermal preference apparatus in our lab as well.

Here, the changes are enormous in terms of blood flow and evaporation. The animal is desperately trying to release that heat. It is a big, big stress. In comparison to regulated hyperthermia, which is this febrile event, where now the set point is elevated and slowly the body temperature is elevated, either by the animal moving to a warm spot or from shivering and a variety of other metabolic.

Here the animal acts as if it is cold. You do feel cold, you want to get warmer, and it moves to a warmer environment as well.

Finally, regulated hypothermia is another interesting event that occurs in a lot of nature, where the set point is suddenly dropped. The animal acts as though it is really hot, digs down into the coldest spot it can find to gradually lower its body temperature to enter the hibernating state.

What is shown over here though is how different so many of these metabolic aspects. What I am most interested in is metabolism and blood flow, how these things are different based on whether you have regulated or forced hyperthermia or hypothermia.

This is the last data that I will show you. Something that we are very surprised by is the fact that all this time, and this is at least 15 years of research in my own life regarding this topic, we always do experiments where we take tumor bearing mice and we heat them up, as we are trying to mimic what is going on in humans.

I always assumed that the control mice that we are not heating were at 37 degrees. As it turns out, as an animal grows a tumor, its body temperature falls. That was very late in coming for me, and I was stunned by that, because I always thought we were comparing hyperthermic animals with normal thermic animals. But as it turns out, if you keep the room temperature constant so they can't change the thermostat or doing anything else, as tumor growth occurs, body temperature falls. So in our hyperthermia experiments, we are not comparing 39.5 to 37. In our mice we are more like 35 or lower as these tumors grow. I think this has some implications that we are trying to get ideas on and follow up on.

One of the things we did was increase the temperature of the room. We changed the thermostat. In our institute we cannot exceed 23 degrees in our animal colonies, but we bought a chamber where we were able to change the temperature. Now we know that if room temperature -- that is how our standard 23 experiments are done -- yes, it still falls as tumors grow. But now if we provide some extra energy in the way of ambient temperature to the mice, their core temperature stayed normal. That became significant out here in terms of the difference in their temperature.

In other words, ambient temperature can provide enough energy so these animals aren't losing so much of their body heat as tumors grow. But what happened was, when we compared those two groups, we found a difference in tumor growth.

These are the animals whose tumors -- and remember, there is no treatment here, this is just B16 melanoma cell line growing in C57 black. This is CT26 growing in BALB. When we compared the same amount of tumors injected in the animals that are given ambient temperature support so that their body temperature is not falling, we find a significant slowing in tumor growth.

This is modest, in my view, the slopes are about the same, but it is a significant difference. It is equivalent to many of the immunotherapies that we are also testing in my lab at the same time, and there is a survival benefit.

So we think it is immunological. When you deplete CD8 cells, there is absolutely no difference in tumor growth based on ambient temperature. When we use SCID mice that have no T cells at all, there is no difference in tumor growth, and if we deplete NK cells. NK cells and CD8 cells in immunocompetent mice play a very important role in controlling tumor growth. So it looks as though the effect is immunological.

I'll stop here. If you want to hear more about a lot of biological work on the effects of hyperthermia, I have to take the time to let you know about the next meeting of the Society for Thermal Medicine, which will be in Clearwater in April. We have a great lineup of keynote speakers. I think it will be a wonderful meeting. Chris Diederich, many of you know, is a physicist and is being awarded a special recognition at this meeting.

To come back to this tumor growth, what we are saying here with our ambient temperature is, it is not so much the temperature of the animal that mattered, but it was the supplanting of energy that appears to be reflected by temperature which mattered. Of course, this is all new data, and we are moving ahead on this as fast as we can, but it looks as though tumor growth creates a stress that temperature can help with, in terms of giving the immune system what it needed to help control the tumor growth.

The most exciting thing we could do of course is to now combine ambient temperature support with therapy, like radiation or chemotherapy or immunotherapy. That is what we are doing now. All of the work that I presented has been done by graduate students in my lab.

Finally, I understand from talking to our editor-in-chief here, Dr. Dewhirst, that many of the presentations here, including a paper that Sara Beachy is putting together on our immunological survey of the literature on hyperthermia, will be part of the special issue. So I would like you to keep your eye on the journal, and submit papers to our journal when you can.

Thank you very much for your attention.

DR. MORRISSEY: Thank you, Betsy. That was a very interesting talk. I am going to open it up to questions, but I have one first, and I get to ask first because I have the microphone.

You showed a great talk on the effects of heat and how you can stimulate immune cells. Certainly you showed that very well, the signaling molecules, the IL-2 production.

My question is, and you had placed the emphasis on increasing the robustness of the immune system. Turning that around 180 degrees and pointing it more in the direction of these RF standards, there is a lot of diseases that are due to autoimmune diseases hypersensitivity. We are talking about thermal influences on the immune system, from asthma to multiple sclerosis, to a lot of diseases that come up in the RF literature. Is it mechanistically or practically within the realm of possibility that increased local or whole body heat deposition could increase, either by the increased migration of some of these innate immune cells or the active immunity cells -- you know what I am getting at.

DR. REPASKY: Yes, I do.

DR. MORRISSEY: Go ahead.

DR. REPASKY: I absolutely concur with your thought. From our own data we have published that mild thermal stress, using our mouse protocols I didn't talk about, can block the onset type one diabetes in the NOD mouse for diabetes.

Now, we have not looked at the effects of active disease, but it is as good as the gold standard's complete adjuvant treatment of those animals. What the mechanism may be there is indirect. We know that heating activates NK cells, which I couldn't talk about today, but that is in our literature.

NK cells are what controls the out of control T cell that is destroying the islets. So they are thought to be defective in that model. So we thought that heat would provoke the NK cells to control better the autoimmune T cells.

We certainly got the effect of blocking the onset of the disease. Whether that is the mechanism or not, we are not sure. But I can also point you to thousands of years of experience that people have with rheumatoid -- not osteo, but rheumatoid arthritis and the use of heat has a lot of beneficial effects.

So I can't say much more than that. Our own data currently supports the use of heat in controlling inflammatory disease. I know of some unpublished data regarding IL-6 and some other inflammatory processes from my colleague at Roswell Park, Sharon Evans. She is doing a lot of great work on the role of proinflammatory cytokines in promoting T cell migration to tumors and things like that.

DR. MORRISSEY: Obviously you haven't gone down the very small increase in temperature. 39 is a low temperature to you, but in our community that is not. What would you hazard to guess might be a threshold for stimulation of the immune response?

DR. REPASKY: I can't guess. I can just look to see what is in nature. It is interesting to me that there is such an enormous conservation of the febrile response temperature. That is one to five degrees Centigrade, so it doesn't seem -- within that range may be the optimal immune provoking aspects of fever, if that is what happens.

DR. MORRISSEY: Any other questions? Thank you very much.

DR. DONALDSON: I just wanted to get right, the first thing you were talking about when you were increasing air temperature and pouring warm water in the amygdala, response in the finger. I would suggest to you that that is a reflex response, because I don't think the circulation -- the blood won't circulate in that time.

DR. REPASKY: You're right, I probably mis-spoke. But there is a need to get rid of the heat, so there is lots of heat that is simultaneous with the addition of heat.

DR. DONALDSON: The other thing I was going to put to you, do you think it would be a good idea to actively warm patients who got some infection? Some published literature, infections were associated with a fever and some aren't. I was just wondering whether you can shorten the duration of the symptoms.

DR. REPASKY: I think that has been done. I think people have published a study like that, especially in Europe. Another way your experiment can be done was also done in Europe, where they measured the length of time people stayed in the hospital, if they were given antipyretics or not. It is very significant, release from the hospital much earlier if you don't treat the fever.

So it is hard to come by data in humans. Your experiment and the one I just mentioned have both been published. So I think humans have a lot of overlapping mechanisms in increased temperature compared to a lizard, where it is almost entirely a behavioral -- it is a little bit of metabolic change.

So it is easier to do those studies in exotherms. But I think there is enough data in humans to suggest a benefit.

DR. DONALDSON: In your mice, normally you might get a suppression with the release of the inflammatory markers.

DR. REPASKY: We don't see them during the course of the heating. We warm the cages, they have water, but we don't put food in the cage during the heating. Then when we take them out of the heating, of course they have food again. We have not done what you suggested, but changing an ambient temperature though is a well-known procedure for changing the appetite of animals. So if you put them in a warmer room they eat less. Our animals, the ones we all use, get fatter because they have to eat so much for thermogenesis. They have a higher metabolic rate. They need to stay warm. That is why you have to have five mice in a cage and a lot of bedding, because one mouse alone under the conditions we keep them in would become more hypothermic.

DR. DONALDSON: And it is not because the tumor suppresses their appetite, and therefore they have --

DR. REPASKY: Yes, but that is beyond what I showed. The animals hunch in the corner if their tumors are too big. I would get a yellow card and get in a lot of trouble if we had a tumor that grew to that size.

At the time we were looking at these experiments, we stopped them before the tumors are a centimeter and a half. They are not eating, but it is close. The end of those experiments is the last point we would keep them, because they do stop eating.

DR. DONALDSON: Thank you.

DR. DEWHIRST: This is just a quickie to point out that this particular issue of the Hyperthermia Journal is a spinoff from a Japanese conference held a year ago on non-oncological uses of hyperthermia. Just to point out the fact that the Society for Thermal Medicine is not just cancer. We are extremely interested in people from this audience attending our meeting. We are expanding far beyond the idea of using heat to treat cancer. We are looking at the whole biology of thermal medicine. Clearly the subject of this meeting today falls right in the middle of that. Thank you.

DR. MORRISSEY: Thank you, Mark. One more question, Betsy, to try to bring this around to the interests of the standards groups. Incredibly interesting stuff.

What adverse effects on the immune system could you see? This could be either direct on the immune system or like we talked, indirect by triggering autoimmunity. Is there anything from a standards perspective that we need to worry about, and what are the thresholds for this?

DR. REPASKY: We did uncover one. It is still unpublished, so I think this is correct. If you have sepsis, if you have a release of bacteria from the gut, for example, it is not a good idea to heat to the levels we are heating. The animal is already struggling to develop a fever.

I think there could be an inflammatory state that should not be heated further. I can't really say much more than that. We have had bad luck in doing some infection models. We may have some bad effects too if you have a lung infection.

Jeff Hasday, whose work I cited here in the paper, I'm pretty sure he has a paper out on lung inflammation being made worse by hot room temperatures, but not necessarily hyperthermia in mouse.

DR. MORRISSEY: But it would be an over stimulation and cytokine shock, I guess.

DR. REPASKY: An over stimulation of proinflammatory cytokines. The one that was an accident we followed up on, we were growing colon tumors in the colon. We were trying to do an orthotopic model. We ended up with some animals that the tumor had grown too big. So we are pretty sure there might have been leakage from the GI tract. When we heated those animals, half of them died.

So it is hard to tell if the experiment was done so poorly or we shouldn't be heating them, but I would say there might be some circumstances where you should not heat.

DR. MORRISSEY: Thank you. Very interesting that you can get that stimulation. What comes to my mind is these autoimmune and hypersensitivity diseases that could be augmented.

With that, are there any more questions? With that, we are going to have Rob McIntosh. Rob is the last speaker of the day. He is going to talk about correlation between SAR and temperature in different tissues.

Agenda Item: Session 4: Energy Deposition and Temperature

DR. MC INTOSH: Thank you very much, Joe. As someone who does mathematical modeling and sits there in his office, tapping away each day and just working on models, I very much appreciate coming and participating in an environment like this, one step closer to the real world. I don't know how much closer it is, but very much appreciated. So thank you to Joe, even though he has just left the room. Also thank you very much to the MMF and GSMA for support of this workshop, and also a lot of the work that I will be presenting here has also been supported by MMF and GSMA.

In Australia the temperature yesterday was 110 degrees Fahrenheit in Melbourne, so I think my wife and family are suffering from a bit of hyperthermia today. I much enjoyed being in Washington, D.C. yesterday, touring around. I much prefer to suffer from a bit of hypothermia rather than hyperthermia.

A lot of this work that I do is with Vitas Anderson. There is a picture of Vitas on the right. I think a lot of the heat up there is due to maybe a bit of the beer that we has been drinking as he is also trying to cope with the hyperthermia in Australia.

Modeling. A lot of good techniques are done with tools such as finite element modeling, molecular moments, finite integration theory, boundary elements and the like. But the modeling that we use that I will be presenting here today is basically finite difference modeling.

Each of those techniques has its pluses and minuses. One of the pluses in finite difference modeling is that it is very good for heterogeneous objects. You can discretise all the different types of tissues that create little cells. So for example the tissue there, the pink color, that represents muscle. You have the brain up here, white matter, brain matter, cerebral spinal fluid, skin up there in the brown, and the like, or it can be modeled in finite difference modeling.

The finite difference modeling we use, we basically use the two different aspects, electromagnetic modeling and thermal modeling. Joe Morrissey has given me the mandate to look at the correlation between the two as a bit of a primer for this workshop, so we will see how we go into that.

This model and the models I will be using are the Norman and Naomi models, have a resolution of about two millimeters, courtesy of the U.K.'s Health Protection Agency. There are a lot of other very good models out there. The visible human, John Serecks in particular has helped us a lot. The other people in the research community are providing us with the visible human. There are a lot of good Japanese models out there at the moment as well, Japanese institutions have provided some good models. There is a virtual family as well, very high resolution models that are coming and just being released into the international research community.

With those models, when you discretise them and you set up the cells, you need to assign the properties. I am going to focus a lot on the types of properties that we were assigned.

With electromagnetic modeling, it is the electroconductivity and the relative permittivity that you need to assign to these tissues. A lot of that work, the measurements for that were done by Camelia Gabriel and Sammy Gabriel. There is a database, I believe it is on the FCC website, which provide the data for that. So there is a well respected set of data.

Unfortunately, there is still one set of data -- a lot of work in this area is lacking from data, so again, not a great reason to come and participate in an environment like this as a modeler. The more input and more data we can get, the better for the modelers of the world.

I put up some parameters there. Down on the bottom there the density, which is also useful, especially when you are working up SAR and water content. It is not needed directly, but if you are lacking in parameters like thermal conductivity specifically capacity, you can use water content to give you some gauge as to what those types of values are going to be. There are formulas that are provided, and researchers have put out papers giving relationships between water content and other properties, not just with thermal, but also for electromagnetic modeling as well, papers like Speld, and they are provided in books like by Duff and the like.

With the electromagnetic modeling, I won't focus on that much here. The idea there is -- well, I should state that the product we use with finite difference modeling is XFTD. The basic idea there is to calculate the electric field and then from that, that gives you the quantity of specific energy absorption rate, which gives you an estimate of the energy absorbed in the mass.

That is SAR. There is a definition of SAR given they are related to the electric field. There is a picture that is going to be a typical picture that I will use for the rest of this presentation, using the Norman model. We just use the example of a plain wave propagating towards the front of Norman's head. Many of the examples I will give is just one gigahertz. Of course, there are many different types of scenarios. So that is SAR.

Thermal modeling. It is a lot more complex, especially with all the thermic effects that go on in the body. One of the very key ones is modeling of blood perfusion. There are different approaches. The simplistic way of doing it is through the approximation afforded by Pennes' bio-heat equation, where you don't include the vascular structure; you assign a property to each tissue and treat the blood flow and the way it dissipates out the heat, and it is a heat sink term.

There are a couple of models there, the Hardy-Stolwijk model which I will talk about more in a moment, fairly simplistic, but it is a very useful method. There is a model over there on the right from David Nelson, who again worked closely with John Seriacs and the people at Fox Air Force Base. There is the visible human and David's very fine resolution on that one, split over different computers to try and get very high definition calculation of the temperature.

I won't say too much about the vascular models, although it is a very detailed and very promising and very thorough way of modeling the human body. I'll mention Lagendijk, who has done a lot of work in the hyperthermia cancer treatment area. As I said, there are four treatments of the vascular system. There are more papers coming out using more vascular based models, although the numbers still aren't right. Most researchers in the field, I would say about 80 percent or 90 percent, use the simplistic method of Pennes' bio-heat equation.

Pennes' bio-heat equation as I said uses each tissue as a sign of a general heat sink term. The workshop for example I went to, just discussing this with Ron Petersen a few minutes ago, we were in Paris, and the researchers there were of the opinion that Pennes' bio-heat equation is quite sufficient for localized exposures at low levels. That includes levels that you would consider if you looked at assessing a mobile phone at the maximum possible levels.

Having said that, with Eleanor Adair and her excellent work, she used the Hardy-Stolwijk model, Eleanor and her team, and the volunteers she used for her measurements. They looked at high exposure levels, and they also used the Hardy-Stolwijk model for the whole body and found good agreement with using Pennes' bio-heat equation.

Mathematically, standard heat equation. We are looking at transient solutions, looking at the change in temperature with time. It might be a steady state solution, but you can still solve it using a transient method. You see specific heat capacity, K, there in front of the special derivative, it is thermal conductivity. Then you can enter in the energy terms. For example, you can enter into this heat equation, the SAR term, and you can also add in -- if you want a fairly detailed model, you can add in terms like lung respiration as well.

Then there is metabolic heat production, and then there is the simplified use equation of B, which is the parameter associated with each tissue adding into this equation. I will talk more in terms of M, the blood perfusion and B, the density related heat sink terms.

Then as I said, most of our work is done with mobile phones, where you might have low levels of exposure. But if you want to incorporate thermoregulatory effects, then you can of course do that as well, add that into your mathematical models. Most of our formulation has been done by the work of Bernardi et al. He has quite an excellent paper, and it talks quite a good deal about how they have incorporated all their different aspects.

For example, sweating. When heat in the body reaches certain levels, the threshold is set for levels at the base of the brain and also levels set for when temperature reaches above a certain level on the skin. You can incorporate sweating, and Bernardi's model incorporates it by changing the boundary conditions of the model. We have a convection off the surface and then have a sweating term.

You can also have those thresholds as to when you get vasodilation or vasoconstriction. You can then have the blood flow there, a constant, and you can change the blood flow and incorporate that into your model as well.

Heat production apparently changes with temperature. So again, instead of being a nice constant factor, you can make it a factor of temperature and do things like that as well in your model.

I spoke about the Gabriel data for the electromagnetic modeling. With thermal modeling it is not quite a data basis for the input values for tissue properties and quite as nicely set out. When you look at the papers you will see a list of properties they have used; it might give specific capacity of such and such. Then you go through their papers, and a paper might refer to another paper, and then you will go, then it will refer to another paper, and that might refer to a book and the like, and you won't quite know where your source of all that data has come from.

As part of our projects supported by MMF and GSMA, we have put a lot of effort into creating a tissue database, and read 300 to 400 papers, and from that thrown away half of those. So we have got about 150 key papers and books that we found that actually have the original measurements. That has been a key promise of this database, to only have the database with the original measurements.

There are about 44 tissues in that database. The six properties are listed up above, giving us those 260 or so values. For each of those, we have got this Microsoft Excel spreadsheet which tabulates each of the references, providing things like averages and means, et cetera. It has got things like whether the measurements were used say on human or pig or dog or whatever, and also humans at rest.

We would like to, once this project is finished, like the Gabriel database, we would like to provide this open to the public, to be accessible easily by the public, in terms of the researchers in the community, make it freely available, certainly make it available for anyone who wants to change any of the tissue database values, certainly make it available for anyone who wants to enter data into this database, and certainly open to those who will have credit for doing that as well. Like the Gabriel database where you have your standardized set, we also would like to think that the standardized set could also be helpful for those in the thermal modeling field.

This is an interesting graph that we generated, just in terms of when measurements were made. In the modeling era, everyone has 69 computers and a dozen models and the like. Back in the '50s and '60 and '70s, everyone used to do real work, used to go into laboratories and make measurements and the like. Of the measurements that we have got in our tissue properties database, you can see this clear Bell curve, measurements back in the 1800s and 1920s and the like, but it is picking more up in the '50s and '60s and '70s, and degenerating into the modern era.

It is quite an interesting exercise in looking at these tissue properties, because you would think there would be quite a lot of these values. It is actually surprising, the scarcity of data. You would think for example the specific K capacity of PAT, you would think there would be a lot of measurements that would be available, but it took us a long time to actually find some.

The first one I found, I found this paper on the measurements for two pigs. That was the only paper that I had for quite awhile; such a common property. Then I found a paper recently which had values that were measured for humans. So if anyone would like to corrected me on this and say there are also these other measurements, I would love to hear from you. But generally this data has seemed to be fairly scarce.

We just did a test comparison. This is just a one-off test. It is not very rigorous in terms of having done against many authors. With the paper that I just spoke of with Bernardi, we just used exactly the same model, exactly the same formulation, and just used the values that Bernardi used as inputs into his model, and then we used the database values that we have in our model, just to see what sort of difference there was, and there was a 25 percent difference. This is not quite representative of what you might find on other papers, but at least it was a bit of a test case, to see what sort of difference in values you might get.

One of the key aspects also in the modeling has emphasized the blood flow. One of the main differences between our inputs and Bernardi's inputs were the values of blood flow. They could have been out by something like 50 to 70 percent. That is going to make an enormous difference in terms of your results from your models.

This is just assuming there is a human at rest. Of course, tissues are going to be very different in terms of blood flow when they are not at rest. For example, blood flow in a human at rest is apparently three minutes per 100 grams, using the unit that they like to use, whereas during exercise the value is more like 60. But also just doing the sensitivity analysis, preliminary data, we are varying the thermal conductivity K and blood perfusion M up and down by ten percent and seeing the results.

I just presented one example there, where the temperature change in the body, when we changed the blood perfusion, the temperature changed, delta T, changed by eight percent. So that seems reasonably significant. But then when you consider authors might have very different results with blood perfusion values up around 50 to 70 percent difference, you are going to get quite different results in your models.

That is basically the background. I would just like now to go on to the mandate Joe gave me, in terms of comparing and temperature and why we are doing that.

This should be familiar. Standards setting bodies use SAR as their metric because of its relationship with thermal effects. They acknowledge the relationship with thermal effects.

In looking at this correlation between SAR and temperature, I have got a plot here. Again, this is just an example. We have just done a similar model to before, change in temperature. I have plotted the SAR, one gram SAR and ten gram SAR.

I have spoken about how to calculate SAR, but in the safety standards and the safety guidelines, the SAR is not specified just as SAR, it is specified as mass average SAR. So for ten grams you take about 21 to 23 millimeters in length, and then you average SAR over that. The reason that is done is because we are trying there to match the diffusion properties of temperature, trying to average out SAR is that it matches the diffusion properties of temperature. We are trying to look at, if we are going to choose SAR as a metric, let's choose an average one so that it better matches up with the temperature change.

A lot of this work has been done by Professor Hirata. He has put out a very good paper on that subject. George Bit-Babik from Motorola did a very thorough analysis in 2007, comparing different sizes. I think he had one, three, five, seven and ten gram as his averaging. We have done a little bit of work on that as well.

In terms of which one correlates best to temperature, just SAR itself, there doesn't appear to be that level of smoothness. With one gram, as you start to average it and start to smooth it, you can see a little bit better correlation. With ten gram it gives you more of that level of smoothness. So the correlation there in a way does not suggest a high level of correlation, but in terms of that level of smoothness, it fits with change in temperature the best.

With people such as Professor Hirata, this is just a graphical representation of that. Professor Hirata has done things like, you take the data point from this and then you match it up with a data point to that, and then you plot them, and then you look at the correlation between the two, and the correlation is best at about ten grams.

As I said, George Bit-Babik found that the correlation was best at around seven grams, although ten grams is what he has used in the standard in the guidelines. Just looking at that idea in a little bit more detail, this time comparing that correlation between the ten gram metric and the change in temperature metric. One thing about this analysis is that when you are doing this analysis, if you take a point, you need to construct a little cube around that point before you perform that ten gram block.

You can only do that with points in the interior. These points on the exterior ribbon here, if you construct that ten gram block, it goes out into the air. So those points at that region of the head are not called valid points. To be able to get that data you need to extrapolate the values from inside the head to that external rhythm that goes along the side of the head. So here is this ribbon that is going up the head, where you are just using extrapolated data.

This to me is a little bit of a negative aspect of this averaging. It lessens the correlation between temperature and SAR. For one thing, temperature on the boundary, you have the convection effects, convection off the surface. So what you find with temperature is, if you draw a line from the inside to the outside, you get the low values. Then you get a bit of a peak here in terms of the change in temperature. Then as you get to the surface you get a lessening of the temperature values.

With SAR, SAR in a way has quite a different effect, just in terms of the general behavior of exponential dropoff from the surface into the body. So that matchup is not that wonderful right on the surface here.

To me, the ten gram size does give one of the best correlations between SAR and temperature, although George Bit-Babik is suggesting that perhaps seven grams is the best correlation. I feel if you were to have something a little bit smaller than the ten grams, something like seven grams, that might be a little bit better of a method. So I would just like to throw that into the audience. But you do need something at least of that order, at least of the order of seven to ten grams to get something of a correlation between SAR and temperature.

With this sort of plot, we are talking about correlation between SAR and temperature. Just looking at those, how well are they correlated? SAR is creating this energy which is giving this temperature change. Yes, there is a reasonable correlation between the two, but clearly you can see points in the body where the correlation is a little bit low. So I just want to delve a little bit more into that and look at specific tissues and try and dig a little bit deeper.

Before I do that, again looking in terms of this correlation between SAR and temperature change, I wanted to look at what happens with frequency as we increase frequency. I've got that standard model. With resonance, you are going to get resonance of the human body of about 40 to 80 megahertz, so 500 is not quite down at that level, but you are getting a little bit more of the fuller body resonance effect. In this case you are getting a higher SAR and a bit more resonance effect down at the bottom of the head, down near the jaw region.

At one gigahertz you are getting a bit more resonance effect happening in the head, so you are getting the higher SAR levels occurring up here in the nose and the front of the brain. As you increase the frequency, the depth of penetration goes down. At three gigahertz you are getting this bit of a ribbon around the front of the head, and then at six gigahertz that ribbon deceases in size quite a bit.

We have been investigating that as well, supported by the MMF and GSMA in terms of how far is SAR an appropriate metric. Our work and Professor Hirata as well has also been looking at this, suggesting that six gigahertz is about the limit for the suitability for ten gram SAR to be a reasonable metric. Three gigahertz is where you get the best correlation with the maximum level of correlation of SAR and temperature change, and from three up to six gigahertz where you are still getting a bit of a correlation, but after six gigahertz and from six gigahertz up to ten gigahertz, the correlation is better between power flux density, which is more a surface based parameter, and temperature change. Not a wonderful correlation, but it appears to be better than SAR in any of its forms.

So in terms of that correlation with temperature, how does SAR correlate with temperature at each particular frequency? At 500 megahertz, in terms of our models, we are seeing that peak occur -- well, in all cases it is going to correspond to where that high SAR level is. So that high SAR level at 500 megahertz was down near the jaw. At one megahertz it was up near the nose region, and at three gigahertz you had more of that ribbon around the front of the head. Then at six gigahertz the ribbon becomes thinner and thinner.

Then you are going to be getting effects at the higher frequencies, because it is becoming much more surface based. You are getting more of that convection of the temperature at the front of the head. The actual temperature change due to the SAR at the higher frequency is going to be lower. In fact, you are getting the higher temperature peak at the one gigahertz and 500 megahertz. I think some authors call that the hot spot region. I saw that referenced in a paper recently.

As I said before, that correlation between SAR and temperature change is reasonable, but I wanted to dig a little bit deeper in terms of each particular tissue. I have just got an example of about four or five different tissues.

I will start with muscle, not so much because it is one of the more critical tissues, but it gives a bit of an example. There is quite a range of all these properties for these different tissues, but I just want to simplify it by looking at electric conductivity and blood flow, just because they are two of the more influential of all those tissue properties.

With muscle, the sort of value you are getting is around about one S/m at one gigahertz, just as a rough number. Blood flow at rest, the value that we are getting from the literature is around about three, a fairly lowish value in this non-SI unit that was used. With those pictures, this is going back to SAR here, the peak temperature is occurring in this muscle region at the back of the nose here.

My simplistic summary for this case and other cases, if I can just say one S/m is the basic level, just call it moderate level in a relative sense, and low blood perfusion, especially when the human is at rest, then that is the tissue that gives you about the peak temperature change for the human body in muscle.

Skin, if I can just talk about skin for a moment, you will get all those convection type effects happening off skin, so it is a bit of a lesser effect.

I am not trying to give numbers here, because I just want to give a general feeling. But if you want some numbers, what I have got here is a field of five watts per meter squared as the standard reference field under ICNIRP guidelines, giving a peak ten gram SAR of .3 watt per kilogram and giving a peak temperature change of .1 of a degree. So if you want to follow numbers, then you can just look down at the bottom for the different types of numbers that are there.

Just picking on gray matter, I circled these tissues with these orange circles. It is a little bit hard to see, but hopefully it will present some picture.

Electrical conductivity, similar to muscle. Blood flow 81 quite high. To give you a feeling as to where those blood flow levels sit, kidneys from the literature database has a blood flow rate of about 600, stomach about 200, so those are the highest numbers that we found. So the blood flow in the brain apparently is very high, but not quite to that extent, so it is going to dissipate away a lot of the energy that has been absorbed in the brain.

The gray matter is characterized here by the orange cells. I just picked a couple of cells here. One gigahertz, it does penetrate fairly deep into the head. So even though this value position is quite deep in the head you are getting higher SAR levels with that high electrical conductivity. With that high blood perfusion however, you are seeing very low temperature change.

Now, of course this is just a mathematical model. This is a very simplistic view of how the body works. This is what modeling is like sometimes.

White matter. The electrical conductivity is lower. The blood flow is lower as well, a quarter of gray matter apparently, so the SAR is not quite as high, but you have got reasonable perfusion. The areas that I have circled there aren't really showing up there as much in terms of the temperature change.

Cerebral spinal fluid, electrical conductivity very high, blood flow negligible, so you are getting very high SAR levels next to the edge of the brain and the interior of the brain, with the lack of blood flow, it is a very low level, although it does show up to a small extent with this blue here. So again, a fairly minor temperature change.

Testes. We had a discussion on testes this morning and learned a lot about that. I was worried about the modeling that I was going to present, especially as the data is very scarce. Blood flow levels. I have always been led to believe over the years that blood flow in the testes was very minimal.

There was only one paper we found, Wax and Peterson, 1967, for dogs, and SAR values lowish. The testes tissue is in here, and the temperature level minor as well in that region also. So a basic model, but it may give you some guidelines.

I wasn't going to present the models that we have done, because we just don't have the resolution in our models at two millimeters. However, the authors, Peter Wainwright, Professor Hirata and his team have written about four or five different papers and done a lot of extensive research in this area.

In terms of the tissues, the electrical conductivity of the lens was about .62, which is quite lowish, although the electrical conductivity of the vitreous and aqueous humor is quite high. In terms of blood perfusion, very negligible for the tissues in the center of the model, although blood perfusion is very high on the periphery of the eyeball.

This might be a detailed model of those tissues, the ciliary body, the iris have still been modeled by those authors as just one tissue.

There are other factors involved that you also need to consider with mathematical modeling. The heat convection in this model, this is from Professor Hirata's paper, where he is looking at the heat convection of the skin in this schematic of the eyeball. H3 represents the skin to the air of about ten watts per meter squared degrees C. There is a different convection coming off the eyeball, much higher. It goes from 20 up to around 50, but I think it sometimes goes up to around 100 or so watts per meter squared, and that needs to be incorporated into the model.

Also, under high stress conditions you can have other things happening, like tears forming. The H value there, the convection off the surface of the eyeball increases, leading to a decrease in the effect of the temperature change. That is from a paper by De Santis and Feliziani. So there are more factors there to be involved.

Professor Hirata spoke about watts per kilogram, talked about the limit with SAR. Their work calculated the temperature change of about .35 of a degree. Just in terms of those resonance type issues, you are getting those peaks happening when you are getting standing waves happening in the eyeball of around about two gigahertz. So much more complex issues happening, as I said.

This is a paper from Professor Hirata, where he looked at rabbits exposed to microwave energy. This is a picture of SAR, directly from his paper. This is the temperature change, this is the actual temperature that happened due to that SAR, although I think that picture of temperature change is probably somewhat similar. You probably get similar effects.

SAR is going to be lowish in the lens, although it is going to be higher in the electrical conductivity ranges of the aqueous and vitreous humor around the edges of the lens.

In terms of temperature change, what effect does that have? We have got blood perfusion happening in the eye. You are going to get the higher temperature changes happening in the lens region and in the vitreous humor and around the back of the eye. It is lesser at the front of the eye where you are getting more of the surface type effects.

So in summary, temperature change due to SAR, of course there is going to be a correlation coming directly by SAR just by definition. You can see from the work today that correlation. I will let you judge whether you feel that is a reasonable correlation or not. Of what they are highly correlated, I'm not sure if correlate is quite the right word; maybe moderately correlated. When you mass average SAR, particularly around that seven gram and ten gram, when you use that type of mass averaging, that is when you are going to get the highest correlation.

However, I have got here some disclaimers. There are several factors that need to be considered. This is a workshop looking at particular tissues and effects of SAR on those tissues, and the following need to be taken into account from a modeling perspective.

The most critical is definitely the blood flow, but in all the work we have done in sensitivity analyses and other tests we have done, blood flow is very important. The position; if it is near the surface then it is going to have less effects. If it is deep within the body it is going to have less just by protection. But if it is closest to the surface but maybe not on the surface, then that is when you are going to have the highest effect in terms of temperature change because of the SAR.

I haven't spoken about thermal conductivity so far in this talk, but that is also going to be a key player in terms of effect on tissues. If a tissue is going to have high thermal conductivity, then it is going to dissipate that heat away from that tissue, so that also needs to be looked at.

The frequency of exposure you see in the plots shown there. It is going to be critical as to what type of frequency and how the SAR is going to affect that. I didn't mention, I think I had it in one of my slides, but that frequency is going to very much affect the thermoregulatory response. I think all the speakers this morning and the rest of the day have been talking about that, and also the work of Eleanor Adair, where you have got the higher frequencies and microwave frequencies, you are going to have these surface type thermoregulatory responses. When you are going to have the lower frequencies, around say 400 megahertz as it was in Eleanor's work, then you are going to have the responses from the base of the brain.

Location of the exposure; I haven't spoken about that as well, but of course that is critical. And with each tissue obviously there is thermoregulatory response.

To finish up, I would like to acknowledge the support of the MMF and GSMA. Their support to us has been very valuable and very much appreciated over the years.

Thank you very much.

DR. MORRISSEY: Thank you for a very nice talk. That is the other side of the issues. With the biologists here, looking at the critical temperatures, the other side of that issue is the engineers trying to model the RF energy into the tissue, which is equally challenging. And certainly the blood flow issue is very complex, and we are not there yet.

DR. MC INTOSH: No, in terms of modeling. For example, you have got Penne's equation, you have got other equations; is that valid? People say that it is valid.

Behind all this modeling you want research to be done out in the laboratories. You want these measurements to be made of tissues. You would like to have different testing of models versus Penne's equation versus other types of equations. It is very limited.

We would like as much as possible to work as closely as possible with the biological community. That is just not happening. Your type of workshop here, Joe, -- I thanked you at the beginning; you had walked out of the room at the moment -- these types of workshops are really critical for the electromagnetic and the modeling community to be able to say I have got a model, and my model actually makes sense.

DR. DEWHIRST: Just to clarify, right now you are not accounting for change in blood flow in response to temperature elevation, are you?

DR. MC INTOSH: Most of the models that we have done look at low levels, so basically it is an approximation. But if you wanted to look at high levels of exposure, then yes, all those equations, you can just make the blood flow as a function of temperature. Like in Bernardi's paper, they have done that. They have got that in there, and yes, if you want to, you can incorporate that into your model. It just comes down to mathematics and coding out.

DR. DEWHIRST: Right, but right now it is incremental. Bernardi is a fixed point, it is not a dynamic.

DR. MC INTOSH: He has both, but as I said, in our work you don't need that. But if you do need it, then we have it into our models. It is easy for us to do that, incorporate it in there, if we are looking at high levels of exposure, very high levels, up where you are looking at challenging the body's thermoregulatory responses.

DR. MORRISSEY: Any other questions? With that, we have a dinner. The shuttle will leave from the lobby at 6:30. I will close the workshop for today. We will be back here bright and early 8 o'clock tomorrow. Eight o'clock tomorrow we will see you bright-eyed and bushy-tailed, and we will get on with the second half of the workshop.

Thank you very much.

(Whereupon, the meeting was recessed at 4:57 p.m., to reconvene Tuesday, January 12, 2010 at 8:30 a.m.)