P-R-O-C-E-E-D-I-N-G-S

(8:34 a.m.)

DR. WEISS: Would everyone please take their seats? We will be beginning in a moment. I would like to call this meeting of the Ophthalmic Devices Panel to order. We will have introductory remarks by Sally Thornton and for the record, I would like to note that there is a quorum present.

MS. THORNTON: Good morning. I'd like to introduce myself. I am Sara Thornton, and I am the Executive Secretary of the Ophthalmic Devices Panel. On behalf of the FDA, I would like to welcome you to the 106th meeting of the Ophthalmic Devices Panel. Before we proceed with today's agenda, I have a few short announcements to make. I'd like to remind everyone to please sign in our at the registration table. There are sheets there for you to fill out, just your name and whether you're from industry or the panel or FDA or the public. Please, we do like to have that filled out.

All public handouts for today's meeting are available at the registration table. There are two new additions to our usual group of handouts. We've put out there information on public participation in open public hearings and copies of a guidance document for FDA and industry on quality system information for certain pre-market application reviews.

Messages for panel members and FDA participants, information or special needs should be directed through Ms. AnnMarie Williams, who is available at the registration table. The phone number to call for the meeting area is 301-590-0044. In consideration of the panel, the sponsor and the Agency we ask that those of you with cell phones and pagers either turn them off or put them on vibration mode while in this room and to make your calls outside the meeting area, please.

Lastly, will all meeting participants please speak into the microphone and give your name clearly so the transcriber will have an accurate recording of your comments? Now, at this time, I'd like to extend a special welcome and introduce to the public the panel and the FDA staff a new panel consultant who is with us at the table for the first time today, Dr. Oliver Schein, to my left, who comes to us from Johns Hopkins University where he holds a joint appointment as the Grossman Professor of Opthamology in the School of Medicine and as a Professor of Epidemiology in the School of Public Health and Hygiene.

His clinical expertise is in the medical and surgical management of patients with corneal disease and problems involving the interior segment of the eye. I'd also like to welcome our acting industry rep, Mr. Michael Crompton, Vice President for Regulatory and Clinical Affairs and Quality Assurance for Carl Zeiss Meditec, Inc. Mr. Crompton is sitting in for Mr. Ronald McCarley, who will not participate in today's proceedings at the request of the PMA sponsor.

Will the remaining panel members please introduce themselves beginning with Glenda?

MS. SUCH: Glenda Such, Consumer Representative.

DR. SUGAR: Joel Sugar, University of Illinois at Chicago.

DR. BANDEEN-ROCHE: Karen Bandeen-Rhodes, Johns Hopkins University.

DR. McMAHON: Tim McMahon, Department of Ophthalmology, University of Illinois at Chicago.

DR. MATOBA: Alice Matoba, Cullen Eye Institute, Baylor College of Medicine.

DR. BRADLEY: Arthur Bradley, Professor of Vision Science, Indiana University.

DR. WEISS: Jayne Weiss, Kresge Eye Institute, Wayne State University, School of Medicine.

DR. MATHERS: Bill Mathers, Oregon Health Sciences University.

DR. HO: Allen Ho, Wills Eye Hospital, Philadelphia.

DR. GRIMMETT: Michael Grimmett, West Palm Beach Florida.

DR. MACSAI: Marian Macsai, Northwestern University, Chicago.

DR. McCULLEY: Jim McCulley, University of Texas, Southwestern Medical School, Dallas.

DR. COLEMAN: Anne Coleman, UCLA.

DR. ROSENTHAL: Ralph Rosenthal, FDA.

MS. THORNTON: Thank you, panel. I'd like to read now the conflict of interest statement for this meeting of October 3rd, 2003. The following announcement addresses conflict of interest issues associated with this meeting and is made part of the record to preclude even the appearance of an impropriety. To determine if any conflict existed, the Agency reviewed the submitted data for this meeting and all financial interest reported by the committee participants. The conflict of interest statutes prohibit special government employees from participating in matters that could effect their or their employer's financial interest.

The Agency has determined, however, that the participation of certain members and consultants, the need for whose services outweigh the potential conflict of interest involved is in the best interest of the government. Therefore, a waiver has been granted for Dr. Oliver Schein for his interest in firms that could potentially be effected by the panel's recommendations. The waiver which allows him to participate fully in today's deliberations involves a pending consulting relationship on a competitor's unrelated product for which he has not received any compensation and also consulting with a competitor on unrelated matters for which he receives between $10,001.00 and $50,000.00 yearly.

Dr. James McCulley has been granted a limited waiver which allows him to participate in the review and discussion but excludes him from voting on the application. Dr. McCulley's waiver involves three consulting arrangements with competing firms. For these consulting services he received greater than $50,000.00 within the past year. Copies of these waivers may be obtained from the Agency's Freedom of Information Office, Room 12A-15 of the Park Loan Building.

We would like to note for the record that the Agency took into consideration other matters regarding Drs. Bradley, Schein and Coleman, Michael Grimmett, Allen Ho and Jayne Weiss. Each of these panelists reported past or current interest involving firms at issue but in matters that are not related to today's agenda. The Agency has determined, therefore, that the panelists may participate fully in the deliberations with the exception of Dr. McCulley, as noted previously.

We would also like to note that the Acting Industry Representative for this meeting, Mr. Michael Crompton, reported that his employer has numerous business relationships with firms at issue. In the event that the discussions involve any other products or firms not already on the agenda for which an FDA participant has a financial interest, the participant should excuse him or herself from such involvement and the exclusion will be noted for the record.

With respect to all other participants, we ask in the interest of fairness that all persons making statements or presentations disclose any current or previous financial involvement with any firm whose products they may wish to comment upon. Thank you.

I'd like to read not at this time the appointment to temporary voting status for this meeting. Pursuant to the authority granted under the Medical Devices Advisory Committee Charter dated October 27th, 1990, and as amended August 18th, 1999, I appoint the following individuals as voting members of the Ophthalmic Devices Panel for this meeting on October 3rd, 2003. Drs. William Mathers, Karen Bandeen-Roche, Joel Sugar, Marian Macsai-Kaplan and Oliver Schein. For the record, these individuals are special government employees and consultants to this panel or other panels under the Medical Devices Advisory Committee.

They have undergone the customary conflict of interest review and have reviewed the materials to be considered at this meeting. Signed, David W. Feigal, Jr. MD, MPH, Director of the Center for Devices and Radiological Health dated September 26th. Thank you. Dr. Weiss.

DR. WEISS: Thank you, Sally. We will now begin the open public hearing. Captain Steven Schallhorn -- I'm sorry, I'm just going to have him approach the podium and then I have a statement. But, I'm sorry, you have a presentation to make to Dr. Matoba. I apologize.

DR. ROSENTHAL: I do thank you very much.

DR. WEISS: That's very important.

DR. ROSENTHAL: I will come over and stand next to her.

MS. THORNTON: Give him a microphone. This is important.

DR. ROSENTHAL: Hi. I get two kisses this time. I'd like to give this presentation to Alice Matoba and read the Associate Commissioner for External Relations' comments. "Dear Dr. Matoba, I would like to express my deepest appreciation for your efforts and guidance during your term member -- your term as a member of the Ophthalmic Devices Panel of the Medical Devices Advisory Committee. The success of this committee's work reinforces our conviction that responsible regulation of consumer products depends greatly on the experience, knowledge and various backgrounds and viewpoints that are represented on the committee.

In recognition of your distinguished service to the Food and Drug Administration, I am pleased to present you with the enclosed plaque". And I am pleased to express my thanks. Alice and I go back a long time.

(Applause)

DR. MATOBA: Well, thank you, Dr. Rosenthal. It was a great honor for me to be asked to serve as a member of the FDA Ophthalmic Devices Panel and it's been such a great pleasure for me to work with the excellent FDA staff and fellow panel members and with you and especially with Sally Thornton, who has done such a great job.

I have been so impressed with the thoroughness and the very high standard of scrutiny that you give to all of the protocols that we have seen and I look forward to continuing to work with you as a consultant in the future. Thank you.

DR. WEISS: Thank you, Alice. Thank you, Dr. Rosenthal. We will now begin the Open Public Hearing but first, I wanted to read a statement that was requested by the FDA. "Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision making. To insure such transparency of the open public hearing session of the Advisory Committee, FDA believes that it is important to understand the context of an individual's presentation. For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement, to advise the committee of any financial relationship that you may have with the sponsor, its product and if known, its direct competitors.

For example, this financial information may include the sponsor's payment of your travel, lodging or other expenses in connection with your attendance at the meeting. Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have such financial relationships. If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.

Dr. Schallhorn, we have your presentation, we have up to a half hour for the open public hearing, but you have 10 minutes at this point.

DR. SCHALLHORN: Well, good morning, and thank you for allowing me to address the panel. My name is Steve Schallhorn. I'm an opthamologist, the Director of Cornea and Refractive Surgery at the Navy Medical Center, San Diego. I have no financial interest in STAAR. I'm not a paid consultant. I've self-funded my travel to come here to address the panel. I am a clinical investigator in the Toric ICL Study, which is ongoing but treatments at our center have not begun.

I'd like to also add that I'm an active duty U.S. Navy Ophthalmologist but the views that I express are not necessarily those of the U.S. Navy.

The reason I'm here is just to address an important issue, I believe and that is that we need options. We need surgical options, surgical options beyond what we can do with keratorefractive surgery in particular, excimer laser ablative procedures, especially to correct high myopia. There are many issues here and they deal with issues such as thin corneas. There are patients who are not good candidates for refractive surgery because of high refractive errors.

Patients with high refractive errors may not be good candidates anyway because current technology induces a number of aberrations on the cornea which can result in visual symptoms. And there are patients or subject that we want to treat that have critical visual demands, especially those again with high refractive error.

Now, my area of expertise and what we've studied to a great extent, deals with the quality of vision after refractive surgery and that's really what I'd like to spend the rest of the time talking about. The -- what I'd like to talk about is a study that we've conducted looking at a 105 consecutive LASIK subjects that we had visual acuities measurement on, questionnaires and a special test, a night-driving simulator. I'll talk more about that.

This was LASIK performed with multiple laser platforms with a six and a half millimeter optical zone size with a transition zone, so it's the latest technology for high myopia. This was also conventional and not customer wavefront-guided. The average preop refraction was relatively high, it was minus six, a little over minus six diopters and it ranged up to minus 11. At six months the results were good and the uncorrected visual acuity results were satisfactory with about three-quarters of the patients achieving 20/20 uncorrected.

The night-driving simulator that we used was a derivative of the simulator that Dr. Ginsberg developed that I believe was required in some earlier investigational studies conducted for intraocular lenses. This test, and it's shown here, you can see the -- it doesn't show up very well, but on the right side, it's looking over the shoulder of a subject in best corrected trial frames right here, looking at a rural night driving scene at 55 mile per hour. It's done in best corrected vision. Each eye is tested independently. There were numerous conditions at that the subject were tested on; that was business signs, traffic signs, pedestrian hazards, et cetera.

Six thresholds were made for each one of those conditions for both detection and identifying what that was and it was conducted with and without a glare source simulating driving which led to 144 measurements that were made, threshold measurements, per patient and so in these 105 subjects that we tested each eye independently, with this unique test, the data represents thousands and thousands of man-hours because it's extremely labor intensive. They're very, very specialized tests, but nonetheless, it's a performance-based task and that's what I'm going to start with.

It is a performance-based task, whereas, other tests, I should say of visual acuity such as contrast sensitivity, you can ask yourself, I certainly pondered this, you know, what does it mean if somebody has a subtle loss of contrast? What does that really mean and that's a very good question? What does that really mean and we're trying to get an answer to that, what does that really mean, but a performance based task built in has some of those answers addressed. This is a task that we are now looking at.

We look at that. Under all conditions, in this population of 105 subjects, we find a decrement in night driving performance. How much of a decrement? A little bit. This is the data shown another way and this shows the seconds improvement or the seconds decreased in the detection or identification distance, preop to post-op, so it's a paired analysis and zero represents no change post-op compared to preop and you can see most patients had no change. But the trend and the significant -- and it is significant that there was a loss. About 40 percent of patients had one second or longer increase in their detection distance.

Now, you could ask also, what does one second mean? Is that significant? We've worked with the National Traffic Safety Administration on the meaning of this and they've conducted studies which have shown that one second is a significant decrement in night driving performance at 55 miles per hour under similar but different circumstances. So it's a -- we're seeing a significant loss in a significant portion of patients treated with LASIK for relatively high levels of myopia.

Now, let's look at the vision. This is best corrected and five percent contrast acuity shown on the same chart. In orange, it's best corrected and this is lines gained or lost and you can see most patients had no change but the curve has shifted to the right meaning more patients had improvement than a decrement, consistent with what we see and that's, perhaps, partly due to reduction in minification from the act of putting that correction on the cornea.

In contrast to what we see with high contrast acuity, we see a shift to the left or worse with five percent contrast acuity, five percent low contrast acuity. It's an ETDRS eye chart, that five percent level and it's backlit. We see a loss, in fact, 25 percent of patients having measurable loss of contrast acuity with this. How about the symptomatology, most patients have no change in their symptomatology, preop to post-op. However, the curve is shifted slighted toward worse. Again, this is a paired analysis. We're looking at all patients and the difference between post-op and preop. It's slightly shifted worse, meaning patients have symptoms. In fact, a subset of patients can have relatively significant symptoms after the surgery.

Now, we tried to find out, okay, what are the factors that now are related to their driving performance decrement, what are those factors and we've done correlation analysis. And we find surprisingly that pupil size placed no factor whatsoever and I'll talk more about the briefly. Pupil size placed no factor in their night driving performance. Where we see a significant decrement pupil size has no effect. One of the strongest effects we see, though, is the level of preop myopia. The higher level of preop myopia, the worse the night driving symptoms. I'll talk, again, more about that.

We also get correlations with symptomatology in night driving performance. We get correlations with the contrast. People who have worse contrast, don't do as well in night driving. That all makes sense. Here's, just quickly, shows the low-light pupil diameter and you can see we had patients that were eight millimeters or larger. We had a wide range of pupils. We did not exclude patients who had large pupils in this study. Just to repeat, we did not exclude patients who had large pupils from the study. We had a broad distribution of pupil size. We found no correlation with pupil size.

And all of the analysis that we've done, other types of analysis with many, many other data sets have shown no correlation with pupil size. However, we do find a significant correlation again, as I mentioned, with preop myopia. Patients who have high levels of preop myopia had a significant decrease in the night-driving performance. You can see on a scatter plot of all the data that there is significant spread. However, there is a significant relationship also.

Now, what are the causes of this, what are the causes of these problems after LASIK and the answer is, I think, has to do with higher order aberrations, the induction of higher order aberrations. This is looking at preop, a distribution of the higher order RMS preop and looking at it post-op in yellow and we see a significant increase in the higher order aberrations.

We do correlation analysis with those higher order aberrations and we find that the level of preop myopia is significantly correlated to induced or an increase in spherical aberration. And again, a lot of scatter, but a significant relationship. Likewise, we find that increase in higher order aberrations, higher order RMS, change in higher order RMS vertically versus change in five-percent contrast horizontally that there also a significant relationship. Patients who have increase in higher order aberrations have an increase or a decrease in their contrast acuity.

Anyway, in conclusion, conventional LASIK works well. Most patients have no symptoms, but in some patients, it can induce visual symptoms, it can reduce low contrast acuity, it can increase higher order aberrations and it can decrease night driving visual performance. Preop myopia is the strongest risk factor. Patients who are especially above six diopters have the greatest risk and, of course, that's also the range where improved algorithms, improved ways to do LASIK, such as wavefront-guided surgery, is not yet -- is not available.

And lastly, we need these kind of surgical options. Surgical options are needed especially to correct higher orders of myopia. Thank you.

DR. WEISS: Thank you, Dr. Schallhorn.

(Applause)

DR. WEISS: We don't usually have questions at this point, but if anyone had any pressing questions for Dr. Schallhorn, we could limit them to a few, otherwise, we'll -- Dr. Bradley does, Dr. Schallhorn.

DR. BRADLEY: Thanks for the presentation, Dr. Schallhorn. One question, you made an emphatic statement that pupil size was not critical. You then inferred from your data that these driving problems were related to higher order aberrations. Well, the one thing we know for use is that as pupil size gets bigger, aberrations get worse. So how can there be a correlation with higher order aberrations but not with pupil size?

DR. SCHALLHORN: Well, aberrations can increase as the pupil size increases. But its effect on visual performance is what I'm saying we don't see that effect on visual performance. For instance, there may be -- I think there are things we really don't understand about the visual system and this comes to the heart of several of them. You can have a very aberrated eye that might have aberrations at seven or eight millimeters but it may not effect visual performance. You can measure it on an aberrometer, but if it doesn't effect visual performance, I'm not sure.

You know, I think the central four, five, six maybe larger than that, millimeters, of the visual system is critical for high quality vision but it may not be that the eye has to be that perfect beyond that range, even though we can measure aberrations in that range.

DR. WEISS: Thank you very much. We are going to move onto the open committee session with the Division update by Dr. Rosenthal, followed by a Branch update by Donna Lochner.

DR. ROSENTHAL: Thank you, Dr. Weiss. This year we are pleased to announce the addition of several members to the staff of our Division and I'd like to introduce them to you. There are actually two from the Ear, Nose and Throat Branch but I will not introduce them. They're not here and probably will not be playing much of a role, though I will comment on them at the end on their -- who they are.

First, I'd like to introduce Lori Austin-Hanberry, who has joined our Division in the position of Project Manager. Amongst her duties will be insuring that the Division meets MDUFA (ph) product review goals. She's a Lieutenant Commander in the Public Health Service, has over 14 years experience as a registered nurse with clinical, instructional and management background. Prior to joining FDA she managed various clinical and administrative operations for the Montgomery County Department of Health and Human Services, most recently managing the Childhood Lead Poisoning and Prevention Program.

She was also a Captain in the Air Force Reserves for 11 years. She obtained her nursing degree from Howard University and her Masters Degree in Health Care Administration from Central Michigan University. Lori?

Dr. Joseph Blustein is a shared hire with the Office of Surveillance and Biometrics and will be working on post-market issues relating to ophthalmic and ENT devices. He is a Board certified ophthalmologist and former Medical Director of the Wisconsin Peer Review Organization. He has two Masters degrees, one in epidemiology and one in food science. He serves on the Wisconsin Public Health Advisory Committee and we welcome Dr. Blustein.

Clay Buttemere went to Virginia Tech to pursue his engineering studies. In 2000 he received his BS in engineering science and mechanic from Virginia Tech. He and his wife, after living in Macedonia, moved to Nashville, Tennessee where he enrolled in graduate studies in the Biomedical Engineering Department at Vanderbilt University. His research in the biomedical optics lab at Vanderbilt involved using optical spectroscopy to assess tissue thermal damage in vivo. In May of 2003, he received an MS degree in Biomedical Engineering from Vanderbilt and in August of this year he joined the FDA as a Biomedical Engineer.

Brad Cunningham is also a Biomedical Engineer, who was hired to work in Donna Lochner's Intraocular and Corneal Implants Branch. He received his undergraduate degree from the University of Maryland in Bioengineering focusing on biomedical instrumentation. After graduation, he is employed full time at Walter Reed Army Institute of Research in the Department of Neuropharmacology in the Division of Neuroscience. Whilst there, he co-authored three papers, two recently published articles focusing on studying the therapeutic intervention window following transient cerebral ischemia and the delayed gene response and he's also in the Public Health Service as you can tell from his uniform.

I'd like to announce that the Office of Science and Technology has brought Dr. Ethan Cohen to work as a staff fellow in the Electrophysiology Branch of the Division of Physical Sciences. This is also a shared hire with OST. He will be working in our Division as well. Dr. Cohen's area of expertise is electrophysiology of the retina and Dr. Saviola usurped me. His position is a shared high with the Office of Device Evaluation.

Dr. Cohn comes to CDRH from Harvard University where he was a visiting professor in the Department of Molecular and Cell Biology. Prior to working at Harvard, Ethan was an assistant professor in the Department of Ophthalmology and Visual Sciences at Yale Medical School. His PhD is in anatomy from the University of Pennsylvania Medical School. As an OST staff fellow, he will continue to research synaptic interactions of retinal cells. His review work with ODE will be in the area of retinal prosthetic devices that are reviewed in the Vitreoretinal and Extraocular Devices Branch of DOED. Dr. Cohen.

And the final two are from ENT. The first is Dr. Nandkumar, who is an electrical engineer with an MS in EE from Tulane University receiving his PhD from Duke in Electrical Engineering. He is an authority on acoustical issues and will be working in the ENT Branch and the final individual is Dr. Antonio Periera, who is a Board certified otolaryngologist, head and neck surgeon who was trained at the University of Puerto Rico and subsequently came to work in private practice in Washington, D.C.

He has been in the Center for Biologics since 1995 and where he had assisted in formulating regulations for the human tissue program and we pinched him from them and I must say we're delighted to have him join our staff, although there may not -- they probably will not be working on ophthalmic issues, they might be if we have issues that relate to their expertise.

So we welcome all seven new people and I hope you will all get a chance to work with them and enjoy their company. Thank you.

DR. WEISS: Thank you. Donna?

MS. LOCHNER: In the spirit of keeping the panel apprised of PMAs that have come before the panel previously, I'd like to discuss two such PMAs. First, P010059 is a PMA for the Morcher endocapsular tension ring used for capsular bag stabilization in patients with pseudo exfoliation syndrome or other situations of compromised zonulas.

This PMA was reviewed by the panel in January of 2002. The panel recommended that the PMA was approvable with requests for essentially a complete reanalysis of the clinical data to resolve discrepancies in the PMA and to clarify information that was presented at the panel meeting. We are in the final stages of review and we expect a decision in the near future.

The second PMA is P030002 for the C&C Vision CrystalLens Accommodating Intraocular Lens. This PMA was reviewed by the panel in May of 2003. The panel recommended that the PMA was approvable with requests that the patient satisfaction data be stratified by pupil size and that certain labeling revisions be made. The panel recommended that the lens provides accommodative amplitude of about one diopter. Again, we are in the final stages of review and expect a decision in the near future. Thank you.

DR. WEISS: Thank you, Donna. I will ask the sponsor to come to the podium. We are going to begin the presentation of PMA P030016. The sponsor has one hour for their presentation. I would request that each presenter speak into the microphone, initially identify yourself and your relationship with the sponsor and any potential financial conflicts.

DR. LAMIELLE: Good morning. My name is Helene Lamielle and I'm Chief Scientific Officer for STAAR Surgical. We are pleased to present you today PMA P030016 for the Collamer Implantable Contact Lens for the correction of myopia. Presenting on behalf of STAAR Surgical today will be Dr. Steven Slade, from Houston, Texas, Dr. John Vukich, a medical monitor from Madison, Wisconsin and Dr. Henry Edelhauser, Director of Ophthalmic Research at Emory University an Director of Specular Microscopy Reading Center.

Dr. Vukich has a financial interest in STAAR Surgical while Dr. Slade and Edelhauser are paid consultants with no financial interest other than compensation for their time. Dr. Donald Sanders will participate in the discussions that follow our presentation. Dr. Sanders has a financial interest in STAAR Surgical.

The STAAR Myopic Implantable Contact Lens is the subject of today's panel meeting, is indicated for the correction of moderate to high myopia between minus three to minus 20 diopters and is intended for placement behind the iris in the posterior chamber of the phakic eye. The design of the ICL is very similar to that of standard plate haptic intraocular lenses used for cataract surgery. However, the ICL has been designed with forward vault to minimize contact with the central anterior capsule of the crystalline lens. The lens material is a hydrophilic biocompatible polymer known as Collamer and has a history of safe use in approved standard posterior chamber intraocular lenses.

Here is a photograph of the ICL in the vault of the crystalline lens. The footplates are approximately 100 microns thick and are intended to rest in the sulcus. At this time, I would like to introduce Dr. Steven Slade, who will present the surgical procedure, study method for the PMA clinical trial and effectiveness outcome.

DR. SLADE: Okay, thank you, Helene. Good morning. My name is Steven Slade and I certainly appreciate the opportunity to present for you today. I'd like to begin my presentation by describing the procedure used to implant the STAAR ICL. The ICL is shipped to the surgeon in a sterile glass vial and hydrated in saline solution. The lens is removed from the vial with forceps. The lens is then loaded by the surgeon into a sterile disposal injector cartridge for insertion into the eye and this injection system is just like the ones we commonly use for small incision cataract surgery.

The injector is specifically designed to minimize surgical manipulation associated with the ICL insertion. Iridotomies are performed up to two weeks before the ICL surgery. The pupil is dilated and the entire surgery is performed under topical anesthesia. Viscoelastic is placed in the anterior chamber. The lens is injected through a spornia (ph) cataract-style incision.

Now, the surgery is completed then by positioning the lens haptics beneath the iris and rinsing out the Viscoelastic. The lens centers extremely well and no sutures were necessary in virtually all cases. The STAAR ICL is specifically designed to vault over the anterior capsule of the human crystalline lens. This vault should be approximately 500 microns or one corneal thickness. This shine through (ph) photograph demonstrates an average vault with the STAAR ICL.

The clinical study of the STAAR ICL described in this PMA was a prospective multi-center clinical trial designed to evaluate the safety and effectiveness of this lens for the correction of moderate to high myopia. Patients with myopia of minus three to minus 20 were enrolled and followed for three years. The study was originally planned for a two-year follow-up under the IDE which was approved in 1995.

During the study, follow up was extended to three years at the FDA's recommendation to be consistent with more recent guidance for studies of phakic refractive intraocular lenses. Our patients were required to be between 21 and 45 years of age and of note, their best corrected vision pre-optively could be as poor as 2100, and they were allowed to enroll with as much as two and a half diopters of refractive cylinder, since moderate to high myopia is associated with lower levels of best corrected visual acuity, and higher amounts of cylinder. Our effective parameters included a decrease in refractive myopia, improvement in uncorrected visual acuity, predictability of the refractive outcomes, refractive stability and patient satisfaction.

Safety parameters included a preservation of best corrected visual acuity. Slit lamp findings, intraocular pressure, contrast sensitivity with and without glare, reports of complications in adverse events. Specular microscopy was also performed and we'll present the results in detail of those studies. Accountability; 539 eyes of 305 patients were implanted with the ICL. Thirteen eyes of 11 subjects did not meet the entry criteria and were excluded from the safety and effectiveness cohort.

This accountability was well within FDA guidance of no more than 10 percent loss per year of follow up. Even though the study was originally planned for only two years of follow up, accountability at three years was 77.2 percent, exceeding the target of 70 percent identified in the FDA's draft guidance for refractive implants. Again, even though the FDA guidance requires a minimum of 80 percent accountability at two years, we had follow up on 91 percent of our cohort and at three years, we were well above the minimum follow up of 70 percent of patients.

The demographics of the study population were fairly unremarkable but it is worth noting that the average mean myopia preoperatively in this population was over 10 diopters, minus 10.1 diopters. Now, I'd like to show you uncorrected visual acuity for the entire study cohort and then uncorrected visual acuity for the eyes that had the potential preoptively to achieve 20/20 uncorrected vision as well as then the eyes that had the potential and were actually able to be targeted to emmetropia or 20/20.

Because we enroll patients with up to 20 diopters of myopia, not all eyes had the potential for 20/20 nor were all eyes able to be targeted to emmetropia. In part, this was the result of limits on the range of lens powers available during this study. If we look at the entire cohort of study patients, the uncorrected visual acuity over time 20/40 or better, excellent uncorrected distance visual was achieved rapidly, 80 percent at one week, 20/40 or better and had excellent stability, 81 percent 20/40 or better uncorrected at the three-year visit.

Again, looking at the entire study cohort, but at the 20/20 level, we see again, a rapid improvement in uncorrected acuity and excellent stability. It should be noted that the total cohort of eyes, this slide, includes those eyes that were not able to be targeted for emmetropia, eyes with preoperative best spectacle corrected visual acuity worse than 20/20 and eyes that had up two and a half diopters of refractive cylinder.

Here's the breakout for uncorrected visual acuity for the entire study cohort at three years showing the 20/20, 20/25, 20/30 and 20/40 levels. Now, if you take that same format, I'd like to show you the results for eyes that had the potential for 20/20 uncorrected vision. In this group, 89 percent, of 250, 89 percent reached 20/40 or better at the three-year visit and 52 percent were 20/20 or better, the eyes that had the potential preoperatively to reach 20/20, and the results get even better if we look at the patients who had both the potential to achieve 20/20 and were able to be targeted to emmetropia. In this population, good visual potential, 59 percent were 20/20 or better at the three-year visit and 95 percent were 20/40 or better uncorrected at their three-year visit. If we take the population and stratify it by preoperative myopia as in this slide, with less than 7, 7 to 10, 10 to 15 and over 15, it's apparent that the uncorrected visual acuity of 20/40 or better and of 20/20 or better was achieved by a lower portion of the eyes with preoperative myopia greater than minus 15.

It's not unexpected given that the majority of these eyes could not be targeted for emmetropia and the lens powers were not available to allow for full correction of all eyes in this group. Further, in this group of the highest myopes, only four eyes had best corrected visual acuity of 20/20 or better preoperatively. In fact, if we look at the patients again stratified by myopia, who had the potential for 20/20 and were targets of emmetropia, we see excellent results at both the 20/40 and the 20/20 levels of uncorrected vision.

But indeed, none of the patients who were in the over 15 group actually even had the potential for 20/20 at the same time they were able to be targeted to emmetropia. We'll have more to say about this group of higher myopes, over 15, later in the presentation since it certainly is a unique population.

From a patient's perspective, this efficacy ratio slide comparing the post-operative uncorrected visual acuity to the preoperative best corrected visual acuity may be the most important data in this part of our presentation, since this is what patients are seeking, uncorrected vision, better than or equal to what they were able to see before surgery with their best spectacle correction. The efficacy ratio for the ICL was excellent with upwards of 60 percent of patients seeing as well or better after surgery with nothing, no correction, than they were able to see before surgery with their very best spectacle correction.

We examined the standard metrics of predictability of refractive outcome as well as refractive stability. As indicated on this slide, our achieved levels of plus or minus a half and plus or minus one attempted versus achieved, were excellent and did exceed FDA targets for both phakic IOLs and refractive lasers were greater than minus seven diopters of myopia. Accuracy of the attempted refractive change was excellent in eyes of pre-operative myopia looking at the cohort stratified by myopia up to minus 15 and then did, indeed, decrease for the myopes with a baseline myopia greater than 15 as you can see in this slide, again, three years looked at the entire cohort stratified by myopia.

This slide pretty much speaks for itself. This is our stability slide. The achieved refractive change was again, both rapid, one week, and sustainable throughout the follow up minus a half, minus a half at 36 months. These outcomes do exceed FDA guidance for stability of manifest spherical equivalent refraction.

A patient survey was administered to all study subjects and I will share the three-year results of that survey with you. Ninety-nine percent of our patients reported very extremely or moderately satisfied. When asked to rate their quality of vision, 77 percent reported very good or excellent quality of vision as compared to 55 percent of patients before the surgery. Indeed, 97 percent of the study patients expressed a willingness to have the ICL surgery again. The unwilling included eyes with refractive errors, hyperopia (ph), myopia, vomiting right after surgery, and one patient who questioned why repeat the surgery when they had already had the surgery and were doing fine.

To summarize, our uncorrected distance visual acuity at three years all eyes in the yellow was excellent. Eighty-one percent of the entire cohort achieved 20/40 or better and 95 percent of the entire cohort stratified now for those people that had the potential to see 20/20 and were able to be targeted for 20/20, 95 -- that group, 95 percent of those patients achieved 20/40 or better uncorrected visual acuity.

Predictability of refractive outcome was also excellent, exceeding FDA targets with a significantly -- a very small amount of patients winding up over-corrected or under-corrected, particularly in view of the very broad range of high to moderate myopia treated and this does, again, exceed FDA targets.

And now I would like to introduce Dr. John Vukich, who was the medical monitor for the ICL clinical trial. Dr. Vukich will present safety outcomes and he'll be followed by Dr. Henry Edelhauser who will discuss the specular microscopy outcomes for the ICL study. Thank you.

DR. VUKICH: Good morning. My name is Dr. John Vukich and I am the medical monitor of the STAAR Surgical Implantable Contact Lens Clinical Trials. I have a financial interest in STARR Surgical. I will be presenting the safety outcomes for the study cohort.

Key safety parameters that were analyzed and will be presented include preservation of best spectacle corrected acuity, complications and adverse events, lens opacities, inflammation, patient symptoms, contrast sensitivity and endothelial cell analysis. There was a rapid and sustained return of best spectacle corrected visual acuity in the study population beginning at one week and continuing through every follow up interval through the three-year period. At every follow up visit the proportion of eyes with 20/40 best corrected acuity was improved over the baseline preoperative level of 97 percent. When we break out the best spectacle corrected acuity at three years, the improvement experienced by the study population is even more notable particularly with regard to the improvement in spectacle correction of 20/20 and 20/25.

Thus, these patients have the potential to benefit not only with regards to uncorrected acuity, but also in terms of best spectacle corrected acuity. This study population is quite different from other populations that have undergone refractive surgery evaluations in that only 69 percent of the preop cohort could be corrected to 20/20 or better. We believe this is a unique feature of this cohort and reflects the high level of myopia and the unique challenges these patients face.

When we stratify postoperative best corrected acuity by baseline myopia, at every level of myopia, the ICL cohort experienced an improvement in best corrected acuity at 20/20 or better as compared to baseline. The highest myopes, those with preoperative myopia greater than 15 diopters, also experienced a substantial improvement at the 20/40 level. The most dramatic increase was observed in those patients with the highest level of myopia. While we acknowledge the contribution of magnification in this group of very highly myopic patients, the visual results are real and are enjoyed by the patients.

When we look at the changes in lines of best spectacle corrected acuity, 49 percent of eyes gained one or more lines of acuity at three years. This contrasts with only eight percent of eyes that lost one or more lines of best corrected acuity. Complications and adverse events are an important aspect of the evaluation of the ICL and we examined this from several perspectives. Perioperative complications were reported for 17 eyes, the most common of which was removal and reinsertion on the day of surgery.

A small number of other perioperative complications was also reported and these included reformation of the anterior chamber, a peripheral iridectomy and repair of iris prolapse. Postoperative complications other than intraocular pressure rises, lens opacities or secondary surgical procedures were reported in five of the 526 eyes in the study cohort for an incidence of less than one percent. Since there were so few of these cases in this category, I think it is useful to describe each of these individually.

One eye experienced a macular hemorrhage at one week and this result without sequelae. An asymptomatic subretinal hemorrhage was observed as an incidental finding at the three-month visit and best corrected visual acuity remained unchanged from baseline in this eye. Three retinal detachments were reported during the three years of follow up in this ICL clinical trial. One eye had a retinal detachment with a macula off. This required repair with silicon oil and a subsequent nuclear pacification was noted with loss of best corrected acuity to count fingers. This patient represents the only case in the study cohort with irreversible loss of acuity to worse than 20/40. This patient had 16 diopters of myopia preoperatively.

Two other retinal detachments were reported during the course of the clinical trial. Both cases were successfully repaired such that the final best corrected visual acuity remained within one line of the preoperative spectacle correction. Based on published reports, and incidents of retinal detachment of .68 percent per year might have been anticipated and we might have anticipated as many as nine retinal detachments in this study cohort that is following 526 eyes over three years. That fact that we had only three retinal detachments in this study suggests that the ICL had limited or no impact on the incidents of this adverse event.

Intraocular pressure rises occurred in 20 eyes or 3.8 percent of the study cohort. The majority of the acture pressure rises occurred during the first one to two days after surgery. Preoperative iridotomies were performed on all study eyes as a routine part of the ICL surgery. Seventeen eyes required additional YAG iridotomy or enlargement of an existing iridotomy for control of intraocular pressure. Irrigation of the anterior chamber for removal of retain viscoelastic was performed in three eyes. Late intraocular pressure rises occurred in five eyes or less than one percent of the cohort. This was defined as a single reading intraocular pressure of 25 millimeters or greater or an increase over baseline of 10 millimeters of mercury at three months or later.

In three of these eyes the intraocular pressures are being monitored without intervention and the most recent pressures are shown on this slide. Two eyes are currently being treated with a topical beta blocker. The most recent intraocular pressure for these patients are 20 millimeters of mercury or less. Secondary surgical procedures were performed in three percent of the study cohort. The most common procedure was removal and replacement as a result of sizing issues.

Repositioning was performed in four study eyes. One ICL was replaced for a power miscalculation. In the entire study cohort only three eyes underwent ICL removal and cataract extraction representing .6 percent of the entire study population. This summary slide shows all of the secondary ICL surgeries. I'd like to point out that only a single eye lost best corrected acuity and this loss was only one line occurring in one eye that underwent and ICL repositioning. It is particularly noteworthy that those patients who underwent cataract extraction maintained their best spectacle corrected acuity relative to their preoperative level prior to insertion of the ICL.

Assessment of the crystalline lens was an area of significant concern and this was monitored carefully throughout the course of the study. Nuclear opacities were observed in five eyes of three patients. In a patient who was previously described, a nuclear opacification occurred following retinal detachment which was repaired with silicon oil. Both eyes of two patients developed simultaneous bilateral nuclear opacities between two and three years postoperatively and one of these four eyes required cataract extraction. Once again, it should be noted that all of these eyes were very highly myopic ranging from minus 14 to minus 17 diopters.

Lens clarity was graded at all patient visits using the LOCS 3 Scale. This scale ranges from zero to 5.9 and under this system a Grade 1 was best described as a trace opacity. Given here is the photographic standard for Grade 1. I'd like you to keep this photograph in mind since over half of the anterior subcapsular opacities we are going to describe were no greater than this clinical standard. In fact only one eye in the study had an anterior subcapsular change at Grade 2 or higher.

Anterior subcapsular opacities were observed in 14 eyes of 13 patients. It is important to note that 12 of these 14 cases were asymptomatic and visually insignificant at the most recent follow up visit. We believe that many of these cases were surgically related and this is supported by the fact that 11 of these cases occurred within the first six months of surgery.

Clinically significant anterior subcapsular opacities were observed in only two eyes. These were defined as LOCS score of less than -- greater than .5 with a loss of two or more lines of best spectacle corrected acuity or an increase in glare or a opacity requiring ICL removal with cataract extraction. One of these cases was a surgical mishap in which a preservative containing topical miotic was inadvertently injected into the anterior chamber.

The second case was an eye in which an opacity was observed six months postoperatively. Cataract surgery was performed and post-cataract best corrected acuity was unchanged from the pre-ICL baseline. To summarize our findings on lens opacities, only three cataract extractions were performed in the study population of 526 eyes.

One was related to the inadvertent injection of a topical preserve miotic into the eye. One was a nuclear cataract and the third case was an anterior subcapsular opacity that did progress to the level of clinical significance. Best corrected visual acuity was unchanged or improved following cataract extraction in all three eyes compared to pre-UCL treatment. Safety may be best summarized in eye -- by examining the eyes with persistent loss of best corrected acuity of two or more lines. There are only five of these eyes and you have seen all of these cases previously in our presentation on safety.

Here is the retinal detachment repaired with silicone oil and the eye irrigated intracamerally with preserve miotic agent. Additionally, three of the nuclear opacities had a persistent loss of two or more lines of best corrected acuity. One of these we've just described had cataract extraction. In the entire clinical trial, these are the only eyes that had a persistent loss of two lines or more of best corrected acuity.

Next I would like to present our findings related to inflammation. Slit lamp examination was performed in all study eyes at all visits and a laser cell-flare meter was used to evaluate information in a sub-study of patients. No inflammatory response was observed after the first week postoperatively either clinically or by the more sensitive laser cell-flare meter. Laser flare measurements following ICL implantation were within the normal range for the first post-operative week, and remained normal throughout the course of the entire clinical trial.

A subjective questionnaire was administered to all study patients preoperatively and at follow up examinations. Patients were asked to rate each of the symptoms listed on this slide as either absent, mild, moderate, marked or severe. When comparing preoperative responses to those attained at three years, there were no significant changes in symptoms rated as absent or mild.

Equally importantly is the fact that there were no significant changes from baseline to three years in symptoms rated as moderate, marked or severe. Contrast sensitivity and glare were evaluated in a sub-group study. Well established techniques were used in our contrast sensitivity testing. After 10 minutes of dark adaptation, measurements were made both with and without a glare source. There was no loss of contrast sensitivity at any spacial frequency when compared to baseline to postoperative results. In fact, at two frequencies there was a significant increase in log units of contrast sensitivity. When contrast sensitivity was repeated in the presence of a glare source, there was a significant improvement at all four spacial frequencies starting at three cycles per degree up to 18 cycles per degree.

I would now like to introduce Dr. Henry Edelhauser who will be presenting the Specular Microscopy Substudy.

DR. EDELHAUSER: Thank you, John. Good morning. I'm Dr. Henry Edelhauser, Director of ophthalmic research at Emory University. I have no financial interest in STAAR Surgical. I serve as Director of the Specular Microscopy Reading Center for the ICL clinical trial and will be presenting the results of a sub-study conducted by STAAR Surgical to evaluate the effects of the ICL implantation on the corneal endothelium. I would like to emphasize the importance of the methods used at the Specular Microscopy Reading Center. Images were received from 12 investigators at nine clinical sites and a signal masked reader analyzed all the images. The images were then scanned and analyzed with the Konan KSS-300 Software. Approximately 1300 images were analyzed in this study and the mean number of cells per image that was counted was 93. This slide shows how the images were handled and that the images were taken with a Konan. They were then sent to us in the reading center as hard copy. We then scanned them. We then resized them and formatted the images. We then calibrated and analyzed, put it in a spreadsheet and then sent the data back for statistical analysis at STARR.

I think it's important when we talk about specular microscopy to review what a good image is because not all specular microscope and reading centers and photographers are able to take good images and this is the real challenge in undertaking specular microscopy. One, it's important to have distinct cells as illustrated on the right. In the specular micrograph one can identify 100 cells and even more in that's essentially what we do at the reading center is to identify as many cells as possible because when you analyze this, it's done by putting a dot in each one of the cells and then the analysis software is the nearest neighbor analysis. So cells in the periphery that don't have a nearest neighbor are not counted.

Cells need to be grouped into form in a contiguous area and then after you have dotted all the cells, it's extremely important that the evaluator or the reader check to see that the cells haven't been double-dotted or cells missing because if you miss three cells, you have a significant change in the end of field cell density because what you see from this specular micrograph is multiplied by 10⁶. Precision of the readings is an important factor in the analysis of any endothelium. We have estimated that the precision to be two percent in the ideal situation which was published in our study of LASIK patients. In this particular case we had one single clinical site, photographer and one single reader. When you undertake multi-center study where you have numerous photographers and then you then send this to a reading center and one single reader, the precision is somewhere between eight to 10 percent.

The outcomes of our analysis of the corneal endothelium are shown in this slide and include endothelial cell density, percent hexagonality, or pleomorphism and coefficient of variation or polymegathism. Studies indicate that stress corneas present -- have a percent hexagonality of less than 45 and a coefficient of variation greater than 45.

Published studies and studies from my own laboratory have shown that morphology is the best indicator of corneal endothelial stress and instability. I would now like to share with you some examples where endothelial morphology has been demonstrated to be the most sensitive measure of corneal endothelium stability. These examples are pseudophakic bullous, diabetes, and contact lens wear.

In this seminal paper, published by Rao and Aquavella in 1984, they studied patients implanted with iris fixated lenses in patients whose corneas made clear shown in the yellow bars, were compared to patients who ultimately developed corneal edema in the blue bars. Interesting, these authors found no difference between the two groups with regard to percent endothelial cell loss. However, there was a marked difference in coefficient of variation indicating that morphology is a more sensitive indicator for the development of bullous keratopathy.

In the second illustration, the corneal endothelium is illustrated in diabetes and this was published from one of our papers in 1984 where we reviewed the endothelium of both Type 1 and Type 2 diabetics. In this study we showed there was no significant difference in endothelial cell density but there was a significant difference -- decrease in percent hexagonality and a significant increase in coefficient of variation.

The next example that shows the importance of morphology is related to endothelial cell density is provided in a study by McRae and Matsuda, et al, and they compared patients who used contact lenses for more than 20 years and compared to age-match controls. Again there was no significant difference in endothelial cell density, a significant decrease in hexagonality and a significant increase in the coefficient of variation.

The three examples I've shown demonstrate the corneal endothelial morphometric changes are the first indicators of endothelial stress. The percent hexagonality and coefficient of variation are more sensitive indicators of endothelial stability than endothelial cell density.

I would now like to review the ICL STAAR PMA data on endothelial morphology. This graph is a scattergram of all pre and post-operative data points. In general, the majority of the points were between 2,000 and 3,000 cells per millimeter square with a small number our outlyers. The dark bars in the center of the scattergram illustrate the mean plus or minus 90 percent of the confidence interval. This slide shows a similar scattergram but with data points for a consistent cohort of 37 eyes with specular microscopic data in all visits from preop to four years.

This slide does show that the endothelial cell density remains unchanged from three to four years.

DR. GRIMMETT: Do you know the confidence intervals at the last visit?

DR. EDELHAUSER: Yes, I have it. It's coming up in the next slide.

DR. GRIMMETT: Thank you.

DR. EDELHAUSER: The table shows the pair-wide comparison of endothelial cell density at consecutive intervals beginning with the preop to three months a minus .2 was measured and cell loss was observed and from three months to one year a minus .9 percent observed. Between three and four years, a plus .1 percent and a narrow confidence limits of 1.4 percent to plus 1.6 percent. The percent hexagonality data shows no change over the course of study in this cohort of patients.

For comparative purposes, a percent hexagonality of 45 would be an indication of a stressed corneal endothelium. The coefficient of variation data also shows no increase over the course of study of this cohort. Again, for comparative purposes, a coefficient of variation of 45 would be an indication of a stressed corneal endothelium. In summary, the specular microscopic data show a cumulative or a total mean endothelial cell loss of 8.4 percent to 9.7 percent over a course of four-year follow up with stabilization suggested at the four years. It should be noted that there is no apparent mechanism for chronic cell loss due to the ICL. This is supported by the absence of changes in the percent hexagonality and coefficient of variation, which do not indicate chronic endothelial cell stress in this study population. This conclusion is supported by the previous reported data on pseudophakic loss, diabetics and contact lens wear.

We don't have a long-term study of endothelium in high myopes in the peer review literature. But we know that extrapolating endothelial cell densities over time is complex. It should be noted that the endothelium is not a homogenous population of cells from central to peripheral and migration of endothelial cells must be considered in any long term modeling of the endothelial cell density.

Recent data published from my laboratory in March of this year in the AJO, addressed the issue of peripheral corneal endothelial cells. In this study we found that if, indeed, you measure the corneal endothelium here and then you go two millimeters off in the paracentral region, there's a five percent increase in the corneal endothelium. And if you go four millimeters off center, there is a 10 percent increase in endothelial cell density.

Now, let's put this into perspective with this. The cell density within a four millimeter button is roughly 34,740. The paracentral region has a cell density of 119,845. And four millimeters off center in this area where we have a high cell density, the cell density is calculated out to be 264,632 cells per millimeter square. Now, this is not a study that doesn't have backup because it had first been identified by Bert Chimifane (ph) in 1984 and subsequently two papers in the German literature in `89 and `90, all showing an increase in the peripheral corneal endothelium. I do want to say that in this study we measured the corneal endothelial cells in four different ways; non-contact specular microscopy, contact specular microscopy, alizarin red staining of corneas we received from the eye bank, and also fixed corneas where we developed the nomogram to correlate with the pathologist the number of corneal endothelial cells as measured by the nuclei in high power field correlated to a nomogram of endothelial cell density.

The higher the endothelial cell density found in the paracentral and peripheral cornea affords an additional reassurance of safety for the endothelium in the patients implanted with the ICL. In summary, stability appears to be achieved between three years and four years in the ICL population. This data -- these data are consistent with endothelial remodeling and stabilization. The absence of any effect on the percent hexagonality, coefficient of variation support the absence of stress on the corneal endothelium. This would be consistent with an implant placed behind the iris and suggests that the endothelial cell loss observed in the ICL clinical trial is related to the initial surgical procedure and not a chronic phenomena.

Ongoing surveillance of the corneal endothelium will be critical to establishing the continual safety of the ICL and the study sponsor is committed to collecting the additional four-year follow up patients. Patients will also be asked to return for five-year specular microscopic exams and the same rigor and precision will be used to evaluate that corneal endothelium by the reading center. I would now like to turn the podium over to Dr. John Vukich.

DR. VUKICH: Once again, I am Dr. John Vukich. A unique group in our clinical trials represented by the patients with more than 15 diopters of myopia. This group deserves special attention since concerns have been expressed by both the FDA and panel reviewers regarding acceptability of study outcomes in this population. I think we all understand the unique challenges represented by this group of extremely myopic patients. These include significant variability in simply determining the end point of the manifest refraction. Many of these patients have poor visual acuity even with their best spectacle corrected acuity. In spite of this, the mean post-operative spherical equivalent was reduced from minus 17.3 diopters to minus 2.2 diopters with the implantable contact lens for an average correction of 88 percent of the pre-existing myopia. At the time of the ICL clinical trial, lens powers were not available to achieve full correction to emmetropia in all cases. Even with this limitation, 39 percent of eyes with greater than 15 diopters of myopia achieved an uncorrected acuity of 20/40 or better.

Substantial improvement was observed in the proportion of eyes with best corrected acuity of 20/40 or better. The proportion of eyes with best corrected acuity of 20/20 or better increased from 13 percent at baseline to 42 percent at three years. We acknowledge that magnification contributes to the observed improvement in best corrected acuity but continue to believe that this improvement in best corrected vision is an important benefit to the patient.

Any analysis of complications and adverse events in this population of high myopes must be viewed relative to their baseline risk. A body of published literature has established that the risk of spontaneous complications such as retinal detachment and nuclear opacities is significantly increased in high myopes. For example, the risk of detachment of the retina is 26 times higher in myopes above minus 6 diopters. A significantly increased risk has also been established for the incidents of nuclear opacities in highly myoptic patients. These complications must be viewed in the context of the increased risk of the population. Given the additional risk it should not be surprising that a higher rate of complications was observed in the subset of highly myopic patients.

Review of these complications which have already been presented as part of the safety data for the total study population revealed that two retinal detachments and four nuclear opacities were observed in six eyes. Only the eye with complicated detachment requiring silicone oil has had an irreversible loss of vision. In fact, this is the only eye in the entire clinical trial in this category.

With the exception of the eye with retinal detachment requiring silicone oil, all of these patients were satisfied with the outcome of ICL implantation and would be willing to undergo surgery again. We have shown that these patients had a substantial improvement uncorrected visual acuity and over half of these eyes experienced a gain in best corrected acuity. We believe that these are the very patients that stand to gain the most from implantation of an ICL particularly in the absence of alternative devices or surgeries for the correction or reduction of their myopia.

In summary, the data presented to you on the outcomes in this PMA serve to establish the safety and effectiveness of the myopic ICL for its intended use in myopia from minus 3 to minus 20 diopters. We believe that the concerns raised by the FDA and panel reviewers can and should be addressed. To this end, we are committed to long-term surveillance of the study population with regard to endothelial cell analysis. We also believe that a comprehensive training program is an essential part of achieving successful outcomes with the ICL and plan to require formal training and certification for all surgeons who use this device.

Finally, we believe that labeling can be developed to adequately communicate the risks as well as the benefits of the ICL and we welcome labeling recommendations from both FDA and panel. This will allow surgeons and patients to make informed decisions on the use of the ICL and the appropriateness of this device for each individual patient. We believe that the data presented to you today and the safeguards we are proposing in terms of long-term patient surveillance, surgeon training and adequate labeling support a panel recommendation for approval of the ICL as an important option in the management of myopia. Thank you and this concludes the formal presentation by the sponsor.

DR. WEISS: I'd like to thank the sponsor for their presentation and if they'd remain at the podium, we will begin for questions from the panel to sponsor on their presentation. Dr. Macsai?

DR. MACSAI: My question is directed at Dr. Edelhauser. The slide you showed of the 37 patients, the standard cohort of endothelial cells changing, on the next slide you said you would address the coefficient variation confidence intervals and that slide was not for that 37 patient cohort. This is new information and I think that data would help us figure out more information about the endothelial cells.

DR. EDELHAUSER: Yes, I'd like to turn this -- this was data that came back to us.

DR. WEISS: Please, would you be able to identify yourself each time you speak in the mike for the transcription.

DR. EDELHAUSER: I'm Dr. Edelhauser. This data came from Dr. Gray, the statistician from the FDA when he sent his review back to STAAR where he then broke out and calculated this cohort of patients from the start or the pre-op all the way to four years.

DR. MACSAI: But what is the -- this is Dr. Macsai.

DR. SANDERS: Dr. Gray did not include that in --

DR. WEISS: Please identify yourself.

DR. SANDERS: Dr. Sanders. We used the analysis that Dr. Gray provided us on the Internet and it did not include the confidence intervals.

DR. MACSAI: Dr. Macsai speaking. But does STAAR have the same patients followed from pre-op all the way through to four years, those 37 patients? Do you have that information, can you provide that information to us?

DR. VUKICH: We do have those patients and again, this is an analysis -- I'm sorry, Dr. John Vukich. We do have that analysis available and can provide that to the panel.

DR. MACSAI: Thank you.

DR. WEISS: Dr. Grimmett?

DR. GRIMMETT: Sure, Dr. Michael Grimmett. I have a number of questions as you can well imagine. The first one to Dr. Edelhauser; I really appreciated your review of the endothelial morphology data and I would just like to ask you regarding that data of endothelial stress, has it ever been stratisfied (sic) by corneal age, that is do younger corneas have a blunted endothelial morphometric alteration as compared to old corneas with less endothelial cushion or reserve?

DR. EDELHAUSER: Dr. Edelhauser. The best data stratification that I can think of to answer the question is the data that we published in `84 on the diabetic corneas. In there we broke it down in terms of decades. And indeed, if you look at the bar graph that is published in that paper, you will find that there is -- as one ages, there is both a progressive decreased in percent hexagonality and an increase in coefficient of variation, so they -- as the cornea does age, you know, you see these changes and that's in a diabetic population, you know, compared to controls.

DR. GRIMMETT: Okay, Dr. Grimmett again. Then can you infer that a younger cornea, because if its higher reserve, higher cushion, will have a blunted response in terms of hexagonality and coefficient of variation?

DR. EDELHAUSER: Yes, I think you can. I think the corneal endothelial cells are certainly more robust in a younger population and I certainly have seen this in laboratory studies where for example, if calcium free solution is placed on a corneal endothelium and you break the endothelial junctions, the -- in an older cornea, you know, about 40 or so, those junctions won't come back in an in vitro situation but they certainly will with younger tissue.

DR. GRIMMETT: Okay. Dr. Grimmett again, just as a reminder, this study ranged to age 21 or so up to 45 and an average age in the 30s I believe. So from the discussion we've just had, this particular cohort may not show as much change in morphometric parameters as a 60, 70-year old cornea, something like that.

DR. EDELHAUSER: Dr. Edelhauser, that's true.

DR. WEISS: I just had a follow-up question as far as that goes. For a patient who's destined to develop corneal edema from continued cell loss, would you say 100 percent of the time they're going have the first sign as a change in the percent hexagonality or coefficient of variation? Is that always the first sign?

DR. EDELHAUSER: From our experience, yes, you see this and let me just illustrate it in terms of patients who undergo cataract surgery for example, the -- when the percent hexagonality and the coefficient of variation start to come back or the cells become more regular, the chances of that cornea going onto a post-operative corneal edema are very much less, so you do see that once stability is established, you do have a normal functioning corneal endothelium

DR. WEISS: But just in relationship to Dr. Grimmett's point, in a younger patient, it would be -- those changes might be more subtle but would they always be able to be picked up, do you think, as a first sign?

DR. EDELHAUSER: They might, but don't forget, this would have to be done with specular microscopy and when you are sampling the cornea, you are taking central corneal endothelial cells in a very, very small population, small area. I mean, you're roughly counting 100, 150 cells and are looking at the endothelium of that out of a population say of 450,000 cells. So you may not pick it up and certainly our past studies have shown that you do see changes in the superior region if you do cataract surgery there. You'll pick that up in the peripheral area very readily where you have damaged the endothelium.

DR. WEISS: So it's possible in a younger patient there might be a subtle change in these -- in the coefficient of variation of the percent hexagonality which might not initially be picked up but then later on as things developed got picked up and that could lead to corneal edema.

DR. EDELHAUSER: Possibly, yeah, and I mean, it goes in hand in hand with total cell analysis, too, because you know that corneal decompensation is going to occur somewhere between 500 and 800 cells per millimeter square.

DR. WEISS: Thank you. Dr. Sugar, Dr. Bandeen-Roche, Dr. Matoba and then Dr. Mathers.

DR. SUGAR: Two things. One is a comment on what Dr. Edelhauser said and what he said in his presentation. Certainly, you didn't measure the peripheral corneal endothelial cell densities in any of these patients and presumably the trauma was greatest in the periphery, so that it's conceivable that the central measurements are a distant reflection of what really counts. And I agree that the increased hexagonality and the decreased coefficient of variation over time implies that the endothelial cells in the center are doing better, but you don't -- your reassurance from the data on the periphery is not specifically appropo of this study because you didn't look at it, correct?

DR. EDELHAUSER: Yes.

DR. SUGAR: The other issue is, I guess for John Vukich. In terms of the powers of the lenses -- I assume we can ask about anything just stick with endothelium.

DR. WEISS: Dr. Sugar, you can ask about anything you want.

DR. SUGAR: I'll limit myself.

DR. WEISS: And that applies to everyone else on the panel.

DR. SUGAR: Okay. When you started this study, did you know that the powers of the lenses that you had were insufficient for totally correcting the patient population that you were investigating? And is -- if that is so, is there an engineering reason or a reason why you didn't have lenses of higher power to correct what you wanted to, that is are there thickness limitations, optic size limitations that keep you from having a higher power?

DR. VUKICH: At the time of the initiation of the study, we had anticipated that we would be able to correct the full range. It became clear that at the higher powers the effective power within the eye was less than the engineering estimates and at that point. Due to manufacturing limitations we found that we could only manufacture at that time up to a minus 20 lens but the effective power within the eye was approximately 16 to 16-1/2 diopters.

At this point those manufacturing limitations are not longer applicable but, of course, that wasn't germane to this clinical trial.

DR. SUGAR: And one other, you said anything?

DR. WEISS: Yes.

DR. SUGAR: You talk about repositioning lenses and you talk about sizing. Repositioning lenses was for haptics that went in front of the iris, for lenses that propellered, what was that and the sizing, are you talking about vaulting or are you talking about something that doesn't go -- that isn't sufficiently long to be stable or so long that it causes iris pombe (ph) or some other problem?

DR. VUKICH: There were four eyes that underwent repositioning. Two of these were for a haptic that was malpositioned, not anterior to the iris but appeared to be folded without flap presentation. One of these was a rotation or actually a decentration, a slight decentration that was recentered without removal and then finally there was one eye that had an edge and one side that captured the pupil in the perioperative area, periopterative period that was readjusted.

DR. SUGAR: Did any lenses ever propeller? Were they ever small enough that they rotated?

DR. VUKICH: No, we did not observe rotational changes in any of our patients throughout the course of the trial.

DR. WEISS: Dr. Bandeen-Roche?

DR. BANDEEN-ROCHE: Karen Bandeen-Roche, and I have a few questions about the specular microscopy. First is a clarification question, so there were 67 eyes followed to four years. As Dr. Grimmett pointed out, two separate 57 patient cohorts preop to four year and three year to four year, and so by my calculations that leads to 47 patients at baseline three years and four years and then two 10-patient cohorts that missed either baseline or three years.

And I just want to make sure, by my calculations, the -- and you may need to get somebody to check on this, the mean cell density in that 47 patient group was 2496, in the group that did not have the three-year visit, 1779 and in the group that did not have the baseline visit 2269 or I guess rounding up to 2270. And so can someone check whether that's correct or --

DR. VUKICH: We will look into that and have an answer for you.

DR. BANDEEN-ROCHE: Okay, now, three quick other questions. First, regarding the plot that Dr. Macsai asked for, what would also be very useful would be to have a plot just like you showed for the 37 patient cohort along with overlaid on the same plot, the patients who had three-year data to just compare. Do you know if it's possible to show the panel something like that?

DR. VUKICH: We do have that available and can give that to the panel as well.

DR. BANDEEN-ROCHE: Okay. I'm interested in how representative the patients with four-year data are of the entire cohort. So that's part of what the first two questions were getting at. Could you tell us the number of investigators who contributed to the 67-patient cohort and anything else that would help us about how representative they are besides the anterior chamber depth which we already know about?

DR. VUKICH: These eyes were done as a sub-study and the number of investigators that actually contributed again, I'll have to look up that particular number for you. There were 12 sites that did participate in the entire trial, however. Nine actually did the specular microscopy.

DR. BANDEEN-ROCHE: Right, and so the number who actually had four-year data, that would be helpful.

DR. VUKICH: Four-year data and we'll get that information. I'm sorry, I don't have that with me.

DR. BANDEEN-ROCHE: Okay, thank you. And finally, I guess a question for Dr. Edelhauser. Certainly an unlimited amount of cell loss would not be benign. I mean, could you give me an idea for the degree of cell loss that would be of concern and that would be expected to cause stress independently of hexagonal cells or CV?

DR. EDELHAUSER: Well, if we go back and look at the literature, the data that we have in the literature, for example, says that in a normal population, not a high myopic population, the cell loss per year is .6 percent, and that seems to be consistent, say .6 to 1 percent per year, which goes. The only other comparative data that I can think about as we -- and this is not really the best comparative data, is the data published from Bill Bourne, and this is 10-year data that he has published with various types of intraocular lenses. He's used three different types of lenses. The only trouble with this is that his average age population was 70 at that particular time and he used a medallion iris suture lens. He used a trans-iridectomy clip lens and he used a posterior chamber lens. He could show no difference in cell loss in any one of those three and the cell loss ranged from 2.8, 2.6 and 2.9 percent. So that's kind of the upper level where we do know that if you have that type of cell loss that you still have clear cornea in a 70-year old population.

DR. BANDEEN-ROCHE: Thank you.

DR. WEISS: Dr. Matoba?

DR. MATOBA: Alice Matoba. My question goes back to the age issue raised by Dr. Grimmett. You enrolled patients, ages 21 to 45. Could you tell us how you selected 45 as the cutoff point?

DR. VUKICH: That was the recommendation and guidance of the FDA for enrollment and I believe this was primarily to look at issues of aging as a compounding variable in the formation of lens opacities.

DR. MATOBA: And then in your labeling, I notice that the patient information states that you must be 21 to 45 to receive this implant. Does that mean that you intend to limit the use of the implant to patients 45 years or younger?

DR. VUKICH: That is the only age range on which we have data to support the safety of this product and would be consistent with our labeling.

DR. MATOBA: Okay, and then as the patient ages, what do you think happens to the vaulting, amount of clearance that you have as the lens becomes more nuclear sclerotic with age.

DR. VUKICH: Well, we do know that over time we can anticipate an increase in the anterior, posterior dimension of the crystalline lens. We do have information internationally with up to 10 years of experience that suggests that there doesn't seem to be a significant change in the vaulting characteristics which is somewhat counter-intuitive. We believe that there is also an age related change in the ciliary sulcus diameter as well. And so there may be several things going on at once that can influence the characteristics that fit within the eye over time.

Nevertheless, we simply have to accept that as an unknowable piece of information until those time periods have been observed in greater quantities and greater patients have been observed.

DR. MATOBA: But you're saying the information you do have indicates that the clearance doesn't change significantly over time.

DR. VUKICH: Throughout the course of our trial, which clearly is the best controlled, we have no evidence but again, this is only three years but at this point, we've been carefully monitoring this internationally where there has been longer data follow-up but at this point, we have not seen that as a trend.

DR. WEISS: Dr. Mathers, Dr. Schein, Dr. McMahon and Dr. Grimmett.

DR. MATHERS: I have a question for Dr. Edelhauser. If the morphologic change in the endothelium is so sensitive, why doesn't it show something when we know that the endothelial cell count is actually falling by these measurements?

DR. EDELHAUSER: Well, I think that you're dealing with essentially a stable -- Dr. Edelhauser -- a stable endothelium and the way cells in a normal population that you would see. For instance, we do know we lose cells over a lifetime that if we say .6 to 1 percent. We don't see marked changed there either because one, it's an apoptotic change that usually occurs. You're losing a cell. The adjacent cell then slides into then cover up the area and I think that what we're seeing here is just a distribution over the whole surface of the cornea.

I can say that we -- I have seen this and we've published papers on this where if you look at the regional areas of corneas. For example, in cataract surgery, if you look superiorly, centrally and inferiorly, you can see these changes markedly and don't forget, in this study, these -- as these were specular micrographs that were taken.

DR. MATHERS: Is your explanation inconsistent with the concept that you could have progressive stable loss rate of one, two, three percent and have a maintenance of a hexagonality as it would be, because the process is essentially just an accelerated but similar to a normal loss rate. It's just faster, so you'd still maintain hexagonality.

DR. EDELHAUSER: You could. I mean, again, it's going to depend upon the -- you're expecting a change to occur over the total corneal endothelium and that may not be the specific case that we're seeing.

DR. MATHERS: Do you think it would be helpful in understanding what's happening to the endothelium to have images that incorporated more than 93 cells on a given patient? It seems to me that when you're looking at the snapshot of the endothelium and as you pointed out, this is a very small area that you're trying to extrapolate then to the entire cornea. Did you only -- for these readings, did you use the single image for each patient time point or do you use five?

DR. EDELHAUSER: We use single image, single image and to answer your question, yes, it would be but the only way that you can get large field or wide field specular micrographs is either with contact specular microscopy and there's no algorithm to go ahead and automatically digitize that other than tracing cells and putting it into a computer, or more recently, there is the possibility of using the confocal and that's certainly a possibility. That gives you a wonderful wide field.

DR. MATHERS: You mentioned that you -- with Bourne's study on endothelium loss in iris fixed lenses, that sort of thing, that he found a loss rate of 2.7 and it was consistent or that -- and also a clear cornea. You're not maintaining that a loss rate of 2.7 would be able to sustain a clear cornea over a long period of time, I would think. I mean, you're not suggesting that.

DR. EDELHAUSER: Well, if you make the assumption that there's not a possibility of some mechanism to produce more corneal endothelial cells, and I think recent evidence has been shown that ARVO -- that we're seeing is that there is the potential that the peripheral corneal endothelial cells have adult stem cells there. This hasn't been confirmed. There's leading indication that you can measure telomerase activity out there which show -- with telomerase activity you only find in cancer cells and stem cells.

You can see that cells do stain with BrdU which is -- and so I think this is a world of research that is developing about the potential of endothelial cells to be replenished.

DR. MATHERS: But there is a loss rate at which eventually you will run out of endothelial cells, I'm sure, you maintain.

DR. EDELHAUSER: Yes, uh-huh, right.

DR. MATHERS: Thank you.

DR. WEISS: Dr. Schein?

DR. SCHEIN: This is Oliver Schein. I'm going to limit questions or comments to the endothelial area at this time. I remember in 1995 when the first data was presented to the panel on photo refractive keratotectomy there was endothelial cell counts and morphology was performed and the sponsor was pleased and amused to see a large and statistically significant improvement in the morphology from pre-PRK to two years.

And this was explained that the majority of individuals before PRK were chronic contact lens users which effected not the cell count but the morphology and the removal of the contact lens allowed the remodeling that appeared favorable over time. Can you please give us some summary of the contact lens wear in this patient population before the surgery and perhaps speculate on how that might impact your estimates of stabilization in the morphology.

DR. VUKICH: Let me lead off by saying that contact lens wear was common in our patient population. However, we do not have an exact number of contact lens wearers pre-operatively. That was not recorded as pre-operative entry criteria other than they had to be out of their lenses for six weeks prior to the entry exam. So we have to make the assumption the majority were. We believe that to be true but we can't give you the percent.

We do know that these patients actually showed stability not improvement over time. And so that when we look at the morphometric analysis through time, we did not see a worsening with improvement of a simpler stable population.

DR. SCHEIN: But if you're presenting comparison of means, you can't actually determine that. You have to look within subgroups to arrive at such a conclusion. Across an entire population if there is an improvement, that would balance worsening and appear as if there were stabilization.

That gets at a second issue. If a majority of the population were wearing contact lenses, that's the acuity that I'd be most interested in as a baseline comparison. It's kind of a habitual vision. It's the vision the patient enters the trial with. The second issue related to endothelial cell count that struck me was not so much the concern about progressive cell loss but of absolute cell loss when you look at the entire cohort. And again, if you simply present a mean, it doesn't get at the safety issues that we're concerned with.

So I interpreted one table as showing that about a third, slightly more than a third of individuals lost 10 percent or more of the central endothelial cell count comparing baseline to three years. Tell me if I've done that correctly. And about 20 percent lost 15 percent or more. Is that correct?

DR. EDELHAUSER: I'd have to -- Dr. Edelhauser. Don, you'll have to --

DR. SLADE: Steve Slade. While they're getting that, I might just address, Oliver, your point about the contact lens being the habitual vision, that's a good point. On the other hand, this was developed with best spectical visual acuity as a target with FDA guidance and, of course, as a standard for refractive surgery and if you look at the patient's satisfaction rates, they were very high and if they were comparing their post-operative vision to what they were used to in contact lenses, and if that were markedly better and we had reduced them, I don't think the satisfaction rates would have been quite so high.

DR. SCHEIN: You've got me now on digression which I wasn't going to raise now. The satisfaction scale that you use appeared to only have three options which doesn't give a lot of room a very strong ceiling and floor effect with only three options for a response. And there are at least two well validated, available, three kinds of visual function questionnaires directed towards populations like this that I think could be used to get more detail.

DR. WEISS: I would ask -- I would have the sponsor just given the advantage of not having to identify themselves any more because I'm told the transcriptist knows your voice. I would also ask the panel members if we could limit our questions because now we're over. So if we could just get to the cogent points quickly and give the sponsor the ability to answer those. Are there any other questions you have Dr. Schein?

DR. SCHEIN: I'm just waiting for the -- was I correct on the endothelial cell count?

DR. SANDERS: Yeah, I believe you were correct on the numbers. I think one has to remember that that was the cumulative -- you were talking the cumulative loss between pre-op to three years.

DR. SCHEIN: Correct.

DR. SANDERS: Yes. And again, I think we have to also keep in mind that some of that -- given that the counting variability is eight to 10 percent, that that enters into the equation, the numbers you did give us are very similar to what we would calculate.

DR. WEISS: Dr. McMahon, Dr. Grimmett, Dr. Coleman, Dr. Ho, then Dr. Macsai and Dr. McCulley.

DR. McMAHON: Dr. McMahon. This is a question for Dr. Vukich. I believe two and a half percent of implanted lenses were implanted initially up side down. And the majority of those, I believe occurred in the first 10 cases, though in Dr. Grimmett's review he pointed out that a number of them occurred downstream. The question I have, is this an issue of surgical training or is this an issue where the device needs to be more clearly labeled as to which is right, left, to minimize those sorts of things?

DR. VUKICH: There's clearly a learning curve in this, in that half of these did occur early in the experience, within the first eight cases of any individual surgeon. This technique is an important part of our training program. We've identified that if this lens is loaded properly and carefully under the microscope in the cartridge, that we can significantly minimize the risk of an uncontrolled entry into the anterior chamber. And so I believe firmly that this is something that can be controlled and in fact, in my personal experience of having put 90 of the lenses in at our site, not a single one went in up side down. I'm also in charge of the training to address this issue. So I think it is something that clearly is a concern but we believe it's an issue that isn't a matter of identifying the right side up. It's just a matter of doing it properly in the first place.

DR. WEISS: Dr. Grimmett.

DR. GRIMMETT: Michael Grimmett. Dr. Slade mentioned that this study was IDE-approved in 1995. What was the first date of the V4 lens implantation? Quite a bit later?

DR. SLADE: All of the -- Steve Slade. I believe all of this data was before.

DR. GRIMMETT: Correct, `89, `99 something like that?

DR. SLADE: `98.

DR. GRIMMETT: `98, okay. Was gonioscopy performed on any patient in this study?

DR. SLADE: Gonioscopy was performed on all patients preoperatively in this study.

DR. GRIMMETT: Preop, okay. I didn't note it on the clinical study report form or in the PMA materials. You have that somewhere then. We just didn't see it; is that correct?

DR. VUKICH: Dr. Vukich. It was on the preoperative checklist for inclusion in the study and gonioscopy was required for every patient as an entry criteria.

DR. GRIMMETT: Okay, good, but just to confirm it was not performed post-op on any patient.

DR. SLADE: It was not a required examination.

DR. GRIMMETT: Okay, was angle anatomy viewed with ultrasound in the ultrasound sub-study?

DR. VUKICH: Yes, it was.

DR. GRIMMETT: It was, good. Was the data in the PMA somewhere?

DR. VUKICH: It was not.

DR. GRIMMETT: It was not, okay. Of the up side down lens insertions that we just heard about from Dr. McMahon, were they related at all to using the plunger versus the screw injector style, like 13 used plunger and 3 used the screw injector?

DR. VUKICH: We did look at that as a variable and we are unable to look at any evidence that the screw -- that the actual injection mechanism itself was a factor. Again, we firmly believe that it was how the lens was loaded in the cartridge as opposed to how it is pushed through the cartridge.

DR. GRIMMETT: Okay.

DR. SLADE: It might be worthwhile -- the lens, it's very apparent, the lens because of the vault, because of the markings before you load is which is right side and which is not. And the cartridge is the same whether you use it with the screw type injector or the plunger.

DR. GRIMMETT: Okay, all right. Your materials indicate your white-to-white measurements had an accuracy of a tenth of a millimeter. My Castroviejo's calipers in the OR have an accuracy in one millimeter increments. What calipers were you using to get .1?

DR. VUKICH: The same ones you are.

DR. GRIMMETT: Oh, okay.

DR. VUKICH: Calibrate them against the steel rule under a microscope.

DR. GRIMMETT: But they're only one millimeter increments. So any unit underneath one millimeter is a shear guess; isn't that correct?

DR. VUKICH: There would be an estimate, yes, below that level, correct.

DR. GRIMMETT: Okay, I've used them and, boy, when I want them at 3.3, it's awful hard to set it at that. And I guess my last question I'll make it, I'll just skip some of these, we'll get to it later, in your materials you stated that your version 4 lens has an additional .13 to .21 millimeters of vault compared to version 3. And I was just curious, did you substantiate that by in vivo measurements or was this a design parameter and you postulated it or how do you know that?

DR. VUKICH: It was a manufacturing and design parameter. This is an engineering issue. External to the eye, this would be the vault that was designed.

DR. GRIMMETT: Okay, thank you very much.

DR. WEISS: Dr. Coleman?

DR. COLEMAN: This is Dr. Coleman and I have a question about two of your subjects developed glaucoma in this study and I was questioning how you define glaucoma. Was that based on optic nerve changes or visual field loss?

DR. VUKICH: Actually, two of our patients were treated with beta blockers, neither of which showed optic nerve changes or visual field changes. So I think they would be best categorized as ocular hypertensive so there was not the diagnosis of visual field loss or glaucoma as we would classically define it.

DR. COLEMAN: You might want to change that. In addition, you said that you did do angle morphology via ultrasound but that's not available in the PMA. Is that --

DR. VUKICH: That angle was observed. There was a sub-study of forty patients that were observed with a P40 unit, a Paradigm unit and we did look at angle morphology. There is within the PMA a description, pictures as well as a commentary of the results of that study.

DR. COLEMAN: Okay, and then in terms of post-operative gonioscopy was not done, not even in the subjects who were diagnosed with glaucoma or is that not available or --

DR. VUKICH: That information was not a required part of the post-operative follow up and I do not have that information as part of the PMA.

DR. WEISS: Dr. Macsai?

DR. MACSAI: My questions are mostly to Dr. Edelhauser again. Sorry, Hank.

DR. EDELHAUSER: No problem.

DR. MACSAI: When this lens is inserted, the most damage to the endothelium should occur in the periphery. If the surgeon is following the technique, they're not supposed to go anywhere near the central cornea. It's defined as a no-touch zone. The manipulation of the haptics is done way at the periphery and they're tucked under the iris with a little lens manipulator, haptic manipulator device as Dr. Slade showed us in his slide. So given that and the presentation you've said about how the peripheral endothelial cell counts is greater than the central in your well-established published articles, and the fact that we're talking about implanting this in 22-year old patients, I have some level of confusion I'm asking you to help me with.

In a guidance draft from a meeting in 10/02 accepted endothelial cell loss rate was 1.5 percent, yet in an ANSI document that I think is also a draft, it was set at two percent. So let me ask this question to you. In your 22-year old child that is being implanted with this intraocular lens because they are minus 13 and too thin for LASIK and contact lens intolerant, what is an acceptable cell loss rate, not coefficient of variation, so that when they are 82 they still have a functioning endothelium?

What is the number? Is it .9? Is it .6? Is it 1.5? Is it 2?

DR. EDELHAUSER: Good question and it's hard to come up with a number because the thing that I think would be ideal to answer that question would be is that if we had a longitudinal study of high myopes and looked and actually measured the cell loss, I think this would be very important. Unfortunately, this is not in the literature. So the -- what you -- in order to answer that question, you know, it has jumped around between 1 to 2 percent as to where we stand with it. One can do all kinds of mathematical calculations to see, you know, how many years would the endothelial cells be depleted, half percent, one percent, two percent, and this all assumes a linear decline which I don't think is completely accurate at this particular stage based on the new information.

So to answer your question, I don't know what the exact level would be with this without some good longitudinal data to be able to make an absolute judgment on. And I think one of the things that all of the endothelial studies suffer from is that we don't have good epidemiological data on various populations for the corneal endothelium with regard to aging and various types of subsets like high myopia for example.

DR. MACSAI: Well, given that lack of security and an absolute number, you know, I'm wary of creating another closed loop anterior chamber IOL disaster that I think most of the cornea surgeons in this room experienced. So what do you think -- I mean, is there a problem with vault and does that correlate with endothelial cell damage? We saw laser flare meter data, not fluoroscopy data, and it looked good but I've posed to the sponsors and I continue to have this concern, vault is good because cataracts are bad but does vault cause posterior chamber -- I mean, posterior iris chafing? Does it release pigment? We don't know. We haven't looked for sample EC lines on gonioscopy.

Does that cause some chronic inflammation that over the 60-year life expectancy of this 22-year old, may effect their endothelial cells. Someone needs to, you know, provide some data from the sponsor regarding this concern.

DR. VUKICH: Well, there's two things that I think we are putting together. One would be the initial -- that could account for some initial cell loss

DR. MACSAI: And then to propose there's an increased rate of loss, there has to be some ongoing irritation to accelerate above baseline. That ongoing accelerated rate, we believe, would be consistent with the morphometric analysis. If we're going to see some sort of insult, whether it be inflammation of which we detected none, whether it would be a mechanical of which again, we would have to postulate some contact with the cornea that we simply have not observed. These chambers have remained well-formed and we have, again, not seen a mechanism by which we can take a posterior chamber lens and equate this into ongoing corneal endothelial trauma. We would really have to propose a new mechanism for a chronic ongoing accelerated loss of endothelial cells that takes into account normal morphology and no other known cause of this accelerated loss. We believe a lot of what we're seeing here is just an extended remodeling period. We have some insult similar to what we'd expect in clear corneal cataract surgery and there is remodeling that stabilizes the population back to its, again, normal redistribution and that, we believe takes as long as three years and we simply can't see an accelerated rate. So I think projecting is difficult but we've certainly accepted the limitations of the data we have and are committed to long-term follow up. It's an important issue and I think it needs monitoring.

DR. MACSAI: Have -- you know, when you talk about long-term insult, my concern is not lens corneal touch. My concern is lens iris touch. My concern is that you know, pigment release and has the sponsor in some way separated those with the good high vault, segregate those out, look at their flare meter, look at their angles, look at their transillumination defects, and look at their endothelial cell loss, that particular group, because I think that would help answer the question.

DR. SANDERS: Well, we do have data on three lenses were replaced because they were too long, which were the highest vault and if you look at the final endothelial cell densities, 3300, 2400, 2700. They were the highest cell densities at the later time periods so it appears that these cases are not the ones that demonstrate cell loss with time.

DR. MACSAI: But they were replaced.

DR. SANDERS: Yes, but they were replaced after a fairly long period in the eye.

DR. SUGAR: Can I ask a clarification from Marian? Are you implying that pigment release causes endothelial cell loss because I'm not aware of that?

DR. MACSAI: I'm not implying pigment release causes endothelial cell loss. I'm implying pigment release implies touch. Touch may insight chronic inflammation and may have some role in this. I don't know. I ask the questions of the sponsor because I don't know.

DR. WEISS: Dr. McCulley and then Dr. Ho.

DR. McCULLEY: I've been around -- Jim McCulley. I've been around since prior to the beginning of clinical specular microscopy. Been through decades of frustrations of trying to listen to people make sense out of and make points based on cell density. And having listened to -- read everything that was provided, having listened to what everyone has said, quite honestly, I'm at a point where it seems to me that what you've presented at least my interpretation of it, would be that we have surgical trauma, endothelial cell loss, and no evidence for anything except continued remodeling. And no evidence for any other mechanism for continued endothelial cell loss or death other than the normal apoptotic death. So I'm not sure where, you know, one could go further with this or what we would ask you to do other than the surveillance that you're doing except to ask is there some other more sensitive way of looking for inflammation which wasn't a part of your PMA. So I'm not even sure how fair that question is.

DR. SANDERS: With regard to the inflammation, ocular inflammation was the subject of my PhD thesis so I did quite a bit of work in this area and that's why we included in the PMA five -- I mean, the laser cell flare meter has been basically thought to be too sensitive a measure of inflammation and it's not even allowed for inflammatory studies because it's too easy to show a decrease in inflammation between groups, and five separate studies in the published literature have shown no inflammatory response after the early post-operative period with this implantable contact lens.

DR. McCULLEY: Well, then I guess what I would hope is I envision potential hours of discussion about small points relative to endothelial cell loss and cell density in something that is less than an ideal science. So I would hope that the panel would really press Hank, who is the world's expert in my experience on endothelial specular microscopy with any other issues rather than us trying to figure out what's what among ourselves. If we can have -- so I guess my plea is -- to the panel is, please press Hank while he's here to give us the information that will be more expert than we're at to be able to generate amongst ourselves and hopefully have a more efficient discussion of this because to me this is surgical trauma remodeling.

DR. WEISS: Dr. Ho.

DR. HO: Allen Ho. But is there any evidence that this sub-clinical inflammation has a deleterious effect on the cornea?

DR. VUKICH: We have not demonstrated any subclinical information, no.

DR. HO: Is there anything in the literature?

DR. McCULLEY: I think -- Jim McCulley. They have no evidence for subclinical inflammation and depending on how you define subclinical which presumably would be what we see at the slip lamp, they've gone another step forward, don't have any. What we could do would be again, intuition. My intuition tells me what I've said. It would be intuitively to go back to some of the closed-loop AC IOLs that actually didn't have sub-clinical, they had clinical inflammation that led to loss of endothelium, so I'm not sure that maybe in some of those eyes some of us didn't see the cell and flare that was going along with those AC IOLs but I think if you have chronic inflammation or chronic rise in intraocular pressure, there is data that suggests there is endothelial cell damage over time.

But we don't have any of that there and that's one of the things that intuitively leads me to my conclusion, we have no proposed -- we have no support for any mechanism for any continued endothelial cell loss beyond the apoptotic aging.

DR. WEISS: Yeah, I would prefer if we could keep the panel discussion in the panel discussion portion and keep the questions while the sponsor is up there because we have limited time. Do you have any other questions specifically for the sponsor?

DR. HO: I do. The only patient that had == Allen Ho -- that had severe sustainable loss of vision in this trial was a patient who had a retinal detachment and in a group of very high myopes we would expect perhaps without intervention by natural history that you might see retinal detachment.

However, one of the predisposing factors to retinal detachment in high myopes is clearly retinal breaks and lattice degeneration. Do you have any data about number one, lattice degeneration retinal breaks pre-operatively and was indirect ophthalmoscopy part of the study procedures pre and post-operatively?

DR. VUKICH: A dilated funduscopic examination was required at several intervals throughout the follow-up period and detailed information was collected by the investigators specific to peripheral retinal findings. We don't have that collated specifically but also entry criteria did require a stable retinal exam. Any pre-existing holes or tears or retinal changes that would be considered high risk, of course, were excluded.

DR. HO: They were excluded. Stable retinal breaks were included in this or were they treated preoperatively with laser, for example?

DR. VUKICH: We do have a patient, I believe, who had -- we had one patient was treated for an acute retinal break.

DR. HO: Yeah, I think that's really important to flesh out for a potential consumer of this kind of technology because, you know, that's where you're losing an eye. However, you may not lose that eye based on your intervention. It simply may be natural history. So I think that's -- I would like to see that information. Thank you.

DR. WEISS: Do you think the optic size of 4.65 had any impact on visual acuity in younger patients who had larger pupils or is this something you didn't look at?

DR. VUKICH: Well, visual acuity and quality were two different things. The visual acuity didn't seem to have an impact in terms of the improvement in best spectacle corrected acuity. Those were the patients who actually had the most improvement quality of vision by subject of symptoms. We can stratify that by level and can provide that, yes.

DR. WEISS: So you would be able to look at the size of the pupils to see if it had any adverse effect. Dr. --

DR. VUKICH: Well, excuse me, let me qualify that by saying, pupil size measurement was not a part of this clinical exam, either preoperatively or during the course of the trial, so we could only stratify it by level of myopia, not by pupil size.

DR. WEISS: Okay, so that's an unknown factor.

DR. VUKICH: Correct.

DR. WEISS: Dr. Grimmett?

DR. GRIMMETT: A quick one for Dr. Sanders. Campbell estimated that pigment particles can be as small as one micrometer in size. Does that laser cell meter detect particles that small?

DR. SANDERS: Yes, it does. The standards that are used are in the two micron range and those are meant to be certainly large enough. One micron sized particles should be detected by the Kowan machine.

DR. GRIMMETT: You had about 20 patients after the three-month period or so, 25 or something like that up to two years, something like that.

DR. SANDERS: Correct, and the cell measurements were essentially below one per area that was seen on average.

DR. GRIMMETT: In those 20, okay, thank you.

DR. SANDERS: Yes.

DR. SLADE: Yeah, Steve Slade, one quick point to address Dr. Macsai's concern about the vaulting, I just want to make it clear that while angle examination, gonioscopy, was not part of the exam, we certainly did slit lamp exams at multiple intervals and at no point did we ever find peripheral touch, so we were looking at grading angles in that fashion and at no point was the vaulting such that it actually caused touch or PAS.

DR. WEISS: Thank you. Dr. Mathers?

DR. MATHERS: For Dr. Edelhauser, if you're going to postulate that remodeling is the process, it might be helpful to know -- to see these cells and watch them remodel because they're not being created. They've got to be out there. Could you help us by letting us know how many cells you like to see on a cornea to understand the remodeling process. You're looking at 93 here. What would you recommend that we try to look at if we're going to actually understand if remodeling is the issue versus cell loss on a given patient?

DR. EDELHAUSER: I think that one, it's important to do more than -- if you want right now the information, more than just central specular microscopy. Obviously, if we have these pooled cells out in the periphery, it would be interesting to see what's happening with those. I mean, and to get a larger cell number, now the -- most of the instruments that we used in specular microscopy you're limited to pretty much about four millimeters in the center, unless you really encourage the patient you can get out to maybe four millimeters off center to look at the periphery. It's not an easy measurement to obtain.

DR. MATHERS: But there's a half a million cells in that area, so --

DR. EDELHAUSER: Yeah, so I mean, one -- if one had to say predict the ideal way to really evaluate it, is I think some of the ways that we -- that article we published in the AJO is that we did take eight or nine readings across the cornea; one central, four paracentral and for far peripheral and then if you do that, you can -- and then the interesting thing when you do that, Bill, is that you find out that there's a higher percentage of corneal endothelial cells in the superior region. And similarly the German Daus all found the same thing. So you have a 16-percent increase in peripheral endothelial cells in the superior region.

DR. MATHERS: Would you recommend that -- matching that against controls as a means to obtain this understanding?

DR. EDELHAUSER: Well, if we're going to really map out what's happening in the cornea, with any type of surgical situation with remodeling one would have to do that.

DR. WEISS: Dr. Bradley has one brief question. I will ask a question and then we're going to have a 10-minute break.

DR. BRADLEY: You might need a 10-minute break after my question. I'm bringing -- I'd like to just go back to the issue that Dr. Weiss raised a few minutes ago about pupil size. There seems to be a certain irony here. I mean, one of the motivations for the product is that there are certain people out there whose myopia level is too high although cornea too thin to perform LASIK simply because -- perform LASIK and have the standard 6.5 millimeter diameter optical zone.

The replacement product is only having potentially a 4.65 millimeter optical zone. And one of the reasons why we have a large optical zone with LASIK is because we are concerned about pupil size issues. And I'm a bit concerned that we have so little information about pupil sizes of these patients even -- we would anticipate for example, with young adults mesopic light levels that at least half of the light would be passing into the eye outside of the optical zone of the ICL.

Under those circumstances, one can only imagine that the image quality would be very poor. Having said all that, the data seems to point that the patients are quite happy with their nighttime driving, your mesopic contrast sensitivity test with a glare source showed perfectly good results and I'm completed confused by that. I wonder if the sponsor could clarify how that could possibly happen with such a small pupil size.

DR. VUKICH: Well, we'll start by looking at pupil size. Certainly, when we developed the protocol in 1995, I don't believe that the interest or the understanding of how these pupil sizes could interact with optical quality were fully understood. That said, pupil size we neither an entry criteria nor a parameter that was measured throughout the course of the trial. I think the only way that we can answer that is to go back to the patient's satisfaction surveys and the quality of vision that they report inasmuch as the patients, in fact, didn't seem to be bothered by the theoretical concerns of an optic size smaller than their pupil. Of course, they didn't know this but what they saw they seemed satisfied with.

I understand and appreciate the concerns even with pupil size. However, there seems to be some variability in the response or the effect of the pupil size that we're understanding now with LASIK where it may not be as much of a correlation as we perhaps, intuitively may expect. So we don't understand the mechanism why a smaller optical size at the level of the lens inside the eye may not have as much influence but yet, we simply have to go back to the results and I believe that they are consistent with patient satisfaction and with the use of this device.

To speak to vision quality, there was a subset in a published report looking at vision quality in patients looking at induced aberrations and we found post-LASIK versus ICL, that the ICL patients had one-third as much spherical aberration and half as much coma. And so we certainly believe that it's at least in comparison to LASIK, probably better in that regard at least.

DR. WEISS: One last question and this is sort of a bottom line question for Dr. Edelhauser because it seems that the main concern of the panel is the impact on the endothelium. Would you be surprised if this lens was a contributory factor in causing corneal edema in any of the patients on whom it was implanted?

DR. EDELHAUSER: At this stage, no, because the cell density of these patients were well above, you know, 23, 2400.

DR. WEISS: I should say eventually. If any of these patients eventually developed corneal edema, in conjunction with having this placed, would that surprise you or do you think that would be totally independent of having this lens placed?

DR. EDELHAUSER: Well, when you think about having a lens behind the iris and not rubbing onto the corneal endothelium, it's hard to imagine, you know, the mechanism of what would cause this -- a marked decrease in corneal endothelial cells.

DR. WEISS: So you would be -- that as a complication would be surprising to you even 20 years down the line.

DR. EDELHAUSER: Yeah.

DR. WEISS: Okay.

DR. SLADE: Just one quick thing, this lens has been implanted outside the U.S., tens of thousands of cases over 10 years and while reporting that experience is not FDA quality, I do believe we would know if this lens ever created a corneal decompensation if the patient had to have a graft and we know of none in that experience.

DR. WEISS: Thank you. We're going to take a 10-minute break and I'd ask you to be back here promptly and then we're going to go onto the FDA presentation.

(A brief recess was taken.)

DR. WEISS: Donna Lochner will be introducing the FDA presentation.

MS. LOCHNER: Thank you, Dr. Weiss. Because this is the first phakic IOL to be brought before the panel, I would like to briefly present how FDA's guidance to industry on the design of phakic IOL studies has evolved beginning with the October `98 panel meeting. In 1999 ANSI standards and later the ISO meetings began and they currently are held every six months or so. Both the ANSI and ISO standards are expected to be submitted for voting in 2004.

Today I'll provide just the highlights of the three panel discussions and then summarize the current ANSI and ISO standards which have incorporated all the major recommendations of the panel with some minor exceptions. FDA issued a draft guidance document in 2000 and expects to issue a final guidance when the ANSI standards are finalized. So this first slide -- I think I went -- this first slide is for the October 23rd, 1998 meeting which, as I said was the first discussion by the panel and at that meeting, the panel recommended that effectiveness criteria generally followed the refractive laser guidance. For example, with respect to the uncorrected VA loss of BSCVA, and also recommended that adverse events in the first year should generally follow the IOL grid for aphakia as a starting point for the study design.

The panel recommended a sample size of 500 subjects and this was primarily because they felt that as a new indication, new technology, they should take a more conservative approach and the 500 subjects was consistent with what was originally done with IOL aphakia studies. Further, they recommended mesopic contrast and sensitivity testing be done and mesopic pupil size measurements be done. That a questionnaire for visual complaints be administered and that pachymetry, dilated lens and fundus evaluations, topography, keratometry and gonioscopy evaluations be performed.

With regards to specular microscopy, the panel recommended a sample size to allow detection of 2.5 percent per year and they obtained this figure from the Bourne article that was referred earlier in the discussion this morning. There was a suggestion that all patients be tested but they felt that FDA should try to power the studies to detect the 2.5 percent per year. They felt PMA data was needed to three years and if there was a loss or the loss was progressing, a five-year study should be performed. With respect to lens opacities, the panel recommended a clinical grading system and three-year data be collected.

The May 12th meeting was held to receive the panel's input prior to publication of FDA's draft guide and at that meeting, the panel generally endorsed our proposals to power the studies to be able to detect a 1.5 percent loss in the specular microscopy study per year and the 1.5 percent figure came after iterating several hypothetical annual losses from a phakic IOL taking an average endothelial cell densities at different age ranges from the literature and determining the age at which the hypothetical annual loss would result in corneal decompensation for the various age groupings. From there we assigned a standard deviation of five percent and sort of arrived at -- which was sort of arrived at as being a reasonable loss so that even young adults would be in their 70s prior to decompensation and that the sample size would still remain reasonable for these studies.

The panel endorsed this approach and also asked for data analysis to include a stratification by age. And they further recommended that the analysis look at the mean rate of loss and a frequency analysis to show the percent of patients losing greater than 10 percent over the course of the study. With respect to lens opacities, the panel again recommended a preoperative and post-operative clinical grading system and at this meeting they also -- there was quite a bit of discussion about control group and felt that that was recommended. The panel also again emphasized gonioscopy and dilated fundus exam.

After another two years of meetings with ANSI and ISO we brought a composite of the standards to the panel but with a focused review of endothelial cell density, lens opacity and the contra-sensitivity study. We assigned primary reviewers for each of these three topics and also invited speakers to address endothelial cell design and lens opacity clinical study design issues. The panel recommended that the cell density studies be able to detect the 1.5 percent annual loss and this, again, was based upon entry criteria on cell density and acceptable density for the life of the patient. Depending upon the standard deviation, they commented that this will equate to about 200 to 300 eyes. They recommended use of a central reading center or other methods with similar precision and validity. They recommended the three-year data was needed for the PMA and also that an intermediate measure between the two and three-year point might be needed to help to establish linearity.

Depending upon the three-year data, the panel recommended that additional two years post-marketing study may be needed. And finally, again, the frequency analysis was requested. With respect to lens opacities, again, the panel recommended a clinical rating system and the three-year data also was needed to address the issue of lens opacity and that consideration will be given to longer term at least a five-year post-marketing study. Once a PMA has been reviewed, the panel felt it was useful to look at laser flare and high resolution ultrasound for source of any opacities. And they felt that two or more lines loss with glare or one line without glare would be the level that would be considered clinically significant for any opacity.

They further recommended that contrast sensitivity testing be done on all patients to document the severity of any future opacity. With respect to the contrast sensitivity discussion, the major recommendation that came out of that was that the panel felt a clinically significant decrease in contrast sensitivity should be set at .3 log units and again, the panel emphasized gonioscopy and further stated at this meeting that consideration should be given to collection of data post-market depending upon how the PMA data looked.

Again, as I said, all of this culminated in the current draft ANSI and ISO standards with recommendations for a three-year, 300-subject preoperative control study. Safety end points from the FDA's aphakic IOL grid are also used as control data in these standards and now I'll just briefly go through the current recommendations and the most current versions of these standards and that is that the following evaluations be performed; in corrected CVA, distance and near, BSCVA distance and near, manifest and cycloplegic refractions, a subject questionnaire, a slit lamp exam including aqueous cell and flare, gonioscopic exam, corneal edema, pupillary irregularities, iris atrophy and pigment dispersion.

These standards recommended a dilated fundus exam, that IOP testing be performed, mesopic pupil size be measured and that pachymetry, preoperative axial length, anterior chamber duct measurement and kerotometry be performed. With respect to specular microscopy, the standards assume a 10 percent surgical loss and recommend that the studies be able to deduct a two-percent loss per year. The standards recommend that all 300 subjects be tested so that at least 200 good images would be obtained.

They recommend use of a central reading center and they recommend that 100 to 150 cells be counted. With respect to lens opacities, again the standards recommend a clinical grading system and they recommend that a change in contrast sensitivity performance from preop to each post-op visit at which an opacity is observed be performed to document any significance to the opacity. The standards recommend contrast sensitivity be performed under mesop and mesopic with glare and the sample size recommended is 61 subjects.

Now, I'd like to thank and acknowledge the PMA review team for this application. Dr. Alexander, who is the lead reviewer for the PMA, Dr. Eydelman, the clinical reviewer, Dr. Gray who performed the statistical review, Don Calogero, our jack of all trades who performed engineering, contrast sensitivity and specular microscopy reviews. Susanna Jones reviewed the toxicology. Susan Gouge, microbiology, Charles Sawyer, patient labeling, Pam Reynolds performed the bio-research monitoring review and Vertleen Covington on the quality systems or good manufacturing practices review. And last but not least, I have to give a special thanks to Sally Thornton, who due to the expedited nature of this PMA really had to do above and beyond the amount of normal running around and we couldn't have gotten here today without her excellent support.

Now, Dr. Eydelman will present the clinical questions.

DR. EYDELMAN: Good morning. This PMA is truly precedent setting and I wanted you to be aware of it for several reasons. First of all, there are currently no phakic intraocular lenses approved in the U.S. There are also no currently approved devices requiring intraocular surgery for correction of refractive error. Thirdly, there are no current FDA approved devices for the correction of myopia greater than 15 diopters. In addition, FDA approved IOLs for use only in adults 60 years of age and older until this year.

Currently, responses may require lowering age for indication to all adults by reference to our recent publication. This is the first time, therefore, that you're going to be considering a PMA for an IOL intended solely for implantation in young adults. As you heard, this PMA received an expedited review status. That truly meant much shorter turnaround time for both the sponsor and us. To make a point of it, I want you to be aware that the last major clinical amendment wasn't received by FDA till September 3rd.

As a result of all this, I haven't been able to receive the sponsor's final panel presentation until today, so please forgive any redundancies that I might have in my presentation. As you have all seen, this was a very large PMA with numerous analysis and I will not try to summarize all of it. I'm merely trying to bring your attention to some information which is relevant to the questions that we ask for your consideration.

Regarding lens opacification, there were five eyes in the whole PMA that developed nuclear opacities of two plus at the LOCS scale at two to three years. There were 14 cases of ASC opacities of trace or more. Eleven of them occurred at or before the six months and three cases at one year to 26 months post-op. In view of these, do you believe that the three-year follow up is sufficient to establish a lens opacification profile associated with this device? If not, what is your recommendation?

Eleven out of the 14 cases of ASC appeared at or before the six-month visit suggesting surgical trauma. Combining surgical experience with V3 and V4 models, 50 percent of 87.5 percent if you exclude the problematic site number 15, of early ASC cases occurred within the first eight surgical cases. In the Canadian trial performed by three inexperienced surgeons, 22.5 percent of cases developed ASC opacification.

The Dominican Republic study which was performed under supervision of a surgical proctor, demonstrated a rate of 4.8 percent. In light of these findings, do you believe surgeon experience to be an important factor in ASC development, secondary to surgical trauma? If yes, do you believe that future users of this lens should be required to undergo special training?

Vault measurements in the study were clinical estimates comparing the slit lamp appearance of the corneal thickness to the interval centrally between the crystalline lens and the ICL. Five hundred micron corneal thickness was assumed for conversion from a percentage of corneal thickness to microns. All measurements in an individual case at every visit were averaged to derive at a vault measurement. So as you can see, it was not a very precise measurement estimate. However, it was done.

Patients were graded as having poor vault if investigators consistently graded the space between ICL and crystalline lens as less than 10 percent of the central corneal thickness and that equated to about 50 microns. Twenty-four cases of the V4 cohort with this technique were determined to have poor vault, 16.7 percent of them or four out of 24 V4 cases with poor vault, subsequently developed ASC opacification in contrast only two percent of cases with good vault had ASC.

All three cases of significant ASC opacification of late onset defined as greater than six months in V4 cohort were in the eyes with poor vault. In V3 cohort, 41 percent of cases with poor vault developed ASC versus nine percent of cases with good vault. Gonvers, et al, in his recent publication further supported the relationship of vaulting to cataract information. In the PMA the sponsor recommended replacement of the ICL only in cases of poor vault that exhibited early ASC in areas of ICL touch in subjects with UCVA worse than 20/50. Do you agree with this recommendation? If not, what would you recommend?

In the clinical trial, sizing was determined by the horizontal white-to-white and ACD, anterior chamber depth measurements. Inherent measurement error associated with caliper measurements was judged by the sponsor to be plus or minus .1 millimeter. Anterior chamber depths in the study was measured by ultrasound, Orbscan and IOL master. From the literature review, the sponsor concluded that results may differ by as much as .3 millimeters between different measurement methods.

Our own literature review revealed lack of correlation of white-to-white measurements and the sulcus-to-sulcus dimension. We also believed that the literature shows that none of the external measurements, including anterior chamber depth and axial length, have been able to accurately predict internal ocular dimensions. The sponsor believes that this literature evidence currently available is anecdotal and they further point out that all the safety and efficacy data available were obtained with a current sizing algorithm based on white-to-white and ACD measurements.

It's interesting to note that looking at the distribution of the ICL implanted, 50 percent were performed with 12.5 millimeters, versus 7.6 percent was 11.5 millimeter lens. In the overall PMA cohort, 1.5 percent of the lenses were replaced due to inappropriate sizing. Do you believe that the method currently recommended by the sponsor for determination of the overall diameter of the ICL to be inserted is appropriate? If not, what do you recommend?

As you heard previously, we asked the sponsor to break up their cohort into four refractive groups. Fifteen to 20 diopter group contained 31 eyes at three years. I want to make sure that you're aware that while preliminary discussion for refractive laser guidance for myopia greater than seven diopters was held a the `97 panel meeting. There was no consensus reached on several issues and therefore, there is no currently available guidance for acceptable safety and efficacy outcomes for high myopes after refractive surgery. For eyes with MRSE greater than 15 diopters, in the ICL cohort, there were 3.8 percent or two eyes that lost greater than two lines, 3.8 percent that lost 2 lines and 17.3 percent that lost 1 line. If you calculate it out, it turns out that at 15 diopters of myopia, magnification factor account for a one-line loss being equivalent to a two-line loss. Therefore, we ask the sponsor to include that in the analysis of their high myopia group.

Thus, if you add it up, total loss of one line or greater was 25 percent for the small cohort.

Some additional safety outcomes for these eyes were retinal detachment at 3.8 percent, ASC opacification of 5.8 percent and as of 9/15, only -- the sponsor informed us that only one eye of these was clinically -- had clinically significant ASC and that is 1.9 percent. Clinically significant nuclear cataract in 7.7 percent, ICL removal/cataract extraction performed in 3.8 percent and again, 3.8 percent had an increase of greater than two diopter cylinder.

As you heard, currently limitation of ICL power is minus 20 diopters. Inadvertently a lot of eyes with MRSE greater than 15 diopters were targeted for under-correction. Eleven point five percent of them were targeted for greater than three diopters, 28.8 for greater than two and 65.4 for greater than one. Looking at predictability, 23.3 percent had accuracy within half diopter, 53.3 was within one diopter. Combining the targeted under-correction was a predictability that you saw resulted in rather large range for resultant MRSE for this group at three years. As you can see, it ranged from minus .85 diopters to plus .5 with 10 percent of the eyes ending up greater than four diopter myopia, 26.6 greater than three diopters.

Looking at all eyes with preop MRSE greater than 15 diopters 38.7 percent of them were able to achieve 20/40 or better. There were no eyes available that were targeted for emmetropia and had preop of 20/20 or better. While all eyes in this sub-group were -- while there were no eyes that were -- there were no patients that were unsatisfied, looking at very extremely satisfied patients, you see that for the group of greater than 15 diopters, the satisfaction percentage drops somewhat to 75 percent.

Does the safety and efficacy data for eyes with preoperative myopia of greater than 15 to 20 diopters support approval of this refractive range? If approval for eyes with preoperative MRSE greater than 15 to 20 is recommended, is the term "correction of" as it relates to this refractive range, appropriate in the indication statement? If not, what alternative term do you recommend?

Any time we at FDA consider risk benefit analysis for each of the refractive groups, we have to consider two factors. First, is a safety and efficacy profile for each refractive group with the device in question. In addition, we look at safety and efficacy profile for the currently approved or alternate devices available; in this case, glasses, contacts, LASIK, for each of the refractive groups? With this in mind, does the safety and effectiveness outcomes support approval of STAAR ICL for the eyes with the following preoperative MRSE, minus 3 to minus 7, greater than 7 to 10, and from greater from 10 to 15 diopters? Twenty patients in overall PMA cohort required treatment other than IOP-lowering meds in the early post-op period. Seventeen of them requiring additional irodotomies, and three requiring additional irrigation/aspiration procedure. In these 20 eyes, IOP ranged as high as 65, with IOP spikes observed between one and 21 days post-op. Most of them, however, were seen in one to two days post-op.

Incidents of early post-op spikes was stratified by study site and was shown to range between zero to 20 percent. The differences were not found to be statistically significant. Do you believe that specific recommendations regarding early post-op follow up are needed in the labeling? I want to bring your attention to the fact that the labeling you currently have is not -- did not undergo final FDA's review. We always correct all the inconsistencies. Patient symptoms and quality of vision assessment stratified by refractive groups would automatically be included. Demographics is always included.

What we are asking your input on is issues unique to ICL that need to be communicated in physician and patient labeling, possibly as a warning or precaution. In addition, we're asking you to consider issues that will be common to all phakic IOLs, such as possible requirement for exclusion of subjects with low endothelial cell density as a function of age. This would be consistent with ANSI PIOL draft standards recommendation for clinical studies. It would, however, imply access to specular microscope for all implanting surgeons.

In addition, recommendations for gonioscopy and mesopic pupil size assessment preop and post-op in all patients. This is consistent, once again, with our standards recommendation for all clinical studies. Overall, we want to know what additional labeling recommendations do you have. Now, I would like to introduce Dr. Gerry Gray who will review all of the endothelial cell data analysis and when the Chair is ready, I'll be happy to project all questions as they appear in your handout.

DR. GRAY: Good morning. My name is Gerry Gray. I'm the team leader for cardiovascular and ophthalmic statistics and I was the statistical reviewer for this PMA. My comments are going to be restricted to the specular microscopy sub-study. This is an overview of the design. We've heard it several times. We're talking about endothelial cell counts and measurements on endothelial cells based on photographs from a specular microscope and all the images were read at a core center with one reader.

The study was originally designed to have a preoperative and then three-month, one-year, two-year follow up. During the course of the study it was modified to add three and four-year visits. And the purpose was to investigate the effects on endothelial cells. There were a total of 306 eyes that were enrolled in this sub-study and it had at least one count. I'm just going to go through a little bit about the accountability of the eyes because it gets a little confusing here.

The pattern of missing is not quite standard where everyone has a preop visit and then people start to drop off after that. It's a fair amount different. In fact, there were -- 94 of the 306 patients had no preoperative visit. Six people had preop and one subsequent. Thirty-four had preop, two subsequent and 172 had preop and then three of them were after that, and the small numbers after that tell you where the person's last visit was.

And all this accountability information is based on a data set that was submitted to me by the sponsor for analysis. So actually, I think it was a SAS formatted data set. A couple of more accountability combinations; 154 patients had preop and three-year visits, 57 comes up a couple of times. It's not the same 57 patients but 57 had three and four-year visits, 57 had preop and four-year visits. A total of 67 people had all the visits up to three years and a total of 37 had all visits up to four years. So there's 37 patients out of these 306 that had all the visits.

So here's a plot that we've seen before. It's the raw results from the data -- from the study, excuse me. The year or the time has been jittered a little bit to show the distribution there. There are preop measurements and then three months, one year, two years, three years and four years. The dashed blue line here just simply connects the means at those time points. And when we look at this, there's really two questions that are key here. The first one is how -- at what point in time can we say that any effect of the actual surgical procedure, whether it would be just lost due to surgical trauma and/or some amount of remodeling, at what point in time would we say that is negligible and we can ignore it and use the data after that to get some estimate of what long-term loss might be? So that's the first key question that we need to think about.

And then the second thing is what happens off to the right-hand side of this graph, what happens after five, 10, 20 years down the road? Just to set the stage a little bit, this is -- these numbers here are the mean cell counts for various cohorts of patients that you might think about using in this study. The first cohort is all eyes. That's just all 306 eyes that were measured whenever, the baseline preoperative measurement, the mean was 2657 and it steadily declined after that to 2355 at the four-year point.

The next cohort, I couldn't fit it in very well, so I call it pre and two plus. Those are all the patients that have a preoperative measurement and then they had at least two measurements after that. So that's 206 of the 306 patients and you can see there, it's fairly similar actually to what we get with all eyes. The next two cohorts are somewhat different. The cohort that only has three and four-year measurements and this is a cohort that in the analysis presented to us by the sponsor for the three to four-year loss they used. You'll note that the main difference here is at the three-year point that measurement of 2355 is somewhat lower and it's actually in fact, lower than the average measurement they got at four years for those 57 patients.

And then finally, an even smaller subset was everyone who had all the visits and that shows a similar pattern to the three and four-year one, and I presume these are the numbers that were used to make that plot that came up in the sponsor's presentation. So over the duration of the study, over the three and four years we're talking about here, the estimates of cell loss are fairly stable regardless of how you calculate them. At three years, the range of estimates is 8.5 to 8.9 percent. If you use the 154 patients who had preoperative and then three-year -- a three-year visit, the estimate is 8.7 percent. And the competence interval for that is anywhere between a 10.3 and 7.1 percent loss. In raw numbers that's 220, 235 cells per millimeter square and that calculation includes anything that happened to the patients between preop and the three-year point which would be any initial operational loss, any kind of remodeling, any normal loss due to aging over that period.

And at four years, we've added on a little bit here and it's anywhere from 8.4 to 9.7 percent loss. Okay, now the big question, of course, is what's the steady-state long-term loss that we can expect to see. What's the long-term rate of change in the endothelial cell density we might think we would see? And it turns out that this estimate depends mostly on those -- on the question of how long we believe the effects of the implantation persist, at what time point can we say whatever remodeling or operative loss we have seen in negligible at this point. And that translates into which of the cohorts we actually used to do that estimation. As you saw in the previous slide, the table of cell densities, the two cohorts on the bottom that only had -- that had three and four-year measurements had a markedly lower three-year cell count than the others and that's the main difference in terms of what you get out in the estimates.

The analysis that was presented to us by the sponsor in this PMA was basically using the percent change between the three and four-year time points, using only those patients who had both three and four-year measurements. That's the 57-patient cohort and it properly did some statistics to account for a correlation within a patient between eyes. And the net result there is an estimated percent change of .07 percent, that is a slight gain. In fact, it was one cell per millimeter squared with a confidence interval between minus 1.4 and positive 1.6 percent.

Now other cohorts you'll recall, have relatively higher three-year counts and you can do a lot of different kinds of analyses but the bottom line is that the various analyses using those other cohorts and using all time points or time points other than just the three and four-year, produce a change of around minus two percent per year. If you go the fancy statistics route and do random coefficients regression, you get a loss of minus 1.9 percent per year. If you believe that whatever -- that the time cutoff for the operational and/or remodeling change is three months and just use the data after three months, and go through an analysis, it's exactly like the one done by the sponsor, in other words, just use the changes from time point T to T plus one, you get an estimate of minus two percent per year.

If you believe that any trauma or remodeling is done after two years and you use the two to three-year difference plus the three to four-year differences, you get an estimate of minus 1.8 percent. And the confidence intervals change a little bit. The one for the -- using the regression is probably the smallest because it has a model to help it make the balance smaller, but those are fairly consistent estimates compared to the difference between them and the one that only uses the three and four-year data.

So the key question, of course, is where is that cutoff between operative and/or remodeling loss and whatever you might call steady-state, long-term loss. And all I can do is statistics, right? I don't have the clinical knowledge but I have the data, so using the data that we do have, the question here from the statistical point of view is we see that there's some amount -- in many of the cohorts, there's some amount of leveling off after the three-year point between three and four years and the question is, is that statistically significantly different than whatever the slope we saw between three months and one year, one year and two years, two years and three years. And the answer to that is no. If you'll recall the previous plot, it showed the dotted line that connected the means, it looked pretty much like a straight line and the statistics confirm that. There's no strong evidence that the rate of endothelial cell loss between three and four years is any different than the rate -- the annual rate before that. So in the data we have, there's not strong evidence that it's different. Of course, we only have 57 people at four years and that could be do to just random fluctuation or we just don't have a big enough sample at four years to have much statistical power but that's what we have.

And just in case you care about the details, this was all based on a piecewise linear model that assumes there's a preoperative loss between zero and three months and then after that, it's steady decline either to three years and then a change to four years or it's straight from three months on. But the implication of all this from the data we have is that is that the steady state loss should be estimated using all the data after three months. And if you'll recall from a previous slide, even if we want to go to two years, it doesn't make that much difference here.

And so my best guess is due to long-term loss would be that we have -- first of all, there's a mean preoperative measure of 2651 and with the first three months, the absolute loss is about 1.9 percent, so about a two-percent loss over the first three months, and then after that, the rate of loss per year is about 1.9 percent.

If we extend this model a little bit to include a three and four-year slope, which again was not warranted by the statistics probably, you do get a pretty similar estimate to what the sponsor had between three and four years of an actual slight gain. So here's the results from the two different fits, the two main different kinds of fit that I'm talking about. First of all, there's a blue line here that's just like the one you saw in the previous plot that's pretty much overlaid by the black line. The black line is the fit that I was describing where we had a linear drop at the three months and then a straight line after that. And the green line out at the end is the analysis that was presented to us by the sponsor which is just using the patients who have three and four-year data. And you look at this plot and you say, well, that's not that much different because you know, the only thing different is maybe the difference between the mean at three years there, but the problem is that we don't really care that much at the four-year point. What we care is what happens after 10, 20, 30 years and when you make the plot -- when you show the time span we're talking about those are quite a bit different results.

And if you believe the three to four -- using the three to four-year data, we're basically a flat line, slight increase over time on the endothelial cell density. If you believe that the loss is going to continue linearly at 1.9 percent per year forever, then after about 20 years you're at the 1500 cells per millimeter squared and somewhere around 35 years you're down to 800. I don't have any -- I don't show any errors around these lines, in the error bars. If you know much -- if you know about errors for regression the errors go, they move outward the further away you get from the center of the data and if I put them on here, they would -- these estimates are pretty much meaningless I think after 15 to 20 years. You don't have very much confidence at all in them.

And that brings me to, of course, the caveats that the statisticians always give about extrapolation. It's always a questionable exercise to extrapolate beyond the range of the data we have and especially when we're talking about the range we have here. It's highly -- any extrapolation you would make would be highly dependent on the model we use and the assumptions we want to make and both those lines that you saw previously assume that whatever linear trend you saw between three and four years is going to continue forever beyond that.

And it's probably in this case a lot more important to think about if it's necessary to obtain good long-term data and if so, how to go about doing that. Okay, now, I'm going to switch gears a little bit and talk about individual patients because maybe more important than the average cell loss through time which is described by the linear fits are questions like what proportions of patients will show a cell loss greater than some critical amount. In other words, what proportion of patients will have cell densities less than 1500 or 800 cells per millimeter squared in 10, 20 or 30 years.

And from my point of view, the problem is you can't really answer this with much confidence using the data we have here. But let me just summarize what we do have here. If you'll recall one of the previous -- the fancy statistical model I used previously actually gives me an estimate for each eye of what the post-operative ECD change for that eye is and then after that, what's the annual change through time, and so you have a distribution of those estimates for each eye.

And using that, you can get -- you can create tables like this that tell you something like in this case four and a half -- excuse me, 10 percent of the patients will have an initial loss of four and a half percent or more and 10 percent of the patients will have an annual loss of 2.9 percent or more. Now, that's based on again, I'm making some assumption that whatever we've seen in the first three or four years is going to continue however far in the future you want to go. Okay, and finally, there were some co-variants that seemed to be significant predictors of endothelial cell loss, notably is the anterior chamber depth which was a statistically significant predictor of cell loss regardless of how you analyze it really. The sponsor presented analysis in the PMA that showed the used binned data, in other words, they broke the ACD into four different groups based on three, three and a half, four millimeter cuts and then presented the cell loss for each of those groups.

A bunch of other co-periods didn't appear to be significant predictors of cell loss. Just to help put the ACD effect into context, I created this graph here that takes -- for each eye, you take all the possible annual differences that you got for that eye and calculate from those the percentage loss for that eye and then average those for that one eye. So on the Y axis is for each eye now an annual percentage ECD change that we see in the four years -- after three months. I threw out the first few months because that seemed to be somewhat different. And then platted on the X axis is the ACD measurement for that eye.

And the point is that, remember the average ACD is around 3.5 and the average cell -- annual cell loss was right here, it's around two percent and down here it says estimated slope is 1.6, so you know that the difference between -- if this right here is about two percent loss, and someone that's a half a unit to the left is going to have a loss that's about 0.8 percent more, 2.8 percent, and someone who is a half a unit to the right is going to have about 0.8 percent, less cell loss. They're loss is going to be about 1.2 percent per year. This is just an attempt to kind of put the -- take the statistical significance of the ACD effects and try to put it in some terms that might be hopefully relevant.

So after all that, there's two main questions here for the panel. The first one is that the mean change between three and four years in that 57-patient cohort that had both of those was an actual gain of .1 percent in endothelial cell density, so is there sufficient data to support the conclusion that the losses in the first three years are reflective of surgical trauma with some prolonged remodeling period that culminates in a stabilization after three years and if not, what minimum eyes in follow up would you try to make a recommendation that we might need to make that assessment?

The second question relating to the anterior chamber depth eyes with the smaller anterior depth of 2.8 to 3 had a greater loss of endothelial cells than the eyes with a greater than 3 millimeter ACD. So the question is, do the outcomes of the ACD analysis provide some assurance of safety in this device for eyes in the lower end and then the upper end of the ACD spectrum? Thank you very much for your attention.

DR. WEISS: Thank you. We will now have questions for the FDA from the panel. I'm just going to start off, just to clarify for myself about the endothelial cell loss in terms of determination whether it levels off or increases between three to four years versus whether it continues dropping. From what I understood you to say, if you look at the cohort of 57 which is what the sponsor was looking at between three to four years, you could possibly say that it was going to level off, but if you look at the other cohorts, it does not show that. Am I misinterpreting it or is that basically --

DR. GRAY: That's correct. Your estimated amount of endothelial cell loss depends primarily on which cohort you use and the one cohort -- the cohort that has either the three and four-year measurement that has three and four-year measurements has a lower three-year count and therefore, you get basically a flat line after that.

DR. WEISS: So we have a choice of basically looking at the cohort of 306 and if we look at the cohort of 306, it does not support leveling off between three to four years. If we look at the cohort of 206, it does not support leveling off at three to four years. And if we look at the cohort of 37?

DR. GRAY: Well, when you say "support" it might mean a different thing to you than to me. When you get down to the 57 or 37 patients, there is more of a leveling off but on the other hand, there's more air because you have fewer patients. So I didn't actually do the test with the 37 patient cohort, but my guess is that you couldn't say statistically that there was a difference, but I didn't actually do that.

DR. WEISS: But certainly for the larger groups, which would have more statistical strength, it shows no leveling off.

DR. GRAY: That's correct. I personally used -- concentrated on the group that had a preoperative measurement and then two or more measurements after that because that was the one that I -- in order to do these tests you have to be able to fit a model of some sort.

DR. WEISS: So we're talking about if you look at the group of 206, which had the preoperative measurement and measurements at each of these time points, or at some of these time points, at least on two.

DR. GRAY: Two or more, yes.

DR. WEISS: At two or more of those time points. If you looked at that group, this did not support leveling off between three to four years.

DR. GRAY: From a statistical point of view doing the test for leveling, that's correct, it did not support it.

DR. WEISS: Okay, thank you. Dr. Grimmett?

DR. GRIMMETT: Michael Grimmett. Dr. Gray, I appreciate your comments. On the group of 37, you may not have run the analysis at the end but did you calculate the rates of or the confidence intervals for the endothelial cell loss, what it ranges between for the 37 eyes at year four? Did you show that? I mean, I know for the 57 it was a 90 percent confidence interval was 1.4 something. Did you do the same thing for the 37 eyes? It's probably wider, right?

DR. GRAY: No, I didn't do that. It would most likely be wider because of the sample size is three-quarters. So that would increase it by some amount, yes.

DR. GRIMMETT: Okay, thank you.

DR. WEISS: Dr. Bradley?

DR. BRADLEY: Dr. Gray, on one of your last slides there you showed us the relationship between anterior chamber depth and cell loss and you did a linear regression that 1.6 percent per millimeter.

DR. GRAY: Yes.

DR. BRADLEY: Did you do the analysis to find out how much of the variance was explained by the linear model? That becomes quite an important number for us.

DR. GRAY: Well, that was part of the analysis but I don't have that number here on me. The reason I -- I guess their point is that there is a statistically -- when you ask how much of the variation is explained, there is a statistically significant -- that slope is significantly different than zero, okay, so from a statistical point of view there is -- that's a significant slope. And what I was trying to get at was that what's the clinical relevance of that and that's where -- why I made the plot that calculated the 1.6 percent per year. But I don't have that number on me.

DR. BRADLEY: Yeah, but it's the clinical significance that's driving my question here in a sense that the linear regression might be highly significant but it may explain a very tiny amount of the variance and thus making policy based upon a parameter which explains only a tiny amount of the variance is really meaningless. So if we had that number or after the meeting somehow that number could be available, that might help policy.

DR. WEISS: Dr. Bandeen-Roche, Dr. McCulley and then Dr. Mathers.

DR. BANDEEN-ROCHE: Thank you for your presentation. I just have a brief clarification question which is that the numbers that you cited for the four-year mean cell counts differ from the calculations that I cited earlier. And so for instance the three, four-year mean that you cited three years and four years is 2355 and 2356 and reading from Volume 4 of 4, page MD19, those numbers are cited as 2455 and 2456. Now, this in a way sounds like a little point but it goes to the representativeness, the relative representativeness of the various cohorts. So I don't know whether it's clear which one of those is right.

DR. GRAY: I'm not, those all differ by exactly 100?

DR. BANDEEN-ROCHE: Yes, yes.

DR. GRAY: So my first guess is somebody has a typo because that's probably not just a coincidence that they're both exactly 100 off. These calculations that you see here, the mean cell, the sponsor sent me a data set at the end of July, July 25th, that has the endothelial cell counts that I later discovered they were rounded -- these are mean so they were rounded off to the nearest cell, the one I got. And that -- the numbers you see here are what I calculated using the data set that I was sent.

Now, if the three and four years -- if the two-year number is correct of 2428, then I would say 2455 and 2456 are probably not correct, because that would mean that there was an increase between two years and three years as well.

DR. BANDEEN-ROCHE: Okay, thank you.

DR. WEISS: Dr. McCulley?

DR. McCULLEY: Yeah, I've already expressed a little bit of skepticism about the emphasis being put on cell density but I know those are the numbers you had when you did your analysis, but from a clinical standpoint just over the years, I'm a little skeptical about putting too terribly much weight on something that can vary depending on where you take the count and the variability over time, the reproducibility, so I remain a little skeptical in that regard based on my clinical experience and what I've seen in reviewing papers and hearing presentations over many years.

So I guess then my question is, did you do any statistical analysis assessing the size and shape variation over time of the cells?

DR. GRAY: No, I did not do that. I used the results that we were submitted to us by the sponsor which seemed to indicate there was really not an issue. So I didn't --

DR. McCULLEY: Not, an issue, I'm sorry, meaning what, that there wasn't a change over time?

DR. GRAY: There did not seem to be a change through time for either the percent hexagonal or the CV and I didn't dig into that further. I used the same thing that you got in the submission, which is the analysis that the sponsor did.

DR. McCULLEY: Yeah, I mean, in the absence of data, I don't really know for sure what's right here and your extrapolation caveats, I think, are good and it would be nice to have the very long-term data, but at least from a cell density standpoint, my impression is that the critical cell density for corneal edema is 800 plus/minus 400 roughly tremendous range and tremendous variability. And that these other factors seem to play a very critical role and it would be more comforting for me to know that we had more data to support the size and shape didn't change over time. The numbers just aren't -- or the density isn't the only thing and there's tremendous variability in the measurement methodologies.

DR. WEISS: One thing, and I hope that we can pull this perhaps on the lunch break is one difficult item is for the August 2002 panel meeting when we had some of the people who were working with sponsor actually consult and guide the panel as far as what the requirements should be for such a study, I do not recall any such emphasis on hexagonality and coefficient of variation. The number -- the cell density is what was emphasized. Dr. Grimmett can comment in terms if your recollection is any different.

DR. GRIMMETT: Yeah, Mike Grimmett. I was the assigned primary reviewer for endothelial analysis at that meeting in August of `02 and in the presentation I made and included in the outline were the references that Dr. Edelhauser was citing regarding the sensitivity of pleomorphism and polymegathism so it was covered. I don't think the sponsor emphasized it or the presenters emphasized it but I did cover it in my presentation, making very similar comments to what Dr. Edelhauser said.

DR. WEISS: Dr. Mathers?

DR. MATHERS: Thank you for the clarity of your presentation. I thought it was very helpful. In the written work that we were given beforehand, you note that the -- by your model one you had an endothelial cell density loss in absolute numbers of about 49 cells per year and 20 percent of the population actually had a cell loss of 100 cells per year. That's what you're saying. Am I correct in assuming then that that 20 percent of the population in this population would then have an endothelial cell loss rate of about 3.8 percent per year by that calculation? If the 4.9 is average and the average is 1.9 by your model 1, it seems to me that would give a 20 percent of this group that were having a loss of 3.8 percent per year. I mean, that's the logical conclusion.

DR. GRAY: That is a conclusion that I didn't actually calculate. It's very difficult -- the problem is it's hard enough to estimate the mean function here and now we're trying to estimate the line below which only 10 percent of the people are going to be. And that actually is not -- is even harder statistically.

DR. MATHERS: Right, okay.

DR. GRAY: The best estimate I can do right now, based on the data we have are what I gave in the presentation, which is that 25 percent of the people will have 2.3 percent or more. Now, if I understand your confidence limits on that, it would be pretty wide.

DR. MATHERS: Right.

DR. GRAY: I'm not sure exactly what they are. I haven't -- I don't have them on me.

DR. WEISS: Seeing no other -- Dr. Macsai?

DR. MACSAI: I have three brief comments and I thought all your presentations were great, thank you. The first, they all revolve around endothelium but the first is to Donna. In all your presentations about ANSI and the guidance documents, nowhere did you mention a history of contact lens work and in light of all this discussion about endothelial cell remodeling, I would ask the agency to consider adding that so that that -- I think it's a critical piece of information to help us in the future on any intraocular device.

So I didn't see it. Maybe it's there.

MS. LOCHNER: It was discussed at some of the earlier panel meetings and the end result was that I think given considerations to the population you're treating and that there is going to be contact lens wear and what's the practical thing to impose on a clinical study, in the end the panel didn't give that emphasis, but I do hear what you're saying and I appreciate the comment.

DR. MACSAI: I'm simply asking for history so that you could segregate out --

MS. LOCHNER: Oh, yes, yes.

DR. MACSAI: -- who wore lenses and who didn't preoperatively. It helps analyze this endothelial cell data.

MS. LOCHNER: Yes, and I think many studies will be able to do that.

DR. MACSAI: Okay, the second two questions are for Dr. Gray. In this data set you received from the sponsor, do you know if patients who had exchanges at the time of implantation or subsequent to the time of implantation were excluded because that would skew this data, I think significantly?

DR. EYDELMAN: I think I actually touched on this in my review. I believe there were two different analysis. In the overall analysis by the sponsor, the data for the eyes that underwent secondary procedure were included, but they were excluded in the analysis where they were determining ACD significance.

DR. MACSAI: But were they excluded in measuring endothelial cell density long term?

DR. EYDELMAN: They weren't excluded from continuation of collection of data if that's what you're asking. We don't have the analysis for those eyes separated out.

DR. MACSAI: Well, do we have an analysis of the eyes that had the lens put in once and only once and never touched again and what happened to the endothelial cells?

DR. EYDELMAN: I believe that would be the analysis where the tables for the ACD depth significance were performed.

DR. MACSAI: And then I would ask Dr. Gray, looking at those tables, does your slope still hold to the green versus the black slide number 15 or whatever it was, 13, sorry?

DR. GRAY: I guess I'm -- first of all, I'm not entirely sure because I don't recall the exact -- I didn't actually do that analysis both ways to compare but the key difference between the estimates that we saw was the fact that the 37 or the 57 patients had a lower, a much lower count at the three-year time point than the other group and that's what is driving most of the difference. All the other methods of analysis and different groups of patients that you include, if you get beyond just the three and four-year data, you have a switch and so all of a sudden, it's about two percent, 1.8, 1.9, 2 and so it really comes down to a question of what time point you think the remodeling is over or whatever happens during the surgery is done with and beyond that, we can consider steady state. And then you get into the whole issue of what does that even mean and how can we extrapolate 20 years down the road which is sort of unanswerable, I think, with the data we have.

DR. MACSAI: Maybe I'm not getting something here.

DR. EYDELMAN: Let me just try to add, we don't have exactly what you're asking for, Dr. Macsai. We don't have the analysis of just the eyes that had secondary intervention, the endothelial cell separated out. What I do want to point out were that there were few eyes to start out with and chances are some of them did not have the analysis all together. As far as I'm aware, PMA did not contain breakdown for the -- on this issue. Certainly your recommendation can look upon it after the panel.

DR. WEISS: Since we're running 40 minutes behind and we haven't gotten into a discussion, I'm going to have one brief comment by Dr. McCulley, and then we're going to go to five minutes of questions for the sponsor and then break for a 45-minute lunch.

DR. McCULLEY: Okay, a critical question seems to be in humans, how long does it take for the endothelium to remodel after an injury and is it degree of injury dependent, is it age dependent? I don't know the answers to those questions but that seems to be absolutely -- the answer to that seems to be absolutely critical in knowing how to interpret the cell density and the cell shape and size change. Do we know that? Do we know how long it takes to -- and maybe when the sponsor comes back, Hank will know. But that's a key question to all of this.

DR. WEISS: I want to thank FDA for an excellent analysis and presentation. Sponsor, would you be able to answer or address some of these issues?

So you have five minutes to answer all our questions. While the sponsor is setting up, when we break for lunch, I'll just point out, this will be abbreviated. It will be 45 minutes, not an hour as listed in deference to the fact that we are running over significantly at this early point in time.

MS. THORNTON: Are you ready, Dr. Vukich?

DR. VUKICH: Pardon me?

MS. THORNTON: Are you ready?

DR. VUKICH: I believe so. For some reason, I believe the projector was changed out from underneath us. Okay. We would like to just take a moment to respond to a couple of the questions that were requested of the sponsor. For the number of sites that were contributing to the four-year analysis, this data was collected at eight of the nine sites that were collecting specular micrographs. We were able to calculate the confidence interval for the 37-eye consistent cohort of eyes at all of the intervals and that will be the graph that follows.

There was clarification that we will need from Dr. Bandeen-Roche on her request for information on an overlie of one of our cohorts, but it may take a little more time than we have available and a little more clarification on exactly what we would like to provide. This is the 90-percent confidence interval of the mean for the 37-eye cohort and at four years, which I think is the point of interest. It was 2244 to 2509. I see we're taking notes here. Okay, good. This is the entire cohort then for the endothelial cell density. For a point of clarification, this cohort did include all patients and these were also -- who were examined and did include patients who had secondary procedures so in some essence it does look at a worse case scenario.

A separate analysis of the data, subtracting those patients out has been done. We can tell you that it shows no difference in our estimation. We were hoping it would, but it didn't.

There was one final question that we'd like to address and that was from Dr. Bradley. There was a question concerning pupil size and quality of vision. We wanted to point out that our contrast sensitivities were all done under mesopic illumination at 3 candelas per meter squared. Although we did not have pupil size to correlate with that, there would be some assumption that the pupils would be at least smaller than under photopic conditions and that with and without glare there was no demonstrable difference at post-operative contrast sensitivity and in fact, at four of the five measured intervals, there was actually an improvement in contrast sensitivity so we hope that speaks to the quality of vision at least under mesopic conditions.

Finally, we'd like to thank the members of the FDA panel for their thoughtful and thorough review of all of this information. Thank you.

DR. WEISS: Thank you for making it brief.

DR. McCULLEY: Does Hank have an answer to my question?

DR. WEISS: We'll find out. Can you make it -- can you give a brief answer and if the answer is, we don't have the information, then that is the answer.

DR. EDELHAUSER: I think that is the answer. We don't really have the information. The only really data that we can rely on is probably the keratoplasty data from Bill Bourne which showed a market drop-off, you know, and that's not really the data we're after. So we don't have the data.

DR. WEISS: No data. Forty-five minutes for lunch and then we'll be starting promptly.

(Whereupon, the proceedings in the above‑entitled matter went off the record at 12:22 p.m. and went back on the record at 1:14 p.m.)

DR. WEISS: Can everyone from the panel take their seat, please. We're going to continue the Committee deliberations on this PMA with presentations from Primary Panel Reviewers, beginning with Dr. Marian Macsai‑Kaplan. I will remind Panel Members and Sponsor, and FDA, et cetera, that we are now about an hour behind, so I would suggest or request that all comments be short, to the point, and have the purpose of moving this PMA ahead.

DR. MACSAI: I'm done.

DR. WEISS: With that non‑intimidating introduction, I have Dr. Macsai.

DR. MACSAI: I would like to first acknowledge a few things. One is, that the Sponsors did an amazing job on a really fast track PMA, and that the FDA did an outstanding job in getting us this information as fast as it could be gotten. And I want to really thank Sally for being in such close communication. This was a difficult PMA to review I think for all of the reviewers.

The Sponsor has gone through a lot, and so has the FDA, so I'm going to try and limit my comments, but I have a few things I just feel obliged to say.

First of all, you saw in the distribution of the patients enrolled in the study, that the vast majority were Caucasian. And from previous devices we looked at, we realized that we do need to look at the affects in non‑Caucasian patients. The Sponsor did supply data from the Dominican Republic data set, and I think it would be important for that to be included in anything made available to the public, segregated by refractive error, to help the non‑Caucasian population with their expectations.

Second of all, exclusion criteria were included, and 65 eyes with pre‑existing conditions were included in the study. The results of what happened to those 65 eyes should also be made available by the Sponsor to the Agency, because from those 65 eyes, we may glean information that would help patients who might be treated in an off‑label manner.

In addition, in the exclusion criteria, limbal pathology was not included, and must be included if a white‑to‑white measurement is required to size this IOL.

Another additional criteria that must be included for exclusion is what the lower limit of endothelial cell counts are per age group. And I would reference Dr. Grimmett's excellent review for that.

I'm going to now address efficacy, and then the questions put forward by the Agency. Efficacy of this device is really good, very good. And I'm going to just limit by comments by saying that I was happy to see the efficacy of this data in the 3 to 7, 7 to 10, and 10 to 15 diopter groups, and leave the over 15 diopter group for later in my discussion.

I would have some questions why a refractive surgeon might use this in a minus 3 diopter group, and until I personally see data that this is superior to refractive surgery already out there, I would personally wonder about that issue.

Regarding the specular microscopy data, which was my question 1 in the original questions provided by Dr. Eydelman to us, I feel uncomfortable, plain and simple. I feel uncomfortable because we haven't set a limit of what is the minimal number of endothelial cells that a patient needs to have. We're talking about implanting a device in a 22 year old patient, taking worse case scenario, as the Sponsor said earlier.

We've segregated out the patients that had complications, replacements, removal, and if you take a 22 year old and assume that they don't become in need of a cataract until they're 62, assuming they're myopic, they have a higher prevalence of nuclear sclerotic cataracts, you're talking about the device remaining in place for 40 years. And at 40 years, according to Dr. Gray's chart, they're going to drop to a dangerous limit. And so my discomfort comes from the fact that the surgeons who participated in this trial are the best of the best. They have the best hands, they have the best experience. I've had the privilege of being taught by some, and observing them, and they are really the best there is, so we're taking a device and releasing it to Joe Q. Average surgeon, and this device will be seen as sort of a drive‑ through procedure, I'm afraid, where you drive in, you get your IOL, you drive out, you move to Outer Mongolia, and we don't know what happens to you. And we don't know what's going to happen in 10, 20, 30, 40 years to the endothelium. So I, of course, having experienced the closed‑loop AC IOL induced pseudo phakic bullous keratopathy, am concerned about this device and its effect on the endothelium. And that, to me, is the biggest issue with this PMA. Everything else is really pretty small in comparison to that.

Along those lines, we were asked to look at the anterior chamber depths. And I think the Sponsor has shown, Dr. Gray has shown, everyone has shown that in the hands of the best, with an anterior chamber depth less than 3, this device induces a 50 percent higher endothelial cell loss. So at this time, my recommendation would be that this device not be labeled to be used in an eye under 3 millimeters anterior chamber depth. And that if the Sponsor has further data, that can, of course, be looked at in the future.

Question 2 is the nuclear opacities. Nuclear opacities in this population developed at two time courses, early‑on, probably surgically‑related. Later on, probably nuclear sclerosis developing in these high myopes.

I didn't have a big problem with this, but it brings very much to the surface the training of surgeons who are going to use this device. If you look at the Canadian data in those three inexperienced surgeons, there was a 22.5 incidence of anterior subcapsular opacities, while the surgeons that were proctored in the Dominican Republic only had a 4.8 percent incidence of anterior subcapsular opacity development. So clearly, that technique used in the Dominican Republic has some effect, so the Sponsors are now left with a huge challenge; how do you take Joe Q. Average surgeon and make him good enough to use this device?

And some of my suggestions would be that this device, this Collamer ICL is very similar to the Collamer posterior chamber intraocular lens and Toric intraocular lens that is currently available, and has been for years, for cataract surgery. And that any surgeon who wants to implant this device must first become proficient using that intraocular lens and loading it in the shooter, which is the exact same, and implanting it in the eye. And only after they're proficient with that device, should they then be able to use this device. And they should be proctored one‑ on‑one in the use of this device.

But it brings to mind another concern, which is, if you look at the analysis of the investigational sites, one surgeon at one site had a significantly higher number of complications, and a significantly higher number of IOL removals and exchanges. And remember, we're dealing with the best of the best, so I raise this question to the Sponsor, pending release to the general public, how is the Sponsor going to monitor this? If the Sponsor has to supply these IOLs to someone who's exchanging them too often, or repositioning them too often, the Sponsor seemingly should have some kind of tracking method for this, and further training required prior to the release of this device. And it's a big, onerous task, but we're talking about putting this in young people with clear lenses, so I think that there's a degree of responsibility the Sponsor will have on this post‑approval.

Regarding the Agency's question about removal, and if there's areas of touch, and if the uncorrected vision is worse than 20/50, I thought these were fine caveats, but I would also raise the question to both the Agency and the Sponsor, if there is an anterior subcapsular cataract in the visual pathway, should that also be added as a reason for removal?

Question 3 regarded the use of the horizontal white‑to‑white in the anterior chamber depth measurements to determine the sizing of the ICL. I too, like Dr. Grimmett, went back to my operating room and looked at what I had available to measure white‑to‑white, and it's just a little, I think, Castroviejo caliper, and mine goes by 1 millimeter increments.

I, like Dr. Vukich and Dr. Slade, was trained in a time that we did extra caps, we measured white‑to‑white. I think my residents have done five extra caps in their entire training. I don't think they know how to measure white‑to‑white. I think the Sponsor is going to either have a huge task of teaching them how to do it, or find a better technique. And for that, I would recommend consideration of the Orbscan, which we now know has been shown in the Wang article from the Development of Ophthalmology Journal to be reproducible. It also supplies your anterior chamber depth.

I'm not endorsing that product. I hold no interest in that product, but it's out there, and it would give a reliable reproducible measurement for the beginning surgeon. Regardless though, if the patient has limbal pathology, you cannot ascertain a white‑to‑white measurement; therefore, that is an exclusion criteria in my mind for this device.

Question 4. There are currently no devices approved in the U.S. for correction of myopia greater than 15 diopters. True. So I feel once again very uncomfortable here.

First of all, clearly this device in that population does not correct myopia, it only reduces it. So in light of Dr. Eydelman's question, we have to change "correction of" to "reduction of". But I worry that we, as a panel, are going to arbitrarily set a standard by approving this in this age range.

I look to the Agency, and ANSI in their wisdom for guidance, and my feeling is once a guidance document is developed in this population, minus 15 to minus 20, and the Sponsor has this engineering thing worked out, that at that time, once the guidance document is set, if the device meets the guidance document criteria, approval is a no‑brainer. But at this time, we have no guidance, and I'm uncomfortable with arbitrary approval, which would set a standard, because I am certain there will be more phakic IOLs to come in the future.

Question 5, does safety and effectiveness data support approval of the STAAR ICL for the eyes with the following pre‑operative MRSE, minus 3 to minus 7, minus 7 to minus 10, minus 10 to minus 15. And in general, my response to this question is yes. However, there remains this outstanding issue regarding endothelial cell loss, sizing of the IOL, cataract information. I'm not uncomfortable with the cataract formation, sizing of the IOL is fixable. And I guess I feel if Dr. Edelhauser doesn't have the answer for endothelial cell loss, I don't know who will. And so, we're functioning in a big old gray zone. And maybe a warning that might be appropriate is that endothelial cell count must be done on these patients pre‑operatively, and should be done on these patients post‑operatively for a very long time. And if there is a decrease long‑term in endothelial cell count, not from an otherwise obvious condition, such as a high fema, trauma, iritis, that perhaps this device should be explanted to protect these patients from pseudo phakic bullous ‑‑ from bullous keratopathy at some time in the future.

The Sponsor Question 6, management of acute intraocular pressure rises in post‑operative period. Well, I'm disappointed that gonioscopy was not performed post‑operatively in these patients, and I think that Dr. Lochner's presentation has addressed this issue. A mistake was made in the development of this PMA protocol, and it will have to be rectified in the future. But perhaps if the PIs were made farther in advance ‑ I don't know, one week seems awfully early to me ‑ the PI would have healed, and not of them might have been included. And there wouldn't be a need for reopening in the future.

In addition, I think the Sponsor must mandate that the surgeon check the pressure within 4 to 6 hours after placement of the device, and again in 24 hours, so that if it's the viscoelastic, this can be addressed.

Question 7, Sponsors have reported that a number of patients noted glare and/or halos post‑operatively. Again, I'm disappointed because though Dr. Schallhorn might feel pupil does not make a difference, and I know this lens is much farther inside the eye, I think we could have learned a great deal from that information. And I would ask the Agency to mandate pupil measurements in the future, so that our patients can have a better idea of what to expect from a device. Without it, we can't answer the question, so we're kind of left ‑‑ we need to include the data about glare and halos, what patients experienced. We need to include the data about the quality of vision pre‑operatively. It was poor at 11.6 percent of patients pre‑operatively, but at 36 months, it was still poor in 5.8 percent of patients. And that's a little disconcerting, because if you read the recently published paper where they compared an eye with an ICL and an eye with LASIK, those patients were doing great. And I have no doubt that the refractive quality with this device for patients will be better than a minus 10 LASIK. And that the higher order aberrations will be less with this device than a minus 10 LASIK. But I'm still wondering why 5.8 percent of the patients rated their vision poor. Who were they, and why was it poor? So that concludes my presentation. Thank you.

DR. WEISS: Thank you very much, Dr. Macsai. We're going to have Dr. Joel Sugar, who's the second primary reviewer.

DR. SUGAR: Thank you. I'm going to just skip through various parts of my review. Of course, I want to thank and compliment the Sponsor and the FDA reviewers for the excellent job they did in both putting the data together, and then analyzing the information.

The accountability was good in the study. The efficacy was good up to the minus 15 diopter range, and beyond that range, certainly reduction of myopia should be the indication, or the labeling should be for reduction of myopia, not for correction of myopia. The stability was good.

In terms of safety, the loss of lines of best corrected visual acuity, I thought was very acceptable. I think that you can play games about the fact that the minification has changed and, therefore, you should lose less lines, but what matters to the patient is how well they see. And if they don't lose lines of vision, even though they should have theoretically gained a line of vision, I think they're still benefitted.

I was concerned about the patients who required enlargement of their laser iridotomies post‑operatively because of elevated interocular pressures. In my review, I had the wrong time periods because I measured from the baseline examination, not from the day of treatment. I'm concerned about the Sponsor developing a better means of assessing the iridotomies, both their spacing and their size, so that these patients won't have the pressure elevations as high as 65, as were noted in the presentations.

The retinal detachments, I think were acceptable given the population that was being assessed. The cataracts, I think, were acceptable given the population that was being assessed. Although I have concern about the recommendation for removing the lens when anterior subcapsular cataract is seen at an acuity of 20/50 or greater, I would be more concerned about removing it when there's progression of cataract. If, however, I had the data that I don't have, which is, is going in and taking the IOL out, putting a new one cause more progression of the ASC or not, and I don't think we've been presented with any information to tell us whether that does or does not happen.

I'm also concerned in terms of the issue of cataracts, since these are patients who will develop cataracts in the long run, like all of us. Is axial length measurable through the IOL easily or not? Does a new algorithm have to be developed for ultrasonic, or whatever technique is used for measuring axial length?

People who have their axial length measured, their anterior chamber depth measured ultrasonically could presumably have that data, their axial length captured concurrently and presented to the patient. And it would, I think, make sense, since this is an implant, that the patients be given a card with the data on the lens implanted. But also, if there's data on their axial length, that that be captured, unless it's easy to measure their axial length with the IOL in place, and it would be nice to know that from the Sponsor. It would also be nice to know whether exchanging the lens in and of itself induces another order of complications.

Endothelial cell loss has, I think, been very well discussed, and I guess I do feel that, contrary to what I wrote in my review, that anterior chamber depth less than 3 should probably be contraindicated for this lens.

There are a few other minor issues. There's some in the labeling that I mentioned in my review. For example, in the brochure it says that surgeons should never touch the center of the optic with instruments when it's in the eye. I don't know if that's because of concern about leaving imprints on the lens, or it's because pushing the lens, pushing the IOL into the crystalline lens could induce cataract. It would be worth having a statement in the brochure saying why that's an issue.

The statements made, again, in the labeling, that this device has "been shown to improve the overall quality of vision", I think that's too broad a brush to paint this with. I think you need specific data saying that some patients have overall vision improvement, some don't, and give data.

The brochure should also, I think, have a picture of the device, and a picture of the positioning so that even if someone's taking a course, they will have some hard copy information, should a question arise about lens positioning; although it seems pretty obvious.

In terms of the specific questions, is there sufficient data to suggest that there's remodeling? I think that there is. I'm concerned that we capture more data in four years, and definitely capture data at five years on endothelial cell loss. I don't think that we should wait for that information to approve the product.

I already talked about the anterior chamber depth. Do I believe surgeon experience is an issue? Absolutely, and that's been addressed by the Sponsor, saying that there will be mandated training. I also talked about the anterior subcapsular cataract, that we need more information on what secondary interventions do.

Do I believe the method for determination of overall diameter is appropriate? I think that it is. I think that white‑to‑white is not as difficult to measure as has been implied. While Orbscan gives it a .1 millimeter on a standard printout, it gives you the white‑to‑white up to .1 millimeter, I don't think that that ‑‑ and that's been shown to be reproducible, I don't know that it's been shown to be any better than manual white‑to‑white measurements. And certainly, hasn't been shown with this device to provide any advantages. And it's a substantial expense for the average practitioner, who may not have the Orbscan.

We talked about the greater than 15. I think that the device should be approved for correction of myopia up to minus 15 diopters, and for reduction of myopia beyond that level. And I think that ends my review. Thank you.

DR. WEISS: Thank you very much, Dr. Sugar. The last reviewer, Dr. Grimmett.

DR. GRIMMETT: I'm pleased to have the privilege to make a few comments about the application. I apologize for any redundancy. I didn't have any of the talks before during the preparation of my talk. Additionally, part of my purpose and mission is to get the information in the public record, so that interested patients in the future can search relevant issues regarding this device.

You've obviously all read my review, the cure for insomnia, and I will try to highlight just a few of those issues, but will not go over the data in excruciating detail. You can be happy about that.

Before I dig into the PMA, I'd like to go over a few background issues regarding the application to help us in our overall analysis. First, I want to review a few issues related to phakic IOL lens vault. Proper lens vaulting is clearly critical to the success of this phakic IOL. Excessive vault over the crystalline lens will push the iris forward, and has the following potential complications; angle closure, angle synechiae, iris chafing with potential complications of pigment dispersion and pigmentary glaucoma, iris sphincter erosion, iris translumination defects, and alteration of the normal aqueous dynamics that is pupilary block.

On the flip side, a poor vault in the ‑‑ over the crystalline lens has the potential to induce cataracts due to IOL crystalline lens contact. Moreover, if the IOL is too short, it's theoretically possible for it to be mobile, with possible rotation or anteroposterior movement. Clearly, the vault has to be just right to minimize complications, and the tolerances are expected to be low.

With an older version of the ICL, Version 3, the Sponsor believes that poor lens vault led to a higher right of anterior subcapsular opacities, quoting results from Sanders, in the Journal of Refractive Surgery in 2002. The current application states that Version 4 has an additional .13 to .21 millimeters of anterior vault, as compared to Version 3. And while I didn't find data in the PMA to substantiate that, the Sponsor clarified today that's a design issue.

In the literature, Gonvers & Associates examined central vaulting with digitized slit lamp photographs in 75 eyes. They had 24 V3s and 51 V4s. At three months, the central vaulting of the 24 V3s was slightly less than the central vault of the 51 V4s, but the difference in their study was not statistically different. And they concluded, "The change in design between models V3 and V4 did not achieve its goal, which was an increase in vaulting." I just bring that up because I didn't see any data in this application to substantiate the assertion in vivo. Certainly, it's important to keep in mind that increased vaulting may reduce cataractogenesis at the expense of iris and angle complications.

In the application, when looking at vaulting, one gets the impression that the phakic IOL vault is a static situation, but I don't ‑‑ this couldn't be further from the truth. Stable phakic IOL vaulting on a day‑to‑day basis is probably not achievable for numerous reasons. Number one, accommodation has been shown to decrease anterior chamber depth by about a quarter of a millimeter, increase the lens thickness by .28 millimeters, and it decreases the radius or curvature of the anterior surface of the crystalline lens.

Number two, lens vaulting may differ, depending on whether the patient is supine or prone; that is, gravitational effects. And number three, the light reflex has been shown to cause a reduction in the phakic IOL anterior capsular distance. Therefore, on a day‑ to‑day basis, the actual lens vault is probably a dynamic variable.

Here's an ultrasonic image from Kim and colleagues in AJO in 1998. The third image on the top shows accommodation on a 30 centimeter target, and displays a decreased distance between the IOL and the crystalline lens right there, due to changes in lens thickness and radius of curvature.

The fourth image shows a relationship of the phakic IOL to the crystalline lens in total darkness right here. And then the relevant change when shining a penlight on this eye. In this particular case, there's IOL lens contact with simply a light reflex. Based upon these data, perfectly static phakic IOL crystalline lens relationships on a day‑to‑day basis are improbable.

Moreover, stable IOL vault over the lifetime of the eye is probably not achievable either for numerous reasons. One, the soft IOL material may flatten with time. Dr. Vukich, I believe, mentioned European or outside the United States data over 10 years, that it may not. There is an article in the literature that indicates that it may. I believe it's from Arne.

Number two, aging has been shown to increase the lens thickness by 1.24 millimeters from age 40 to age 65. Number three, plate phakic IOLs may rotate or have mobility. And number four, the ciliary sulcus diameter has been shown to decrease by approximately 1 millimeter in diameter from age 40 to 80.

All of these day‑to‑day and lifetime issues may lead to intermittent or permanent IOL crystalline lens contact, and may lead to cataractogenesis, pigment dispersion, subclinical inflammation, and/or disruption of the normal aqueous humer dynamics. Given these factors, I can't imagine that ICL positioning will be stable and problem‑free for the lifetime of a given patient, especially since this device is intended for young recipients.

Let's talk about issues related to the sizing of these IOLs. The sizing of the ICL myopic lenses was determined by the horizontal white‑to‑white and the anterior chamber depth measurements in the following fashion. For anterior chamber depths 2.8 to 3‑1/2, they added half millimeter to the white‑to‑white, and for anterior chambers greater than 3‑1/2, they added 1 millimeter to the white‑ to‑white. For in‑between sizes, there was a rounding down and rounding up protocol.

Hence, STAAR's sizing methodology is based upon white‑to‑white measurements. However, valid scientific evidence exists saying that white‑to‑white measurement do not correlate to the sulcus dimension. So white‑to‑white measurement does not ‑‑ is not a good surrogate marker of the variable of interest, the sulcus diameter.

Here is just one piece of information from Reinstein's study, in which he examined white‑to‑white values with calibrated photographs and sulcus‑to‑sulcus dimensions with high frequency ultrasound. All this information is in the public domain. It's right off the Internet.

The top value shows that of myopic eyes, plotting white‑to‑white on the X axis, and sulcus‑to‑sulcus on the Y axis, that there's no correlation for myopic eyes. The same was true for hyperopic eyes.

These data imply that a one‑size fits all phakic IOL would seemingly have just a good chance of success or failure as basing the ICL upon the horizontal corneal diameter.

Let's go ahead and look at a few examples of basing the ICL on white‑to‑white measurements to display this fact. Here's a case where white‑to‑white is 11‑1/2 OU. Put the ICL based on that, bravo, it looks pretty good ‑ adequate lens vault in both eyes, left and right, so we're pleased with ourselves on this case.

The next one we have an asymmetric white‑to‑white, 11‑1/2 on the right, and 12 on the left. However, despite differing white‑to‑ white measurements, the lenses were over‑sized in both by about the same amount, rather than an asymmetric amount, and the vault is excessive, causing angle closure, as you can probably see.

Here's a case where the same white‑to‑white existed on both sides, but the vault was excessive on the right, and non‑existent on the left, with lens IOL touch. I simply would say that because there's valid scientific evidence indicating there's a lack of correlation between white‑to‑white and sulcus‑to‑sulcus, that physician labeling should include relevant material facts indicating the lack of the correlation. In fact, in knowing this data now, it's amazing to me that the vault data within the application is as good as it looks.

We'll review a few issues related to glaucoma. And please pardon me, Dr. Coleman. I will defer to your judgment on these issues. I'm just the cornea guy. Projected glaucoma risks for this device include pigment dispersion syndrome, angle narrowing, and angle closure.

Regarding pigment dispersion syndrome, it's important to realize that STAAR's study cohort fit squarely within the known risk factors for pigment dispersion syndrome; that is, myopia young age in Caucasian race. Pigment dispersion syndrome is at least as common in women as in men.

One study quoted about a 2‑1/2 percent incidence of pigment dispersion in Caucasians. Simply using this figure based on the number of Caucasians in the STAAR PMA, we'd expect six in this study to have pigment dispersion. The Sponsor reports zero, both before and after ICL implantation. Let's look to the literature.

A published study found pigment dispersion in the angle in 9 of 58 eyes, or 15‑1/2 percent at 18 months. The authors postulated that the STAAR ICL pushes the iris anteriorly, and optic iris chafing leads to pigment dispersion syndrome in a subset of patients.

A 1998 study, using ultrasound after ICL implantation, found angle narrowing in all eyes, and peripheral anterior synechiae in 2 out of 9 eyes, or 22 percent. The ICL was in wide contact with the iris in all eyes.

For this study, I reviewed the submitted PMA materials, and reviewed both the pre‑op and post‑op clinical study report forms. I didn't find any gonioscopy data, which I was shocked to see that. I also didn't find any ultrasound data presented to determine angle anatomy alterations following the ICL. It's my opinion that the lack of these data is a disservice to present and future patients with the STAAR ICL, and represents a major study design error.

Gonioscopy can assess angle pigment deposition, a sensitive and common finding in pigment dispersion syndrome. Perhaps no patient was diagnosed with pigment dispersion syndrome because no one looked at the angle post‑op.

Moreover, gonioscopy can determine angle narrowing and synechiae. Further, if no gonioscopy examinations were performed, other relevant features could be missed, vascularization and other preoperative abnormalities.

The theoretical risk to the angle can be easily surmised given the design and intended use of this phakic IOL, and it's my belief that the initial study design should have included gonioscopy, whether or not it was mandated by the FDA.

Let's review issues related to pupil diameter and the lens optic diameter. It's well known that dim illumination mydriasis can be robust in the young. Dr. Vukich indicated that when this study was designed, that those parameters were not well known. Being an old guy, I beg to differ. Back in about 1993‑94, I reviewed issues related to pupil diameter with small optical zone radial keratotomy. My literature review at that time revealed that the mydriasis being robust in the young was documented, well know, and in the literature at that time. I believe that predates the design of this particular study.

STAAR's study cohort ranged from 22 to 45 years of age, and we've heard that the lens optic diameter is 465 to 55. Given the young age of the cohort, as Dr. Bradley already noted, it's reasonable to expect that some patients will have dim illumination pupil diameters that exceed the lens optic diameter. We, therefore, have an expectation that some patients may experience halos and dim illumination, or have nighttime visual aberrations.

Looking to the literature, Arne found a higher frequency of halos with small optic diameter ICLs. The rate of halos correlated to the difference between the scotopic pupil diameter and the optical zone size. Due to these halos, these authors recommended intentional under‑correction for high myopia; that is, using a larger optic diameter lens followed by LASIK.

Hence, another study design error in this PMA is the absence of pupil size measurements. Relevant analysis should have included the rate of visual aberrations with increasing optic pupil mismatch. Regrettably, this was not performed for our review.

In the absence of this pupil size information, the best we can do is stratify the patient's symptoms by the lens optic diameter. I couldn't find this information in the materials given to me, but it should be required for later FDA review. Also, each symptom category should be reported separately; that is, separately none and mild, rather than lumping the categories in the current tables. Of course, this information presumes that the small lens optic patients do not have skewed pupil sizes one way or the other. We'll simply never know.

Let's go on to endothelial cell loss. The threshold analyses that I presented in Appendix 1 of my written review show maximum rates of annual cell loss to reach various target levels at the time of death. Clearly, there's many assumptions that are made, including an annual instantaneous cell loss, and that it's linear, and it doesn't include information regarding stem cell repopulation. However, using these figures, if we desire a 1500 cell for millimeter square density at death, a .9 percent annual loss rate is the maximum, inclusive of all age ranges; that is, the 20 to 30 year old range. And if we desire an 800 cell per millimeter square density at death, a 1.9 percent annual loss is the maximum.

It's important to remind ourselves that 50 percent of patients will have endothelial cell densities that fall below the normal mean cut‑off values; and, therefore, younger patients, that 20 to 30 age group, have a significantly higher risk of running out of endothelial cells during their lifetime if these rates are continuous. And now to the PMA itself.

Regarding the study population, the total eyes show with the blue bars indicate very good follow‑up. I certainly recognize the difficulty of carrying out such a large study for an extended period of time, and commend the Sponsor for their efforts. The purple bars show endothelial data on approximately 200 eyes, with a large drop‑off at the 48 month interval shown out here as 67 eyes.

I find it ironic, some studies we reveal at panel only have 6 and 12 month data, and we're always wrestling with not enough data. And here a Sponsor has run a 3 and 4 year study, and we're still wrestling with not enough data. I just found that amazing.

Unfortunately, the endothelial data in the written PMA have varying ends, and there's no consistent cohort of eyes followed through each and every examination interval. The data we've seen today with that 37 eye consistent cohort was not provided to me in the materials that I reviewed. That made the evaluation difficult.

Just one housekeeping item, and I believe Malvina already alluded to this. The inclusion criteria had a stable refraction within a half diopter over the prior year. The indications for use statement had a 1 diopter over the prior year, obviously, needs to be matched or reconciled.

Regarding the exclusion criteria, we know that phakic IOLs can alter the corneal endothelial. Dr. Macsai alluded to this. The corneal endothelial status was omitted from the exclusion criteria, and given the young age of these patients, I believe it would be a relevant material fact to be considered prior to implantation of this device.

Certainly, if a young patient had an abnormal endothelial layer, I would not recommend this device as a clinician. There is no question that I wanted pre‑op specular endothelial analysis for this cosmetic elective procedure, where the alternative is glasses or contact lenses.

On to some safety issues. Let's discuss the learning curve associated with phakic IOL implantation. I believe we're all in agreement that the labeling should include relevant learning curve issues. Of the 13 upside down lens insertions, 11 occurred within the investigator's first 22 procedures, 6 out of 13 developed in AST in the early post‑op period. Of the 14 eyes that developed anterior subcapsular cataracts, most occurred within each investigator's first 8 surgical cases. One investigator accounted for a disproportionate share of the ASCs, a 9.4 percent rate, and that same investigator accounted for both cataract extractions in the study. To lessen the impact of learning curve issues for the patient, I'd favor specialized course training or case supervision by an experienced surgeon for early cases.

On to change in best spectacle corrected visual acuity. As compared to the lower dioptic groups, there are larger post‑op gains of best corrected visual acuity, 20/20 or better, in the high myopia group. For the less than 7 diopter group shown in the orange, there is an 8.3 percent gain pre‑op among 36. For the 7 to 10 diopter group shown in the maroon, there's a 15.6 gain pre‑op among 36, and for the greater than 10 diopter group shown down here in the blue, there's about a 20.4 percent gain pre‑op among 36. These are findings strongly argued for an induced magnification effect as a result of the surgery.

In looking at greater than or equal to one line of best corrected visual acuity loss, high myopes have an increased rate of vision loss with time as compared to lower myopes. And we've already heard that for this particular group, a one line loss is the equivalent of a two line loss due to induced magnification as a result of the surgery.

The rate of greater than one line loss goes up to about 16 percent. I'm not sure why that would exactly be. I don't know if that has to do with lens optic pupil mismatch or other issues, but I'm not sure it's well delineated. It's certainly not clear in my mind as the ultimate etiology of that.