Date: Thu, 30 Mar 2000 12:03:15 -0700

To: Recipient List Suppressed:;

Beginning on page 280

CHAIRMAN WHALEN: Next, Dr. Wendie Berg will discuss considerations of imaging patients with breast implants.

Dr. Berg.

DR. BERG: Thank you, Mr. Chairman, members of the panel.

If I can have the lights down a little bit, as a radiologist.(Laughter.)

DR. BERG: Can I have the next slide, please?

I'm going to be presenting imaging considerations largely focusing on breast cancer diagnosis in women with breast implants. Rupture, particularly with saline implants, is really a clinical diagnosis. Periprosthetic fluid is a common finding on imaging, but we dismissed this. It's not thought to represent leakage on the whole, and again, I'm going to focus my comments on detection of breast cancer.

Could I have the next slide, please?

We can argue about the specific number of women who have undergone breast implantation, but approximately two million women in the United States are affected by it, and again, most of my comments are going to be directed to women with augmentation rather than reconstruction since we do not generally image the breast after mastectomy.

If one considers the rate of breast cancer to be approximately one in nine over a course of a lifetime, we can estimate that roughly 200,000 women with breast implants will develop breast cancer.

Next slide, please.

Mammography remains the standard for early detection of breast cancer.

The goal, of course, is to detect breast cancer before it becomes palpable at an earlier, more curable stage. We know from the literature that 90 to 95 percent cure rates are achievable when breast cancer is detected at Stage 0 or Stage 1, and this is nonpalpable disease, largely found by mammographic screening.

Survival rates and disease free survival, in particular, drops to 60 to 70 percent when lymph nodes are involved by the tumor. That number further drops to approximately 40 percent when the lymph nodes are involved and the primary tumor is palpable at presentation.

Next slide, please.

There have been several studies looking at the risk of breast cancer in women with implants, and they have rather conclusively demonstrated to date that there is no increased risk of breast cancer as a result of the presence of the implant, and in fact, in several smaller studies there has been actually a slightly decreased rate of breast cancer compared to that expected.

May I have the next slide, please?

Some general considerations first, and then I'll get into specific data that is available.

The American College of Radiology Standards require the performance of routine views, as well as implant displaced views in order to adequately evaluate the breast tissue in patients with implants. As a result, we can expect at least double the radiation dose to the breast tissue per mammogram obtained in such patients.

Further, the presence of implants in and of themselves is an indication for diagnostic mammography, which would allow the mammograms to be reviewed by the radiologist when the patient is there in the suite. The reason for this is that many times the technologist is unable to obtain an optimal mammogram at the first pass, and additional views would be needed to adequately compress or evaluate the breast tissue.

As a result, we again anticipate at least more than double the cost of annual surveillance mammography.

Next slide, please.

These are rather difficult to project, but just to illustrate, this is a mammography with routine views first in a patient with silicone implants, and the next slide, please. The corresponding images are obtained when the implant is pushed back out of the field of view, allowing better compression of the implant -- of the parenchyma itself.

Next slide, please.

Even with such techniques, there is a reduction in the visualized breast tissue in patients who have breast implants. It's difficult to answer absolutely how much that reduction would be in any given patient. There is some data from a series where patients' mammograms were measured both prior to and after implantation, and overall it was found that 30 percent reduction in the visualized breast parenchyma in the absence of any contracture.

If contracture is present, it's more difficult to compress the breast. As a result, greater reduction, on the order of 50 percent, was observed in the amount of visualized parenchyma. Even with implant displacement techniques, the amount of breast tissue that we see is still decreased compared to a patient without implants, and in fact, on average that was 25 percent still obscured with implant displacement; greater, on the order of 35 percent, if the implants are subglandular compare to subpectoral locations.

Next slide, please.

To illustrate, this is a woman who had silicone implant placed behind the muscle and there's very little breast tissue visible on the routine views.

Next slide please.

And this is difficult to project, but on the implant displaced views, a subtle cluster of calcifications was noted, and it's actually right here in the middle of a spot magnification view. This patient had a very small focus of ductile carcinoma in situ that was detected despite the presence of the implants.

Next slide, please.

However, it's not always so easy to displace the implant. This woman has a saline implant, and you can see that it's still quite dense, although you can see a little bit of the internal structure and the folds of the edge of the implant.

And despite every attempt at implant displacement, this is the best mammogram that could be obtained. She had very little breast tissue.

Next slide, please.

She underwent an ultrasound. I don't know if we can have the lights down any further -- underwent an ultrasound that showed the implant itself, and there was a very subtle mass anterior to the implant that was an early infiltrating ductile carcinoma, completely invisible on mammography as a result of the implant.

Next slide, please.

In general, as I've mentioned, the implant can hide the breast tissue directly and, as a result, can hide lesions as well in the breast tissue. Adequate compression is sometimes difficult to achieve due to contracture, pain, and the mass effect of the implant itself. It can displace the tissue and cause overlap in the normal parenchyma.

It can be difficult to visualize lesions in both projections. You might see that lesion inferiorally in the breast, and yet it's hidden by the implant in the craniocaudal projection, despite implant displacement techniques, and this can confound interpretation as well as limit the biopsy options and make it more difficult to biopsy any lesions that are seen.

And finally, in the woman who had undergone removal of an implant, there can be extensive scarring, not always, but there can be. That can confound interpretation. There can be residual calcifications, particularly if the capsule is left behind after removal of the implant, and both of these can mimic cancer.

Next slide, please.

To illustrate, again, this is a patient who has a silicone implant behind the breast tissue, and this was the best mammogram that could be obtained. Very poor compression was achieved in the tissue itself, and you can see there's a rather large density. This is approximately four centimeter invasive ductile carcinoma was visible, but if there were any other lesions in this breast, it would be very difficult to assess that.

Next slide, please.

And, again, this doesn't project well in this lighting, but this was a patient who was found to have a subtle cluster of calcifications in the inferior breast.

Next slide, please.

But she had ruptured saline implants bilaterally, and we were unable to localize the calcifications in the other plane because they really proved to be on the inferior breast directly underneath the implant. We were able to biopsy these with stereotactic technique and found fibrocystic change in this case.

Next slide, please.

After implant removal we can see a variety of changes that are very suspicious. This particular patient had explantation of an intact saline implant, but remained with a spiculated density at the chest wall which, if you didn't know the history, would be considered highly suspicious.

She then underwent ultrasound -- next slide, please -- and was found to have a seroma.

Next slide, please.

Another patient who had undergone explantation, again, had a spiculated density of the chest wall, and there were actually calcifications evident within this, a lot of deformity in the breast tissue, difficult to get an adequate mammogram, especially in the inferior breast.

This proved to be an infective collection. We're having fund. I think you need a few more microphones.

Next slide, please.

I mentioned that the capsule itself can cause some problems with interpretation, and the main reason for that is the presence of calcifications in the capsule itself, a rather common finding. Judy Distouet and colleagues found about a quarter of the patients have some degree of calcification in the capsule. Usually it's relatively easy to identify because it's relatively coarse and typically benign, but when it's first starting it can, again, mimic early cancer.

Next slide, please.

Just an illustration of these calcified capsules. You can see it really can get quite extensive. This patient had severe contracture, as well.

Next slide, please.

And the calcification in that capsule can be visible, easily distinguished from ligament calcification or not. If it's left behind, this capsule itself can form the pocket for collecting fluid, and as I mentioned that one case, infection.

Next slide, please.

I think the overwhelming question which I was asked to address is really will the diagnosis of breast cancer be delayed in women with implants as a result of suboptimal mammography. Unfortunately I'm not sure I can answer this question. There are only several small, retrospective studies that have been performed which are really inadequate to answer this question at this time.

Next slide, please.

I'm going to present a literature review, but there is, again, minimal data and keep in mind most serious report results from silicone implants, not saline.

Next slide.

I think there is some evidence to suggest, however, that the results may be generalizable between silicone and saline implants. A study again from Washington University in 1989, using the American College of Radiology and Mammography Phantom, which includes a variety of artifacts, including dense specks, which mimic calcifications, and densities which mimic early cancerous manifests as masses, was used with a variety of types of implants positioned on top of the Phantom and a normal mammography exposure performed.

In their study, they found the shell alone minimally altered the ability to detect the various artifacts, but the shell filled with either silicone or saline completely obscured all artifacts.

Next slide, please.

What kind of performance are we expecting from mammography? Well, this is a good question. I'm not sure we have the absolute answer, but in the American Health Care Policy Research Manual from 1994, we do have benchmarks that were established by a variety of experts in the field suggesting that with routine screening, we should be able to achieve detecting of the majority of cancers at Stage 0 or Stage 1, over 50 percent, and that node positivity should be under 25 percent of the patients diagnosed, and overall sensitivity of mammography on the order of 85 percent should be achievable.

I think that last number may be a little optimistic. There have been multiple studies showing performance. In practice it's closer to 78 to 80 percent detection of breast cancer, allowing for a variety of factors, including errors in interpretation.

Next slide, please.

These are the references on which I have drawn, the literature that does exist on implants and breast cancer detection.

Next slide, please.

There's a lot of information here, but just to summarize, you can see across these studies very small numbers of patients, and I think these are patients who had augmented breasts with usually silicone implants and were not undergoing annual mammographic screening. So this is simply at the time of detection looking at results.

They had ten patients, six patients, seven who had implant displacement views, as well as routine views, a total of 41 patients in the study of Silverstein, but all small numbers of patients in these studies.

Overall, the degree to which cancers were visible mammographically ranged from 55 percent up to a high of 86 percent. Overall palpability of the lesions detected was quite high across all these series. The lowest was the first study here with Leibman and Kruse, where six out of ten cancers were palpable at presentation, but the vast majority of the cancers in these studies were palpable, and again, this may reflect the lack of routine screening in these patients.

Nodal positivity was also higher than that benchmark of 25 percent across most of these series. One study in particular I want to call your attention to was that of Laurie Fajardo and colleagues done at Arizona. At the time 18 patients all had implant displaced mammography views, as well as routine views, and in that series the sensitivity was only 67 percent, and in fact, 39 percent had positive lymph nodes at presentation.

Next slide, please.

So to summarize, the majority of patients in these studies that have been done to date had only routine mammographic views without implant displacement, and we've already shown, I think, that it's mandatory that the implant displacement views be obtained in order to adequately evaluate the parenchyma.

More cancers were palpable at diagnosis than in general. We expect that number to be about 40 percent palpable at presentation. In these series it was from 80 to 90 percent in the majority of the studies.

The stage distribution of cancers, however, in the papers that had control groups was not found to be significantly different in women with implants, nor was the survival found to be different.

Okay. Next slide, please.

Overall, where it could be assessed, 66 percent of cancers were visible on mammography with implants, and when implant displacement views were included, again, very small numbers, but 72 percent of those cancers were then visible on mammography.

That's lower than the expected sensitivity, as I mentioned.

Next.

I think of greatest concern though, it's difficult to make recommendations because the performance of mammographic screening in women with saline implants or any other implants, for that matter, has really not been adequately evaluated, particularly with high quality mammography and implant displacement views.

We can suspect from the data that does exist that there's at least a ten to 20 percent decrease in mammographic sensitivity, and that alone, even that relatively conservative number, has the potential for delayed diagnosis of cancer in 20 to 40,000 women.

Are there other alternatives to mammography? Very briefly, yes, there are, but they all have their limitations as well. Ultrasound is being used more and more widely. It has a clear role in evaluating palpable abnormalities in all patients, including those with implants. We also use it when there's a mammographic density that we're concerned about.

It's easy to guide biopsy lesions as we see under ultrasound, but the problem is screening ultrasound is less sensitive to the very early carcinomas, particularly ductile carcinoma in situ, than is mammography.

Further, it's technically extremely demanding. It requires a lot of expertise on the part of the person doing the ultrasound, usually requires a physician to perform the task, and at least in the United States the costs of screening ultrasound are on the order of $300 per patient compared to approximately $75 for mammography.

Further, lesions behind the implant will not be well seen even on

ultrasound.

Next.

Just another slide that illustrates a cancer adjacent to the implant on ultrasound.

Next slide, please.

It has been suggested that MRI may be appropriate in these patients. In fact, one recent reference suggested it's the modality of choice for detection of primary breast cancer in the augmented breast.

Well, it's clearly a very sensitive test. The implant does not obscure detection of lesions. However, it does require injection of intravenous contrast. It's extremely expensive. A billed cost is about $1,000 for a contrast enhanced MRI.

It's difficult to guide biopsy lesions seen only on MRI. It's technically very demanding and not widely available.

Next slide, please.

Just to illustrate though, it is very nice to demonstrate cancers on MRI. We have here an implant at the lower right-hand corner of the slide, and you can see the area of enhancement just above it is a spiculated mass with associated rim enhancing lesion, and these were two adjacent cancers that were nonpalpable in a woman with implants.

Next slide, please.

MRI done improperly, however, still doesn't help obviously and it can be very demanding. This is a woman that we saw in our practice with a saline

implant, some periprosthetic fluid inferiorally; several cysts in the breast, but no contrast had been administered and, therefore, no lesions were detected of significance, and she had a breast cancer that went undetected for another year.

Next slide, please.

One other potential method for screening would be nuclear medicine techniques, such as Sestamibi or Miraluma, as it's more commonly known. However, again, the sensitivity is not very good. It's an expensive test, again, and in particular, I call your attention to the statistic that nonpalpable lesions under a centimeter, only 48 percent of these were detected.

And so I don't think there are very many good alternatives to mammography, in summary, but I think we also have at least reason to consider that there may be some limitations of mammography in patients with implants.

Thank you for your attention.

CHAIRMAN WHALEN: Do any of the panel have questions for Dr. Berg?

DR. ROBINSON: I have a question.

CHAIRMAN WHALEN: Yes.

DR. ROBINSON: Just a question. How do you think MRI will evolve as an imaging technique in the evaluation of women with implants or one where you really cannot get good imaging by your other --

DR. BERG: I think more and more we're finding from data from international studies and grant sponsored research trials now that it is an extremely effective method at finding early cancer. I think the problem is going to be who's going to foot the bill.

It's an extremely expensive test. It's very demanding, and if insurance companies will foot that bill, great, we can do the test. But I think that as a society, we really can't afford to screen women with breast MRI at this point. So we've got that double edged sword.

DR. ROBINSON: What numbers would be involved if you weren't screening them per se, but just doing women where you could not get good imaging by another technique?

DR. BERG: Well, again, I think you're looking at least probably 30 percent of women with implants where you've got significant limitations. You've got women with dense breasts, women who are at high risk. We're probably looking overall at the population of maybe, again, probably 30 percent of the overall population who has mammography routinely where MRI would stand to benefit them.

It is routinely done in women who are at high risk at some centers

already, and it's being more and more widely used.

DR. ROBINSON: Yeah, that's what my impression was.

The last question. I'm sorry. For lesions behind implant, is spherical CT or helical, anything in that area going to have any implication?

DR. BERG: I thought about including CT. The reason I did not is that it's got very high radiation dose to the patient, on the order of two to three rads as opposed to mammography is on the order of .2. You don't want to irradiate the breast with tenfold as much radiation. You're going to be causing a significant number of cancers.

So it also is not -- it's clearly not as sensitive a test for certainly not in situ disease and may pick up invasive cancers with the injection of contrast, but, again, I don't think anybody wants to advocate CT for that purpose.

Any other questions?

DR. BURKHARDT: I have a question. Most of the studies that you quoted here are of necessity a few years old.

DR. BERG: Right.

DR. BURKHARDT: In the last five years or so, there's been a tremendous shift in the placement of these implants in the plastic surgery community. They're almost all put behind the muscle now --

DR. BERG: Right.

DR. BURKHARDT: -- in nonreconstructive cases.

DR. BERG: That's right.

DR. BURKHARDT: Do the ACR standards still require double the radiation dose?

DR. BERG: Yes, they do. I think it's very explicit. I looked in the most recent ACR standards that I have, which is 1998, and it does require the implant displacement views be obtained as part of routine practice, and I think you'd be very hard-pressed to defend if you missed a cancer as a result.

It's very difficult. Even with subpectoral implants it's very difficult to adequately compress the entire tissue, depending how much tissue the patient has.

DR. BURKHARDT: Do you have any sense of what percentage of women in the eligible and in the recommended cancer screening group actually have mammograms according to the ACR standards?

DR. BERG: Good question. I was thinking about discussing that. I don't really have good data on that, but I can tell you that many women with implants hesitate to have mammography even once they become of that age because it's a painful exam, and it's more involved.

DR. BURKHARDT: How about women without implants?

DR. BERG: Well, without implants, we know it's about 60 to 70 percent who do under -- have had a mammogram within the last two to three years.

DR. BURKHARDT: Thank you.

DR. BERG: Un-huh.

CHAIRMAN WHALEN: Ms. Brinkman.

MS. BRINKMAN: For routine screening mammography, does insurance pay for the extra views then for the displacement of the implant?

DR. BERG: As a rule, insurance does pay the additional cost, although oftentimes a woman will still have a deductible and still bear a greater cost as a result of having to have a diagnostic mammogram on a yearly basis for what amounts to screening.

CHAIRMAN WHALEN: Thank you, Dr. Berg.

We are now going to proceed to the review of the first PMA, and that is going to be the one of Mentor Corporation. So I would ask those who are going to be making that presentation to come forward.

I would like to remind all of the public observers at this meeting that while this portion of the meeting is open to your public observation, you as public attendees may not participate unless there were to be a specific request of the panel.

We now turn it over to Mentor Corporation who, if necessary, can take upwards of a full hour for their presentation.

I'm being outvoted by the mutiny here. I was going to hold off on the break until afterwards, but it seems that all of the panel has bladders the size of walnuts -- (Laughter.)

CHAIRMAN WHALEN: -- we will take about a seven minute break and then resume.

(Whereupon, the foregoing matter went off the record at 2:43 p.m. and went back on the record at 3:04 p.m.)