When the FDA re-approved the general use of silicone breast implants in 2006, there were some “strings” attached. One of these was the recommendation for follow-up testing of the breast, using an MRI scan, at years 3, 5, 7 (and so on) to look for clinically “silent” implant leakage.
This was a new recommendation – and a change from the earlier FDA policy of 1992, which had not recommended any screening tests at all for asymptomatic women with silicone implants.
So, does this new FDA recommendation make sense?
Well, let’s look at the science, shall we? It really breaks down into two different questions:
1) Is an MRI scan a good way to look at the breast – or would a breast ultrasound be a reasonable, inexpensive first-line alternative?
2) What are the odds of having a leaking silicone implant at 3 years, or 5 years?
This is going to get a little technical, but please, bear with me.
Question one: MRI or not.
Prior to the 2006 ruling, MRI was not in routine use as a “screening” exam for women with silicone breast implants – it was used if there was a specific reason, such as a mammogram or ultrasound that was indeterminate, or a clinical situation that required investigation.
An excellent paper published in 1998 by Chung and associates, from the University of Michigan, actually developed an algorithm for decision-making for women with suspected silicone breast implant rupture, based on a statistical method called Bayes’ theorem.
They found that in asymptomatic women, if they had a “normal” result on a screening ultrasound test (i.e. no rupture reported), the probability of an actual rupture was low – 2.2%. If the ultrasound reported a rupture, and this was confirmed by MRI, the likelihood of a true rupture rose to 86% – a reasonable threshold to operate.
So, Chung et al. recommended breast ultrasound as their first-line test, since ultrasound is widely available, relatively cheap, and doesn’t involve radiation. They only used the expensive MRI if the ultrasound was abnormal. Makes sense to me – I like this idea because the patient saves money, and is more likely to actually get the test done, because of the lower financial barrier.
In a separate paper published in 2001, Cher and associates did a meta-analysis of MRI results in 1039 women. They found that when the MRI was used for women with specific complaints like breast hardness, shape change, etc., that it was reasonably accurate in detecting rupture – >80%. In women without these issues, the positive predictive value of an MRI was “insufficient to warrant use as a screening tool”. They also recommended using the MRI to confirm the results of a screening ultrasound.
The Royal College of Radiologists (UK) echoed these views in their recently published guidelines on breast imaging. A normal ultrasound examination was highly predictive (91%) for an intact implant. Ultrasound was recommended as an initial investigation. According to the RCR, patients with an abnormal ultrasound examination should then proceed to an MRI examination, provided the more powerful 1.5 Tesla MRI machines are used with the specialized “breast coil” imaging device. Lower power MRI machines or those that don’t have the breast coil are not nearly as reliable.
Question two: how often does leakage really happen?
The best way to determine whether there is implant leakage is to perform surgery, and look at the implant directly. But, as you might imagine, it’s pretty darn difficult to convince a women who is feeling perfectly fine, without any symptoms at all, to undergo surgery – just so we can open her up to look and see whether her implants are OK!!
So, this data is hard to come by. Instead, we use MRI scans – to estimate the deflation rate. In Mentor’s Core study, at 3 years out, just 0.5% of the 420 patients had a rupture in the primary augmentation group.
In other words, at the three year point, 199 women will have to pay to have an MRI that is completely normal, for one woman to find a rupture. That’s probably not the best utilization of health-care dollars.
In the study by Sharpe and Collis (UK), no ruptures were noted until 7 years postoperatively.
Data like this really make me wonder whether it makes sense to insist on a scan at years 3 and 5, when not much is happening.
In summary: the screening recommendations of the FDA are certainly controversial. I feel that they are scientifically questionable, financially imposing, and a bit of a “CYA” move by the FDA, particularly at years 3 and 5. However, they are the “official party line”, so I tell all my patients about them. The FDA rules were probably written more with political expediency in mind. I hope that, in time, they will be modified to better reflect the actual science!