Thursday, April 5, 2007

Two cases where increased imaging of the breast does not improve outcomes


There's another paper today challenging conventional wisdom on how effective screening studies are in the prevention of breast disease. In this instance it's computer-aided detection (CAD) mammograms, which were billed as a way to increase the effectiveness of interpreting mammograms. It uses advanced image recognition software to screen for abnormalities. Despite little data, this technolopgy was quickly adopted. Many mammography centers adopted this technology and were incentivized by medicare with additional payments ( ~ $20 per study) to use CAD. Note: When you do thousands of mamograms a year, this $20 per study can be serious money in someone's budget

CAD is apparently more sensitive then humans to abnormalities, but less able to distinguish malignant from benign. It also tends to identify a number of ductal carcinmona in situ (DCIS) lesions, which if left undiscovered may never turn into invasive breast cancer (aka. "real" breast cancer). Much like small prostate cancers in males, where many would die with it rather then from it, we as yet have no predictable way of tell patients that nothing will come of it and are forced to offer more aggressive surgical and medical treatments and their associated morbidity.




This report is the mirror image of the other large counter-intuitive finding we saw recently with screening CT scans for lung cancer. (ie. we find more disease but don't improve outcomes about death from cancer, while actually increasing morbidity and anxiety for patients.

While I'm off on this I'd like to dovetail on imaging studies for breast implants. With the 2006 FDA approval for silicone gel breast implants came a recommendation for routine screening by MRI of all implant patients starting 3 years post-op then every 2 years. This illogical and unenforceable (and unfunded) suggestion is surely going to widely ignored by patients and I'm not sure I'd blame them. For the period which would encompass the first 2 MRI's (five years out) the rupture rate of an implant in augmentation patients of "regular" silicone implants is likely under 1-2% while after 4 MRI's (10 years out)it's likely only 6 or 8%.

What drove this recommendation by the FDA? Most point to the highly charged political environment that still exists in America over this. One month prior to the FDA approval, Canada released the last of it's token restrictions on the devices and commented that while MRI screening was discussed, routine use was not evidence-based medicine. What do other countries do? They image implants selectively and usually start with ultrasound and reserve MRI for equivocal findings. As consequences of rupture tend to be confined to the breast and we know many woman may go years or decades with asymptomatic rupture, this is the position that makes the most sense to me. If you were to start MRI screen in asymptomatic women it would make sense to do this at 10-12 years when you sit and crunch the numbers.


The form-stable cohesive gel implants from Inamed
(picture below), Mentor, & other companies will make this topic be revisited when they are likely (you never know with the politics of the FDA)approved for general use later this year. When you have a device like Inamed's 410 implant with rupture rates so low as there are no data points to even do projections on rupture rates will the unsound MRI recommendation be attached to it as well? Keep in mind there has also been an issue of a number of false positive MRI's read out with the Inamed 410 and some of the dual-lumen devices (part saline-part silicone)in some of their clinical trials as they look somewhat different on MRI

Stay tuned!

Rob Oliver Jr. MD
www.oliverplasticsurgery.com

No comments:

Post a Comment