With some trepidation, I'd like to discuss a topic that I think has significant implications for our profession, the over use of and over billing for immunohistochemistry stains.
I do not think this is a widespread abuse, but it definitely does exist in isolated practice foci. There's no doubt whatsoever that immunohistochemical stains, properly applied, are one of the most valuable, if not THE most valuable diagnostic tools in our surgical pathology armamentarium. However, it is also clear when viewing cases from other institutions and speaking to colleagues across the country, that a few surgical pathologists are overly and inappropriately "generous" in their application. In some cases this may well be the result of nothing more than a haphazard and not well thought out approach to trying to diagnose a difficult a case. In other instances, however, I suspect a more clear-cut and less noble motive. Why, for example does one need to order a panel of nearly a dozen cytokeratin subtypes on a pleomorphic malignancy before initially determining that the tumor is entirely cytokeratin negative with an appropriate cocktail? Why order a large panel of hematopoietic markers as an initial diagnostic foray before demonstrating positivity or negativity with a CD45 study? Why order basal cell markers on every prostate biopsy? I could go on at length, but I think you get the picture.
The reality is that these abuses constitute a form of inappropriate self-referral. They're coming under increasing scrutiny, and the result may well be some Draconian ruling outside of our control that, for example, may state that no payer will reimburse for more than three immunostains on any given case. Witness the recent radical cut in reimbursement for the technical component of 88305 (and other codes) by the CMS, based on perceived abuses. Even worse, imagine if no immunostains could be ordered without clinician approval.
In order to avoid the above we really need to police ourselves, and soon! This will require developing and using "best practice" algorithms for working up cases needing immunohistochemical studies. My friend and collague Dr. Mark Wick has spent a signficant portion of his career developing such algorithmic approaches to light microscopically undifferentiated tumors. I'm not remotely suggesting that immunostains be ordered one by one, day after day, until the answer is realized, an approach that would add tremendous delay to the diagnosis. Nor am I suggesting that we can or should eliminate cases where, once the diagnosis is known, a shorter immunohistochemical path to the diagnosis becomes obvious. Hindsight is always 20:20! Rather, I'm advocating that staining panels proceed in a logical, one, two or even three-step process from the broader to the more specific markers, instead of "bathing" the specimen up front with huge numbers of markers. And, needless to say, I'm suggesting that immunostains should not be used unless they are truly needed.