In recent years, immunohistochemical analysis for KIT (CD117) has made the once troublesome diagnosis of gastrointestinal stromal tumor (GIST) almost routine in most cases. This is particularly important now that an effective therapy—treatment with the kinase inhibitor imatinib mesylate (Gleevec)—is available for patients with unresectable or widely metastatic disease.1,2 Oncologists still sometimes fret that a diagnosis of GIST has been overlooked, but the recent attention given to KIT staining probably has resulted more in overdiagnosis than in missed opportunities to treat.3 Although there are GISTs with unusual morphologic (epithelioid, pleomorphic) and immunophenotypic (KIT weak or negative) characteristics, these represent only a minority of cases.4 The greater problem with GISTs today is in determining their prognosis rather than making their diagnosis. Great debates of past decades about the distinction between “benign” and “malignant” GISTs largely have been resolved in favor of the view that all GISTs have malignant potential; it is just a question of degree. Thus, the phrase “stromal tumor of uncertain malignant potential (STUMP)” still applies to the typical primary GIST, although it must be emphasized that it is the pathologist who is stumped, not the tumor. To stress the malignant potential of GISTs, many experts in the field routinely sign them out as gastrointestinal stromal sarcoma—a prudent diagnosis that nevertheless offers relatively little guidance to clinicians and their patients. How best, then, to assess the prognosis of a patient with a newly diagnosed primary GIST? There are a number of parameters that have been defined reproducibly as prognostically important in GISTs, including size, mitotic index, mucosal ulceration, necrosis, site of origin, and aneuploidy. Among these, size and mitotic index are the features emphasized in most published schemes, including the consensus guidelines published after a National Institutes of Health–National Cancer Institute–sponsored workshop held in April 2001.5 These guidelines recently were used in a large (600 cases), retrospective epidemiologic study by Kindblom and colleagues,6 in which it was found that the long-term survival of patients with low-risk and very-low-risk tumors was almost the same as that of the general population. Not surprisingly, high-risk and overtly malignant tumors, which constituted 20% to 30% of the total, had a much poorer prognosis. Yet even among these groups, there were long-term survivors (>15 years). For patients with intermediate-risk lesions, the survival was closer to that of the low-risk group than the highrisk group. This important study, the first in the post-KIT era, confirms 2 general observations about GISTs: (1) The great majority of very-low-risk, low-risk, and intermediate-risk GISTs behave in a benign manner. (2) Among these tumors, there is an unpredictable subset that behaves aggressively. Similar studies are ongoing and perhaps will result in fine-tuning of the consensus guidelines. It is doubtful, however, that any scheme based solely on tumor size and mitotic index will permit the clinician, when meeting a patient with a typical primary GIST, to interpret the pathology report with anything more than vaguely legalistic phrases (“Mr Smith, it is more likely than not that....”). Such uncertainty was tolerable when there were no effective therapies for recurrent or metastatic GISTs (chemotherapy and radiation treatment do not work in this disease). With the introduction of imatinib, however, risk assessment of GISTs has taken on new importance. The prognosis of patients with advanced GISTs has changed dramatically with imatinib therapy.1,2 Based on an
Read full abstract