ABSTRACT
New lung cancer cases for 1997 are estimated at 178,100 by the American Cancer Society with 160,400 deaths from this disease (1 ). Most of these (75-80%) are caused by non-small cell lung cancer (NSCLC). The majority of the patients who die of this disease are at some point candidates for palliative chemotherapy, but this chemotherapy is toxic, expensive, and of limited benefit. A recent review including results from 4 meta-analyses suggests a median survival improvement of only 1-3 months for treated compared with untreated control patients (2). Most tumors respond only partially to chemotherapy, and the chances of such responses range from less than 30% in most large multi-institutional trials to over 60% in some large single-institution trials, where the treated population is probably made up of patients with better prognosis. An ideal in vitro test would allow us to select for treatment patients with a 100% response rate. In reality, however, any approach that would allow the reliable prediction of a significant number of the nonresponders-sparing the expense of their treatment and at the same time increasing the probability of response for the treated population-would seem worthy of consideration in an era in which cost control has become a major focus of attention. We will present evidence that in vitro chemosensitivity testing might play such a role.
