Usefulness of longitudinal evaluation of Cyfra 21-1 variations in advanced lung cancer monitoring

To investigate the usefulness of Cyfra 21-1 as an indicator of therapy effectiveness and prognosis in advanced primary lung cancer, sixty-three patients were selected on the basis of a high Cyfra 21-1 serum level (> 3.3 ng/ml) at the time of diagnosis. Serial assays of Cyfra 21-1 were performed during the first three courses of chemotherapy among 63 patients. The serial values were analysed according to response to treatment and overall survival. After three courses of chemotherapy, a 70% reduction under the initial marker's value or a return to normal was observed for 36 patients. Twenty-two (61%) of these patients presented an objective response to therapy, making Cyfra 21-1 a moderate indicator in terms of positive predictive value (PPV). However, a significant decrease of Cyfra 21-1 was observed in 88% (sensitivity) of the 25 objective responders. Cyfra 21-1 changes after one course of chemotherapy (61 patients) were not sufficient to predict the future response after three courses (sensitivity 52%, specificity 56%, PPV 45%). Among 30 clinical or radiological relapses, a 10% increase or a return upper reference limit in Cyfra 21-1 level was observed in 18 cases (sensitivity 60%, specificity 100%, PPV 100%). Survival data were available for 61 patients. No significant statistical difference (P > 0.05) was found between survival curves depending on a significant decrease of Cyfra 21-1 after the first course of chemotherapy. We can conclude that the only interest of serial Cyfra 21-1 assays may be the detection of relapse, where one observes a significant decrease of the marker correlated with an objective response to first treatment. (C) 1997 Elsevier Science Ireland Ltd.