SERUM TUMOR-MARKER DECLINE IS AN EARLY PREDICTOR OF TREATMENT OUTCOME IN GERM-CELL TUMOR PATIENTS TREATED WITH CISPLATIN AND IFOSFAMIDE SALVAGE CHEMOTHERAPY

Background. Serum tumor marker regression (alpha-fetoprotein [AFP] and human chorionic gonadotrophin [hCG]) was studied in patients treated with ifosfamide-based chemotherapy for cisplatin-resistant germ cell tumors (GCT) to investigate the role of marker regression as a predictor of treatment outcome. Methods. Fifty-four patients treated with cisplatin and ifosfamide-containing therapy were the subject of this retrospective analysis. The serum tumor marker half-life (T1/2) for the first two cycles of therapy was calculated for each patient using all marker values Day 7 through the end of the second treatment cycle. A calculated T1/2 for hCG of less than or equal to 3 days or a calculated T1/2 for AFP of less than or equal to 7 days was defined as appropriate marker regression; any T1/2 greater than these values was considered prolonged. A variable designated ''marker decline'' was defined to indicate whether the serum tumor marker half-life of AFP and/or hCG was satisfactory or unsatisfactory for each individual patient. Both univariate and multivariate analyses were conducted to investigate ''marker decline'' as a predictor for response, event-free survival (time to death or relapse), and overall survival. Results. Satisfactory marker decline predicted an improved event-free survival and overall survival. The median event-free survival for patients with an unsatisfactory marker decline was 5.8 months versus 20.7 months for patients with a satisfactory marker decline. Survival for patients with an unsatisfactory marker decline was 6.3 months versus 20.7 months for those patients with a satisfactory marker decline. Further evaluation demonstrated that hCG decline was a stronger predictor for improved survival than AFP decline. A multivariate analysis performed on selected clinical variables using a Cox regression model demonstrated that marker decline, pretreatment hCG, and primary site were independent predictors for event-free and overall survival. Conclusions. The rate of serum AFP and/or hCG decline during the first two cycles of therapy was predictive for event-free and overall survival in GCT patients treated with ifosfamide-based salvage therapy. Those patients with an appropriate serum tumor marker decline had a longer event-free and overall survival. When evaluated separately, the rate of hCG decline was more predictive of treatment outcome than decline of AFP. The rate of serum tumor marker regression during the first two cycles of therapy is a clinically useful tool in assessing treatment outcome at an early point in therapy and may thereby identify patients who could benefit from a change to more intensive therapy.