Prognostic factors in patients progressing after cisplatin-based chemotherapy for malignant non-seminomatous germ cell tumours.

The aim of this study was to define prognostic parameters for survival in patients with malignant germ cell tumours progressing after platinum-based induction chemotherapy with or without surgery. A total of 164 progressing patients (testicular: 83%, extragonadal: 17%) were identified out of 795 patients treated with platinum-based induction chemotherapy for metastatic germ cell malignancy with or without surgery. 'Progressive disease' included patients who had progressed after a previous partial or complete remission as well as patients who failed primary therapy. Salvage chemotherapy consisted of 'conventional' platinum-based chemotherapy. Prognostic factors for survival were assessed by uni- and multivariate analyses. The resulting prognostic model was validated in an independent data set of 66 similar patients. For all 164 patients the median time from start of induction chemotherapy to progression was 10 months (range: 0-99). Thirty-eight (23%) patients relapsed after 2 years. The 5-year survival rate for all progressing patients was 30% (95% confidence interval 23-38%). In the univariate analysis the following factors most importantly predicted a poor prognosis: progression-free interval < 2 years: initial poor prognosis category (MRC criteria), < CR to induction chemotherapy, initial treatment early in the 1980s and treatment given at a 'small' centre. Three prognostic factors remained in the multivariate analysis: progression-free interval, response to induction treatment and the level of serum human chronic gonadotrophin (hCG) and alpha fetoprotein (AFP) at relapse. One hundred and twenty-four patients could be classified on the basis of these characteristics, Those patients with progression-free interval < 2 years, < CR to induction chemotherapy and high markers at relapse (AFP >100 kU l(-1) or hCG >100 IU l(-1)) formed a poor prognosis group of 30 patients, none of whom survived after 3 years. Patients with at most two of these three risk factors formed a good prognosis group of 94 patients (76%) with a 47% (37-56%) 5-year survival. Thirty-eight patients from the good prognosis group with a progression-free interval of >2 years had a 2-year survival of 74% (60-88%) and 5-year survival of 61%. These prognostic groups were validated in the independent data set, in which 5-year survival rates in the good and poor risk groups were 51% and 0% respectively. One-third of patients progressing during or after platinum-based induction chemotherapy for metastatic germ cell malignancy may be cured by repeated 'conventional' platinum-based chemotherapy. Good prognosis parameters are: progression-free interval of > 2 years, CR to induction treatment and normal or low serum markers at relapse (hCG < 100 IU l(-1) and AFP < 100 kU l(-1)). The results of high-dose salvage chemotherapy should be interpreted on the background of these prognostic factors.