Modified cisplatin, etoposide, and ifosfamide (PEI) salvage therapy for male germ-cell tumors (GCT): Long-term efficacy and safety outcomes.
328 Background: Second-line chemotherapy (CT) of advanced GCT is vividly debated and no standard of care has been established. Since 1985 we introduced the modified PEI combination with the aim to reduce toxicity while maintaining efficacy over the original regimen. Methods: We retrieved data of pts who received ifosfamide at 2.5 gr/m2 (with mesna) on days 1-2, etoposide and cisplatin at 100 mg/m2 and 33 mg/m2, respectively, on days 3-5 every 21 days for a maximum of 4 cycles, followed by surgery of residuals. Multivariable analysis (MVA) was undertaken for pre-specified variables and IGCCCG-2 score was analysed as a continuous covariate. ITT analysis was applied. Results: From 02/85 to 01/12, 189 pts failing cisplatin, bleomycin, and vinblastine (PVB, N=25) or etoposide (PEB, N=164) were treated. 87% had a gonadal primary, 50.3% were MSKCC poor risk, 74.7% were IGCCCG-2 intermediate-to-very high risk, 21.8% had liver, bone or brain (LBB) mets, 16.7% were late relapses. 35 pts (18.5%) had a complete response (CR), 68 (35.9%) a marker normalization (PRm-) (major response-rate: 54.4%). 41/68 PRm- were rendered disease-free (total NED-rate: 40.2%). After a median follow-up of 122.1 mos (IQR 71.4-232), median (95%CI) PFS was 7.2 mos (6.2-9.5) and median OS was 21.7 mos (16.7-69.5). 2yr PFS was 34.3% (28.1-41.9) and 5yr OS was 42.1% (35.3-50.2). Mediastinal primary (HR 3.06, 95%CI, 1.68-5.58), LBB mets (HR 2.02, 95%CI, 1.30-3.16), and AFP>1000 ng/mL (HR 2.78, 95%CI, 1.51-5.14) were negative prognostic factors for OS at MVA while no effect was seen according to IGCCCG-2 category. 70.3% of grade 3-4 neutropenia (25% febrile neutropenia), 48.1% thrombocytopenia, 21.1% anemia, 3% neurotoxicity, and no severe renal toxicity were recored. 81 pts (78% before 2000) had a dose reduction of ≥1 drug for ≥1 cycle. No discontinuations for toxicity occurred. Conclusions: Dose-modified Italian PEI showed activity comparable with the best ones achievable by conventional-dose CT (CDCT) in an unselected patient population, and a favorable toxicity profile. Results are among the most robust available for a CDCT, and should be considered as an appropriate benchmark to be compared with high-dose CT.