Epirubicin/paclitaxel/carboplatin (TEC) vs paclitaxel/carboplatin (TC) in first-line treatment of ovarian cancer (OC) FIGO stages IIB-IV. An AGO-GINECO Intergroup phase III trial.

5007 Background: The addition of further drugs to TC is regarded as an option for improving 1st-line therapy in OC. Anthracyclines have shown activity in OC and are among the drugs suitable for evaluation in triple drug regimens. METHODS We performed a randomized phase III trial in pts. with advanced OC stages FIGO IIB-IV to determine the role of epirubicin (E) when added to TC. Pts were randomized to receive 6 cycles of either T 175 mg/m2 3 h iv + C AUC 5 (according to Calvert formula) + E 60 mg/m2 iv (TEC) or TC at same doses, both regimens given every 3 weeks. Pts. were stratified per center and according to FIGO stage and post-operative residual tumor (RT): stratum I 'FIGO IIB-IIIC and RT =/<1 cm'; stratum II 'FIGO IIB-IIIC and RT > 1 cm or FIGO IV'. Primary endpoint was overall survival (OS), secondary endpoints were progression-free survival (PFS), overall response (OR; complete + partial), toxicity, quality of life, and stratified analysis of OS. RESULTS From 11/97 to 2/00 1,282 pts. were randomized and 647 and 635 were enrolled in the TEC and TC arm, respectively. 743 pts (58%) were classified as stratum 1. More than 86% of pts received 6 or more courses. TEC induced more myelotoxicity, e.g. grade (G) 3/4 neutropenia was observed in 75.6% and 55.5% after TEC and TC (p<.0001), but G 3/4 febrile neutropenia was rare and occurred in 5.5% pts in TEC and in 1.3% pts in TC (p<.0001). Addition of E induced more G3 nausea (6.9 vs 3.3%, p=.004), G3/4 emesis (6.4% vs 2.8%, p=.002), and mucositis (1.8% vs 0.2%, p=.004), but did not increase cardiac or other non-hematological toxicity. Response data were available from 137 and 158 pts. with measurable disease in the TEC and TC arm. TEC was associated with a 74% OR and TC showed 70% OR (ns). All pts were analyzed for PFS and OS; 968 pts. had progressed and 732 pts (57%) had died. The median follow-up for pts. who remained alive was 52 mos. Median PFS for pts with/without E was 18.4 vs 17.9 mos (HR 0.94; 95% CI: 0.83-1.07). Median OS was 45.8 and 41.0 mos for TEC and TC, respectively (HR 0.93; 95% CI 0.83-1.07). Median OS in stratum 1 and stratum 2 were 57.3 and 28.1 mos (HR 0.46, 95% CI 0.40-0.53; p<.0001). However, the stratified comparison of TEC vs TC did not show any significant differences with respect to PFS or OS. Exploratory post hoc subgroup analysis did not show differences with respect to residual tumor (0 vs 1-10 vs > 10 mm), but for histological subtypes; the latter suggested superior efficacy for TEC in endometrial OC. CONCLUSIONS The addition of E to TC did not result in superior OR, PFS and OS, but induced more toxicity and costs. Therefore, TEC cannot be recommended as first-line therapy of advanced OC in clinical routine. Further research should focus on the role of anthracylines in endometrial OC. [Table: see text].