Treatment-free interval (TFI) after first-line therapy (FLT) and disease control rate (DCR) on second-line therapy (SLT): Impact on overall survival (OS) in relapsed malignant pleural mesothelioma (MPM)

e18528 Background: Both TFI (the time elapsing from FLT completion to SLT initiation) and DCR (the rate of nonprogression at first tumor evaluation) have been found to predict OS in other tumor types. Methods: The impact on outcome of TFI after FLT and DCR on SLT was tested in an individual patient pooled analysis of 261 MPM patients (pts) who had radiologic progressive disease (PD) after a pemetrexed-based FLT before entering a ph 2 trial with single-agent NGR-hTNF (n=50) and a ph 3 trial with single-agent gemcitabine, vinorelbine or doxorubicin plus NGR-hTNF/placebo (n=211). In both trials, response to SLT was assessed every 6 weeks by MPM-modified RECIST. Progression-free survival (PFS) and OS were computed from SLT start. By ROC analysis, the cutpoint for estimating TFI in relation to OS was set at 6 months (AUC=0.59; p=.009). Results: After FLT, 60 pts (23%) had partial response (PR), 135 (52%) stable disease (SD), for a DCR of 75%, and 66 (25%) early PD. Median time to PD was 7.0 months (95% CI, 6.2-7.4) and median TFI was 4.4 months (3.8-5.0), with 97 pts (37%) having a TFI > 6 months. Among baseline factors (age, sex, PS and histology) used in logistic regression, only younger age was related to higher odds to attain a TFI > 6 months (OR=1.9; 95% CI, 1.2-3.3). A TFI > 6 months (vs ≤ 6) was weakly related to DCR (Spearman's r=0.16; p=.01) and strongly associated with longer PFS and OS, with HR of 0.48 (0.36-0.65) and 0.59 (0.42-0.83), respectively. After SLT, 9 pts (3%) had PR, 130 (50%) SD, for a DCR of 53%, and 122 (47%) early PD. Of 60 pts responding to FLT, 23 (38%) progressed early on SLT, while of 66 pts progressing early during FLT, 32 (48%) experienced disease control on SLT. Baseline factors did not relate with DCR at week 6. On landmark analysis set at the 6-week time point, DCR (vs PD) was associated with OS benefit (HR=0.46; 0.32-0.65), which persisted after adjusting for baseline factors (HR=0.40; 0.27-0.57). Conclusions: With a smaller HR, an early-look measure such as 6-week DCR on SLT shows some advantage over TFI after FLT in predicting subsequent OS. In relapsed MPM, TFI as stratification factor and DCR as surrogate endpoint may be considered. Clinical trial information: NCT00484276-NCT01098266.