A randomized phase II study of pemetrexed (PEM) with or without sorafenib (S) as second-line therapy in advanced non-small cell lung cancer (NSCLC) of nonsquamous histology: NCCTG N0626 study.

Julian R. Molina, Grace K. Dy, Nathan R. Foster, Katie L. Allen Ziegler, Alex A. Adjei, K. M. Rowland, Marie Christine Aubry, Patrick J. Flynn, S. J. Mandrekar, Steven E. Schild,

Journal of Clinical Oncology · 2011 · 24 citations · 0 references

Concepts

Abstract

7513 Background: Based on preclinical data of additive cytotoxicity with sorafenib (S) and pemetrexed (PEM), this randomized phase II study was done to evaluate the safety and efficacy of the combination of S plus PEM vs PEM alone, as second line treatment of non-squamous NSCLC (determined by central pathology review). Methods: A lead-in phase to determine the tolerability of S plus PEM was followed by a 1:1 randomization of either PEM 500 mg/m2 q21 days plus S 400 mg p.o. BID for 21 days (arm A) or PEM 500 mg/m2 q21 days (arm B). The primary endpoint was progression-free survival (PFS). 47 patients (pts) per arm provided 82% power to detect an increase in the median PFS from 3.0 to 5.5 months (mos), with a 1-sided alpha of 0.05. Polymorphisms in VEGFA, VEGFR1-3, FPGS, GGH, SLC19A1 and TYMS were evaluated in germline DNA and correlated with outcomes. Results: One-hundred evaluable pts were randomized (49-A, 51-B). Baseline age, gender, and ECOG performance status were balanced between arms (p > 0.24). Prior exposure to bevacizumab (BEV) was 43% in A and 59% in B. Median number of cycles was 2 (1-31) on A and 4 (1-22) on B. A had a significantly increased rate of grade 3 non-hematologic (NH) adverse events (AE) (76% vs. 39%, p < 0.001). Grade 4/5 NH AE rates were similar for A and B (14% vs. 16%, p=1.00). No significant difference in grade 4 hematologic AE (20% vs 10.0%, p = 0.17). The overall median follow-up was 13.6 mos (1.6-27.8). PFS and OS was similar (PFS: A=3.4 mos vs B=4.1 mos, p = 0.22; OS: A=9.4 mos, B=9.7 mos, p=0.49). In arm A, the median PFS for patients with no prior BEV exposure was 2.8 mos compared to 5.0 mos in those with prior BEV exposure (p=0.06). In arm B, the median PFS was 3.5 vs 4.2 mos (p=0.91). The CT+TT genotypes of VEGFA rs3025039 was associated with worse PFS in A than in B (interaction p=0.03). Conclusions: Although the trial did not meet the protocol specified success criteria, post-hoc sub group analyses suggests a marginal improvement in PFS for patients on S plus PEM with prior exposure to BEV. Polymorphisms in VEGFA appear to predict for a worse outcome in patients receiving S. These preliminary findings warrant further investigation.