Brain metastases as the primary site of relapse in two randomized phase III pemetrexed trials in advanced non-small-cell lung cancer.

BACKGROUND: Symptomatic brain metastases (BM) frequently occurs after initial treatment of non-small-cell lung cancer (NSCLC). Therefore, 2 large randomized trials that involved pemetrexed were retrospectively analyzed to determine the pattern of symptomatic relapse in the brain and to gauge if pemetrexed could influence the incidence.

METHODS: Two large phase III studies of pemetrexed in advanced NSCLC were included. One study compared pemetrexed with docetaxel in previously treated patients (n = 571); the other study tested cisplatin plus pemetrexed vs. cisplatin plus gemcitabine in chemotherapy-naive patients with advanced NSCLC (n = 1725). Patients with known BM at study entry were excluded from this analysis. Each study was analyzed separately, then jointly to determine the rate of BM reported as the only site of progressive disease by treatment group and histology. Logistic regression was used to obtain an odds ratio for the treatment effect on the overall occurrence of BM while controlling for potential confounding factors.

RESULTS: Overall, 71.5% of patients in pemetrexed-containing arms (819 of 1145), and 68.2% of patients in non-pemetrexed-containing arms (785 of 1151) experienced progressive disease. BM recurrence rates were 3.2% (95% confidence interval [CI], 2.1%-4.6%) in the pemetrexed-containing arms vs. 6.6% (95% CI, 5.0%-8.6%) in the non-pemetrexed-containing arms (P = .002). The odds ratio for BM recurrence associated with exposure to pemetrexed was 0.49 (95% CI, 0.32-0.76; P = .001). The beneficial effect of pemetrexed on BM was confined to patients with nonsquamous NSCLC.

CONCLUSIONS: Patients with advanced nonsquamous NSCLC treated with pemetrexed either in first-line or second-line therapy may reduce the risk of developing BM as the first site of progressive disease. This retrospective analysis is limited due to the lack of baseline and periodic brain scans, and it reflects symptomatic BM only. Regardless, these findings suggest a potential beneficial effect of pemetrexed-based treatments on the control of BM.