Response Rate And Local Recurrence After Concurrent Immune Checkpoint Therapy And Radiation For Non-Small Cell Lung Cancer And Melanoma Brain Metastases

Abstract

Growing evidence suggests that immune checkpoint therapy (ICT) and radiation (RT) may be synergistic in the treatment of brain metastases (BrM), but the optimal timing of therapy to maximize synergy remains unclear. We evaluated the impact of treatment timing on response rate and local recurrence in patients with non-small cell lung cancer (NSCLC) or melanoma BrM. We reviewed the records of BrM patients who received ICT and brain-directed RT between 2007-2016 at our institution. ICT consisted of ipilimumab, pembrolizumab, and/or nivolumab; RT included stereotactic radiosurgery/radiotherapy (SRS/T) or whole brain RT (WBRT). We evaluated response rate and local recurrence on a per-metastasis basis (unidimensional RANO-BM). Individual metastases were included only if they underwent RT within 90 days of ICT, to minimize selection biases while analyzing the impact of treatment timing on outcomes. Within this cohort, concurrent treatment was defined as receiving RT on the same day as a dose of ICT or in between doses of an ICT course; all other treatment was non-concurrent (but still within 90 days). Multivariable Cox proportional hazards models were used to assess time to response and local recurrence on a per-metastasis basis, adjusting for confounding variables and using a sandwich estimator to account for intra-patient correlation. The final cohort included 110 patients treated for 340 BrM, with 102 BrM treated concurrently and 238 BrM treated non-concurrently. In total, 68% had melanoma and 32% had NSCLC; 59% of BrM were treated with SRS/T and 41% with WBRT; and 44% received ipilimumab, 46% received anti-PD-1 therapy, and 9% received both simultaneously. Median radiographic follow-up after RT was 6.9 months (IQR 3.5-13.3 months). Rates of any response (complete or partial) were higher with concurrent treatment (70% vs 47%, p<0.001), with correspondingly lower rates of progressive disease (5% vs 26%). On multivariable analysis, time to response was improved with concurrent treatment (HR 1.78, 95% CI 1.18-2.66, p = 0.005), and local recurrence was also decreased with concurrent treatment (HR 0.46, 95% CI 0.26-0.82, p = 0.009). Only 1/103 BrM with a complete response later progressed. In a sensitivity analysis comparing concurrent treatment (n = 102 BrM) to treatment that was non-concurrent but within 30 days (n = 110 BrM), there remained improved rates of any response with concurrent treatment (70% vs 53%, p<0.001). Concurrent brain-directed RT and ICT improves response rates and decreases local recurrence of NSCLC and melanoma BrM, even when compared to treatment that was non-concurrent but within 90 days. Further study of this combination is warranted, particularly in patients in whom rapid treatment response is needed.