2017 Research Grant: NF1 and KEAP1 mutations are correlated to increased rates of local failure after radiation therapy for spine metastases from non-small cell lung cancer.

The spine is a common site of metastasis from non-small cell lung cancer (NSCLC), a disease hallmarked by heterogeneity in molecular alterations. Traditional spine radiation treatment (RT) with conventionally fractionated external beam radiation therapy (cEBRT) is associated with high local failure (LF) rates in patients surviving more than 12 months. Treatment with advanced techniques including stereotactic body radiation therapy (SBRT) significantly reduces LF. We hypothesized that specific genomic mutations in NSCLC metastases impart increased risk of LF, and that treatment with SBRT would prevent LF in metastases harboring such alterations. Retrospective review, single institution. NSCLC patients with tumor molecular profiling data who underwent spine RT with follow-up at a single institution. Local control of irradiated lesions (radiographic), overall survival. Under an IRB-approved retrospective research protocol, we examined the records of 189 pathologically-confirmed lung adenocarcinoma patients who had undergone radiation therapy for spine metastases at our institution and had tumor molecular profiling via the FDA-authorized MSK-IMPACT tumor-profiling multiplex panel platform. Mutation profiles, radiation treatment details and patient follow-up were recorded. Local control was defined radiographically using spine MRI when available and CT or PET-CT in a minority of cases. cEBRT was defined as a biologically effective dose (BED) <45Gy in 2Gy equivalents (BED 2Gy [α/β=10] <45Gy), while SBRT was defined as treatment in ≤5 fractions with BED 2Gy >45Gy. Cumulative incidence of LF was calculated using a competing risk analysis with R Statistical Software and examined for significance using Gray's test, stratified by alteration in canonical NSCLC driver genes. Survival after RT was calculated by the Kaplan-Meier method using GraphPad Prism software and examined for significant differences using the Log-rank (Mantel-Cox) test. Finally, radiation sensitivity was tested by clonogenic survival assay in isogenic NSCLC cell lines with CRISPR-mediated disruption of candidate genes. We identified 189 NSCLC patients with spine metastases who underwent RT (cEBRT = 81, SBRT = 108), had available tumor profiling by MSK-IMPACT, and had at least one follow-up imaging study. There was an overall 12-month LF rate of 13.9% for all patients. When stratified by RT type, LF at 12 months after cEBRT was 24.4% vs 5.7% after SBRT (p<0.001). Median survival after RT was not significantly different between cEBRT and SBRT groups (13.0 vs 19.9 months, p=0.06). We reviewed molecular alteration rates in 20 canonical NSCLC driver genes and found that 14 had alteration rates > 5%. Mutations in NF1 were associated with increased LF after RT (HR 3.63; 95% CI 1.58-8.33; p=0.002), and mutations in KEAP1 trended toward correlation with LF (HR 1.94; 95% CI 0.92-4.10; p=0.08). When stratified by RT type, mutations in NF1 and KEAP1 were significantly associated with LF after cEBRT (NF1 : HR 3.72, 95% CI 1.58-8.73, p=0.003; KEAP1 : HR 2.29, 95% CI 1.00-5.26, p=0.05), but not after SBRT (NF1: HR 2.52, 95% CI 0.31-20.7, p=0.39; KEAP1 : HR 1.31, 95% CI 0.26-6.56, p=0.74). CRISPR-mediated disruption of NF1 with two independent guide sequences resulted in reduced radiation sensitivity in two NSCLC cell lines (H661 and H1299), as assayed by clonogenic survival assay. LF after RT for spine metastases from lung adenocarcinoma is common, with 12-month incidence of LF=13.9% in this cohort. Progression at 12 months is more common after cEBRT as compared to SBRT (24.4% vs 5.7%), suggesting that a large subset of patients may benefit from upfront SBRT. Alterations in NF1 and KEAP1 are associated with increased risk of local progression after cEBRT, while failure after SBRT remains low even in tumors harboring these mutations. Importantly, these findings are supported by our novel preclinical data that NF1 loss of function reduces radiation sensitivity in NSCLC cells. Furthermore, it confirms preclinical results from independent laboratories showing a role for KEAP1 in radiation sensitivity. We propose that upfront SBRT for NSCLC patients harboring mutations in NF1 and KEAP1 is appropriate for prospective study. Finally, we suggest that future radiation dose escalation trials contain pre-specified stratification on molecular alteration patterns. MSK-IMPACT Tumor-Profiling Multiplex Panel. FDA-authorized. [ABSTRACT FROM AUTHOR]