부산대학교 도서관

Purpose: NSCLCs with activating EGFR mutations or EML4-ALK translocation respond to tyrosine kinase (TK) inhibitors and may also be sensitive to ionizing radiation. Because radiotherapy for intracranial metastasis utilizes a standardized planning approach, we examined local control for this clinical scenario to determine if TK mutations are associated with increased radiosensitivity. Methods: Analysis was performed on 239 NSCLC patients prospectively enrolled in an IRB approved GammaKnife (GK) database between 2005 and 2012. Tumor mutations for 81 patients were determined through sequencing (EGFR or KRAS) or FISH (EML4- ALK), and revealed 21 EGFR, 9 ALK, 17 KRAS, and 34 tumors wild type for these mutations. Excluding resection cavities, 469 metastatic brain lesions in 79 patients were treated with GK (164 EGFR mutant, 61 ALK, 105 KRAS mutant, and 139 wild type). The median prescribed dose was 18-20 Gy for all groups (range 16-24 Gy), and mean lesion diameter was 0.68 to 0.88 cm (standard deviation = 0.49 cm). Treatment plans and surveillance MRIs were reviewed to assess local control. Recurrence was defined as ≥ 20% increase in lesion diameter. Consensus amongst a multi-disciplinary team or pathology after resection superseded radiographic criteria. Estimates of local control were calculated using a Cox proportional hazards model with a robust sandwich covariance estimator to account for intra-cluster dependence. Results: 9/79 (11%) patients developed in-field recurrence after GK radiotherapy. 0/164 metastases with EGFR mutation and 0/61 metastases with ALK translocation recurred in-field, corresponding to 0/30 patients with activating TK mutations that recurred in- field. In contrast, 10/139 (7%) lesions in 6/32 (19%) wild type patients recurred in-field. 3/105 (3%) lesions in 3/17 (18%) KRAS mutant patients recurred in-field. Rates of distant brain recurrence were similar between patients with and without TK mutations (16/30 [53%] and 23/49 [47%], respectively). The median in-field recurrence time was significantly longer for TK-activated compared to wild type patients (not reached for TK- mutated vs. 18.4 months for wild type, p < 0.0001), whereas median distant brain recurrence time was equivalent (7.7 months for both groups, p = 0.97). Conclusions: Local control of NSCLC brain metastases treated with radiation therapy was superior for tumors harboring either an EGFR TK domain mutation or the EML4-ALK translocation, while tumors that tested negative had expected rates of in-field failure. This data suggests that TK driver mutations, in particular those of the EGFR, may identify a subset of tumors with increased sensitivity to radiation therapy. Disclosure: No significant relationships. [ABSTRACT FROM AUTHOR]