KEY POINTS
- This single-center retrospective study screened 473 patients with stage IV NSCLC and brain metastases treated with SRS from 2018–2023; 161 with available next-generation sequencing comprised the analysis, including 81 KRAS-mutant and 80 KRAS-wild-type patients. EGFR- and ALK-mutant disease was excluded.
- Intracranial treatment was CyberKnife SRS with a 1-mm PTV margin. 70% received single-fraction SRS; common schedules were 20 Gy ×1 in 52%, 18 Gy ×1 in 17%, 27 Gy/3 in 13%, 30 Gy/5 in 10% and 25 Gy/5 in 8%. Fifty-two percent had undergone craniotomy.
- CNS recurrence was similar with and without KRAS mutation: 31% versus 36% (p≈0.4). At six months, CNS progression-free survival was 54% versus 49%, and at 12 months 40% versus 34%, with neither comparison statistically significant.
- Lesion-level analysis included 412 treated brain metastases. Only 17 lesions (4.1%) recurred locally; actuarial local control was 96.4% at six months and 92.3% at 12 months, with virtually identical 12-month control for KRAS-mutant versus wild-type disease (92.6% vs 91.8%; p=0.78).
- Survival likewise did not demonstrate a KRAS penalty. Median overall survival was 19 versus 15 months (p=0.11) for KRAS-mutant versus wild-type disease, while extracranial progression-free survival was 14 versus 9 months (p=0.61). KRAS status remained nonsignificant after propensity matching.
- Radiation dose emerged as more informative than molecular status. After propensity matching, BED10 ≥48 Gy was independently associated with better CNS progression-free survival (HR 0.38; p=0.01) and overall survival (HR 0.44; p=0.03). Younger age, good performance status, smaller metastases and controlled extracranial disease were also favorable in broader analyses.
- Important limitations include retrospective design, absence of specific KRAS subtype data such as G12C/G12D/G12V, incomplete systemic-treatment information and molecular testing in only 161/473 patients (34%). Residual imbalance remained after matching, so the study supports non-exclusion from SRS rather than proving identical intrinsic radiosensitivity.
CLINICAL TAKEAWAY
These data provide no reason to withhold or modify SRS simply because an NSCLC brain metastasis is KRAS-mutant. Intracranial outcomes were much more closely related to established clinical factors and dose intensity, although prospective studies incorporating individual KRAS variants and modern KRAS-targeted therapy are still needed.