RESUMO
OBJECTIVE: Publications on adjuvant stereotactic radiosurgery (SRS) are largely limited to patients completing SRS within a specified time frame. The authors assessed real-world local recurrence (LR) for all brain metastasis (BM) patients referred for SRS and identified predictors of SRS timing. METHODS: The authors retrospectively identified BM patients undergoing resection and referred for SRS between 2012 and 2018. Patients were categorized by time to SRS, as follows: 1) ≤ 4 weeks, 2) > 4-8 weeks, 3) > 8 weeks, and 4) never completed. The relationships between timing of SRS and LR, LR-free survival (LRFS), and survival were investigated, as well as predictors of and reasons for specific SRS timing. RESULTS: In a cohort of 159 patients, the median age at resection was 64.0 years, 56.5% of patients were female, and 57.2% were in recursive partitioning analysis (RPA) class II. The median preoperative tumor diameter was 2.9 cm, and gross-total resection was achieved in 83.0% of patients. All patients were referred for SRS, but 20 (12.6%) did not receive it. The LR rate was 22.6%, and the time to SRS was correlated with the LR rate: 2.3% for patients receiving SRS at ≤ 4 weeks postoperatively, 14.5% for SRS at > 4-8 weeks (p = 0.03), and 48.5% for SRS at > 8 weeks (p < 0.001). No LR difference was seen between patients whose SRS was delayed by > 8 weeks and those who never completed SRS (48.5% vs 50.0%; p = 0.91). A similar relationship emerged between time to SRS and LRFS (p < 0.01). Non-small cell lung cancer pathology (p = 0.04), earlier year of treatment (p < 0.01), and interval from brain MRI to SRS (p < 0.01) were associated with longer intervals to SRS. The rates of receipt of systemic therapy also differed significantly between patients by category of time to SRS (p = 0.02). The most common reasons for intervals of > 4-8 weeks were logistic, whereas longer delays or no SRS were caused by management of systemic disease or comorbidities. CONCLUSIONS: Available data on LR rates after adjuvant SRS are often obtained from carefully preselected patients receiving timely treatment, whereas significantly less information is available on the efficacy of adjuvant SRS in patients treated under "real-world" conditions. Management of these patients may merit reconsideration, particularly when SRS is not delivered within ≤ 4 weeks of resection. The results of this study indicate that a substantial number of patients referred for SRS either never receive it or are treated > 8 weeks postoperatively, at which time the SRS-treated patients have an LR risk equivalent to that of patients who never received SRS. Increased attention to the reasons for prolonged intervals from surgery to SRS and strategies for reducing them is needed to optimize treatment. For patients likely to experience delays, other radiotherapy techniques may be considered.