Repeat stereotactic radiosurgery for treatment of brain metastases locally recurrent following initial radiosurgery.

Purpose: We report outcomes of repeat stereotactic radiosurgery (rSRS) to sites of tumor progression following initial SRS. Additionally, we sought to determine if, at the time of recurrence following initial SRS, surgical resection of the tumor followed by SRS (surgery + rSRS) provided benefit compared to rSRS alone.

Methods: We retrospectively reviewed patients treated with rSRS for local recurrence after initial SRS. The cumulative incidences of LF and adverse radiation effect (ARE), with death as a competing risk, were estimated.

Results: From 2004 to 2022, we identified 77 patients with 429 brain metastases treated with initial SRS, of which 97 metastases were treated with rSRS for salvage of LF following initial SRS; 49 metastases had resection prior to rSRS. Of the 429 brain metastases treated with initial SRS, the cumulative incidence of LF was 12.6% [95% confidence interval (CI) 9.7-15.9] at 1 year; in 97 sites treated with rSRS, LF was 14.6% (95%CI 8.4-22.4) at 1 year. There was no significant difference (p = 0.3) in 1-year LF after surgery + rSRS [11.1% (95%CI 4.0-22.31)] versus rSRS alone [18.4% (95%CI 9.0-30.5)]. The 1-year rates of ARE were: 3.0% (95%CI 1.7-5.0%) for initial SRS (overall grade 1-4), 15.6% (95%CI 9.2-23.6) for rSRS (overall grade 1-4), and 12.6% (95%CI 6.8-20.1%) for rSRS (symptomatic grade 2-4).

Conclusion: Given that the 1-year local progression of 15% with rSRS is similar to the 13% of initial SRS, our data do not support that tumors recurrent after initial SRS are inherently radioresistant to salvage SRS. Tumor control must be balanced by the 1-year rates of adverse radiation effect (16% overall, 13% symptomatic).