RTHP-35. LOCAL RECURRENCE PATTERN OF PATIENTS ENROLLED ON A RANDOMIZED STUDY OF POST-OPERATIVE STEREOTACTIC RADIOSURGERY VS OBSERVATION FOR COMPLETELY RESECTED BRAIN METASTASIS Free

To evaluate the difference in patterns of local recurrence (LR) on a prospective randomized trial of stereotactic radiosurgery (SRS) or observation (OBS) after complete resection of a brain metastasis.

All patients who experienced LR within or adjacent to the surgical cavity were identified and the failure pattern was evaluated. Failures were categorized as completely In-Field, Out-Field or Both based on the prescription isodose for the SRS arm or potential target volume if OBS. Characteristics including time to failure, initial size, histology, radiation dose are evaluated.

Of the 128 eligible patients enrolled on study, 46 (15 SRS, 31 OBS) experienced LR (p=0.005). Six patients (5 SRS, 1 OBS) had LR before SRS or by post-op day 30 and were not included in this analysis. The median time to LR in the remaining group was 250d vs 115d (range: 126-488d vs 39-424d) in the SRS vs OBS arms, respectively (p=0.02). The median initial tumor diameter was 3.4cm vs 2.7cm for the LR vs non-LR, respectively, in both SRS and OBS. There was no statistically significant difference in recurrence rates between histologies, but, LR incidence after SRS was 0/3 and 0/12 in patients with renal and lung cancer vs 4/9 and 4/9 patients in breast and GI cancers (p=0.07). Recurrence pattern in the SRS patients was noted to be In-Field, Out-Field and Both in 40%, 30% and 30%, respectively. Similarly, for the OBS patients, these rates were 31%, 21% and 48%, respectively.

Post-operative SRS for brain metastasis decreases and delays LR. It should be noted that 6/128 (4.6%) patients had LR by post-op day 30. Based on the outcomes of the remaining patients, improvements in dose and target volume delineation may be possible for higher risk patient cohorts based on tumor size and histology.