Hypofractionated Stereotactic Radiotherapy (HFSRT) of the Resection Cavity vs. Whole-Brain Radiotherapy (WBRT) Following Brain Metastasis Resection Results of the ESTRON Randomized Phase 2 Trial

Purpose/Objective(s) Radiotherapyof the resection cavity following brain metastasis (BM) resection improves local control. Single-session stereotactic radiosurgery (SRS) can reduce toxicity compared to whole-brain radiotherapy (WBRT). Hypofractionated stereotactic radiotherapy (HFSRT) can deliver a higher biologically effective dose, potentially improving local control over SRS. The ESTRON trial compared HFSRT of the cavity with post-operative WBRT in patients with 1-10 BM. Materials/Methods Eligible patients with 1-10 newly diagnosed BM and following the resection of one BM regardless of histology were randomized 1:1 to receive HFSRT of the cavity and SRS of all unresected lesions or WBRT. A dose of 7 × 5 Gy was delivered to the cavity with a safety margin of 3-4 mm based on highly sensitive and standardized magnetic resonance imaging (MRI). The primary endpoint was intracranial control. Key secondary endpoints were overall survival (OS), safety, quality of life and neurocognitive function. Results 54 patients were randomized; HFSRT n=27, WBRT n=27. Most common histologies were non-small cell lung cancer (54%) and breast cancer (13%). Median cavity diameter was 28 mm (Q1-Q3: 22-33). 27 patients (52%) had at least one and 14 patients (26%) had >3 unresected BM. 15 patients (29%) had residual tumor at the cavity. These characteristics did not differ significantly between arms. At 12 months, intracranial control (IC) was 44% in the HFSRT and 59% in the WBRT arm. Median intracranial progression-free survival was 4.7 months (Q1-Q3: 1.7-18.0) in the HFSRT arm and 15.2 months (Q1-Q3: 5.6-12.7) in the WBRT arm (HR 1.72, 95%-CI: 0.94; 3.17, p = 0.080). In multivariable Cox-Regression, incomplete resection was predictive of inferior IC (HR 2.18, 95%-CI: 0.99; 4.79, p = 0.052). 3-year Local control (LC) at the cavity was 96% in the HFSRT arm and 89% in the WBRT arm (p=0.116) with a median follow-up of 19.7 months. Progression occurred distant from the cavity in 96% of cases (n = 26 out of 27). Leptomeningeal disease (LD) occurred in 7 patients (26%) in the HFSRT arm and 2 patients (7%) in the WBRT arm. LD occurred out of field in 6/7 cases in the HFSRT arm. 4 patients (15%) in the HFSRT arm and none in the WBRT arm developed blood brain barrier disruption at the cavity; 2 of those were symptomatic requiring treatment. Median OS was 10.3 months (Q1-Q3: 3.7-19.5) in the HFSRT arm and 18.6 months (Q1-Q3: 7.2-26.6) in the WBRT arm (HR 1.38, 95%-CI: 0.72; 2.66, p = 0.336). Conclusion Following the resection of BM, HFSRT of the resection cavity provides excellent local control. Compared to WBRT, distant intracranial progression and LD remain a relevant risk and rigorous follow-up is warranted. Detailed analyses on neurocognitive function are pending. Radiotherapyof the resection cavity following brain metastasis (BM) resection improves local control. Single-session stereotactic radiosurgery (SRS) can reduce toxicity compared to whole-brain radiotherapy (WBRT). Hypofractionated stereotactic radiotherapy (HFSRT) can deliver a higher biologically effective dose, potentially improving local control over SRS. The ESTRON trial compared HFSRT of the cavity with post-operative WBRT in patients with 1-10 BM. Eligible patients with 1-10 newly diagnosed BM and following the resection of one BM regardless of histology were randomized 1:1 to receive HFSRT of the cavity and SRS of all unresected lesions or WBRT. A dose of 7 × 5 Gy was delivered to the cavity with a safety margin of 3-4 mm based on highly sensitive and standardized magnetic resonance imaging (MRI). The primary endpoint was intracranial control. Key secondary endpoints were overall survival (OS), safety, quality of life and neurocognitive function. 54 patients were randomized; HFSRT n=27, WBRT n=27. Most common histologies were non-small cell lung cancer (54%) and breast cancer (13%). Median cavity diameter was 28 mm (Q1-Q3: 22-33). 27 patients (52%) had at least one and 14 patients (26%) had >3 unresected BM. 15 patients (29%) had residual tumor at the cavity. These characteristics did not differ significantly between arms. At 12 months, intracranial control (IC) was 44% in the HFSRT and 59% in the WBRT arm. Median intracranial progression-free survival was 4.7 months (Q1-Q3: 1.7-18.0) in the HFSRT arm and 15.2 months (Q1-Q3: 5.6-12.7) in the WBRT arm (HR 1.72, 95%-CI: 0.94; 3.17, p = 0.080). In multivariable Cox-Regression, incomplete resection was predictive of inferior IC (HR 2.18, 95%-CI: 0.99; 4.79, p = 0.052). 3-year Local control (LC) at the cavity was 96% in the HFSRT arm and 89% in the WBRT arm (p=0.116) with a median follow-up of 19.7 months. Progression occurred distant from the cavity in 96% of cases (n = 26 out of 27). Leptomeningeal disease (LD) occurred in 7 patients (26%) in the HFSRT arm and 2 patients (7%) in the WBRT arm. LD occurred out of field in 6/7 cases in the HFSRT arm. 4 patients (15%) in the HFSRT arm and none in the WBRT arm developed blood brain barrier disruption at the cavity; 2 of those were symptomatic requiring treatment. Median OS was 10.3 months (Q1-Q3: 3.7-19.5) in the HFSRT arm and 18.6 months (Q1-Q3: 7.2-26.6) in the WBRT arm (HR 1.38, 95%-CI: 0.72; 2.66, p = 0.336). Following the resection of BM, HFSRT of the resection cavity provides excellent local control. Compared to WBRT, distant intracranial progression and LD remain a relevant risk and rigorous follow-up is warranted. Detailed analyses on neurocognitive function are pending.