Radiation necrosis in renal cell carcinoma brain metastases treated with checkpoint inhibitors and radiosurgery: An international multicenter study
Eric J. Lehrer, Jason Gurewitz, Kenneth Bernstein, Dev Patel, Douglas Kondziolka, Ajay Niranjan, Zhishuo Wei, L. Dade Lunsford, Timothy D. Malouff, Henry Ruiz-García, Samir Patel, Phillip A. Bonney, Lindsay Hwang, Cheng Yu, Gabriel Zada, David Mathieu, Claire Trudel, Rahul N. Prasad, Joshua D. Palmer, Brianna M. Jones, Sonam Sharma, Kareem Fakhoury, Chad G. Rusthoven, Christopher P. Deibert, Piero Picozzi, Andrea Franzini, Luca Attuati, Cheng‐Chia Lee, Huai‐Che Yang, Manmeet S. Ahluwalia, Jason P. Sheehan, Daniel M. Trifiletti
Abstract
BACKGROUND Patients with renal cell carcinoma (RCC) brain metastases are frequently treated with immune checkpoint inhibitors (ICIs) and stereotactic radiosurgery (SRS). However, data reporting on the risk of developing radiation necrosis (RN) are limited. METHODS RN rates were compared for concurrent therapy (ICI/SRS administration within 4 weeks of one another) and nonconcurrent therapy with the χ 2 test. Univariable logistic regression was used to identify factors associated with developing RN. RESULTS Fifty patients (23 concurrent and 27 nonconcurrent) with 395 brain metastases were analyzed. The median follow‐up was 12.1 months; the median age was 65 years. The median margin dose was 20 Gy, and 4% underwent prior whole‐brain radiation therapy (WBRT). The median treated tumor volume was 3.32 cm 3 (range, 0.06‐42.38 cm 3 ); the median volume of normal brain tissue receiving a dose of 12 Gy or higher (V12 Gy) was 8.42 cm 3 (range, 0.27‐111.22 cm 3 ). Any‐grade RN occurred in 17.4% and 22.2% in the concurrent and nonconcurrent groups, respectively ( P = .67). Symptomatic RN occurred in 4.3% and 14.8% in the concurrent and nonconcurrent groups, respectively ( P = .23). Increased tumor volume during SRS (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.01‐1.19; P = .04) was associated with developing RN, although V12 Gy (OR, 1.03; 95% CI, 0.99‐1.06; P = .06), concurrent therapy (OR, 0.74; 95% CI, 0.17‐2.30; P = .76), prior WBRT, and ICI agents were not statistically significant. CONCLUSIONS Symptomatic RN occurs in a minority of patients with RCC brain metastases treated with ICI/SRS. The majority of events were grade 1 to 3 and were managed medically. Concurrent ICI/SRS does not appear to increase this risk. Attempts to improve dose conformality (reduce V12) may be the most successful mitigation strategy in single‐fraction SRS.