Effect and safety of immune checkpoint inhibitors for brain metastases from non-small cell lung cancer

Abstract

Abstract Background Immune checkpoint inhibitors (ICPi) have dramatically changed the treatment of non-small cell lung cancer (NSCLC), but the effect of these agents on brain metastases (BM) has not been clearly elucidated. We retrospectively evaluated the effect of ICPi on BM. Methods NSCLC patients who harbored BM treated with ICPi in out institution were enrolled. History of radiation therapy (RT) to the brain was not included in the selection criteria. The primary endpoint was overall survival after the initiation of ICPi (OS), and the secondary endpoints were duration of ICPi administration, causes of withdrawal, responses of BM, time to the first response of BM, and the incidence of radiation necrosis (RN). Results 35 NSLC with BM were treated with ICPi: nivolmab in 21, pembrolizumab in 13, and athezolizumab in 1 case. Among these cases, BM were progressive in 25 but stable in 10 cases at the initiation of ICPi. At the time of evaluation, ICPi were already withdrawn in 26 cases and the median administration time was 2.4 (0.5 to 25.9) months. The cause of withdrawal was progression of extracranial lesions in 8, progression of BM in 6, and adverse events or PS deterioration in the others. In competing risk analysis, the times to progression of extra- and intra-cranial lesions were not different. The responses of BM were evaluable in 28 cases: CR in 3, PR in 9, SD in 10 and PD in 6 cases. ICPi had been sequentially administrated after RT in one of the 3 CR and all of the 9 PR cases. The objective response rate was 76.9% after sequential RT/ICPi, but only 13.3% after ICPi alone (P\xa0=\xa00.002). The median time to the first response of BM was 1.0 (0.4 to 3.0) months after RT/ICPi and 1.3 (0.3 to 3.6) months after ICPi alone (n.s.). The median OS was 14.8 (95% CI: 8.8 to 22.7) months after the initiation of ICPi, and majority (90.5%) of the deaths were owing to the progression of extracranial disease. Among the 24 cases with history of RT to the brain, RN was observed in 7 (29.2%) and required necrotomy in 2 cases, even though ICPi could safely be continued after craniotomy in these 2 cases. Conclusions ICPi alone showed only limited efficacy on BM and were recommended to be administrated sequentially with RT. RN was the major neurological adverse event, and necrotomy was one of the choices to manage RN during ICPi. Legal entity responsible for the study The authors. Funding Has not received any funding. Disclosure All authors have declared no conflicts of interest.