Session Item

Immunotherapy with combined nivolumab and ipilimumab (CNI) is active in patients with asymptomatic melanoma brain metastases (MBM), providing a significant and durable intracranial response. Just as the treatment paradigms of MBM evolved over the last decade, overall survival (OS) also improved based both on local and systemic therapies. In this context, we evaluated the impact of cranial stereotactic radiotherapy (SRT) on OS of MBM patients treated with CNI in a contemporary and real-world setting.

The study was performed by using TriNetX, a global health network that provided a de-identified dataset of electronic medical records from approximately 68 million patients in 49 large Healthcare Organizations (HCO). We queried for patients with the terms: secondary malignant neoplasm of brain and malignant melanoma of skin andnivolumab and ipilimumab and cranial SRT (included only in the irradiated cohort) between 1 January 2016 and 31 December 2020. Exclusion criteria were defined by the terms: vemurafenib or cobimetinib or dabrafenib or trametinib or craniectomy for excision of brain tumor or stereotactic body radiation therapyor cranial SRT (excluded only in the non-irradiated cohort).

After building and retrieving the initial query for CNI and CNI+SRT cohorts, baseline characteristics were compared with chi‐squared tests for categorical variables and independent‐sample t tests for continuous variables. A propensity score matching was run to balance cohorts with the following baseline characteristics: age, sex, known ECOG status, extracranial metastases (secondary malignant neoplasms of adrenal gland, bone, liver, lung and lymph nodes) and use of glucocorticoids. OS was defined from the date on which the patient met all selected criteria for the cohort to date of death and estimated by Kaplan-Meier. Log-rank test was applied to compare the survival curves. Hazard ratio (HR) with 95% confidence intervals, together with the test for proportionality, was calculated. Statistical significance was prespecified as P <.05

The initial query retrieved 355 patients from 31 HCO in CNI cohort and 114 patients from 16 HCO in CNI+SRT cohort. After propensity score matching, 114 patients were selected in each cohort. Baseline characteristics are listed in Table 1. There were 52 and 42 deaths in CNI and CNI+SRT cohort, respectively. The median OS was 327 days in CNI and not reached in patients in CNI+SRT cohort, with a probability of survival at end of time window of 54.4% and 40.9%, respectively (log-rankP = .0057; Figure 1). Treatment with CNI+SRT was associated with a significantly decreased mortality (HR, 0.57; 95% CI 0.377-0.853; proportionality P = .0034).

This real-world, selective approach and propensity score matching analysis showed that CNI+SRT led to an improvement in OS compared to CNI, corroborating prior findings that local therapy with SRT or surgery improved OS in patients treated with CNI for MBM, regardless of the timepoint of local therapy.