RESUMEN
BACKGROUND: Most patients with advanced non-small cell lung cancer (NSCLC) treated with first-line pembrolizumab monotherapy will experience progressive disease (PD). Only a minority will go on to receive subsequent systemic anticancer therapy for which outcomes are guarded. We investigated the prognostic significance of biomarkers of systemic inflammation following failure of first-line pembrolizumab for NSCLC to aid subsequent management decisions. METHODS: Patients with radiological and/or clinical evidence of PD on first-line pembrolizumab for advanced NSCLC at a regional Scottish cancer centre were identified. Inflammatory biomarkers at the time of PD, including serum albumin, neutrophil count and the Scottish Inflammatory Prognostic Score (SIPS; combing albumin and neutrophils), and clinicopathological factors, including age, sex, histology, PDL1 expression and time to PD were recorded. The relationship between these and post-progression overall survival (ppOS) were examined. RESULTS: Data were available for 211 patients. Median ppOS was 2.1 months. Only SIPS was predictive of ppOS on multivariate analysis (HR2.54 (95 %CI 1.81-3.56) (<0.001)), stratifying ppOS from 0.8 months (SIPS2), to 1.8 months (SIPS1), to 8.1 months (SIPS0) (p < 0.001). Thirty (14 %) patients received second-line systemic anticancer therapy with median ppOS 8.7 months. These patients had lower levels of systemic inflammation, as defined by albumin (p < 0.001), neutrophil count (p = 0.002), and SIPS (p = 0.004)), than all other patients. CONCLUSIONS: SIPS, a simple biomarker of systemic inflammation, predicts ppOS after first-line pembrolizumab and may be useful alongside routine assessments of patient fitness to inform individualised discussions about subsequent treatment. We highlight poor outcomes in this patient group and a role for SIPS in signposting transition to best supportive care and early referral to palliative care. It may also help identify a small group of patients most likely to benefit from further lines of therapy.