P3.02c-083 Treatment Related Adverse Events Predict Improved Clinical Outcome in NSCLC Patients on KEYNOTE-001 at a Single Center: Topic: IT Biomarkers

Abstract

It has been suggested that certain patients could be primed to respond to anti-programmed cell death-1 (PD-1) therapy due to heightened baseline “immunocompetence,” but data supporting this is limited as is our ability to measure it. The experience with ipilumumab suggests that immune related adverse events (irAEs) experienced by melanoma patients may predict improved clinical outcomes (Weber et al, J Clin Oncol 2012). We retrospectively analyzed NSCLC patients from a single center on the KEYNOTE-001 trial and evaluated the association between treatment related adverse events (trAE) and clinical outcomes. We performed a retrospective analysis of the 97 NSCLC patients treated on KEYNOTE-001 at UCLA with either 2 mg/kg Q3W or 10 mg/kg Q2/3W of pembrolizumab (data cut-off 3/2016). Investigators reported AEs and graded according to CTCAE v4.0, labeling them as unlikely, possibly, or probably treatment related. AEs labeled as possibly/probably related were considered trAEs. The initial scan was at 9 weeks and subsequent scans were every 9 weeks. Investigator assessed irRC was the radiographic assessment used for clinical decisions at individual sites. Progression-free survival (PFS) and overall survival (OS) were calculated using the Kaplan-Meier method and compared using the log-rank test. 10% (85/826) of AEs reported on trial were considered trAEs. The most frequent trAEs were rash (29%), fatigue (9%), and pneumonitis (8%). The occurrence of a trAE was associated with higher objective response rate (ORR) (OR=0.1509, P=0.0009), PFS (HR=0.3004, P<0.0001) and OS (HR=0.4391, P<0.0001). To assess whether the shorter duration of follow-up in those progressing earlier biased this analysis, additional analyses were performed. The relationship remained, particularly for longitudinal outcomes, when assessed only in patients that continued on trial >9 weeks. This was true both when including trAEs over the entire trial duration (ORR: OR=0.1839, P=0.005; PFS: HR=0.3525, P<0.0001; OS: HR=0.4526, P=0.0008) or when including only trAEs occurring within the first 9 weeks (ORR: OR=0.4063, P=0.1047; PFS: HR=0.5568, P=0.0211; OS: HR=0.6404, P=0.0465). Neither number of prior lines of therapy nor age, gender, or smoking history predicted frequency of trAE occurrence. This single center, retrospective analysis, revealed that a trAE predicted for improved clinical outcome with pembrolizumab. When controlling for the inherent bias of asymmetric follow-up, these associations remained. Although this analysis has the weakness of being conducted at a single center representing less than 20% of patients on trial, the strength is that a limited number of investigators assessed if an event was an AE and was treatment related.