Introduction: Immunotherapy Response Score (IRS) is a tissue-based multivariable algorithm previously validated to predict anti-PD-(L)1 benefit pan-tumor (PMID:36750617) from routine tissue samples. We evaluated IRS in an independent cohort of Kaiser Permanente Northern California (KPNC) patients with advanced non-small cell lung cancer (NSCLC) following a pre-specified statistical analysis plan for the primary objective.

Methods: NSCLC patients treated with pembrolizumab with IRS were included (timeframe 2018-2022). Primary endpoint included time to next treatment (TTNT). The primary objective was to determine if IRS status (High[-H] or -Low[L]) is associated with 1^st line pembrolizumab TTNT independent of mono-/combination therapy and PD-L1 Tumor Proportion Score (TPS) comparing full vs. reduced (excluding IRS) Cox proportional hazards models. Post-hoc analysis assessed IRS association with ≥1^st line monotherapy treated patients within PD-L1 TPS>50% to determine if IRS provided benefit beyond TPS.

Results: 214 1^st line NSCLC patients were treated with pembrolizumab monotherapy or combination therapy and IRS was significantly associated with pembrolizumab TTNT independent of therapy type and TPS​ (HR=0.70; LRT p=0.03). Post-hoc analysis demonstrated that IRS provided benefit within TPS≥50% as IRS-H patients had longer pembrolizumab monotherapy TTNT than IRS-L patients (Fig. 1, n= 66 median TTNT 21.8 95%CI [12.3, NA] vs. 9.0 95%CI [5.2 – 21.3] months).

Conclusions: IRS supports informed clinical decisions beyond TPS for determining if a patient with NSCLC will benefit from pembrolizumab monotherapy (IRS-H) or could consider combination therapy (IRS-L). Importantly, IRS stratified monotherapy benefit within PD-L1 TPS≥50% which may support clinical decision making.