938P Nutritional status (NS) as a strong predictor for immunotherapy (IT) outcome in recurrent or metastatic (R/M) head and neck squamous cell carcinoma (HNSCC) patients (pts)

Lung Immune Prognostic Index (LIPI) based on derived neutrophil–lymphocyte ratio (dNLR) and LDH levels is a validated prognostic factor in IT exposed populations. We investigated whether LIPI and additional clinical and tumor characteristics predict outcomes in R/M HNSCC pts treated with IT. Review of a prospective database of R/M HNSCC pts treated with IT at VHIO from 2014-2019 was conducted. dNLR, LDH and other variables related to host NS [Prognostic Nutritional Index (PNI) = 10 × albumin (g/dL) + 0.005 × lymphocyte count (cells/mm)], body mass index and phosphate (Phos) levels were collected at baseline and treated as continuous variables. The primary endpoint was progression free survival (PFS) calculated with the Kaplan-Meier method. Univariate and multivariable Cox models were fitted and the proportion of explained variation of each factor in the final PFS model was calculated. Out of 101 pts, median age was 64y, 96% ECOG ≤1, treated with single agent IT 38% or IT combinations 62%. Only 8% were HPV-related oropharynx cancer. Baseline NGS was available for 44 tumors: 28 (64%) were TP53 mutated and 10 (23%) had cyclin pathway alterations. Median prior lines were 1 (range 1 - 5) with a median follow-up of 23.8 months (m) and a median OS of 15 m (CI95% 11.6 - 2.5). Median PFS on IT was 4 m (CI95% 3 – 6.5). Significantly better PFS was observed for the absence of liver metastases (HR: 0.37; 0.16 - 0.85; p= 0.02), ECOG 0 vs ≥1 (HR: 0.64; 0.4 – 1; p= 0.05), PNI index (HR: 0.94; 0.9 – 0.97; p= 0.001) and Phos levels (HR: 0.45; 0.25 – 0.8; p= 0.007). However, dNLR (HR: 1.0; 0.94 – 1.1; p= 0.9) and LDH (HR: 1.0; 0.91 – 1.1; p= 0.93) were not significantly associated with PFS. In a multivariable model, ECOG, PNI and Phos levels remained independent prognostic factors. Of note, the relative contribution of NS (PNI + Phos) in the multivariable PFS prediction model was 67%. In our cohort, ECOG, PNI and phosphate levels were the strongest variables associated with IT PFS in R/M HNSCC pts. These results suggest that parameters informative of NS should be considered when stratifying pts in IT trials. The lack of predictive power for LIPI determinants deserves further investigation.