Key points are not available for this paper at this time.
e20591 Background: Immune check point inhibitors (ICIs) are the cornerstone of treatment for patients with advanced NSCLC. However, not all patients benefit from ICIs and some develop significant immune-mediated toxicities. There is limited data regarding prognostic factors in NSCLC patients treated with ICIs. Imaging features can complement clinical variables in predicting response to ICIs. This study aims to evaluate CEA, blood leukocyte markers, and novel CT-derived radiomic features as factors predictive of survival in this patient population. Methods: In this retrospective study, clinical data from patients with advanced stage NSCLC treated with ICIs at Stony Brook University Hospital from 2016 to 2021 were collected. Baseline and subsequent CEA (carcinoembryonic antigen), ANC (absolute neutrophil count), ALC (absolute lymphocyte count), NLR (neutrophil to lymphocyte ratio), AEC (absolute eosinophil count), and AMC (absolute monocyte count) were evaluated as independent prognostic factors. 3-D segmentation of ipsilateral lung lobe was obtained using U-Net and radiologist report of tumor location. 600 stable IBSI-standardized features were extracted from the region using PyRadiomics. The discriminative ability of clinical and image-derived variables was assessed using Kaplan-Meier survival and Cox proportional hazards regression analysis. Results: This study included 80 patients. Objective response rate (ORR) was 25.0% with 1-year OS of 58.0%. PD-L1 expression ≥ 50% was significantly associated with survival (p < 0.02). Decrease in CEA (p < 0.01) and 5% increase in AEC (p = 0.027) significantly correlated with survival. Age, sex, clinical stage, ANC, ALC, NLR, AMC did not significantly predict survival. Three image biomarkers that correlated with median enhancement of tissue on chest CT, and degree of tissue heterogeneity were significant predictors of mortality (p < 0.02). A multivariable model with clinical inputs outperformed one with radiomic features only (C-index 0.70 vs. 0.64). Clinical and radiomic inputs together performed well in predicting survival (C-index 0.76). Conclusions: Changes in CEA and AEC were the strongest predictors of survival in advanced NSCLC patients treated with ICIs. This study suggests that combined clinical and radiomic biomarkers (quantifying local lobar heterogeneity) can identify patients most likely to benefit from immunotherapy. Table: see text
Huang et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: