Stage-dependent divergence of immune prognostic signatures in non–small cell lung cancer.

P Praveen Thapa (Chirayu National Hospital, Kathmandu, Nepal) T Trilok Shrivastava (University of Arkansas Medical Sciences, Little Rock, AR) A Ashish Nepal (WellSpan York Hospital, York, PA) A Anupam Raj Upadhyay (Stupa Community Hospital, Kathmandu, Nepal)

Abstract

8066 Background: Tumor mutational burden (TMB) and immune activation are widely studied biomarkers in non–small cell lung cancer (NSCLC), yet their prognostic relevance outside the context of immunotherapy remains unclear. We investigated how age, histology, and disease stage modify the relationships between tumor genomics, immune contexture, and overall survival in NSCLC. Methods: We analyzed The Cancer Genome Atlas NSCLC cohort from the “Pan-cancer atlas” (2018) with a total of 1053 patients evaluating TMB, CD8A expression, cytolytic activity score (CYT), age at diagnosis, histology, and overall survival. Associations among continuous variables were assessed using Spearman correlation. Multivariable Cox proportional hazards models adjusted for age, sex, and tumor mutational burden were performed and stratified by stage (I–II vs III–IV). Results: Age demonstrated an inverse correlation with tumor mutational burden (ρ = −0.12, P < 0.001) but was not associated with CD8A expression and showed a weak positive association with CYT, driven by non-squamous tumors. Tumor mutational burden was not associated with overall survival in either early- or late-stage disease. In contrast, immune biomarkers demonstrated marked association with stage of disease. Higher CD8A expression was associated with improved survival in early stage (HR 0.77 per SD, P = 0.021) but not in advanced disease. CYT exhibited a significant interaction with stage with higher CYT showing association with improved survival in early-stage NSCLC (HR 0.79 per SD, P = 0.026) but with worse survival in late-stage disease (derived HR ≈ 1.14). Tumor stage remained the dominant prognostic factor across all models. Conclusions: In untreated NSCLC, immune prognostic signatures are strongly stage dependent. Immune infiltration and cytolytic activity confer survival benefits in early-stage disease but not in advanced disease, where cytolytic activity may reflect aggressive tumor-associated inflammation. Tumor mutational burden does not provide prognostic value across disease stages. These findings highlight the importance of disease context when interpreting immune biomarkers and provide critical baseline insight for immunotherapy-era studies. Patient characteristics. Characteristic Overall (N=1053) Early Stage (I–II) (N=517) Late Stage (III–IV) (N=479) Age, median (IQR) 67 68 67 Sex, % male 40.24 (402) 44.68 (231) 35.69 (171) Histology, %  Squamous 46.24 (487) 46 (238) 51.35 (246)  Non-squamous 53.7 (566) 56 (279) 50 (233) Stage  I–II 47%  III–IV 53% TMB, median (IQR) 7.3 (3.9-12.1) 7.35 (3.7-12.1) 7.36 (4.2-12.1) CD8A expression, median (IQR) 232 (112-469) 240 (122-471) 208 (103-460) CYT score, median (IQR) 183 (103.9-368.3) 191.3 (109.6-368.3) 175.9 (98-370)

Article Details

Volume / Issue Vol. 44, Issue 16_suppl
Published June 01, 2026
Pages 8066-8066
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (4)

P

Praveen Thapa

Chirayu National Hospital, Kathmandu, Nepal

T

Trilok Shrivastava

University of Arkansas Medical Sciences, Little Rock, AR

A

Ashish Nepal

WellSpan York Hospital, York, PA

A

Anupam Raj Upadhyay

Stupa Community Hospital, Kathmandu, Nepal