Lipid metabolic gene expression and association with decreased overall survival and immunogenicity in <i>KRAS-STK11</i> NSCLC.

J Joshua Pothen (University of Illinois Chicago, Chicago, IL) J Jen Godden (Tempus AI, Chicago, IL) D Denise Shieh (TEMPUS AI, Chicago, IL) S Stamatina Fragkogianni (Tempus AI, Inc.) J Jacob Mercer (Tempus AI, Inc., Chicago, IL) M Melissa Conrad Stoppler (Tempus AI, Chicago, IL) D Daniel Principe (Department of Medicine, Physician Scientist Training Program, University of Wisconsin, Madison, WI) R Ryan Huu-Tuan Nguyen (University of Illinois College of Medicine at Chicago, Division of Hematology and Oncology, Chicago, IL) K Kamya Sankar K Koosha Paydary (Rush University Medical Center, Chicago, IL) M Mary J. Fidler (Rush University Medical Center, Chicago, IL) F Frank Weinberg (University of Illinois College of Medicine at Chicago, Division of Hematology and Oncology, Chicago, IL)

Abstract

8534 Background: Approximately 30% of patients (pts) with non-small cell lung cancer (NSCLC) have alterations (alt) in KRAS . How co-alt such as STK11 affect the tumor microenvironment and survival requires further characterization. Recently, our group found that lower expression of lipid metabolic genes in KRAS G12C co-alt tumors was associated with worse overall survival (OS). Here, we seek to confirm these findings utilizing a large real-world (rw) patient de-identified database. Methods: Approximately 2,187 pts (61%, stage IV) with KRAS G12C alt NSCLC who underwent sequencing via the Tempus xT/xR assay with co-alt in TP53 (44%) , STK11 (17%), or LRP1B (4%) were selected. The groups are mutually exclusive. Single-sample GSEA (ssGSEA) based on 775 lipid metabolic genes (LMG) was used to calculate enrichment scores (LMG ES) for each pt. Pts were dichotomized into low vs. high groups based on their median LMG ES. Immune cell infiltration predicted from gene expression patterns, TMB, and PD-L1 from IHC was evaluated. Risk-set adjusted rwOS was calculated from sample collection date to death from any cause. Hazard ratios (HR) were calculated using Cox proportional hazards model, and p-values were calculated using the Wald test. Results: Among pts with KRAS G12C alt, the median age was 68, 58% were female, and 84% were White. Pts with KRAS G12C/ STK11 alt had the lowest TMB, neoantigen burden, and PDL-1 positivity compared to other cohorts (p&lt;0.001 for all). Importantly, the proportion of total immune cells, M1, M2, NK cells, CD8 T cells and regulatory T cells was lowest in tumors with KRAS G12C/ STK11 alt (p&lt;0.001 for all). To determine if lipid genes were associated with immunogenic changes, LMG ES was compared to immune infiltration. The ES was associated with immune cell infiltration percentages for M1 macrophages (OR 1.11 (1.03-1.21) p=0.012), M2 macrophages (OR 1.27 (1.15-1.40) p&lt;0.001) and neutrophils (OR 1.12 (1.04-1.22) p=0.005), with a trend towards significant association with CD4 T cells (OR 1.08 (1.00-1.17) p=0.062). Pts with KRAS G12C/STK11 alt and low LMG ES had decreased median rwOS ( 5.4 vs 18.2 months, p=0.0002 ) compared to pts with a high ES. Multivariate analysis demonstrated that lower LMG ES correlated with reduced rwOS (HR = 1.75 (1.22-2.51, p = 0.002) compared to pts with high ES. Individual gene analysis showed that low LPL (HR = 1.85 (1.147-2.97) p=0.012), LDLRAD4 (HR = 1.72 (1.082-2.72) p = 0.022) and LDLR (HR = 1.58 (1.009-2.46) p=0.045) expression was associated with poorer rwOS. Conclusions: Low lipid gene expression in KRAS - STK11 NSCLC was associated with decreased OS. Lipid gene expression and tumor immune cell infiltration were associated, suggesting that lipid metabolism may regulate tumor immunogenicity. These data suggest that lipid metabolic genes should be further explored as potential therapeutic targets for pts with NSCLC and KRAS - STK11 alt.

Article Details

Volume / Issue Vol. 43, Issue 16_suppl
Published June 01, 2025
Pages 8534-8534
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (12)

J

Joshua Pothen

University of Illinois Chicago, Chicago, IL

J

Jen Godden

Tempus AI, Chicago, IL

D

Denise Shieh

TEMPUS AI, Chicago, IL

S

Stamatina Fragkogianni

Tempus AI, Inc.

J

Jacob Mercer

Tempus AI, Inc., Chicago, IL

M

Melissa Conrad Stoppler

Tempus AI, Chicago, IL

D

Daniel Principe

Department of Medicine, Physician Scientist Training Program, University of Wisconsin, Madison, WI

R

Ryan Huu-Tuan Nguyen

University of Illinois College of Medicine at Chicago, Division of Hematology and Oncology, Chicago, IL

K

Kamya Sankar

K

Koosha Paydary

Rush University Medical Center, Chicago, IL

M

Mary J. Fidler

Rush University Medical Center, Chicago, IL

F

Frank Weinberg

University of Illinois College of Medicine at Chicago, Division of Hematology and Oncology, Chicago, IL