Integrative analysis of tumor microenvironment in advanced pancreatic cancer: Unraveling genomic and immune landscape for targeted therapies.

C Catia Fava Gaspar (Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada) G Grégoire Marret (Division of Medical Oncology and Hematology, Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto, Canada) B Benson Z. Wu (1Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada) J Jeffrey Bruce (Princess Margaret Cancer Centre) S Simone C. Stone (Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada) B Ben X Wang (Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada) L Lillian L. Siu (Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto) G Gun Ho Jang A Amy Zhang A Anna Dodd J Julie Wilson G G. Zogopoulos (PanCuRx Translational Research Initiative, Ontario Institute for Cancer Research, Toronto, ON, Canada) E Elena Elimova (Princess Margaret Cancer Centre, Toronto) R Raymond Woo-Jun Jang (Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada) R Robert C. Grant F Faiyaz Notta S Steven Gallinger J Jennifer J. Knox G Grainne M. O'Kane (St Vincent's University Hospital, Dublin, Ireland) E Erica S. Tsang

Abstract

4182 Background: An understanding of genotype and immunophenotype interactions in advanced pancreatic ductal adenocarcinoma (PDAC) is important in designing combination strategies. In addition, PDAC subtypes may harbor unique tumor immune microenvironments (TMEs) and confer differential sensitivity to KRAS inhibitors (KRASi). We aimed to characterize the baseline TME in advanced PDAC and its relationship with genomic, transcriptomic and clinical data. Methods: The COMPASS trial (NCT02750657) investigated whole genome (WGS) and transcriptome (RNA-Seq) sequencing in patients (pts) receiving first line therapy for advanced PDAC. We performed multiplex immunohistochemistry (mIHC) to identify 5 immune cell subtypes (CD8+/CD4+ T cells, Tregs, B cells and macrophages) and CIBERSORT, a deconvolution method that uses gene expression profiles. Statistical analyses were performed using STATA/R software and significance was defined as p - value < 0.05. Multivariate logistic regression was used and Kaplan Meier analyses evaluated impact on survival. Results: Of 268 pts, 62 had available tissue samples with mIHC, WGS, and RNA-Seq data (n = 21 primary biopsies, n = 41 metastases, 34/41 liver). All 62 pts had KRAS mutations (28 G12D, 19 G12V, 10 G12R, 5 other) and 29 had KRAS major or minor imbalances. 55 cases (88.7%) were classified as classical subtype and HRDetect hi was seen in 10 pts, including 5 with BRCA1/2 mutations (4 germline, 1 somatic). In the overall cohort, differences between tumor and stroma were evident with increased infiltration of CD8 and CD4 Tcells and Tregs in stroma ( p < 0.001) and increased macrophages (p= 0.0343) in tumor. CIBERSORT in a subset of 51 pts demonstrated increased M0 (p= 0.0035) and M2 macrophages ( p = 0.0067) in liver metastases compared to primary samples, suggesting a more immunosuppressive TME. A higher number of B cells were seen in lung metastases (median 207.5 vs. 43.2 vs. 3.7 vs. 1.8 cells/mm2, p = 0.011) compared to abdominal wall, peritoneum and liver, respectively. Pts with KRAS major imbalance (n =14, 3 basal like) were found to have higher median numbers of CD8+ (114.5 vs. 25.1 vs. 27.5 cells/mm2, p = 0.038) and CD4+ Tcells (292.1 vs 133.4 vs. 97.4 cells/mm2, p = 0.005) when compared to minor/balanced samples, respectively. Basal-like PDAC had fewer macrophages than classical subtype (median 13.2 vs. 28.2 cells/mm2, p = 0.0103). On survival analysis, pts with HRDetect lo and classical subtype with higher macrophage counts had a tendency towards increased survival (median OS: 11.8 vs. 9.5 months, p = 0.066). Conclusions: We identified increased CD8/CD4 T cell infiltration in PDAC stroma, as well as in pts with KRAS major imbalance. Immune cell profiling may complement molecular profiling as potential biomarkers and warrants further study in this context.

Article Details

Volume / Issue Vol. 43, Issue 16_suppl
Published June 01, 2025
Pages 4182-4182
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (20)

C

Catia Fava Gaspar

Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada

G

Grégoire Marret

Division of Medical Oncology and Hematology, Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto, Canada

B

Benson Z. Wu

1Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada

J

Jeffrey Bruce

Princess Margaret Cancer Centre

S

Simone C. Stone

Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada

B

Ben X Wang

Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada

L

Lillian L. Siu

Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto

G

Gun Ho Jang

A

Amy Zhang

A

Anna Dodd

J

Julie Wilson

G

G. Zogopoulos

PanCuRx Translational Research Initiative, Ontario Institute for Cancer Research, Toronto, ON, Canada

E

Elena Elimova

Princess Margaret Cancer Centre, Toronto

R

Raymond Woo-Jun Jang

Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada

R

Robert C. Grant

F

Faiyaz Notta

S

Steven Gallinger

J

Jennifer J. Knox

G

Grainne M. O'Kane

St Vincent's University Hospital, Dublin, Ireland

E

Erica S. Tsang