Abstract 573P
Background
Tumour infiltrating CD8+ cytotoxic T cells confer favourable prognosis in colorectal cancer (CRC). The added prognostic value of other immune cells is unclear.
Methods
We used multiparameter co-immunofluorescence to determine the density of CD8+, CD20+, FoxP3+ and CD68+ cells in the intraepithelial (IE) and intrastromal (IS) compartments in stage II/III CRCs from the SCOT and QUASAR2 trials. We determined marker prognostic and predictive value by Cox regression of SCOT cases, with validation in the QUASAR2 cohort.
Results
Our study included 2,340 SCOT and 1,069 QUASAR2 cases. Univariable analysis of recurrence-free interval (RFI) in SCOT revealed expected prognostic value of intraepithelial CD8 (CD8IE) (continuous variable: HR for cases at 75th vs 25th percentile (HR75vs25)= 0.73, 95% CI=0.68-0.79, P=2.5e-16) and, less expectedly FoxP3 within tumour stroma (FoxP3IS) (HR75vs25)= 0.71, 95% CI=0.64-0.78, P=1.5e-13), but not epithelium. CD8IE and FoxP3IS were independently prognostic in multivariable analysis, and a composite marker including both (CD8IE-Foxp3IS) was superior to either alone as a continuous variable (adjHR75vs25= 0.70; 95%CI=0.63–0.78; P=5.1e-11), and when categorised into three groups (Low/Int/High) using sample medians (Int vs High HR = 1.60, 95%CI=1.24–2.05, P=2.5e-04; Low vs High HR = 2.30, 95%CI=1.80–2.95; P=3.6e-11). These results were validated in the QUASAR2 trial. Pooled analysis of both studies revealed interactions between CD8IE-FoxP3IS and tumour sidedness (adjP INT=0.010), stromal proportion (adjP INT=0.018), and MMR status (adjP INT=0.079). CD8IE-FoxP3IS refined stage III risk groups, with clinical low risk (T1-3, N1), CD8IE-FoxP3IS low cases demonstrating similar 3-year recurrence-free probability to clinical high risk (T4 and/or N2), CD8IE-FoxP3IS high cases (79.2 vs 76.4%, adj log-rank P=0.16).
Conclusions
Combined evaluation of CD8IE and FoxP3IS is superior to single markers and refines CRC risk stratification. FoxP3IS prognostic value is inconsistent with an immunosuppressive role; further investigation of these cells as an immunotherapy target is merited.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
D.N. Church.
Funding
Cancer Research UK, National Institutes for Health Research, Promedica Foundation.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
518P - Artificial intelligence real-world applications in pediatric neuro-oncology: The AICCELERATE project
Presenter: Federica D'Antonio
Session: Poster session 10
519P - The landscape of PDGFRA mutation in Chinese patients with glioma
Presenter: Qiang Lv
Session: Poster session 10
520P - Copy number variation spectrum analysis of primary glioblastoma
Presenter: Chuandong Cheng
Session: Poster session 10
521P - Deciphering a three-miRNA signature as a prognostic biomarker in glioma patients: Correlation with DFS and OS
Presenter: Ana Belen Diaz Mendez
Session: Poster session 10
522P - Galectin 3 binding protein as potential biomarker in glioma diagnosis
Presenter: Rashmi Rana
Session: Poster session 10
523P - Analysis of DNA damage response (DDR) gene expression as a prognostic factor for glioblastoma patient mortality
Presenter: Alessia-Tara Droesse
Session: Poster session 10
524P - Cell line study of nucleosome-based biomarkers in the diagnosis and detection of relapses in glioblastoma
Presenter: Jonathan Decarpentrie
Session: Poster session 10
525P - Immuno markers in newly glioblastoma patients underwent Stupp protocol after neurosurgery
Presenter: Lorena Gurrieri
Session: Poster session 10
526P - In silico evaluation of the mutational profile of glioblastomas with high expression of PD1, CTLA4 and LAG3 identifies the ERBB-PI3K pathway as a druggable vulnerability target
Presenter: Cristina Saiz-Ladera
Session: Poster session 10
527P - Targetable gene fusions and other alterations in central nervous system tumors assessed by RNA and DNA-based next-generation sequencing
Presenter: LEIMING WANG
Session: Poster session 10