Variation in FOLFOX, FOLFIRI, and FOLFOXIRI effects on CD8+ T cell and PDL1 levels in MSS CRC patients.

Abstract

3526 Background: Colorectal cancer (CRC) is a heterogenous disease treated with FOLFOX, FOLFIRI, and FOLFOXIRI chemotherapy regimens. About 85% of CRC patients have microsatellite stable (MSS) tumors that resist immune checkpoint blockade (ICB). Some studies suggest that chemotherapy modulates the MSS CRC tumor microenvironment (TME) to enhance CD8+ T cell infiltration, but ICB remains ineffective. We evaluated the TME of MSS CRC patients to investigate chemotherapy impact on immune markers. Methods: CRC patient samples (n = 16,827) underwent DNA (592-gene or whole exome)/RNA (whole transcriptome) sequencing at Caris Life Sciences. Immune cell fractions within TMEs were estimated from RNA deconvolution (quanTIseq; Finotello, 2019; n = 11,109). PDL1 expression was assessed by IHC (SP142; ≥2+/5%). Patients who received FOLFOX (n = 425), FOLFIRI (n = 88), or FOLFOXIRI (n = 19) < 1 year prior to tumor collection or who didn’t receive these treatments > 4000 days before tumor collection (untreated, n = 6,608) were analyzed. Statistical significance was determined using chi-square, Fisher’s exact, and Mann-Whitney U tests, where appropriate. Results: FOLFOX-treated CRC (n = 213) had a higher CD8+ T cell fraction vs untreated (n = 3,449, FC = 1.9, p < 0.01) CRC and there was no difference among FOLFIRI-treated (n = 37, FC = 1.1, p = 0.13) or FOLFOXIRI-treated (n = 9, FC = 2.2, p = 0.49) CRC. Survival outcomes were similar in FOLFOX-treated CD8+ T cell-high (n = 82) vs low (n = 161, HR = 1.3, p = 0.12) and in FOLFOX-treated PDL1+ (n = 13) vs PDL1- (n = 338) CRC (HR = 1.6, p = 0.14). The CD8+ T cell fraction was similar in untreated KRAS wild-type (WT, n = 1714) vs untreated KRAS-mutated (mut, n = 1718) CRC (FC = 1.3, p < 0.01). FOLFOX-treated KRAS-mut CRC had a higher CD8+ T cell fraction (n = 1718 untreated, 96 treated, FC = 2.3, p < 0.01) and CD69 expression (n = 1724 untreated, 97 treated, FC = 1.7, p < 0.01) vs untreated KRAS-mut CRC. This FOLFOX-dependent effect on CD8+ T cells was lower in WT KRAS CRC (n = 114 FOLFOX-treated, n = 1714 untreated, FC = 1.6, p < 0.01) but survival outcomes were similar in FOLFOX-treated WT (n = 210) vs KRAS-mut (n = 210) CRC (HR = 1.0, p = 0.82). The CD8+ T cell fraction was similar in untreated right- (R; n = 803) vs left-sided CRC (L; n = 927, FC = 1.3, p < 0.01) but PDL1+ IHC rates were significantly higher in R-CRC (n = 1340 L, 1124 R, FC = 2.2, p < 0.01). The CD8+ T cell fraction was similar in untreated WT (n = 3144) vs BRAF-mut CRC (n = 248, FC = 0.8, p < 0.01), but BRAF-mut were more frequently PDL1+ (n = 4622 WT, 340 mut, FC = 4.2, p < 0.01). PDL1+ rates were similar in untreated (n = 1124) vs FOLFIRI-treated R-CRC (n = 12, FC = 2.0, p = 0.46) but were higher in FOLFIRI-treated (n = 9) vs untreated (n = 1340) L-CRC (n = 9, FC = 6.0, p = 0.04). Conclusions: We describe an association between FOLFOX and subsequent infiltration by CD8+ T-cells and high PDL1+ rates in R-CRC, BRAF-mut CRC, and FOLFIRI-treated L-CRC. The findings provide insights for future therapeutic strategies.