Clinical and molecular characterization of <i>FAP</i> and <i>SPP1</i> in colorectal cancer (CRC), CALGB (Alliance)/SWOG 80405 and real-world data.

Abstract

276 Background: FAP and SPP1 contribute to immune modulation in the CRC tumor microenvironment (TME). FAP is expressed by cancer associated fibroblasts, aiding in tissue remodeling and tumor invasion. SPP1 is an integrin-binding protein expressed by tumor associated macrophages that promotes tumor growth, adhesion, and metastasis. We present a clinical and molecular characterization of FAP and SPP1 in CRC. Methods: We analyzed 24,257 CRC samples tested at Caris Life Sciences (Phoenix, AZ) with WTS (Illumina NovaSeq), NextGen DNA sequencing (NextSeq), and PD-L1 expression (SP142, positive ≥ 2+, 5%). RNA deconvolution analysis estimated cell infiltration in the TME. Data from the phase 3 CALGB/SWOG 80405 trial (NCT00265850) on 433 metastatic CRC patients treated with bevacizumab (Bev, n = 226) or cetuximab (Cet, n = 207) in combination with first-line chemotherapy were also evaluated. RNA isolated from FFPE tumor samples were sequenced with HiSeq 2500 (Illumina). Overall survival (OS) and progression-free survival (PFS) were compared using Cox regression in categorical gene expression tertiles (high (T3), medium (T2), and low (T1)) for FAP and SPP1 . Results: FAP -T3 and SPP1 -T3 had increased PD-L1 positivity (q&lt;0.05), while SPP1 -T3 also demonstrated increased MSI-H (8.4% vs 5.3%) and TMB-H(&gt;10 mt/mb, 15.0% vs 10.1%) status relative SPP1 -T1 (all q&lt;0.001). T3 of both genes of correlated with increased M1/M2 macrophages, NK cells and T cell inflamed score (q&lt;0.001); dendritic cells and neutrophils were increased in SPP1 -T3 and FAP -T1 tumors (q&lt;0.05). FAP -T3 and SPP1 -T3 had increased pathway activation of epithelial-mesenchymal transition (EMT), inflammatory response, TNF-a signaling, angiogenesis and KRAS signaling (all q &lt; .005). In Caris cohorts, FAP -T3 demonstrated worse OS in Cet/panitumumab treated CRC (T3: 23.6 vs T1: 21.0 months [mo], P = .005; HR 0.85, 95% CI [0.77-0.95]), but SPP1 did not correlate with OS. In 80405, FAP -T3 showed shorter PFS (T3: 9.5 vs T2: 11.5 vs T1: 12.6 mo, T3 vs T1 (reference) adjusted HR 1.27 [1.07-1.51]) and OS (25.2 vs 29.4 vs 35.5 mo, adjusted HR 1.31 [1.10-1.57]). Similarly, SPP1 -T3 demonstrated worse PFS (9.0 vs 12.7 vs 14.0 mo, adjusted HR 1.29 [1.12-1.48]) and OS (20.9 vs 34.0 vs 36.3 mo, HR 1.24 [1.07-1.44], P &lt; 0.001). Treatment interaction tests noted FAP -T1 CRC benefited from Cet over Bev with respect to PFS ( P = 0.003) and OS ( P = 0.044), while SPP1-T1 tumors demonstrated a PFS benefit with Cet ( P = 0.009). Conclusions: Our results indicate that increased FAP and SPP1 expression is associated with immune cell infiltration, EMT, and inflammatory signaling. Additionally, their expression may be prognostic and predictive of targeted therapy. These data support the evaluation of FAP and SPP1 as predictive markers and therapeutic targets in CRC.