In colorectal cancer (CRC), attempts to identify cancer cell-specific markers to guide antibody-mediated therapeutics have failed to uncover markers that are both exclusive to cancer tissues and abundant across CRCs. Alternatively, cancer-associated fibroblasts (CAFs), which are abundant in the tumor microenvironment and upregulate unique surface markers, are not found in healthy tissues. Here, we evaluated the expression patterns of CAF-associated proteins α-smooth muscle actin (αSMA), fibroblast activation protein (FAP), podoplanin (PDPN), matrix metalloproteinase-2 (MMP2), transgelin (TAGLN), and THY1. While αSMA and THY1 were abundant in cancer tissues, high abundance in normal tissues limited their targeting potential. FAP was present in 94.5% of primary and metastatic CRC tissues and absent in 93.7% of adjacent normal colon and liver tissues assessed. These results indicate that FAP is a promising target for antibody conjugates with potential for broad application in CRC. Co-expression analyses showed that CRCs simultaneously expressing high levels of PDPN, MMP2, and THY1 were enriched for immune-related signatures, indicating potential for antibody-mediated immune engagers. Overall, this work highlights the potential of CAF proteins to act as therapeutic targets for novel anticancer agents and become important therapeutic biomarkers.

Phosphatidylinositol-3-kinase (PI3K)/AKT/mammalian target of rapamycin (MTOR) pathway hyperactivation is seen in a multitude of malignancies. Due to the importance of this pathway in numerous critical cellular functions, preclinical and clinical investigations have aimed to target this pathway as an anti-cancer therapeutic strategy. This has led to the development of PI3K, AKT, and MTOR inhibitors for use in cancer patients, leading to multiple FDA approvals over the past decade. In this review, we outline therapeutic targets in PI3K/AKT/MTOR signaling in solid tumors, the current state of using inhibitors of this pathway to treat patients whose cancers possess activating mutations in PIK3CA, AKT1/2, or MTOR, and exciting new inhibitors that are entering clinical trials.

Targeting PIK3CA-mutant colorectal cancers with precision medicine strategies is of great clinical interest. However, resistance to single-agent PI3K pathway inhibitors has been observed across multiple clinical trials, necessitating the identification of combination therapies that overcome or prevent resistance to precision medicine strategies. Previously, our group identified that inhibition of mTORC1/2 is necessary to induce a response in PIK3CA-mutant colorectal cancers. The PI3K/mTORC1/2 inhibitor copanlisib has demonstrated some clinical activity in PIK3CA-mutant solid tumors as part of the NCI-MATCH trial. In this study, we evaluate potential combination therapies that could enhance the efficacy of copanlisib and other similar inhibitors in PIK3CA-mutant colorectal cancers. Using a novel high-throughput drug screen method in Apc- and Pik3ca-mutant mouse-derived cancer organoids, we identify navitoclax, a BCL-2 family inhibitor, as a drug that could potentially enhance the response to copanlisib. Across multiple in vitro and in vivo colorectal cancer models, navitoclax enhanced PI3K/mTOR inhibition (copanlisib, sapanisertib, and dactolisib) and induced apoptosis. Furthermore, we examine these combination therapies across a panel of patient-derived cancer organoids with a range of mutation profiles. These studies indicate that KRAS mutations could confer resistance. Furthermore, we identify BCL-xL as the major BCL-2 family target important for the response to this combination in this setting. This provides a strong rationale for mTORC1/2 and BCL-2 family inhibition as a potential treatment strategy for PIK3CA-mutant colorectal cancers.

Background/Objectives: Recent clinical trials in breast cancer have demonstrated that some patients benefit from immune checkpoint blockade, though better predictive markers are needed. The accumulation of the immunomodulatory matrix proteoglycan versican (VCAN) can predict the exclusion of CD8⁺ tumor-infiltrating lymphocytes (TILs) in some settings and, thus, is evaluated in breast cancer here. Methods: A total of 230 breast cancers were analyzed for VCAN accumulation, VCAN proteolysis, and CD8⁺ TILs. CD8⁺ TILs were categorized based on their localization in the tumor epithelial or stromal compartments. Results: VCAN accumulation was detected in 90% of breast cancers, more commonly in ER+ tumors (93% vs. 77%; p < 0.001). MCF7 cells treated with estrogen upregulate VCAN without an enhanced expression of ADAMTS-proteases. VCAN-undetectable tumors demonstrate greater CD8⁺ TILs compared to VCAN-detectable tumors (p = 0.012). CD8⁺ T cells within TNBC tumors with high VCAN proteolysis infiltrated the epithelial compartment more often than in tumors with low VCAN proteolysis (91% vs. 42% respectively; p = 0.008). In the TCGA cohort, a strong inverse correlation between CD8A and VCAN expression was observed across subtypes. Conclusions: VCAN accumulation correlates with the exclusion of CD8⁺ TILs across subtypes of breast cancer, warranting further validation of VCAN accumulation and proteolysis as predictive biomarkers for breast cancer immunotherapy.

Purpose: Versican (VCAN) is an immunoregulatory matrix proteoglycan that, when cleaved, releases an immunostimulatory fragment, versikine. In this study, we evaluate the impact of VCAN on immune surveillance and immunotherapy response in a prospective clinical trial. Patients and Methods: T-cell function was assayed in the setting of VCAN. Normal and matched tissues were acquired from 243 patients across colorectal cancer stages. A phase Ib clinical trial enrolled 15 subjects with microsatellite stable oligometastatic colorectal cancer and examined the safety and efficacy of the sequential combination of stereotactic body radiotherapy, pembrolizumab, and resection (NCT02837263). The outcomes were correlated with the VCAN status and circulating immune phenotype by single-cell RNA sequencing. Results: VCAN significantly accumulates in 59% of colorectal cancers and is heavily proteolyzed [VCAN proteolytic predominant (VPP)] in 27% (40% of oligometastatic colorectal cancers). VCAN decreased CD8+ T-cell trafficking (P < 0.01) and activated CD8+ T-cell effector function [decreased IL2RA (P < 0.05), PD-1 (P < 0.05), and granzyme B (GZMB; P < 0.001)]. The clinical trial met the primary endpoint with a 1-year recurrence-free survival (RFS) of 60%. Of those enrolled, 40% had the VPP phenotype, which was associated with a trend toward improved RFS (P = 0.053), and all are still alive with a median follow-up of 4.1 years. The VPP phenotype was associated with enhanced circulating CD8+ T-cell effector function at baseline, which was further enhanced with study treatment. Conclusions: VCAN limits T-cell trafficking and effector function, and its proteolysis correlates with an effector phenotype of circulating CD8+ T cells, greater tumor-infiltrating lymphocytes, and improved clinical outcomes with this immunotherapy-based clinical trial treatment.