Future Directions in Integrating Genomics and Immunotherapy for Gastrointestinal Cancers

A New Era for Precision Immunotherapy in GI Cancers

The Global Burden of Gastrointestinal Cancers

Gastrointestinal (GI) cancers represent a formidable global health challenge, accounting for 26% of all cancer incidence and 35% of cancer-related deaths worldwide. In the United States, colorectal, pancreatic, and liver cancers are the top three causes of GI cancer mortality. This alarming burden underscores an urgent need for novel therapeutic strategies that go beyond conventional chemotherapy and radiation. The integration of genomics and immunotherapy offers a promising path forward, aiming to tailor treatments to the unique molecular and immune profile of each tumor.

Proof of Principle: Success in Immunogenic Subtypes

Immunotherapy has already transformed outcomes for a subset of GI cancer patients. Tumors with high microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR), most commonly seen in colorectal and gastric cancers, are highly responsive to PD-1 blockade. In advanced MSI-H colorectal cancer, pembrolizumab has become a first-line standard, and dual checkpoint blockade with nivolumab plus ipilimumab yields even deeper responses. For locally advanced rectal cancer, neoadjuvant PD-1 blockade achieves clinical complete response rates of 75–100%, enabling organ-sparing approaches. In esophageal squamous cell carcinoma, first-line chemo-immunotherapy has become the standard of care based on phase 3 trials. These successes, however, are limited to a small fraction of patients—only 5–10% of metastatic GI tumors are MSI-H, and overall objective response rates to single-agent checkpoint inhibitors hover around 10–20%.

The Challenge of Immune-Cold Tumors

Most GI cancers, particularly pancreatic ductal adenocarcinoma (PDAC) and microsatellite-stable (MSS) colorectal cancer, are immunologically ‘cold’—they lack sufficient T-cell infiltration and harbor an immunosuppressive tumor microenvironment (TME). In PDAC, single-agent checkpoint inhibitors have shown negligible activity. This resistance arises from multiple factors: low tumor mutational burden, dense desmoplastic stroma enriched with cancer-associated fibroblasts (CAFs), dominance of M2-polarized tumor-associated macrophages (TAMs), and metabolic reprogramming that fuels immune evasion. Overcoming these barriers requires rationally designed combination strategies informed by detailed genomic and TME profiling.

Emerging Genomic Tools for Patient Stratification

The development of genomic signatures promises to extend immunotherapy benefits beyond the MSI-H subgroup. The Gastrointestinal Immune Prognostic Signature (GIPS), a six-gene mutation model (RNF43, CREBBP, CDKN2A, TP53, SPEN, NOTCH3), robustly predicts survival and durable clinical benefit in patients treated with checkpoint inhibitors across multiple cohorts. In validation studies, GIPS-high patients had median overall survival of 13–15 months compared to 5–6 months for GIPS-low, outperforming PD-L1 expression and TMB as predictive biomarkers. Importantly, combining GIPS with TMB further refines stratification, identifying a subgroup with the highest response rates and longest survival. Other promising tools include gene expression profiles (e.g., 18-inflammatory-gene signatures), the Immunoscore (based on CD3+ and CD8+ densities), and algorithms that integrate microbiome composition (e.g., Pseudoxanthomonas abundance in PDAC).

Combination Strategies to Convert Cold to Hot

Armed with these genomic insights, researchers are designing multi-pronged combinations. Targeting TAMs through CSF1R inhibition or reprogramming with agents like pomalidomide can convert an immunosuppressive TME into a responsive one. In preclinical PDAC models, combining anti-PD-1 with CSF1R or TGFβ inhibitors achieved higher cure rates. For pMMR/MSS colorectal cancer, the CAPability-01 trial showed that a triplet of sintilimab (anti-PD-1), chidamide (HDAC inhibitor), and bevacizumab (anti-VEGF) improved objective response to 44% versus 13% with doublet therapy. Bispecific T-cell engagers, CAR-T cells targeting Claudin18.2 (e.g., CT041), and neoantigen vaccines are also entering the clinic. Notably, in PDAC, an mRNA-lipoplex neoantigen vaccine (autogene cevumeran) combined with atezolizumab and chemotherapy induced durable T-cell responses and prolonged recurrence-free survival.

The Road Ahead: Multi-Omics and Dynamic Monitoring

The future of precision immunotherapy in GI cancers lies in integrating multi-omics data—genomics, transcriptomics, metabolomics, and metagenomics—with artificial intelligence to build predictive models. Dynamic monitoring using liquid biopsies (ctDNA, circulating tumor cells) will enable real-time adjustments. The shift from static biomarkers (PD-L1, MSI) to comprehensive, dynamic signatures will allow clinicians to select the right combination at the right time for each patient. Given the heterogeneity of GI cancers, large prospective trials are essential to validate these integrated approaches. The ultimate goal is to transform more immunologically cold tumors into hot ones, expanding the reach of immunotherapy beyond the current minority of responders.

Current Progress and Persistent Challenges in GI Immunotherapy

Is there immunotherapy for gastric cancer?

Yes, immunotherapy for gastric cancer is available for gastric cancer. The FDA has approved several options, including checkpoint inhibitors such as nivolumab, pembrolizumab, and dostarlimab, as well as targeted antibodies like ramucirumab, trastuzumab, and trastuzumab deruxtecan. These treatments harness the immune system to fight cancer cells, particularly in advanced stages where surgery or chemotherapy may no longer be effective. Immunotherapy can stabilize disease progression and improve prognosis for patients with advanced stomach cancer, and ongoing clinical trials continue to explore additional approaches.

Challenges in Single-Agent Checkpoint Inhibitor Therapy

Despite successes in specific subtypes, most gastrointestinal cancers remain resistant to single-agent checkpoint inhibitors. Microsatellite-stable (MSS) colorectal cancer and pancreatic ductal adenocarcinoma (PDAC) are notable examples. These tumors often feature low mutational burdens and immunosuppressive tumor microenvironments, limiting response rates to around 0–2% in MSS CRC and minimal efficacy in PDAC. Overcoming this resistance requires rationally designed combination strategies, such as pairing checkpoint blockade with targeted therapies, metabolic inhibitors, or immunomodulators.

Success Stories: Where Genomics Already Guides Treatment

Genomic testing for microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR) has transformed treatment for a subset of bowel cancer patients. This biomarker predicts exceptional responses to immunotherapy.

In a landmark phase II trial, neoadjuvant pembrolizumab (anti–PD-1) eliminated all signs of cancer in 59% of patients with dMMR/MSI-H colorectal cancer. For the remainder, any residual tumor was removed during surgery, leaving all 32 participants cancer‑free after the procedure. This far exceeds the less than 5% pathological complete response rate seen with standard chemotherapy.

Further success comes from dual checkpoint blockade. The NICHE‑2 trial achieved a 100% 3‑year disease‑free survival rate with neoadjuvant nivolumab plus ipilimumab, yielding pathological complete responses in 60–68% of patients. These outcomes demonstrate that genomics‑guided immunotherapy can deliver deep, durable remissions, sparing many patients the need for adjuvant chemotherapy.

Beyond Traditional Biomarkers: The Rise of Genomic Signatures

Limitations of Established Biomarkers

Current biomarkers for immunotherapy response in gastrointestinal (GI) cancers have significant shortcomings. Microsatellite instability-high (MSI-H) occurs in only 0–5% of metastatic cases, excluding most patients. PD-L1 expression suffers from interobserver variability and yields inconsistent results across clinical trials. Tumor mutational burden (TMB) lacks a standardized cutoff and weights all mutations equally, ignoring the distinct immunogenic impact of specific alterations (e.g., TP53, ARID1A). These limitations create an urgent need for more precise predictive tools.

The Gastrointestinal Immune Prognostic Signature (GIPS)

A newly developed genomic signature, GIPS, integrates mutations in six genes (RNF43, CREBBP, CDKN2A, TP53, SPEN, NOTCH3) into a cost-effective panel. Validated in multiple independent cohorts, GIPS-high status predicts significantly longer overall survival and higher durable clinical benefit rates compared to GIPS-low. Moreover, combining GIPS with TMB improves patient stratification, identifying those most likely to benefit from immunotherapy. This signature offers a promising future direction beyond traditional biomarkers.

Systematic Approaches: Stratifying Tumors by Immunogenicity

For these "cold" tumors, novel combinations aim to convert the immunosuppressive microenvironment into an immune-responsive state. Strategies include dual checkpoint blockade (anti-PD-1 plus anti-CTLA-4), targeting stromal components (CSF1R or TGFβ inhibitors), and epigenetic modulation combined with anti-VEGF (e.g., HDAC inhibitor plus sintilimab and bevacizumab). These approaches, informed by genomic and immune profiling, seek to improve outcomes in difficult-to-treat gastrointestinal malignancies.

Novel Frontiers: CAR-T, Vaccines, and Multi-Omics Integration

What novel immunotherapeutic approaches are being developed for gastrointestinal cancers, and how will genomics be integrated?

CAR-T cell therapies targeting antigens like Claudin18.2 (e.g., CT041) and EphA2 have shown objective response rates up to 48.6% in later-line gastrointestinal cancers. Personalized neoantigen vaccines, such as the mRNA-based autogene cevumeran in pancreatic ductal adenocarcinoma, induce long-lived CD8+ T cells with an average lifespan of 7.7 years and prolong recurrence-free survival. Integration of genomics is central: multi-omics profiling (genomics, transcriptomics, metabolomics) guides antigen selection, patient stratification, and combination design. Artificial intelligence builds predictive models (e.g., for tumor microenvironment subtypes), while dynamic monitoring via circulating tumor DNA and liquid biopsies enables real-time treatment adaptation. These technologies aim to rationally combine checkpoint inhibitors, CAR-T cells, vaccines, and oncolytic viruses based on each tumor's unique genomic and immune landscape.

The Path Forward

Prospective trials validating multi-modal strategies are essential. Combining genomic biomarkers with immune profiling and AI will allow precision immunotherapy that overcomes resistance and expands benefit to more patients with gastrointestinal cancers.

A Personalized Future Rooted in Hope and Science

The trajectory of immunotherapy in gastrointestinal (GI) cancers is accelerating from broad, one-size-fits-all approaches toward precise, genomics-driven strategies. This shift, guided by biomarkers such as MSI-H and emerging signatures like GIPS, promises to tailor treatment to each tumor’s unique biology and immune microenvironment.

Realizing the promise requires sustained effort

Continued research and rigorous clinical trials remain essential to overcoming resistance and extending durable benefits to the majority of patients who currently do not respond. Future exploration of multi-omics integration, dynamic monitoring via liquid biopsies, and novel combinations—such as CAR-T cell and neoantigen vaccines—will be critical in translating scientific insights into durable clinical outcomes.