KEY POINTS
- This narrative review evaluated 63 publications on radiotherapy combined with immune checkpoint inhibitors in localized, locally advanced, or potentially resectable gastrointestinal malignancies. A structured search of PubMed/MEDLINE, Embase, and Web of Science through September 2025 identified 6,289 records, with 134 full-text articles assessed and 63 publications included: 25 addressing esophageal, gastric, or gastroesophageal junction cancers, 15 hepatobiliary or pancreatic cancers, and 23 rectal or anal cancers. No formal risk-of-bias assessment or quantitative meta-analysis was performed.
- Radiotherapy provides several mechanisms for enhancing antitumor immunity, including immunogenic cell death, tumor antigen release, increased MHC-I expression, and activation of the cGAS–STING/type I interferon pathway. However, radiation can also promote immunosuppression through PD-L1 upregulation, regulatory T-cell enrichment, hypoxia, and lymphocyte depletion. Preclinical evidence suggests that single doses above approximately 12–18 Gy may induce TREX1-mediated degradation of cytosolic DNA, whereas fractionated regimens such as 8 Gy × 3 or 6 Gy × 5 may better preserve immune stimulation, although no optimal clinical RT–ICI fractionation has been established.
- The most mature esophageal cancer evidence supports adjuvant nivolumab after chemoradiotherapy and surgery, rather than proving enhanced efficacy from concurrent RT–ICI. In the randomized phase III CheckMate-577 trial, patients with residual disease after neoadjuvant chemoradiotherapy and R0 resection achieved median disease-free survival of 22.4 months with nivolumab versus 11.0 months with placebo (HR 0.69; p<0.001). In contrast, studies adding PD-1/PD-L1 inhibitors directly to neoadjuvant chemoradiotherapy reported pathological complete response rates of approximately 25–55%, but most were single-arm or early-phase studies without mature comparative survival outcomes.
- Rectal cancer provides randomized evidence of improved pathological response with selected immunotherapy-containing neoadjuvant regimens, although results are not uniformly positive. In the phase III UNION trial, short-course RT (25 Gy in five fractions) followed by camrelizumab and chemotherapy achieved 39.8% pathological complete response versus 15.3% with conventional long-course chemoradiotherapy and chemotherapy (OR 3.7; p<0.001). However, the experimental arm differed in both RT schedule and systemic treatment, while NRG-GI002 found no significant pCR improvement from adding pembrolizumab (31.9% vs 29.4%; p=0.75) and did not improve its primary neoadjuvant rectal score.
- Hepatobiliary studies demonstrated encouraging responses but offered limited controlled evidence. In a multicenter phase II study of hepatocellular carcinoma with portal vein tumor thrombus, 30–50 Gy in ten fractions combined with sintilimab and bevacizumab produced an objective response rate of 58.7%, median progression-free survival of 13.8 months, and median overall survival of 24.0 months. In 36 patients with unresectable intrahepatic cholangiocarcinoma, radiotherapy followed by camrelizumab achieved 61.1% objective response, median PFS of 12.0 months, and median OS of 22.0 months. Neither study established the independent benefit of adding RT to immunotherapy-based treatment.
- Pancreatic cancer evidence remains particularly heterogeneous and predominantly early-phase. Neoadjuvant RT–ICI regimens reported pathological complete response rates up to 23.1%, R0 resection rates of approximately 82–88.5%, and median overall survival of 27.8–30.5 months in selected studies. In nonresected localized disease, objective response rates ranged from 0–28.1%, median PFS from 8.1–12.0 months, and median OS from 12.9–22.3 months. Different chemotherapy regimens, radiation prescriptions, surgical eligibility, and immunotherapy schedules prevent reliable cross-study comparisons.
- Evidence maturity and treatment-related lymphopenia remain central limitations. Of the 63 included publications, only two reported phase III randomized trials, alongside eight randomized phase II trials, while most other evidence came from nonrandomized prospective or retrospective studies. Neither phase III trial independently randomized the contribution of radiotherapy, and no biomarker has been prospectively validated to predict RT–ICI-specific benefit across gastrointestinal cancers. The review emphasizes minimizing unnecessary low-dose exposure to circulating lymphocytes and lymphoid tissues, standardizing RT–ICI sequencing, and incorporating long-term survival, toxicity, functional outcomes, and patient-reported quality of life into future trials.
CLINICAL TAKEAWAY
Combining radiotherapy with immune checkpoint inhibition has a strong biological rationale and has improved pathological response in selected gastrointestinal cancer trials, particularly in rectal cancer, while adjuvant immunotherapy after chemoradiotherapy and surgery is established in specific esophageal cancer settings. However, the independent contribution of RT to immunotherapy efficacy, optimal sequencing and fractionation, and consistent survival benefits remain unproven across gastrointestinal malignancies, making further randomized evaluation essential.