Rectal cancer remains a common gastrointestinal malignancy, with approximately 30%–40% of patients presenting with locally advanced disease (1). Standard management for locally advanced rectal cancer includes neoadjuvant chemoradiotherapy (nCRT) followed by total mesorectal excision (TME); more recently, total neoadjuvant therapy (TNT) has been adopted to increase pathological complete response (pCR) rates and improve systemic disease control (2, 3). Despite these advances, pCR after nCRT or TNT in mismatch repair–proficient (pMMR) non metastatic rectal cancer generally remains below 20%, and clinical complete response (cCR) rates are typically 15%–30%, leaving substantial unmet need for strategies that increase organ preservation and long-term disease control (4, 5).

Programmed cell death protein 1 (PD−1) inhibitors produce durable antitumor responses in several solid tumors and show marked activity in colorectal cancers with deficient mismatch repair (dMMR) or high microsatellite instability (MSI−H) (6). Those phenotypes, however, represent a minority of colorectal cancers overall and an even smaller fraction of locally advanced rectal cancers, while the majority of tumors are pMMR or microsatellite stable (MSS) and derive limited benefit from PD−1 monotherapy, likely because of low baseline CD8+ T−cell infiltration and an immunologically “cold” tumor microenvironment (7). Preclinical and early clinical data indicate that radiotherapy and certain chemotherapies can increase tumor immunogenicity by promoting antigen release, dendritic cell activation, and T−cell recruitment, providing a biological rationale for combining PD−1 blockade with nCRT to convert non−inflamed tumors into responders (8).

Several early−phase clinical trials have tested PD−1 inhibitors added to standard nCRT in pMMR non−metastatic rectal cancer with promising signals for increased tumor response, but results have been heterogeneous and no pooled quantitative synthesis of randomized data is available (9–11). We therefore performed a systematic review and meta−analysis of phase II and III randomized controlled trials to evaluate whether adding PD−1 inhibitors to standard nCRT improves pCR and cCR rates without substantially increasing high−grade toxicity or surgical morbidity, and to explore whether radiotherapy strategy (short−course versus long−course) modifies the treatment effect.

All analyses were performed in R (version 4.4.1) using the readxl, dplyr, tidyr, meta, and grid packages. Outcome data were imported from an Excel workbook and reshaped to wide format with pivot_wider (tidyr). Dichotomous endpoints (pCR, cCR, R0 resection, sphincter preservation, neoadjuvant treatment-related adverse events, or surgery-related adverse events) were pooled as RRs with 95% confidence intervals (CIs) using the metabin function (meta package) under an inverse-variance–weighted DerSimonian–Laird random-effects model. For outcomes with rare or zero events, a continuity correction of 0.5 was applied where appropriate to enable RR estimation. Sensitivity analyses were additionally performed using the Hartung–Knapp adjustment and REML estimator to assess robustness in the context of small-study meta-analysis. Between-study heterogeneity was quantified by the I² statistic. Forest plots, including 95% CIs and prediction intervals, were generated via the forest function and exported to PDF. Subgroup analyses were prespecified for radiotherapy modality (short course versus long course radiotherapy). All tests were two-sided, with p < 0.05 denoting statistical significance.

This meta-analysis pooled data from six randomized controlled trials and, for the first time in a quantitative synthesis, demonstrated that adding PD-1 inhibitors to nCRT significantly increases pCR rates in pMMR non metastatic rectal cancer. A greater pCR benefit was observed in the short-course radiotherapy subgroup; however, given the limited number of trials, this finding should be considered hypothesis-generating and warrants further investigation in future prospective studies. The pCR improvement was not accompanied by a significant increase in grade ≥3 toxicity or surgical complications, suggesting that the combined approach has an acceptable tolerability and safety profile. However, adverse event reporting was highly heterogeneous across studies, with differences in grading criteria, attribution, and reporting timepoints. Many trials did not distinguish between immune-related and chemoradiotherapy-related toxicities, and some reported adverse events across the entire treatment course without separating neoadjuvant and adjuvant phases. Due to these inconsistencies, we did not perform pooled analyses of specific adverse event types. This limitation highlights the need for future trials to adopt standardized definitions and timepoint-specific AE reporting to enable more meaningful safety comparisons.

Radiotherapy can induce immunogenic cell death within tumors, releasing tumor antigens and promoting dendritic cell activation, thereby increasing the pool of effector T cell targets for PD-1 blockade. Short-course, high fraction radiotherapy may provoke a stronger local immune activation and a more intense inflammatory response in the tumor microenvironment, which could explain the greater pCR improvement compared with long-course CRT (18).

Clinically, pCR is an important predictor of postoperative survival and organ preservation. The observed increase in pCR in our analysis may expand eligibility for watch and wait strategies and sphincter preserving procedures, although the impact on long-term oncologic and functional outcomes remains to be established. Interestingly, while the addition of PD-1 inhibitors significantly improved pCR rates, the increase in cCR did not reach statistical significance. This discrepancy may reflect differences in the sensitivity and specificity of clinical versus pathological response assessments. Clinical complete response is typically determined by a combination of endoscopy, digital rectal examination, and imaging, which may underestimate residual microscopic disease or be influenced by interobserver variability. In contrast, pCR is based on histopathologic examination of the resected specimen and provides a more definitive assessment of tumor eradication. These findings underscore the need for standardized and validated criteria for cCR assessment, particularly in the context of organ-preserving strategies such as watch-and-wait. Future studies should explore whether improved pCR with immunoradiotherapy translates into more reliable cCR detection and durable non-operative management.

Although the included trials used different PD-1 agents and radiotherapy fractionation schemes, the consistent efficacy signal supports the generalizability of a radiotherapy–immunotherapy synergistic effect. Radiotherapy dose, fractionation, and sequencing with chemotherapy or PD-1 inhibitors varied across studies. LCRT was generally delivered as 45–50.4 Gy in 25–28 fractions with concurrent capecitabine, while SCRT was delivered as 25 Gy in 5 fractions followed by systemic therapy. These differences may influence the degree of immune activation and contribute to clinical heterogeneity. Moreover, several included regimens incorporated multi-cycle chemotherapy before or after radiotherapy and may be considered TNT, which could further contribute to clinical heterogeneity. Given the small number of included trials, further stratified analyses by chemotherapy regimen, PD-1 agent, or sequencing were not feasible. However, these factors likely contribute to the observed heterogeneity and should be addressed in future trials with harmonized protocols. Future work should integrate biomarkers such as tumor mutational burden, PD L1 expression, and measures of immune infiltration to more precisely identify pMMR patients most likely to benefit from combination therapy (19).

Further studies are needed to define optimal sequencing and dosing—whether concurrent with radiotherapy, as post radiotherapy maintenance, or as consolidation after chemoradiotherapy—and should prioritize long term endpoints including disease free survival, overall survival, and functional organ preservation. Multi omics and single cell sequencing approaches will be valuable for elucidating the dynamic effects of combined radiotherapy and immunotherapy on the tumor microenvironment (20).

The main strength of this analysis is its adherence to PRISMA guidelines and the Cochrane assessment framework, systematically synthesizing available multicenter randomized trial data. However, the predominance of early-phase (phase II) trials with limited sample sizes and exploratory designs introduces phase-related heterogeneity, which may affect the maturity and generalizability of the findings. Moreover, most studies lacked long-term oncologic endpoints such as disease-free and overall survival, as well as functional outcomes like sphincter preservation, limiting the ability to assess the durability and clinical relevance of the observed pCR improvements. These limitations underscore the need for confirmatory phase III trials with standardized treatment protocols, longer follow-up, and comprehensive outcome reporting to validate the benefits of immunoradiotherapy in pMMR rectal cancer.

Adding PD-1 inhibitors to nCRT appears to safely increase pCR rates in pMMR non metastatic rectal cancer, with the benefit most pronounced in the short course radiotherapy setting. These results provide a rationale for designing and conducting phase III trials and support the further clinical investigation of immunoradiotherapy combinations in the neoadjuvant treatment of rectal cancer, with particular emphasis on long-term survival, functional outcomes, and quality of life.