Radiotherapy-free neoadjuvant strategy using sintilimab plus XELOX in locally advanced rectal cancer: A single-arm phase II trial

Abstract

BACKGROUND

Neoadjuvant chemoradiotherapy is the standard treatment for locally advanced rectal cancer (LARC) and effectively reduces local recurrence; however, its survival benefit remains limited and radiotherapy is associated with substantial acute and long-term toxicities. Increasing evidence suggests that not all patients benefit equally from radiotherapy, particularly those without high-risk features such as cT4b disease, involved mesorectal fascia, or lateral lymph node metastasis. Meanwhile, immunotherapy alone shows limited efficacy in mismatch repair-proficient tumors. We hypothesized that combining chemotherapy with immunotherapy could provide a feasible radiotherapy-free neoadjuvant strategy for selected patients.

AIM

To evaluate the feasibility, efficacy, and safety of radiotherapy-free neoadjuvant sintilimab plus chemotherapy in resectable LARC.

METHODS

This multicenter, single-arm phase II trial enrolled patients with resectable LARC. Participants received 2-4 cycles of neoadjuvant sintilimab combined with XELOX, followed by total mesorectal excision. Patients with radiotherapy-mandating high-risk features were not the primary target population. Pathological response and safety were assessed. Multivariate logistic regression was used to explore predictors of major pathological response (MPR), and multiplex immunofluorescence was performed to evaluate tumor immune microenvironmental features.

RESULTS

A total of 41 patients underwent surgery, and all achieved R0 resection. The pathological complete response was observed in 11 patients (26.8%), and MPR was in 19 patients (46.3%). The objective response rates was 85.4%. T-stage and N-stage downstaging occurred in 41.5% of patients, with concurrent TN downstaging in 26.8%. Treatment-related adverse events were predominantly grade 1-2, with one case each of grade 3 immune-related rash and colitis. Multivariate analysis identified tumor size > 4 cm (odds ratio = 6.51, 95% confidence interval: 1.23-45.2; P = 0.036) and elevated baseline carcinoembryonic antigen (odds ratio = 0.12, 95% confidence interval: 0.01-0.73; P = 0.035) as independent negative predictors of MPR. Immunofluorescence analysis showed increased stromal CD8+ T-cell infiltration and reduced M2 macrophage polarization in responders.

CONCLUSION

Radiotherapy-free neoadjuvant sintilimab plus XELOX shows promising pathological responses and manageable toxicity in selected patients with LARC.

Key Words: Locally advanced rectal cancer; Immunotherapy; Sintilimab; Radiotherapy-free; Pathological complete response

Core Tip: This phase II single-arm study evaluated a radiotherapy-free neoadjuvant strategy using sintilimab plus XELOX in patients with locally advanced rectal cancer without mesorectal fascia involvement. The regimen achieved favorable pathological responses with acceptable toxicity. Tumor size > 4 cm and elevated baseline carcinoembryonic antigen levels were associated with a lower likelihood of major pathological response. Exploratory immune microenvironment analysis suggested a potential association between CD8+ T-cell infiltration and treatment response. These findings support further investigation of radiotherapy-sparing strategies in carefully selected patients.

INTRODUCTION

Rectal cancer is among the most common gastrointestinal tract malignancies worldwide. According to the 2022 China–United States cancer statistics, the incidence and mortality of colorectal cancer in China are projected to rank second and fifth places, respectively, closely mirroring the disease burden in developed countries[1]. Despite advances in multidisciplinary management, the treatment of locally advanced rectal cancer (LARC) remains challenging.

Currently, the standard treatment for LARC consists of neoadjuvant chemoradiotherapy (nCRT) followed by total mesorectal excision and adjuvant therapy[2]. Although this approach significantly reduces local recurrence rates, its impact on progression-free survival and overall survival (OS) remains limited[3]. Moreover, controversy persists regarding the necessity of radiotherapy and the optimal patient population. In high-risk patients, such as those with mesorectal fascia-positive (MRF+) or cT4b disease, nCRT effectively reduces tumor burden and improves surgical resectability[4]. However, in low-risk patients [(MRF-) or cT1-3N1], conventional radiotherapy may induce acute toxicities, such as radiation enteritis and pelvic fibrosis, and is associated with long-term complications, including anastomotic leakage, sexual dysfunction, and urinary disorders[5]. Increased tissue fragility, fibrosis, and adhesions resulting from radiotherapy alter pelvic anatomy and significantly increase surgical difficulty[6,7]. Multiple reports indicate that approximately 40%-50% of patients are unable to tolerate the adverse effects of postoperative adjuvant radiotherapy and therefore fail to complete the planned treatment, mainly due to treatment-related toxicities and deterioration in quality of life[8,9]. Given its substantial toxicity and limited survival benefit, the role of radiotherapy in this subset of patients has been increasingly questioned. For low-risk patients, trials such as the FOWARC study have suggested that neoadjuvant chemotherapy (nCT) alone may be feasible[10]; however, its efficacy requires further validation.

In recent years, immunotherapy has made breakthrough progress in the neoadjuvant treatment of solid tumors. Programmed cell death protein 1 (PD-1) immune checkpoint inhibitors have been shown to significantly improve prognosis in patients with microsatellite instability-high (MSI-H) or mismatch repair-deficient (dMMR) colorectal cancer and were approved by the Food and Drug Administration for treating such metastatic cases in 2017[11,12]. However, the dMMR/MSI-H subtype accounts for only 10%-15% of LARC cases, and the majority of tumors are mismatch repair-proficient (pMMR) and microsatellite stable (MSS), with an immunotherapy monotherapy response rate of < 5%[13]. To address this challenge, combining chemotherapy with immunotherapy has emerged as a promising research focus. Preclinical studies have demonstrated that cytotoxic agents such as oxaliplatin can induce immunogenic cell death, releasing tumor-associated antigens that activate antigen presentation and recruit CD8+ T-cell infiltration[14], while simultaneously reversing the immunosuppressive microenvironment[15]. Additionally, chemotherapy can reduce tumor-associated M2 macrophages, enhance immune activation, and overcome resistance to immunotherapy[16]. Clinically, the combination of chemotherapy with immune checkpoint inhibitors has yielded favorable outcomes across multiple malignancies, including non-small cell lung cancer (KEYNOTE-189), triple-negative breast cancer (IMpassion130), and gastric cancer (CheckMate-649), with improvements in progression-free survival and objective response rates (ORR)[16-18].

Based on this background, we designed a multicenter phase II clinical trial to evaluate the pathological response rate and safety of sintilimab (a PD-1 inhibitor) combined with the XELOX regimen (oxaliplatin + capecitabine) as a radiotherapy-sparing neoadjuvant treatment for LARC. This study targeted patients with LARC and aimed to provide a novel therapeutic option for those who cannot tolerate radiotherapy or who seek organ preservation.

MATERIALS AND METHODS

Study design and participants

This clinical trial was conducted at the First and Second Affiliated Hospitals of Anhui Medical University (ClinicalTrials.gov identifier: NCT07127497). This was a multicenter, phase II, single-arm study designed to evaluate the efficacy and safety of sintilimab (a PD-1 inhibitor) combined with the XELOX regimen (oxaliplatin + capecitabine) as neoadjuvant therapy for LARC. The trial was performed in strict accordance with the International Council for Harmonisation of Good Clinical Practice, the Declaration of Helsinki, and applicable local laws and regulations. All patients were informed of the nature of the study and provided written informed consent before enrollment.

Patient population

Eligible participants were adults (≥ 18 years) with histologically confirmed rectal adenocarcinoma, staged as cT3-4aN0M0 or cT1-4aN1-2M0, and deemed resectable by colorectal surgeons prior to neoadjuvant treatment. Patients undergoing conversion therapy for initially unresectable or borderline resectable disease were excluded to maintain a relatively homogeneous study population and to specifically evaluate the feasibility of a radiotherapy-free strategy in patients considered appropriate for upfront surgical resection. Detailed inclusion and exclusion criteria are provided in Supplementary Table 1.

Patient assessments

Radiotherapy-related anatomical risk factors, including T4b invasion, involved MRF+/circumferential resection margin positive (CRM+), lateral lymph nodes ≥ 7 mm, anal canal involvement, and sphincter involvement were systematically evaluated by baseline pelvic magnetic resonance imaging (MRI). For selected patients with low rectal tumors (< 5 cm from the anal verge) or threatened CRM involvement, the decision to omit radiotherapy was made after a multidisciplinary team (MDT) discussion and thorough consideration by the patient and their family. Given the increased risk of anastomotic leakage and pelvic complications associated with preoperative radiotherapy in patients with low anterior resection rates, several patients chose a radiotherapy-free regimen after informed consent.

Study procedures

At baseline, all patients underwent a complete medical history review, physical examination, colonoscopy, pelvic MRI, and chest/abdominal/pelvic computed tomography (CT) scans. All patients received sintilimab (3 mg/kg IV every 3 weeks for patients < 60 kg, or 200 mg IV every 3 weeks for patients ≥ 60 kg) combined with the XELOX regimen (oxaliplatin 130 mg/m² IV on day 1 plus capecitabine 1000 mg/m² orally twice daily on days 1-14 of a 21-day cycle). Treatment was administered every 3 weeks for two to four cycles. Within 3-4 weeks of the last dose, patients underwent preoperative imaging evaluation (Figure 1). If patients experienced treatment-related adverse events of grade ≥ 3, dose reductions in chemotherapy were permitted, and treatment interruption could be applied if necessary. Although the protocol permitted 2-4 cycles of neoadjuvant treatment, the actual number was individualized based on the tumor regression rate, toxicity, surgical scheduling, and patient preference.

Figure 1 The treatment procedures for patients with locally advanced rectal cancer. cCR: Clinical complete remission; PR: Partial response rate; SD: Disease stability rate; TME: Total mesorectal excision; LARC: Locally advanced rectal cancer.

Surgical resection was performed according to total mesorectal excision principles. For mid/Low rectal tumors with a distal margin > 5 cm, low anterior resection was performed; for ultra-low tumors, abdominoperineal resection was performed by the attending surgeon. All surgeons involved in this study had substantial experience (at least 30 open and 30 laparoscopic rectal resections) and held formal robotic surgery qualifications.

Assessments and outcome measures

Baseline assessments included demographic data, medical history, and disease characteristics. Patients underwent a systematic physical examination, laboratory tests (including blood counts and biochemistry), and imaging (chest CT, liver MRI, pelvic CT, or MRI) before and after treatment. Safety was monitored via treatment-related adverse events and graded according to NCI CTCAE version 5.0 (grades 1-5). Surgical specimens were evaluated by the Pathology Department of the First Affiliated Hospital of Anhui Medical University. Tumor regression grade (TRG) was assessed using the Ryan criteria[19], TRG 0 [complete response (CR): No viable cancer cells, only fibrosis or acellular mucin], TRG 1 (near-CR: Single cells or small clusters of residual cancer), TRG 2 [partial response (PR): Residual cancer with evident regression but more than TRG 1], and TRG 3 (poor response: Minimal regression with extensive residual cancer). Tumor response was evaluated according to RECIST 1.1 criteria[20], as CR, PR, stable disease, or progressive disease. ORR was defined as the sum of CR and PR, and the disease control rate was defined as the sum of CR, PR, and stable disease. Clinical TNM and pathological staging were determined per the American Joint Committee on Cancer 8^th edition[21]. Clinical CR (cCR) was defined as the absence of any residual tumor on digital rectal examination, no visible tumor or ulceration on colonoscopy (the presence of a flat white scar or telangiectasia was allowed), and no evidence of a residual tumor or suspicious lymph nodes on pelvic MRI. Normalization of serum carcinoembryonic antigen (CEA) levels was also considered supportive evidence[22]. Successful downstaging was defined as a reduction in the clinical TNM stage from stage III to stage II or I, following neoadjuvant sintilimab + XELOX. According to the patient’s neoadjuvant rectal (NAR) score, the patient was classified into high-, medium-, and low-risk categories: < 8, low risk, 8-16 is medium risk; and > 16, high risk[23].

After surgery, patients were followed up every 3 months for 2 years by telephone or outpatient visits under a standardized oncology follow-up protocol. Follow-up was performed every 6 months until 5 years postoperatively or until recurrence or death. Follow-up assessments included clinical evaluation, routine laboratory tests, serum tumor marker measurements (including CEA and carbohydrate antigen 19-9), as well as imaging and endoscopic examinations, in accordance with institutional guidelines. Patients with MDT-confirmed cCR were offered a watch-and-wait strategy after detailed counselling and written informed consent. Patients who declined the watch-and-wait approach proceeded directly to total mesorectal excision. During the first 2 years, surveillance was performed every 3 months with MRI and every 3-6 months with endoscopy according to institutional guidelines.

Study endpoints

The primary endpoint was the pathological CR (pCR) rate. Secondary endpoints included the major pathological response (MPR) rate, ORR, downstaging rate, R0 resection rate, and safety. Exploratory analyses were conducted to assess correlations between tumor response and features of tumor immune microenvironment or pretreatment biomarkers.

Sample size and statistical analysis

We assumed a historical pCR rate of approximately 10% with standard nCRT and targeted a pCR rate of 25% with the current regimen[24]. We calculated that 41 patients would provide 80% power at a two-sided 95% confidence level (accounting for 10% dropouts). Univariate logistic regression analyses were performed to identify the potential predictors of MPR. Variables included in the multivariate models were selected based on a combination of clinical relevance, biological plausibility, and statistical significance in univariate analysis. To avoid model overfitting given the limited sample size, only variables with established clinical importance and a univariate P value < 0.10 were considered for entry into the multivariate analysis[25]. Odds ratios (ORs) with 95% confidence intervals (CIs) were reported. Prespecified subgroup analyses were conducted according to baseline clinical factors, including tumor diameter (≤ 4 cm vs > 4 cm), pretreatment CEA level (≤ 5 ng/mL vs > 5 ng/mL), and threatened CRM (positive vs negative). The interaction terms between tumor size and CEA levels were explored using multivariate models to assess potential effect modifications.

Variables with missing values (< 5%) were excluded from the corresponding analyses and no imputation was performed. All variables in this study were complete. No patient was lost to follow-up during the neoadjuvant treatment or postoperative follow-up period. Due to the limited sample size of this phase II trial, no formal sensitivity analysis was performed. All statistical analyses were conducted using SPSS v15. A two-sided P < 0.05 was considered statistically significant. Continuous data are reported as mean ± SD or median (range).

Biomarker analysis and multiplex immunofluorescence

Tumor tissue specimens resected during surgery were collected for biomarker analysis. Samples were fixed in 10% neutral-buffered formalin, routinely paraffin-embedded, and sectioned at a thickness of 4-5 μm. After deparaffinization with xylene and rehydration with graded ethanol and distilled water, antigen retrieval and blocking were performed.

The slides were incubated with a blocking solution at room temperature in the dark, rinsed with phosphate buffered saline, and divided into two panels for multiplex immunohistochemical staining: Panel 1 consisted of PD-1, programmed death ligand 1 (PD-L1), and CD8; panel 2 consisted of CD68, CD163, and CD8. Primary antibodies against PD-L1 (ab205921, Abcam, MA, United States), CD68 (ab955, Abcam, MA, United States), CD163 (ab182422, Abcam, MA, United States), CD8 (ab237709, Abcam, MA, United States), and PD-1 (MAB-0734, MAB-BIO, Guangzhou, China) were used. Sections were incubated overnight at 4 °C. The following day, after washing, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies at room temperature. Nuclear counterstaining was performed with DAPI solution (Beyotime, C1005, Shanghai, China) for 5 minutes, followed by a 10-minute rinse in running water before mounting. Stained slides were scanned and digitized using a 3DHISTECH Pannoramic SCAN digital slide scanner.

The proportion of positively stained cells for each marker was quantified in high-power fields. M2 macrophage polarization was evaluated by dual staining for CD68 (a pan-macrophage marker) and CD163 (an M2 subtype marker). The percentage of CD163+ cells among CD68+ cells was calculated to assess the degree of M2 polarization.

RESULTS

Patient characteristics

From August to December 2025, 57 patients with LARC were screened for inclusion (Figure 2). Sixteen patients did not complete the planned treatment: Five received more than the prescribed cycles, three withdrew consent for personal reasons, three were found to have distant metastases on subsequent evaluation and received palliative care, and four chose alternative treatment strategies. Ultimately, 42 patients completed the full neoadjuvant therapy, of whom 1 achieved clinical cCR by endoscopic and MRI assessment after treatment but refused surgery. Thus, 41 patients who underwent surgery were included in the final study cohort. None of the patients received radiation therapy, and patients with a cCR were closely followed up.

Figure 2 Consort diagram. PCR: Pathological complete response rate.

The baseline characteristics of the 41 patients are summarized in Table 1. All baseline clinical and pathological variables were complete, with no missing data. The median age was 61 years, and 73.2% (30/41) of patients were male. The mean pretreatment body mass index (BMI) was 22.99 ± 3.05 kg/m². The median tumor distance from the anal verge was 6.5 cm (range, 2.0-12.0 cm), and the mean tumor diameter was 4.85 cm (range, 2.7-8.3 cm). Before treatment, 39 patients (95.1%) had stage III disease. MRI showed threatened CRM in eight patients (19.5%) and positive extramural vascular invasion in 10 patients (24.4%). Pretreatment T-stage distribution was T2 in 2 patients (4.9%), T3 in 26 (63.4%), and T4 in 13 (31.7%). Nodal stage was: N0 in 2 (4.9%), N1 in 14 (34.1%), and N2 in 25 (61.0%). Regarding treatment cycles, 20 patients (48.8%) completed 2 cycles, 17 patients (41.5%) completed 3 cycles, and 4 patients (9.7%) completed 4 cycles. None of the patients had dMMR/MSI-H disease, and all 41 patients (100%) had pMMR/MSS.

Table 1 Baseline characteristics, mean ± SD/n (%).

Characteristic	All patients (n = 41)
Age (years)	61.29 ± 9.48
Gender
Male	30 (73.2)
Female	11 (26.8)
Before treatment BMI (kg/m2)	22.99 ± 3.05
Hypertension	14 (34.1)
Diabetes	5 (12.2)
Smoking	15 (36.6)
Drinking	9 (22.0)
Before treatment clinical T stage
cT2	2 (4.9)
cT3	26 (63.4)
cT4a	13 (31.7)
cT4b	0 (0)
Before treatment clinical N stage
cN0	2 (4.9)
cN1	14 (34.1)
cN2	25 (61.0)
Lateral lymph node ≥ 7 mm	0 (0)
Clinical disease stage before treatment
Stage II (cT3-4N0)	2 (4.9)
Stage III (cT1-4N1-2)	39 (95.1)
CRM
Threatened	8 (19.5)
Negative	33 (80.5)
EMVI
Positive	10 (24.4)
Negative	31 (75.6)
Maximum length of tumor involvement (cm)	4.82 ± 1.55
Less than 4	18 (43.9)
Not less than 4	23 (56.1)
Distance from anal verge (cm)	6.71 ± 3.13
Less than 5	17 (41.5)
Not less than 5	24 (58.5)
Anal canal involvement	0 (0)
Sphincter involvement	0 (0)
Initial CEA level (ng/mL)	7.59 ± 14.68
Less than 5	26 (63.4)
Not less than 5	15 (36.6)
Neutrophil/Lymphocyte ratio	2.57 ± 1.17
Less than 2.2	20 (48.8)
Not less than 2.2	21 (51.2)
Number of neoadjuvant treatments
2	20 (48.8)
3	17 (41.5)
4	4 (9.7)
Patients’ mismatch repair status
dMMR/MSI-H	0 (0)
pMMR/MSS	41 (100)

BMI: Body mass index; CRM: Circumferential resection margin; EMVI: Extramural vascular invasion; CEA: Carcinoembryonic antigen; MSI-H: Microsatellite instability-high; dMMR: Mismatch repair-deficient; pMMR: Mismatch repair-proficient; MSS: Microsatellite stable.

The primary reasons for completing only 2-3 cycles of neoadjuvant treatment were as follows: Early tumor regression prompting MDT recommendation for earlier surgery (15/37, 40.5%), patient preference for earlier surgical resection (10/37, 27.0%), mild but persistent adverse events such as diarrhea or leukopenia (8/37, 21.6%), and scheduling considerations to avoid delaying total mesorectal excision (4/37, 10.8%). Importantly, none of the patients discontinued the treatment due to severe toxicity.

Treatment efficacy

All 41 patients completed neoadjuvant therapy. One patient achieved a cCR on colonoscopy/MRI and declined surgery, thus requiring close surveillance. All 41 patients who underwent surgery achieved R0 resection (100% R0 rate). Surgical procedures included miles procedure (abdominoperineal resection) in 7 patients (17.1%) and Dixon procedure (low anterior resection) in 31 patients (75.6%); among the Dixon cases, 25 patients (86.2%) had a temporary ileostomy fashioned. According to the NAR score, 22 patients (53.7%) were classified as low risk (NAR ≤ 8), 9 patients (22.0%) as high risk (NAR > 16), and the remainder as intermediate risk. Pathological examination revealed that 11 patients (26.8%) achieved ypT0, which constituted a pCR. By TRG, 11 patients (26.8%) were classified as TRG 0, 8 (19.5%) as TRG 1, 10 (24.4%) as TRG 2, and 12 (29.3%) as TRG 3, yielding a MPR (TRG 0-1) rate of 46.3%. All resection margins were tumor-free (R0) with no tumor involvement at the proximal or distal margins.

Radiographically, the ORR was 85.4%, with 4 patients (9.8%) achieving CR and 31 (75.6%) achieving PR; 6 patients (14.6%) had SD, and no patient had progressive disease. On MRI, 4 patients (9.8%) achieved a radiologic CR. Comparing the pre- and post-treatment imaging stages, 17 patients (41.5%) had a decrease in the T stage, 17 (41.5%) had a decrease in the N stage, and 11 (26.8%) had both T and N downstaging (Figure 3; Table 2).

Figure 3 Analysis of multifactorial clinical risk assessment factors. A: Changes of before treatment T stage to after treatment T stage in 41 patients who reached the surgery; B: Changes of before treatment N stage to after treatment N stage in 41 patients who reached the surgery; C: Rectal magnetic resonance imaging evaluation of pathological complete response rate patients before and after treatment; D: Rectal magnetic resonance imaging evaluation of non-pathological complete response rate patients before and after treatment. In C and D, the orange arrows indicate tumor location and size, with the left arrows representing pre-treatment findings and the right arrows representing post-treatment findings.

Table 2 Treatment outcomes and surgical parameters, n (%).

Characteristic	All patients (n = 41)
Clinical T stage after treatment
cT0	4 (9.8)
cT1	2 (4.9)
cT2	8 (19.5)
cT3	23 (56.0)
cT4a	4 (9.8)
Clinical N stage after treatment
cN0	11 (26.8)
cN1	17 (41.5)
cN2	13 (31.7)
NAR
Less than 8	22 (53.7)
8 to 16	10 (24.4)
Not less than 16	9 (21.9)
ORR	35 (85.4)
DCR	41 (100)
CR	4 (9.8)
PR	31 (75.6)
SD	6 (14.6)
PD	0 (0)
Reduced TMN stage
T stage reduction	17 (41.5)
N stage reduction	17 (41.5)
TN stage reduction	10 (24.4)
Pathological T stage
ypT0	11 (26.8)
ypTis	1 (2.4)
ypT1	7 (17.1)
ypT2	10 (24.4)
ypT3	10 (24.4)
ypT4	2 (4.9)
Pathological N stage
ypN0	30 (73.2)
ypN1	9 (21.9)
ypN2	2 (4.9)
PCR	11 (26.8)
TRG classification
TRG 0	11 (26.8)
TRG 1	8 (19.5)
TRG 2	10 (24.4)
TRG 3	12 (29.3)
MPR	19 (46.3)
Resection range
R0	41 (100)
R1	0 (0)

NAR: Neoadjuvant rectal; ORR: Objective response rate; DCR: Disease control rate; CR: Complete remission; PR: Partial response rate; SD: Disease stability rate; DP: Disease progression; PCR: Pathological complete response rate; MPR: Major pathological remission rate; TRG: Tumor regression grade.

Safety

During neoadjuvant therapy (Table 3), the most common haematological adverse event was leukopenia (7/41, 17.1%), followed by elevated transaminases (alanine aminotransferase and aspartate aminotransferase, 9/41, 22.0%), anemia (5/41, 12.2%), and elevated bilirubin (3/41, 7.3%). All hematologic adverse events were of grades 1-2, and no patient discontinued treatment or died due to toxicity. Among non-hematological events, diarrhea was the most common (9/41, 22.0%), followed by rash (2/41, 4.8%), decreased renal function (2/41, 4.9%), and hypothyroidism (2/41, 4.8%). One patient experienced a grade III immune-related rash that resolved with medication and did not affect the treatment course.

Table 3 Adverse events during neoadjuvant therapy, n (%).

Treatment-related adverse events	Grade I-II	≥ grade III	All grade
Hematology related
WBC reduction	7 (17.1)	0 (0)	7 (17.1)
Anemia	5 (12.2)	0 (0)	5 (12.2)
PLT reduction	0 (0)	0 (0)	0 (0)
ALT increase	9 (22.0)	0 (0)	9 (22.0)
AST increase	9 (22.0)	0 (0)	9 (22.0)
Elevated bilirubin	3 (7.3)	0 (0)	3 (7.3)
Renal failure	2 (4.9)	0 (0)	2 (4.8)
Total	21 (51.2)	0 (0)	21 (51.2)
Non hematology
Weak	1 (2.4)	0 (0)	1 (2.4)
Nausea	1 (2.4)	0 (0)	1 (2.4)
Diarrhea	9 (22.0)	0 (0)	9 (22.0)
Constipation	1 (2.4)	0 (0)	1 (2.4)
Rash	2 (4.8)	0 (0)	2 (4.8)
Abdominal pain	0 (0)	0 (0)	0 (0)
Hypothyroidism	1 (2.4)	1 (2.4)	2 (4.8)
Hyperthyroidism	0 (0)	0 (0)	0 (0)
Lumbago	0 (0)	0 (0)	0 (0)
Total	15 (36.6)	1 (2.4)	16 (39.0)

WBC: White blood cell; PLT: Platelets; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase.

The postoperative complications were generally manageable. The most common complication was transient abdominal pain (11/41, 26.8%). Anastomotic leakage occurred in 3 patients (7.3%); two of these had concurrent intra-abdominal infection and underwent reoperation during the same hospitalization. One patient developed a grade III rectovaginal fistula requiring a second surgery. Additionally, one patient had postoperative prolapse of the ileal anastomosis (> 5 cm) but remained asymptomatic with normal stoma function. The mean total length of hospital stay was 13.9 ± 6.8 days. To date, no treatment-related deaths have occurred during follow-up, and the overall treatment was well tolerated (Table 3; Supplementary Table 2).

Multiplex immunofluorescence

Multiplex immunofluorescence (MIF) analysis of surgically resected tissues and evaluation of immune cell marker proportions (Figure 4) showed that among patients receiving the same nCT combined with immunotherapy, non-pCR patients exhibited a significantly lower proportion of CD8+ T cell infiltration within the tumor stroma than pCR patients (P = 0.0208). Concurrently, the proportions of CD163+ and CD68+ cells, markers of M2 macrophages, were significantly higher in the stroma of non-pCR patients (P = 0.0138). These findings indicate that enhanced stromal CD8+ T cell infiltration and reduced density within the tumor immune microenvironment are closely associated with favorable therapeutic responses in patients with LARC treated with nCT combined with immunotherapy.

Figure 4 Multiplex immunofluorescence analysis of the tumor immune microenvironment and immune cell infiltration in patients receiving neoadjuvant sintilimab plus chemoradiotherapy. A-D: Representative multiplex immunofluorescence images of the tumor immune microenvironment at 20 × or 80 × magnification. Image showing the same pathological complete response rate (pCR) patient (A and C); Image showing the same non-pCR patient (B and D). The yellow boxed regions in the 20 × images (upper left) are further magnified at 80 × (lower right). A and B demonstrate programmed cell death protein 1 and CD8 co-localization in CD8+ T cells, while panels C and D demonstrate CD163 and CD8 co-localization in M2 macrophages; E-J: Immune cell infiltration levels in the tumor tissue after neoadjuvant sintilimab and chemoradiotherapy between patients achieving pCR (n = 4) and non-pCR (n = 7) assessed by multiplex immunofluorescence. PD-1: Programmed cell death protein 1; PD-L1: Programmed death ligand 1; pCR: Pathological complete response rate.

Assessment of clinical risk factors

Patients were stratified into MPR and non-MPR groups based on clinical high-risk factors. Univariate analysis demonstrated that maximum tumor diameter (OR = 9.38, 95%CI: 2.29-50.1, P = 0.0036) and pretreatment serum CEA level (OR = 0.183, 95%CI: 0.0352-0.740, P = 0.025) were significant independent factors influencing MPR status. BMI before treatment (OR = 3.54, 95%CI: 0.979-14.0, P = 0.060) and threatened CRM assessed by MRI (OR = 5.25, 95%CI: 1.03-39.9, P = 0.063) approached statistical significance, suggesting they may be potential influencing factors that require validation in larger cohorts. Other baseline characteristics, including age, sex, smoking history, and neutrophil-to-lymphocyte ratio before and after treatment, were not significantly associated with MPR status.

In multivariate logistic regression analysis, tumor size > 4 cm (OR = 6.51, 95%CI: 1.23-45.2, P = 0.036) and baseline CEA > 5 ng/mL (OR = 0.12, 95%CI: 0.01-0.73, P = 0.035) were identified as independent predictors of MPR. Higher BMI (> 24 kg/m²) and threatened CRM showed trends toward lower MPR rates but did not reach statistical significance (Table 4, Supplementary Table 3).

Table 4 Multivariable logistic regression analysis of clinical risk assessment factors.

Characteristic	P value	OR	95%CI
BMI before treatment (> 24 kg/m2)	0.06	6.72	1.04-66.3
CRM (threatened)	0.11	5.97	0.732-68.1
Tumor size (> 4 cm)	0.036	6.51	1.23-45.2
Before treatment CEA (> 5 ng/mL)	0.035	0.12	0.01-0.73

BMI: Body mass index; CRM: Circumferential resection margin; CEA: Carcinoembryonic antigen; OR: Odds ratio; CI: Confidence interval.

DISCUSSION

Previous phase II/III clinical trials explored the value of checkpoint inhibitors combined with nCRT in patients with dMMR/MSI-H or pMMR/MSS LARC[26,27]. This study was a prospective investigation focusing on a “radiotherapy-free” strategy, evaluating the efficacy of sintilimab combined with XELOX in resectable LARC. The trial demonstrated encouraging results, with a pCR rate of 26.8% (11/41) and MPR rate of 46.3% (19/41). One patient achieved a cCR but refused surgery and opted for close surveillance. During treatment, only one patient developed hypothyroidism requiring medication. Univariate and multivariate analyses indicated that maximum tumor diameter and pre-treatment serum CEA level were key clinical factors for patient prognosis. Whether additional cycles would further improve the pathological response remains uncertain and warrants further evaluation.

Importantly, most patients enrolled in our study did not exhibit high-risk features that mandate radiotherapy, such as T4b invasion, MRF(+), or lateral lymph node involvement. None of the patients had T4b disease and only 19.5% had threatened CRM, which was markedly lower than that in the high-risk LARC population traditionally treated with nCRT. Therefore, our cohort represents a relatively lower-risk subgroup in which radiotherapy omission may be oncologically acceptable.

In the CONVERT study, the efficacy of XELOX-based nCT was compared with that of nCRT in LARC patients with MRF-negative tumors, showing comparable outcomes with no significant differences in pCR rates (nCT, 11% vs nCRT, 13.8%) or MPR rates (nCT, 23.2% vs nCRT, 36.8%)[10]. In contrast, in the present study, combining sintilimab with the same chemotherapy regimen resulted in a higher pCR (26.8%) and MPR (46.3%) rates, with 24.4% of patients achieving simultaneous T and N downstaging. This improvement suggests that the addition of immune checkpoint inhibitors to low-risk LARC cells enhances tumor immunogenicity and accelerates tumor cell apoptosis, thereby further increasing the pathological response. Chemotherapy-induced immunogenic cell death releases tumor antigens into the microenvironment, enhancing CD8+ T cell infiltration[28,29], This phenomenon was also confirmed by MIF analysis.

Studies on tumor-infiltrating immune cells and the tumor microenvironment have shown that patients with pCR tend to have increased numbers of CD8+ T cells and decreased numbers of CD163+ tumor-associated macrophages[30]. This aligns with our findings that elevated stromal CD8+ T cells and reduced M2 macrophages correlate with better treatment outcomes in LARC, suggesting that remodeling of the immune microenvironment may play a crucial role in neoadjuvant immunotherapy. CD8+ T cells exert direct antitumor effects through perforin, granzyme, and interferon-γ secretion, and can activate macrophages toward an M1 phenotype via interferon-γ, further enhancing antitumor immunity[31,32]. In addition, the inhibition of the PD-1/PD-L1 pathway can restore T cell function while suppressing the immunosuppressive activity of CD163+ M2 macrophages, thereby improving the overall immune microenvironment[33]. These interactions may explain our observations and provide a rationale for future strategies that combine the induction of macrophage polarization with T cell activation. Despite the biological plausibility of these associations, the findings derived from MIF analysis should be interpreted with caution. Further validation is still needed in large-scale, independent cohorts.

Although several short-course radiotherapy-chemo-immunotherapy regimens have reported higher pCR rates (30%-50%)[34], these approaches inevitably involve pelvic radiotherapy, which increases bowel, urinary, and sexual dysfunction and is associated with greater hematologic toxicity[35]. In contrast, our radiotherapy-free regimen achieved a clinically meaningful pCR of 26.8% with almost no grade ≥ 3 hematologic adverse events. Xu et al[36] also noted that higher radiotherapy doses were associated with increased proctitis, myelosuppression, and gastrointestinal complications, indicating that radiotherapy carries a non-negligible cost. Moreover, Piffkó et al[37] described a “badscopal effect” in which non-irradiated metastatic tumors paradoxically grew after radiotherapy. By comparison, our radiotherapy-free strategy combining nCT with sintilimab achieved a pCR of 26.8% and an MPR of 46.3%, with all patients achieving R0 resection. Although the pCR rate is slightly lower than in NCT04304209, the safety profile was superior, with no grade ≥ 3 hematologic toxicity, manageable adverse events, no treatment discontinuations, and no treatment-related deaths. A meta-analysis of seven studies, including 2707 patients with LARC, comparing nCT and nCRT found no significant differences in OS or disease-free survival; however, nCT was associated with lower rates of infection, anastomotic leakage, and comparable postoperative pathological responses[8]. This suggests that a chemoimmunotherapy-only approach may serve as a practical alternative for patients seeking organ preservation, those unable to tolerate radiotherapy, or those wishing to avoid long-term pelvic toxicity.

Multivariate logistic regression analysis identified tumor diameter > 4 cm (OR = 6.51, 95%CI: 1.23-45.2, P = 0.036) as an independent negative predictor of MPR. Tumor burden may directly influence histopathological response, as larger primary tumor volumes are associated with more aggressive biology and reduced treatment sensitivity[38]. Elevated pretreatment CEA levels were also significantly associated with poorer TRG (OR = 0.12, 95%CI: 0.01-0.73, P = 0.035), suggesting its potential as a biomarker for tumor biology and treatment response[39]. Other factors reported in previous studies, such as age < 50 years, pre-treatment neutrophil-to-lymphocyte ratio > 2.2, or tumor encircling > 1/2 of the rectal circumference, were not significant predictors in our cohort, possibly due to small sample size and older age distribution. Although factors such as BMI and threatened CRM approached statistical significance in univariate analysis, they were not retained in the final multivariate model. This may be partly attributable to limited statistical power due to the small sample size, as well as potential collinearity with other tumor-related variables. Moreover, the effect of threatened CRM on pathological response may be less pronounced in this selected cohort with a low proportion of high-risk features. These findings warrant further investigation in larger cohorts to clarify their independent predictive value.

From a clinical standpoint, these findings suggest that baseline tumor burden and tumor biology play important roles in determining response to neoadjuvant chemoimmunotherapy[40]. Larger tumors (> 4 cm) may exhibit greater intratumoral heterogeneity and immunosuppressive microenvironments, which could reduce sensitivity to systemic treatment. An increase in tumor size or volume is associated with lower rates of pathological response after neoadjuvant therapy in LARC[41]. Similarly, elevated pretreatment CEA levels have been linked to aggressive tumor behavior and inferior response to neoadjuvant treatment, potentially reflecting an unfavorable biological phenotype[42,43]. Together, tumor size and baseline CEA may serve as practical indicators for risk stratification in the neoadjuvant setting. Patients with lower tumor burden and normal CEA levels may be more suitable candidates for radiotherapy-free strategies, whereas those with larger tumors or elevated CEA may require closer monitoring or intensified approaches. These findings remain exploratory and require prospective validation.

This study has several limitations. First, as a single-arm, non-randomized phase II trial, selection bias may have occurred because only patients who were eligible for intensive chemotherapy plus immunotherapy were included. The exclusion of high-risk features limits the generalizability of the findings to broader LARC populations requiring mandatory radiotherapy. Second, although baseline demographic and clinical characteristics were well documented, unmeasured confounders, such as molecular alterations or heterogeneity of the total mesorectal excision, may have influenced treatment response. Third, despite achieving the prespecified pCR target, the follow-up period was relatively short. As a result, long-term oncological outcomes, including disease-free survival, OS, and late treatment-related toxicities, could not be adequately assessed. Larger randomized studies are warranted to determine whether this radiotherapy-free strategy offers durable oncological control.

CONCLUSION

In summary, this study evaluated the efficacy and safety of a radiotherapy-free approach using sintilimab combined with XELOX as neoadjuvant therapy in patients with LARC. The regimen demonstrated encouraging efficacy, as reflected by pCR, MPR, and radiological downstaging, along with acceptable toxicity and manageable postoperative complications.

ACKNOWLEDGEMENTS

We thank the patient and multidisciplinary team for their support.