BPG is committed to discovery and dissemination of knowledge
Retrospective Study Open Access
Copyright: ©Author(s) 2026. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial (CC BY-NC 4.0) license. No commercial re-use. See permissions. Published by Baishideng Publishing Group Inc.
World J Gastrointest Oncol. Nov 15, 2026; 18(11): 124022
Published online Nov 15, 2026. doi: 10.4251/wjgo.124022
Propensity score-based real-world study of immune checkpoint inhibitor combination therapy in microsatellite stable/proficient mismatch repair metastatic colorectal cancer
Chang Zhang, Lei Jin, Liang Lu, Shui Jin, Department of Gastroenterology, The Fourth Affiliated Hospital of Anhui Medical University, Hefei 238000, Anhui Province, China
Chang Zhang, Shi-Hao Wu, Lei Jin, Anhui Medical University, Hefei 230000, Anhui Province, China
Shi-Hao Wu, Department of Burns, The First Affiliated Hospital of Anhui Medical University, Hefei 230000, Anhui Province, China
Ji Li, Northeastern University, Shenyang 110000, Liaoning Province, China
ORCID number: Chang Zhang (0009-0005-9896-3239); Shui Jin (0009-0009-9782-4148).
Author contributions: Zhang C, Wu SH, Jin L, and Li J contributed to formal analysis; Wu SH, Li J, and Jin S participated in visualization; Zhang C and Wu SH participated in methodology; Zhang C and Jin L participated in data curation; Wu SH, Jin S, and Lu L participated in writing - review & editing; Zhang C participated in conceptualization, software, and wrote the original draft; Wu SH participated in investigation; Jin S participated in validation and accessed and verified the study data; Lu L participated in supervision. All authors critically reviewed and provided final approval of the manuscript.
AI contribution statement: The authors declare that no artificial intelligence tools were used in the preparation, analysis, or writing of this manuscript.
Supported by Key Project of Natural Science Research in Universities of Anhui Province, No. 2023AH050581.
Institutional review board statement: This study was approved by the Ethics Committee of the Fourth Affiliated Hospital of Anhui Medical University, No. KYXM-202605-018.
Informed consent statement: The need for patient consent was waived due to the retrospective nature of the study.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: The authors will provide the raw data that backs the conclusions without extra limits.
Corresponding author: Shui Jin, Department of Gastroenterology, The Fourth Affiliated Hospital of Anhui Medical University, No. 64 Chaohu North Road, Juchao District, Chaohu, Hefei 238000, Anhui Province, China. e-ch2009640010@fy.ahmu.edu.cn
Received: June 5, 2026
Revised: July 27, 2026
Accepted: September 10, 2026
Published online: November 15, 2026
Processing time: 153 Days and 18.7 Hours

Abstract
BACKGROUND

Immune checkpoint inhibitors (ICIs) show limited efficacy as monotherapy in microsatellite stable/proficient mismatch repair (MSS/pMMR) metastatic colorectal cancer (mCRC), largely due to an immunosuppressive tumor microenvironment with low T-cell infiltration. Preclinical studies suggest chemotherapy induces immunogenic cell death, while anti-angiogenic agents normalize tumor vasculature, potentially synergizing with ICIs to enhance antitumor immunity. However, real-world evidence for this combination remains insufficient, and whether clinical factors like tumor location or metastasis pattern influence outcomes requires validation. This study evaluated ICI-based combination therapy in MSS/pMMR mCRC using propensity score-based methods.

AIM

To evaluate ICI-based combination therapy efficacy and explore prognostic modeling and dynamic inflammatory biomarkers in MSS/pMMR mCRC.

METHODS

A total of 123 patients with MSS/pMMR mCRC were retrospectively enrolled. ICI-based combination therapy was compared with chemotherapy plus bevacizumab alone. Propensity score matching and inverse probability of treatment weighting were applied to reduce confounding, and E-value analysis assessed unmeasured confounding. The primary endpoint was progression-free survival (PFS). A prognostic nomogram was developed and temporally validated. Associations between post-treatment changes in inflammatory biomarkers and PFS were also evaluated.

RESULTS

After propensity score matching, ICI-based therapy was associated with longer PFS (hazard ratio = 0.486, P = 0.026), with consistent inverse probability of treatment weighting findings. Multivariable regression identified combination therapy as independently favorable and liver metastasis as unfavorable for PFS. Subgroup analyses suggested numerically greater benefits in right-sided and non-liver-metastasis patients, but interaction tests were not significant. A prognostic nomogram showed moderate discrimination temporally (C-index: 0.676-0.695). In exploratory analyses, elevated post-treatment neutrophil-to-lymphocyte ratio and systemic immune-inflammation index were associated with worse PFS in the overall cohort. Grade ≥ 3 adverse event rates were comparable between groups.

CONCLUSION

ICI-based combination therapy improved PFS with manageable safety in MSS/pMMR mCRC. Nomogram and dynamic neutrophil-to-lymphocyte ratio/systemic immune-inflammation index monitoring may facilitate risk stratification and warrant prospective validation.

Key Words: Metastatic colorectal cancer; Microsatellite stable; Immune checkpoint inhibitors; Anti-angiogenic therapy; Inflammatory biomarkers

Core Tip: Immune checkpoint inhibitors have limited efficacy in microsatellite stable/proficient mismatch repair metastatic colorectal cancer. This real-world study showed that immune checkpoint inhibitor-based combination therapy with chemotherapy and anti-angiogenic agents was associated with improved progression-free survival after propensity score adjustment. Liver metastasis and tumor sidedness showed exploratory prognostic and treatment-related signals, although subgroup findings require further validation. Post-treatment neutrophil‑to‑lymphocyte ratio and systemic immune‑inflammation index may provide additional risk stratification value. These findings offer clinically relevant evidence supporting further prospective evaluation of immune checkpoint inhibitor-based combination strategies in microsatellite stable/proficient mismatch repair metastatic colorectal cancer.



INTRODUCTION

Globally, colorectal cancer (CRC) is a common type of cancer, ranking high in both incidence and mortality rates. It poses a major health risk and leads to a significant economic burden[1,2]. According to epidemiological data, CRC is the third most common cancer and the second leading cause of cancer-related deaths worldwide. This burden is particularly pronounced in China, where both incidence and mortality rates remain high[3-5]. The survival outcomes for patients with early-stage CRC have significantly improved in recent years, owing to the widespread use of screening techniques such as colonoscopy and ongoing improvements in surgery, radiotherapy, and chemotherapy[6,7]. Still, a considerable number of patients are diagnosed at an advanced stage or already have distant metastases when initially diagnosed. Individuals with metastatic CRC (mCRC) typically face a grim outlook, with survival rates under 15% over five years[8-10]. Therefore, exploring more effective treatment strategies remains a major focus of current clinical research.

At present, the standard treatment for mCRC mainly includes combination chemotherapy regimens based on fluoropyrimidines, oxaliplatin, and irinotecan, with or without bevacizumab[11,12]. Despite the increasing availability of treatment options, the overall improvement in patient outcomes remains limited. The median progression-free survival (PFS) is approximately 11 months in the first-line setting and 5-7 months in the second-line setting, while the benefit from third-line and later therapies is even more modest[13,14]. Even with the introduction of targeted therapies and immunotherapy-based combinations in recent years, treatment efficacy in unselected populations remains suboptimal. These findings suggest that the current treatment paradigm still faces clear limitations, and more effective combination strategies are urgently needed to further improve clinical outcomes.

Immune checkpoint inhibitors (ICIs) have emerged as a major breakthrough in cancer therapy and have demonstrated durable antitumor effects across multiple solid tumors[15-17]. In CRC, ICIs have shown significant effectiveness in patients with deficient mismatch repair (MMR) or microsatellite instability-high tumors, becoming a standard treatment for this group[18,19]. Most cases of advanced or mCRC are associated with the microsatellite stable (MSS) or proficient MMR (pMMR) subtype. Characterized as “immune cold”, these tumors have a low mutation burden and limited lymphocyte infiltration, resulting in poor response to ICI monotherapy[20,21]. A study by Le et al[22] reported an extremely low objective response rate (ORR) in patients with MSS/pMMR mCRC treated with PD-1/PD-L1 inhibitors alone, indicating that single-agent immunotherapy is insufficient to overcome the immunosuppressive tumor microenvironment (TME) in this population[22].

The TME plays a critical role in determining the efficacy of ICIs and has attracted increasing attention in recent years[23]. MSS/pMMR CRC is generally characterized by an immunosuppressive TME, with low effector immune cell infiltration, increased presence of myeloid-derived suppressor cells and regulatory T cells, and abnormal vasculature, all of which contribute to poor immunotherapy response[24]. Therefore, modulating the TME to convert “immune cold” tumors into “immune hot” tumors has become an important research direction.

Previous investigations have found that some chemotherapeutic agents can cause immunogenic cell death, aid in the release of tumor antigens, and enhance antigen presentation, thereby triggering antitumor immune responses[25]. Meanwhile, anti-angiogenic therapy can transiently normalize tumor vasculature, improve perfusion and hypoxia, and promote T-cell infiltration while reducing immunosuppression, thereby enhancing the TME for immunotherapy[26]. Based on these mechanisms, combining ICIs with chemotherapy and anti-angiogenic therapy is believed to have potential synergistic effects. Several clinical studies have explored this strategy. The AtezoTRIBE study, for instance, showed that including atezolizumab with FOLFOXIRI and bevacizumab notably extended PFS[27]. Conversely, the CheckMate 9X8 trial failed to achieve its primary endpoint of PFS[28]. These findings indicate that the efficacy of such combination therapies may depend on specific treatment regimens and patient characteristics.

Despite these advances, several key questions remain unresolved. First, results from randomized controlled trials are inconsistent. Second, most studies are conducted in highly selected populations, which limits their generalizability to real-world patients who often present with more complex disease and greater treatment heterogeneity.

Systemic inflammation has emerged as a critical determinant of cancer initiation, progression, and therapeutic response. Peripheral blood-based inflammatory biomarkers, including the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII), have been extensively investigated as readily accessible prognostic indicators across various malignancies. In CRC, accumulating evidence has demonstrated the prognostic relevance of these inflammatory parameters, with elevated inflammatory indices consistently associated with unfavorable clinical outcomes[29-31]. However, in patients with mCRC, the clinical significance of longitudinal changes in peripheral inflammatory biomarkers during treatment remains insufficiently characterized.

To fill these gaps, this study evaluates ICIs combined with chemotherapy and bevacizumab in real-world MSS/pMMR mCRC. Compared with previous studies, this work has several notable strengths. First, propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) were used to reduce measured confounding. Second, post-treatment changes in inflammatory biomarkers (NLR, SII) were analyzed as exploratory prognostic indicators in the overall cohort. Together, these findings may provide clinically relevant evidence to guide future validation of immunotherapy-based combination strategies in MSS/pMMR mCRC.

MATERIALS AND METHODS
Patient selection

The study focused on patients with MSS/pMMR mCRC from two different periods. The primary group comprised 123 patients treated at the Fourth Affiliated Hospital of Anhui Medical University from May 2018 to January 2022, according to set criteria. In the experimental group, 31 participants were treated with chemotherapy, bevacizumab, and anti-PD-1 inhibitors, whereas in the control group, 92 participants received only chemotherapy and bevacizumab. A temporal validation cohort from the same institution, consisting of 61 patients treated between January 2023 and May 2025 according to the same eligibility criteria, was established to assess the nomogram over time. The validation cohort included 20 patients receiving ICI-based combination therapy and 41 patients receiving chemotherapy plus bevacizumab alone.

MSS/pMMR status was determined using a two-step diagnostic strategy. First, immunohistochemistry was performed to evaluate the expression of four MMR proteins, including MLH1, MSH2, MSH6, and PMS2, in tumor tissue samples. Patients with preserved expression of all four MMR proteins were classified as pMMR. For cases showing loss of one or more MMR proteins by immunohistochemistry, further molecular testing was performed using PCR-based microsatellite instability analysis. Patients with stable microsatellite markers were classified as MSS. Only patients with confirmed MSS/pMMR status according to the combined pathological and molecular evaluation were included in this study.

Inclusion criteria: (1) Pathologically confirmed unresectable metastatic colorectal adenocarcinoma, classified per American Joint Committee on Cancer (8th ed.); (2) ≥ 1 measurable lesion according to RECIST 1.1; (3) Treatment with either chemotherapy plus bevacizumab or the same regimen combined with anti-PD-1 therapy; (4) Eastern Cooperative Oncology Group Performance Status (ECOG PS) 0-2; (5) Age ≥ 18 years; and (6) Completion of ≥ 1 treatment cycle with ≥ 1 response assessment.

Exclusion criteria: (1) Previous exposure to ICIs; (2) Concurrent malignancies or severe systemic disorders (e.g., significant cardiac, pulmonary, hepatic, or renal impairment); (3) Inadequately controlled hypertension (> 140/90 mmHg); (4) Significant proteinuria (≥ 2+ g/24 hour or > 1 g/24 hour); (5) Coagulopathy or bleeding disorders; (6) Prolonged systemic corticosteroid use (prednisone ≥ 10 mg/day for > 2 weeks), interstitial lung disease, or active autoimmune conditions; and (7) Recent major surgery or radiotherapy (within 21 days) without sufficient recovery.

The main chemotherapy regimens included oxaliplatin-based regimens (such as FOLFOX and CAPEOX) and irinotecan-based regimens (such as FOLFIRI). In the ICI-based combination therapy group, anti-PD-1 antibodies, including sintilimab, tislelizumab, toripalimab, and camrelizumab, were administered in combination with chemotherapy and bevacizumab. Although ICI-based therapy is not routinely recommended for MSS/pMMR mCRC outside specific clinical contexts, a subset of patients received ICIs in routine practice based on physician judgment and individualized assessment. Factors influencing this decision may have included perceived disease aggressiveness, tumor burden, expected treatment tolerance, patient preference, treatment accessibility, and multidisciplinary discussion.

Data collection

The collected clinicopathological data included the following: (1) Initial traits: Age, gender, and ECOG PS; (2) Tumor-associated factors cover the primary tumor’s site, its resection, histological type, differentiation, lymphovascular or perineural invasion, liver and lung metastasis, the number of metastatic sites, and RAS status; (3) Treatment outcomes: PFS, overall survival (OS), follow-up time, treatment response, and survival status; and (4) Safety data: Adverse events during treatment and their grades.

Peripheral blood samples from all enrolled patients (n = 123) were collected within one week prior to treatment as baseline and again at 6 weeks to evaluate dynamic changes. The inflammatory markers assessed included NLR, lymphocyte-to-monocyte ratio (LMR), PLR, and SII. Posttreatment changes were classified as increased or decreased relative to baseline values for subsequent analysis.

PFS was the primary endpoint, measured from enrollment to disease progression or death from any cause. Secondary endpoints included OS, calculated from enrollment to death or last follow-up; ORR and disease control rate (DCR), representing the proportions of patients achieving tumor response or disease stabilization; and treatment-related adverse events, assessed according to CTCAE criteria. Follow-up was completed in January 2026. Follow-up for the primary cohort was conducted until December 2025, with data cutoff in January 2026. The temporal validation cohort was followed up through January 2026.

Statistical analysis

Confounding was addressed using PSM (1:2 matching ratio, caliper width 0.05) and IPTW. The propensity score was estimated using a logistic regression model incorporating clinically relevant baseline variables, including age, sex, ECOG PS, primary tumor location, primary tumor resection status, histological subtype, RAS status, liver metastasis, lung metastasis, and number of metastatic sites. Covariate balance before and after adjustment was assessed using standardized mean differences (SMD), with an SMD < 0.1 considered indicative of adequate balance. To quantify the potential impact of unmeasured confounding, an E-value analysis was performed based on the IPTW-adjusted hazard ratio (HR) for PFS. Survival was analyzed with Kaplan-Meier curves and log-rank tests, while Cox models were applied to determine factors associated with PFS and OS. Subgroup patterns were explored using interaction terms, with false discovery controlled using the Benjamini-Hochberg method. An exploratory nomogram integrating significant variables was built to estimate 3-, 6-, and 12-month PFS. Its performance was evaluated by C-index, time-dependent area under the curve (AUC), and calibration, with bootstrap resampling (1000 times) for internal validation and a later same-center cohort for temporal validation. Risk scores were used to divide patients into two groups based on an optimal cutoff, followed by survival comparison. Changes in inflammatory markers were analyzed in the overall cohort (n = 123) and classified as increased or decreased relative to baseline at the 6-week post-treatment time point, with associations to PFS assessed using survival analysis. All analyses were conducted in R (4.4.2) and SPSS (25.0), with two-sided P < 0.05 indicating significance.

RESULTS
Baseline characteristics of the study participants

A total of 123 patients were included, comprising 92 patients in the control group and 31 patients in the ICI-based combination therapy group. Treatment characteristics were generally well balanced between the two groups. The majority of patients received first-line therapy (77.4% in the ICI group and 76.1% in the control group), and oxaliplatin-based chemotherapy represented the predominant backbone regimen (74.2% and 76.1%, respectively). Among patients receiving ICI-based combination therapy, camrelizumab (16/31, 51.6%) and sintilimab (10/31, 32.3%) were the most frequently administered PD-1 inhibitors. Detailed distributions of treatment lines, chemotherapy regimens, ICI agents, and subsequent therapies after disease progression are summarized in Supplementary Table 1.

Propensity score distributions before and after matching are shown in Supplementary Figure 1. Before PSM, although conventional P value testing showed no significant differences between groups (all P > 0.05; Table 1), SMD evaluation revealed notable imbalances in several key covariates, particularly primary tumor location (SMD = 0.256) and primary tumor resection (SMD = 0.265), highlighting the limitation of P values in small cohorts and the necessity of propensity score-based adjustment (Supplementary Table 2; Supplementary Figure 2). After 1:2 PSM, overall balance improved substantially, with the mean SMD decreasing from 0.081 to 0.029 and most covariates achieving SMD < 0.1. Notably, the SMD for liver metastasis, the strongest unfavorable prognostic factor in our cohort, increased slightly from 0.075 to 0.101, marginally exceeding the threshold, likely reflecting a trade-off due to limited sample size during matching. This residual imbalance was further attenuated after IPTW, with the mean absolute SMD reduced to 0.015 (Supplementary Figure 3). Collectively, these diagnostic assessments confirm that both PSM and IPTW substantially improved baseline comparability, providing a robust foundation for subsequent efficacy analyses, while the mild residual imbalance in liver metastasis was explicitly addressed in additional sensitivity analyses.

Table 1 Baseline clinicopathological characteristics of patients before and after propensity score matching, n (%).
Variables
Subcategories
All patients (n = 123)
Before PSM
After PSM
Control group (n = 92)
Experimental group (n = 31)
P value
Control group (n = 62)
Experimental group (n = 31)
P value
Age< 6559 (47.97)45 (48.91)14 (45.16)0.87827 (43.55)14 (45.16)1.000
≥ 6564 (52.03)47 (51.09)17 (54.84)35 (56.45)17 (54.84)
GenderFemale48 (39.02)35 (38.04)13 (41.94)0.86427 (43.55)13 (41.94)1.000
Male75 (60.98)57 (61.96)18 (58.06)35 (56.45)18 (58.06)
ECOG PS0-180 (65.04)59 (64.13)21 (67.74)0.88341 (66.13)21 (67.74)1.000
243 (34.96)33 (35.87)10 (32.26)21 (33.87)10 (32.26)
Primary tumor locationRight colon48 (39.02)33 (35.87)15 (48.39)0.30628 (45.16)15 (48.39)0.941
Left colon75 (60.98)59 (64.13)16 (51.61)34 (54.84)16 (51.61)
Primary tumor surgeryNo43 (34.96)35 (38.04)8 (25.81)0.30918 (29.03)8 (25.81)0.935
Yes80 (65.04)57 (61.96)23 (74.19)44 (70.97)23 (74.19)
Histological typeAdenocarcinoma104 (84.55)78 (84.78)26 (83.87)1.00051 (82.26)26 (83.87)1.000
Mucinous adenocarcinoma19 (15.45)14 (15.22)5 (16.13)11 (17.74)5 (16.13)
Tumor differentiationPoorly differentiated44 (35.77)31 (33.70)13 (41.94)0.54123 (37.10)13 (41.94)0.821
Moderately to well differentiated79 (64.23)61 (66.30)18 (58.06)39 (62.90)18 (58.06)
Lymphovascular perineural invasionNo80 (65.04)61 (66.30)19 (61.29)0.77338 (61.29)19 (61.29)1.000
Yes43 (34.96)31 (33.70)12 (38.71)24 (38.71)12 (38.71)
Liver metastasisNo51 (41.46)39 (42.39)12 (38.71)0.88121 (33.87)12 (38.71)0.818
Yes72 (58.54)53 (57.61)19 (61.29)41 (66.13)19 (61.29)
Lung metastasisNo64 (52.03)48 (52.17)16 (51.61)1.00031 (50.00)16 (51.61)1.000
Yes59 (47.97)44 (47.83)15 (48.39)31 (50.00)15 (48.39)
Number of metastatic organs168 (55.28)52 (56.52)16 (51.61)0.79033 (53.23)16 (51.61)1.000
≥ 255 (44.72)40 (43.48)15 (48.39)29 (46.77)15 (48.39)
RASUnknown27 (21.95)19 (20.65)8 (25.81)0.81616 (25.81)8 (25.81)0.943
Wildtype44 (35.77)33 (35.87)11 (35.48)20 (32.26)11 (35.48)
Mutation52 (42.28)40 (43.48)12 (38.71)26 (41.93)12 (38.71)
Efficacy analysis

The median follow-up duration of the overall cohort was 26.51 months [95% confidence interval (CI): 24.51-33.25 months]. Before PSM, Kaplan-Meier survival analysis showed that the experimental group had a significantly longer median PFS than the control group (6.61 months vs 4.85 months; log-rank P = 0.027; Figure 1A). Cox regression analysis further demonstrated a significantly reduced risk of disease progression in the experimental group (HR = 0.551, 95%CI: 0.323-0.939, P = 0.029). In contrast, although the median OS was longer in the experimental group than in the control group (25.04 months vs 21.45 months), the difference was not statistically significant (log-rank P = 0.331; HR = 0.762, 95%CI: 0.441-1.317, P = 0.331; Figure 1B).

Figure 1
Figure 1 Kaplan-Meier curves for progression-free survival and overall survival before propensity score matching. A: Progression-free survival; B: Overall survival.

After PSM, the median PFS remained 4.85 months in the control group and 6.61 months in the experimental group. Compared with the control group, ICI-based combination therapy significantly prolonged PFS (log-rank P = 0.015). Cox regression analysis confirmed a significantly lower risk of disease progression in the experimental group (HR = 0.486, 95%CI: 0.257-0.918, P = 0.026). For OS, the median survival was 21.45 months in the control group and 25.04 months in the experimental group. Although a numerical improvement was observed, the difference did not reach statistical significance (log-rank P = 0.306). Consistently, Cox regression analysis did not show a significant difference (HR = 0.741, 95%CI: 0.402-1.366, P = 0.336; Figure 2).

Figure 2
Figure 2 Kaplan-Meier curves for progression-free survival and overall survival after propensity score matching. A: Progression-free survival; B: Overall survival.
Sensitivity analyses and assessment of residual confounding

Since liver metastasis remained slightly imbalanced after PSM (SMD = 0.101) and was identified as a strong adverse prognostic factor for PFS in our cohort, we performed an additional multivariable Cox regression analysis within the matched cohort with further adjustment for liver metastasis. After adjustment, ICI-based combination therapy remained associated with a numerically reduced risk of disease progression (HR = 0.604, 95%CI: 0.345-1.057, P = 0.078). Although the statistical significance was attenuated compared with the primary PSM analysis (HR = 0.486, 95%CI: 0.257-0.918, P = 0.026), the direction and magnitude of the treatment effect remained consistent. Liver metastasis remained independently associated with inferior PFS (HR = 2.691, 95%CI: 1.733-4.180, P < 0.001).

To further assess the robustness of the primary findings, IPTW was performed to adjust for measured baseline confounders. After weighting, the effective sample sizes were 90.1 (control) and 27.7 (treatment). Covariate balance improved substantially, with all SMDs < 0.1 and the mean absolute SMD decreasing from 0.050 to 0.015 (Supplementary Figure 3). Consistent with the primary analysis, the treatment group remained associated with longer PFS (median, 6.61 months vs 4.76 months; HR = 0.511; 95%CI: 0.296-0.882; P = 0.016), whereas no significant difference in OS was observed (HR = 0.750; 95%CI: 0.446-1.261; P = 0.278; Figure 3).

Figure 3
Figure 3 Survival analysis after inverse probability of treatment weighting. A: Progression-free survival comparison between treatment groups after inverse probability of treatment weighting adjustment; B: Overall survival comparison after inverse probability of treatment weighting. Weighted Cox regression analysis was used to estimate hazard ratios.

To further assess the potential impact of unmeasured confounding, an E-value analysis was performed. Based on the IPTW-adjusted HR for PFS (0.511), the E-value was 3.33, indicating that an unmeasured confounder would need to be associated with both treatment assignment and PFS by a risk ratio of at least 3.33 to fully explain away the observed association. These findings further support the robustness of the observed PFS benefit.

Exploratory subgroup analyses

Subgroup analyses were performed to explore potential treatment-effect heterogeneity across baseline characteristics (Figure 4). In the PFS analysis, patients with right-sided primary tumors (P = 0.002) and those without liver metastases (P = 0.002) showed numerically greater treatment benefit. However, formal tests for treatment-by-subgroup interaction were not statistically significant (all P for interaction > 0.05). These findings suggest that, although stronger treatment signals were observed in certain subgroups, there was no statistically significant evidence of differential treatment effects across subgroups. Therefore, these results should be considered exploratory and hypothesis-generating, requiring further validation in prospective studies.

Figure 4
Figure 4 Exploratory subgroup analyses of treatment effects on progression-free survival and overall survival. A: Progression-free survival; B: Overall survival. P values were adjusted using the Benjamini-Hochberg method. Formal interaction tests were not statistically significant, and subgroup findings should be interpreted as exploratory. ECOG: Eastern Cooperative Oncology Group.
Tumor response and adverse events

Tumor response: Tumor response was evaluated according to RECIST version 1.1 criteria. The ORR is the rate of patients achieving complete response or partial response, and the DCR is the rate of those achieving complete response, partial response, or stable disease. As shown in Figure 5A, substantial heterogeneity in tumor response was observed. Most patients exhibited tumor shrinkage (negative change), while a subset showed disease progression (positive change). Between the two groups, the response distribution did not differ significantly in a statistical sense (P = 0.650). The ORR was 32.26% for the experimental group and 23.91% for the control group, yet this difference lacked statistical significance (P = 0.356). Similarly, the DCR was 74.19% in the experimental group and 69.57% in the control group, with no significant difference (P = 0.820), as shown in Table 2. Figure 5B shows that the experimental group had numerically higher ORR and DCR. Despite the lack of statistical significance, the numerical pattern was consistent with the observed PFS association. These observations suggest that any potential effect of PD-1 inhibitors on short-term tumor response may be limited and may be reflected more in disease stabilization than in major tumor reduction.

Figure 5
Figure 5 Tumor response evaluation. A: Waterfall plot showing the maximum percentage change in target lesion size for each patient according to RECIST 1.1 criteria; B: Bar chart comparing objective response rate and disease control rate between the two groups. ORR: Objective response rate; DCR: Disease control rate; PD: Progressive disease; PR: Partial response; SD: Stable disease.
Table 2 Tumor response according to RECIST version 1.1 in the two groups.
Group
n
PR
SD
PD
ORR
DCR
Control group9222 (23.91)42 (45.65)28 (30.43)22 (23.91)64 (69.57)
Experimental group3110 (32.26)13 (41.94)8 (25.81)10 (32.26)23 (74.19)
P value0.6500.3560.8350.8200.3560.820

Adverse events: Throughout the treatment, 111 patients encountered at least one adverse event related to the treatment, with 29 in the experimental group and 82 in the control group. A summary of adverse events is provided in Table 3. The overall incidence of adverse events of any grade was comparable between the two groups. Bone marrow suppression was the most frequent hematological side effect, occurring in 53.26% of the control group and 58.06% of the experimental group (P = 0.798). All other adverse events, including gastrointestinal reactions and non-hematologic toxicities, did not differ significantly between the two groups (all P > 0.05). Notably, reactive cutaneous capillary endothelial proliferation (rCCEP) was observed in 16 patients (51.61%) in the experimental group and was exclusively observed among patients receiving camrelizumab-containing regimens. No cases were reported in the control group (P < 0.001). rCCEP is a characteristic adverse event associated with camrelizumab treatment[32]. In general, adding PD-1 inhibitors to chemotherapy with bevacizumab did not notably raise the risk of most side effects. Although rCCEP occurred at a relatively high frequency, it was predominantly grade 1-2 and clinically manageable. No substantial increase in severe adverse events was observed, indicating that the combination therapy had an acceptable safety profile.

Table 3 Treatment-related adverse events in the two groups, n (%).
Adverse events
Control group (n = 92)
Experimental group (n = 31)
P value
Myelosuppression49 (53.26)18 (58.06)0.798
Nausea27 (29.35)13 (41.94)0.284
Diarrhea18 (19.57)7 (22.58)0.918
Fatigue18 (19.57)6 (19.35)1.000
Proteinuria16 (17.39)3 (9.68)0.397
Hypertension7 (7.61)5 (16.13)0.302
Hepatic dysfunction24 (26.09)10 (32.26)0.666
Thyroid dysfunction13 (14.13)7 (22.58)0.411
Hand-foot syndrome8 (8.70)2 (6.45)1.000
rCCEP0 (0.00)16 (51.61)< 0.001

As shown in Figure 6, the overall distribution of adverse event grades was comparable between the two groups. Most adverse events were grade 1-2, while the incidence of grade 3-4 events was low in both groups. The occurrence of individual grade ≥ 3 adverse events was also rare, and no statistically significant differences were observed between the two groups in bone marrow suppression, nausea, diarrhea, fatigue, and other events (all P > 0.05; Table 4). Overall, the incidence of any grade ≥ 3 adverse events was comparable between the experimental and control groups (16.13% vs 17.39%, P = 1.000; Supplementary Table 3).

Figure 6
Figure 6 Adverse events and safety profile. Distribution of treatment-related adverse events by grade in the control and experimental groups. Most adverse events were grade 1-2, with comparable incidence of grade ≥ 3 events between groups.
Table 4 Grade ≥ 3 treatment-related adverse events in the two groups, n (%).
Grade ≥ 3 AEs
Control group (n = 92)
Experimental group (n = 31)
P value
Myelosuppression3 (3.26)3 (9.68)0.341
Nausea3 (3.26)0 (0.00)0.730
Diarrhea3 (3.26)0 (0.00)0.730
Fatigue1 (1.09)1 (3.23)1.000
Proteinuria2 (2.17)0 (0.00)0.995
Hypertension2 (2.17)0 (0.00)0.995
Hepatic dysfunction3 (3.26)1 (3.23)1.000
Thyroid dysfunction0 (0.00)1 (3.23)0.566
Hand-foot syndrome0 (0.00)0 (0.00)1.000
Prognostic factors and predictive model

Prognostic factors for PFS: Univariate and multivariate Cox regression analyses identified treatment group, primary tumor resection, tumor location, and liver metastasis as independent prognostic factors for PFS (all P < 0.05). For OS, only liver metastasis remained significant in multivariate analysis (P < 0.001; Table 5). Kaplan-Meier survival analysis confirmed these associations (Figure 7), with liver metastasis showing the strongest adverse prognostic impact.

Figure 7
Figure 7 Kaplan-Meier survival curves for key prognostic factors. A: Treatment group; B: Liver metastasis status; C: Primary tumor location; D: Primary tumor resection. PFS: Progression-free survival.
Table 5 Univariate and multivariate Cox regression analysis of prognostic factors for progression-free survival and overall survival.
Variables
Subcategories
PFS-univariate analysis
PFS-multivariate analysis
OS-univariate analysis
OS-multivariate analysis
P value
HR (95%CI)
P value
HR (95%CI)
P value
HR (95%CI)
P value
HR (95%CI)
Age< 65
≥ 650.1281.414 (0.905-2.211)0.3471.248 (0.787-1.981)
GenderFemale
Male0.4891.175 (0.745-1.853)0.1321.447 (0.895-2.339)
ECOG0-1
20.9961.001 (0.640-1.567)0.5240.856 (0.532-1.379)
Primary tumor locationRight colon
Left colon< 0.0010.456 (0.294-0.708)< 0.0010.385 (0.243-0.611)0.4360.833 (0.525-1.320)
Primary tumor surgeryNo
Yes0.0060.532 (0.341-0.831)0.0230.591 (0.376-0.929)0.3480.797 (0.496-1.281)
Histological typeAdenocarcinoma
Mucinous adenocarcinoma0.1931.454 (0.827-2.556)0.6091.176 (0.632-2.187)
Tumor differentiationPoorly differentiated
Moderately to well differentiated0.5670.877 (0.560-1.374)0.7511.083 (0.663-1.766)
Lymphovascular perineural invasionNo
Yes0.1091.455 (0.920-2.302)0.0111.819 (1.144-2.892)0.1351.430 (0.894-2.286)
Liver metastasisNo
Yes0.0031.996 (1.260-3.164)0.0121.830 (1.141-2.937)< 0.0013.772 (2.162-6.579)< 0.0013.516 (1.998-6.185)
Lung metastasisNo
Yes0.3620.813 (0.521-1.268)0.0870.668 (0.422-1.060)
Number of metastatic organs1
≥ 20.3021.261 (0.812-1.959)0.8491.046 (0.660-1.656)
RASUnknown
Wildtype0.3881.313 (0.707-2.436)0.9231.034 (0.528-2.024)
Mutation0.1341.566 (0.871-2.815)0.1721.532 (0.831-2.826)
GroupControl group
Experimental group0.0290.551 (0.323-0.939)0.0200.509 (0.288-0.901)0.3310.762 (0.441-1.317)

Development of the prognostic nomogram: A prognostic nomogram was developed from the entire pre-PSM cohort (n = 123) incorporating the four independent PFS predictors (Figure 8). During follow-up, 82 PFS events (66.7%) were observed in the development cohort, yielding an events-per-variable ratio of approximately 20.5, which substantially exceeds the recommended minimum threshold of 10 events per variable, suggesting a low risk of model overfitting. Each variable in the nomogram was given a score reflecting its influence on the result. The total score was calculated by summing the individual scores of all predictors, and this total score was then used to estimate the probabilities of PFS at 3, 6, and 12 months. As illustrated in Figure 8, higher total scores were associated with lower probabilities of longer PFS. To illustrate, a patient scoring approximately 211 had a 77.4% likelihood of PFS being greater than 3 months, 40.1% for more than 6 months, and 28.8% for more than 12 months, pointing to a moderate risk of disease progression.

Figure 8
Figure 8 Nomogram for predicting progression-free survival. The nomogram was constructed based on independent prognostic factors to predict 3-, 6-, and 12-month progression-free survival probabilities. PFS: Progression-free survival.

Risk stratification based on the nomogram: With the aid of R software, the nomogram model was used to assign patients to risk groups based on Cox-derived risk scores. An optimal cutoff value of 0.144 was determined, and patients were classified into low-risk and high-risk groups (Figure 9A). In the low-risk group, 40 of 80 patients experienced disease progression or death, corresponding to an event rate of 50.0%. In contrast, 42 of 43 patients in the high-risk group experienced events, yielding an event rate of 97.67%. The Kaplan-Meier survival analysis revealed that PFS was longer for patients in the low-risk group than for those in the high-risk group (log-rank P < 0.001; Figure 9B). The survival curves showed clear separation between the two groups. These findings suggest that the exploratory risk stratification model may help distinguish patients with different prognostic profiles. Because the cutoff value was derived and applied within the same development cohort, potential optimism in risk stratification performance should be considered.

Figure 9
Figure 9 Risk stratification based on the nomogram. A: Distribution of risk scores and optimal cutoff value; B: Kaplan-Meier curves comparing progression-free survival between low-risk and high-risk groups. HR: Hazard ratio; CI: Confidence interval.

Validation of the prognostic model: Internal validation via bootstrap resampling (1000 iterations) revealed a minimal optimism of 0.014 (apparent C-index: 0.695; optimism-corrected C-index: 0.681). The negligible discrepancy between apparent and corrected estimates indicates that the model is not overfitted to the development cohort and maintains stable discriminative performance. The AUCs for predicting 3-, 6-, and 12-month PFS were 0.723, 0.748, and 0.728, respectively. Calibration curves showed agreement between predicted and observed outcomes, and decision curve analysis suggested potential net clinical benefit (Figure 10).

Figure 10
Figure 10  Internal validation of the exploratory nomogram. A: Receiver operating characteristic curves for predicting 3-, 6-, and 12-month progression-free survival; B: Calibration curves showing agreement between predicted and observed probabilities; C: Decision-curve analysis suggesting potential clinical net benefit. AUC: Area under the curve; PFS: Progression-free survival.

Temporal validation was then performed using a cohort from the same center treated in a later period. The temporal validation cohort included 61 patients, among whom 48 PFS events were observed during follow-up. The C-index was 0.676, and the corresponding AUCs for 3-, 6-, and 12-month PFS were 0.742, 0.764, and 0.716, respectively. Similar calibration and decision curve results were observed (Figure 11). Overall, the exploratory nomogram showed acceptable discrimination and moderate performance over time in this temporal validation cohort. However, because this validation was conducted at a single center, these findings should be considered preliminary. Further multicenter validation is needed to confirm the model’s generalizability to broader populations.

Figure 11
Figure 11  Temporal validation of the exploratory nomogram. A: Receiver operating characteristic curves in the temporal validation cohort; B: Calibration curves; C: Decision-curve analysis suggesting potential clinical net benefit. AUC: Area under the curve; PFS: Progression-free survival.
Exploratory analysis of inflammatory biomarkers and clinical outcomes

In this exploratory analysis of the overall cohort (n = 123), posttreatment changes in systemic inflammatory biomarkers were evaluated for their association with PFS. Cox regression analysis showed that elevated NLR and SII were significantly associated with worse PFS (NLR: HR = 3.146, 95%CI: 1.147-8.634, P = 0.026; SII: HR = 2.526, 95%CI: 1.014-6.292, P = 0.047), while LMR and PLR showed no significant associations (both P > 0.05). Kaplan-Meier analyses confirmed these findings, with elevated NLR or SII corresponding to lower PFS rates, whereas no differences were observed for LMR or PLR (Figure 12). Overall, elevated NLR and SII at the single post-treatment assessment were significantly associated with shorter PFS. Although these exploratory findings do not establish them as validated prognostic biomarkers, they highlight the potential utility of routine inflammatory indices as dynamic, real-time prognostic indicators during combination therapy. Given their low cost and wide availability, post-treatment shifts in NLR and SII may inform early risk stratification, warranting further prospective investigation with standardized serial sampling and formal testing for treatment-by-biomarker interaction.

Figure 12
Figure 12  Kaplan-Meier curves for progression-free survival according to post-treatment changes in inflammatory biomarkers in the overall cohort. A: Neutrophil-to-lymphocyte ratio; B: Systemic immune-inflammation index; C: Lymphocyte-to-monocyte ratio; D: Platelet-to-lymphocyte ratio. Elevated neutrophil-to-lymphocyte ratio and systemic immune-inflammation index were associated with shorter progression-free survival, whereas lymphocyte-to-monocyte ratio and platelet-to-lymphocyte ratio were not significantly associated with progression-free survival. These findings are exploratory and based on a single post-treatment assessment. NLR: Neutrophil-to-lymphocyte ratio; SII: Systemic immune-inflammation index; LMR: Lymphocyte-to-monocyte ratio; PLR: Platelet-to-lymphocyte ratio; HR: Hazard ratio; CI: Confidence interval.
DISCUSSION

The limited efficacy of ICI monotherapy in MSS/pMMR mCRC is now widely recognized[33,34]. To overcome this therapeutic barrier, combining ICIs with chemotherapy and antiangiogenic agents has emerged as a promising strategy[35-37]. However, prospective clinical trials evaluating this approach have yielded inconsistent results. The AtezoTRIBE study demonstrated a PFS benefit with the addition of immunotherapy[27], whereas the CheckMate 9X8 trial failed to meet its primary endpoint[28]. These discrepancies suggest that the efficacy of combination strategies may be influenced by patient selection, treatment regimens, and trial design. Nevertheless, robust real-world evidence with rigorous adjustment for treatment selection bias remains limited. Therefore, in this study, we used real-world data and applied two complementary approaches, PSM and IPTW, to systematically evaluate the effectiveness, safety, and prognostic factors associated with ICI combined with chemotherapy and bevacizumab in patients with MSS/pMMR mCRC.

Our primary finding was that ICI-based combination therapy was associated with significantly prolonged PFS, whereas no significant improvements were observed in ORR or OS. The association between combination therapy and improved PFS remained highly consistent across the original cohort, PSM cohort, and IPTW-adjusted analysis, with HRs ranging from 0.486 to 0.551 (all P < 0.05), supporting the robustness of this observation. Because liver metastasis was identified as the strongest adverse prognostic factor in our cohort and showed a marginal residual imbalance after PSM (SMD = 0.101), we further performed a multivariable Cox regression analysis within the matched cohort with additional adjustment for liver metastasis. After this adjustment, the treatment effect remained directionally consistent (HR = 0.604, 95%CI: 0.345-1.057), although statistical significance was attenuated, likely owing to the reduced sample size after matching. These findings suggest that the observed PFS benefit was unlikely to be fully explained by residual imbalance in liver metastasis. Furthermore, the E-value analysis suggested that the observed association would require an unmeasured confounder with a relatively strong effect to fully explain away the observed benefit. However, neither ORR nor DCR differed significantly between groups. Similarly, no statistically significant OS benefit was observed (HR range, 0.741-0.762; all P > 0.05).

The observed PFS benefit is generally consistent with the biological rationale underlying the combination of ICIs with chemotherapy and antiangiogenic therapy in MSS/pMMR mCRC. Several prospective studies have investigated similar treatment strategies, although their outcomes have varied considerably. In the CheckMate 9X8 trial, the primary PFS endpoint was not achieved; however, the nivolumab-containing arm demonstrated a numerically higher 12-month PFS rate. The NIVACOR study reported encouraging activity in patients with RAS/BRAF-mutated mCRC[38], while the CAPability-01 trial suggested potential efficacy of immunotherapy-based combinations in previously treated MSS/pMMR disease[39]. Similarly, the AtezoTRIBE study provided evidence that adding immunotherapy to a backbone of chemotherapy plus bevacizumab may improve disease control in selected patients. Differences among these studies may reflect variations in treatment line, chemotherapy backbone, ICI agents, metastatic patterns, molecular characteristics, and patient selection criteria.

The pattern of improved PFS without significant improvement in ORR or OS warrants further consideration. Chemotherapy may induce immunogenic cell death, promote tumor antigen release, and enhance antigen presentation and T-cell priming. Meanwhile, bevacizumab-mediated vascular normalization may improve immune cell infiltration and reduce immunosuppressive signaling mediated by myeloid-derived suppressor cells and regulatory T cells[40,41]. These complementary mechanisms may create a more favorable immune microenvironment for checkpoint blockade. However, immune remodeling is likely a gradual process, and its early clinical impact may be reflected primarily through delayed disease progression rather than rapid tumor shrinkage. This may explain why PFS improvement was not accompanied by a corresponding increase in ORR.

The lack of OS benefit may have several potential explanations. First, in real-world settings, patients in the control group may receive effective subsequent therapies after progression, including salvage chemotherapy, targeted therapy, or even immunotherapy, thereby attenuating the survival difference attributable to initial treatment[42]. Second, the intrinsic immunologically “cold” nature of MSS tumors may limit the durability of immune-mediated tumor control, preventing early disease stabilization from translating into long-term survival improvement. Collectively, these findings suggest that current combination strategies may delay disease progression but may not yet overcome the fundamental mechanisms of immune escape in MSS CRC.

Subgroup analyses provided additional insights into potential treatment heterogeneity but should be interpreted cautiously. Patients with right-sided primary tumors and those without liver metastases showed numerically greater PFS benefits; however, formal interaction tests did not demonstrate statistically significant differences between subgroups. Therefore, these findings should be considered hypothesis-generating rather than evidence of differential treatment efficacy. Nevertheless, they are biologically plausible. Right-sided CRCs exhibit distinct molecular and immune characteristics, including higher mutation burden and increased immune infiltration in certain molecular subsets, which may contribute to enhanced sensitivity to immune activation[43,44]. In contrast, liver metastasis emerged as the most consistent adverse prognostic factor in our cohort. The liver can promote systemic immune tolerance through mechanisms involving antigen-presenting cells, regulatory immune populations, and impaired function of tumor-reactive T cells, thereby reducing responsiveness to immune checkpoint blockade[45-47]. Thus, hepatic metastasis may represent not only a marker of poor prognosis but also a biological barrier to effective antitumor immunity. These hypotheses require validation in adequately powered prospective studies incorporating biomarker-based stratification.

Beyond efficacy and potential treatment heterogeneity, safety represents a critical consideration for clinical implementation of this combination strategy. The overall incidence of adverse events and grade ≥ 3 toxicities was comparable between treatment groups, suggesting that the addition of PD-1 blockade did not substantially increase the toxicity burden of chemotherapy plus bevacizumab. Most adverse events were grade 1-2 and manageable, while severe toxicities were uncommon. rCCEP was observed exclusively in patients receiving camrelizumab-containing regimens, consistent with its established toxicity profile, and was generally low-grade and clinically manageable[32]. Taken together, the observed PFS benefit was achieved without a substantial increase in severe toxicity, supporting the feasibility and tolerability of this therapeutic approach.

In addition to evaluating treatment efficacy and safety, we explored whether routinely available clinical variables could improve prognostic stratification in patients with mCRC. An exploratory prognostic nomogram incorporating treatment group, primary tumor location, primary tumor resection status, and liver metastasis demonstrated moderate but acceptable discrimination, with a concordance index (C-index) of 0.695 in the development cohort and 0.676 in the temporal validation cohort. Considering the real-world heterogeneity of the study population and the fact that the model was constructed using readily available clinical variables, this level of discrimination was considered acceptable for exploratory risk stratification and was within the range reported for similar prognostic models in advanced CRC. Bootstrap correction demonstrated limited optimism, suggesting no substantial evidence of overfitting within the study population. Furthermore, model-based risk stratification was able to identify patient groups with significantly different PFS outcomes, supporting its potential value as an exploratory prognostic assessment tool. However, this nomogram should be interpreted as a prognostic stratification tool rather than a predictive model for treatment benefit. Because the validation cohort was derived from a later period at the same institution, this evaluation represents temporal validation rather than independent external validation. Therefore, although the model provides a clinically accessible framework for risk stratification, its broader clinical applicability remains preliminary and requires further validation in larger, independent multicenter cohorts.

Systemic inflammatory biomarkers may provide additional prognostic information. Markers including NLR, PLR, LMR, and SII reflect the interaction between systemic inflammation, immune status, and tumor biology and have been associated with outcomes across multiple malignancies. Previous studies in CRC have demonstrated that elevated NLR and SII are associated with unfavorable survival outcomes, supporting the biological relevance of systemic inflammation[29-31]. In our cohort, elevated post-treatment NLR and SII were associated with shorter PFS, whereas PLR and LMR were not significantly associated with outcomes. Biologically, elevated NLR and SII reflect a shift toward a pro-inflammatory and immunosuppressive state, which may facilitate tumor progression and immune evasion, providing a plausible biological basis for the observed associations. These findings suggest that readily available inflammatory markers may help identify patients at higher risk of disease progression. However, because these analyses were conducted in the overall cohort and formal treatment-by-biomarker interaction testing was not performed, NLR and SII should be interpreted as prognostic markers reflecting baseline risk and on-treatment inflammatory changes rather than predictive biomarkers capable of identifying patients who derive specific benefit from ICI-based therapy. Furthermore, these analyses were based on a single post-treatment measurement and should therefore be considered exploratory. Future prospective studies incorporating standardized sampling schedules, longitudinal monitoring, and formal interaction analyses are warranted to determine whether dynamic changes in inflammatory markers can independently track treatment response over time and provide incremental prognostic value beyond routinely available clinical variables.

This study has several strengths. First, it provides real-world evidence regarding ICI-based combination therapy in a population with substantial unmet clinical need. Second, the use of both PSM and IPTW approaches improved baseline comparability, and the consistency of PFS findings across multiple analytical frameworks strengthens confidence in the observed association. Third, beyond treatment effect estimation, this study further explored potential treatment heterogeneity, developed a practical prognostic model, and evaluated clinically accessible inflammatory biomarkers.

Several limitations should be acknowledged. First, the major limitation of this study is the potential for confounding by indication due to the retrospective treatment allocation. Since ICI-based therapy is not a standard treatment for MSS/pMMR mCRC, patients receiving ICI-containing regimens may have been selectively treated based on physician judgment, patient preference, treatment accessibility, and other clinical considerations. Although propensity-based methods adjusted for measured confounders, residual confounding from unmeasured factors, such as tumor burden, disease trajectory, socioeconomic status, and undocumented clinical characteristics, cannot be completely excluded. The E-value analysis provided sensitivity assessment for this potential bias but cannot eliminate the possibility of residual confounding. Second, treatment heterogeneity existed regarding chemotherapy backbone, ICI agents, treatment lines, and subsequent therapies, which may have influenced PFS and OS outcomes. Third, the prognostic model underwent only internal and single-center temporal validation and requires independent external validation in larger multicenter datasets. Fourth, inflammatory biomarkers were evaluated at only one post-treatment time point, limiting assessment of longitudinal changes. Finally, subgroup analyses, nomogram development, and biomarker analyses were exploratory and should be interpreted as hypothesis-generating.

CONCLUSION

In conclusion, the addition of PD-1 inhibitors to chemotherapy plus bevacizumab was associated with longer PFS without a substantial increase in severe toxicity in patients with mCRC. The absence of a significant OS benefit and the lack of significant subgroup interactions highlight the need for cautious interpretation. Nevertheless, the findings support further investigation of ICI-based combination strategies and suggest that metastatic pattern, routinely available clinical variables, and post-treatment inflammatory biomarkers may contribute to future prognostic stratification. Prospective multicenter studies with standardized treatment regimens, serial biomarker assessments, and independent validation are required to confirm these observations.

ACKNOWLEDGEMENTS

The authors thank the doctors and patients who provided data for this study.

References
1.  Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, Jemal A. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229-263.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 16785]  [Cited by in RCA: 18093]  [Article Influence: 9046.5]  [Reference Citation Analysis (55)]
2.  Chen L, Ma X, Dong H, Qu B, Yang T, Xu M, Sheng G, Hu J, Liu A. Construction and assessment of a joint prediction model and nomogram for colorectal cancer. J Gastrointest Oncol. 2022;13:2406-2414.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in RCA: 11]  [Reference Citation Analysis (0)]
3.  Dekker E, Tanis PJ, Vleugels JLA, Kasi PM, Wallace MB. Colorectal cancer. Lancet. 2019;394:1467-1480.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 4063]  [Cited by in RCA: 3732]  [Article Influence: 533.1]  [Reference Citation Analysis (40)]
4.  Kazemi E, Zayeri F, Baghestani A, Bakhshandeh M, Hafizi M. Trends of Colorectal Cancer Incidence, Prevalence and Mortality in Worldwide From 1990 to 2017. Iran J Public Health. 2023;52:436-445.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in RCA: 11]  [Reference Citation Analysis (0)]
5.  Wang JY, Huang JC, Chen G, Wei DM. Expression level and potential target pathways of miR-1-3p in colorectal carcinoma based on 645 cases from 9 microarray datasets. Mol Med Rep. 2018;17:5013-5020.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 2]  [Cited by in RCA: 16]  [Article Influence: 2.0]  [Reference Citation Analysis (4)]
6.  Van Cutsem E, Cervantes A, Adam R, Sobrero A, Van Krieken JH, Aderka D, Aranda Aguilar E, Bardelli A, Benson A, Bodoky G, Ciardiello F, D'Hoore A, Diaz-Rubio E, Douillard JY, Ducreux M, Falcone A, Grothey A, Gruenberger T, Haustermans K, Heinemann V, Hoff P, Köhne CH, Labianca R, Laurent-Puig P, Ma B, Maughan T, Muro K, Normanno N, Österlund P, Oyen WJ, Papamichael D, Pentheroudakis G, Pfeiffer P, Price TJ, Punt C, Ricke J, Roth A, Salazar R, Scheithauer W, Schmoll HJ, Tabernero J, Taïeb J, Tejpar S, Wasan H, Yoshino T, Zaanan A, Arnold D. ESMO consensus guidelines for the management of patients with metastatic colorectal cancer. Ann Oncol. 2016;27:1386-1422.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 2828]  [Cited by in RCA: 2624]  [Article Influence: 262.4]  [Reference Citation Analysis (24)]
7.  Zang J, Sun J, Xiu W, Liu X, Chai Y, Zhou Y. Low Expression of AGPAT5 Is Associated With Clinical Stage and Poor Prognosis in Colorectal Cancer and Contributes to Tumour Progression. Clin Med Insights Oncol. 2022;16:11795549221137399.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in RCA: 4]  [Reference Citation Analysis (1)]
8.  Liu Z, Li Y, Zhu Y, Li N, Li W, Shang C, Song G, Li S, Cong J, Li T, Xiu Z, Lu J, Ge C, Yang X, Li Y, Sun L, Li X, Jin N. Apoptin induces pyroptosis of colorectal cancer cells via the GSDME-dependent pathway. Int J Biol Sci. 2022;18:717-730.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 79]  [Cited by in RCA: 78]  [Article Influence: 19.5]  [Reference Citation Analysis (1)]
9.  Chin CC, Yeh CY, Tang R, Changchien CR, Huang WS, Wang JY. The oncologic benefit of high ligation of the inferior mesenteric artery in the surgical treatment of rectal or sigmoid colon cancer. Int J Colorectal Dis. 2008;23:783-788.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 86]  [Cited by in RCA: 82]  [Article Influence: 4.6]  [Reference Citation Analysis (1)]
10.  Ricke J, Mohnike K, Pech M, Seidensticker M, Rühl R, Wieners G, Gaffke G, Kropf S, Felix R, Wust P. Local response and impact on survival after local ablation of liver metastases from colorectal carcinoma by computed tomography-guided high-dose-rate brachytherapy. Int J Radiat Oncol Biol Phys. 2010;78:479-485.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 105]  [Cited by in RCA: 90]  [Article Influence: 5.6]  [Reference Citation Analysis (1)]
11.  Aslan F, Günaydın E, Yukruk F, Güvenç İ, Gençler OS. Glioblastoma Multiforme With CDKN2A, Loss of PTEN and EGFR Amplification, and Diffuse Distant Organ Metastasis Treated With Six Lines of Therapy: A Case Report and Literature Review. Cureus. 2022;14:e23796.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 3]  [Reference Citation Analysis (0)]
12.  Suenaga M, Mizunuma N, Matsusaka S, Shinozaki E, Ozaka M, Ogura M, Yamaguchi T. Phase II study of reintroduction of oxaliplatin for advanced colorectal cancer in patients previously treated with oxaliplatin and irinotecan: RE-OPEN study. Drug Des Devel Ther. 2015;9:3099-3108.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 30]  [Cited by in RCA: 50]  [Article Influence: 4.5]  [Reference Citation Analysis (4)]
13.  Kuriyama S, Yamada T, Matsuda A, Takahashi G, Iwai T, Takeda K, Ueda K, Miyasaka T, Yokoyama Y, Shinji S, Sonoda H, Ohta R, Yonaga K, Kanaka S, Yoshida H. Biomarkers for anti-vascular endothelial growth factor drugs. Oncol Lett. 2022;24:463.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in RCA: 5]  [Reference Citation Analysis (0)]
14.  Zhang H, Guo J, Gao S, Zhang P, Chen H, Wang X, Li X, Zhu X. Prognostic factors for transarterial chemoembolization combined with sustained oxaliplatin-based hepatic arterial infusion chemotherapy of colorectal cancer liver metastasis. Chin J Cancer Res. 2017;29:36-44.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 6]  [Cited by in RCA: 11]  [Article Influence: 1.2]  [Reference Citation Analysis (0)]
15.  Huang J, Li J, Zheng S, Lu Z, Che Y, Mao S, Lei Y, Zang R, Liu C, Wang X, Fang L, Sun N, He J. Tumor microenvironment characterization identifies two lung adenocarcinoma subtypes with specific immune and metabolic state. Cancer Sci. 2020;111:1876-1886.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 9]  [Cited by in RCA: 20]  [Article Influence: 3.3]  [Reference Citation Analysis (0)]
16.  Senju H, Kumagai A, Nakamura Y, Yamaguchi H, Nakatomi K, Fukami S, Shiraishi K, Harada Y, Nakamura M, Okamura H, Tanaka Y, Mukae H. Effect of IL-18 on the Expansion and Phenotype of Human Natural Killer Cells: Application to Cancer Immunotherapy. Int J Biol Sci. 2018;14:331-340.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 31]  [Cited by in RCA: 78]  [Article Influence: 9.8]  [Reference Citation Analysis (0)]
17.  Zeng G, Jin L, Ying Q, Chen H, Thembinkosi MC, Yang C, Zhao J, Ji H, Lin S, Peng R, Zhang M, Sun D. Regulatory T Cells in Cancer Immunotherapy: Basic Research Outcomes and Clinical Directions. Cancer Manag Res. 2020;12:10411-10421.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 6]  [Cited by in RCA: 20]  [Article Influence: 3.3]  [Reference Citation Analysis (0)]
18.  Li L, Chen C, Liu C, Niu L, Pan C. Clinical and molecular impacts of tumor mutational burden in histological and cytological specimens from cancer patients. Ann Transl Med. 2022;10:214.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 2]  [Reference Citation Analysis (0)]
19.  Huang J, Liu W, Kang W, He Y, Yang R, Mou X, Zhao W. Effects of microbiota on anticancer drugs: Current knowledge and potential applications. EBioMedicine. 2022;83:104197.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 66]  [Reference Citation Analysis (0)]
20.  Li DD, Tang YL, Wang X. Challenges and exploration for immunotherapies targeting cold colorectal cancer. World J Gastrointest Oncol. 2023;15:55-68.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in CrossRef: 21]  [Cited by in RCA: 24]  [Article Influence: 8.0]  [Reference Citation Analysis (0)]
21.  Liu JL, Yang M, Bai JG, Liu Z, Wang XS. "Cold" colorectal cancer faces a bottleneck in immunotherapy. World J Gastrointest Oncol. 2023;15:240-250.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in CrossRef: 8]  [Cited by in RCA: 15]  [Article Influence: 5.0]  [Reference Citation Analysis (2)]
22.  Le DT, Uram JN, Wang H, Bartlett BR, Kemberling H, Eyring AD, Skora AD, Luber BS, Azad NS, Laheru D, Biedrzycki B, Donehower RC, Zaheer A, Fisher GA, Crocenzi TS, Lee JJ, Duffy SM, Goldberg RM, de la Chapelle A, Koshiji M, Bhaijee F, Huebner T, Hruban RH, Wood LD, Cuka N, Pardoll DM, Papadopoulos N, Kinzler KW, Zhou S, Cornish TC, Taube JM, Anders RA, Eshleman JR, Vogelstein B, Diaz LA Jr. PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N Engl J Med. 2015;372:2509-2520.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 8030]  [Cited by in RCA: 7666]  [Article Influence: 696.9]  [Reference Citation Analysis (35)]
23.  Chen Y, Lu J, Xie Z, Tang J, Lian X, Li X. The Mechanism of Alisol B23 Acetate Inhibiting Lung Cancer: Targeted Regulation of CD11b/CD18 to Influence Macrophage Polarization. Drug Des Devel Ther. 2022;16:3677-3689.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 8]  [Reference Citation Analysis (0)]
24.  Zhang W, Jin C, Liu S, Wan X, Li Y, Liu J, Duan Z, Ma J, Gao Y. Myeloid-derived suppressor cells and regulatory T cells in colorectal cancer: a synergistic immunosuppressive axis and emerging therapeutic opportunities. Front Immunol. 2026;17:1757513.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 2]  [Cited by in RCA: 13]  [Article Influence: 13.0]  [Reference Citation Analysis (0)]
25.  Zhang W, Wang P, Pang Q. Immune checkpoint inhibitors for esophageal squamous cell carcinoma: a narrative review. Ann Transl Med. 2020;8:1193.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 30]  [Cited by in RCA: 32]  [Article Influence: 5.3]  [Reference Citation Analysis (1)]
26.  Li S, Zhang Q, Hong Y. Tumor Vessel Normalization: A Window to Enhancing Cancer Immunotherapy. Technol Cancer Res Treat. 2020;19:1533033820980116.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 19]  [Cited by in RCA: 38]  [Article Influence: 6.3]  [Reference Citation Analysis (0)]
27.  Antoniotti C, Rossini D, Pietrantonio F, Catteau A, Salvatore L, Lonardi S, Boquet I, Tamberi S, Marmorino F, Moretto R, Ambrosini M, Tamburini E, Tortora G, Passardi A, Bergamo F, Kassambara A, Sbarrato T, Morano F, Ritorto G, Borelli B, Boccaccino A, Conca V, Giordano M, Ugolini C, Fieschi J, Papadopulos A, Massoué C, Aprile G, Antonuzzo L, Gelsomino F, Martinelli E, Pella N, Masi G, Fontanini G, Boni L, Galon J, Cremolini C; GONO Foundation Investigators. Upfront FOLFOXIRI plus bevacizumab with or without atezolizumab in the treatment of patients with metastatic colorectal cancer (AtezoTRIBE): a multicentre, open-label, randomised, controlled, phase 2 trial. Lancet Oncol. 2022;23:876-887.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 236]  [Cited by in RCA: 231]  [Article Influence: 57.8]  [Reference Citation Analysis (4)]
28.  Lenz HJ, Parikh A, Spigel DR, Cohn AL, Yoshino T, Kochenderfer M, Elez E, Shao SH, Deming D, Holdridge R, Larson T, Chen E, Mahipal A, Ucar A, Cullen D, Baskin-Bey E, Kang T, Hammell AB, Yao J, Tabernero J. Modified FOLFOX6 plus bevacizumab with and without nivolumab for first-line treatment of metastatic colorectal cancer: phase 2 results from the CheckMate 9X8 randomized clinical trial. J Immunother Cancer. 2024;12:e008409.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 58]  [Cited by in RCA: 70]  [Article Influence: 35.0]  [Reference Citation Analysis (3)]
29.  Chen JH, Zhai ET, Yuan YJ, Wu KM, Xu JB, Peng JJ, Chen CQ, He YL, Cai SR. Systemic immune-inflammation index for predicting prognosis of colorectal cancer. World J Gastroenterol. 2017;23:6261-6272.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in CrossRef: 618]  [Cited by in RCA: 572]  [Article Influence: 63.6]  [Reference Citation Analysis (16)]
30.  Cai A, Zha B, Zhang C, Guo S, Wang G, He S. Unveiling the predictive role of systemic inflammation marker in mortality outcomes: a nationwide analysis of cancer survivorship in the United States. Discov Oncol. 2025;16:1356.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 1]  [Reference Citation Analysis (0)]
31.  Menyhart O, Fekete JT, Győrffy B. Inflammation and Colorectal Cancer: A Meta-Analysis of the Prognostic Significance of the Systemic Immune-Inflammation Index (SII) and the Systemic Inflammation Response Index (SIRI). Int J Mol Sci. 2024;25:8441.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 86]  [Reference Citation Analysis (14)]
32.  Yao J, Tan X, Sha Y, Chen Y, Chen R, Shi D. An updated review of immunotherapy in esophageal cancer: PD-L1 footprint. Cent Eur J Immunol. 2024;49:77-90.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 10]  [Cited by in RCA: 9]  [Article Influence: 4.5]  [Reference Citation Analysis (1)]
33.  Wei XL, Zhang Y, Zhao HY, Fang WF, Luo HY, Qiu MZ, He MM, Zou BY, Xie J, Jin CL, Zhou XF, Wang F, Wang FH, Li YH, Wang ZQ, Xu RH. Safety and Clinical Activity of SHR7390 Monotherapy or Combined With Camrelizumab for Advanced Solid Tumor: Results From Two Phase I Trials. Oncologist. 2023;28:e36-e44.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 2]  [Reference Citation Analysis (0)]
34.  Bai H, Wang WH, Zhou FF, Yang D, Li RJ. Feasibility and Tolerability of Anlotinib Plus PD-1 Blockades for Patients with Treatment-Refractory Metastatic Colorectal Cancer: A Retrospective Exploratory Study. Cancer Manag Res. 2024;16:73-86.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 4]  [Reference Citation Analysis (0)]
35.  Fang X, Zhu N, Zhong C, Wang L, Li J, Weng S, Hu H, Dong C, Li D, Song Y, Xu D, Wang J, Sun L, Wang J, Wang Z, Cao H, Liao X, Yu N, Xiao Q, Mi M, Zhang S, Ding K, Yuan Y. Sintilimab plus bevacizumab, oxaliplatin and capecitabine as first-line therapy in RAS-mutant, microsatellite stable, unresectable metastatic colorectal cancer: an open-label, single-arm, phase II trial. EClinicalMedicine. 2023;62:102123.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 32]  [Cited by in RCA: 31]  [Article Influence: 10.3]  [Reference Citation Analysis (0)]
36.  Thibaudin M, Fumet JD, Chibaudel B, Bennouna J, Borg C, Martin-Babau J, Cohen R, Fonck M, Taieb J, Limagne E, Blanc J, Ballot E, Hampe L, Bon M, Daumoine S, Peroz M, Mananet H, Derangère V, Boidot R, Michaud HA, Laheurte C, Adotevi O, Bertaut A, Truntzer C, Ghiringhelli F. First-line durvalumab and tremelimumab with chemotherapy in RAS-mutated metastatic colorectal cancer: a phase 1b/2 trial. Nat Med. 2023;29:2087-2098.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 88]  [Cited by in RCA: 97]  [Article Influence: 32.3]  [Reference Citation Analysis (0)]
37.  Damato A, Iachetta F, Antonuzzo L, Nasti G, Bergamo F, Bordonaro R, Maiello E, Zaniboni A, Tonini G, Romagnani A, Berselli A, Normanno N, Pinto C. Phase II study on first-line treatment of NIVolumab in combination with folfoxiri/bevacizumab in patients with Advanced COloRectal cancer RAS or BRAF mutated - NIVACOR trial (GOIRC-03-2018). BMC Cancer. 2020;20:822.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 20]  [Cited by in RCA: 24]  [Article Influence: 4.0]  [Reference Citation Analysis (4)]
38.  Damato A, Bergamo F, Antonuzzo L, Nasti G, Iachetta F, Romagnani A, Gervasi E, Larocca M, Pinto C. FOLFOXIRI/Bevacizumab Plus Nivolumab as First-Line Treatment in Metastatic Colorectal Cancer RAS/BRAF Mutated: Safety Run-In of Phase II NIVACOR Trial. Front Oncol. 2021;11:766500.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 19]  [Cited by in RCA: 17]  [Article Influence: 3.4]  [Reference Citation Analysis (11)]
39.  Wang F, Jin Y, Wang M, Luo HY, Fang WJ, Wang YN, Chen YX, Huang RJ, Guan WL, Li JB, Li YH, Wang FH, Hu XH, Zhang YQ, Qiu MZ, Liu LL, Wang ZX, Ren C, Wang DS, Zhang DS, Wang ZQ, Liao WT, Tian L, Zhao Q, Xu RH. Combined anti-PD-1, HDAC inhibitor and anti-VEGF for MSS/pMMR colorectal cancer: a randomized phase 2 trial. Nat Med. 2024;30:1035-1043.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in Crossref: 161]  [Cited by in RCA: 149]  [Article Influence: 74.5]  [Reference Citation Analysis (2)]
40.  Li Y, Liang X, Li H, Chen X. Efficacy and safety of immune checkpoint inhibitors for advanced non-small cell lung cancer with or without PD-L1 selection: A systematic review and network meta-analysis. Chin Med J (Engl). 2023;136:2156-2165.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Cited by in RCA: 12]  [Reference Citation Analysis (0)]
41.  Brueckl WM, Reck M, Rittmeyer A, Kollmeier J, Wesseler C, Wiest GH, Christopoulos P, Tufman A, Hoffknecht P, Ulm B, Reich F, Ficker JH, Laack E. Efficacy of Docetaxel Plus Ramucirumab as Palliative Third-Line Therapy Following Second-Line Immune-Checkpoint-Inhibitor Treatment in Patients With Non-Small-Cell Lung Cancer Stage IV. Clin Med Insights Oncol. 2020;14:1179554920951358.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 22]  [Cited by in RCA: 29]  [Article Influence: 4.8]  [Reference Citation Analysis (6)]
42.  Dimopoulos MA, Oriol A, Nahi H, San-Miguel J, Bahlis NJ, Usmani SZ, Rabin N, Orlowski RZ, Suzuki K, Plesner T, Yoon SS, Ben Yehuda D, Richardson PG, Goldschmidt H, Reece D, Ahmadi T, Qin X, Garvin Mayo W, Gai X, Carey J, Carson R, Moreau P. Overall Survival With Daratumumab, Lenalidomide, and Dexamethasone in Previously Treated Multiple Myeloma (POLLUX): A Randomized, Open-Label, Phase III Trial. J Clin Oncol. 2023;41:1590-1599.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 123]  [Reference Citation Analysis (0)]
43.  Yin H, Li H, Xu J, Wu J. Primary tumor location impacts survival in colorectal cancer patients after primary resection: a populationbased propensity score matching cohort study. J Gastrointest Oncol. 2023;14:886-899.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in Crossref: 8]  [Cited by in RCA: 10]  [Article Influence: 3.3]  [Reference Citation Analysis (1)]
44.  Gao Z, Wang XY, Shen ZG, Liu JH, Wang XY, Wu SK, Jin X. Real-world comparison of chemotherapy plus bevacizumab with or without immunotherapy as first-line therapy in colorectal cancer. World J Gastroenterol. 2025;31:108298.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 2]  [Reference Citation Analysis (0)]
45.  Kim HS, Kim CG, Hong JY, Kim IH, Kang B, Jung S, Kim C, Shin SJ, Choi HJ, Cheon J, Chon HJ, Lim HY. The presence and size of intrahepatic tumors determine the therapeutic efficacy of nivolumab in advanced hepatocellular carcinoma. Ther Adv Med Oncol. 2022;14:17588359221113266.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 17]  [Reference Citation Analysis (0)]
46.  Zeng ZM, Mo N, Zeng J, Ma FC, Jiang YF, Huang HS, Liao XW, Zhu GZ, Ma J, Peng T. Advances in postoperative adjuvant therapy for primary liver cancer. World J Gastrointest Oncol. 2022;14:1604-1621.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in CrossRef: 2]  [Cited by in RCA: 28]  [Article Influence: 7.0]  [Reference Citation Analysis (1)]
47.  Kuwano A, Yada M, Koga Y, Tanaka K, Ohishi Y, Masumoto A, Motomura K. Dynamics of the neutrophiltolymphocyte ratio during lenvatinib treatment for unresectable hepatocellular carcinoma. Oncol Lett. 2024;28:309.  [RCA]  [PubMed]  [DOI]  [Full Text]  [Full Text (PDF)]  [Cited by in RCA: 2]  [Reference Citation Analysis (0)]
Footnotes

Peer review: Externally peer reviewed.

Peer-review model: Single blind

Specialty type: Oncology

Country of origin: China

Peer-review report’s classification

Scientific quality: Grade B, Grade B, Grade C

Novelty: Grade B, Grade C, Grade C

Creativity or innovation: Grade B, Grade B, Grade C

Scientific significance: Grade C, Grade C, Grade C

P-Reviewer: Kashiv P, Assistant Professor, DM, MD, India; Zha B, Researcher, China S-Editor: Wu S L-Editor: A P-Editor: Yang YQ

Write to the Help Desk