Overstaging of the mesorectal fascia following neoadjuvant therapy and its impact on therapeutic management: a single-center retrospective cohort study of 506 mesorectal fascia positive patients

Introduction

Colorectal cancer (CRC) is the third most common malignancy globally and poses substantial socioeconomic burdens (1). Notably, rectal cancer accounts for approximately one-third of CRC cases, with an increasing incidence (1,2). Due to anatomical complexity, locally advanced rectal cancer (LARC) presents greater surgical challenges and poorer prognosis than colon cancer (3). Standard LARC management includes neoadjuvant therapy (NAT) followed by total mesorectal excision (TME) and adjuvant treatment (4). As the cornerstone of multimodal therapy, TME is critical. The circumferential resection margin (CRM) status serves as a key indicator of TME completeness, with CRM-positive independently predicting adverse survival outcomes (5,6).​​ On magnetic resonance imaging (MRI), the mesorectal fascia (MRF) serves as a key anatomical surrogate for the potential CRM, theoretically corresponding to the surgical resection plane in standard TME. Persistent MRF involvement on post-therapy MRI suggests residual tumor threatening the CRM. These patients may require intensified NAT or extended resection to mitigate local recurrence risk (7,8).

High-resolution pelvic MRI is the most important method to evaluate MRF involvement. On T2-weighted imaging (T2WI), untreated rectal adenocarcinoma typically exhibits intermediate signal intensity compared to the muscularis propria (9). However, radiation-induced changes, including fibrosis, desmoplasia, and inflammatory edema, often mimic the signal characteristics of viable residual tumor on T2WI. These biological processes arise from NAT’s effects on the tumor microenvironment: radiation triggers fibroblast activation and extracellular matrix deposition, leading to dense fibrotic scarring that appears isointense to tumor tissue. Inflammatory responses further complicate interpretation by causing edema and vascular changes that enhance signal overlap (10,11). This leads to low specificity and frequent overstaging, which we operationally define in this context as persistent MRF involvement on post-NAT MRI in patients who subsequently achieve a pathologically negative CRM (ymrMRF+/ypCRM−). Such overstaging can potentially result in prolonged chemotherapy or unnecessary extended surgeries, increased perioperative morbidity, and delayed recovery without improving oncologic outcomes.

The concept of “threatened” MRF margins remains controversial, as they may represent either true viable tumor infiltration or benign treatment sequelae like fibrotic pseudotumor, lacking a clear distinction as a biological entity (12,13). While National Comprehensive Cancer Network (NCCN) guidelines recommend radical resection for locally advanced disease after NAT (14), emerging evidences suggest that patients with imaging-suggested MRF involvement but ypCRM− may achieve favorable outcomes without escalation, highlighting the potential for overtreatment (15). Consequently, high-quality data precisely linking the post-therapy MRF status to pathological outcomes are urgently needed to clarify these issues and guide clinical decision-making. This retrospective study aimed to investigate the association between post-NAT MRF involvement and CRM positivity, with a particular focus on quantifying the overstaging rate and its impact on survival outcomes. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-792/rc) (16).

Methods

Study design and population

This retrospective cohort study identified patients with primary rectal cancer who were treated at a university-affiliated cancer center between 2014 and 2024. The inclusion criteria were as follows: (I) biopsy-confirmed rectal adenocarcinoma; (II) baseline MRI demonstrating MRF invasion (tumor-to-MRF distance ≤1 mm). The exclusion criteria were: (I) failure to undergo rectal surgery; (II) incomplete NAT; (III) unavailability of pre- and post-treatment MRI; (IV) receipt of concurrent immunotherapy or targeted agents; (V) confirmed non-adenocarcinoma histologic subtypes. The flow diagram of study is shown in Figure 1. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Research Ethics Committee of The Fudan University Shanghai Cancer Center in China (No. 050432-4-1212B). All patients provided written informed consent for the analyses of their clinicopathological information.

Figure 1 Flow diagram of this study. CRM, circumferential resection margin; MRF, mesorectal fascia; MRI, magnetic resonance imaging.

Imaging protocol, pathological evaluation and follow-up

All patients underwent 3.0T pelvic MRI before and after NAT using standardized protocols: primary sequence: high-resolution T2-weighted turbo spin-echo [repetition time/echo time (TR/TE): 3,000–4,000/85 ms; slice thickness: 3 mm]. Two abdominal radiologists independently assessed tumor-to-MRF distance using routine clinical MRI protocols. In cases of disagreement between the two radiologists, a third senior radiologist was consulted to adjudicate and reach a consensus. This process ensured consistency and reliability in MRF positivity determinations across the study period. MRF positivity was defined as tumor infiltration within ≤1 mm of the MRF on MRI. All patients underwent TME by specialized colorectal surgeons. All surgical specimens underwent standardized processing beginning with ≥24 hours of formalin fixation. Following fixation, specimens were systematically sectioned at 5-mm intervals perpendicular to the rectal axis. Two specialized gastrointestinal pathologists independently performed macroscopic and microscopic evaluations to determine the minimal tumor-to-margin distance. CRM involvement was histologically defined as tumor presence within ≤1 mm of the resection plane, consistent with established prognostic criteria (17).

Patients underwent standardized surveillance comprising quarterly contrast-enhanced abdominopelvic computerized tomography (CT) with concurrent non-contrast chest CT, semiannual rectal MRI including high-resolution T2-weighted and diffusion-weighted sequences, and annual colonoscopy. Serum tumor markers [carcinoma embryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9)] were measured at each visit. Survival data were collected through medical records and telephone interviews, with the final follow-up cutoff date being December 2024. The median follow-up time for all patients was 36.56667 months. Overall survival (OS) was defined as time from treatment initiation to death from any cause or last contact.

Statistical analysis

Categorical variables were analyzed using χ² or Fisher’s exact tests, while continuous variables were compared via Student’s t-test or Mann-Whitney U test based on distributional assumptions. Confusion matrices visualization the predictive accuracy of MRF status for CRM involvement. Logistic regression identified independent risk factors for CRM positivity, with mixed effect models accounting for potential clustering effects. Survival analysis was performed using Kaplan-Meier analysis with log-rank testing for group comparisons, while Cox proportional hazards (PH) regression was used to determine independent prognostic factors. PH assumption was tested using Schoenfeld residuals. Violations were addressed with time-dependent covariates or stratification. Post-hoc power analysis evaluates if the current sample size suffices to detect the expected effect size. All analyses were two-sided with statistical significance defined as P<0.05, implemented in R version 4.4.2 (https://www.r-project.org/) (for Windows).

Results

Patient demographics

We enrolled 506 consecutive patients with baseline MRI-confirmed MRF involvement in this retrospective cohort study. Following NAT, post-treatment MRI revealed persistent MRF positivity in 50.2% of patients (n=254), while 49.8% achieved MRF conversion to negative status (n=252). Comparative analysis demonstrated a significantly lower BMI in the MRF-positive cohort than in the MRF-negative group (P=0.004). The MRF-positive group had a significantly higher rate of abdominoperineal resection and exhibited higher rates of advanced pathological T-stage (ypT3–4) and perineural invasion (P<0.001) as detailed in Table 1.

Discordance between MRI and CRM

Following NAT, CRM positivity rates were significantly higher in persistently MRF-positive patients (10.2%) compared to those achieving MRF clearance (1.6%, P<0.001). While MRF status demonstrated high sensitivity (86.7%) for predicting CRM involvement, its specificity was limited (52.1%), resulting in poor overall agreement (κ=0.086; Figure 2). Multivariable logistic regression confirmed persistent MRF positivity as an independent predictor of CRM compromise [odds ratio (OR): 6.23; 95% confidence interval (CI): 2.35–21.51; P<0.001; Table 2]. In robustness analyses accounting for potential metastatic confounding through mixed-effects modeling, MRF positivity remained a significant risk factor (OR: 6.58; 95% CI: 2.27–19.07; P=0.001; Table S1), though its standalone predictive value was modest (marginal R²: 0.194).

Figure 2 Confusion matrix for MRF and CRM. CRM, circumferential resection margin; MRF, mesorectal fascia.

Survival outcomes stratified by metastatic status

Univariate Cox regression identified MRF involvement as a significant prognostic factor for OS [hazard ratio (HR): 2.31; 95% CI: 1.57–3.39; P<0.001]. However, after adjusting for other risk covariates and addressing PH violations of CA19-9, this association was attenuated in multivariate analysis (Table S2). Given the independent prognostic value of metastatic (M) stage, we performed metastasis-stratified analyses. In the non-metastatic (M₀) cohort, persistent MRF positivity independently predicted inferior OS (adjusted HR: 2.37; 95% CI: 1.10–5.09; P=0.03). No significant survival difference was observed by MRF status in metastatic (M₁) patients (adjusted HR: 1.61; 95% CI: 0.86–3.03; P=0.14; Table 3), with a marginal PH violation post-adjustment (global P=0.02) attributed to limited events (n=72, events =30) and CA19-9’s covariate role. Table 3 and Table S3 present the HRs in the M1 subgroup before and after excluding CA19-9, with minimal changes confirming the robustness of the overall conclusions. Kaplan-Meier analysis confirmed this M-stage-dependent prognostic pattern (all: P<0.001, Figure S1A; M₀: P<0.001, Figure S1B; M₁: P=0.13, Figure S1C). However, the power analysis results of MRF positivity in the M1 group indicate a power value of 0.338 (Table S4).

Prognostic refinement using RAS mutation status

Cox regression analysis established RAS status as an independent predictor of survival across metastatic subgroups (M₀: HR: 3.15; 95% CI:1.34–7.41; P=0.008; M₁: HR: 2.96; 95% CI: 1.33–6.57; P=0.008). Building on this finding, we integrated RAS status with persistent MRF involvement to develop a comprehensive prognostic nomogram for M₀ patients (Figure S2A). This combinatorial model demonstrated significantly enhanced predictive accuracy over MRF-alone assessment at all evaluated timepoints [area under the curve (AUC): 1-year: 0.74 vs. 0.59; 3-year: 0.66 vs. 0.58; 5-year: 0.75 vs. 0.63; Figure S2B,S2C].

Discussion

This study evaluated the reliability of post-neoadjuvant MRI assessments for MRF involvement. We observed substantial discordance between radiologically assessed MRF status and histopathological CRM involvement (κ=0.086), aligning with prior literatures (18,19). Despite this limitation, persistent MRF positivity demonstrated an independent predictive value for both CRM compromise and survival outcomes in M₀ patients, underscoring its utility for identifying high-risk cohorts.

Demographic and clinical features showed several notable differences between the groups. Patients in the MRF-positive group had a significantly lower body mass index (BMI) compared to the MRF-negative group. This lower BMI may reflect underlying cancer cachexia or increased tumor burden, as nutritional depletion and systemic inflammatory responses associated with advanced disease stages can lead to weight loss in rectal cancer patients (20,21). Furthermore, a higher proportion of MRF-positive patients underwent Mile’s, which may have been preferentially selected to ensure complete tumor resection. Pathologically, the MRF-positive group exhibited higher rates of advanced T staging and perineural invasion. These findings indicate greater tumor aggressiveness and local invasiveness, which are associated with increased risk of MRF involvement.

Current guidelines uniformly recommend NAT followed by radical resection for LARC (14,22). However, evidence remains scarce regarding CRM positivity and survival outcomes, particularly in patients with persistent MRF involvement after NAT. Oberholzer et al. demonstrated that while MRF status reliably predicts resectability in patients who don’t undergo NAT [positive predictive value (PPV) =80%, negative predictive value (NPV) =89%], its predictive value deteriorates significantly after NAT (PPV =42%) (10). Our findings corroborate this deterioration, showing a markedly reduced specificity for CRM prediction and poor MRF-CRM concordance (κ=0.086). This diagnostic limitation primarily stems from post-therapeutic desmoplastic reactions and inflammation that obscure MRF evaluation on MRI. Specifically, radiation-induced fibrosis can mimic residual tumor on MRI, leading to overstaging by altering signal intensity and making it challenging to differentiate fibrotic scar tissue from viable cancer cells (23,24). Mucinous histology further complicates this, as mucinous tumors often exhibit high signal intensity on T2WI, which can overestimate tumor extent and MRF involvement due to the extracellular mucin pools that persist post-therapy and reduce concordance with histopathological CRM (25). These factors contribute to the low kappa value observed, highlighting the need for advanced imaging techniques to improve post-NAT accuracy.

Inaccuracies may prompt unnecessary extended resections or result in positive margins, which is a critical therapeutic dilemma that requires resolution. Despite suboptimal concordance between MRF status and CRM involvement, this study confirms that persistent MRF positivity remains an independent predictor of CRM compromise, with MRF-positive patients exhibiting a ​4-fold higher​ CRM positivity rate compared to MRF-negative counterparts (10.2% vs. 2.6%). This significant difference underscores MRF’s critical role in stratifying CRM risk. The integration of functional MRI sequences offers a promising solution for these diagnostic limitations. However, in this retrospective cohort, we did not have access to specific DWI data for direct comparison of its impact on MRF assessment specificity versus T2WI alone. Prior studies have demonstrated that incorporating DWI into standard MRI protocols can significantly enhance diagnostic accuracy in post-NAT settings, such as improving viable tumor detection and response evaluation in rectal cancer (11). Future prospective studies integrating DWI could address these gaps and further refine MRF-based risk stratification (26). Deep learning automates extraction of high-dimensional imaging features, demonstrating a significant potential to overcome post-NAT ambiguities through end-to-end lesion characterization. These approaches are well-established for predicting pathological complete response (pCR) in rectal cancer (27,28), yet their application to MRF assessment remains unexplored.

Persistent MRF involvement signifies locally aggressive disease and independently impacts survival outcomes (29). Our metastasis-stratified analysis revealed divergent prognostic outcomes. In M₀ patients, MRF involvement indicates that the tumor is locally aggressive, and complete tumor resection can significantly improve prognosis. In M₁ patients, the systemic biological behavior of tumor cells emerges as the predominant determinant of prognosis, consequently diminishing the prognostic utility of MRF status. The primacy of systemic disease over local factors in metastatic disease is well supported, since distant metastasis is a leading cause of death, and local control confers limited benefit when the tumor is widely disseminated (30). Factors such as nerve infiltration, extranodal tumor deposition, and systemic inflammation further exacerbate the poor outcome in M1 cases, masking the role of MRF (31). These findings indicate that systemic cancer control should be prioritized for M₁ patients. The SYNCHRONOUS and CCRe-IV trials confirmed that primary tumor resection offers no survival benefit for patients with unresectable metastases and may cause harm through surgical risks and delayed/omitted chemotherapy (32). Although multivariable Cox models initially showed PH assumption violations, these were addressed through time-dependent covariates or stratification, achieving compliance in overall and M0 models. In the M1 subgroup, a marginal PH violation persisted post-adjustment (global P=0.02), likely due to limited events (n=72, events=30) and CA19-9’s covariate role. Sensitivity analyses yielded minimal HR changes, confirming the robustness of primary findings. However, the lower value from the post-hoc power analysis indicates that the M1 subgroup analysis may be underpowered, which could have led to the negative results (Table S4). Future studies with larger sample sizes are needed to clarify its impact on the prognosis of M1 patients.

Moreover, RAS mutation status constitutes a key prognostic variable; our combinatorial model integrating RAS with MRF significantly enhanced the 5-year OS prediction in M₀ patients. This improvement likely stems from RAS-driven pathophysiology: MAPK/PI3K pathway hyperactivation induces therapeutic resistance (33). while concurrent recruitment of immunosuppressive cells and cancer-associated fibroblast activation promote immune evasion (34).

There are several limitations in this study. First, as a single-center retrospective study, our findings require validation in broader multicenter cohorts to ensure generalizability. Additionally, MRF manifests through various pathways (direct tumor invasion, tumor deposit invasion, metastatic lymph node involvement), but our study did not stratify by invasion subtype due to insufficient sample size in non-direct invasion categories. Finally, patients receiving neoadjuvant immunotherapy were excluded, given historical regimen heterogeneity. Given the proven efficacy of immunotherapy in CRC, future studies should assess the reliability of MRF in immunotherapeutic context.

Conclusions

In conclusion, this study demonstrates suboptimal concordance between post-neoadjuvant MRF status and CRM involvement. This finding implies that select MRF-positive patients may be candidates for surgical intervention rather than escalated chemotherapy. Notwithstanding these discrepancies, MRF involvement serves as a significant independent predictor of CRM positivity, thereby necessitating comprehensive multidisciplinary deliberation in the development of post-neoadjuvant treatment strategies. Furthermore, the integration of RAS mutation status with MRF evaluation refines prognostic stratification in non-metastatic rectal cancer, facilitating tailored, risk-adapted therapeutic approaches.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-792/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-792/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-792/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-792/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Research Ethics Committee of The Fudan University Shanghai Cancer Center in China (No. 050432-4-1212B). All patients provided written informed consent for the analyses of their clinicopathological information.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
2. Eng C, Yoshino T, Ruíz-García E, et al. Colorectal cancer. Lancet 2024;404:294-310. [Crossref] [PubMed]
3. Schrag D, Shi Q, Weiser MR, et al. Preoperative Treatment of Locally Advanced Rectal Cancer. N Engl J Med 2023;389:322-34. [Crossref] [PubMed]
4. Dossa F, Baxter NN. Changing the Treatment Paradigm for Locally Advanced Rectal Cancer. N Engl J Med 2023;389:375-7. [Crossref] [PubMed]
5. Diagnostic accuracy of preoperative magnetic resonance imaging in predicting curative resection of rectal cancer: prospective observational study. BMJ 2006;333:779. [Crossref] [PubMed]
6. Glimelius B, Beets-Tan R, Blomqvist L, et al. Mesorectal fascia instead of circumferential resection margin in preoperative staging of rectal cancer. J Clin Oncol 2011;29:2142-3. [Crossref] [PubMed]
7. Glynne-Jones R, Mawdsley S, Novell JR. The clinical significance of the circumferential resection margin following preoperative pelvic chemo-radiotherapy in rectal cancer: why we need a common language. Colorectal Dis 2006;8:800-7. [Crossref] [PubMed]
8. Detering R, Saraste D, de Neree Tot Babberich MPM, et al. International evaluation of circumferential resection margins after rectal cancer resection: insights from the Swedish and Dutch audits. Colorectal Dis 2020;22:416-29. [Crossref] [PubMed]
9. Beets-Tan RG, Beets GL. Rectal cancer: review with emphasis on MR imaging. Radiology 2004;232:335-46. [Crossref] [PubMed]
10. Oberholzer K, Junginger T, Heintz A, et al. Rectal Cancer: MR imaging of the mesorectal fascia and effect of chemoradiation on assessment of tumor involvement. J Magn Reson Imaging 2012;36:658-63. [Crossref] [PubMed]
11. Park MJ, Kim SH, Lee SJ, et al. Locally advanced rectal cancer: added value of diffusion-weighted MR imaging for predicting tumor clearance of the mesorectal fascia after neoadjuvant chemotherapy and radiation therapy. Radiology 2011;260:771-80. [Crossref] [PubMed]
12. Gollub MJ, Lall C, Lalwani N, et al. Current controversy, confusion, and imprecision in the use and interpretation of rectal MRI. Abdom Radiol (NY) 2019;44:3549-58. [Crossref] [PubMed]
13. Lambregts DMJ, Bogveradze N, Blomqvist LK, et al. Current controversies in TNM for the radiological staging of rectal cancer and how to deal with them: results of a global online survey and multidisciplinary expert consensus. Eur Radiol 2022;32:4991-5003. [Crossref] [PubMed]
14. Benson AB, Venook AP, Al-Hawary MM, et al. Rectal Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:1139-67. [Crossref] [PubMed]
15. Kim KH, Park MJ, Lim JS, et al. Circumferential resection margin positivity after preoperative chemoradiotherapy based on magnetic resonance imaging for locally advanced rectal cancer: implication of boost radiotherapy to the involved mesorectal fascia. Jpn J Clin Oncol 2016;46:316-22. [Crossref] [PubMed]
16. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453-7. [Crossref] [PubMed]
17. Taylor FG, Quirke P, Heald RJ, et al. Preoperative magnetic resonance imaging assessment of circumferential resection margin predicts disease-free survival and local recurrence: 5-year follow-up results of the MERCURY study. J Clin Oncol 2014;32:34-43. [Crossref] [PubMed]
18. Dent OF, Chapuis PH, Haboubi N, et al. Magnetic resonance imaging cannot predict histological tumour involvement of a circumferential surgical margin in rectal cancer. Colorectal Dis 2011;13:974-81. [Crossref] [PubMed]
19. Sautter-Bihl ML, Hohenberger W, Fietkau R, et al. Rectal cancer : when is the local recurrence risk low enough to refrain from the aim to prevent it? Strahlenther Onkol 2013;189:105-10. [Crossref] [PubMed]
20. Kasi PM, Zafar SY, Grothey A. Is obesity an advantage in patients with colorectal cancer? Expert Rev Gastroenterol Hepatol 2015;9:1339-42. [Crossref] [PubMed]
21. Shahjehan F, Merchea A, Cochuyt JJ, et al. Body Mass Index and Long-Term Outcomes in Patients With Colorectal Cancer. Front Oncol 2018;8:620. [Crossref] [PubMed]
22. Glynne-Jones R, Wyrwicz L, Tiret E, et al. Rectal cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2017;28:iv22-40. [Crossref] [PubMed]
23. Horvat N, El Homsi M, Miranda J, et al. Rectal MRI Interpretation After Neoadjuvant Therapy. J Magn Reson Imaging 2023;57:353-69. [Crossref] [PubMed]
24. Awiwi MO, Kaur H, Ernst R, et al. Restaging MRI of Rectal Adenocarcinoma after Neoadjuvant Chemoradiotherapy: Imaging Findings and Potential Pitfalls. Radiographics 2023;43:e220135. [Crossref] [PubMed]
25. Horvat N, Hope TA, Pickhardt PJ, et al. Mucinous rectal cancer: concepts and imaging challenges. Abdom Radiol (NY) 2019;44:3569-80. [Crossref] [PubMed]
26. Fornell-Perez R, Vivas-Escalona V, Aranda-Sanchez J, et al. Primary and post-chemoradiotherapy MRI detection of extramural venous invasion in rectal cancer: the role of diffusion-weighted imaging. Radiol Med 2020;125:522-30. [Crossref] [PubMed]
27. Zhang XY, Wang L, Zhu HT, et al. Predicting Rectal Cancer Response to Neoadjuvant Chemoradiotherapy Using Deep Learning of Diffusion Kurtosis MRI. Radiology 2020;296:56-64. [Crossref] [PubMed]
28. Zheng Y, Jiang Z, Shi J, et al. Encoding histopathology whole slide images with location-aware graphs for diagnostically relevant regions retrieval. Med Image Anal 2022;76:102308. [Crossref] [PubMed]
29. Zhang G, Xu Z, Zheng J, et al. Prognostic value of multi b-value DWI in patients with locally advanced rectal cancer. Eur Radiol 2023;33:1928-37. [Crossref] [PubMed]
30. Chen L, Xu F, Chen L. Diagnostic accuracy of artificial intelligence based on imaging data for predicting distant metastasis of colorectal cancer: a systematic review and meta-analysis. Front Oncol 2025;15:1558915. [Crossref] [PubMed]
31. Yu L, Yuan J, Lou S, et al. Factors affecting survival in patients with colorectal cancer: an umbrella review. J Transl Med 2025;23:1185. [Crossref] [PubMed]
32. Rahbari NN, Biondo S, Frago R, et al. Primary Tumor Resection Before Systemic Therapy in Patients With Colon Cancer and Unresectable Metastases: Combined Results of the SYNCHRONOUS and CCRe-IV Trials. J Clin Oncol 2024;42:1531-41. [Crossref] [PubMed]
33. Liu C, Ye D, Yang H, et al. RAS-targeted cancer therapy: Advances in drugging specific mutations. MedComm (2020) 2023;4:e285.
34. Hsu WH, LaBella KA, Lin Y, et al. Oncogenic KRAS Drives Lipofibrogenesis to Promote Angiogenesis and Colon Cancer Progression. Cancer Discov 2023;13:2652-73. [Crossref] [PubMed]