A commentary on “Predicting disease-free survival following curative-intent resection of right-sided colon cancer using a pre-and post-operative nomogram: a prospective observational cohort study”

Dear Editor,

We recently read with great interest the article by James Lucocq and colleagues published in the International Journal of Surgery, which presented a model for predicting disease-free survival (DFS) in patients with right-sided colon cancer after radical resection^[1]. Based on a prospective cohort of 822 patients treated in a tertiary-care setting from 2010 to 2020, the study gathered data on demographics, clinical status, biochemistry, surgical procedures, and pathology. It employed Kaplan-Meier curves and Cox proportional hazards models to identify factors influencing DFS. The authors are to be congratulated on their work, which provides a new tool for predicting post- operative DFS and could assist clinicians in risk stratification, treatment planning, and surveillance. However, we have some questions to help strengthen the research. In line with the TITAN Guideline 2025, we carefully considered AI use in this study^[2]. We declare that no AI tools were used in any stage of the research, including research question development, data collection, analysis, and manuscript writing. We believe traditional methods and manual analysis better suited this study, ensuring its rigor and result reliability.

Study population and validation methods

The single-center design may limit the generalizability of the findings. Differences across centers in patient populations, resources, surgical techniques, and post-operative care can affect outcomes. The lack of external validation, with only bootstrap validation used, raises questions about the model’s external validity.

Model development

During model development, variable selection using log-rank analysis and stepwise regression might have caused bias. Some important clinical variables could have been left out, or some included variables might not directly relate to the outcome, which can affect the model’s accuracy and interpretability. The article largely overlooks the impact of anesthetic agents on cancer prognosis (Figure 1). Research indicates that while intravenous anesthetics can maintain safe and stable therapeutic drug concentrations during surgery and reduce surgical trauma, they can also affect the immune system. Immune system disturbances or suppression during the perioperative period can lead to postoperative complications, especially in cancer patients. Post-surgical immune suppression can speed up the spread of residual cancer cells and promote new metastases^[3]. Therefore, the type, dose, and duration of anesthetic drugs are factors that need more attention in the future. In addition, CEA is a common tumor marker for colorectal cancer, and its level is closely related to tumor burden and prognosis (Figure 1). Many studies have shown that elevated preoperative CEA levels are associated with an increased risk of postoperative recurrence. One study pointed out that preoperative CEA levels are an independent risk factor for postoperative recurrence in colorectal cancer, with high sensitivity and specificity, which can help predict patients’ prognosis more accurately^[4]. The degree of tumor differentiation, which reflects the similarity between tumor cells and normal cells, is another important factor (Figure 1). The poorer the differentiation, the higher the tumor malignancy and the worse the prognosis. In colorectal cancer research, it has been confirmed that the degree of tumor differentiation is one of the key factors affecting patient survival rates and is closely related to tumor recurrence and metastasis^[5]. Including tumor differentiation as a variable can help assess patients’ prognosis more comprehensively. Research also shows that depression may increase the risk of death in cancer patients (Figure 1). On the one hand, depression can produce pathophysiological effects by influencing neuroendocrine and immune functions related to mortality, such as hypothalamic-pituitary-adrenal axis dysfunction, especially the circadian changes in cortisol and melatonin^[6]. On the other hand, depressed patients may be less likely to adhere to preventive screening programs, cancer treatments, or healthy advice. For example, depressed patients may not exercise regularly, may be more likely to smoke and drink excessively, and may not follow prescribed medication regimens or miss treatment appointments^[7]. Including patients’ depression status in the model can help predict their DFS more accurately. Tumor size is related to tumor stage and malignancy (Figure 1). Larger tumors may indicate a higher tumor burden and relatively worse prognosis^[8]. Preoperative clinical staging, assessed through imaging and other tests, can reflect the initial tumor state. When combined with postoperative pathological staging, it can provide a more comprehensive assessment of tumor progression and prognosis.

Figure 1.

Visualization of risk factors on the prognosis of colorectal cancer patients.

Molecular biomarkers

The absence of molecular biomarkers such as microsatellite instability (MSI), KRAS, and BRAF mutations is a limitation. These markers are linked to prognosis and treatment responses (Figure 1) Incorporating them could improve the model’s predictive accuracy. For instance, a study delved into the impact of KRAS and BRAF mutations on colorectal cancer prognosis, asserting that these molecular markers enhance prognostic accuracy and guide treatment^[9]. Thus, integrating such molecular-biological markers into predictive models is crucial for boosting their personalized predictive capacity.

Model updating and maintenance

As medicine advances and new therapies emerge, the model’s predictive power may change over time. The study does not mention ways to update and maintain the model to ensure its future applicability. Also, the model fails to account for post-surgical patient changes at different times, like responses to adjuvant therapy and new complications. These factors can influence long-term DFS but are not reflected in the model.

Clinical application value

While the study hints at the model’s potential for clinical decision-making, it did not adequately explore how the model would be applied in real-world settings and its impact on treatment decisions.

This study adds to existing literature by offering a method to develop preoperative and postoperative nomograms based on prospective cohort studies. Researchers can learn about variable selection, data management, and statistical analysis techniques. The study highlights key factors affecting disease-free survival in right-sided colon cancer patients. It also shows how to apply prediction models in clinical practice for risk assessment and treatment decisions. Furthermore, it emphasizes the limitations of single-center designs and the importance of external validation. Future research directions, including incorporating molecular markers and model maintenance, are also clarified.

In conclusion, we appreciate the valuable contribution of James Lucocq and colleagues to the field of colorectal cancer prognosis. We believe that considering these comments could further enhance the scientific validity and utility of the research.

Ethical approval

Not applicable.

Consent

Not applicable.

Sources of funding

None.

Author contributions

H.X. C. and Z.L.Q: searched literatures and wrote the first draft; H.X. C.: revised the final manuscript. All authors approved the last version of the manuscript.

Conflicts of interest disclosure

The authors declare no conflicts of interest.

Guarantor

The Guarantors of this manuscript are Xiangcheng Hu.

Research registration unique identifying number (UIN)

Not applicable.

Provenance and peer review

Not applicable.

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.