Published online Oct 21, 2023. doi: 10.3748/wjg.v29.i39.5483
Peer-review started: July 11, 2023
First decision: September 18, 2023
Revised: September 21, 2023
Accepted: September 28, 2023
Article in press: September 28, 2023
Published online: October 21, 2023
Colorectal cancer (CRC) poses a serious threat to human life and health. Previous studies have shown that tumor deposits (TDs) are significantly associated with early metastasis and poor prognosis. However, research on related risk factors is limited, and accurate prediction of TDs remains challenging. We developed a model based on preoperative clinical and pathological features to accurately predict the likelihood of TDs in CRC patients.
At present, the diagnosis of TDs in CRC requires postoperative pathology, which is passive for clinicians. If TDs can be accurately identified before patients receive treatment, it is of great importance for evaluating clinical staging, selecting reasonable treatment plans, and judging the prognosis of CRC patients.
To develop and validate a nomogram with good predictive ability for the preoperative prediction of TDs in patients with CRC.
We retrospectively collected the data of 645 eligible patients with CRC. SPSS 27.0 and R (version 4.2.1) were used for statistical analysis, and a prediction model for TDs in CRC patients was established.
A total of 51 patients with CRC had TDs in this study. The areas under the curve of the training cohort and the validation cohort were 0.839 and 0.853, respectively. The model showed good accuracy and discrimination ability and has broad clinical practicability. The results of this study suggest the value of preoperative indicators in predicting TDs in CRC patients and can assist in guiding clinical decision making.
This nomogram based on preoperative indicators can effectively predict the preoperative TD status of CRC.
In the future, we will try to apply radiomics combined with clinical indicators to construct a model to predict the status of TDs in CRC patients.