Competing risk and random survival forest models for predicting survival in post-resection elderly stage I–III colorectal cancer patients
Abstract Elderly patients with colorectal cancer (CRC) face an elevated risk of cardiovascular and cerebrovascular death (CVD), yet few studies have explicitly addressed CVD as a competing risk event. Traditional survival analyses often overlook competing risks, potentially biasing prognostic estima...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-05824-1 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Elderly patients with colorectal cancer (CRC) face an elevated risk of cardiovascular and cerebrovascular death (CVD), yet few studies have explicitly addressed CVD as a competing risk event. Traditional survival analyses often overlook competing risks, potentially biasing prognostic estimates. This study aimed to evaluate cancer-specific survival (CSS) in elderly patients with stage I–III CRC after surgery using Fine-Gray subdistribution hazard model and a random survival forest (RSF) approach, thereby improving clinical decision-making. Older patients (≥ 65 years) with stage I–III CRC between 2010 and 2015 were selected from the Surveillance, Epidemiology and End Results (SEER) database. In addition, data from 2018–2021 in the database is extracted as an external validation set. In this study, CVD was considered as a competing risk event of CRC specific death, and Fine-gray regression analysis was used to construct the Fine-Gray subdistribution hazard model and a competing risk-based random survival forest (RSF) model were used to analyze postoperative cancer-specific survival (CSS) in elderly patients with stage I–III CRC as the best mechanism to obtain more precise results and help make clinical management decisions. Predictors included age, sex, race, marital status, grade, T stage, N stage, histological type, primary site, carcinoembryonic antigen (CEA), perineural invasion, tumor deposits, tumor size. Model performance was assessed through discrimination[C-index, area under the receiver operating curve (AUC)], accuracy[Brier score (BS)], and clinical utility[decision curve analysis (DCA)]. In addition, we also visualized the Fine-Gray subdistribution hazard model with a nomogram and compared it with the nomogram of the Cox model. A total of 19195 elderly (≥ 65 years) patients with stage I–III CRC who underwent primary site surgery between 2010 and 2015 were included in the study. There were 10305 deaths among all patients, including 4253 deaths specific to CRC, 2571 deaths due to cardiovascular and cerebrovascular diseases, 379 deaths due to other neoplastic diseases and 3120 deaths due to other non neoplastic diseases. The Fine-Gray subdistribution risk and RSF models we developed have good discrimination power and accuracy. The Fine-Gray subdistribution risk model:the 1-year, 3-year and 5-year C-index was 0.771, 0.775 and 0.759 in the train set, and 0.744, 0.762 and 0.753 in the internal test set . The 1-year, and 3-year C-index in the external validation set was 0.762 and 0.775.The RSF model:the 1-year, 3-year and 5-year AUC was 0.782 (95% CI 0.765, 0.798), 0.8 (95% CI 0.79, 0.811) and 0.786 (95% CI 0.776, 0.796) in the train set, and 0.754 (95% CI 0.727, 0.782), 0.786 (95% CI 0.769, 0.802) and 0.782 (95% CI 0.766, 0.797) in the internal test set. The 1-year and 3-year AUC was 0.77 (95% CI 0.749, 0.79) and 0.83 (95% CI 0.786, 0.82) in the external verification set. The 1-year, 3-year and 5-year BS was 0.053 (95% CI 0.050, 0.056), 0.104 (95% CI 0.101, 0.107) and 0.128 (95% CI 0.124, 0.132) in the train set, and 0.050 (95% CI0.044, 0.056), 0.106 (95% CI 0.098, 0.112) and 0.130 (95% CI 0.124, 0.136) in the internal test set. The 1-year and 3-year BS was 0.042 (5% CI 0.038, 0.044) and 0.085 (95% CI 0.078, 0.092) in the external verification set. The RSF model we established has good discrimination power and accuracy.The 1-year, 3-year, 5-year C-index was 0.801, 0.788 and 0.769 in the train set, and 0.744, 0.754 and 0.745 in the internal test set of the RSF model. The 1-year, and 3-year C-index in the external validation set was 0.761 and 0.771.The 1-year, 3-year and 5-year AUC was 0.792 (95% CI 0.776, 0.807), 0.813 (95% CI 0.802, 0.823) and 0.801 (95% CI 0.791, 0.811) in the train set and 0.749 (95% CI 0.721, 0.777), 0.779 (95% CI 0.762, 0.796) and 0.782 (95% CI 0.767, 0.798) in the internal test set (Fig. 6a, b). The 1-year and 3-year AUC was 0.767 (95% CI 0.747, 0.788) and 0.8 (95% CI 0.783, 0.817) in the external verification set (Fig. 7c). The 1-year, 3-year and 5-year BS was 0.053 (95% CI 0.51, 0.057), 0.105 (95% CI 0.102, 0.108) and 0.131 (95% CI 0.128, 0.134) in the train set, and 0.051 (95% CI0.45, 0.055), 0.109 (95% CI 0.102, 0.116) and 0.132 (95% CI 0.125, 0.140) in the internal test set (Fig. 7d, e). The 1-year and 3-year BS was 0.042 (95% CI 0.038, 0.045) and 0.086 (95% CI 0.082, 0.091) in the external verification set (Fig. 7f).DCA showed that models could lead to higher clinical benefits for patients. Through the nomogram we constructed, it can be calculated that the traditional Cox model overestimated the CSS of patients compared with the Fine-Gray subdistribution risk model. Based on the SEER database, the Fine-Gray subdistribution hazard model and the competing risk-based RSF model were used to predict CSS after CRC surgery in elderly patients, and models performed well. Incorporating competing risk events in survival analysis improves result accuracy and supports personalized clinical decision-making for elderly CRC patients. |
|---|---|
| ISSN: | 2045-2322 |