Multicenter validation of the PREDICT score for prediction of local recurrence after total mesorectal excision of rectal cancer

Abstract

Background Total mesorectal excision (TME) is the gold standard treatment for rectal cancer. Although TME has managed to decrease the rates of local recurrence after rectal cancer resection, local recurrence is still recorded at varying rates. The present study aimed to validate the PREDICT score in the prediction of local recurrence of rectal cancer after TME with curative intent. Methods This was a retrospective multicenter study on patients with nonmetastatic low or middle rectal cancer who underwent TME. The total PREDICT score was calculated for every patient and related to the onset of local recurrence. According to the final score, patients were allocated to one of three risk groups: low, moderate, and high, and the rates of local recurrence in each group were calculated and compared. Results The present study included 262 patients (50.4% males) with a mean age of 47.1 years. The overall local recurrence rate was 12.6%. 29.4% of patients were in the low‐risk group, 63.7% in the moderate‐risk group, and 6.9% in the high‐risk group. The local recurrence rate was 3.9% (95% confidence interval [CI]: 0.8–10.9) in the low‐risk group, 13.2% (95% CI: 8.4–19.3) in the moderate risk group, and 44.4% (95% CI: 21.5–69.2) in the high‐risk group (p < 0.0001). The sensitivity of the PREDICT score was 72.7%, the specificity was 88.1%, and the accuracy was 86.3%. Conclusions The PREDICT score had good diagnostic accuracy in the prediction of local recurrence after TME and a good discriminatory ability in the differentiation between patients at different risks to develop local recurrence.