[Clinicopathological features and prognosis of colorectal stromal tumor]

The incidence of colorectal stromal tumor is low among digestive tract tumors, therefore the literatures about clinicopathological features and prognosis of colorectal stromal tumor are few at home and abroad. In this study, we performed survival analyses for colorectal stromal tumor. The nomogram made by prognostic factors provided basis for evaluation of prognosis.

The clinico-pathological and prognostic data of colorectal stromal tumor between January 1992 and December 2015 were collected from the surveillance, epidemiology, and end results (SEER) database. The survival analyses were made by SPSS 24.0 software. The nomogram and calibration curve were made by RMS package in R 3.5.2 software.

In the study, 546 patients with colorectal stromal tumor were included. The median age of onset was 64 years. The regional lymph node metastasis (LNM) rate was 9.4%. The multivariate Cox regression analyses of the 546 cases showed that the older age of onset (>64 years), single or divorce, colon tumor (compared with rectal tumor), non-surgery, high histological grade, LNM and distant metastasis were associated with worse cancer specific survival (CSS) and overall survival (OS), P < 0.05 for all. The treatment district was independent prognostic factor of OS (P = 0.027). The C-index of independent prognostic factors predicting CSS and OS probability were 0.76 (95%CI: 0.72-0.80) and 0.75 (95%CI: 0.72-0.78), respectively. Multivariate analyses were further carried out in the 174 patients with definite histological grade and tumor location, which revealed that the age of onset, histological grade, surgery or not were independent prognostic factors of CSS and OS (P < 0.05 for all). Tumor location was associated with CSS (P = 0.041) but not OS (P = 0.057) among the 174 cases. Four independent prognostic factors influencing the 174 patients' prognosis were used to make nomogram for predicting survival probability of 546 cases. The C-index of four prognostic factors predicting probability of CSS and OS of the 546 cases were separately 0.71 (95%CI: 0.66-0.75) and 0.73 (95%CI: 0.70-0.77). The nomogram had more accuracy for predicting OS probability of colorectal stromal tumors.

The prognosis of colorectal stromal tumor was affected by multiple clinicopathological factors. The nomogram provided the basis for predicting the survival probability of patients with colorectal stromal tumor.