Prognostic Factors Affecting Survival in Patients with Colorectal Carcinomatosis

Abstract

This study evaluates prognostic factors that may influence survival in patients who present with carcinomatosis from colorectal cancer. Patients may present with curcinomatosis as the pattern of metastases at the initial diagnosis of colorectal cancer. Little is known about the natural history of carcinomatosis and the prognostic factors affecting outcome. All patients treated at Roswell Park Cancer Institute from 1988 to 1994 who presented with carcinomatosis at the initial diagnosis of colorectal cancer were identified. A retrospective review of the medical records for patient and tumor demographics was performed. Estimated survival distributions were calculated by the method of Kaplan and Meier. Tests of significance with respect to survival distribution were based on the log-rank test. Cox proportional hazards model was used for the multivariate analysis. There were 31 males and 38 females. The median age was 61 years (range, 26–80). The primary cancers were in the sigmoid with 24 patients (35%), cecum with 14 patients (20%), and transverse colon with 11 patients (16%). The remainder were distributed throughout the colon and rectum. The most common presentation was large bowel obstruction in 29 patients (42%). T3 and T4 cancers were present in 39 (57%) and 13 patients (19%), respectively. Lymph nodes were positive in 39 patients (57%), and mucin-producing tumors were also present in 39 patients (57%). Twelve patients (17%) had one site of disease, 17 patients (25%) had two sites of disease, and 37 (54%) patients had three or more sites of disease. Ascites was present in 29 patients (42%). Residual disease was present at the completion of surgery in 45 patients, absent in 13 patients, and status unknown in 11. The presence of residual disease (p = 0.0001), presence of ascites (p = 0.02), stage greater than T3 (p = 0.02), and increasing number of carcinomatosis sites (p = 0.006) were found to have a negative impact on survival on univariate analysis. On multivariate analysis, only the presence of residual disease at the completion of surgery was found to be an independent predictor of survival (p = 0.04). Overall median survival was 14 months with a 26% estimated 2-year survival. The presence of gross residual disease at the completion of surgery was shown to be the only independent factor negatively affecting survival. This has potential implications for the operative management of patients presenting with colorectal carcinomatosis.