Survival of colorectal cancer patients in Singapore by anatomic subsite: a population-based study.

INTRODUCTION

Studies of the prognostic value of anatomic subsite in colorectal cancer survival have yielded conflicting results. Two explanations for possible differences in survival patterns between proximal and distal lesions in the colorectum are biological difference between subsites and the presence of more early-stage lesions in distal than in proximal large bowel.

MATERIALS AND METHODS

A total of 435 cases with proximal lesions and an equal number with distal lesions diagnosed between 1990 and 1992 were randomly selected from the Singapore Cancer Registry. Information on vital status at 31 December 1996 were obtained by computerised matching with data from the National Registry of Births and Deaths.

RESULTS

Persons with proximal cancers in our study population did not present at a later stage than persons with distal cancer, local lesions (Dukes' Stage A + B) being 45.5% and 45.1%, respectively. Our analysis showed no significant differences in survival between subsites on a stage-for-stage basis. The 5-year survival rates were 42% and 44% for proximal and distal lesions, respectively (median survival times 3.98 and 4.27 years). Stage at diagnosis was the strongest predictor of survival. Among proximal lesions, 5-year survival rates were 57%, 36% and 12% for local, regional and metastatic lesions, respectively. The corresponding figures for the distal group were 65%, 37% and 10%. Age at diagnosis had a significant influence on survival.

CONCLUSIONS

Our results, based on population-based figures on survival of colorectal cancer patients where the impact of screening has not been large, do not support an independent influence of anatomic subsite in predicting survival of colorectal cancer.