Relationship of diagnostic and therapeutic delay with survival in colorectal cancer: a review.

BACKGROUND

Early diagnosis of colorectal cancer before the onset of symptoms improves survival. Once symptoms have occurred, however, the effect of delay on survival is unclear. We review here evidence on the relationship of diagnostic and therapeutic delay with survival in colorectal cancer.

METHODS

We conducted a systematic of Medline, Embase, Cancerlit and the Cochrane Database of Systematic Reviews to identify publications published between 1962 and 2006 dealing with delay, survival and colon cancer. A meta-analysis was performed based on the calculation of the relative risk (RR) and on a model of random effects.

RESULTS

We identified 40 studies, representing 20,440 patients. Fourteen studies were excluded due to excessively restricted samples (e.g. exclusion of patients with intestinal obstruction, with tumours at stage C or D at the time of diagnosis, or who died 1-3 months after surgery); or because they studied only a portion of the delay. Of the 26 remaining studies, 20 showed no association between delay and survival. In contrast, four studies showed that delay was a factor contributing to better prognosis, and two showed that it contributed to poorer prognosis. There was no association between delay and survival when the colon and rectum were considered separately, when a multivariate analysis was performed, and when the effects of tumour stage and degree of differentiation were taken into account. To perform a meta-analysis, 18 additional studies were excluded, since the published articles did not specify the absolute numbers. In the remaining eight studies, the combined relative risk (RR) of delay was 0.92 (confidence interval (CI) 95%: 0.87-0.97).

CONCLUSIONS

The results of the review suggest that there is no association between diagnostic and therapeutic delay and survival in colorectal cancer patients. Colon and rectum should be assessed separately, and it is necessary to adjust for other relevant variables such as tumour stage.