Background: The wide variability of reported local recurrence rates after curative resection of rectal cancer without adjuvant therapy may be a consequence of many different factors. However, few studies have investigated the potential effects of such factors on local recurrence by multivariate analysis. The present study examined clinical and tumour characteristics, operation type and operative technique as potential predictors of local recurrence in patients treated by surgery alone.
Methods: Prospective data were analysed by bivariate and multivariate methods including actuarial survival and proportional hazards regression.
Results: Local recurrence (pelvic or pelvic and systemic) was diagnosed in 59 of 596 patients. The 5-year actuarial local recurrence rate was 11.1 (95 per cent confidence interval 8.7-14.3) per cent. Independent predictive factors for local recurrence were: positive nodes (hazard ratio (HR) 5.5, P < 0.01); distal margin of 1 cm or less (HR 3.8, P < 0.01); venous invasion (HR 2.0, P = 0.02) and total anatomical dissection of the rectum (HR 2.0, P = 0.01). There was no difference in local recurrence between patients who had the mesorectum divided and those in whom it was totally excised.
Conclusion: Nodal involvement is the most potent predictor of local recurrence. In patients in whom total anatomical dissection is performed, total mesorectal excision confers no additional protection against local recurrence.