Does use of a metallic colon stent as a bridge to surgery modify the pathology data in patients with colonic obstruction? A case-matched study.

BACKGROUND

In a recent propensity score study, we established that overall- and disease-free survival were worse after use of a colonic stent (CS) than after emergency surgery for colonic obstruction. The present study sought to explain the association between CS use and poor survival by analyzing pathological specimens.

METHODS

From January 1998 to December 2011, all patients with left obstructive colon cancer and having been operated on with curative intent were included in the study. The primary end point involved a comparison of pathological data from the CS- and the surgery-only groups in a case-matched analysis (with the groups matched for the T stage). In a series of secondary analyses, we studied a range of factors known to be associated with adverse outcomes (microscopic perforation, vascular embolism, perineural invasion, and lymph node invasion) in the study population as a whole (in order to evaluate stenting as an adverse factor) and in the CS group alone (in order to identify factors associated with a poor prognosis in this specific group of patients).

RESULTS

A total of 84 patients were included in the study (50 in the CS group). Stenting was mentioned in only 70 % of the pathology lab reports (n = 35/50). Twenty-five patients in the CS group were matched with 25 patients of the surgery-only group. Tumor ulceration (p < 0.0001), peritumor ulceration (p < 0.0001), perineural invasion (p = 0.008), and lymph node invasion (p = 0.005) were significantly more frequent in the CS group. In a multivariate analysis of the CS group, T4 status and tumor size were significant risk factors for microscopic perforation, perineural invasion, and lymph node invasion.

CONCLUSION

The CS- and surgery-only groups differed significantly in terms of ulceration at or near the tumor, perineural invasion, and lymph node invasion. Explanation of the adverse outcomes associated with CS use will probably require further investigation.