Perineal Wound Closure Following Abdominoperineal Resection and Pelvic Exenteration for Cancer: A Systematic Review and Meta-Analysis

Abstract

Background: Abdominoperineal resection (APR) and pelvic exenteration (PE) for the treatment of cancer require extensive pelvic resection with a high rate of postoperative complications. The objective of this work was to systematically review and meta-analyze the effects of vertical rectus abdominis myocutaneous flap (VRAMf) and mesh closure on perineal morbidity following APR and PE (mainly for anal and rectal cancers).

Methods: We searched PubMed, Cochrane, and EMBASE for eligible studies as of the year 2000. After data extraction, a meta-analysis was performed to compare perineal wound morbidity. The studies were distributed as follows: Group A comparing primary closure (PC) and VRAMf, Group B comparing PC and mesh closure, and Group C comparing PC and VRAMf in PE.

Results: Our systematic review yielded 18 eligible studies involving 2180 patients (1206 primary closures, 647 flap closures, 327 mesh closures). The meta-analysis of Groups A and B showed PC to be associated with an increase in the rate of total (Group A: OR 0.55, 95% CI 0.43-0.71; p < 0.01/Group B: OR 0.54, CI 0.17-1.68; p = 0.18) and major perineal wound complications (Group A: OR 0.49, 95% CI 0.35-0.68; p < 0.001/Group B: OR 0.38, 95% CI 0.12-1.17; p < 0.01). PC was associated with a decrease in total (OR 2.46, 95% CI 1.39-4.35; p < 0.01) and major (OR 1.67, 95% CI 0.90-3.08; p = 0.1) perineal complications in Group C.

Conclusion: Our results confirm the contribution of the VRAMf in reducing major complications in APR. Similarly, biological prostheses offer an interesting alternative in pelvic reconstruction. For PE, an adapted reconstruction must be proposed with specialized expertise.

Keywords: abdominoperineal resection; flap; mesh; perineal morbidity; perineal wound healing; rectal cancer; surgical oncology.

Figure 1

Flow diagram showing the search strategy for articles included in the systematic review and meta-analysis reported following the Preferred Reporting Systems for Systematic Reviews and Meta-Analysis statement.

Figure 2

Forest plot for total perineal wound complications between patients with primary closure and with flap closure (A); forest plot for total perineal wound complications between patients with primary closure and with mesh closure (B); forest plot for total perineal wound complications between patients with primary closure and with flap closure in pelvic exenteration (C). Odds ratio: overall ratio of total perineal wound complications when comparing primary closure with flap closure. All weights are from random effects analysis. M–H, Mantel–Haenszel.

Figure 3

Forest plot for major perineal wound complications between patients with primary closure and with flap closure (A); forest plot for major perineal wound complications between patients with primary closure and with mesh closure (B); forest plot for major perineal wound complications between patients with primary closure and with flap closure in pelvic exenteration (C). Odds ratio: overall ratio of major perineal wound complications when comparing primary closure with flap closure. All weights are from random effects analysis. M–H, Mantel–Haenszel.

Figure 4

Forest plot for minor perineal wound complications between patients with primary closure and with flap closure (A); forest plot for minor perineal wound complications between patients with primary closure and with mesh closure (B); forest plot for minor perineal wound complications between patients with primary closure and with flap closure in pelvic exenteration (C). Odds ratio: overall ratio of minor perineal wound complications when comparing primary closure with flap closure. All weights are from random effects analysis. M–H, Mantel–Haenszel.

Figure 5

Forest plot for length of stay between patients with primary closure and with flap closure. Odds ratio: overall ratio of length of stay when comparing primary closure with flap closure. All weights are from random effects analysis. M–H, Mantel–Haenszel.