Optimizing Surgical Outcomes After Complex Reconstruction of Pelvic Cancer Defects with the Vertical Rectus Abdominis Myocutaneous Flap

Introduction: Radiated pelvic defects after cancer resection require obliteration of deadspace with well-vascularized tissue to avoid perineal wound healing complications. The vertical rectus abdominis myocutaneous flap (VRAM) is the first choice for complex pelvic reconstruction in this high-risk population. We evaluate and compare the benefit of technical advances designed to improve donor and recipient site outcomes in this high-risk population. Methods: Prospectively collected data from all consecutive patients undergoing resection of pelvic cancers and immediate VRAM reconstruction from 2001 to 2009 were included in the study. Specific techniques used to potentially improve outcomes were evaluated and include: fascial-sparing VRAM flap harvest, de-epithelialized VRAM skin paddle with primary perineal skin closure, extended VRAM (across ipsilateral costal margin), addition of a pedicled omental flap to pelvic VRAM reconstruction and use of inlay mesh reinforcement of the fascial donor site, compared to standard VRAM flap. Patient defect and treatment characteristics were compared for equivalence between groups. Primary outcome measures included surgical outcomes, particularly donor and recipient site complications. Results: A total of 185 patients were included in the study with an average follow-up of 2 years. Patient demographics, co-morbidities, and treatment characteristics were equivalent between groups. The fascial-sparing VRAM harvest resulted in a significantly lower hernia rate compared to traditional VRAM harvest (1.5% to 11.5%, p<0.001), with a trend towards lower rates of dehiscence, bulge, and evisceration. Patients receiving donor site mesh inlay had greater incidence of abdominal laxity/bulge (7.7% versus 0%, p=0.001) but fewer post-operative hernias (2.6% versus 5.5%) than patients without mesh repair. Less recipient site dehiscence was noted for patients with de-epithelialized skin paddles (20.4%), those with the addition of an omental flap (11.1%, p<0.05) and patients with extended VRAM (7.7%, p<0.05) compared to traditional VRAM (32%). Multivariate logistic regression identified the addition of an omental flap as protective against recipient site dehiscence (OR=0.3, CI 0.09-0.95, p<0.05), and BMI as a positive predictor (OR 1.07, CI 1.017-1.125, p=0.009) of recipient site dehiscence. Conclusions: Several technical advances of VRAM flap reconstruction improve surgical outcomes after pelvic reconstruction. Adjunctive techniques should be considered in high risk patients to improve outcomes in these challenging cases. Further prospective studies will be important to elucidate patient selection criteria and indications for these techniques.