Original article Multicentre propensity score-matched analysis of conventional versus extended abdominoperineal excision for low rectal cancer H. Ortiz 1 , M. A. Ciga 2 , P. Armendariz 2 , E. Kreisler 3 , A. Codina-Cazador 4 , J. Gomez-Barbadillo 5 , E. Garcia-Granero 6 , J. V. Roig 7 and S. Biondo 3 on behalf of the Spanish Rectal Cancer Project Departments of Surgery, 1 Public University of Navarra and 2 Complejo Hospitalario de Navarra, Pamplona, 3 Colorectal Unit, Department of General and Digestive Surgery, Bellvitge University Hospital, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), Barcelona, 4 Colorectal Unit, Department of General and Digestive Surgery, Josep Trueta University Hospital, Girona, 5 Colorectal Unit, Department of Surgery, Hospital Universitario, Reina Sofia, Cordoba, and 6 Colorectal Unit, Department of Surgery, Hospital la Fe, Universidad de Valencia, and 7 Department of Surgery, Hospital General Universitario, Valencia, Spain Correspondence to: Dr S. Biondo, Department of General and Digestive Surgery, Bellvitge University Hospital, C/ Feixa Llarga s/n, L’Hospitalet de Llobregat, 08907 Barcelona, Spain (e.mail: sbn.biondo@gmail.com) Background: Abdominal perineal excision (APE) was originally described with levator ani removal for rectal cancer. An even wider, more aggressive extralevator resection for APE has been proposed. Although some surgeons are performing a very wide ‘extralevator APE (ELAPE)’, there are few data to recommend it routinely. This multicentre study aimed to compare outcomes of APE and ELAPE. Methods: A multicentre propensity case-matched analysis comparing two surgical approaches (APE and ELAPE) was performed. All patients who underwent abdominoperineal resection of a rectal tumour were considered for the analysis. Tumour height was defined by magnetic resonance imaging measurement and patients with stage II–III tumours had neoadjuvant radiochemotherapy. Involvement of the circumferential resection margin (CRM) and intraoperative tumour perforation were the main outcome measures. A logistic regression model was used to study the relationship between the surgical approaches and outcomes. Results: From January 2008 to March 2013 a total of 1909 consecutive patients underwent APE or ELAPE, of whom 914 matched patients (457 in each group) formed the cohort for analysis. Intraoperative tumour perforation occurred in 7·9 and 7·7 per cent of patients during APE and ELAPE respectively (P = 0·902), and there was CRM involvement in 13·1 and 13·6 per cent (P = 0·846). There were no differences between APE and ELAPE in terms of postoperative complication rates (52·3 versus 48·1 per cent; P = 0·209), need for reoperation (7·7 versus 7·0 per cent; P = 0·703), perineal wound problems (26·0 versus 21·9 per cent; P = 0·141), mortality rate (2·0 versus 2·0 per cent; P = 1·000) and local recurrence rate at 2 years (2·7 versus 5·6 per cent; P = 0·664). Conclusion: ELAPE does not improve rates of CRM involvement, intraoperative tumour perforation, local recurrence or mortality. Paper accepted 28 February 2014 Published online in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.9522 Introduction In spite of an increasing trend towards sphincter-saving surgery for low rectal cancer 1,2 , abdominoperineal excision (APE) is still commonly performed 3–5 . Experience with rectal tumours requiring APE suggests a poorer prognosis than for patients in whom anterior resection is possible 6 . It is likely that there are tumour-specific reasons for this rather than being due to the procedure itself 7–9 . Indeed, good results can be achieved with standard APE without radiotherapy for early disease 10 . A wider ‘extralevator APE (ELAPE)’ was described for advanced tumours to reduce the risks of circumferential resection margin (CRM) involvement and intraoperative tumour perforation 11,12 . This approach has gained some momentum, although there is no clear evidence to sug- gest the superiority of a more radical operation in all patients 13,14 . There may be indications for extended resec- tion in patients with disease beyond the mesorectum 15 . Moreover, some studies 16,17 found that ELAPE did not confer an advantage, and meticulous standard APE results have similar outcomes to those following anterior  2014 BJS Society Ltd BJS 2014; 101: 874–882 Published by John Wiley & Sons Ltd