Endocoil magnetic resonance imaging quanti®cation of external anal sphincter atrophy A. B. Williams*²³, C. I. Bartram*, D. Modhwadia*, T. Nicholls*, S. Halligan*, M. A. Kamm*, R. J. Nicholls² and W. A. Kmiot³ Departments of *Intestinal Imaging and Physiology and ²Surgery, St Mark's Hospital, Harrow and ³Department of Surgery, St Thomas' Hospital, London, UK Correspondence to: Professor C. I. Bartram, Department of Intestinal Imaging, Level 4V, St Mark's Hospital, Northwick Park, Harrow HA1 3UJ, UK e-mail: c.bartram@ic.ac.uk) Background: Anal function depends on the integrity and quality of the sphincter muscles. The diagnosis of external anal sphincter atrophy on endocoil magnetic resonography has been associated with poor outcome from sphincter repair, although the imaging criteria for atrophy remain unclear. Methods: Women with intact sphincters on endosonography and either normal more than 60 cmH 2 O) n = 9) or low n = 16) squeeze pressures had endocoil magnetic resonography and electromyography. The area and fat content of the external anal sphincter and puborectalis were measured on mid-coronal magnetic resonography and images were graded as showing normal, intermediate or advanced atrophy. The de®nition of the external anal sphincter on endosonography and the thickness of the internal anal sphincter were also assessed. Results: Women with a normal anal squeeze pressure had a larger external anal sphincter cross- sectional area means.d.) 24056) versus 19362) mm 2 ; P = 0´01) with a lower mean fat content means.d.) 234) versus 306) per cent; P < 0´001) than those with low squeeze pressures. There was an overall correlation between squeeze pressure, cross-sectional area r = 0´32, P = 0´02) and fat content r = ± 0´51, P < 0´001). Patients with a thin less than 2 mm) internal anal sphincter and/or a poorly de®ned external sphincter on endosonography were more likely to have atrophy positive predictive value 74 per cent). Conclusion: Potential endosonographic markers for external anal sphincter atrophy are suggested, and a visual scale for endocoil magnetic resonographic assessment has been validated. Paper accepted 15 February 2001 British Journal of Surgery 2001, 88, 853±859 Introduction Trauma to the anal sphincter and its nerve supply during vaginal delivery is the most common cause of faecal incontinence in women 1 .Endosonographyhasbecomean accepted technique for selecting those with anal sphincter disruptionwhomightbeeligibleforanterioranalsphincter repair. Anal sphincter repair has a favourable outcome in 60±80 per cent of patients 2±4 , although the long-term success rate falls to 50 per cent 5 . Endosonographic evidence of restoration of an intact ring of external anal sphincter EAS) has been associated with a good result 6±8 , although this is not the only determinant of outcome.Manometricvariableshavenotbeenshowntobe useful 8,9 , but evidence of pelvic ¯oor nerve damage, inferring EAS atrophy, has been associated with a poor outcome 10±12 . Endocoil magnetic resonance imaging MRI) has provided supporting evidence that atrophy is associated with a poor clinical outcome 13 , suggesting that qualitativeEASimagingmaybeanimportantdeterminant ofsurgicaloutcome. AtrophyoftheEASisconsideredtobetheendresultof denervation.Theultimateeffectsofdenervationonstriated muscle depend on axonal distribution, as well as on the extent and duration of nerve damage. Following denerva- tion, individual muscle ®bres either regain a nerve supply from the branching of adjacent intact axons, or undergo degeneration and atrophy with fatty replacement. Single- The Editors have satis®ed themselves that all authors have contributed signi®cantly to this publication Original article ã 2001 Blackwell Science Ltd British Journal of Surgery 2001, 88, 853±859 853 Downloaded from https://academic.oup.com/bjs/article/88/6/853/6267819 by guest on 20 October 2023