LAUSR

Dear Sir, We read with interest the paper by Rezaie et al. on the use of 3D high definition anorectal manometry (3DARM) to detect anal sphincter defects in patients with faecal incontinence [1]. In their series of 39 patients, they described a new metrics to define anal pressure defect (a defect of at least 18° of the 25 mmHg isobaric contour on anal resting pressures) and then compared the results of pressure defects determined by 3DARM and 3D anal ultrasound results. They found a rather good negative predictive value of manometry to eliminate the presence of ultrasound anal sphincter defects (92%) and suggested the possibility of using 3DARM to rule out anal sphincter defects and avoid the need for anal ultrasound in selected patients. We described in a recent paper a similar method to identify a pressure defect, although we considered a pressure defect only when it encompassed the entire length of the anal high pressure zone, to take into account the physiological asymmetry of 3D anal pressures [2]. We report here our results obtained in a cohort of 120 consecutive patients with faecal incontinence or chronic constipation [101 women, mean age 55 (range 18–84) years], who underwent 3DARM (Medtronic, Shoreview, Minnesota, USA). We found the presence of pressure defects in 22 cases (18%). The inter-observer agreement for this metric was excellent, with a kappa coefficient of 0.91 (95% CI: 0.85–0.94). Resting pressures were significantly lower in the defect group [34 (SD 4) vs 87 (SD 2) mmHg, P < 0.0001 vs no defect], and all patients with a pressure defect had a mean anal resting pressure below 60 mmHg. Maximal and mean sustained squeeze pressures were also lower in the defect group (P < 0.0001). The physiological lambda aspect of 3D pressures during voluntary squeeze was significantly less frequent in the defect group (12/20, 60% vs 84/93, 90% in the no defect group, P = 0.0019) [3]. The manometric results could be compared to the detection of ultrasound anal sphincter defects in 36 patients (32 females) who underwent endoanal ultrasound (EAUS) (rigid rotating probe, Br€ uel & Kjaer, Naerum, Denmark) immediately after 3DARM. Anal sphincter defects were found in 24 patients on EAUS (12 isolated external sphincter defects, six isolated internal sphincter defects and six combined defects) and pressure defects in nine cases only. Our results were different from those of Rezaie et al. [1], with an excellent positive predictive value (the nine patients with a pressure defect had an EAUS defect, positive predictive value 100%), while the negative predictive value was only 44% (among the 27 patients without pressure defect, 15 had an EAUS defect). Comparing the manometric results in patients with and without EAUS anal sphincter defects, anal resting and squeezing pressures were significantly lower in patients with anal sphincter defects [56 (SD 4) vs 83 (SD 6) mmHg, P = 0.001, and 61 (SD 7) vs 94 (SD 9), P = 0.008, respectively]. The absence of the typical lambda aspect of 3D pressures was similar in both groups (21% in patients with EAUS anal sphincter defect and 20% in patients without). There may be several reasons to explain the differences between the studies. First, Rezaie et al. used 3D EAUS to identify anal sphincter defects, while we used 2D EAUS: 3D EAUS may be more specific than conventional EAUS for the diagnosis of anal sphincter defects, although this has not been clearly documented in the literature, and probably more reproducible than 2D EAUS [4,5]. This could explain why their series had a much lower prevalence of sphincter defects (21%) than ours (67%). Second, we used slightly different metrics for the diagnosis of pressure defect, to take into account the physiological asymmetry of pressures within the anal canal. This could explain why the prevalence of pressure defects was lower in our series (25%) than in the Rezaie study (36%) [1]. Finally, both series had a relatively small number of observations, which clearly limited their statistical value. We therefore recommend caution using 3DARM pressure defects as a surrogate marker for EAUS anal sphincter defects. In our view, it should not be surprising that pressure defects (or functional defects) do not correlate well with anatomical defects: a fibrous scar (anatomical defect) may exert a higher pressure on the probe, whether at rest or during voluntary squeeze, and thus may not correspond to a pressure defect. By definition, pressure defects are linked to low resting or squeezing pressures, and they must be analysed taking into account the physiological asymmetry of pressures within the anal canal. Previous studies have also shown the poor correlation between 3D pressure and EAUS defects [6,7].