Br. J. Surg. 1992, Vol. 79, February, 107-113

G . S. Duthie and D. C. C . Bartolo Ward 1 1 J 12, Royal Infirmary of Edinburgh, 1 Lauriston Place, Edinburgh EH3 9YW, UK Correspondence to: Mr G . S. Duthie

Abdominal rectopexy for rectal prolapse: a comparison of techniques T o compare the methods of abdominal rectopexy and to elucidate the mechanism by which rectopexy restores continence inpatients with rectal prolapse, the role of sphincter recovery, rectal morphological changes and improved rectal sensation were assessed in 68 patients (eight men, 60 women) of median age 63 (range 18-83) years undergoing resection rectopexy ( n = 29), anterior and posterior Marlex@ rectopexy (n = 20), posterior Ivalon@ rectopexy (n = 9 ) or suture rectopexy (n = 10). Preoperative and postoperative manometry, radiology and electrosensitivity measurements were made. Age and duration of follow-up were similar in all groups and the prolapse was controlled in all patients. Signijicantly improved continence was seen in all but the Ivalon group. There was no evidence of increasing postoperative constipation. Sphincter length and voluntary contraction were unaltered, but improved resting tone was seen in the resection and suture groups. This was not seen in the prosthetic groups. Improved continence correlated with recovery of resting pressure. Upper anal sensation was improved in all groups. Radiological changes did not correlate with improved continence. W e conclude that continence is improved by all rectopexy procedures but seems better without prosthetic material. Sphincter recovery seems to be the most important factor.

Faecal incontinence commonly accompanies complete rectal prolapse. Although abdominal rectopexy corrects the prolapse in the majority of cases, incontinence may persist in a number of patients. In this respect, abdominal procedures are superior to the perineal ones in that impaired continence is more prone to persist after operations like the Altmeier procedure, which involves a perineal rectosigmoidectomy sometimes accompanied by reefing of the pelvic floor muscles'!'. Overall, approximately 75 per cent of patients can expect their continence to be improved following abdominal rectopexy. There is debate surrounding the mechanism of this improvement. In a recent report, postoperative constipation affected 47 per cent3 and it was suggested that this was the mechanism of improved postoperative continence. Another report ascribed restoration of continence to recovery of the internal sphincter with an increase in postoperative anal canal pressures4. This study examined anorectal function before and after surgery prospectively in an attempt to elucidate the mechanisms of postoperative continence.

Patients and methods In the 3-year period 1986-1989,68 abdominal rectopexies were carried out by one of the authors (D.C.C.B.) at the Bristol Royal Infirmary (Table I ). The ma1e:female ratio was 8:60, with a median age of 63 (range 18-83 ) years. Using standard physiological history sheets, preoperative and postoperative assessment was carried out by an independent assessor (G.S.D.) to avoid operator bias. Objective preoperative and postoperative physiological variables (anal canal manometry, proctometrography, proctography and electrosensitivity ) were measured. Postoperative physiological assessment was carried out at 6 months. Continence Continence was assessed on a scale similar to that described by Browning and Parks5. Grade 1 is normally continent; grade 2 is normally continent to liquids and solids but incontinent to flatus ;grade 3 is incontinent to liquids and flatus but normally continent to solids; and grade 4 patients are incontinent to flatus, liquids and solids. In

0007-1323/92/020107-07

6 1992 Butterworth-Heinemann

Ltd

Table 1 General details

Suture rectopexy Ivalon rectopexy Marlex rectopexy Resection rectopexy

of

patients undergoing rectopexy

n

Age (years)

Duration of symptoms (months)

Follow-up (months)

10 9 20 29

70 (45-76) 60 (27-71) 62 (45-78) 60 (18-83)

7.5 (2-38) 8 (2-25) 6.5 (1-24) 8 (1-27)

11.5 (6-29) 13 (5-15) 6 (2-12) 6 (2-12)

Values are median (range). There were no significant differences between the groups assessing adequate preoperative and postoperative continence, grades 1 and 2 have been accepted as continent, but grades 3 and 4 are defined as incontinent. We believe this to be a socially acceptable definition. Manometry The closed water-filled microballoon system used by this unit is described elsewhere6. A station pull-through technique was used to define sphincter length, and maximum resting and maximum voluntary contraction anal canal pressures. The recorder was calibrated before each use. Preoperative and postoperative values have been paired for analysis. Proctometrography A similar catheter was covered by a thin latex balloon and placed in the rectum. This was filled at a constant rate (60 ml/min) with water at 37°C to determine the volume required for first awareness of rectal filling. Anal canal mucosal electrosensitivity The threshold sensitivity to an electrical stimulus was measured in the lower, mid and upper anal canal as defined by manometry. A graduated catheter bearing two platinum electrodes 1 cm apart and connected to a pulsed constant current generator' was used to determine the sensitivity thresholds by slowly increasing the current delivered until a tapping or prickling sensation was reported by the patient. The mean

107

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo

Figure 1 Colorectal mobilization for rectopexy :a f o r non-resectional procedures (lateral ligaments divided); b for resection and rectopexy (lateral ligaments divided) of three readings was taken as the threshold value at each level in the anal canal. The reliability and reproducibility of this investigation has been confirmed'. Transit studies All patients underwent a standard 5-day whole-gut transit study. They were instructed to refrain from the use of stimulant laxatives. Bulking agents were allowed if these were already being used as part of a high-fibre diet. Twenty radio-opaque markers were given on day zero and a plain radiograph of the abdomen was taken on day 5. Patients were considered to have slow transit if >20 per cent of the markers was retainedg. Proctoyraphy Patients were examined in the left lateral position with hips and knees flexed to 90". The degree of perineal descent and alterations in the anorectal angle were noted with the patients at rest, on maximal contraction of the pelvic floor, and while straining down. Barium and starch paste, simulating rectal contents, was introduced to the previously unprepared rectum while the anal canal was delineated by a radio-opaque ball and chainlo. The anorectal angle was defined as the angle between a line bisecting the rectum and the line of the anal canal. Perineal descent was assessed from the pubococcygeal line, as the perpendicular distance from this to the anorectal angle. An angle lying below this line was given a negative value for descent. In addition to these static studies, evacuation studies were undertaken with the patient seated on a commode at the same examination and used to confirm complete rectal prolapse. Statistics All investigative values were expressed as medians with ranges in parentheses. Paired preoperative and postoperative data were assessed by Wilcoxon's paired signed rank test. Correlations were assessed using Spearman's ranked correlation coefficient. P values of < 0.05 were considered as significant for this study. In comparisons between types of treatment, significance was accepted only when confirmed by the KruskalLWallis analysis of variance. Surgical procedures All operations were performed via a transabdominal route. Initially midline incisions were used, but we have now routinely adopted a low transverse muscle-cutting approach. A full mechanical bowel preparation and perioperative antibiotic regimen was used.

108

Non-resecrional rectopexy. Mobilization of the colon and rectum after initial laparotomy for all these procedures was similar (Figure l a ) . The sigmoid was fully mobilized after division of the congenital adhesions. The dissection was carried on round the extraperitoneal rectum to the level of the pelvic floor; Denonvilliers' fascia was divided well above the seminal vesicles in men to avoid injury to the nervi erigentes. The lateral ligaments were always divided. In the suture rectopexy group, after full mobilization, four non-absorbable sutures ( P r o h e % ;Ethicon Ltd., Edinburgh, U K ) were used to fix the rectum to the pelvic floor and presacral fascia (Figure 2a). In the group undergoing posterior Ivalon sponge rectopexy, the procedure followed was as described in 1959 by Wells". Ivalon" (polyvinyl alcohol sponge; Downs Surgical Ltd., U K ) was secured to both the rectum and the presacral fascia (Figure Zb) by non-absorbable sutures. The rationale for using Ivalon lies in the dense fibrous reaction it produces, resulting in firm fixation of the rectum within the sacral curve. The Ivalon itself is considered to be absorbed over a number of years". The anterior and posterior Marlex rectopexy procedure is similar to that described by Nicholls and SimsonI3. In addition to the sheet of Marlex'"' mesh (C. R. Bard Incorporated, Massachusetts, USA) sutured posteriorly, a strip of the material was also placed in the rectovaginal septum to provide greater support anteriorly (Figure 2c). Resection and rectopexy. The technique adopted was modified from that of Frykman and G ~ l d b e r g ' The ~ . extent of the rectal and colonic dissection and the resection lines are shown in Figure l h . The splenic flexure was always mobilized ; we believe that this ensures that healthy well vascularized descending colon is used for the anastomosis rather than the sigmoid, which is often diverticular and which may have a poor blood supply. As previously noted, the lateral ligaments were always divided. The colorectal anastomosis was performed without tension where convenient, usually at about 12 cm from the anal verge (Figure 2d). Low anastomosis was avoided. Finally, the lower rectum was fixed with non-absorbable sutures to the pelvic floor and presacral fascia.

Results Patient age, duration of symptoms and length of follow-up were similar for the different procedures (Table 1 ). Initially older patients were selected for simple suture rectopexy because it is a less extensive procedure. With increasing experience we have been happy to resect even in the elderly. In all cases the prolapse was well controlled by surgery and there has been no significant morbidity and no deaths. Furthermore, there has been no

Br. J. Surg., Vol. 79, No. 2, February1992

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo

Figure 2 a Simple suture rectopexy. Four SuturesJix the rectum to thepresacralfascia. b Posierior Ivalon sponge rectopexy. c Anierior and posterior Marlex rectopexy. d Resection and rectopexy. The rectum is sutured as in a, but the sigmoid colon and upper rectum have been excised with colorectal anastomosis

anastomotic breakdown, nor has any prosthetic material needed to be removed. Three patients in the Marlex group and three in the resection group had slow transit constipation as defined by retention of > 20 per cent of ingested markers at 5 days'. Bowel frequency was not significantly altered in any of the groups ( T a b l e 2 ) , nor was there any indication that postoperative constipation was a problem, as incomplete evacuation and straining at stool were unchanged by operation, except that in the resection group emptying was significantly improved. None of our patients required regular laxatives as a consequence ofthe surgery, but many remained on a high-fibre diet to maintain a regular bowel habit.

Br. J. Surg., Vol. 79, No. 2, February1992

Continence was significantly improved in all except the Ivalon group. This was most marked after resection and suture procedures. In the patients in whom continence was successfully restored, straining at stool was significantly reduced after operation in the resection and rectopexy group (0.01 < P < 0.05), but not after any other procedure. The awareness of the call to stool was unaffected by surgery. Anal manometry Table 3 shows the paired preoperative and postoperative manometric data. There was a significant rise in the resting anal canal pressures in the resection group (Figure 3) and the

109

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo Table 2 Functional outcome after four types of rectopexy

Bowel frequency (Per day 1

Continence to solid and liquids Preop.

Postop.

Straining at stool

Preop.

Postop.

Preop.

Incomplete emptying

Postop.

Preop.

Postop.

Suture rectopexy ( n = 10)

2 (20)

9 (90)*

2.5 (0.5-5)

1.5 (1-3)

5 (50)

4 (40)

5 (50)

2 (20)

Ivalon rectopexy (n = 9) Marlex rectopexy ( n = 20) Resection rectopexy ( n = 29)

4 (44)

6 (67)

1 (0.5-3)

1.5 ( 1 - 5 )

7 (78)

4 (44)

6 (67)

5 (56)

5 (25)

15 ( 7 5 ) t

3 (0.1-5)

1.75 (1-5)

13 (65)

11 (55)

13 (65)

8 (40)

6 (21)

24 (83)$

1.5 (0.1-6)

2 (0.5-5)

22 (76)

17 (59)

16 (55)

9 (31)*

Bowel frequencies are median (range); all other values are number (percentage) of patients. * P < 0.05; t P < 0.01; $ P < 0.001 (Wilcoxon's paired signed rank test)

Table 3 Manometric data before and after rectopexy by four techniques Suture rectopexy

Sphincter length (cm) Maximum resting pressure (cmH20 1 Maximum voluntary contraction (cmH*O)

Ivalon rectopexy

Marlex rectopexy

Preop.

Postop.

Preop.

Postop.

Preop.

Postop.

Resection rectopexy _ _ Preop. Postop.

3 (1-4) 25 (10-100) 90 (40-200)

3 (2-3.5) 55 (20-100) 87 (40-175)

3 (2-4) 60 (35-85) 180 (70-295)

3 (2'5-4) 55 (30-125) 135 (85-313)

3 (0-4) 60 (20-160) 113 (45-265)

3 (0-4) 52 (20-125) 115 (30-225)

3 (0-4) 45 (12-145) 1 I5 (32-215)

~

3 (1-4) 65 (25-185)* 110 (50-230)

Values are median (range). * P < 0.001 (Wilcoxon's paired signed rank test)

60

O'

-

1

W

3

:

200

L

0.

1

m

,F + !A

150

Y

2 ON I $E -5

E

100

.-

m .-

.-__ ....

50

+m L 0)

a

0

-2

-1

0

1

2

3

4

L a,

a

Change in continence

Preoperative resting pressure correlates inversely with change in conlinence. rs = -0.32, P = 0.008 (Spearman',. ranked correlation Figure 4 tesl)

Suture rectopexy

I valon

rectopexy

Marlex rectopexy

Resection rectopexy

Figure 3 Changes in maximum restiny pressure before (0) and after ( ) operation. * There was a significant improvement o w r preoperative

values after resecrion and rectopexy ( P < 0.001, Wilcoxon's paired signed rank test)

same trend towards improvement was seen in the suture group, although this did not achieve statistical significance. There was no significant alteration in pressures in the implant groups. Indeed, there was a trend towards reduction in anal pressures after Marlex and Ivalon rectopexy. In addition, the improvement in continence for all patients was correlated with the preoperative resting pressure ( r s = -0.32, P = 0.008; Figure 4 ) and with the improvement in resting pressure after operation (rs = 0.26, P = 0.034; Figure 5 ).

110

Anorectal sensation Rectal sensation was assessed in only three patients in the suture and Ivalon groups, and hence no data are presented ( T a b l e 4 ) . The volume required to produce a sensation of rectal filling was reduced in the other two groups but reached statistical significance only in the resection group. Lower and mid-anal canal sensation was unchanged except in the Ivalon and suture groups. However, if the patients in whom continence was restored were analysed separately, mid-anal canal sensitivity was also improved in those undergoing Marlex or resection rectopexy ( P < 005).In the upper anal canal, sensation was improved in all except the Marlex group ; similarly, those with restored continence in this group and the other three groups showed significant improvements ( P < 0.01 ). Despite these significant improvements, there was no direct correlation between improved sensation and restoration of continence.

Br. J. Surg., Vol. 79, No. 2, February1992

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo Table 4 Sensory data before and after rectopexy ~~

Suture rectopexy Preop. First rectal n.d. sensation (ml) Anal sensation (mA) Lower 8.4 (3-26) Middle 9 (6-26) 23 (5-26) Upper

Ivalon rectopexy

Marlex rectopexy

Resection rectopexy

Postop.

Preop.

Postop.

Preop.

Postop.

Preop.

Postop.

n.d.

n.d.

n.d.

65 (10-415)

50 (10-300)

53 (10-250)

35 (10-100)*

7.5 (3-26) 7.5 (3-15)* 12.5 (9-23)*

8 (3-19) 9 (3-26) 15 (3-26)

6.3 (2-11) 7.3 (3-12)* 11 (4-23)*

7.5 (2-20) 9 (2-26) 17 (3-26)

7.5 (1-14) 9 (3-22) 13 (3-26)

10 (4-25) 15 (5-26)

9 (3-26)

7 (4-14) 8 (3-15) 13 (1-24)"

Values are median (range). n.d., Not done; * P < 0.05; t P < 0.01 (Wilcoxon's paired signed rank test) Table 5

Radiographic changes after Marlex and resection rectopexy

Marlex rectopexy

Resection rectopexy

~

Preop

Postop

Anorectal angle (degrees) At rest 104 (81-123) 125 (98-138)* On contraction 91 (72-116) 105 (95-138)* During straining 117 (76-140) 137 (74-160)* Perineal descent (cm) (negative result means below pubococcygeal line) At rest -1.1 (0 to -2.2) -1.1 (0 to -2.6) On contraction -1.2 (0.2 to - 3 ) -0.2 (-3.7 to - 1.2)* During straining -4.5 (-6.6 to -1.5) -2.7 (-6.8 to -0.1) ~

Preop.

Postop

92 (89-146) 92 (83-137) 113 (97-148)

112 (93-135) 104 (95-130) 135 (89-144)

-1 (-3.9 to 1.8) -1 (-4 to 0.5) -3.2 (-9.7 to -1.5)

-1.8 (0 to -4.5) -1.8 ( - 5 to -0.2)

-4

(-6.3 to -0.3)

~

Values are median (range). * P < 0.05 (Wilcoxon's paired signed rank test)

L

3 'm A

100

1

W L

Q

c !-

20

....-....-____.-.-0 ____.-----

1

-20

!-

c ._ W

m

5 2

-

-60 -2

0

t -1

I 0

I

I

I

I

1

2

3

4

Change in continence

Figure 5 Improvement in continence correlates with improvement in maximum resting pressure. rs = 0.26, P = 0,034 (Spearman's ranked correlation test)

Proctography

Radiological correlations were unhelpful in predicting the success of the prolapse repair or the return of continence. Table5 shows a significant opening of the anorectal angle associated with Marlex rectopexy and a similar trend was seen in the resection group. O n contraction, the pelvic floor was significantly higher after Marlex rectopexy but there were no significant changes in the other measurements of pelvic descent.

Discussion Transabdominal rectopexy is the operation of choice for the control of complete rectal prolapse in patients fit enough for an abdominal procedure. There are various techniques available which, in addition to correcting the procidentia, restore continence in the majority of individuals with prolapse and incontinence'5,'6. The majority of authors use some form of prosthetic material to induce fibrosis and ensure fixation"-". There is, however, a recognized incidence of sepsis" in relation to the use of prosthetic material, which we have not to date encountered in our series. It has also been suggested that absorbable mesh may be appropriate and may

Br. J. Surg., Vol. 79, No. 2, February1992

reduce the risk of sepsis without increasing the recurrence ratez3.The argument in favour of synthetic implants is the low recurrence rate after their use and the reported higher incidence of recurrence for fixation-only procedures. Disordered bowel habit is considered to be one of the predisposing factors in the aetiology of prolapse, and was considered to have contributed to recurrences in one series 1 9 . Furthermore, since constipation frequently complicates both Ivalon3 and Marlex rectopexy, there is a case for adopting an operation which specifically aims to minimize this complication. This series describes the progression and evolution of our practice from prosthetic to non-prosthetic rectopexy and as such is not a randomized trial. There was initially a tendency to select suture rectopexy for patients considered to be at risk from a more extensive procedure, but with experience this selection has been found to be unnecessary. The precise mechanism of constipation following rectopexy is uncertain. Most previously reported studies have not carried out preoperative transit studies, so those reporting a high incidence of constipation may have contained numbers of subjects with slow transit. However, this cannot account for all the reports of postoperative constipation, and the operation may be responsible in some way. High fixation of the rectum allows the redundant sigmoid colon to prolapse into the pouch of Douglas and may create a mechanical obstruction at this site. Indeed, Ripstein has now abandoned Marlex encirclement in favour of posterior fixation ; the original Ripstein procedure was complicated by a 16.5 per cent incidence of faecal impactionz4. Despite this modification, if the sigmoid is redundant it may still be kinked at its junction with the rectum. However, Mann and Hoffman3 reported that almost 50 per cent of their patients had troublesome constipation after posterior Ivalon rectopexy. An alternative hypothesis is that the rectal mobilization divides the parasympathetic neural inflow to the left colon through the pelvic autonomic fibres, thereby altering bowel function. Resection rectopexy as advocated by Frykman and GoldbergI4 aims to prevent constipation by removing the partially denervated sigmoid and avoiding rectosigmoid distortion, which besets the fixation procedures. They also advocate low fixation, primarily to re-create the anorectal angle. Although we would not consider the angle to be of major

111

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo

importance in restoring continence”, low fixation does avoid distorting the anatomy. In addition, the Mayo Clinic group has for many years advocated high anterior resection without fixationz6. Clearly, simple sigmoid resection is not a good procedure for treating severe constipation, but as these groups did not find constipation to be a problem after operation there may be a good case for adopting routine resection. I n addition, the long-term recurrence rates are among the lowest reported’. We were initially concerned that diarrhoea might be a consequence and that this could cause incontinence, but this did not prove to be the case. The major drawback of the resection procedures is the risk of anastomotic leakage. For this reason we prefer to mobilize the splenic flexure routinely and to use well vascularized descending colon for the anastomosis. We have had no anastomotic complications using this technique. We avoid low anterior resection with its increased propensity to leakage, and consider this to be an irrational operation because it removes the compliant rectal reservoir, which we believe is an important component of the continence mechanism. Removal of the rectum and its replacement by non-compliant and often diverticular sigmoid may explain the lower continence rates after perineal rectal excision for prolapse. Certainly it may be appropriate to perform resection and rectopexy in patients with preoperative constipation, and in fact the resection may be usefully extended to good effect in those with proved delayed colonic transit’. The incidence of restored continence associated with successful prolapse repair is generally high16~’8~19,’7~2a, and a proportion of those with persistent incontinence may be helped by postanal repair”. In our experience, continence was restored in the majority of patients in all operative groups, although interestingly this was more marked in the resection and rectopexy group compared with the prosthetic operations. This seems to suggest that restoration of continence is not dependent on the induction of postoperative constipation. In fact, in all our groups, there was a trend to fewer features of constipation. Straining at stool was most reduced after Ivalon and resection rectopexies, and a feeling of incomplete evacuation improved after all rectopexies. All but the resection procedures have been considered to increase constipation3. Sayfan et a / . reported an increase in constipation after posterior Marlex rectopexy and a reduction after resection rectopexy”. It may well be that, in patients without preoperative constipation, the risks of an anastomosis have to be balanced against the risks of sepsis from prosthetic implants in coming to a decision about the type of procedure to select. We also recommend that colonic transit studies to diagnose delayed transit should be performed as a routine before all procedures, because this may affect the choice of procedure. Both anal sphincter and pelvic floor muscle impairment are well known features of prolapse3’, and there is good evidence of somatic denervation, both h i s t ~ l o g i c a l l yand ~ ~ electromyographically3’. Internal (visceral) anal sphincter function is also impaired and recovery of function has been cited as a major factor in improved c o n t i n e n ~ e ~ , ~ It’ , ~seems ~ . probable that this impaired function is related to activation of the rectoanal inhibitory reflex by the advancing prolapse. We have recorded significant improvements in resting pressures after resection and rectopexy and have seen a similar trend after suture rectopexy, which suggests that recovery of the internal anal sphincter was important in the restoration of continence. Keighley et a / . have reported a series with no evidence of sphincter recovery”, but they used a Marlex implant. Our study is in agreement with their findings in that we did not record an improvement in resting sphincter pressures in the Ivalon and Marlex groups. The reasons for these differences are unclear and it could be that the implants, by producing an intense tissue reaction, prevent recovery of internal anal sphincter function. This is an important finding, and clearly warrants further investigation. Recovery of continence is obviously complex and multifactorial and cannot be dependent only on internal sphincter recovery or we would have poorer results in the implant groups. It is possible that with larger numbers the improved physiological

112

results would demonstrate a definite advantage for avoiding implants. We could not explain recovery of continence in our implant patients on the basis of improved anal pressures ; the only measurement that improved was anal sensation, which would make patients more aware of impending leakage. It is interesting to note the correlations between the recovery of continence and initial sphincter pressures. This suggests that the preoperative pressures reflect the degree of sphincter inhibition and that patients with the lowest sphincter pressures have potentially the most to gain in terms of improvement of continence once the sphincter inhibition is removed. This finding is contrary t o that of one other but our observations seem to be confirmed by the correlation of improved continence with improvement in resting sphincter pressure. Spencer3’ suggested that rectoanal inhibition, combined with a failure to appreciate descent of the prolapse because of poor sensation, was an important aetiological factor in incontinence. Our patients show reduced rectal awareness compared with normal controls ( P < 0.01 ) and, although there was some evidence of recovery of rectal sensation in the postoperative period, this seems unlikely to be instrumental in restoring continence per se. In addition, we have recorded significantly improved sensitivity in the upper and mid-anal canal, which are the important areas for discrimination of rectal contents during the sampling reflex36. This overall improved anorectal sensation may have an important role to play in improved postoperative continence, and we believe that successful rectopexy may work by restoring normal upper anal and lower rectal anatomy, thus allowing a return to more normal function. There does not, however, seem to be any direct correlation between the recovery of continence and the improvement in sensation. Despite this we believe that the recovery of sensation and recovery of the sphincter allow a more appropriate response to any perceived threat to continence. Unlike Yoshioka et a / . 3 7 we , have found no predictive value in assessing radiological parameters : either anorectal angles or perineal descent. In fact, in our series, there was significant opening of the anorectal angle in the postoperative period, which confirms previous observations that the acuity of the anorectal angle has little to do with continencez5.We have also reported a lack of correlation between radiological findings and recovery of continence after sphincter repair3*. In conclusion, restoration of continence and correction of the prolapse is satisfactory after any of these abdominal procedures. Constipation does not seem to be the underlying reason for recovery of continence, whereas sphincter recovery and perhaps improved sensory awareness have important roles to play. For patients with evidence of preoperative constipation, resection and rectopexy seems most logical, but for all other patients a balance must be struck between the minimally higher risk of recurrence in simple suture rectopexy, the risk of sepsis and slightly poorer improvements in bowel function with the implant procedures, and the risk of anastomotic complications with resection and rectopexy. We currently favour resection and rectopexy, and reserve suture rectopexy for those at risk from a longer and more complicated procedure.

Acknowledgements Mr G. S. Duthie is supported by a grant from the Wellcome Trust. Both authors would like to acknowledge the generous assistance of Mrs Anne Mills and Ms J. C. Locke-Edmunds during this work.

References 1.

2.

Solla JA, Rothenberger DA, Goldberg S M . Colonic resection in the treatment of complete rectal prolapse. NcJrhJ Sury 1989; 41 : 132-5. Watts JD, Rothenberger DA, Buls JG, Goldberg SM, Nivatvongs S. The management of procidentia: 30 years experience. Dis Colon Rectum 1985; 28: 96-102.

Br. J. Surg., Vol. 79, No. 2, February1992

Abdominal rectopexy for rectal prolapse: G. S. Duthie and D. C. C. Bartolo

3. 4. 5.

6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

Mann CV, Hoffman C. Complete rectal prolapse: the anatomical and functional results of treatment by extended abdominal rectopexy. Br J Surg 1988; 75: 34-7. Broden G, Dolk A, Holmstrom B. Recovery of the internal anal sphincter following rectopexy : a possible explanation for continence improvement. Inr J Color Dis 1988; 3 : 23-8. Browning G G P , Parks AG. Postanal repair for neuropathic faecal incontinence : correlation of clinical result and anal canal pressures. Br J Sury 1983; 70: 101-4. Miller R, Bartolo DCC, Roe AM, Mortensen NJMcC. Assessment of microtransducers in anorectal manometry. Br J Surg 1988; 75: 40-3. Roe AM, Bartolo DCC, Mortensen NJMcC. New method for assessment of anal sensation in various anorectal disorders. Br J Surg 1986; 73: 310-12. Roe AM. Intractable idiopathic constipation : diagnosis and surgical management. M.Ch. Thesis, University of Bristol, UK, 1987. Hinton J M , Lennard-Jones JE, Young AC. A new method for studying gut transit times using radio-opaque markers. Gut 1969; 10: 842-7. Bartolo DCC, Roe AM, Virjee J, Mortensen NJMcC, Locke-Edmunds JC. An analysis of rectal morphology in obstructed defaecation. Int J Color Dis 1988; 3: 17-22. Wells C. New operation for rectal prolapse. Proc R Soc Med 1959; 52: 602-3. Penfold JCB, Hawley PR. Experience of Ivalon sponge implant for complete rectal prolapse at St. Mark’s Hospital. Br J Sury 1972; 59: 846-8. Nicholls RJ, Simson JNL. Anteroposterior rectopexy in the treatment of solitary rectal ulcer syndrome without overt rectal prolapse. Br J Surg 1986; 73: 222-4. Frykman HM, Goldberg SM. The surgical treatment of rectal procidentia. Surg Gynecol Obstet 1969 ; 129: 1225-30. Ejerblad S, Krause U . Repair of rectal prolapse by rectosacral suture fixation. Actu Chir Scand 1988; 154: 103-5. Graham W, Clegg J F , Taylor V. Complete rectal prolapse : repair by a simple technique. Ann R Coll Surg Engl 1984; 66: 87-9. Kuijpers JHC, de Morree H. Towards a selection of the most appropriate procedure in the treatment of complete rectal prolapse. Dis Colon Rectum 1988; 31 : 355-7. Anderson JR, Wilson BG, Parks TG. Complete rectal prolapse: results of Ivalon sponge rectopexy. Posrgrad Med J 1984; 60: 411- 14. Boulos PB, Stryker SJ, Nicholls RJ. The longterm results of polyvinyl alcohol (Ivalon) sponge for rectal prolapse in young patients. Br J Sury 1984; 71: 213-14. Keighley MRB, Fielding JWL, Alexander-Williams J. Results of Marlex mesh abdominal rectopexy for rectal prolapse in 1 1 consecutive patients. Br J Surg 1983; 70: 229-32.

Br. J. Surg., Vol. 79, No. 2, February 1992

21.

Notaras MJ. The use of mersilene mesh in rectal prolapse repair. Proc R Soc Med 1973; 1: 123-7. 22. Thompson JPS, Ross AHMcL. Management of infection after prosthetic abdominal rectopexy (Wells procedure). Br J Sury 1989; 76: 610-12. 23. Arndt M, Pircher W. Absorbable mesh in the treatment of rectal prolapse. Int J Color Dis 1988; 3 : 141-3. 24. Gordon PH, Hoexter B. Complications of the Ripstein procedure. Dis Colon Recturn 1978; 21 : 277-80. 25. Bartolo DCC, Roe AM, Locke-Edmunds JC, Virgee J, Mortensen NJMcC. Flap valve theory of anorectal incontinence. Br J Surg 1986; 73: 1012-14. 26. Beahers OH, Vandertoll DJ, Baker MH. Complete rectal prolapse: an evaluation of surgical treatment. Ann Surg 1965; I61 : 221. 27. Duthie GS, Bartolo DCC. A comparison between Marlex and resection rectopexy. Neth J Surg 1989; 41: 136-9. 28. Lehtola A, Salo JA, Fraki 0, Lempinen M. Treatment of rectal prolapse: a clinical study of 50 consecutive patients. Ann Chir Gynaecol 1987; 76: 150-4. 29. Sayfan J, Pinho M, Alexander-Williams J, Keighley MRB. Sutured posterior abdominal rectopexy with sigmoidectomy compared with Marlex rectopexy. Br J Surg 1990; 77: 143-5. 30. Porter NH. A physiological study of the pelvic floor in rectal prolapse. Ann R Coll Surg Engl 1962; 31 : 379-404. 31. Parks AG, Swash M, Urich H. Sphincter denervation in anorectal incontinence and rectal prolapse. Gut 1977; 18: 656-65. 32. Neil1 ME, Parks AG, Swash M. Physiological studies of the anal musculature in faecal incontinence and rectal prolapse. Br J Surg 1981; 68: 531-6. 33. Holmstrom B, Broden G , Dolk A, Freckner B. Increased anal resting pressure following the Ripstein operation : a contribution to continence? Dis Colon Recrum 1986; 29: 485-7. 34. Williams JG, Wong WD, Jensen LL, Rothenberger DA, Goldberg SM. Incontinence and rectal prolapse : a prospective manometric study. Dis Colon Rectum 1990; 33: 23. 35. Spencer RJ. Manometric studies in rectal prolapse. Dis C ( J ~ C J ~ Rectum 1984; 27: 523-5. 36. Miller R, Lewis G T , Bartolo DCC, Cevero F , Mortensen NJMcC. Sensory discrimination and dynamic activity in the anorectum: evidence using a new ambulatory technique. Br J Surg 1988; 75: 1003-7. 37. Yoshioka K, Hyland G , Keighley MRB. Anorectal function after abdominal rectopexy : parameters of predictive value in identifying return of continence. Br J Surg 1989; 76: 64-8. 38. Miller R, Orrom WJ, Cornes H, Duthie GS, Bartolo DCC. Anterior sphincter plication and levatorplasty in the treatment of faecal incontinence. Br J Sury 1989; 76: 1058-60. Paper accepted 8 October 1991

113

Abdominal rectopexy for rectal prolapse: a comparison of techniques.

To compare the methods of abdominal rectopexy and to elucidate the mechanism by which rectopexy restores continence in patients with rectal prolapse, ...
998KB Sizes 0 Downloads 0 Views