- Email: [email protected]
The preliminary study of modified Swenson procedure in Hirschsprung disease
Journal of Pediatric Surgery (2009) 44, 1560–1563
www.elsevier.com/locate/jpedsurg
The preliminary study of modified Swenson procedure in Hirschsprung disease Akiko Yokoi ⁎, Shiiki Satoh, Shigeru Takamizawa, Toshihiro Muraji, Chikara Tsugawa, Eiji Nishijima Kobe Children's Hospital, Department of Pediatric Surgery, Kobe, Hyogo 654-0081, Japan Received 30 July 2008; revised 23 November 2008; accepted 25 November 2008
Key words: The Swenson Procedure; Hirschsprung disease; Laparoscopy
Abstract Purpose: We have been using the Swenson procedure for more than 3 decades for Hirschsprung disease (HD). Recently, we modified this procedure, leaving the anterior wall below the peritoneal reflection undissected (mSwen). In 2000, we introduced mSwen with laparoscopic guidance (LapmSwen). We hypothesized that (1) omitting anterior wall dissection would not affect postoperative anorectal function, and (2) reduced dissection with better visualization via laparoscopy would reduce operative risks. Method: Charts of 89 patients with Hirschsprung disease operated on between 1990 through 2005 were retrospectively reviewed. Comparisons between Swen and mSwen, as well as between mSwen and LapmSwen, were analyzed in terms of operating time, blood loss, and complications. Results: Mean operating times (minutes) were 312 for Swen, 284 for mSwen (P = .152), and 302 for LapmSwen ( mSwen, P = .218). Mean blood loss (mL) were 64.8 for Swen, 60.3 for mSwen (P = .669), and 8.7 for LapmSwen (as compared to mSwen, P = .001). We noted leakage in 7 Swen, 2 mSwen, and no LapmSwen patients. There were no significant differences between Swen and mSwen, or between mSwen and LapmSwen, in the incidence of enterocolitis, constipation, and soiling. Conclusion: LapmSwen appeared to be comparable to the standard and modified Swenson procedures in most measures. Operative blood loss may be reduced in the LapmSwen approach. © 2009 Elsevier Inc. All rights reserved.
The Swenson procedure for Hirschsprung disease (HD) was reported in 1948, in which almost all the aganglionic bowel is removed [1]. Since then, this procedure has been considered a gold standard for HD. However, because there is concern about injury to pelvic nerves and vessels with this operative approach, others such as the Duhamel-Grob and the Soave procedures were developed [2,3]. Recently, these 3
⁎ Corresponding author. Tel.: +81 78 732 6961; fax: +81 78 735 0910. E-mail address: [email protected] (A. Yokoi). 0022-3468/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2008.11.056
major procedures were performed laparoscopically with better display of pelvic anatomy [4-6]. We have been using the Swenson procedure for 3 decades and have been quite satisfied with operative outcomes. However, dissection of the anterior wall of the rectum below the peritoneal reflection, where meticulous dissection is required to avoid injury to nerves and vessels of the urogenital system, still remained challenging. Thus, we hypothesized that (1) leaving the anterior wall of the rectum undissected (mSwen) would not affect postoperative function, and (2) reduced dissection with better
Preliminary study of modified Swenson procedure in HD
1561
Fig. 1 A and C, Swen; dissecting the anterior wall of the rectum. B and D, mSwen; leaving the anterior wall of the rectum undissected. Dissecting the posterior wall of the rectum with laparoscopic guidance in LapmSwen.
visualization provided by laparoscopy (LapmSwen) would result in a safer operation.
institution. Patients with total and extensive aganglionosis were excluded from this study. A retrospective chart review on all 89 cases was conducted. Colostomy was used on 57 patients (64.0%) before the Swenson procedure. Sixty-four patients underwent the standard Swenson procedure (Swen) from 1990 thorough 2005, 10 underwent the mSwen from 1991 through 2005, and 15 underwent the LapmSwen from
1. Materials and methods From 1990 to 2005, we had 89 patients with biopsyproven HD including 71 boys and 18 girls treated in our Table 1
Types of aganglionosis
Age (mo; mean ± SD) Aganglionic segment (n) Short Rectosigmoid Long
Swen (n = 64)
mSwen (n = 10)
P (Swen: mSwen)
LapmSwen (n = 15)
P (mSwen: LapmSwen)
10.0 ± 7.7
18.1 ± 10.7
.562 .809
2.3 ± 2.5
.003 .924
18 (28.1%) 37 (57.8%) 9 (14.1%)
2 (20%) 6 (60%) 2 (20%)
4 (26.7%) 8 (53.3%) 3 (20%)
1562 Table 2
A. Yokoi et al. Length of the anterior wall of the rectum
Length of anterior wall (mm; mean ± SD)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
15.5 ± 3.3
19.4 ± 4.6
.010
20.4 ± 3.7
.412
2000 through 2005. The median follow-up period of each group were 115 months ranging from 9 to 212 months in the Swen, 91 months ranging from 29 to 126 months in the mSwen, and 61 months ranging from 9 to 70 months in the LapmSwen. In all procedures, after pelvic dissection, the rectum was everted and incised to mobilize the proximal colon through the anus. Full-thickness biopsy was performed on the pull-through colon. In the mSwen and the LapmSwen (Fig. 1), the anterior wall of the rectum was incised above the dentate line, 1.5 to 2.0 cm in infants (as in the Swen) and more than 2.0 cm in older patients. The posterior wall was incised 0.5 cm above the dentate line (as in the Swen), creating an oblique anastomosis with several centimeters above the most proximal normal biopsy site of the pulledthrough colon. All procedures without laparoscopy were performed by one of the authors who were all trained in our institution and well experienced in Swen. LapmSwen was performed by 1 of the 4 authors (TM, SS, ST, and AY). Operating time, blood loss during operation, and incidence of postoperative complications (anastomotic leakage, enterocolitis, constipation, and soiling) were compared between those who had the Swen and the mSwen, as well as between the mSwen and the LapmSwen. Statistical analysis was performed using Mann-Whitney test and Fisher's Exact test, and values of P b .05 were considered significant. This study was approved by the institutional review board at Kobe Children's Hospital.
2. Results We assessed the incidence of urinary or sexual dysfunction through medical interview and found no detectable postoperative urinary and sexual dysfunction in this study. The mean age at operation was lower in the Swen compared to the mSwen and also was significantly lower in the LapmSwen than in the mSwen. The types of HD were not significantly different in each group (Table 1). The length of remaining aganglionic anterior wall of the rectum was significantly shorter in the Swen than in the mSwen. However, the mean length as long as 19.4 mm in mSwen and 20.4mm in LapmSwen were about the
Table 3
same as that in the original Swenson's oblique anastomosis (Table 2) [1]. The mean operating time was not significantly different either between the Swen and the mSwen or between the mSwen and the LapmSwen. On the other hand, the mean blood loss during operation was significantly less in the LapmSwen than in the mSwen, whereas it was not significantly different between the Swen and the mSwen (Table 3). No anastomotic leakage occurred in the LapmSwen, although this difference from the mSwen did not reach statistical significance. The incidence of other postoperative complications such as enterocolitis, constipation, and soiling were neither significantly different between the Swen and the mSwen nor between the mSwen and the LapmSwen (Table 4).
3. Discussion Our modification, which leaves the anterior wall of the rectum below the peritoneal reflection undissected, reduces concern about pelvic nerve and vessel injury. However, the question might be raised about whether leaving the aganglionic segment longer compromises the procedure and increases the incidence of postoperative complications. Our results indicated that this modification does not increase postoperative complications nor anorectal malfunction as compared with the original Swenson procedure. In fact, during infancy, the length of the anterior wall of the rectum below the peritoneal reflection is short enough to create an anastomosis 1.5 to 2.0 cm above the dentate line, where this is performed in the standard Swenson procedure. Recently, the transanal Soave procedure has been widely used with a shorter muscular cuff because of concerns of cuff stricture [7,8]. It is performed with only 1- to 2-cm submucosal dissection, which theoretically prevents injury to the prostate and vagina. Compared to this procedure, our technique of leaving the anterior wall of the rectum below the peritoneal reflection undissected may be simpler and easier than this approach, with the almost indistinguishable coloanal anastomosis.
Operating time and blood loss
Operating time (min; mean ± SD) Blood loss (mL; mean ± SD)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
312 ± 65.7 64.8 ± 50.8
284 ± 62.0 60.3 ± 48.5
.152 .669
302 ± 57.5 8.7 ± 6.6
.218 .001
Preliminary study of modified Swenson procedure in HD Table 4
1563
Postoperative complications
Leakage (n) Enterocolitis (n) Constipation (n) Staining (n)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
7 (10.9%) 20 (31.3%) 9 (14.1%) 11 (17.2%)
2 (20%) 2 (20%) 3 (30%) 1 (10%)
.415 .469 .204 .530
0 (0%) 5 (33.3%) 1 (6.7%) 5 (33.3%)
.071 .537 .119 .181
In the Swenson procedure, dissection of the posteriorlateral wall of the rectum down to the deep pelvic floor also requires meticulous attention, keeping the dissection directly on the rectal wall. Laparoscopic approaches have been reported for the Swenson procedure [4,9] as well as the Duhamel [6] and the Soave [5] procedures, suggesting better visualization of pelvic structures may contribute to a safe and effective operation, in addition to a cosmetic advantage. This study has shown that with laparoscopic guidance, our modified Swenson operation may have the additional safety advantage of reduced blood loss. Even for long lesions, for which mobilization of the proximal colon is necessary, pelvic dissection may be done with laparoscopic guidance first. If necessary, mobilization of the proximal colon may then be done in the open manner if less experienced surgeons are reluctant to perform this portion of the procedure laparoscopically. Several authors have described the non-Swenson oblique anastomosis not to injure the pelvic nerves and vessels. Grob described modification of Duhamel's method in 1959 [2,10]. In this procedure, side-to-side anastomosis was performed followed by resection of excess rectum at the lever of the peritoneal reflection creating shorter aganglionic rectal porch than the original Duhamel's method. However, the remaining aganglionic rectum might be longer with intraperitoneal resection than with perineal resection in Swenson's method. Wang et al [11] reported heart-shaped anastomosis, which is very akin to our modification. They described that a point of anastomosis in anterior wall of everted rectum was marked at 2 cm above the anal verge and 0.5 cm above the dentate line in posterior wall, and then, at completion of the procedure, the anterior anastomosis was 4 cm above the anal verge and the posterior anastomosis was 2cm above the anodermal junction. The mean age in their study was 25 months. In our study, the mean age was 18.1 months in mSwen and 2.3 months in LapmSwen. For small infants, eversion of the rectum was easy to leave the aganglionic rectum as short as about 2 cm without dissecting the anterior wall of the rectum below the peritoneal reflection. Therefore, we created oblique anastomosis with our modification almost indistinguishable from the original Swenson's method. Paul et al also described an oblique anastomosis with Soave endoanal pull-through [12]. They started mucosectomy 1.5 cm anteriorly and 0.5 cm posteriorly above the dentate line. They did not mention how far this mucosectomy went proximally but stated just until the appropriate level. We suppose they created regular length of muscle cuff, which might cause cuff stricture or cuff abscess, although they performed oblique anastomosis. In our
modified Swenson operation, we do not create muscle cuff or aganglionic pouch so that we could preserve the concept of original Swenson's procedure, almost complete removal of aganglinonic bowels. Although long-term follow-up may be necessary to compare the incidence of enterocolitis and anorectal function, our data suggest that our modifications to the Swenson procedure (leaving the anterior wall of the rectum below the peritoneal reflection undissected while dissecting the posterolateral rectum to the deep pelvis with laparoscopic guidance) may eliminate a potential disadvantage of this validated approach to HD.
References [1] Swenson O, Bill AH. Resection of rectum and recto sigmoid with preservation of the sphincter for benign spastic lesions producing megacolon. Surgery 1948;24:212-20. [2] Dorman GW, Votteler TP, Graivier L. A preliminary evaluation of the results of treatment of Hirschsprung's disease by the Duhamel-Grob modification of the Swenson pull-through operation. Ann Surg 1967; 166:783-91. [3] Swenson O. Hirschsprung's disease—a complicated therapeutic problem: some thoughts and solutions based on data and personal experience over 56 years. J Pediatr Surg 2004;39:1449-53 [discussion 1454-1447]. [4] Kumar R, Mackay A, Borzi P. Laparoscopic Swenson procedure—an optimal approach for both primary and secondary pull-through for Hirschsprung's disease. J Pediatr Surg 2003;38:1440-3. [5] Georgeson KE, Cohen RD, Hebra A, et al. Primary laparoscopicassisted endorectal colon pull-through for Hirschsprung's disease: a new gold standard. Ann Surg 1999;229:678-82 [discussion 682-673]. [6] Travassos DV, Bax NM, Van der Zee DC. Duhamel procedure: a comparative retrospective study between an open and a laparoscopic technique. Surg Endosc 2007;21:2163-5. [7] Nasr A, Langer JC. Evolution of the technique in the transanal pullthrough for Hirschsprung's disease: effect on outcome. J Pediatr Surg 2007;42:36-9 [discussion 39-40]. [8] Langer JC, Durrant AC, de la Torre L, et al. One-stage transanal Soave pullthrough for Hirschsprung disease: a multicenter experience with 141 children. Ann Surg 2003;238:569-83 [discussion 583-565]. [9] Curran TJ, Raffensperger JG. Laparoscopic Swenson pull-through: a comparison with the open procedure. J Pediatr Surg 1996;31:1155-6 [discussion 1156-1157]. [10] Grob M. Intestinal obstruction in the newborn infant. Arch Dis Child 1960;35:40-50. [11] Wang G, Sun XY, Wei MF, et al. Heart-shaped anastomosis for Hirschsprung's disease: operative technique and long-term follow-up. World J Gastroenterol 2005;11:296-8. [12] Paul A, Fraser N, Chhabra S, et al. Oblique anastomosis in Soave endoanal pullthrough for Hirschsprung's disease—a way of reducing strictures? Pediatr Surg Int 2007;23:1187-90.
www.elsevier.com/locate/jpedsurg
The preliminary study of modified Swenson procedure in Hirschsprung disease Akiko Yokoi ⁎, Shiiki Satoh, Shigeru Takamizawa, Toshihiro Muraji, Chikara Tsugawa, Eiji Nishijima Kobe Children's Hospital, Department of Pediatric Surgery, Kobe, Hyogo 654-0081, Japan Received 30 July 2008; revised 23 November 2008; accepted 25 November 2008
Key words: The Swenson Procedure; Hirschsprung disease; Laparoscopy
Abstract Purpose: We have been using the Swenson procedure for more than 3 decades for Hirschsprung disease (HD). Recently, we modified this procedure, leaving the anterior wall below the peritoneal reflection undissected (mSwen). In 2000, we introduced mSwen with laparoscopic guidance (LapmSwen). We hypothesized that (1) omitting anterior wall dissection would not affect postoperative anorectal function, and (2) reduced dissection with better visualization via laparoscopy would reduce operative risks. Method: Charts of 89 patients with Hirschsprung disease operated on between 1990 through 2005 were retrospectively reviewed. Comparisons between Swen and mSwen, as well as between mSwen and LapmSwen, were analyzed in terms of operating time, blood loss, and complications. Results: Mean operating times (minutes) were 312 for Swen, 284 for mSwen (P = .152), and 302 for LapmSwen ( mSwen, P = .218). Mean blood loss (mL) were 64.8 for Swen, 60.3 for mSwen (P = .669), and 8.7 for LapmSwen (as compared to mSwen, P = .001). We noted leakage in 7 Swen, 2 mSwen, and no LapmSwen patients. There were no significant differences between Swen and mSwen, or between mSwen and LapmSwen, in the incidence of enterocolitis, constipation, and soiling. Conclusion: LapmSwen appeared to be comparable to the standard and modified Swenson procedures in most measures. Operative blood loss may be reduced in the LapmSwen approach. © 2009 Elsevier Inc. All rights reserved.
The Swenson procedure for Hirschsprung disease (HD) was reported in 1948, in which almost all the aganglionic bowel is removed [1]. Since then, this procedure has been considered a gold standard for HD. However, because there is concern about injury to pelvic nerves and vessels with this operative approach, others such as the Duhamel-Grob and the Soave procedures were developed [2,3]. Recently, these 3
⁎ Corresponding author. Tel.: +81 78 732 6961; fax: +81 78 735 0910. E-mail address: [email protected] (A. Yokoi). 0022-3468/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2008.11.056
major procedures were performed laparoscopically with better display of pelvic anatomy [4-6]. We have been using the Swenson procedure for 3 decades and have been quite satisfied with operative outcomes. However, dissection of the anterior wall of the rectum below the peritoneal reflection, where meticulous dissection is required to avoid injury to nerves and vessels of the urogenital system, still remained challenging. Thus, we hypothesized that (1) leaving the anterior wall of the rectum undissected (mSwen) would not affect postoperative function, and (2) reduced dissection with better
Preliminary study of modified Swenson procedure in HD
1561
Fig. 1 A and C, Swen; dissecting the anterior wall of the rectum. B and D, mSwen; leaving the anterior wall of the rectum undissected. Dissecting the posterior wall of the rectum with laparoscopic guidance in LapmSwen.
visualization provided by laparoscopy (LapmSwen) would result in a safer operation.
institution. Patients with total and extensive aganglionosis were excluded from this study. A retrospective chart review on all 89 cases was conducted. Colostomy was used on 57 patients (64.0%) before the Swenson procedure. Sixty-four patients underwent the standard Swenson procedure (Swen) from 1990 thorough 2005, 10 underwent the mSwen from 1991 through 2005, and 15 underwent the LapmSwen from
1. Materials and methods From 1990 to 2005, we had 89 patients with biopsyproven HD including 71 boys and 18 girls treated in our Table 1
Types of aganglionosis
Age (mo; mean ± SD) Aganglionic segment (n) Short Rectosigmoid Long
Swen (n = 64)
mSwen (n = 10)
P (Swen: mSwen)
LapmSwen (n = 15)
P (mSwen: LapmSwen)
10.0 ± 7.7
18.1 ± 10.7
.562 .809
2.3 ± 2.5
.003 .924
18 (28.1%) 37 (57.8%) 9 (14.1%)
2 (20%) 6 (60%) 2 (20%)
4 (26.7%) 8 (53.3%) 3 (20%)
1562 Table 2
A. Yokoi et al. Length of the anterior wall of the rectum
Length of anterior wall (mm; mean ± SD)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
15.5 ± 3.3
19.4 ± 4.6
.010
20.4 ± 3.7
.412
2000 through 2005. The median follow-up period of each group were 115 months ranging from 9 to 212 months in the Swen, 91 months ranging from 29 to 126 months in the mSwen, and 61 months ranging from 9 to 70 months in the LapmSwen. In all procedures, after pelvic dissection, the rectum was everted and incised to mobilize the proximal colon through the anus. Full-thickness biopsy was performed on the pull-through colon. In the mSwen and the LapmSwen (Fig. 1), the anterior wall of the rectum was incised above the dentate line, 1.5 to 2.0 cm in infants (as in the Swen) and more than 2.0 cm in older patients. The posterior wall was incised 0.5 cm above the dentate line (as in the Swen), creating an oblique anastomosis with several centimeters above the most proximal normal biopsy site of the pulledthrough colon. All procedures without laparoscopy were performed by one of the authors who were all trained in our institution and well experienced in Swen. LapmSwen was performed by 1 of the 4 authors (TM, SS, ST, and AY). Operating time, blood loss during operation, and incidence of postoperative complications (anastomotic leakage, enterocolitis, constipation, and soiling) were compared between those who had the Swen and the mSwen, as well as between the mSwen and the LapmSwen. Statistical analysis was performed using Mann-Whitney test and Fisher's Exact test, and values of P b .05 were considered significant. This study was approved by the institutional review board at Kobe Children's Hospital.
2. Results We assessed the incidence of urinary or sexual dysfunction through medical interview and found no detectable postoperative urinary and sexual dysfunction in this study. The mean age at operation was lower in the Swen compared to the mSwen and also was significantly lower in the LapmSwen than in the mSwen. The types of HD were not significantly different in each group (Table 1). The length of remaining aganglionic anterior wall of the rectum was significantly shorter in the Swen than in the mSwen. However, the mean length as long as 19.4 mm in mSwen and 20.4mm in LapmSwen were about the
Table 3
same as that in the original Swenson's oblique anastomosis (Table 2) [1]. The mean operating time was not significantly different either between the Swen and the mSwen or between the mSwen and the LapmSwen. On the other hand, the mean blood loss during operation was significantly less in the LapmSwen than in the mSwen, whereas it was not significantly different between the Swen and the mSwen (Table 3). No anastomotic leakage occurred in the LapmSwen, although this difference from the mSwen did not reach statistical significance. The incidence of other postoperative complications such as enterocolitis, constipation, and soiling were neither significantly different between the Swen and the mSwen nor between the mSwen and the LapmSwen (Table 4).
3. Discussion Our modification, which leaves the anterior wall of the rectum below the peritoneal reflection undissected, reduces concern about pelvic nerve and vessel injury. However, the question might be raised about whether leaving the aganglionic segment longer compromises the procedure and increases the incidence of postoperative complications. Our results indicated that this modification does not increase postoperative complications nor anorectal malfunction as compared with the original Swenson procedure. In fact, during infancy, the length of the anterior wall of the rectum below the peritoneal reflection is short enough to create an anastomosis 1.5 to 2.0 cm above the dentate line, where this is performed in the standard Swenson procedure. Recently, the transanal Soave procedure has been widely used with a shorter muscular cuff because of concerns of cuff stricture [7,8]. It is performed with only 1- to 2-cm submucosal dissection, which theoretically prevents injury to the prostate and vagina. Compared to this procedure, our technique of leaving the anterior wall of the rectum below the peritoneal reflection undissected may be simpler and easier than this approach, with the almost indistinguishable coloanal anastomosis.
Operating time and blood loss
Operating time (min; mean ± SD) Blood loss (mL; mean ± SD)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
312 ± 65.7 64.8 ± 50.8
284 ± 62.0 60.3 ± 48.5
.152 .669
302 ± 57.5 8.7 ± 6.6
.218 .001
Preliminary study of modified Swenson procedure in HD Table 4
1563
Postoperative complications
Leakage (n) Enterocolitis (n) Constipation (n) Staining (n)
Swen
mSwen
P (Swen: mSwen)
LapmSwen
P (mSwen: LapmSwen)
7 (10.9%) 20 (31.3%) 9 (14.1%) 11 (17.2%)
2 (20%) 2 (20%) 3 (30%) 1 (10%)
.415 .469 .204 .530
0 (0%) 5 (33.3%) 1 (6.7%) 5 (33.3%)
.071 .537 .119 .181
In the Swenson procedure, dissection of the posteriorlateral wall of the rectum down to the deep pelvic floor also requires meticulous attention, keeping the dissection directly on the rectal wall. Laparoscopic approaches have been reported for the Swenson procedure [4,9] as well as the Duhamel [6] and the Soave [5] procedures, suggesting better visualization of pelvic structures may contribute to a safe and effective operation, in addition to a cosmetic advantage. This study has shown that with laparoscopic guidance, our modified Swenson operation may have the additional safety advantage of reduced blood loss. Even for long lesions, for which mobilization of the proximal colon is necessary, pelvic dissection may be done with laparoscopic guidance first. If necessary, mobilization of the proximal colon may then be done in the open manner if less experienced surgeons are reluctant to perform this portion of the procedure laparoscopically. Several authors have described the non-Swenson oblique anastomosis not to injure the pelvic nerves and vessels. Grob described modification of Duhamel's method in 1959 [2,10]. In this procedure, side-to-side anastomosis was performed followed by resection of excess rectum at the lever of the peritoneal reflection creating shorter aganglionic rectal porch than the original Duhamel's method. However, the remaining aganglionic rectum might be longer with intraperitoneal resection than with perineal resection in Swenson's method. Wang et al [11] reported heart-shaped anastomosis, which is very akin to our modification. They described that a point of anastomosis in anterior wall of everted rectum was marked at 2 cm above the anal verge and 0.5 cm above the dentate line in posterior wall, and then, at completion of the procedure, the anterior anastomosis was 4 cm above the anal verge and the posterior anastomosis was 2cm above the anodermal junction. The mean age in their study was 25 months. In our study, the mean age was 18.1 months in mSwen and 2.3 months in LapmSwen. For small infants, eversion of the rectum was easy to leave the aganglionic rectum as short as about 2 cm without dissecting the anterior wall of the rectum below the peritoneal reflection. Therefore, we created oblique anastomosis with our modification almost indistinguishable from the original Swenson's method. Paul et al also described an oblique anastomosis with Soave endoanal pull-through [12]. They started mucosectomy 1.5 cm anteriorly and 0.5 cm posteriorly above the dentate line. They did not mention how far this mucosectomy went proximally but stated just until the appropriate level. We suppose they created regular length of muscle cuff, which might cause cuff stricture or cuff abscess, although they performed oblique anastomosis. In our
modified Swenson operation, we do not create muscle cuff or aganglionic pouch so that we could preserve the concept of original Swenson's procedure, almost complete removal of aganglinonic bowels. Although long-term follow-up may be necessary to compare the incidence of enterocolitis and anorectal function, our data suggest that our modifications to the Swenson procedure (leaving the anterior wall of the rectum below the peritoneal reflection undissected while dissecting the posterolateral rectum to the deep pelvis with laparoscopic guidance) may eliminate a potential disadvantage of this validated approach to HD.
References [1] Swenson O, Bill AH. Resection of rectum and recto sigmoid with preservation of the sphincter for benign spastic lesions producing megacolon. Surgery 1948;24:212-20. [2] Dorman GW, Votteler TP, Graivier L. A preliminary evaluation of the results of treatment of Hirschsprung's disease by the Duhamel-Grob modification of the Swenson pull-through operation. Ann Surg 1967; 166:783-91. [3] Swenson O. Hirschsprung's disease—a complicated therapeutic problem: some thoughts and solutions based on data and personal experience over 56 years. J Pediatr Surg 2004;39:1449-53 [discussion 1454-1447]. [4] Kumar R, Mackay A, Borzi P. Laparoscopic Swenson procedure—an optimal approach for both primary and secondary pull-through for Hirschsprung's disease. J Pediatr Surg 2003;38:1440-3. [5] Georgeson KE, Cohen RD, Hebra A, et al. Primary laparoscopicassisted endorectal colon pull-through for Hirschsprung's disease: a new gold standard. Ann Surg 1999;229:678-82 [discussion 682-673]. [6] Travassos DV, Bax NM, Van der Zee DC. Duhamel procedure: a comparative retrospective study between an open and a laparoscopic technique. Surg Endosc 2007;21:2163-5. [7] Nasr A, Langer JC. Evolution of the technique in the transanal pullthrough for Hirschsprung's disease: effect on outcome. J Pediatr Surg 2007;42:36-9 [discussion 39-40]. [8] Langer JC, Durrant AC, de la Torre L, et al. One-stage transanal Soave pullthrough for Hirschsprung disease: a multicenter experience with 141 children. Ann Surg 2003;238:569-83 [discussion 583-565]. [9] Curran TJ, Raffensperger JG. Laparoscopic Swenson pull-through: a comparison with the open procedure. J Pediatr Surg 1996;31:1155-6 [discussion 1156-1157]. [10] Grob M. Intestinal obstruction in the newborn infant. Arch Dis Child 1960;35:40-50. [11] Wang G, Sun XY, Wei MF, et al. Heart-shaped anastomosis for Hirschsprung's disease: operative technique and long-term follow-up. World J Gastroenterol 2005;11:296-8. [12] Paul A, Fraser N, Chhabra S, et al. Oblique anastomosis in Soave endoanal pullthrough for Hirschsprung's disease—a way of reducing strictures? Pediatr Surg Int 2007;23:1187-90.