- Email: [email protected]
intermediate-type imperforate anus: prospective comparative study between laparoscopy-assisted and posterior sagittal anorectoplasty
Journal of Pediatric Surgery (2008) 43, 158–163
www.elsevier.com/locate/jpedsurg
Midterm postoperative clinicoradiologic analysis of surgery for high/intermediate-type imperforate anus: prospective comparative study between laparoscopy-assisted and posterior sagittal anorectoplasty Chizue Ichijo a , Kazuhiro Kaneyama a , Yutaka Hayashi a , Hiroyuki Koga a , Tadaharu Okazaki a , Geoffrey J. Lane a , Yoshihisa Kurosaki b , Atsuyuki Yamataka a,⁎ a
Department of Pediatric General and Urogenital Surgery, Juntendo University School of Medicine, Tokyo 113-8421, Japan Department of Radiology, Juntendo University School of Medicine, Tokyo 113-8421, Japan
b
Received 29 August 2007; accepted 2 September 2007
Index words: High-type imperforate anus; Intermediate-type imperforate anus; Anorectoplasty; Anal endosonography; Magnetic resonance imaging; Continence evaluation questionnaire
Abstract Purpose: The objective of this study is to analyze the outcome of surgery for high/intermediate-type imperforate anus using anal endosonography (AES), magnetic resonance imaging (MRI), and a continence evaluation questionnaire (CEQ). Methods: In this study, 24 cases of high/intermediate-type imperforate anus were studied. Fifteen of 24 had Georgeson's laparoscopy–assisted anorectoplasty (GLA), and 9 of 24 had Pena's posterior sagittal anorectoplasty (PPA). All subjects had AES and MRI postoperatively. On AES, differences in thickness of the external sphincter (ES) and puborectalis (PR) at 3 and 9 o'clock were compared, and if pull-through colon was central, AES was 0. On MRI, differences in thickness were analyzed semiquantitatively and scored; if muscles were of even thickness, the score was 0; slight difference was 1, and marked difference was 2. A 5-parameter CEQ questionnaire (maximum score, 10) was administered to 16 of 24 subjects followed up for more than 3 years (9 GLA, 7 PPA). Surgical stress was assessed using mean febrile period, duration of raised white blood cell count, and peak C-reactive protein level. Results: Mean age at surgery and mean postoperative period for both groups were not statistically different. There were no differences in mean muscle thickness for ES or PR on AES according to procedure (ES: GLA = 0.19 ± 0.15 mm, PPA = 0.16 ± 0.09 mm, P = .59; PR: GLA = 0.19 ± 0.19 mm, PPA = 0.22 ± 0.15 mm, P = .69). Magnetic resonance imaging scores were also not significantly different according to procedure (GLA: 0.77 ± 0.83; PPA: 0.75 ± 0.50, P = .97). When CEQ were compared annually, scores for GLA were generally higher throughout the study but only statistically significant at 3 and 4 years (P b .05). Differences in parameters of surgical stress were not significant. Conclusion: Although there were no differences in muscle thickness and parameters of surgical stress observed according to technique, GLA would appear to provide better outcome based on CEQ scores. © 2008 Published by Elsevier Inc.
Presented at the 38th annual meeting of the American Pediatric Surgical Association, Orlando, Florida, May 24-27, 2007. ⁎ Corresponding author. Tel.: +81 3 3813 3111; fax: +81 3 5802 2033. E-mail address: [email protected] (A. Yamataka). 0022-3468/$ – see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.jpedsurg.2007.09.037
High/intermediate-type imperforate anus and postoperative outcome In recent reports [1-3] about short- to midterm outcome after Georgeson's laparoscopy–assisted anorectoplasty (GLA) for high/intermediate-type imperforate anus (HIIA) [4], GLA appeared to be as good as or even better than Pena's posterior sagittal anorectoplasty (PPA). However, there are no prospective studies comparing GLA and PPA in the literature, and past reports about continence evaluation have usually been retrospective. At our institute, we have been assessing our patients with HIIA who underwent surgery (GLA or PPA) after 1999 prospectively with a structured continence evaluation questionnaire (CEQ) administered at follow-up outpatient clinic attendance. Because of this, we were prompted to compare midterm outcome of GLA and PPA on an annual basis as well as use imaging technology (anal endosonography [AES] and magnetic resonance) to obtain information on postoperative anorectal position that might influence continence and correlate position data with outcome to compare surgical techniques.
159
1.2. Anal endosonography and pelvic magnetic resonance imaging Both AES and magnetic resonance imaging (MRI) were performed between 6 months and 7 years postoperatively in all 24 subjects to evaluate anorectal anatomy. For AES, we used rotating ultrasonographic scanners to examine the muscles of the pelvic floor using 360° crosssectional imaging: RU-75M-R1 (7.5 MHz) and RU-12M-R1 (12 MHz) (OLYMPUS, Tokyo, Japan); penetration depth, 2
1. Materials and methods 1.1. Patient demographics and surgical techniques From January 1999 to June 2006, we treated 28 patients with HIIA. All 28 were treated initially with colostomy during the neonatal period. Consent to perform general anesthesia for AES could not be obtained for 4 of these (2 GLA, 2 PPA), and they were excluded, leaving 24 patients as subjects for this study. Of the 24, 15 had GLA and 9 had PPA. High/intermediate-type imperforate anus cases treated by GLA were 5 males with rectoprostatic urethral fistula, 4 males with rectobulbar urethral fistula, 1 male with anorectal agenesis without fistula, 1 female with rectovaginal fistula, 2 females with rectovestibular fistula with absent vagina, and 2 females with cloacal anomaly. High/intermediate-type imperforate anus cases treated by PPA were 2 males with rectovesical fistula, 4 males with rectobulbar urethral fistula, 1 male with anorectal agenesis without fistula, and 2 females with rectovestibular fistula. All HIIA surgery (GLA or PPA) was performed under the direct supervision of a single surgeon (AY). Conventional techniques as described elsewhere by Georgeson et al [4,5] were used to perform GLA. Posterior sagittal procedure was performed as previously described by de Vries and Pena [6]. In this study, there were 9 PPA cases; 3 before GLA was introduced, and 6 after GLA was introduced. After GLA was introduced, 3 of the 6 cases who had PPA had a history of other abdominal surgery for malrotation and duodenal atresia. The remaining 3 cases (1 case of rectoprostatic urethral fistula and 2 cases of rectobulbar urethral fistula) had PPA because we experienced a complication secondary to residual fistula during GLA. We have since refined our GLA procedure to include intraoperative colonoscopy and cystoscopy to deal with such cases.
Fig. 1 Postoperative AES of the PR (arrows). A, The PR is symmetric, 2.6 mm at 3 o'clock and 2.5 mm at 9 o'clock, and the difference of thickness is 0.1 mm. B, The PR is asymmetric, 2.1 mm at 3 o'clock and 1.4 mm at 9 o'clock, and the difference is 0.7 mm.
160
C. Ichijo et al.
1.3. Assessing continence and surgical stress Of the 24 subjects, 16 (9 GLA, 7 PPA) who had been followed up for more than 3 years and had no chromosomal anomalies were evaluated using a structured CEQ by a single experienced surgeon (AY). Table 1 shows the CEQ we used. Five parameters were scored on a scale of 0 to 2, giving a maximum score of 10. In this study, we used “staining” to refer to fecal markings on underclothing and “soiling” to mean actual feces on underclothing. Annual CEQ scores were determined and compared between the 2 groups. To assess surgical stress, we compared the presence of fever, raised white blood cell (WBC) count (N10,000 μL), and peak serum C-reactive protein (CRP) level (N0.3 mg/dL) between the 2 groups.
1.4. Statistical analysis and ethics Data were expressed as mean ± SD. Differences in muscle thickness measured using AES, semiquantitative muscle thickness scores calculated using MRI, annual CEQ scores, and parameters of surgical stress were compared and analyzed using the unpaired Student's t test. A P value of less than .05 was considered to be statistically significant.
Fig. 2 Postoperative MRI of the PR (arrows). A, Obvious muscle masses at both 3 and 9 o'clock, indicating that MRI score is 0 (symmetric). B, Obvious muscle mass at 9 o'clock and poor muscle mass at 3 o'clock, indicating that MRI score is 2 (asymmetric).
to 5 cm; 12 mm in diameter. The endoprobe had a hard sonolucent plastic cone with a balloon at the tip to allow the probe to fit snugly within the pull-through canal once instilled with normal saline. Thicknesses of the external sphincter (ES) and puborectalis (PR) at 3 and 9 o'clock were measured, and differences in muscle thickness were compared between the 2 groups. If the pull-through colon was central, the difference in muscle thickness was 0 (Fig. 1). Magnetic resonance imaging was performed using a 1.5-T scanner (VISART v4.10; Toshiba, Japan) to obtain perpendicular axial planes for contiguous 4- to 7-mm slices along the anal canal parallel to the pubococcygeal line. Magnetic resonance imaging images at the level of the ES and PR were reviewed separately and blindly by a well-experienced radiologist (YK). Muscle thickness was classified as good, fair, or poor, according to whether muscle mass was obvious, thin, or virtually invisible. Differences at 3 and 9 o'clock were analyzed semiquantitatively for symmetry. Scoring was determined as follows: if muscles seemed to be symmetric (ie, thickness was good-good), the score was 0. Slight difference in thickness (ie, good-fair or fair-poor) scored 1 point, and marked difference (ie, good-poor) scored 2 points. The worst score was 2 for each muscle, that is, 4 (ES = 2 + PR = 2) (Fig. 2).
Table 1
Continence evaluation questionnaire score Score
1. Frequency of defecation 1-2/d 3-5/d ≥6/d
2 1 0
2. Staining/soiling None Occasional staining Staining Staining always Soiling
2 1.5 1 0.5 0
3. Perianal erosion Nil Occasionally Often
2 1 0
4. Anal shape Normal looking Scar or skin tag visible Mucosa visible Prolapse needing surgery
2 1.5 1 0
5. Medication Nil Laxative/enema/suppository Antidiarrheals needed Maximum score
2 1 0 10
High/intermediate-type imperforate anus and postoperative outcome This study was approved by the Juntendo University School of Medicine Ethics Committee and complies with the Helsinki Declaration of 1975 (revised 1983).
2. Results Of the 15 GLA subjects, 10 were boys and 5 were girls, and of the 9 PPA subjects, 7 were boys and 2 were girls. The mean age at surgery was 9.6 ± 7.1 months for GLA and 7.1 ± 4.8 months for PPA (P = .45), and the mean length of followup after surgery was 60.0 ± 15.1 months for GLA and 66.3 ± 24.0 months for PPA (P = .53). On AES, the mean difference in ES thicknesses was 0.19 ± 0.15 mm for GLA and 0.16 ± 0.09 mm for PPA (P = .59). Also, the mean difference in PR thicknesses was not statistically significant between the 2 groups (0.19 ± 0.19 mm for GLA, 0.22 ± 0.15 mm for PPA, P = .69). The mean semiquantitative MRI scores were not significantly different (GLA, 0.77 ± 0.83; PPA, 0.75 ± 0.50, P = .97), indicating there were no differences in muscle thicknesses according to surgical technique. The maximum duration of follow-up in the GLA group was 6 years—but only for 2 subjects, so direct comparison between the 2 groups was only possible for the first 5 years after surgery. The mean annual CEQ scores for the first 5 years after GLA were 6.1 ± 1.6 (n = 9), 7.2 ± 1.3 (n = 9), 8.1 ± 0.8 (n = 9), 8.2 ± 0.7 (n = 8), and 8.2 ± 1.1 (n = 5), and for the first 5 years after PPA, the scores were 5.1 ± 1.9 (n = 7), 5.9 ± 1.8 (n = 7), 6.3 ± 1.6 (n = 7), 6.2 ± 1.9 (n = 5), and 6.5 ± 2.3 (n = 4) (Fig. 3). Georgeson's laparoscopy–assisted anorectoplasty had generally higher CEQ scores throughout the study, but differences were only statistically significant at 3 and 4 years after surgery (P b .05).
Fig. 3 Continence evaluation questionnaire scores after GLA and PPA. Scores for GLA are generally higher throughout the study but only statistically significant at 3 and 4 years (*P b .05 in both).
161
The mean peak WBC and duration of raised WBC for GLA were 14,720 ± 3940/μL, 5.3 ± 3.3 days, and for PPA were 11,730 ± 4200/μL, 4.1 ± 7.1 days, respectively (P = .08 and .59). The mean peak CRP and duration of raised CRP for GLA were 8.19 ± 6.33 mg/dL, 10.5 ± 5.4 days, and for PPA were 3.92 ± 3.36 mg/dL, 6.5 ± 5.9 days, respectively (P = .07 and .16). The mean duration of fever was 0.9 ± 2.1 days for GLA and 1.1 ± 2.2 days for PPA (P = .87), indicating that all parameters for surgical stress were not statistically different between the 2 groups. Postoperative complications occurred in 2 GLA patients with rectobulbar urethral fistula (one had an enlarged residual fistula [7] and the other had temporary dysuria) and 3 PPA patients (minor wound infection in all).
3. Discussion In this study, we applied AES to evaluate postoperative anorectal anatomy accurately. Essentially, our aim was to compare muscle thickness along the pull-through route for symmetry rather than for actual thickness. Thus, we did not focus too much on differences in muscle thickness between causes of HIIA, but generally, lower-type disorders such as rectobulbar fistula appear to have thicker muscles than higher-type disorders such as cloaca and rectovesical fistula. A recently published report about AES in a pediatric population by Jones et al [8] investigated the value of AES compared with MRI after the repair of anorectal malformations and recommended that AES was an accurate alternative to MRI in the assessment of anorectoplasty because AES had a better correlation with results of perineal muscle stimulation (MRI failed in some cases). In our study, we performed both AES and MRI for thoroughness and confirmed that AES was accurate for measuring the thicknesses of ES and PR in all cases. However, MRI alone was not reliable for measuring thickness because only the thickness on a particular image can be measured. Thus, MRI was only used for making a semiquantitative assessment because subjective factors would appear to influence the interpretation of MRI images. Using our AES results, we found that GLA was just as accurate as PPA for positioning the colon accurately along the pull-through route; in other words, pelvic floor muscles were found to surround the pull-through route nicely after GLA as they do after PPA in most of cases. Thus, we suggest that the main reason for the better continence achieved after GLA might be related to the less impact on the pelvic nerve system and levator muscles during GLA, allowing post-GLA patients to achieve better continence at an earlier period (especially 3-4 years postoperatively). Iwanaka et al [9] also recommended GLA as an alternative procedure because PPA was considered to cause damage to the sphincter muscles as well as the tiny nerves that maintain anorectal sensation and motility as a consequence of the large incision used in the pelvis, a conclusion that supports our findings.
162 Our CEQ results showed that GLA subjects had higher scores than PPA subjects, suggesting that GLA subjects do achieve better midterm postoperative fecal continence. Specifically, frequency of motions, staining/soiling, and incidence of erosions were better in GLA subjects, and the difference between GLA and PPA for staining/soiling was statistically significant 4 years after surgery. However, in the literature, there are only 2 reports comparing midterm outcome of GLA and PPA [1,2]. Kudou et al [1] found that both GLA and PPA patients had similar bowel habits. In their study, age at the time of evaluation of fecal continence was significantly lower in the GLA group (51 months for GLA vs 73 months for PPA), and their results appear not to be age matched, suggesting that anorectal function in the GLA group would probably improve in the long term. Wong et al [2] also compared fecal continence according to procedure (GLA vs PPA) and reported that significantly more patients in the GLA group had acceptable defecation status. However, they investigated only frequency of motions as an indicator of fecal continence, which we believe is insufficient because fecal continence is multifactorial. Thus, we assessed severity of staining/soiling and requirement for medications as conventional factors but added presence of perianal erosions and anal shape to provide a complete picture because, if there is anal mucosa prolapse, there is increased irritation/stimulation of the anus, which can be the cause of unstable bowel control. We found no statistical difference for anal shape or perianal erosions according to technique throughout the period of our study. In addition, the median duration of follow-up in their study (as well as Kudou et al) was not matched (27 months for GLA and 52 months for PPA). Thus, although our study is the first to compare age/sex-matched patients with regard to postoperative fecal continence according to procedure, there is a risk for bias because of small patient numbers; for example, there were 2 cases of rectovesical fistula in the PPA group that could lower CEQ scores, and increasing patient numbers will help diminish this risk. One challenging issue revealed by this study is the role of GLA for patients with rectobulbar urethral fistula, because the 2 subjects with complications both had rectobulbar urethral fistula type. Briefly, one had a large cystic mass behind the bladder, suggestive of a posterior urethral diverticulum, and we diagnosed it as an enlarged residual fistula at laparotomy. With this in mind, we began to perform intraoperative colonoscopy and cystoscopy during GLA to evaluate the size of the fistula lumen and facilitate excision of the fistula as completely as possible [7]. However, we had one patient develop temporary dysuria after intraoperative colonoscopy and cystoscopy, probably because of surgical manipulation of the posterior urethral nerve plexus and/or as a result of indirect thermal injury to the posterior urethral nerve plexus when monopolar diathermy was used. Thus, for rectobulbar urethral fistula, dissection must be meticulous and careful during GLA, and in fact, we use PPA in such cases to decrease physical impact on pelvic nerves.
C. Ichijo et al. In conclusion, for the treatment of HIIA (except rectobulbar urethral fistula), GLA appears to have a better postoperative outcome than PPA, based on CEQ results, with no difference in the amount of surgical stress inflicted. However, the small but statistically significant difference in outcome may not have any great clinical application at the present time because it is unlikely to guide surgeons in planning their management. Thus, although both procedures would appear to be equivalent in terms of muscle thickness and surgical stress, GLA may appear to have a small advantage over PPA functionally. Further follow-up research is currently being performed to increase the number of patients and assess long-term outcome.
References [1] Kudou S, Iwanaka T, Kawashima H, et al. Midterm follow-up study of high-type imperforate anus after laparoscopically assisted anorectoplasty. J Pediatr Surg 2005;40:1923-6. [2] Wong KKY, Khong PL, Lin SCL, et al. Post-operative magnetic resonance evaluation of children after laparoscopic anorectoplasty for imperforate anus. Int J Colorectal Dis 2004;20:33-7. [3] Lin CL, Wong KKY, Lan LCL, et al. Earlier appearance and higher incidence of the rectoanal relaxation reflex in patients with imperforate anus repaired with laparoscopically assisted anorectoplasty. Surg Endosc 2003;17:1646-9. [4] Georgeson KE, Inge TH, Albanese CT. Laparoscopically assisted anorectal pull-through for high imperforate anus—a new technique. J Pediatr Surg 2000;35:927-31. [5] Yamataka A, Yoshida R, Kobayashi H, et al. Intraoperative endosonography enhances laparoscopy-assisted colon pull-through for high imperforate anus. J Pediatr Surg 2002;37:1657-60. [6] de Vries PA, Pena A. Posterior sagittal anorectoplasty. J Pediatr Surg 1982;17:638-43. [7] Koga H, Okazaki T, Yamataka A, et al. Posterior urethral diverticulum after laparoscopic-assisted repair of high-type anorectal malformation in a male patient: surgical treatment and prevention. Pediatr Surg Int 2005; 21:58-60. [8] Jones NM, Humphreys MS, Goodman TR, et al. The value of anal endosonography compared with magnetic resonance imaging following the repair of anorectal malformations. Pediatr Radiol 2003;33:183-5. [9] Iwanaka T, Arai M, Kawashima H, et al. Findings of pelvic musculature and efficacy of laparoscopic muscle stimulator in laparoscopy-assisted anorectal pull-through for high imperforate anus. Surg Endosc 2003;17: 278-81.
Discussion Marc Levitt, MD (Cincinnati, Ohio): Thank you so much for trying to draw comparisons between these 2 techniques. I think we realize that they are not so different. My question to you is: You call these cases high imperforate anuses. What is included in that? Because I would consider bladder neck, prostatic, and bulbar fistula 3 very different anatomic problems, each with their own prognosis. I would also ask, what was the impact of the status of the sacrum of the spine in those patient groups? Before you draw any conclusions about continence, I think it is important to note those characteristics of the patients.
High/intermediate-type imperforate anus and postoperative outcome A Yamataka, MD (response): Thank you for your question. In answer to the first part of your question, in our paper, intermediate types, such as rectobulbar urethral fistula and rectovaginal fistula, were already included in our high types. So, you are very correct in pointing out that bladder neck type, prostatic type, and bulbar type will all have different prognoses. In our paper, PSARP was done in 2 cases of bladder neck type and none in the Georgeson procedure, which we know may lead to a slight bias in our results. However, we still believe our findings sufficiently reflect the outcome in the 2 groups and may prove useful. In answer to the second part of your question, from x-ray images of the spine and sacrum, they were shown to be normal in all cases. But if we took a more in-depth look from MRI, a few cases might be shown to have tethering of the spinal cord, which may affect our results and conclusion as you pointed out.
163
Jean Martin Laberge, MD (Montreal, Canada): Thank you for a beautiful presentation. It was very clear. I have a question: I noted that the age of operation was around 8 or 9 months. I think in recent years we tend to do the operation in younger patients. Would you like to comment on what is your preferred age for the definitive pull-through? A Yamataka, MD (response): Thank you very much for your question. Currently, during the Georgeson procedure, I perform intraoperative colonoscopy and intraoperative cystoscopy to evaluate the size of the fistula and excise the fistula as much a possible, especially in the imperforate anus with an intermediate type, ie, rectobulbar urethral fistula. So, I need an adequate-sized baby in order to best perform this. This is why I prefer to operate around the age that is mentioned.
www.elsevier.com/locate/jpedsurg
Midterm postoperative clinicoradiologic analysis of surgery for high/intermediate-type imperforate anus: prospective comparative study between laparoscopy-assisted and posterior sagittal anorectoplasty Chizue Ichijo a , Kazuhiro Kaneyama a , Yutaka Hayashi a , Hiroyuki Koga a , Tadaharu Okazaki a , Geoffrey J. Lane a , Yoshihisa Kurosaki b , Atsuyuki Yamataka a,⁎ a
Department of Pediatric General and Urogenital Surgery, Juntendo University School of Medicine, Tokyo 113-8421, Japan Department of Radiology, Juntendo University School of Medicine, Tokyo 113-8421, Japan
b
Received 29 August 2007; accepted 2 September 2007
Index words: High-type imperforate anus; Intermediate-type imperforate anus; Anorectoplasty; Anal endosonography; Magnetic resonance imaging; Continence evaluation questionnaire
Abstract Purpose: The objective of this study is to analyze the outcome of surgery for high/intermediate-type imperforate anus using anal endosonography (AES), magnetic resonance imaging (MRI), and a continence evaluation questionnaire (CEQ). Methods: In this study, 24 cases of high/intermediate-type imperforate anus were studied. Fifteen of 24 had Georgeson's laparoscopy–assisted anorectoplasty (GLA), and 9 of 24 had Pena's posterior sagittal anorectoplasty (PPA). All subjects had AES and MRI postoperatively. On AES, differences in thickness of the external sphincter (ES) and puborectalis (PR) at 3 and 9 o'clock were compared, and if pull-through colon was central, AES was 0. On MRI, differences in thickness were analyzed semiquantitatively and scored; if muscles were of even thickness, the score was 0; slight difference was 1, and marked difference was 2. A 5-parameter CEQ questionnaire (maximum score, 10) was administered to 16 of 24 subjects followed up for more than 3 years (9 GLA, 7 PPA). Surgical stress was assessed using mean febrile period, duration of raised white blood cell count, and peak C-reactive protein level. Results: Mean age at surgery and mean postoperative period for both groups were not statistically different. There were no differences in mean muscle thickness for ES or PR on AES according to procedure (ES: GLA = 0.19 ± 0.15 mm, PPA = 0.16 ± 0.09 mm, P = .59; PR: GLA = 0.19 ± 0.19 mm, PPA = 0.22 ± 0.15 mm, P = .69). Magnetic resonance imaging scores were also not significantly different according to procedure (GLA: 0.77 ± 0.83; PPA: 0.75 ± 0.50, P = .97). When CEQ were compared annually, scores for GLA were generally higher throughout the study but only statistically significant at 3 and 4 years (P b .05). Differences in parameters of surgical stress were not significant. Conclusion: Although there were no differences in muscle thickness and parameters of surgical stress observed according to technique, GLA would appear to provide better outcome based on CEQ scores. © 2008 Published by Elsevier Inc.
Presented at the 38th annual meeting of the American Pediatric Surgical Association, Orlando, Florida, May 24-27, 2007. ⁎ Corresponding author. Tel.: +81 3 3813 3111; fax: +81 3 5802 2033. E-mail address: [email protected] (A. Yamataka). 0022-3468/$ – see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.jpedsurg.2007.09.037
High/intermediate-type imperforate anus and postoperative outcome In recent reports [1-3] about short- to midterm outcome after Georgeson's laparoscopy–assisted anorectoplasty (GLA) for high/intermediate-type imperforate anus (HIIA) [4], GLA appeared to be as good as or even better than Pena's posterior sagittal anorectoplasty (PPA). However, there are no prospective studies comparing GLA and PPA in the literature, and past reports about continence evaluation have usually been retrospective. At our institute, we have been assessing our patients with HIIA who underwent surgery (GLA or PPA) after 1999 prospectively with a structured continence evaluation questionnaire (CEQ) administered at follow-up outpatient clinic attendance. Because of this, we were prompted to compare midterm outcome of GLA and PPA on an annual basis as well as use imaging technology (anal endosonography [AES] and magnetic resonance) to obtain information on postoperative anorectal position that might influence continence and correlate position data with outcome to compare surgical techniques.
159
1.2. Anal endosonography and pelvic magnetic resonance imaging Both AES and magnetic resonance imaging (MRI) were performed between 6 months and 7 years postoperatively in all 24 subjects to evaluate anorectal anatomy. For AES, we used rotating ultrasonographic scanners to examine the muscles of the pelvic floor using 360° crosssectional imaging: RU-75M-R1 (7.5 MHz) and RU-12M-R1 (12 MHz) (OLYMPUS, Tokyo, Japan); penetration depth, 2
1. Materials and methods 1.1. Patient demographics and surgical techniques From January 1999 to June 2006, we treated 28 patients with HIIA. All 28 were treated initially with colostomy during the neonatal period. Consent to perform general anesthesia for AES could not be obtained for 4 of these (2 GLA, 2 PPA), and they were excluded, leaving 24 patients as subjects for this study. Of the 24, 15 had GLA and 9 had PPA. High/intermediate-type imperforate anus cases treated by GLA were 5 males with rectoprostatic urethral fistula, 4 males with rectobulbar urethral fistula, 1 male with anorectal agenesis without fistula, 1 female with rectovaginal fistula, 2 females with rectovestibular fistula with absent vagina, and 2 females with cloacal anomaly. High/intermediate-type imperforate anus cases treated by PPA were 2 males with rectovesical fistula, 4 males with rectobulbar urethral fistula, 1 male with anorectal agenesis without fistula, and 2 females with rectovestibular fistula. All HIIA surgery (GLA or PPA) was performed under the direct supervision of a single surgeon (AY). Conventional techniques as described elsewhere by Georgeson et al [4,5] were used to perform GLA. Posterior sagittal procedure was performed as previously described by de Vries and Pena [6]. In this study, there were 9 PPA cases; 3 before GLA was introduced, and 6 after GLA was introduced. After GLA was introduced, 3 of the 6 cases who had PPA had a history of other abdominal surgery for malrotation and duodenal atresia. The remaining 3 cases (1 case of rectoprostatic urethral fistula and 2 cases of rectobulbar urethral fistula) had PPA because we experienced a complication secondary to residual fistula during GLA. We have since refined our GLA procedure to include intraoperative colonoscopy and cystoscopy to deal with such cases.
Fig. 1 Postoperative AES of the PR (arrows). A, The PR is symmetric, 2.6 mm at 3 o'clock and 2.5 mm at 9 o'clock, and the difference of thickness is 0.1 mm. B, The PR is asymmetric, 2.1 mm at 3 o'clock and 1.4 mm at 9 o'clock, and the difference is 0.7 mm.
160
C. Ichijo et al.
1.3. Assessing continence and surgical stress Of the 24 subjects, 16 (9 GLA, 7 PPA) who had been followed up for more than 3 years and had no chromosomal anomalies were evaluated using a structured CEQ by a single experienced surgeon (AY). Table 1 shows the CEQ we used. Five parameters were scored on a scale of 0 to 2, giving a maximum score of 10. In this study, we used “staining” to refer to fecal markings on underclothing and “soiling” to mean actual feces on underclothing. Annual CEQ scores were determined and compared between the 2 groups. To assess surgical stress, we compared the presence of fever, raised white blood cell (WBC) count (N10,000 μL), and peak serum C-reactive protein (CRP) level (N0.3 mg/dL) between the 2 groups.
1.4. Statistical analysis and ethics Data were expressed as mean ± SD. Differences in muscle thickness measured using AES, semiquantitative muscle thickness scores calculated using MRI, annual CEQ scores, and parameters of surgical stress were compared and analyzed using the unpaired Student's t test. A P value of less than .05 was considered to be statistically significant.
Fig. 2 Postoperative MRI of the PR (arrows). A, Obvious muscle masses at both 3 and 9 o'clock, indicating that MRI score is 0 (symmetric). B, Obvious muscle mass at 9 o'clock and poor muscle mass at 3 o'clock, indicating that MRI score is 2 (asymmetric).
to 5 cm; 12 mm in diameter. The endoprobe had a hard sonolucent plastic cone with a balloon at the tip to allow the probe to fit snugly within the pull-through canal once instilled with normal saline. Thicknesses of the external sphincter (ES) and puborectalis (PR) at 3 and 9 o'clock were measured, and differences in muscle thickness were compared between the 2 groups. If the pull-through colon was central, the difference in muscle thickness was 0 (Fig. 1). Magnetic resonance imaging was performed using a 1.5-T scanner (VISART v4.10; Toshiba, Japan) to obtain perpendicular axial planes for contiguous 4- to 7-mm slices along the anal canal parallel to the pubococcygeal line. Magnetic resonance imaging images at the level of the ES and PR were reviewed separately and blindly by a well-experienced radiologist (YK). Muscle thickness was classified as good, fair, or poor, according to whether muscle mass was obvious, thin, or virtually invisible. Differences at 3 and 9 o'clock were analyzed semiquantitatively for symmetry. Scoring was determined as follows: if muscles seemed to be symmetric (ie, thickness was good-good), the score was 0. Slight difference in thickness (ie, good-fair or fair-poor) scored 1 point, and marked difference (ie, good-poor) scored 2 points. The worst score was 2 for each muscle, that is, 4 (ES = 2 + PR = 2) (Fig. 2).
Table 1
Continence evaluation questionnaire score Score
1. Frequency of defecation 1-2/d 3-5/d ≥6/d
2 1 0
2. Staining/soiling None Occasional staining Staining Staining always Soiling
2 1.5 1 0.5 0
3. Perianal erosion Nil Occasionally Often
2 1 0
4. Anal shape Normal looking Scar or skin tag visible Mucosa visible Prolapse needing surgery
2 1.5 1 0
5. Medication Nil Laxative/enema/suppository Antidiarrheals needed Maximum score
2 1 0 10
High/intermediate-type imperforate anus and postoperative outcome This study was approved by the Juntendo University School of Medicine Ethics Committee and complies with the Helsinki Declaration of 1975 (revised 1983).
2. Results Of the 15 GLA subjects, 10 were boys and 5 were girls, and of the 9 PPA subjects, 7 were boys and 2 were girls. The mean age at surgery was 9.6 ± 7.1 months for GLA and 7.1 ± 4.8 months for PPA (P = .45), and the mean length of followup after surgery was 60.0 ± 15.1 months for GLA and 66.3 ± 24.0 months for PPA (P = .53). On AES, the mean difference in ES thicknesses was 0.19 ± 0.15 mm for GLA and 0.16 ± 0.09 mm for PPA (P = .59). Also, the mean difference in PR thicknesses was not statistically significant between the 2 groups (0.19 ± 0.19 mm for GLA, 0.22 ± 0.15 mm for PPA, P = .69). The mean semiquantitative MRI scores were not significantly different (GLA, 0.77 ± 0.83; PPA, 0.75 ± 0.50, P = .97), indicating there were no differences in muscle thicknesses according to surgical technique. The maximum duration of follow-up in the GLA group was 6 years—but only for 2 subjects, so direct comparison between the 2 groups was only possible for the first 5 years after surgery. The mean annual CEQ scores for the first 5 years after GLA were 6.1 ± 1.6 (n = 9), 7.2 ± 1.3 (n = 9), 8.1 ± 0.8 (n = 9), 8.2 ± 0.7 (n = 8), and 8.2 ± 1.1 (n = 5), and for the first 5 years after PPA, the scores were 5.1 ± 1.9 (n = 7), 5.9 ± 1.8 (n = 7), 6.3 ± 1.6 (n = 7), 6.2 ± 1.9 (n = 5), and 6.5 ± 2.3 (n = 4) (Fig. 3). Georgeson's laparoscopy–assisted anorectoplasty had generally higher CEQ scores throughout the study, but differences were only statistically significant at 3 and 4 years after surgery (P b .05).
Fig. 3 Continence evaluation questionnaire scores after GLA and PPA. Scores for GLA are generally higher throughout the study but only statistically significant at 3 and 4 years (*P b .05 in both).
161
The mean peak WBC and duration of raised WBC for GLA were 14,720 ± 3940/μL, 5.3 ± 3.3 days, and for PPA were 11,730 ± 4200/μL, 4.1 ± 7.1 days, respectively (P = .08 and .59). The mean peak CRP and duration of raised CRP for GLA were 8.19 ± 6.33 mg/dL, 10.5 ± 5.4 days, and for PPA were 3.92 ± 3.36 mg/dL, 6.5 ± 5.9 days, respectively (P = .07 and .16). The mean duration of fever was 0.9 ± 2.1 days for GLA and 1.1 ± 2.2 days for PPA (P = .87), indicating that all parameters for surgical stress were not statistically different between the 2 groups. Postoperative complications occurred in 2 GLA patients with rectobulbar urethral fistula (one had an enlarged residual fistula [7] and the other had temporary dysuria) and 3 PPA patients (minor wound infection in all).
3. Discussion In this study, we applied AES to evaluate postoperative anorectal anatomy accurately. Essentially, our aim was to compare muscle thickness along the pull-through route for symmetry rather than for actual thickness. Thus, we did not focus too much on differences in muscle thickness between causes of HIIA, but generally, lower-type disorders such as rectobulbar fistula appear to have thicker muscles than higher-type disorders such as cloaca and rectovesical fistula. A recently published report about AES in a pediatric population by Jones et al [8] investigated the value of AES compared with MRI after the repair of anorectal malformations and recommended that AES was an accurate alternative to MRI in the assessment of anorectoplasty because AES had a better correlation with results of perineal muscle stimulation (MRI failed in some cases). In our study, we performed both AES and MRI for thoroughness and confirmed that AES was accurate for measuring the thicknesses of ES and PR in all cases. However, MRI alone was not reliable for measuring thickness because only the thickness on a particular image can be measured. Thus, MRI was only used for making a semiquantitative assessment because subjective factors would appear to influence the interpretation of MRI images. Using our AES results, we found that GLA was just as accurate as PPA for positioning the colon accurately along the pull-through route; in other words, pelvic floor muscles were found to surround the pull-through route nicely after GLA as they do after PPA in most of cases. Thus, we suggest that the main reason for the better continence achieved after GLA might be related to the less impact on the pelvic nerve system and levator muscles during GLA, allowing post-GLA patients to achieve better continence at an earlier period (especially 3-4 years postoperatively). Iwanaka et al [9] also recommended GLA as an alternative procedure because PPA was considered to cause damage to the sphincter muscles as well as the tiny nerves that maintain anorectal sensation and motility as a consequence of the large incision used in the pelvis, a conclusion that supports our findings.
162 Our CEQ results showed that GLA subjects had higher scores than PPA subjects, suggesting that GLA subjects do achieve better midterm postoperative fecal continence. Specifically, frequency of motions, staining/soiling, and incidence of erosions were better in GLA subjects, and the difference between GLA and PPA for staining/soiling was statistically significant 4 years after surgery. However, in the literature, there are only 2 reports comparing midterm outcome of GLA and PPA [1,2]. Kudou et al [1] found that both GLA and PPA patients had similar bowel habits. In their study, age at the time of evaluation of fecal continence was significantly lower in the GLA group (51 months for GLA vs 73 months for PPA), and their results appear not to be age matched, suggesting that anorectal function in the GLA group would probably improve in the long term. Wong et al [2] also compared fecal continence according to procedure (GLA vs PPA) and reported that significantly more patients in the GLA group had acceptable defecation status. However, they investigated only frequency of motions as an indicator of fecal continence, which we believe is insufficient because fecal continence is multifactorial. Thus, we assessed severity of staining/soiling and requirement for medications as conventional factors but added presence of perianal erosions and anal shape to provide a complete picture because, if there is anal mucosa prolapse, there is increased irritation/stimulation of the anus, which can be the cause of unstable bowel control. We found no statistical difference for anal shape or perianal erosions according to technique throughout the period of our study. In addition, the median duration of follow-up in their study (as well as Kudou et al) was not matched (27 months for GLA and 52 months for PPA). Thus, although our study is the first to compare age/sex-matched patients with regard to postoperative fecal continence according to procedure, there is a risk for bias because of small patient numbers; for example, there were 2 cases of rectovesical fistula in the PPA group that could lower CEQ scores, and increasing patient numbers will help diminish this risk. One challenging issue revealed by this study is the role of GLA for patients with rectobulbar urethral fistula, because the 2 subjects with complications both had rectobulbar urethral fistula type. Briefly, one had a large cystic mass behind the bladder, suggestive of a posterior urethral diverticulum, and we diagnosed it as an enlarged residual fistula at laparotomy. With this in mind, we began to perform intraoperative colonoscopy and cystoscopy during GLA to evaluate the size of the fistula lumen and facilitate excision of the fistula as completely as possible [7]. However, we had one patient develop temporary dysuria after intraoperative colonoscopy and cystoscopy, probably because of surgical manipulation of the posterior urethral nerve plexus and/or as a result of indirect thermal injury to the posterior urethral nerve plexus when monopolar diathermy was used. Thus, for rectobulbar urethral fistula, dissection must be meticulous and careful during GLA, and in fact, we use PPA in such cases to decrease physical impact on pelvic nerves.
C. Ichijo et al. In conclusion, for the treatment of HIIA (except rectobulbar urethral fistula), GLA appears to have a better postoperative outcome than PPA, based on CEQ results, with no difference in the amount of surgical stress inflicted. However, the small but statistically significant difference in outcome may not have any great clinical application at the present time because it is unlikely to guide surgeons in planning their management. Thus, although both procedures would appear to be equivalent in terms of muscle thickness and surgical stress, GLA may appear to have a small advantage over PPA functionally. Further follow-up research is currently being performed to increase the number of patients and assess long-term outcome.
References [1] Kudou S, Iwanaka T, Kawashima H, et al. Midterm follow-up study of high-type imperforate anus after laparoscopically assisted anorectoplasty. J Pediatr Surg 2005;40:1923-6. [2] Wong KKY, Khong PL, Lin SCL, et al. Post-operative magnetic resonance evaluation of children after laparoscopic anorectoplasty for imperforate anus. Int J Colorectal Dis 2004;20:33-7. [3] Lin CL, Wong KKY, Lan LCL, et al. Earlier appearance and higher incidence of the rectoanal relaxation reflex in patients with imperforate anus repaired with laparoscopically assisted anorectoplasty. Surg Endosc 2003;17:1646-9. [4] Georgeson KE, Inge TH, Albanese CT. Laparoscopically assisted anorectal pull-through for high imperforate anus—a new technique. J Pediatr Surg 2000;35:927-31. [5] Yamataka A, Yoshida R, Kobayashi H, et al. Intraoperative endosonography enhances laparoscopy-assisted colon pull-through for high imperforate anus. J Pediatr Surg 2002;37:1657-60. [6] de Vries PA, Pena A. Posterior sagittal anorectoplasty. J Pediatr Surg 1982;17:638-43. [7] Koga H, Okazaki T, Yamataka A, et al. Posterior urethral diverticulum after laparoscopic-assisted repair of high-type anorectal malformation in a male patient: surgical treatment and prevention. Pediatr Surg Int 2005; 21:58-60. [8] Jones NM, Humphreys MS, Goodman TR, et al. The value of anal endosonography compared with magnetic resonance imaging following the repair of anorectal malformations. Pediatr Radiol 2003;33:183-5. [9] Iwanaka T, Arai M, Kawashima H, et al. Findings of pelvic musculature and efficacy of laparoscopic muscle stimulator in laparoscopy-assisted anorectal pull-through for high imperforate anus. Surg Endosc 2003;17: 278-81.
Discussion Marc Levitt, MD (Cincinnati, Ohio): Thank you so much for trying to draw comparisons between these 2 techniques. I think we realize that they are not so different. My question to you is: You call these cases high imperforate anuses. What is included in that? Because I would consider bladder neck, prostatic, and bulbar fistula 3 very different anatomic problems, each with their own prognosis. I would also ask, what was the impact of the status of the sacrum of the spine in those patient groups? Before you draw any conclusions about continence, I think it is important to note those characteristics of the patients.
High/intermediate-type imperforate anus and postoperative outcome A Yamataka, MD (response): Thank you for your question. In answer to the first part of your question, in our paper, intermediate types, such as rectobulbar urethral fistula and rectovaginal fistula, were already included in our high types. So, you are very correct in pointing out that bladder neck type, prostatic type, and bulbar type will all have different prognoses. In our paper, PSARP was done in 2 cases of bladder neck type and none in the Georgeson procedure, which we know may lead to a slight bias in our results. However, we still believe our findings sufficiently reflect the outcome in the 2 groups and may prove useful. In answer to the second part of your question, from x-ray images of the spine and sacrum, they were shown to be normal in all cases. But if we took a more in-depth look from MRI, a few cases might be shown to have tethering of the spinal cord, which may affect our results and conclusion as you pointed out.
163
Jean Martin Laberge, MD (Montreal, Canada): Thank you for a beautiful presentation. It was very clear. I have a question: I noted that the age of operation was around 8 or 9 months. I think in recent years we tend to do the operation in younger patients. Would you like to comment on what is your preferred age for the definitive pull-through? A Yamataka, MD (response): Thank you very much for your question. Currently, during the Georgeson procedure, I perform intraoperative colonoscopy and intraoperative cystoscopy to evaluate the size of the fistula and excise the fistula as much a possible, especially in the imperforate anus with an intermediate type, ie, rectobulbar urethral fistula. So, I need an adequate-sized baby in order to best perform this. This is why I prefer to operate around the age that is mentioned.