Transperineal Sonography for Determination of the Type of Imperforate Anus

Transcription

1 Sonography of Imperforate Anus Pediatric Imaging Original Research Hans P. Haber 1 Guido Seitz 2 Steven W. Warmann 2 Jörg Fuchs 2 Haber HP, Seitz G, Warmann SW, Fuchs J Keywords: anorectal anomaly, imperforate anus, pediatric radiology, sonography DOI: /AJR Received April 26, 2007; accepted after revision June 10, Department of Pediatric Radiology, University of Tübingen, Hoppe-Seyler-Str. 1, D Tübingen, Germany. Address correspondence to H. P. Haber (peter.haber@med.uni-tuebingen.de). 2 Department of Pediatric Surgery, University Children s Hospital, Tübingen, Germany. AJR 2007; 189: X/07/ American Roentgen Ray Society Transperineal Sonography for Determination of the Type of Imperforate Anus OBJECTIVE. The purpose of this study was to assess whether transperineal sonography is valid and accurate for discriminating the low (translevator) type of imperforate anus from the intermediate and high (supralevator) types. The distinction is critical in determination of the surgical approach. MATERIALS AND METHODS. Fifty-six consecutively enrolled infants (0 90 days old) with imperforate anus underwent transperineal gray-scale sonography with a 12-MHz linear array transducer. The distance perineum was measured, and a cutoff distance for differentiating types of imperforate anus was identified. The sonographic findings were compared with the final diagnosis reached with radiographic and surgical findings. RESULTS. Transperineal sonography was feasible in all children without specific preparation. The mean distance perineum in the 22 infants with low imperforate anus was 10 ± 4 (SD) mm compared with a mean of 24 ± 6 mm in the 34 infants with intermediate or high anomalies (p < 0.001). The sensitivity of transperineal sonography was 100%; all 34 cases of intermediate or high imperforate anus were identified with a cutoff distance perineum of 15 mm. The specificity of sonography was 86% and the accuracy, 95%. CONCLUSION. Transperineal sonography is a valid and accurate noninvasive imaging technique for differentiating low and intermediate or high imperforate anus. mperforate anus is a common congenital anomaly necessitating ex- I tensive surgery in neonates [1]. The low type of imperforate anus passes through the levator ani muscle group and through its central puborectalis muscle, the major determinant of fecal continence. This translevator type of imperforate anus is managed with one-step transperineal anoplasty soon after birth [2]. The intermediate and high types of imperforate anus do not traverse the puborectalis sling. In the intermediate type, the rectum ends at the level of the puborectalis sling; in high (supralevator) imperforate anus, the rectum ends above the puborectalis muscle. The appropriate surgical procedure for both types consists of initial diverting colostomy followed by the posterior sagittal anorectoplasty as described by Peña [2]. Early assessment and accurate diagnosis of the type of imperforate anus are essential for determining treatment. In addition to clinical indexes [1 3], several imaging techniques are used, including sonography [4 7], CT [8, 9], and MRI with the patients under anesthesia [10, 11]. Noninvasive procedures without ionizing radiation are desirable, particularly in the case of neonates and infants. Initial studies of sonography showed that measurement of the distance between the distal rectal pouch and the perineum may be helpful in differentiating high and low imperforate anus [4 6]. However, the studies had only small sample sizes (fewer than 20 infants) and were conducted with different approaches to the sonographic examination, including the transabdominal and transperineal approaches. Reported cutoff values to differentiate low from intermediate and high imperforate anus range from 10 to 25 mm [4 6]. The diagnostic criteria are not well defined, and the sensitivity and specificity have not been calculated, to our knowledge.for the past 10 years, we have been using transperineal sonography to determine the type of imperforate anus in infants by measuring the the perineum. The purpose of our study was AJR:189, December

2 Fig. 1 1-day-old boy with imperforate anus. Photograph shows placement of linear transducer for transperineal sonographic examination. Fig. 2 1-day-old boy with low imperforate anus without fistula. sonogram shows 5-mm distance (asterisks, dashed line) between distal rectal pouch and perineum (P). Scale segment distance, 2 mm. S = pubic symphysis, U = urethra (arrow), CS = corpus spongiosum. Fig. 3 6-day-old girl with intermediate imperforate anus and rectovestibular fistula. sonogram shows relation between small fistula (arrow) and posterior wall of vagina. Distance (asterisks, dashed line) between distal rectal pouch and perineum (P) is 17 mm. Scale segment distance, 5 mm. S = pubic symphysis, U = urethra (arrow), V = vestibule of vagina (arrow), P = perineum. to assess whether transperineal sonography is valid and accurate in discriminating low from intermediate and high imperforate anus in neonates and infants. Materials and Methods Patients Between January 1996 and August 2006, 56 consecutively enrolled infants (30 boys, 26 girls; median age, 2 days; range, 0 90 days) with the diagnosis of imperforate anus were examined with sonography as part of routine clinical care to determine the type of imperforate anus and to evaluate associated neural and spinal anomalies. The infants were referred to our institution for surgical treatment. The final diagnosis, including the type of imperforate anus, was made on the basis of radiographic and surgical findings according to the international classification of anorectal anomalies, which is based on the relation between the level of the distal rectal pouch and the puborectalis sling of the levator muscle [12]. Forty-six of the patients were 28 days old or younger. Eleven infants had undergone colostomy before the sonographic examination and were examined during a second hospitalization for definitive treatment. Imaging All sonograms were obtained with a real-time scanner with a high-resolution 12-MHz linear array transducer (Sonoline Antares, Siemens Medical Solutions) through the transperineal approach as described by Teele and Share [13]. The gray-scale sonograms were obtained by two sonologists, each with more than 15 years of experience in diagnostic sonography. Transperineal sonography was performed before radiographic studies and definitive surgical repair. The infants were examined in the supine position without specific preparation (Fig. 1). Sonographic scans in the midsagittal plane through the perineum were used to identify the base of the bladder, the urethra, and the vagina between the urethra and the distal rectal pouch as previously described [14] (Figs. 2 and 3). The distal rectal pouch was identified on the basis of the presence of hypoechoic meconium or gas within the rectum. The distance perineum was measured in millimeters with electronic calipers as described by Teele and Share [13] (Figs. 2 4). All measurements were performed in the resting state while the child was not crying. Care was taken not to press or indent the skin or to diminish the distance between the distal rectal pouch and the perineum. For each infant, the distance perineum was recorded in the medical report. Thereafter, the distance was compared with the 1526 AJR:189, December 2007

3 Sonography of Imperforate Anus Fig. 4 1-day-old boy with high imperforate anus and small rectourethral fistula (arrow). sonogram shows distance (asterisks, dashed line) between distal rectal pouch and perineum is 22 mm. Scale segment distance, 2 mm. S = pubic symphysis, U = urethra, CS = corpus spongiosum. TABLE 1: Patient Data and Final Diagnoses Confirmed with Radiographic and Surgical Findings Type of Imperforate Anus Characteristic Low Intermediate High Number Sex (boys/girls) 15/7 3/14 12/5 Age (d) Median Range Fistula No fistula Anocutaneous 16 Rectovestibular 14 Rectovaginal 4 Rectourethral 1 a 10 Rectovesical 1 a One rectobulbar fistula. Fig. 5 Scattergram shows distance (mean ± SD) between distal rectal pouch and perineum measured on transperineal sonography in infants with low, intermediate, and high imperforate anus. Dotted line indicates cutoff distance of 15 mm. Black dots indicate data for infants with internal or anocutaneous fistula; white dots, data for infants without internal or anocutaneous fistula. Distance Between Distal Rectal Pouch and Perineum (mm) Low Intermediate High Cut off type of imperforate anus confirmed on the basis of radiographic and surgical findings. Before final repair, a distal colostomy study was performed to determine the precise type of anatomic defect in children initially treated with diverting colostomy. The study consisted of injection of hydrosoluble contrast material through the distal stoma. The study was performed under fluoroscopic guidance with the patient in the lateral position for identification of a rectourogenital fistula. In rare cases, the internal fistula was found on voiding cystourethrography or retrograde urethrography. Invertograms were not used. The studies were performed by a radiologist with more than 20 years of experience in pediatric radiology. Statistical Analysis A cutoff distance between the distal rectal pouch and the perineum was identified to separate cases of high or intermediate imperforate anus from those of low imperforate anus. The theoretic sensitivity and specificity for each potential cutoff point were analyzed. A receiver operator characteristic curve was used to a establish a cutoff point for obtaining the maximum diagnostic accuracy. A Student s t test was used to test for statistically significant differences in distance between the distal rectal pouch and the perineum for the types of imperforate anus. Results The patient data and final diagnoses confirmed on the basis of radiographic and surgical findings are summarized in Table 1. Transperineal sonography was feasible in all patients. The mean examination time was 11 minutes (range, 7 20 minutes). The distances perineum are shown in Figure 5. A final diagnosis of low imperforate anus was made for 22 patients. In 16 infants, the anocutaneous fistula was detected at inspection. At sonographic examination, the mean the perineum in this group was 10 ± 4 (SD) mm (range, 5 17 mm). A final diagnosis of intermediate or high imperforate anus was made for 34 patients (Figs. 2 and 3). At sonographic examination, the mean distance between the distal rectal pouch and the perineum in this group was 24 ± 6 mm (range, mm). The difference between the pouch-to-perineum distance in the low imperforate anus group and that in the intermediate or high group was statistically significant (p < 0.001). The 17 infants with intermediate imperforate anus had a mean pouch-to-perineum distance of 23 ± 6 mm (range, mm). The 17 infants with high imperforate anus had AJR:189, December

4 Fig day-old girl with low mimicking high imperforate anus. sonogram shows gasfilled distal rectal pouch decompressed by large anocutaneous fistula (arrowheads). Distal end of pouch is difficult to delineate and pouch-toperitoneum (P) distance (asterisks) is more than 15 mm, fulfilling criterion for high anomaly. In general, anocutaneous fistula indicates low anomaly, making pouchto-peritoneum measurements unnecessary. S = pubic symphysis, V = vagina. a mean distance of 25 ± 5 mm (range, mm). At a cutoff pouch-to-perineum distance of 15 mm, the maximum diagnostic accuracy was estimated to be 95%. Sensitivity and specificity were estimated to be 100% and 86%. The positive predictive value of sonography was 92% in this series, and the negative predictive value was 100%. On the basis of the 15-mm cutoff point, 19 of 22 cases of low imperforate anus, including the six cases without anocutaneous fistula, were correctly classified at sonographic examination (Fig. 1). Three infants had a pouch-toperineum distance of mm, and therefore the anomaly was erroneously classified at sonographic examination as intermediate or high imperforate anus (Fig. 6). All three of these patients had an anocutaneous fistula detectable at physical examination. All 34 cases of intermediate or high imperforate anus, including the four without internal fistula, were correctly classified at sonographic examination. Discussion Imperforate anus occurs with a frequency of approximately one to four cases per 5,000 neonates [1]. An accurate diagnosis usually can be made when typical clinical features are present [1 3]. The presence of an anocutaneous fistula indicates the presence of the low type of imperforate anus. Neonates with this phenotype do not need further diagnostic procedures on the anorectum and undergo a surgical procedure in the first days of life. The presence of perineal pearls or corrugated perineum also indicates low imperforate anus [3]. The absence of an anocutaneous fistula and passage of meconium through the vagina or urethra indicate the presence of the intermediate or high type of imperforate anus. However, accurate determination of the type of imperforate anus on the basis of clinical data can be difficult, especially in the 10 20% of boys and approximately 10% of girls with imperforate anus who do not have a fistula at clinical examination [1]. In some cases, external fistulas may not become apparent until the neonate is hours old, at which time the meconium moves distally into the rectum. Peña [1] proposed performing inverted lateral radiography (invertography) for determining the surgical approach in the care of these infants. The type of fistula is commonly taken into account in defining the type of imperforate anus. In this series, however, we focused on sonographic measurement of the distance between the distal rectal pouch and the perineum. We report data on the sensitivity and specificity of conventional transperineal sonographic examination in the preoperative diagnosis of imperforate anus. Our experience has shown that unlike the transverse scans used in the infracoccygeal approach [7], conventional midline sagittal sonographic scans are easy to perform without artifacts caused by gluteal tissue. In our series, there were three false-positive findings. Three (14%) of the 22 infants with low imperforate anus had a pouch-to-perineum distance greater than 15 mm. All three of these patients, however, had an anocutaneous fistula, and the diagnosis was made on clinical grounds. Sonographic measurements may not be accurate in infants in whom the distal pouch is decompressed by a large fistula, causing the distance to be exaggerated (Fig. 6). The results in these patients emphasize that sonographic findings in patients with imperforate anus must be correlated with clinical findings. In our series, however, imperforate anus was correctly classified in all infants who had undergone colostomy before sonographic examination. There were no false-negative results in our series. None of the infants with surgically proven intermediate or high imperforate anus had a pouch-to-perineum distance less than 15 mm, which we consider the cutoff value for discriminating low from intermediate and high anomalies. In our experience, it has been extremely important to perform the sonographic examination when the child is not crying. Crying and straining increase intraabdominal pressure, and the distal rectal pouch moves toward the perineum, shortening the the perineum. Because there is no associated pain or discomfort, sedation for the procedure is not necessary. Our results with this series of patients indicate that in infants without an anocutaneous fistula and with a greater than 15-mm distance perineum, the diagnosis of intermediate or high imperforate anus can be reliably made. If desired by the surgeon, the sonographic findings can be elaborated with CT and MRI to show the anatomic features of the levator ani muscles and external anal sphincter before surgery [8 10]. However, the studies of CT and MRI had only small sample sizes, and owing to poor development of the sphincter muscles in most cases of high imperforate anus, both techniques are limited in depicting the relation between the distal rectal pouch and the puborectalis muscle [9, 10]. The best radiologic techniques for discerning the type of imperforate anus are antegrade contrast studies performed through the distal limb of a colostomy and cystography and genitography in selected cases. The limitations of our study included institutional case-selection bias and lack of data on intraobserver and interobserver variability of measurements. Further work with series of infants with this disorder is necessary to overcome these shortcomings AJR:189, December 2007

5 Sonography of Imperforate Anus Transperineal sonography enables determination of the type of imperforate anus and therefore may assist in determining the appropriate surgical approach. This noninvasive imaging method allows the surgeon to identify low imperforate anus, particularly when a fistula is not detected at clinical examination, sparing some infants unnecessary colostomy. References 1. Peña A. Important basic considerations. In: Peña A, ed. Atlas of surgical management of anorectal malformations. New York, NY: Springer-Verlag, 1989: Peña A. Management of anorectal malformations during the newborn period. World J Surg 1993; 17: Seibert JJ, Golladay ES. Clinical evaluation of imperforate anus: clue to type of anal rectal anomaly. AJR 1979; 133: Schuster SR, Teele RL. An analysis of ultrasound scanning as a guide in determination of high or low imperforate anus. J Pediatr Surg 1979; 14: Oppenheimer DA, Carroll BA, Shochat SJ. Sonography of imperforate anus. Radiology 1983; 148: Donaldson JS, Black CT, Reynolds M, Shermann JO, Shkolnik A. Ultrasound of the distal pouch in infants with imperforate anus. J Pediatr Surg 1989; 24: Han TI, Kim IO, Kim WS. Imperforate anus: US determination of the type with infracoccygeal approach. Radiology 2003; 228: Kohda E, Fujioka M, Ikawa H, Yokohama J. Congenital anorectal anomaly: CT evaluation. Radiology 1985; 157: Ikawa H, Yokohama J, Sanbonmatsu T, et al. The use of computerized tomography to evaluate anorectal anomalies. J Pediatr Surg 1985; 20: Sato Y, Pringle KC, Bergmann RA, et al. Congenital anorectal anomalies: MR imaging. Radiology 1988; 168: McHugh K, Dudley NE, Tam P. Pre-operative MRI of anorectal anomalies in the newborn period. Pediatr Radiol 1995; 25: Santulli TV, Kiesewetter WB, Bill AH. Anorectal anomalies: a suggested international classification. J Pediatr Surg 1970; 5: Teele RL, Share JC. Transperineal sonography in children. AJR 1997; 168: Kim IO, Han TI, Kim WS, Yeon KM. Transperineal ultrasonography in imperforate anus: identification of the internal fistula. J Ultrasound Med 2000; 19: AJR:189, December