Langenbecks Arch Surg (2011) 396:10351040 DOI 10.

1007/s00423-011-0742-6

ORIGINAL ARTICLE

Sonographic findings predictive of the need for surgical management in pediatric patients with small bowel intussusceptions
Yao Zhang & Yu-Zuo Bai & Shi-Xing Li & Shou-Jun Liu & Wei-Dong Ren & Li-Qiang Zheng

Received: 20 September 2010 / Accepted: 17 January 2011 / Published online: 28 January 2011 # Springer-Verlag 2011

Abstract Purpose This study aims to evaluate ultrasound findings that are predictive of the need for surgical management in pediatric patients with small bowel intussusceptions (SBIs). Methods A retrospective review of pediatric patients with SBIs treated from 2004 to 2009 was conducted. Patients were divided into surgical and non-surgical groups. Demographic data, ultrasound findings, treatments, and outcomes were collected and analyzed. Results There were 56 cases of SBIs in 31 males and 25 females ranging in age from 4 months to 9 years; 39 patients were managed conservatively and 17 patients underwent surgery. The mean length and diameter of the intussusception in the surgical group were 6.53 and 2.78 cm, respectively, and 3.21 and 1.81 cm, respectively in the non-surgical group (both, P <0.001). Multivariate logistic regression analysis indicated that diameter, length, and thickness of the outer rim were independent predictors of surgery. Receiver operating characteristic curve analysis indicated an intussusception diameter 2.1 cm, length 4.2 cm, and thickness of the outer rim 0.40 cm were optimal cutoff values for predicting the need for surgery.
Y. Zhang : S.-X. Li : S.-J. Liu : W.-D. Ren Department of Ultrasound, Shengjing Hospital, China Medical University, Shenyang 110004, China Y.-Z. Bai (*) Department of Pediatric Surgery, Shengjing Hospital of China Medical University, No. 36 Sanhao St, Heping District Shenyang 110004, China e-mail: baiyz@sj-hospital.org L.-Q. Zheng Department of Library, Shengjing Hospital of China Medical University, Shenyang 110004, China

Conclusions A diameter 2.1 cm, length 4.2 cm, and thickness of the outer rim 0.40 cm predict the need for surgical management in pediatric patients with SBIs. Keywords Intussusception . Small bowel intussusception . Ultrasound . Intestinal diseases . Infants

Introduction Acute intussusception is the most common condition causing an acute abdomen in infants, and typical clinical manifestations include paroxysmal crying, abdominal pain, abdominal mass, and bloody stool [1]. Intussusceptions are usually located in the ileocolic region and ileocecal junction, and these two types comprise approximately 80% of the intussusceptions in pediatric patients [2]. Small bowel intussusceptions (SBIs) are relatively rare and accounts for <10% of the intussusceptions in pediatric patients [2]. Clinical manifestations of SBIs are not typical, patients may present with non-specific signs and symptoms, an abdominal mass and bloody stool occur infrequently, and diagnosis may be delayed resulting in intestinal necrosis and a potential life-threatening situation [13]. Ultrasound is highly accurate for the diagnosis of ileocolic intussusceptions with a reported sensitivity of 98100%, and the ultrasound detection rate for SBIs is approximately 76% [4]. Diagnosis of an intussusception is made by the appearance of characteristic findings on ultrasound [5, 6]. Because some SBIs reduce spontaneously, it is controversial whether or not surgical treatment is necessary for all cases in pediatric patients [7, 8]. Doi et al. [8] reported that spontaneous reduction happens in most cases of SBI in children, and only clinical observation, rather than surgical intervention, is needed and has termed these transient

1036

Langenbecks Arch Surg (2011) 396:10351040

intussusceptions, benign SBIs. Snmez et al. [9] consider that SBIs secondary to HenochSchonlein purpura can be reduced spontaneously, and conservative treatment is feasible. However, Ko et al. [2] report that persistent SBIs are often associated with intestinal ischemia, necrosis, and perforation, and surgical intervention is warranted once diagnosed. Koh et al. [3] consider that though spontaneous reduction can be achieved in most cases of SBIs, surgical treatment is inevitable in some patients with intestinal ischemia or a pathological lead point. Therefore, it is clinically very important to determine whether an SBI is likely to reduce spontaneously in order to avoid complications and perform surgery in a timely manner, as well as avoid surgery when not necessary. In the present study, we carried out a retrospective analysis of pediatric patients with SBIs who required surgery and in whom the intussusception resolved spontaneously in order to identify ultrasound characteristics predictive of the need for surgical management.

Collected ultrasound image data included the location of the intussusception (left upper quadrant, left lower quadrant, right upper quadrant, and right lower quadrant of the abdomen), length, diameter, thickness of the sheath, the presence of enlarged lymph nodes (defined as 1.00.5 cm), the presence of intestinal expansion (diameter 3.0 cm), the presence of a visible pathological lead point, and the presence of free fluid in the abdomen. Two physicians, each with more than 5-year experience in reading pediatric ultrasounds, performed the ultrasounds and confirmed the diagnoses, i.e., both physicians read each ultrasound, and agreement between the physicians was arrived at before a final diagnosis was made. The diagnosis of SBI was based on the presence of a target sign on cross section and sleeve sign on vertical section in ultrasound images (Figs. 1 and 2). Ultrasound machines used were either a GE V730 (General Electric Healthcare, USA) or PHILIPS iu22US (Philips Medical Systems, Holland), and probes used were a 25-MHz convex array probe and a 612-MHz linear array probe.

Materials and methods Statistical analyses This retrospective study was conducted in the Pediatric Department of Shengjing Hospital of the China Medical University. This study was approved by the institutional review board of the hospital, and the requirement of informed consent was waived because of its retrospective nature. The records of pediatric patients (from birth to 14 years of age) with single SBIs diagnosed by ultrasound who were admitted between January 2004 and December 2009 were reviewed. Patients were divided into a group that received surgery (surgical group) and a group that did not receive surgery (non-surgical group). In the non-surgical group, spontaneous reduction was proven by ultrasound, and surgical intervention was not required. Data extracted from the medical records and evaluated included patient demographic characteristics, clinical symptoms and signs, ultrasound imaging features, operative findings, and treatment outcomes. Categorical data were presented by number and percentage and tested by chi-square test or Fisher's exact test. Continuous data were presented with mean and standard deviation or median and interquartile range and tested by t test or Wilcoxon rank sum test. Diameter, length, and thickness of the outer rim were separately tested in a logistic regression model adjusted for age, gender, and the presence of free fluid and enlarged lymph nodes, with results expressed as a parameter estimate (log odds) and 95% confidence interval (CI) for surgery as the outcome. A receiver operating characteristic (ROC) was used to determine an optimal value for deciding whether a SBI patient required surgery. Data were analyzed using SPSS 15.0 (SPSS Inc., Chicago, IL, USA). A value of P <0.05 was considered to indicate statistical significance.

Fig. 1 a A 22-month-old female. Ultrasound shows the target sign, and the diameter of the abdominal mass was 2.8 cm. Ileal intussusception was confirmed at surgery. b A 2-year-old male with transient small bowel intussusception. Ultrasounds show the target sign, and the diameter of the abdominal mass was 2.0 cm. The mass had disappeared at reexamination

Langenbecks Arch Surg (2011) 396:10351040 Fig. 2 a A 22-month-old female. Ultrasounds show the sleeve sign, and the length of the mass was 5.5 cm. Ileal intussusception was confirmed at surgery. b A 3-year-old male with transient small bowel intussusception. Ultrasounds show the sleeve sign on vertical section, and the length of the mass was 2.1 cm. The mass had disappeared at reexamination

1037

Results Between January 2004 and December 2009, there were 56 cases of single SBIs diagnosed by abdominal ultrasound findings of a target sign and sleeve sign. There were 31 males and 25 females ranging in age from 4 months to 9 years. There were 39 patients who were managed conservatively and did not receive surgery (non-surgical group), and 17 patients who underwent surgery (surgical group). Patient data by group are presented in Table 1. In the surgical group, 14 patients had various degrees of intestinal obstruction, and symptoms included vomiting, abdominal distention, and the absence of flatus and defecation. Bloody stool was present in one case, and plain abdominal radiograph showed a stepped liquidgas interface. There were six secondary SBIs (four associated with intestinal polyps and two ileocolic intussusceptions secondary to an inverted Meckel's diverticulum), five postoperative SBIs, and four were primary SBIs without an identifiable cause. In two cases, preoperative ultrasound
Table 1 Patient demographic and sonographic data

showed an SBI, and repeated ultrasound scanning was carried out every 23 h in which a SBI was observed every time; however, no SBI was found during exploratory laparotomy. Intraoperative manual reduction of the SBI was performed in the 15 cases in which an SBI was found during surgery. In addition, polypectomy was performed in four cases, and Meckel's diverticulum resection was performed in two cases. Intestinal necrosis occurred in one case, and resection and anastomosis were performed. Ultrasound-guided hydrostatic reduction using a saline enema was attempted in five cases and failed in four cases. In one case, recurrence occurred after successful reduction, and surgical treatment was carried out after the intussusception developed four times in 2 days. Surgery and postoperative recovery were without complications in all cases, and patients were discharged in good condition. In the non-surgical group, the main clinical manifestation was diarrhea in 21 cases, paroxysmal crying and abdominal pain in 11, and vomiting in 15. In five cases, there were no clinical symptoms, and the SBIs were found
Non-surgery group (n =39) Surgery group (n =17) 6 (35.3) 22 [948] 17 15 14 1 0 6 2.780.41 6.532.60 0.550.10 2:6:5:4 13 (76.5) 7 (41.2) 11 (64.7) P value 0.3528 0.2861

LU left upper, LL left lower, RU right upper, RL right lower *P <0.05, statistical significance
Chi-square test or Fisher's exact test
b Median [interquartile range], Wilcoxon rank sums test, or meanstandard deviation, t test a

Female (n, %)a Age (months)b Abdominal pain Vomiting No flatus or defecation Bloody stool No clinical symptoms Secondary small bowel intussusception Diameter (cm)b Length (cm)b Thickness of outer rim (mm)b Location (LU/LL/RU/RL)a Free fluid present (n, %) Enlarged lymph nodes present (n, %)a Bowel dilatation present (n, %)a

19 (18.7) 24 [2148] 21 15 2 0 5 0 1.810.28 3.210.86 0.350.05 24:6:8:1 2 (5.1) 15 (42.2) 2 (5.1)

<0.0001* <0.0001* <0.0001* <0.0001* <0.0001* 0.8483 <0.0001*

1038

Langenbecks Arch Surg (2011) 396:10351040

incidentally (two underwent ultrasound reexamination after abdominal tumor resection, one underwent ultrasound reexamination after appendectomy, and two underwent routine abdominal ultrasound examination after a car accident). No abdominal distention, abdominal mass, bloody stool, or signs of acute intestinal obstruction were noted in any of the patients in this group. Spontaneous reduction of the SBI was proven by ultrasound in all 39 patients. In 6 cases, spontaneous reduction was observed during scanning, in 31 cases, reduction occurred 3060 min after ultrasound scanning, and in 2 cases, reduction occurred 23 h after scanning. Abdominal ultrasound, performed 1224 h after the apparent reduction in all patients, revealed no signs of SBI. In 11 patients, carbon powder was administered orally, and discharge through the anus 824 h later was noted, indicating that the SBI had resolved. Ultrasound findings of both groups are presented in Table 1. No pathologic lead point was found in any patient in either group. The mean length and diameter of the intussusception in the surgical group were 6.53 and 2.78 cm, respectively, and both were statistically greater than the corresponding measurements in the non-surgical group (3.21 and 1.81 cm, respectively; both, P <0.001). Results of the multivariate logistic regression analysis are presented in Table 2, and indicate that diameter, length, and thickness of the outer rim were all independent predictors of the necessity of surgery with estimate values (95% CIs) of 6.2 (1.511.0), 2.4 (0.74.1), and 24.2 (4.6 43.8), respectively. Age, gender, and the presence of free fluid, enlarged lymph nodes, and bowel dilatation had no independent correlation with whether or not a patient required surgical treatment. ROC curve analysis was performed to determine the best critical values for diameter, length, and thickness of the outer rim for evaluating whether or not an SBI patient requires surgical management, and results are presented in Fig. 3 and Table 3. The diameter was 2.1 cm (area under the ROC curve [AUC] 0.973, 95% CI 0.9391.000), the length was 4.2 cm (AUC 0.964, 95% CI 0.9201.000), and
Table 2 Parameter estimates of logistic regressions and 95% CIs for surgery as a function of diameter, length, and thickness of the outer rim, adjusted by age, gender, free fluid present, enlarged lymph nodes present, and bowel dilatation present Ultrasound feature Diameter Length Thickness of the outer rim CI confidence interval *P <0.05, statistical significance Estimates 6.2 2.4 24.2 95% CI (1.511.0) (0.74.1) (4.643.8) P value 0.0099* 0.0058* 0.0153*

Fig. 3 Diagnostic accuracy of diameter, length, and thickness of the outer rim on a continuous scale in the prediction of surgery using a receiver operating characteristic curve. Diameter: AUC1 (95% CI), 0.973 (0.9391.000), P <0.001. Length: AUC2 (95% CI), 0.964 (0.9201.000), P <0.001. Thickness of the outer rim: AUC3 (95% CI), 0.968 (0.9291.000), P <0.001. AUC area under the curve, CI confidence interval

the thickness of the outer rim was 0.40 cm (AUC 0.968, 95% CI 0.9291.000). The sensitivity, specificity, positive predictive value, and negative predictive value of the values are shown in Table 3. If the length of the intussusception is 4.2 cm, the sensitivity and specificity for the need of surgical management are 94.1% and 84.6%, respectively.

Discussion In the present study, we retrospectively analyzed the clinical manifestations and ultrasound imaging features of all cases of SBIs managed surgically and conservatively. Our results showed that diameter, length, and thickness of the outer rim of the intussusception were all independent predictors of the need for surgery. ROC analysis indicated that when the length of intussusception is 4.2 cm, the sensitivity and specificity of the need of surgical management are 94.1% and 84.6%, respectively. Diagnosis of most SBIs relies on ultrasound scanning, and types of intussusceptions can be readily distinguished [5, 6, 10]. In addition, with the development and wide

Table 3 Specific ultrasound features as predictors of the necessity of surgery

Ultrasound feature (cm) Diameter 2.1 Length 4.2 Thickness 0.4 Sensitivity Specificity Positive predictive value 70.8 72.7 71.4 Negative predictive value 100 97.1 94.3

100 94.1 88.2

82.1 84.6 84.6

Langenbecks Arch Surg (2011) 396:10351040

1039

application of ultrasound technology, the presence of temporary or transient small bowel intussusceptions has become clinically apparent [1, 4, 5, 8, 11]. Doi et al. [8] have suggested the term benign small bowel intussusception to describe those that are found incidentally and/or resolve spontaneously. Reports have suggested that 50% or more of SBIs will resolve spontaneously [1, 8, 12], thus identifying those that will resolve with conservative management, and those that require surgical management is of clinical importance. Delay of surgery can result in intestinal necrosis, and surgical management of an SBI that is likely to resolve spontaneously exposes the patient to the risks of surgery [2]. While the findings of this study indicated that a large percentage of SBIs resolved spontaneously, we are confident of the results as the diagnosis of SBI was confirmed by two experienced physicians in each case. Kim [4] reviewed the ultrasound findings of 34 SBIs diagnosed in 32 infants and children and found that transient SBIs typically exhibited a small size (mean diameter 1.5 cm) without wall swelling, a short segment (mean length 1.8 cm), preserved wall motion, and the absence of a lead point. Mateen et al. [11] reviewed the records of 108 consecutive patients (adults and children) with intestinal intussusceptions, of which 41 were diagnosed as transient SBIs. They found that SBIs without an identifiable pathological lead point, normal wall thickness, normal non-dilated proximal bowel, normal vascularity on color Doppler ultrasound, and a length of <3.5 cm reduced spontaneously and were not of clinical significance. On the other hand, Munden et al. [1], in a study of 35 cases of SBIs in adults and children, found that a length >3.5 cm was a strong indicator for the need of surgical intervention. Possible reasons for the difference between the results of Munden et al. and those of the current study are ethnic differences in the subjects of the two studies, differences in the ultrasound equipment used, and differences in the statistical methods used. The presence of a pathological lead point is suggestive of the need for surgical intervention. Koh et al. [3] reported the characteristics of six patients with SBIs who required surgical intervention, and five exhibited a pathological lead point on ultrasound. Ko et al. [2] reported the presence of pathological lead points in 8 of 19 surgically proven cases of SBIs in pediatric patients. Navarro et al. [13] found that surgery was required in 32 of 43 children with intussusceptions with pathological lead points. Interestingly, a pathological lead point was not identified in any of the patients, surgical or non-surgical, in our study. In the surgical group, ultrasound-guided hydrostatic reduction using a saline enema was attempted in five cases, and failed in four cases. In one case, recurrence occurred after successful reduction, and surgical treatment was

carried out. Mirilas et al. [14] studied the sonographic features that correlated with hydrostatic reducibility of intestinal intussusceptions in infants and children. They found that when the head of the intussusception appeared as a target-like mass, the hydrostatic reduction rate was 100%. When the head appeared as a donut-like mass, hydrostatic reproducibility depended on the thickness of the hypoechoic external ring of the donut, with larger thicknesses correlating with failure of hydrostatic reduction and the need for surgery. In addition, when a small amount of fluid was present within the head of the intussusception, hydrostatic reduction was unsuccessful in all cases. Though ultrasound imaging features are useful for determining the necessity of surgical intervention, the need for surgical intervention also depends on clinical signs and symptoms that may suggest intestinal obstruction or ischemia. Our results suggest that if the patient has no signs of mechanical ileus, the length of intussusception is short (<4.2 cm), the diameter of intussusception is small (<2.1 cm), and there is no swelling of the intestinal wall, it should be considered a transient SBI, and close observation and repeat ultrasounds are warranted. Conversely, if the length of intussusception is 4.2 cm, the diameter of intussusception is 2.1 cm, the thickness of the outer rim is 0.4 cm, there is swelling of the bowel wall, and there is evidence of mechanical ileus (e.g., vomiting, abdominal distension), surgical management should be considered. There are some limitations to the current study that should be considered. This was a retrospective study with a relatively small number of cases performed at one center. A larger number of cases are necessary to confirm the findings. Enema reduction was not attempted primarily in all of the 17 cases that were managed surgically; although 5 cases once underwent ultrasound-guided hydrostatic reduction using a saline enema, all failed. Thus, evaluation of a relationship between the reducibility of intussusception with non-surgical methods vs. spontaneous reduction or the need for laparotomy was not possible. Lastly, unavoidable error in ultrasound measurements may potentially affect the results.

Conclusion In summary, our data indicate that an intussusception diameter 2.1 cm, length 4.2 cm, and thickness of the outer rim 0.40 cm predict the need for surgical management in pediatric patients with SBIs. These values may be of assistance to clinicians when determining if surgery is required in pediatric patients with SBIs. Values below those determined from these data should be interpreted with caution in patients with signs and symptoms of mechanical ileus or intestinal ischemia.

1040 Acknowledgments This work was financially supported by Shengjing Outstanding Scientific Foundation (grant no. m850) from The Shengjing Hospital of China Medical University. Conflicts of interest None.

Langenbecks Arch Surg (2011) 396:10351040 7. Kornecki A, Daneman A, Navarro O, Connolly B, Manson D, Alton DJ (2000) Spontaneous reduction of intussusception: clinical spectrum, management and outcome. Pediatr Radiol 30:5863 8. Doi O, Aoyama K, Hutson JM (2004) Twenty-one cases of small bowel intussusception: the pathophysiology of idiopathic intussusception and the concept of benign small bowel intussusception. Pediatr Surg Int 20:140143 9. Snmez K, Turkyilmaz Z, Demirogullari B, Karabulut R, Aral YZ, Konu O, Baaklar AC, Kale N (2002) Conservative treatment for small intestinal intussusception associated with HenochSchnlein's purpura. Surg Today 32:10311034 10. Wiersma F, Allema JH, Holscher HC (2006) Ileoileal intussusception in children: ultrasonographic differentiation from ileocolic intussusception. Pediatr Radiol 36:11771181 11. Mateen MA, Saleem S, Rao PC, Gangadhar V, Reddy DN (2006) Transient small bowel intussusceptions: ultrasound findings and clinical significance. Abdom Imag 31:410416 12. Strouse PJ, DiPietro MA, Saez F (2003) Transient small-bowel intussusception in children on CT. Pediatr Radiol 33:316320 13. Navarro O, Dugougeat F, Kornecki A, Shuckett B, Alton DJ, Daneman A (2000) The impact of imaging in the management of intussusception owing to pathologic lead points in children. A review of 43 cases. Pediatr Radiol 30:594603 14. Mirilas P, Koumanidou C, Vakaki M, Skandalakis P, Antypas S, Kakavakis K (2001) Sonographic features indicative of hydrostatic reducibility of intestinal intussusception in infancy and early childhood. Eur Radiol 11:25762580

References
1. Munden MM, Bruzzi JF, Coley BD, Munden RF (2007) Sonography of pediatric small-bowel intussusception: differentiating surgical from nonsurgical cases. AJR Am J Roentgenol 188:275279 2. Ko SF, Lee TY, Ng SH, Wan YL, Chen MC, Tiao MM, Liang CD, Shieh CS, Chuang JH (2002) Small bowel intussusception in symptomatic pediatric patients: experiences with 19 surgically proven cases. World J Surg 26:438443 3. Koh EP, Chua JH, Chui CH, Jacobsen AS (2006) A report of 6 children with small bowel intussusception that required surgical intervention. J Pediatr Surg 41:817820 4. Kim JH (2004) US features of transient small bowel intussusception in pediatric patients. Korean J Radiol 5:178184 5. Park NH, Park SI, Park CS (2007) Ultrasonographic findings of small bowel intussusception, focusing on differentiation from ileocolic intussusception. Br J Radiol 80:798802 6. Tiao MM, Wan YL, Ng SH, Ko SF, Lee TY, Chen MC, Shieh CS, Chuang JH (2001) Sonographic features of small-bowel intussusception in pediatric patients. Acad Emerg Med 8:368373

Copyright of Langenbeck's Archives of Surgery is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.