Ultrasonographic Signs of Intestinal Intussusception Associated With Acute Enteritis or Gastroenteritis in 19 Young Dogs

Introduction

Intestinal intussusception is created by the invagination of a portion of the intestine (intussusceptum) into the lumen of an adjacent intestinal segment (intussuscipiens) in the direction of normal peristalsis or occasionally in a retrograde direction.¹² It occurs more often as an ileocolic intussusception, although gastroduodenal, duodenojejunal, jejunojejunal, and ileoileal intussusceptions have been described in young dogs.^13–6 It is believed that most intussusceptions in young animals are idiopathic, but a number of conditions including ingestion of bones, linear foreign bodies, prior abdominal surgery, and intestinal parasitism have been suggested as predisposing factors for intestinal intussusception.¹ Acute enteritis or gastroenteritis has been demonstrated as the most likely predisposing factor for intestinal intussusception in young dogs.²

Intestinal intussusception in young dogs is usually suspected on the basis of abdominal palpation. The intussusception feels like a defined, firm, tubular structure that should be differentiated from feces and foreign bodies. Abdominal radiography, including survey radiographs and contrast studies, may be diagnostic; however, the radiological findings are often nonspecific. Ultrasonography is an accurate diagnostic method that has totally replaced conventional radiology in the diagnosis of intestinal intussusception in children;^7–9 however, few reports of ultrasonographic findings of intestinal intussusception have been published in the veterinary literature.^610–12

The aims of this study were to describe the ultrasonographic signs of intestinal intussusception associated with acute enteritis or gastroenteritis in young dogs and to determine the sensitivity, specificity, and accuracy of abdominal ultrasonography in the diagnosis of intestinal intussusception.

Materials and Methods

Between March 1998 and February 2001, abdominal radiography and ultrasonography were performed in 65 young dogs with a clinical diagnosis of acute enteritis or gastroenteritis, in order to confirm or rule out the clinical suspicion of intestinal intussusception.

All abdominal ultrasonographic examinations were performed by the same individual (Patsikas) using a 6-MHz mechanical sector transducer.^a After ventral abdominal hair coat clipping and with the dog in right lateral recumbency, the entire abdomen was examined from cranial to caudal in longitudinal, transverse, and oblique sections. The ultrasonographic findings were considered diagnostic of intestinal intussusception when multiple hyperechoic and hypoechoic concentric rings were observed in transverse sections, parallel lines in longitudinal sections, or both. Surgery was performed only in dogs with positive ultrasonographic findings. In dogs with negative ultrasonographic findings, the accuracy of the ultrasonographic diagnosis was assessed by at least 7 days of follow-up that included clinical and abdominal ultrasonographic re-examination of the dogs, telephone communication with the owners, or both. The results of ultrasonographic examinations were compared with the surgical findings. The sensitivity (i.e., the proportion of dogs with intestinal intussusception that had an ultrasonographic diagnosis of intussusception), specificity (i.e., the proportion of normal dogs that did not have an ultrasonographic diagnosis of intussusception), and accuracy (i.e., the proportion of dogs that had a correct ultrasonographic diagnosis) were calculated.¹³

Results

Intussusception, which was diagnosed with ultrasonography in 20 of 65 dogs, was confirmed surgically in 19 dogs. A false-positive diagnosis was made in one dog. In 45 dogs with no ultrasonographic evidence of intussusception, the diagnostic accuracy was assessed based on the follow-up of the cases. The sensitivity, specificity, and accuracy of ultrasonography in the diagnosis of intestinal intussusception were 100%, 97.8%, and 98.4%, respectively.

The 65 dogs included mixed-breed dogs (n=36), German shepherd dogs (n=17), rottweiler (n=7), rough collie (n=3), Dogue Argentino (n=1), and fox terrier (n=1). The mean age of the dogs was 6.32 months (range, 2 to 15 months). Vomiting was recorded in 34 dogs, small-bowel diarrhea was in 13, and vomiting and small-bowel diarrhea were in 18. Diarrhea was frank hemorrhagic in 16 dogs. The mean duration of the clinical signs was 7.1 days (range, 3 to 20 days). Abdominal palpation revealed pain in 11 dogs and a mass at the cranial and middle aspect of the abdomen in 12 dogs.

The breeds of the 19 dogs with confirmed intestinal intussusceptions included mixed-breed dogs (n=10), German shepherd dogs (n=5), Dogue Argentino (n=1), rottweiler (n=1), rough collie (n=1), and fox terrier (n=1). There were 13 males and six females. The mean age of the dogs was 8.42 months (range, 2 to 15 months). Vomiting was recorded in four dogs, small-bowel diarrhea was in seven, and vomiting and small-bowel diarrhea were in eight. Diarrhea was frank hemorrhagic in 10 dogs. The mean duration of the clinical signs was 8.63 days (range, 3 to 20 days). Palpation of the abdomen revealed pain in one dog and a mass in 12 dogs.

In 12 of the 19 dogs with intussusception, survey abdominal radiographs revealed one or more of the following findings: jejunal and colonic dilatation, a tubular viscus, reduced serosal detail, and a rounded, soft-tissue edge within the intestine that corresponds with the leading edge of the intussusceptum (i.e., meniscus sign).⁶ Abdominal radiographs in five of 19 dogs were unremarkable. The remaining two dogs were not radiographed [see Table]. Contrast radiography was not performed.

The ultrasonographic and surgical findings are summarized in the Table. A target-like mass consisting of concentric rings surrounding a hyperechoic center was observed in all transverse sections, except in case no. 1 in which the center was anechoic because of fluid accumulation into the lumen of the intussusceptum [Figure 1]. The hyperechoic center was surgically defined as mesenteric fat and compressed intussusceptum. The hyperechoic center was circular [Figure 2] in six of 19 cases (case nos. 3, 11, 14–17) or had a G or reverse G or semilunar shape [Figure 3] in seven cases (case nos. 6, 7, 9, 12, 13, 18, 19). In the remaining five cases, the hyperechoic center had no clear shape. Small, circular anechoic areas within the hyperechoic center representing mesenteric vessels and dilated mesenteric lymphatics [Figure 4] were found in nine cases (case nos. 2, 4–6, 8, 10, 12, 14, 15). In 12 of 19 cases (case nos. 2, 4–9, 12, 13, 17–19), the part of the intestine located close to the hyperechoic center and surrounded by concentric rings, represented the inner intussusceptum [Figures 3, 4]. More than five well-differentiated, thin hyperechoic and hypoechoic concentric rings [Figure 4] were found in two of 19 cases (case nos. 2, 8). In the remaining 17 cases, one (case nos. 7, 9, 13, 18, 19), two (case nos. 3, 5, 6, 10, 12, 14, 15–17), or three (case nos. 1, 4, 11) poorly differentiated concentric rings were observed [Figures 2, 3]. In seven of these 17 cases (case nos. 3, 4, 10, 12, 14–16), an anechoic ring just inside the outer ring was detected and represented fluid accumulation between the wall of the intussuscipiens and the wall of the outer intussusceptum. In one case (case no. 10) with a very long intussusception, a double target-like sign (an hourglass configuration) was detected in transverse sections [Figure 5].

An ovoid mass with an atypical nonstructured pattern, giving the impression of a tumor (case nos. 5, 7) or kidney (case no. 19), was found in longitudinal sections in three dogs [Figure 6]. In the remaining 16 dogs, longitudinal sections revealed multiple hyperechoic and hypoechoic parallel lines [Figure 7]. However, in three of these 16 dogs (case nos. 2, 6, 8), the walls of the inner and outer intussusceptum and the wall of the intussuscipiens were forming a trident-like configuration [Figure 8].

Exploratory surgery revealed one colocolic (case no. 9) and 18 ileocolic intussusceptions. Intestinal resection and anastomosis were performed in all except four dogs (case nos. 2, 6, 8, 10) in which manual reduction was done. Multiple, full-thickness intestinal biopsies from dogs in which resection was done subsequently did not reveal any evidence of bowel disease. In particular, there was no evidence of idiopathic inflammatory bowel disease.

In the single false-positive case of the present study, a target-like mass consisting of a semilunar, hyperechoic center surrounded by poorly differentiated concentric rings was observed in transverse sections, and multiple hyperechoic and hypoechoic parallel lines were seen in longitudinal sections [Figure 9]. A cranial mesenteric lymphadenomegaly was found on exploratory surgery 5 days after the ultrasonographic examination. The enlarged lymph node was examined ultrasonographically after surgery. The ultrasonographic signs were compatible with a normal echotexture, enlarged lymph node and did not resemble the target-like mass that was observed before surgery.

Discussion

The ultrasonographic patterns of intestinal intussusception result from the juxtaposition of the walls of the inner and outer intussusceptum and the wall of intussuscipiens.¹² The patterns may vary with the quality of the image, the length of bowel involved, the duration of the process, and the orientation of the scan plane in relation to the axis of the intussusception.⁶¹²

The following have been described in transverse, ultrasonographic sections of intestinal intussusception in dogs: hypoechoic concentric rings; multilayered, target-like lesions; or hypoechoic and hyperechoic concentric rings, usually with a hyperechoic center.^56111214 In humans, the target-like or bull’s eye-like lesion is the most frequently described pattern in transverse sections, although atypical patterns have been reported.⁹¹⁵ Although variations among the cases were found, a target-like mass consisting of concentric rings surrounding a center was detected in all transverse sections in the present study.

An echogenic center caused by the presence of the mesentery into the lumen of the intussuscipiens and compression of the intussusceptum was a common finding in this study. However, it was observed that fluid accumulation into the lumen gave an anechoic center, as has been reported in both humans and foals.¹⁵¹⁶ In cases with a non-compressed wall of the inner intussusceptum, the mesentery within the lumen of the intussuscipiens formed a G shape, a reverse G shape, or a semilunar, hyperechoic center. In cases with a compressed wall of the inner intussusceptum, a wide hyperechoic center was evident. In humans, a semilunar or G-shape hyperechoic center is typical of a loose or reducible intussusception, in contrast to a circular homogeneous, hyperechoic center that is usually a sign of a nonreducible intussusception.¹⁷¹⁸ However, in the present study, the G-shape hyperechoic center was found in both reducible and nonreducible intussusceptions.

Five layers are normally visible ultrasonographically in the intestinal wall: the hyperechoic mucosal surface, hypoechoic mucosa, hyperechoic submucosa, hypoechoic muscularis, and hyperechoic serosa.¹⁴ In transverse sections of an intussuscepted bowel, the juxtaposition of the wall layers of the intussusceptum and intussuscipiens creates more than five hyperechoic and hypoechoic concentric rings. These concentric rings are usually indistinctly visible in vivo, probably because of compression of the mucosal and serosal surfaces and the presence of bowel edema.¹⁷ Therefore, the most commonly described ultrasonographic pattern consists of two or three thick, concentric rings representing the hypoechoic edema and the hyperechoic compressed serosal and mucosal layers. The detection of more than five thin hyperechoic and hypoechoic concentric rings might suggest a lack of edema and compression of the intestinal walls and possibly the presence of an uncomplicated and reducible intussusception. Two of four reducible intussusceptions in the present study had multiple, thin, well-differentiated concentric rings. However, in the remaining two reducible cases and in all nonreducible cases, two or three, thick, poorly differentiated concentric rings were observed. In humans, the thicker and more hypoechoic the peripheral rings, the more complicated the intussusceptions, whereas the presence of well-differentiated wall layers is usually associated with loose and reducible intussusceptions.¹⁸ The authors suggest that the lack of clear intestinal wall layers should not exclude the presence of uncomplicated intussusceptions.⁶ In cases with clearly delineated layers in the wall of the intussusceptum and intussuscipiens, a reducible intussusception should be expected.

In cases of a very long intussusception, a double target-like sign and an hourglass configuration were observed in transverse sections. These signs appear when scanning two different parts of the intussusception located close to each other, because of the curvature of the intussusception caused by the tension of the mesentery.

Multiple parallel layers in longitudinal sections of the intestine are the most common described patterns in both veterinary and human literature.⁶⁹¹⁹ Other signs including a mass, a pseudokidney mass, an hourglass, a fused target-like configuration, a hayfork, and a trident have been reported in humans.^15172021 Multiple hyperechoic and hypoechoic parallel lines were the predominating ultrasonographic patterns in longitudinal sections of the present study. However, a mass in two dogs, a kidney-like mass in one dog, and a trident-like configuration in three dogs have also been detected in this study. In humans, a trident-like configuration is a sign of uncomplicated or loose intussusception, and an attempt to reduce the intussusception with an enema before surgery is strongly recommended in such cases.¹⁸ In the present study, all trident-like configurations were associated with reducible intussusceptions, and such a finding in animals may therefore signify a reducible intussusception. Since only three dogs showed the trident sign, further investigations are needed to assess the significance of the trident sign in dogs. In humans, color and pulse-wave Doppler ultrasonography was used successfully to determine the viability and predict the reducibility of the intussuscepted bowel.²² Similar studies in animals have yet to be reported.

In the present study, the ultrasonographic findings were considered diagnostic of intestinal intussusception when a concentric ring sign, multiple parallel lines, or both were detected in transverse and longitudinal sections, respectively. It should be noted that these patterns are not specific for intussusception.⁶¹²¹⁷ Bowel diseases that cause bowel wall thickening such as inflammation, edema, hematoma, primary or metastatic tumors, and even normal bowel and normal lymph nodes may simulate this appearance.^8172324 Observation of more than five wall layers and the semilunar echogenic mesentery within the intussusception have been reported as extremely useful differentiating features in transverse sections.⁸¹² The findings of this study support these reports. The authors suggest that the recognition of the inner intussusceptum in the center of a target-like mass could be another useful, differentiating feature of the intestinal intussusception in transverse sections.

An enlarged mesenteric lymph node may mimic the intussusception pattern;²⁴ however, the ultrasonographic signs in the false-positive case in the present study did not resemble those of the enlarged mesenteric lymph node. The authors suggest that spontaneous reduction of the intussusception had occurred rather than a failure to recognize lymphadenomegaly on the initial ultrasonographic examination.

The process of confirming a disease requires a test with a high specificity.¹³ Although the number of the dogs in this series is not large enough to allow firm conclusions, the high specificity (97.8%) in this study confirms the results obtained in humans, where specificity was reported to range from 88% to 100%.⁷⁹ However, in young animals, the paucity of bowel diseases that may be mistaken for an intussusception may have increased the specificity of ultrasonography in the present study. Moreover, in dogs that did not have surgery (negative group), the accuracy of the ultrasonographic examination was confirmed by the follow-up of the cases. This could not exclude the presence of an intussusception that was missed by ultrasound examination and spontaneously reduced after the examination. However, spontaneous reduction of intussusceptions is not common in small animals, and if it had occurred in this study, it would have slightly decreased the sensitivity value. It is the opinion of the authors of the present study that ultrasonography may provide a sensitive and specific method of diagnosing intestinal intussusception in young dogs despite the weaknesses inherent to the statistical analysis.

Conclusion

Ultrasonography was a sensitive, specific, and accurate method for the diagnosis of intestinal intussusception in young dogs. Transverse and longitudinal sections were necessary for an accurate diagnosis. The ultrasonographic appearance of the intussusception varied. In transverse sections, the target-like mass consisting of a hyperechoic or anechoic center surrounded by well or poorly differentiated concentric rings was observed. In longitudinal sections, the multiple hyperechoic and hypoechoic parallel lines, the tumor-like mass, the pseudokidney mass, and the trident-like configuration were all documented. The trident configuration may suggest an uncomplicated and reducible intussusception. However, the reducibility and viability of the intussusception should be investigated further with Doppler ultrasonography.