Spontaneous Cholecystocutaneous Fistula in a Dog
The purpose of this case report was to describe the surgical correction of a cholecystocutaneous fistula in a dog. A 6 yr old vizsla presented with a 2 mo history of a chronic draining wound on the right ventral thorax. Diagnostics revealed numerous fistulous tracts opening at a single site on the right ventrolateral chest wall, extending caudodorsally through the chest wall and diaphragm to the region of the right medial liver lobe. Exploratory laparotomy revealed the apex of the gallbladder adhered to the diaphragm with a tract of fibrous tissue extending along the diaphragm laterally to the right thoracic wall. Cholecystectomy was performed. The fistulous tract was incised to expose the lumen of the fistula, and the fistula was omentalized. Twenty-eight months after surgery, the dog had had no recurrence of the fistulous tract. Exploratory laparotomy allowed excellent visualization of the intra-abdominal path of the fistula and facilitated the ease of resection of the source. Cholecystectomy resulted in rapid and complete resolution of the fistula without the need for excision of the fistula. Although rare, gallbladder disease should be a differential for chronic fistulous tracts.
Introduction
Spontaneous cholecystocutaneous fistula is a rare complication of biliary tract disease in humans. It has been reported most commonly in geriatric female patients. The condition was first described in humans in 1670.1–3 In recent history, the incidence of cholecystocutaneous fistula formation in humans has markedly decreased, largely due to the advent of antibiotics, the early diagnosis and treatment of biliary disease, and improved surgical technique. To the authors’ knowledge, only one recent report of spontaneous cholecystocutaneous fistula exists in the veterinary literature.4 A single case of an umbilicobiliary fistula (considered to be a congenital condition) has also been reported in a dog.5 Therefore, because of the paucity of information in the veterinary literature, most of the understanding of this rare condition has been extrapolated from human medicine. Given the lack of information on diagnosing and treating cholecystocutaneous fistulas, this report describes the successful treatment and resolution of a cholecystocutaneous fistula in a dog following cholecystectomy and omentalization of the fistulous tract.
Case Report
A 6 yr old intact female vizsla weighing 21.4 kg was referred for evaluation of a chronic draining wound on the right thorax. The wound had been present for 2 mo. Palpation of the wound suggested a fistulous tract that extended to, but did not penetrate, the chest wall. Prior to referral, the tract had been irrigated and surgically debrided, followed by placement of a Penrose drain. Purulent exudate had been observed draining from the wound, but no foreign body was found during debridement. Aerobic culture of the wound had revealed Escherichia coli. The dog was treated with cephalexina (24 mg/kg per os [PO] q 8 hr for 14 days), firocoxibb, tramadolc, and sulfadimethoxine/ormetoprimd (27 mg/kg PO q 24 hr once, then 13 mg/kg PO q 24 hr for 10 days).
At the time of referral, physical examination revealed a nonpainful 0.5 cm opening surrounded by firm granulation tissue on the right caudal ventrolateral thorax with purulent discharge. No other abnormalities were observed. The owner reported a “honey-like” fluid exuding from the wound on various occasions. Results of a complete blood count and serum biochemistry analysis were within normal limits. Thoracic radiographs revealed a convex soft-tissue opacity dorsal to the second and third sternebrae consistent with mild sternal lymphadenopathy. There was also a soft-tissue mass seen arising from the ventrocentral aspect of the cupula of the diaphragm (Figures 1A, B). Possible differential diagnoses for the diaphragmatic mass included either a membranous diaphragmatic defect with herniation of liver or diaphragmatic neoplasia.
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
The dog was anesthetized for a fistulogram and possible CT scan. Prior to induction of anesthesia, the dog was administered glycopyrrolatee (0.01 mg/kg intramuscularly [IM]), xylazinef (0.5 mg/kg IM), and morphineg (0.5 mg/kg IM). Anesthesia was induced with propofolh (6 mg/kg IV to effect). The dog was intubated, and anesthesia was maintained with isofluranei and oxygen. Normosol-R was administered IV at a rate of 10 mL/kg/hr until completion of the procedure.
A 3.5-French tomcat catheter was primed with ionic iodinated contrast mediumj diluted with sterile saline to a total iodine concentration of 185 mg/mL. The catheter tip was inserted into the external fistulous opening of the draining tract located at the level of the right fifth intercostal space just dorsolateral to the sternum. Approximately 6 mL of the dilute contrast medium was injected while pressure was applied around the wound to close the opening. Radiographs of the thorax were immediately obtained following contrast medium administration. The fistulogram revealed multiple, variably sized, irregularly marginated, tortuous draining tracts involving the superficial soft tissues of the right ventrolateral thoracic wall. The fistulous tracts extended from the level of the fifth intercostal space to the tenth rib in the lateral view and from the level of the fifth intercostal space to the first lumbar vertebra in the ventrodorsal view (Figures 2A, B). In the lateral view, the fistulous tracts, which were filled with contrast medium, were seen at the same level as the previously described diaphragmatic mass, raising the suspicion of hepatic involvement. There was no obvious evidence of communication of the draining tracts with the thorax; however, communication with the peritoneal space could not be ruled out.
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
Following fistulography, a CT examination of the affected region was performed to further delineate the paths of the fistulous tracts. Contiguous 5 mm and 1 mm axial images were obtained from the midthorax to the cranial abdomen after injecting an additional 12 mL of the dilute contrast medium into the fistulous opening. The fistulous tracts extended from the right ventrolateral chest wall at the level of the cupola of the diaphragm, cranially to the level of the heart, and caudally to the level of the stomach (Figure 3). The contrast medium appeared to dissect between the subcutaneous and intramuscular tissues of the chest wall, but did not communicate with the pleural space. A small focal accumulation of contrast medium was observed within the right ventral aspect of the liver and was connected to the area of subcutaneous contrast medium by a thin line of contrast medium in one image (Figure 4). Contrast medium was never seen entering any other part of the liver or any identifiable part of the biliary tract. Differential diagnoses now included a chronic inflammatory tract, migrating foreign body resulting in a liver abscess, hematogenous liver abscess with fistulation, and, less likely, chronic penetrating trauma.
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
Abdominal ultrasonographic examination identified the tract and suggested the tract extended through the caudal margin of the right caudal lung lobe and into the right medial liver lobe. An exploratory laparotomy was recommended to further characterize the nature and source of the fistula. The dog was induced and maintained under general anesthesia as described above. During surgery, the dog received cefazolink (20 mg/kg IV q 1.5 hr), hydromorphonel (0.5 mg/kg IV twice), and dobutaminem (2.5–5 μg/kg/hr IV via constant rate infusion). The dog was placed in dorsal recumbency, and the entire ventral abdomen and ventral and right hemithorax were aseptically prepared for surgery and draped. A 30 cm ventral midline incision was made extending from the xyphoid process to the pelvis. Upon entering the abdomen, the fistulous tract was identified lateral to the liver along the right abdominal wall (Figure 5). The liver was investigated to determine the source of the fistulous tract and was found to be unremarkable. The gallbladder was of normal size, but the gallbladder wall was densely opaque, thickened, and the apex adhered to the right ventral aspect of the diaphragm between the pars sternalis and right pars costalis (Figure 6). Numerous omental adhesions were observed along the gallbladder and common bile duct (Figures 7A, B). The common bile duct was normal in size and appearance once the adhesions were removed. The gallbladder expressed easily, and no leakage was observed from the common bile duct. A tract of fibrous tissue could be followed from the apex of the gallbladder adhesion, along the diaphragm laterally to the right thoracic wall, and extending cranially. There were no other adhesions of the liver to the diaphragm. No choleliths were palpated at the time of surgery. Because the diagnostic imaging suggested the fistulous tract extended through the caudal aspect of the right caudal lung lobe and that assertion could not be confirmed or denied from the abdominal approach, the ventral skin incision was extended cranially and the thoracic cavity was entered by a standard median sternotomy of the caudal third of the sternum. The thoracic cavity and right caudal lung lobe were normal.
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
Citation: Journal of the American Animal Hospital Association 48, 1; 10.5326/JAAHA-MS-715
After bluntly and sharply separating the apex of the gallbladder from the diaphragm, a cholecystectomy was performed by bluntly dissecting the gallbladder from the cystic fossa and transecting the neck of the gallbladder after occlusion of the neck with a medium Hemoclipn. The fistulous tract along the diaphragm was incised to expose the lumen of the fistula so the omentum could adhere to the inflamed tissue, drain any residual bile, and aid in the reparative process. Culture swabs were taken of the fistulous tract and gallbladder during surgery and submitted for aerobic and anaerobic culture and sensitivity. The gallbladder was submitted for histopathologic examination. The abdomen was explored for other abnormalities and was found to be unremarkable. A parasternal 8-French thoracostomy tube was placed to evacuate air from the thoracic cavity after surgery. The skin and subcuticular tissue surrounding the opening of the fistula were left intact with the intent that removal of the source of irritation (i.e., the draining bile) would allow resolution of the fistulous tract and skin wound. Omentum was placed over the opened fistulous tract, and the abdomen was closed routinely. The dog recovered uneventfully from anesthesia.
Postoperative Care
The dog's recovery was monitored in the intensive care unit. Normosol-R was administered at a maintenance rate of 60 mL/kg/day. The thoracostomy tube was sterilely evacuated q 4 hr followed by injection of 10 mg of 0.5 mg/kg bupivicaineo diluted 1:9 with sodium bicarbonatep. The thoracostomy tube was removed after 24 hr. Pain was managed with a fentanylq/lidocainer/ketamines IV constant rate infusion at a rate of 10 mL/hr (based on 5 μg/kg/hr of fentanyl) for the first 48 hr, then tramadolc (3.5 mg/kg PO q 12 hr). Cefazolin was continued (20 mg/kg IV q 8 hr for 24 hr) and then transitioned to cefpodoximet (10 mg/kg PO q 24 hr). The dog was discharged 72 hr after surgery with tramadolc (3.5 mg/kg PO q 12 hr for 5 days) and cefpodoximet (200 mg PO q 24 hr for 28 days).
Outcome
Histopathologic examination of the gallbladder revealed cholecystitis with lymphoplasmacytic and regionally neutrophilic inflammation with distinct lymphoid aggregate formation and marked serosal fibrosis. Some sections of the gallbladder were lined by fibrin strands admixed with erythrocytes on the lateral smooth muscle margins, consistent with a tract formation. In the opinion of the pathologist, the inflammation was not considered severe enough to cause the gallbladder to rupture. The cause of the rupture and subsequent fistula could not be determined. The fistulous tract was not biopsied because the tract was considered a result of the gallbladder disease. Therefore, it was thought that histologic examination of the tract would have contributed little to the case management. Aerobic and anaerobic cultures of the tract and gallbladder revealed E. coli and Enterococcus spp., and both were susceptible to cefpodoxime (200 mg PO q 24 hr for 28 days). The owner reported the fistula resolved within a few days after discharge from the hospital. Twenty-eight months after surgery, the owner reported the dog was completely normal in all respects and had had no recurrence of the fistulous tract.
Discussion
Spontaneous cholecystocutaneous fistula in humans is most commonly associated with cholecystitis secondary to cholelithiasis and only rarely with carcinoma.1 There has been one documented case in the human medical literature of idiopathic cholecystocutaneous fistula.3 Cholecystitis and cholelithiasis are rare in dogs, and only one other documented case of a spontaneous cholecystocutaneous fistula has been reported.4,6 Typically, the inflammation and distention of the gallbladder associated with the formation of choleliths predispose the gallbladder to adhesion to the peritoneum and abdominal wall.1 Adhesions occur most often at the fundus of the gallbladder.1 If the gallbladder ruptures, an abscess can occur and form a fistulous tract that eventually exits the skin.2 The tract is usually a single tract and occurs most frequently either in the right cranioventral quadrant of the abdomen (48%) or at the umbilicus (27%).3 Occasionally, the tract can occur elsewhere such as the groin, dorsum, or thoracic wall, which was the presentation in this dog. Discharge from the fistula can be pus, bile, or a combination of both. Rarely, gallstones may be present.7
Cholecystocutaneous fistula is usually associated with a history of chronic biliary disease or an acute episode of cholecystitis. On occasion, it occurs with no history typical of biliary disease.8 In the case described herein, there was no history of previous health problems. Given the appearance of the fistulous lesions, they can be confused with pyogenic granuloma, inclusion cysts, or metastatic carcinoma.1,9 Physical examination findings in humans are highly variable and will depend on overall health and underlying health problems. Identification of bile or gallstones within the fistulous tract allows a definitive diagnosis; however, this scenario is rare, and further diagnostics are usually required. In the dog of this report, a “honey-like” fluid had been seen on several occasions prior to referral. Although initially of unknown and unappreciated significance, in retrospect, this fluid was likely bile, which should have increased the index of suspicion for a cholecystocutaneous fistula.
Results of a complete blood count and serum biochemistry will also vary and can be within reference limits. Alkaline phosphatase can be elevated depending on chronicity.1 Ultrasound may demonstrate a mass with mixed echogenicity adjacent to the abdominal wall, but separate from the gallbladder.2 Rarely, gallstones can be seen within the mass. The path of the fistulous tract can be delineated with a fistulogram. Water-soluble, iodinated contrast material is recommended.2 In the dog in this report, CT scan was performed following injection of iodinated contrast to further define the path of fistula. CT imaging of the fistula was very helpful in identifying the region of the source of the fistula; however, the CT failed to definitively rule in or out the passage of the fistulous tract through the caudal aspect of the right caudal lung lobe. Why CT fistulography was not completely sufficient for identifying the path of the fistula is unknown, but the authors speculate that the small diameter of the fistula and/or an insufficient volume of contrast in the fistula lumen did not allow adequate resolution to visualize the fistula along its entire length. Had the authors been able to confidently affirm that the tract did not violate the pleural space and lung by CT, the authors would not have relied as heavily on the ultrasonographic findings that suggested the tract passed through the right caudal lung lobe. This would have avoided the caudal median sternotomy.
In humans, the treatment of choice will depend on the underlying pathophysiology. Acute inflammatory episodes are treated with broad-spectrum antibiotics pending culture and incisional drainage of the fistula. Once the drainage has decreased, cholecystectomy and excision of the fistula are warranted; however, spontaneous healing has been reported in as many as 20% of cases in humans following conservative management.2,9 Given the paucity of cases of cholecystocutaneous fistula reported in the veterinary literature and the likelihood that differences in gallbladder disease exist between humans and dogs, caution should be exercised when extrapolating treatment methods for humans to dogs. For example, large amounts of biliary drainage in humans indicates a large fistula and common bile duct obstruction, which should be treated immediately by cholecystectomy and common bile duct exploration.2,9 Yet in both cases reported in the veterinary literature, the common bile duct was patent, and, in one of the cases, the fistula resolved without cholecystectomy. The authors of this report chose to open the fistula along its diaphragmatic path because the source of the fistula (i.e., the gallbladder) was removed, the fistula diameter was small, and omentalizing the opened fistula was considered a more viable method than resection for obliterating the fistula while minimizing tissue trauma and operative morbidity. Although considered by the authors to be highly unlikely, had the fistula contained neoplastic cells or been epithelialized, leaving the tract in place would have been inappropriate. Given the extremely rare nature of this condition, a clear treatment regimen with aerobic and anaerobic culture is always indicated to provide proper antimicrobial therapy.
In the case recently reported, the fistulous tract was surgically explored to the origin of the tract at the gallbladder.4 In the dog described in this report, the authors chose to approach the region of the source of the bile leakage directly through a ventral midline celiotomy because the location of the source was not established on the CT and fistulogram. Approaching the source allowed excellent visualization of the intra-abdominal path of the fistula and facilitated the ease of resection of the source (i.e., the gallbladder). Interestingly, in the previous report, significant hemorrhage occurred during excision of the fistulous tract.4 It is unknown whether similar hemorrhage would have occurred if a similar surgical approach had been selected in the dog described in this report, but the distinct advantages offered by directly approaching the source of the bile leakage, combined with the rapid and complete resolution of the fistula following cholecystectomy, support the authors decision to avoid excision of the fistula and fistulous tract.
Why the fistulous tract exited the ventrolateral chest after beginning within the abdomen is unknown, but many fistulous tracts tend to pursue “a path of least resistance.” It may be that the diaphragm provides a surface for fistulous tract development, and the tract then dissects through the thoracic wall where the diaphragm joins the wall, then continues superficial to the ribs and ventrolaterally to exit the skin at a somewhat ventral point on the chest wall. Why the fistulous tract exited the ventral chest wall at the midpoint of the thorax instead of dissecting further cranially is unknown.
Conclusion
Cholecystic disease is a very rare cause for cutaneous fistulation in the dog. Of the two reported cases, the ventrolateral thorax appears to be the typical site of the fistula. In the case of this report, a clear, amber fluid, typical of bile, was observed and should increase the diagnostic suspicion for cholecystocutaneous fistulation. Routine fistulography and CT were sufficient to identify the course of the fistula. Exploratory laparotomy identified the source as the gall bladder. Cholecystectomy and omentalization of the fistulous tract was curative. Excision of the fistulous tract potentially incurs unwarranted surgical morbidity and is seemingly unnecessary.
Lateral (A) and ventrodorsal (B) thoracic radiographs showing a slight enlargement of the sternal lymph node (black arrows) and a large soft-tissue mass seen arising from the ventrocentral aspect of the cupula of the diaphragm (white arrow).
Ventrodorsal (A) and right lateral (B) radiographs obtained during the fistulogram. Multiple irregularly shaped and wide-ranging draining tracts were identified in the right ventrolateral chest wall, extending from the level of the fifth intercostal space to the tenth rib in the lateral view and from the level of the fifth intercostal space to the first lumbar vertebra in the ventrodorsal view.
The transverse CT scan shows a large irregular accumulation of iodinated contrast present in the subcutaneous tissue of the right ventrolateral chest wall at the level of the cupola of the diaphragm. The injection catheter can be seen as a line in the contrast (white arrow).
The transverse CT scan shows a small focal accumulation of contrast present within the liver in the region of the gallbladder. A faint line (black arrows) connects the gallbladder to the accumulated contrast agent in the subcutaneous space and represents the fistula. The fistula can be seen extending from the subcutaneous pool of contrast (dotted arrow) to the apex of the gallbladder (white arrow).
Photograph of the initial view of the fistulous tract (white arrow) and liver during surgery.
Photograph of the apex of the gallbladder adhered to the right ventral aspect of the diaphragm (white arrows).
Photographs of the numerous omental adhesions along the gallbladder (A, white arrows) and common bile duct (B, white arrow).
Contributor Notes
J. Johnson-Neitman's present affiliation is VCA Alameda East Veterinary Hospital, Denver, CO.
