Enterocutaneous Fistulae in Horses: 18 Cases (1964 to 1992)

ORIGINAL RESEARCH Veterinary Surgery 23: 167-1 7 I , 1994 Enterocutaneous Fistulae in Horses: 18 Cases ( 1964 to 1992) DAVID G. BRISTOL, DVM, Diplomate ABVP, Diplomate ACVS Enterocutaneous fistulae are rare in horses and occur most commonly as a complication of umbilical hernias or their treatment. Horses with enterocutaneous fistulae may be successfully treated by en bloc resection of the body wall and intestine or by allowing second intention healing. Complications associated with surgical intervention include fever, colic, incisional prob- lems, and recurrence of the fistula. Nonsurgical management of two horses with presumptive large colon fistulae resulted in resolution of the fistulae without complications. @Copyright 1994 by The American College of’ Veterinary Surgeons NTEROCUTANEOUS fistulae occur rarely in E horses; only six cases have been reported in the North American l i terat~re.’-~ These were associated with Richter’s hernias at the umbilicus (two horses), ’ trauma (two horses),* injection of an irritant around an umbilical hernia (one h ~ r s e ) , ~ and application of a hernia clamp (one h ~ r s e ) . ~ The portion of the digestive tract involved was defined at surgery in five horses, and included the pelvic flexure (one horse),‘ cecum (one h o r ~ e ) , ~ ventral colon (two horses) (M.D. Markel, personal communication, January 1992),’ and distal jejunum (one horse) (M.D. Markel, per- sonal communication, January 1992). The portion of the intestine involved was not identified in one horse because the peritoneal cavity was not opened; however, it was thought that the large colon was in- volved because of the location of the injury.’ Surgical correction of enterocutaneous fistulae was attempted with variable success in the previously re- ported cases. These attempts were associated with considerable postoperative morbidity and prolonged hospitalization. A lactating mare had multiple fis- tulae from the pelvic flexure through the caudal ven- tral abdominal wall and into the parenchyma of the mammary gland. The pelvic flexure had herniated through a rent in the prepubic tendon dorsal to the mammary gland. At necropsy, in addition to the cutaneous fistulae, fistulae into the parenchyma of the mammary gland were identified. Because of the extensive nature of the abnormalities, the mare was euthanatized during surgery.2 Another horse was re- ported to have a residual “bulge” at the surgical site when it was examined 6 months after surgery. Both horses with Richter’s hernias developed subcuta- neous infections after surgery. The two horses that recovered from surgery without reported compli- cations had hospitalization periods of 7 and 2 1 days, respective~y.~.~ The successful management by nonsurgical means of two horses with enterocutaneous fistulae prompted this retrospective study in an attempt to better define this disorder. MATERIALS AND METHODS The Veterinary Medical Data Base (VMDB) at Purdue University, West Lafayette, IN, was used to identify horses with enterocutaneous fistulae that had been admitted to From the Department of Food Animal and Equine Medicine, College of Veterinary Medicine. North Carolina State University, Address reprint requests to David G. Bristol, DVM, Diplomate ABVP. Diplomate ACVS, Department of Food Animal and OCopyright 1994 by The American College of Veterinary Surgeons Raleigh, NC. Equine Medicine, College of Veterinary Medicine, North Carolina State University, Raleigh, NC 27606. 0 16 I -349919412303-0002$3.00/0 167 168 ENTEROCUTANEOUS FISTULAE IN HORSES institutions participating in that database. Records that described fistulae of portions of the digestive tract outside the peritoneal cavity (mouth, salivary glands, esophagus, and rectum) or intentionally created fistulae were excluded from this study. In addition, five clinicians identified six horses with enterocutaneous fistulae, the records for which had not been entered into the VMDB. Data from these six records were also included in this report. RESULTS The VMDB identified 357,67 1 equine medical records for review: 12 were records of horses with enterocutaneous fistulae involving segments of the intestine normally located within the peritoneal cav- ity. Including the 6 case records identified above, the medical records of 18 horses were available for study. Eight horses had fistulae involving the small intes- tine and 10 horses had fistulae involving the large intestine. Signalment Young horses were most commonly affected. Eight horses were younger than 1 -year-old, three were yearlings, two were 2-years-old, and five were 3-years-old. Nine quarter horses, four standardbreds, two Arabians, one Thoroughbred, one Shetland pony, and one Warmblood were affected. History Sixteen horses had fistulae that involved umbilical hernias; 8 of these horses had a history of increased swelling at the umbilicus I to 3 weeks before ad- mission. The swelling was followed by leakage of ingesta in 9 to 18 days. Often, purulent material drained from the umbilicus before leakage of ingesta occurred. One horse developed drainage from its umbilical hernia without premonitory signs. Four horses had umbilical masses lanced, and 2 horses developed leakage of ingesta after sloughing tissue strangulated by hernia clamps. The remaining horse had trauma to its umbilical hernia with subsequent leakage of ingesta. Two horses had fistulae unrelated to umbilical hernias. One had a history of previous ventral me- dian celiotomies for colic with subsequent abscess formation at the incision line. The other horse had no known inciting cause before development of a fistula in the right flank. The fistulae of 14 horses were present from 1 day to 1 month before admission to veterinary hospitals. One horse had a fistula of 1.5 years duration before admission and the duration of three fistulae was unknown. Physical Examination and Laboratory Data Eight of 10 horses with fistulae involving the large intestine were otherwise normal on physical exam- ination. One horse had a respiratory tract infection. One horse that had a fistula secondary to infection of a celiotomy incision had intermittent clinical signs of colic. Eight horses had fistulae of the distal jejunum or ileum; clinical signs in these horses were variable. Six horses appeared otherwise normal and two horses with ileal fistulae were in poor condition and were dehydrated. One of these horses had an increased heart rate (80 beats per minute), injected mucous membranes, and a prolonged capillary refill time. Complete blood cell counts were obtained for 10 horses before surgery. The leukocyte count ranged from 7,400 to 29,500 cells/pL (mean = 15,030 cells/ pL). Neutrophil counts were noted in 9 records and ranged from 2,150 to 27,435 cells/pL (mean = 1 1,17 1 neutrophils/pL). Six of these horses had neu- trophil counts > 6,000 cells/pL. Blood gas and elec- trolyte analyses were completed on 3 horses before surgery; all had fistulae of the ileum. Two horses had severe acidemia (average pH = 7.04, PC02 = 19 mm Hg, HC03 = 6.6 mEq/L, base excess (BE) = - 17.5 mEq/L) with hyponatremia ( 104 mEq/L), and hypochloremia (89.5 mEq/L). Blood gas and elec- trolyte values were normal in the other horse. Treatment und Outcome Two horses with presumptive large intestinal fis- tulae were treated without surgical intervention. A diagnosis of a large intestinal fistula in these two horses was based on digital palpation through the fistula and characteristics of the discharged ingesta. One horse was not hospitalized, whereas the other was hospitalized for 4 days to evaluate a concurrent respiratory tract infection. The owners of both horses were instructed to observe the horses closely for clin- ical signs of colic and to apply petroleum jelly around the fistulae to prevent scalding of the skin. The fis- tulae healed within 1 month of the date of exami- DAVID G. BRISTOL 169 nation without further treatment and without com- plications. One horse was shown for 3 years before being sold. The owner of the other horse could not be contacted for long-term follow-up. Sixteen horses were treated by surgical exploration and attempted closure of the fistulae. Closure tech- nique varied depending on the size of the fistula, the segment of bowel involved, and the surgeon’s pref- erence. Surgical closure of fistulae usually included oversewing the external defect, followed by an ellip- tical incision through the skin surrounding the body wall defect. Sharp dissection was used to enter the abdominal cavity. Several surgical reports included comments about thick fibrous tissue present in the body wall surrounding the fistulae. In some horses, adhesions had to be broken down to identify the bowel segment involved in the fistula. Intestinal re- pair was performed by one of three methods. In 3 horses, the bowel was dissected from the body wall and the intestinal defect simply oversewn (two small intestinal fistulae and one large intestinal fistula). In 7 horses, an incision through the bowel wall was made around the fistulated portion. The resulting enterotomy was then closed (one small intestinal fis- tula and six large intestinal fistulae). In 4 horses, small intestinal fistulae were removed by en bloc resection and the bowel was anastomosed by an end- to-end method. One horse had its small intestinal fistula removed with closure of the resultant enter- otomy, but also required resection of adhered intes- tine, followed by intestinal anastomosis. Horses undergoing surgery were hospitalized for 0 to 36 days (18.3 * 1 1.0 days [mean t SD]). Com- plications associated with anesthesia and surgery in- cluded intraoperative death ( 1 horse), severe cardiac arrhythmia successfully treated by open chest heart massage ( 1 horse), fever (2 horses), hemorrhage from the incision site resulting in anemia ( 1 horse), re- currence of the fistula ( 1 horse), colic resulting in a second surgery ( 1 horse), and jugular phlebitis ( I horse). Incisional infection, discharge, or abscess formation occurred in 4 horses. Dehiscence requiring immediate surgery occurred in one horse. Three months later, this horse had a third surgery to repair a ventral hernia at the surgical site. Three horses died while hospitalized: 1 under anesthesia, 1 of peritonitis caused by a leaking anas- tomosis, and 1 of electrolyte disturbances when the fistula recurred after surgery. The horse that died under anesthesia had a fistula of the right flank. At necropsy, the cecum and right colon was adhered to the body wall around the area of the fistula. A di- agnosis of “purulent granuloma initiated by rupture of a strongyle abscess” was made. Thirteen of 16 horses that had surgery were discharged, but 2 horses died within I month of discharge. One horse had severe colic and was euthanatized, and I horse rup- tured its intestine at the anastomosis site. Long-term follow-up (22 years) was available for 7 of the 1 1 horses that survived > 1 month after surgery: no ad- ditional complications occurred. DISCUSSION Enterocutaneous fistulae are rare in horses, with most fistulae occurring as a complication of an um- bilical hernia. Incarceration of part of the circum- ference of the bowel (Richter’s hernia) may result in necrosis of the herniated segment with subsequent abscess and fistula formation. Iatrogenic causes of enterocutaneous fistulae included misapplication of hernia clamps and lancing bowel in what was pre- sumed to be an umbilical abscess. Careful evaluation of umbilical masses by palpation, aspiration, and ultrasonography can prevent these errors. In most cases, surgical repair of enterocutaneous fistulae is not an emergency procedure, and imme- diate attempts at repair may not be indicated. In humans with enterocutaneous fistulae, 6 to 8 weeks of medical therapy is recommended before attempt- ing surgical intervention. Longer periods may benefit patients with complex fistulae.’ The fact that a horse appears in good health, despite the prcsence of a fistula, confirms that intra-abdominal adhesions have isolated the leaking ingesta from the peritoneal cavity. Two horses in this study healed their fistulae without surgical intervention and others may have healed had surgery not been attempted. However, surgery may be indicated if the fistula does not close spontaneously or if the fistula results in metabolic or physical deterioration of the animal patient. Whereas it was not stated in the medical record, it is presumed that the horse with a fistula of 18- months duration had healed the intestinal mucosa to the skin. Once a fistulae has epithelialized, spon- taneous closure will not occur.’ Similar “permanent” fistulae of the rumen in cattle can be closed by de- briding the edges of the fistula and allowing second 170 ENTEROCUTANEOUS FISTULAE IN HORSES intention healing to occur. It is important not to debride the wound edges beyond the area where the viscus is adhered to the body wall, or contamination of the peritoneal cavity may occur. In humans, most enterocutaneous fistulae occur as complications of abdominal surgery.6 Small-in- testinal fistulae are treated by maintenance of fluid and electrolyte balance and nutritional management. Spontaneous healing without surgical intervention has occurred in 29% to 70% of cases6 Colonic fistulae in people usually do not result in fluid and electrolyte abnormalities, but local infection is common.6 An- tibiotics and low residue diets are recommended, and most fistulae heal without surgical intervention.6 However, therapy must be individualized for each patient. “End fistulae” occur when there is loss of patency of the bowel lumen distal to the fistula, as occurs with strictures or obstruction.’ Loss of distal luminal patency forces all intestinal contents to exit through the fistula. This type of fistula requires sur- gical intervention for closure and to re-establish in- testinal continuity.’ Prognosis for humans with enterocutaneous fis- tulae depends on the portion of intestine involved, the volume of fluid escaping through the fistula, the presence of concurrent disease, most notably neo- plasia, and the presence of sepsis.’ More proximal intestinal fistulae tend to have the greatest output and are associated with the most severe fluid and electrolyte imbalances5 Sepsis may result in multiple organ failure and is the most common cause of death in humans with enterocutaneous f i~ tu lae .~ Five horses in the current report died of complications associated with their fistulae. Three of these deaths were attributed to problems at anastomosis sites, one death was caused by undiagnosed colic, and one death was caused by anesthetic complications. Recent advances in the treatment of enterocuta- neous fistulae in humans include the use of extra- peritoneal approaches to close fistulae, total paren- teral nutrition, and administration of somatostatin analogs. Extraperitoneal approaches use tissue flaps to close the enteric defect without entering the ab- dominal Total parenteral nutrition has been used to maintain nutritional support while resting the bowel.5 Octreotide acetate, a somatostatin ana- log, is used to decrease production of gastrointestinal hormones and to decrease net intestinal fluid secre- tion.’.‘’ Whereas acid-base and electrolyte variables were not evaluated in most horses in this study, severe abnormalities were recorded in two horses. CoIlec- tion of acid-base and electrolyte data was obviously biased by the attending clinician’s perceived need for this data. The severe acidemia present in two horses was presumably caused by loss of alkaline intestinal contents from the ileum and severe de- hydration. The low number of case records with this data precludes conclusions regarding blood gas changes when fistulae occur in other portions of the gastrointestinal tract. The acid-base and electrolyte status of the animal patient would be expected to vary depending on the loss of intestinal fluid, which would in part be related to the size of the fistula, its duration, and the segment of intestine involved. Surgery to correct enterocutaneous fistulae in horses can be complicated by intra-abdominal adhesions. Attempts to disrupt adhesions can cause intestinal rupture. The existence of adhesions may play a role in the occurrence of colic after surgery. The owners of horses that have surgery to correct enterocutaneous fistulae should be warned that the horse is at risk of postoperative colic, and that surgery for colic may be associated with a higher incidence of complications because of pre-existing intra-ab- dominal adhesions. Therapy for horses with en- terocutaneous fistulae should be individualized, based on the location of the fistula, the animal’s metabolic status, and the observed responses to medical therapy. ACKNOWLEDGMENT 1 wish to acknowledge the assistance of the following individuals in reviewing case records: Drs Gordon J. Baker, Alicia Bertone, Mike Collier, William J. Don- awick, John F. Fessler, Peter Fretz, Susan L. Fubini, Larry Galuppo, Dallas Goble, Eleanor M. Green. Neil Hooper, Robert Hunt, J.D. Lillich, Mike Livesey, Eric Reinertson, John Stick, Gayle Trotter, and Ava Trent. REFERENCES I . Markel MD, Pascoe JR, S a m AE: Strangulated umbilical hernias in horses: 13 cases (1974-1985). J Am Vet Med 2. Bailey JV, Fretz PB: Attempts at surgical correction of un- usual colonic fistulae in the horse. Can Vet J 24:222-223, 1983 ASOC I90:692-694, 1987 DAVID G. BRISTOL 171 3. Adams OR: External colonic fistula in the horse. J Am Vet Med Assoc 145:29-3 1, 1964 4. Brown MP, Meagher DM: Repair of an equine cecal fistula caused by application of a hernia clamp. Vet Med Small Anim Clin 73:1403-1407, 1978 5. Rubelowsky J, Machiedo GW: Reoperative vs. conservative management for gastrointestinal fistulas, in Fry DE (ed): Surgical Clinics of North America. Philadelphia, PA, Saunders, I99 1, pp 147- 157 6. Schwartz SI: Complications, in Schwartz SI, Shires GT, Spencer FC, et al (eds): Principles of Surgery. New York, NY, McGraw-Hill, 1979, pp 5 18-520 7. Chitale VG: Enterocutaneous fistulas treated successfully by a turnover flap from surrounding tissue reinforced by musculocutaneous flaps and other local flaps. Plast Re- constr Surg91:1171-1172, 1993 8. Saref IJ, Jakowatz JG: Surgical treatment of enteric “bud’ fistulas in contaminated wounds. A riskless extraperitoneal method using split thickness skin grafts. Arch Surg 127: 9. Borison DI, Bloom AD, Pritchard TJ: Treatment of entero- cutaneous and colocutaneous fistulas with early surgery or somatostatin analog. Dis Colon Rectum 35:635-639, I992 10. Boike GM, Sightler SE, Averette HE: Treatment of small intestinal fistulas with octreotide, a somatostatin analog. J Surg Oncol 49:63-65, I99 1 1027-1030. 1992 Abstract of Current Literature Platysma Myocutaneous Flap for Head and Neck Reconstruction in Cats Smith MM, Shults S, Waldron DR, and Moon ML Head Neck 15:433-439, 1993 Squamous cell carcinoma is the fourth most common neoplasm in cats. As in man, local resective surgery of stage 111 and IV carcinoma often results in recurrence related to compromised margins. Radical resective procedures may be performed when reconstructive techniques are available to restore cosmesis and function. A platysma myocutaneous flap that was based on a cutaneous branch of the caudal auricular artery and vein was developed to fulfill this requirement. Control flaps, which included ligation and division of the caudal auricular artery and vein, were similarly developed on the contralateral aspect of the neck. Mean survival of all platysma myocutaneous flaps (86.7%), compared with control flaps (62.9%), was significantly different ( p < .05). Flaps grouped in lengths of 6, 9, and 12 cm had mean survival lengths of 93.896, 8 1.996, and 84.496, respectively. The mean survival length of flaps measuring 12 cm in length was significantly different ( p < .05) compared with flaps measuring 6 and 9 cm. On the basis of the results of this study, the platysma myocutaneous flap based on a cutaneous branch of the caudal auricular artery and vein may be a source of tissue for reconstructive procedures of the head and neck in cats. 01993 John Wile!, & Sons, Inc.