Diaphragmatic herniation in the horse: 31 cases from 2001–2006

Abstract

Diaphragmatic rent and visceral herniation in the horse is seldom diagnosed, but historically carries a poor prognosis. The objective of this study was to document the presentation and surgical management of all diaphragmatic rents as presented to 2 referral institutions over a 5-year period. A review of 31 cases demonstrated that even with advances in surgical management of abdominal and thoracic conditions, little has been done to change the prognosis for this condition. Success rate was 23% for all horses presented for colic and were ultimately diagnosed with a diaphragmatic hernia, and 46% for those cases for which surgical management was elected.

Introduction

A diaphragmatic rent is relatively uncommon in the horse, and is associated with a poor prognosis. One report noted that the incidence was 3 diaphragmatic hernias in 140 cases of laparotomies (1). Most commonly described as a diaphragmatic hernia, the term diaphragmatic rent, or defect is a more accurate description in congenital cases as it is occasionally discovered without visceral herniation. Determining the cause of the defect can be as challenging as making the diagnosis. Most often, the lesion is associated with trauma or congenital defects. The time from formation by traumatic event to identification of the defect, often during an episode of colic, depends more on the migration and strangulation of viscera through the rent at a later time than discomfort from the formation of the rent itself. There have been no substantial surveys of cases of diaphragmatic rents or defects seen at referral hospitals. Publications have focused on small numbers of cases, primarily reporting on individual cases or on a compilation of previously reported cases (1–6). Several reports in the past outline the successful correction of a diaphragmatic hernia in a single case or small case series. This approach lends itself to the possibility of bias toward positive outcomes by utilizing reports focusing on the documentation of a single successful surgical technique (7,8). This report outlines the presentation, surgical management, postoperative complications, and outcome of all cases of diaphragmatic hernia presented to 2 referral hospitals over a 5-year period.

Materials and methods

Records from Hagyard Equine Medical Institute and UC Davis Veterinary Medical Teaching Hospital (VMTH) were reviewed for the period January 2001 through June 2006. Paper records at Hagyard Equine Medical Institute were examined, and signalment, side of hernia, treatment and outcome were recorded. Electronic records (VMACS) at UC Davis VMTH were reviewed for the same time and same data. Cases in which a diaphragmatic rent was repaired were evaluated for technique and long-term outcome.

Results

A total of 31 cases of diaphragmatic rents with herniation were identified over the time period examined (Figure 1). Twenty-five adults, ranging in age from 1 to 28 y were identified (mean 12 y, median 13 y). Six cases occurred in foals < 1 y of age. Twenty-eight cases occurred in Thoroughbreds (22 adults, 6 foals), 2 cases in Quarter horses, and 1 case in an Arabian. All horses were presented for colic of varying severity and duration. Preoperative diagnostics were performed and included varying degrees of examination as dictated by the horse’s degree of discomfort and index of suspicion. Diagnostics included physical examination, abdominal ultrasound, thoracic ultrasound, abdominal radiographs, thoracic radiographs, abdominocentesis, and/or rectal examination. Of the 31 cases identified, 25 were taken to surgery for exploratory laparotomy. The remaining 6 cases that were not taken to surgery were all adults; 5 were euthanized for a variety of reasons, including economic constraints and poor prognosis for long-term survival and return to work. One adult horse that was diagnosed with diaphragmatic herniation via thoracic radiographs and ultrasound was managed medically until it was stable; it was removed from the hospital against medical advice.

Figure 1.

Ventral diaphragmatic rent identified during surgery.

Horses that were determined to be acceptable candidates for surgery were pre-medicated with gentamicin (Gentavid 100; Vedco, St. Joseph, Missouri, USA), 6.6 mg/kg body weight (BW), IV, penicillin procaine G (Aquacillin; Vedco), 22 000 IU/kg, BW, IM, and tetanus toxoid, IM. Flunixin meglumine (Banamine; Schering-Plough, Kenilworth, New Jersey, USA), 1.1 mg/kg, BW, IV was administered pre-operatively or immediately post-operatively if not administered earlier. Anesthesia was induced with xylazine (Tranquived, Vedco), 1.1 mg/kg, BW, IV, diazepam (Diazepam; Roxane Labs, Columbus, Ohio, USA), up to 0.05 mg/kg, BW, IV, and ketamine (Ketaset; Fort Dodge, Overland Park, Kansas, USA), 2.2 mg/kg, BW, IV. Foals were induced with diazepam (0.5 cc to 1.5 cc, IV) and an isoflurane mask. Anesthesia was maintained with either isoflurane or sevoflurane via an endotracheal tube. A mechanical ventilator was not utilized in any case. Those cases in which ventilation was required were ventilated manually. All the horses were placed in dorsal recumbency. Surgical preparation was performed as per a standard protocol of hair removal via a 40-blade clipper and skin preparation with beta-dine surgical scrub.

Post-operative care consisted of gentamicin (Gentaved 100, Vedco), 6.6 mg/kg, BW, IV, penicillin procaine G (Aquacillin; Vedco), 22 000 IU/kg, BW, IM, flunixin meglumine (Banamine; Schering-Plough), 1.1 mg/kg, IV, and fluid therapy (Plasmalyte 48; Baxter, Deerfield, Illinois, USA), 2 mL/kg/h, for 48 h. A complete blood (cell) count and a serum chemistry profile were run on surgically managed horses daily for 3 consecutive days or until the horses were stable.

In 24 cases, the side of the hernia was identified either in surgery or on postmortem examination (Figure 2). Seventeen of these cases had the defect on the left side, and 7 had the defect on the right side. Details on the organs herniated were recorded in 27 cases: small intestine (16 cases), large intestine (3 cases), small intestine and large intestine (4 cases), small intestine, large intestine, spleen, stomach (1 case), small intestine, liver and spleen (1 case), large intestine, small intestine and liver (1 case), large intestine, liver, and pancreas (1 case).

Figure 2.

Radiograph of thorax. Distended loops of bowel can be seen in the caudal dorsal lung field.

Of the 19 adult horses taken to surgery, 8 were immediately euthanized for a variety of problems. Of the 8 horses immediately euthanized, 2 were euthanized upon discovery of severely strangulated bowel. Two horses were euthanized following discovery of the diaphragmatic rent, and based on extreme size and nature of the defect. One horse was euthanized after attempts at removal of the large colon from the thorax resulted in rupture of the large colon. Three horses were euthanized in surgery after discovery of the defect and upon the owner’s request based on a historical poor prognosis for such a condition. One horse died intra-operatively after replacement of the bowel. Surgical correction of the rent was attempted in 10 horses. Three of these horses also had subsequent small intestinal resection and anastomoses.

Five diaphragmatic rents were corrected via a direct suturing technique after replacement of herniated organs into the abdomen. Placing some horses in the reverse Trendelenburg position assisted exposure of the rent. Direct suturing technique included partial removal of the fibrous edge of the rent with mayo scissors. Closure of the rent was performed via suturing with absorbable or non-absorbable suture in a simple continuous or cruciate pattern. Incorporation of the omentum into the closure was also performed in some cases.

Five diaphragmatic rents were corrected via direct suturing and/or the application of a polypropylene mesh secured to the diaphragm over the defect. In cases where the defect was easily visualized and the mesh was utilized, a hand-stapling device was used. In cases where the defect was not easily visualized (dorsal, at the insertion of the diaphragmatic crura) a laparoscopic stapling device was utilized for its length to apply staples to a location directed by the surgeon’s hand. Pneumothorax was reduced in all cases in which the defect was ultimately repaired. This was performed via the application of a teat cannula through the diaphragm from an intra-abdominal approach. The cannula was attached to suction and negative pressure placed. Closure of the cannula portal was performed if necessary.

Two horses (1 closed with direct suture, 1 closed with the aid of polypropylene mesh) died in recovery after re-formation of the pneumothorax. This condition was recognized quickly in both cases and attempts to correct it via the removal of air from the thorax by the placement of an Argyll trocar catheter were unsuccessful.

The defect in 1 horse was repaired by a combination of direct suturing and polypropylene mesh, but closure of the most dorsal aspect was unsuccessful due to its close relationship to the ribs and unsubstantial substrate for securing either suture or staples. This horse recovered from anesthesia, but ultimately reherniated within 24 h of recovery. No option for a second surgery was available. This horse was euthanized.

One horse died of unknown causes 3 d after surgery and further examination was not performed.

Seven adult horses were discharged from the hospital. Six of these 7 were discharged following exploratory surgery and correction of the rent. One horse was discharged after medical management and against medical advice. One horse returned for repeat colic within 30 d of discharge. This horse had been treated by exploratory celiotomy and closure of the defect via direct suturing and polypropylene mesh. Examination at this time revealed repeated hernation of bowel into the right thorax. This horse underwent a second surgical exploration of its abdomen. A strangulated cecal apex was identified and resected. The defect in the diaphragm was identified and found to continue to be patent, albeit smaller than originally described. A second attempt at closure of the defect was performed with direct suturing and polypropylene mesh. Recovery was uneventful and the horse was discharged after standard postoperative care. A second horse presented 84 d after surgical correction of a rent via direct suturing. Severe signs of abdominal pain were the chief complaint. A second exploratory celiotomy was performed. Adhesions were identified throughout the abdomen, and involved the small intestine, liver, and diaphragm. Euthanasia was elected after discussion of the increased chance of post-operative complications and chronic colic. Seven horses are currently still alive and 5 are performing successfully as broodmares.

Diaphragmatic rents were identified in 6 foals, 3 of which were immediately euthanized upon surgical exploration. Reasons for euthanasia all centered on the presence and amount of compromised herniated intestine. All cases had > 40% of small intestine involved. Three foals were found to be good candidates for surgical correction due to minimal gastrointestinal damage and general systemic health. These were all corrected by direct suturing of the diaphragmatic rent. Two of the 3 foals taken to surgery died in the post-operative period, and 1 foal survived and was discharged from the hospital. This horse has now successfully performed as a racehorse.

Discussion

With an overall rate of survival of 23% for all horses that presented with a diaphragmatic hernia, and a surgical success rate of 46%, the prognosis for these horses still remains low. While not previously investigated in a study with this number of horses, this condition has always been thought to carry a relatively poor prognosis. Most reports focus on individual case reports of successful surgical corrections of a diaphragmatic rent after reduction of a hernia, or of a case series including small numbers of cases and their outcomes. Based on these publications, reported success rates for this condition vary greatly.

In adults, a diaphragmatic rent and subsequent herniation are thought to be primarily associated with a traumatic event, such as impact injury during racing, trailer accidents, or foaling (9–11). In cases in which the rent is formed via a traumatic event, there can be a prolonged interval between the trauma and the onset of clinical signs, as often the migration of viscera through the rent is not directly related to the formation of the defect. In foals, it is most often associated with either congenital defects or difficult foaling leading to rib fractures (2,4,11,12).

Most commonly diagnosed during an episode of colic or on postmortem examination, diaphragmatic hernias are known to involve a variety of organs. Previous reports have described a variety of configurations of viscera herniating through the defect, ranging from single organ herniation, to the herniation of multiple organs including liver, stomach, small intestine, large intestine, and spleen (1,5,6,13). Clinical signs from the physical examination often encountered with diaphragmatic hernia in the horse can range from very mild to severe signs of colic, and often evidence of respiratory distress. The signs observed in each case are more closely associated with the amount of damage done to the herniated viscera, than with the defect itself (11). The size of the rent is also associated with the severity of clinical signs, as smaller rents are more likely to be constrictive once intestine does enter the ring. These smaller defects are also associated with the herniation of small intestine. Larger defects appear to be associated with the herniation of larger visci, such as large colon (14). These larger defects may also be more closely associated with clinical signs of respiratory distress (2,10). Other clinical signs can include tachycardia, and signs of respiratory distress in the form of tachypnia, and shallow breathing. The wide range of clinical signs and severity of those signs can make identification and diagnosis difficult.

Non-invasive methods most useful for the diagnosis of a diaphragmatic hernia other than the physical examination include rectal palpation, transthoracic ultrasonography, and thoracic radiography. None of these are consistently productive, however, and expected findings such as tachypnea are not always present (1). Rectal palpation findings range greatly, from normal findings to the absence of recognizable structures in the caudal abdomen, leaving the clinician the impression of an “empty” abdomen. More invasive methods such as laparoscopy, thoracoscopy, and exploratory celiotomy are more productive but carry an increased risk of complications.

In surgery, the defect is often surrounded by a significantly thickened ring, indicating some degree of chronicity or congenital origin (2,13). Only histopathologic interpretation can definitively ascertain the origin, especially in adults (4). As previously discussed, the correlation between the colic, identification of the defect, and an inciting cause is difficult as the time from traumatic events such as trailer accidents or racing accidents can be quite prolonged. When associated with a foaling event, more certain connections can be made. Correction of the defect has been documented through a variety of methods, including rib resection for access via thoracoscopy, direct suturing of defects, and application of polypropylene mesh over the defect (1,5,6,10,11,14,15). While successful outcomes have been reported using all these techniques, this review of all cases presented to 2 referral centers identifies an overall poor prognosis for survival. As reported, 23% (7/31) of all cases that presented for colic and 46% (6/13) of cases in which correction was attempted survived long-term. While advances in equine surgery over the last few decades have made dramatic increases in survival rate for abdominal and thoracic procedures, additional factors associated with diaphragmatic herniation such as ischemic bowel and pneumothorax appear to be confounding efforts to improve the prognosis for this condition.

An additional factor that may contribute to the success or failure of an attempt to surgically correct a diaphragmatic hernia is the presence of optimal anesthetic equipment. One major limitation of the surgical cases reviewed is the limited monitoring equipment (ECG and arterial blood pressure) or lack of ventilation equipment to aid in proper ventilation of animals that are prone to inadequate oxygenation due to a pneumothorax and compression of the lungs by herniated viscera (3). While the incidence of euthanasia due to identification of ischemic bowel upon entering the abdomen would not change, oxygenation may have affected the 2 horses that died in recovery. In both cases, re-formation of a pneumothorax was identified and treated quickly. Their survival may have been aided by additional intraoperative monitoring equipment leading to earlier application of appropriate treatment in the form of a chest tube or drain prior to recovery. This would have allowed for more rapid evacuation of air from the thoracic cavity and re-inflation of the lungs should collapse occur during the post-operative period.

Ultimately, the prognosis for a horse presenting for surgical correction of a diaphragmatic hernia remains poor at 23% overall and 46% for those in which surgical correction is attempted. While often the focus in these cases is the size and shape of the diaphragmatic defect and the ultimate correction, it appears that there are many other pertinent issues to consider when presented with such a case, such as the amount and types of viscera involved or degree of respiratory insult experienced. Further investigation into anesthetic complications associated with these cases and the effect of these complications on their outcome should be undertaken. As advances in abdominal surgery are made, and the prognosis for horses that experience intestinal ischemia improve, this subject should be revisited as it appears that prognosis is dependent on additional factors beyond the diaphragmatic defect. CVJ

Table 1.

Summary of features and outcome of cases diagnosed with diaphragmatic hernia

Number of horses presented	Age	Breeds affected	Side of hernia		Horses to surgery	Organs herniated (17 cases, surgery, or postmortem)	Surgical correction	Overall survival
			Right	Left
6 foals, 25 adults	1 mo to 28 y (mean 12 y, median 13 y)	TB 28, QH 2, Arabian 1	7	17	19 adult, 3 foals	small intestine, large intestine, spleen, stomach, liver, pancreas	10 adults, 3 foals, 8 via direct suturing, 5 via suturing and mesh	7 adults, 1 foal

TB — Thoroughbred; QH — Quarter horse.