Chronic Diaphragmatic Hernia in 34 Dogs and 16 Cats
Medical records of 34 dogs and 16 cats undergoing surgical repair of diaphragmatic hernia of >2 weeks’ duration were reviewed, and long-term follow-up information was obtained. The most common clinical signs were dyspnea and vomiting; however, many of the animals were presented for nonspecific signs such as anorexia, lethargy, and weight loss. Thoracic radiographs revealed evidence of diaphragmatic hernia in only 66% of the animals, and additional imaging tests were often needed to confirm the diagnosis. Thirty-six hernias were repaired through a midline laparotomy; 14 required a median sternotomy combined with a laparotomy. In 14 animals, division of mature adhesions of the lungs or diaphragm to the herniated organs was necessary to permit reduction of the hernia. Fourteen animals required resection of portions of the lungs, liver, or intestine. All hernias were sutured primarily without the use of tissue flaps or mesh implants. Twenty-one of the animals developed transient complications in the postoperative period; the most common of these was pneumothorax. The mortality rate was 14%. Thirty-four (79%) of the animals that were discharged from the hospital had complete resolution of clinical signs, and none developed evidence of recurrent diaphragmatic hernia during the follow-up period. Nine were lost to follow-up.
Introduction
Diaphragmatic hernia in the dog and cat is a common sequela to automobile accidents. Many of these hernias are diagnosed soon after the traumatic incident and subsequently are surgically repaired. Several reports in the veterinary literature have described the pathophysiology, surgical approaches, and surgical outcomes of traumatic diaphragmatic hernia.1–7 All case series have consisted mainly of acute diaphragmatic hernias but have included a small number of chronic diaphragmatic hernias (CDH).1–7 In one report, the mortality rate among eight animals undergoing surgical repair of CDH >1 year’s duration was 62.5%.6 None of the reports have included long-term follow-up information for patients with CDH. In humans, CDH is a common condition and has an excellent prognosis with appropriate treatment.8–11 However, in veterinary patients, both diagnosis and treatment of the condition can be challenging.
The purpose of this study was to provide information regarding the history, diagnostic evaluation, surgical findings, complications, and long-term outcomes for a large series of dogs and cats undergoing surgical repair of traumatically induced CDH.
Materials and Methods
The medical records of patients that underwent surgical repair of a traumatically induced diaphragmatic hernia at Tufts University School of Veterinary Medicine or Angell Memorial Animal Hospital between 1987 and 1996 were reviewed. Selection criteria were identified to differentiate those with acute traumatic diaphragmatic hernia from those with CDH. A diagnosis of CDH was made when one or more of the following criteria were met: a known history of trauma at least 2 weeks prior to surgery; clinical signs referable to the hernia that were present for at least 2 weeks prior to surgery; or obvious signs of chronicity identified at the time of surgery. Clinical signs that may be referable to a diaphragmatic hernia include coughing, tachypnea, dyspnea, vomiting, or weight loss. Surgical findings that were considered to be indicative of chronicity included absence of bruising in the body wall or diaphragm, absence of blood in the thoracic or abdominal cavities, presence of mature adhesions to or between the herniated tissues, and presence of mature fibrosis of the hernia margins.
Information obtained from the medical records included signalment, history, physical examination findings, results of clinicopathological and imaging tests, anesthetic protocols, surgical findings and repair techniques, surgical time, length of hospitalization, and outcome during the period of hospitalization. Long-term follow-up information was obtained through telephone interviews with owners or referring veterinarians or through mailed questionnaires. Interviews and questionnaires identified any postoperative problems. General questions included recovery assessment such as dyspnea, coughing, gagging, weight and appetite improvement, and subsequent veterinary examinations. Questions specific to individual cases were also included, and cause of death was identified if the animal was no longer alive. Five of these animals underwent a follow-up physical examination. Follow-up thoracic ultrasonography or radiographs were obtained in two and three cases, respectively. Nine animals were lost to follow-up.
Results
Of the 250 animals identified as undergoing diaphragmatic hernia repair during the study period, 34 dogs and 16 cats met the inclusion criteria [Table 1]. Of the 34 dogs, there were eight males, 11 neutered males, two females, and 13 spayed females. Of the 16 cats, there were nine neutered males, six spayed females, and one intact female. Among both dogs and cats, a wide range of breeds was represented. Thirty-four animals had a known history of trauma at least 2 weeks prior to surgery, eight had clinical signs of at least 2 weeks’ duration without a definitive trauma history, and eight animals had surgical findings consistent with CDH. Clinical signs led the owners to seek veterinary care in 45 (90%) of the 50 animals, and in five (10%) animals, signs referable to CDH were found during routine physical examination. Clinical signs and physical examination findings are given in Tables 2 and 3. The duration of clinical signs ranged from 1 day to 7 years (mean, 16 weeks; median, 2 weeks).
Preoperative complete blood counts (CBC) were performed in 38 animals. Abnormalities included anemia (10%) (range, 25% to 36%; reference range, 37% to 55%) and leukocytosis (5%) (range, 24 to 41 × 103/μL; reference range, 6.0 to 17.0 × 103/μL). Preoperative serum biochemical profiles were performed for 39 animals. Abnormalities included elevations in alanine aminotransferase (ALT; 53%) (range, 96 to 2,490 U/L; reference range, 10 to 70 U/L), aspartate aminotransferase (AST; 41%) (range, 48 to 985 U/L; reference range, 10 to 40 U/L), and total bilirubin (15%) (range, 0.47 to 10.0 mg/dL; reference range, 0.04 to 0.40 mg/dL), and hypoproteinemia (7%) (range, 4.5 to 4.8 g/dL; reference range, 5.2 to 7.2 g/dL).
Thoracic radiography was performed in 48 animals. Diaphragmatic hernia was documented in 34 (70%) of these or was suspected due to loss of detail combined with trauma history. Other abnormalities included pleural effusion (15/48; 31%) and rib fractures (3/48; 6%). Abdominal radiography was performed in five animals. Abnormalities included cranial displacement of the liver (n=1; case no. 4), peritoneal effusion (n=1; case no. 34), displacement of the stomach into the thorax (n=2; case nos. 28, 32), and absence of the diaphragmatic line (n=1; case no. 2). Abdominal ultrasonography was performed in 10 animals. Herniation of abdominal organs and loss of the diaphragmatic line11 were identified in all but one (case no. 13) of these. Gastrointestinal contrast studies with orally administered barium were performed in two animals. One dog (case no. 2) had pleural effusion obscuring the diaphragmatic line with possible displacement of the stomach. After barium was given, the pyloric antrum and proximal intestines were identified within the thorax. A cranial abdominal mass effect was palpated in a cat (case no. 38), and barium was administered. Radiographs revealed that the small intestinal tract and a portion of the colon were in the chest cavity. Positive-contrast celiography was performed in one dog (case no. 13) and was inconclusive; this was the same dog that had an inconclusive sonogram.
Anesthetic protocols varied in this series of cases. Premedication was used in 31 (62%) animals. The agents used were butorphanol (n=21; 42%), diazepam (n=3; 6%), acepromazine (n=3; 6%), buprenorphine (n=2; 4%), and oxymorphone/diazepam combination (n=2; 4%). No premedication was used in 19 (38%) of the cases. Induction agents used included ketamine/diazepam (n=25; 50%), thiopental (n=15; 30%), propofol (n=5; 10%), methohexital (n=2; 4%), and oxymorphone/diazepam (n=1; 2%). Two (4%) animals underwent mask induction with isoflurane. All animals were placed on mechanical ventilators intraoperatively.
All animals underwent abdominal exploratory surgery. In the animals with a known history of trauma, the time to surgical correction ranged from 2 weeks to 84 months (mean, 14.5 months). Five dogs underwent emergency surgery. Two of these surgeries (case nos. 18, 31) were due to severe respiratory distress that did not resolve with thoracocentesis, and three (case nos. 17, 24, 25) were operated because of acute intestinal obstruction. The remaining 45 surgeries were performed between 1 and 4 days after diagnosis. Fourteen (28%) animals had a partial median sternotomy in addition to standard midline laparotomy to facilitate breakdown of adhesions and reduction of herniated contents.
Surgical time ranged from 25 to 265 minutes (mean, 87 minutes; median, 100 minutes). Surgical reports indicated that 17 animals had left-sided hernias, 19 had right-sided hernias, and three hernias were reported to be ventral. Location of herniation was not cited in 11 animals. In 14 (28%) of the animals, mature adhesions between the lungs or edges of the diaphragm and herniated organs were noted in the surgery report; these were divided in order to reduce the hernia. The following abdominal structures were herniated at the time of surgery: liver (32/50; 64%), small intestine (28/50; 56%), omentum (17/50; 34%), spleen (16/50; 32%), stomach (17/50; 34%), colon (8/50; 16%), and pancreas (5/50; 10%). In the 14 animals with radiographic evidence of pleural effusion, strangulation of the liver (n=12) or small intestine and omentum (n=2) was found at surgery. In 14 patients, herniated tissues were resected at the time of herniorrhaphy. Complete or partial liver-lobe resections were performed in nine animals due to hemorrhage, advanced adhesions, liver-lobe torsion, or necrosis. Complete or partial lung lobectomies were performed in two animals. One dog (case no. 16) had a right middle lung-lobe torsion, and a complete right-sided lobectomy was performed in a cat (case no. 35) due to fibrosis and contamination from a perforated megacolon caused by strangulation in a scarred hernia ring. Entrapped hemorrhagic omentum was removed from a dog (case no. 13). Adhesions between the small intestine and pericardium required a pericardectomy to reduce the herniated intestine (case no. 14). One dog (case no. 7) underwent a nephrectomy for hydronephrosis unrelated to CDH. Intraoperative pulmonary atelectasis was noted in the anesthetic and surgical records of 15 animals. Following hernia reduction, spontaneous reexpansion of the lungs was noted in the record to have occurred in six animals with ≤ 18 cm H2O of positive-pressure ventilation. The edges of the diaphragm were freshened prior to herniorrhaphy in all cases. No animals required a muscle flap or mesh graft in order to close the hernia defect. Monofilament nonabsorbable suture material was used for hernia repair in 45 (90%) patients; monofilament absorbable suture was used in four (8%) patients; and absorbable braided suture was used in one (2%) patient. Suture patterns varied and included simple-continuous, simple-interrupted, cruciate, and mattress patterns. Chest tubes were placed intraoperatively in 40 (80%) patients.
Forty percent of the animals developed transient complications during the postoperative hospitalization period [Table 4]. Postoperative pneumothorax (5/50; 10%) was the most common of these. Only one of these animals (case no. 3) did not have a chest tube placed at the time of surgical repair and required manual thoracocentesis 4 and 8 days postoperatively. The pneumothorax resolved within 2 to 5 days in all cases using either intermittent or continuous aspiration. Four of the five animals with pneumothorax were noted to have pulmonary adhesions at surgery, and these adhesions were divided to permit reduction of the hernia. All animals with acute anemia (3/50; 6%) caused by intraoperative blood loss received transfusions.
Seven (14%) animals died or were euthanized either intraoperatively (2/50; 4%) or during the early postoperative period (5/50; 10%). All animals that developed fatal complications were dogs. One dog (case no. 25) was found at surgery to have strangulation necrosis of the small intestine; this dog died intraoperatively. A second dog (case no. 18) died intraoperatively due to multiorgan failure and sepsis secondary to liver abscess and necrosis. This dog presented in shock with dyspnea, pyothorax, pyrexia, and leukocytosis. Another dog (case no. 24) developed acute liver failure from severe hepatic congestion and died 8 hours postoperatively. Two dogs died of severe pneumothorax within 24 hours of surgery. One of these deaths (case no. 1) was caused by failure to recognize a poor connection in the chest tube apparatus. The other dog (case no. 28) developed a tension pneumothorax secondary to a suspected pulmonary bulla and died after a second thoracic exploratory for lung-lobe removal. One dog (case no. 21) died of aspiration pneumonia 4 days after surgery. One dog (case no. 22) developed a persistent postoperative pneumothorax and pyothorax and underwent a second surgery 8 days later. The entire right lung fields were removed due to necrosis. He was euthanized 1 day following the second surgery because of intractable seizures and a grave prognosis. A complete necropsy suggested that circulatory compromise to the lung occurred as the result of compression from the herniated liver, and secondary bacterial infection and pulmonary compromise led to his clinical deterioration.
Forty-three (86%) animals were discharged from the hospital. Nine were lost to follow-up, and long-term follow-up information was available for the remaining 34 [Table 5]. All of these animals had a complete resolution of clinical signs of CDH, and none developed recurrent clinical signs. Thoracic ultrasonography in two cats (case nos. 37, 39) at 6 months and 1 year postsurgery found no evidence of recurrence of CDH. Thoracic radiographs in three dogs (case nos. 3, 12, 16) taken at 1, 2, and 6 years postCDH surgery, respectively, were normal.
Discussion
Although the majority of traumatic diaphragmatic hernias are diagnosed shortly after a traumatic incident, a significant proportion may escape detection initially. The 50 cases of CDH reported here represent approximately 20% of all diaphragmatic hernias repaired at the authors’ institutions during this time period. Acute traumatic diaphragmatic hernias may escape detection because owners may be unaware that a traumatic incident has occurred, because the condition may be asymptomatic, and because a definitive radiographic diagnosis can be difficult to establish. Because CDH can be physically debilitating and challenging to diagnose and treat, all animals that have suffered automobile trauma or other blunt and penetrating trauma to the chest or abdomen should have thoracic radiographs at the time of initial evaluation to screen for acute diaphragmatic hernia as well as other thoracic injuries.
The diagnosis of CDH can be complicated by the tendency of the associated clinical signs to be insidious and nonspecific. Only 38% of animals in the present study were dyspneic; many were presented for vague signs such as anorexia, lethargy, or weight loss. Approximately half of the animals were presented for vomiting, which may have resulted from slow contraction of the hernia, causing mechanical obstruction of the stomach or small intestine. A history of chronic, nonspecific clinical signs, possibly combined with a history of vomiting, dyspnea, or past trauma, should prompt consideration of CDH.
Confirmation of CDH requires diagnostic imaging, and clinical pathology usually does not contribute to the diagnosis. False-negative results are possible with any of the commonly available imaging techniques.12–14 In the present series of cases, approximately one-third of the hernias could not be conclusively demonstrated with thoracic radiography. The presence of pleural effusion13 or herniation of relatively small volumes of abdominal contents can make diaphragmatic hernias difficult to recognize on thoracic radiographs, particularly if the herniated structures are not gas filled.15 Other imaging options include upper gastrointestinal (GI) contrast studies, positive-contrast celiography, and ultrasonography. Because any of these can be inconclusive, more than one technique may be necessary to confirm the diagnosis.13 Repeat thoracic radiographs following thoracocentesis may definitively identify a CDH. In seven of eight cases in this study that underwent additional radiographic studies, a confirmed diagnosis of CDH was obtained. Ultrasonography was the most frequently utilized supplemental imaging study, followed by an upper GI contrast study. Upper GI contrast studies may not be diagnostic if the stomach or small intestines are not significantly displaced.12 Positive-contrast celiograms may appear normal if chronic scarring of the hernial ring to the herniated organs prevents flow of the contrast agent into the thorax.15 Although ultrasonography demonstrated the hernias in nine of 10 cases in the present series, accurate diagnosis of diaphragmatic hernia requires an experienced ultrasonographer. The ultrasonographic diagnosis is especially difficult when extensive pulmonary contusions are present, causing the lung to mimic liver, or when the omentum is the only structure passing through a small rent in the diaphragm.11 It is suggested that thoracic radiography be performed and repeated if thoracocentesis is necessary. If a diagnosis cannot be made, an upper GI contrast study may be helpful if there is displacement of the abdominal organs. One film may be diagnostic and cost effective. If there is an experienced ultrasonographer available, the diagnosis may be confirmed on ultrasound. On occasion, exploratory surgery may be necessary to obtain a definitive diagnosis.11 Only one dog (case no. 13) in this series required exploratory surgery to confirm the diagnosis of CDH.
Because clinical signs of traumatic CDH usually develop slowly, emergency surgery is rarely necessary; only 10% of animals in the present series required emergency surgery. The principal indications for emergency surgery include severe respiratory distress that is not relieved by thoracocentesis and oxygen administration, the presence of the stomach in the thoracic cavity, and obstruction of the GI tract. Animals with prolonged, severe dyspnea may develop muscle fatigue and can succumb to respiratory failure. These animals may need to be intubated and ventilated to stabilize them prior to surgery. Chronically debilitated or dehydrated animals should be adequately rehydrated with intravenous fluids prior to anesthesia.
Anesthetic premedication with phenothiazine is often unnecessary in these cases and contraindicated in debilitated animals. Preemptive analgesia is recommended in all animals undergoing surgery; however, drug doses may have to be reduced for critically ill animals. Anesthesia should be induced rapidly in order to gain control of ventilation; either a barbiturate or propofol in the healthier patient or a ketamine/diazepam combination is recommended in debilitated animals. Mask induction with an anesthetic gas should be avoided, because this technique is often prolonged, uncontrolled, and difficult, resulting in respiratory distress.17 Preoxygenation is recommended for dyspneic animals. All animals with diaphragmatic hernias require positive-pressure ventilation, because negative intrathoracic pressure will be lost once the abdominal cavity is opened. To prevent pulmonary barotrauma and reexpansion pulmonary edema, positive pressure should not exceed 20-cm H2O.17
Chronic diaphragmatic hernias often are more difficult to reduce and repair than acute diaphragmatic hernias.4 Most CDH can be reduced through a standard midline laparotomy incision extending caudally from the xiphoid. If the herniated organs cannot easily be withdrawn into the abdomen, the hernial opening should be enlarged to improve visualization and facilitate reduction. In some instances, extensive intrathoracic adhesions may impede reduction; visualization and division of these adhesions may require a median sternotomy.16 Because the need for median sternotomy cannot be assessed preoperatively, the area of surgical preparation should always extend from the manubrium to the pubis. Fourteen animals in this series required median sternotomy. Care should be taken to control hemorrhage during division of adhesions. Vascular clips or electrocautery can be time-saving and convenient methods to control bleeding during division of adhesions. If the omentum has herniated and is difficult to return to the abdomen because of adhesions, the omentum can be divided in a convenient location, and the adherent portion can be left in the thorax. Clearance of omentum from the surgical field can markedly facilitate reduction and repair of the hernia.
The surgeon should be prepared to perform complete or partial lung or liver lobectomies and to resect and anastomose portions of the intestinal tract. Fourteen animals in this series required organ resections. Lung lobectomies may be necessary if areas of the lung are severely atelectic and fibrotic, are extensively involved with adhesions, are torsed, or are inadvertently lacerated during reduction of the hernia. Organs that are torsed should always be removed without untwisting the organ. Atelectic lung lobes should be allowed to slowly expand naturally; forcible manual reexpansion may produce pulmonary barotrauma and result in reexpansion pulmonary edema.1819 Liver lobectomies may be required if segments of the liver are bleeding extensively or appear necrotic or abscessed. Intestinal resection is indicated when segments of bowel are devitalized or perforated secondary to incarceration. Copious intrathoracic and intraabdominal lavage should be used as well as culture and sensitivity testing when abcessation or intestinal perforation is present.
Most CDH are contracted and are small enough to be sutured primarily. None of the hernias in this series required more complex closure techniques, such as muscle or omental flaps or repair with polypropylene mesh. The edges of the rent should be freshened prior to closure to promote healing. Closure may be accomplished with either absorbable or nonabsorbable suture material, in a simple-interrupted or simple-continuous pattern. It is generally easiest to close the dorsal aspect of the opening first and then progress ventrally. An excessively tight closure may produce necrosis of the diaphragm edges, resulting in reherniation.2 Thoracic drains should be placed routinely. These may enter the chest through the ventral abdominal wall and diaphragm or through an intercostal space.
Postoperative complications were common in this series of cases; however, most complications were transient and easily managed. The most common complication was pneumothorax. Four of the dogs with pneumothorax had adhesions to the lung that were manually divided. Because dissection of pulmonary adhesions can result in inadvertent laceration of the lung, partial lung lobectomy should be considered whenever fibrous pulmonary adhesions are present. Severe hypothermia and prolonged anesthetic recovery time are potential complications of CDH repair, because surgery times are often prolonged and there is an occasional need to open both the abdominal and thoracic cavities. Measures that reduce the risk of hypothermia include the use of fluid warmers, circulating hot air and waterbeds, wrapping of the extremities, heat lamps, and warm saline intrathoracic and intraabdominal lavage. Other transient postoperative complications in this series of cases included anemia secondary to blood loss, fever, shock, and cardiac arrhythmias. Whole-blood transfusions were given to three animals, fever was treated with fluid antibiotics where appropriate, and shock was treated with colloid fluid boluses, hetastarch, and dopamine when necessary. Cardiac arrhythmias were asymptomatic and transient and were not treated. Reexpansion pulmonary edema, a reported complication of CDH repair in humans, was not observed on the authors’ series of cases;18 however, it is possible that some of the animals in this series developed pulmonary edema that was not clinically apparent. Reexpansion pulmonary edema may not have been observed in this study, because slow reexpansion was allowed to occur and ventilatory pressures were never >18-cm H2O. One dog’s lungs were forcibly reexpanded with pressures up to 35-cm H2O, and then the dog developed a bulla, subsequent pneumothorax, and died. There is only one report in the veterinary literature of fatal reexpansion pulmonary edema after pectus excavatum surgery in a kitten.19 It is possible that cats and dogs are not at risk as compared to humans undergoing similar procedures.
The 14% mortality rate observed in this series of patients compares favorably with the reported mortality among animals undergoing repair of acute diaphragmatic hernias. Of the 43 patients that were discharged from the hospital, 34 had a permanent resolution of clinical signs, and none were documented to develop hernia recurrence during the follow-up period. In conclusion, the authors’ data indicates that with appropriate surgical technique and perioperative care, dogs and cats with CDH can have an excellent prognosis.
