Clinical and Prognostic Findings in Dogs with Suspected Extrahepatic Biliary Obstruction and Pancreatitis
ABSTRACT
Pancreatitis in dogs may lead to extrahepatic bile duct obstruction as a result of local inflammation. Medical records of 45 client-owned dogs with clinical suspicion of extrahepatic bile duct obstruction secondary to pancreatitis were reviewed to determine clinical findings, outcome, and factors associated with survival. Survival times were determined using the Kaplan-Meier product limit method. Cox multivariable survival methods were employed to determine factors associated with survival time following diagnosis. The median survival time was 241 days (95% confidence interval [CI] 25–631), with 34 of 45 dogs (76%) surviving to discharge. Dogs 9 yr of age or older with azotemia at presentation had a 9.9 greater hazard for death (95% CI 2.5–38.1; P = .001) compared with dogs younger than 9 yr old without azotemia at presentation. Dogs without subjective ultrasonographic gallbladder distension had a 4.4 greater hazard for death (95% CI 1.3–15.4; P = .018) compared with dogs with subjective gallbladder distension. Dogs with a body temperature ≥102.5°F at admission had a 3.1 greater hazard for death (95% CI 1.3–7.7; P = .013) than dogs with a body temperature <102.5°F at admission. This information may help clinicians discuss prognosis with owners of affected dogs.
Introduction
Extrahepatic bile duct obstruction (EHBDO) prevents the flow of bile from the liver and gallbladder to the duodenum. Biliary stasis and decreased bile acid recirculation have deleterious effects, including decreased absorption of fat-soluble vitamins, coagulopathies, and increased risk of bacterial translocation and endotoxemia.1–3 In dogs, pancreatitis is a common cause of EHBDO, owing to the anatomic proximity of the distal end of the common bile duct (CBD) and the pancreas.2,4 The CBD may be compressed by inflamed pancreatic and/or surrounding tissue.5
Information regarding outcome in dogs with EHBDO secondary to pancreatitis is limited. Pancreatitis in dogs varies in severity, with reported mortality rates of 19–58%.6–10 Although several treatment strategies may be employed to treat EHBDO secondary to pancreatitis, there is limited evidence to support the optimal treatment.6,11 Medical management including IV fluid therapy, antiemetics, antacids, analgesics, antithrombotics, nutritional intervention, and possibly corticosteroids has been recommended for managing canine acute pancreatitis.7,12,13 Indications for surgery in dogs with EHBDO remain poorly defined; however, increasing serum bilirubin with ultrasonographic evidence of obstruction is generally considered an indication.5 Potential surgical procedures include cholecystoenterostomy, tube cholecystostomy, and CBD stenting.4,14,15 Outcome of surgical intervention of EHBDO secondary to pancreatitis has been described in a small number of dogs, with mortality ranging from 27 to 75%.6,15–18 Biliary surgery in dogs with pancreatitis has been associated with a worse long-term outcome compared with biliary surgery in dogs without pancreatitis.15 Outcome of medical management alone for this condition has not been reported aside from a case series describing decompressive percutaneous ultrasound-guided cholecystocentesis, which was successful in two of three dogs.19
Prognostic factors in dogs with EHBDO secondary to pancreatitis have not been reported. Risk factors for the development of pancreatitis include age (>7 yr, 5–9 yr, ≥10 yr), sex (spayed female), breed (Yorkshire terrier, terrier group, nonsporting group), dietary indiscretion, overweight body condition, prior gastrointestinal disease, endocrinopathies, and epilepsy, although these have not been evaluated for association with outcome in a clinical population of dogs with pancreatitis.8,20,21 Factors associated with death have been reported for dogs undergoing biliary surgery, including preoperative clinicopathologic indicators of liver, renal, or hemostatic dysfunction, postoperative hypotension, postoperative dyspnea, septic bile peritonitis, and cholecystoenterostomy procedures.15,16 A small percentage of these dogs had EHBDO secondary to pancreatitis (4 of 34 dogs in one study15 and 12 of 60 dogs in another16). Factors associated with outcome have also been evaluated for dogs undergoing surgical management of pancreatitis. No prognostic factors were identified, and fewer than 30% of the dogs had EHBDO.6
The aims of this study were to describe the clinical findings and outcome of dogs who were clinically suspected of having EHBDO secondary to pancreatitis and to identify factors associated with survival.
Materials and Methods
Case Selection Criteria
The medical records computerized database at the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania was searched from June 1, 1989, through June 30, 2016, to identify dogs with the following diagnostic codes: EHBDO, pancreatitis, and hyperbilirubinemia. Records identified via this search were reviewed by a single author (S.M.P.) to identify patients that met inclusion criteria for the study. All patients meeting study criteria were treated during the years 2002–2016.
Inclusion Criteria
Cases were included if they had a diagnosis of pancreatitis and EHBDO. Pancreatitis was diagnosed based on one or more clinical signs including inappetence, vomiting, and diarrhea as well as ultrasonographic evidence of a hypoechoic pancreas with hyperechoic adjacent mesentery. Distinction between acute, chronic, and acute-on-chronic forms of pancreatitis could not be made on the basis of these inclusion criteria. EHBDO was defined based on diagnostic coding of hyperbilirubinemia and ultrasonographic evidence of CBD dilation greater than 3 mm.1 These inclusion criteria support a clinical suspicion of EHBDO, but this can only be confirmed via laparotomy or autopsy, which were not performed in all study dogs.
Exclusion Criteria
Dogs having ultrasonographic or surgical identification of conditions other than pancreatitis that may have contributed to EHBDO were to be excluded from the study. Conditions for which ultrasound reports and surgical reports were screened included cholelithiasis, gallbladder or CBD masses, gallbladder mucocele, or suspected CBD stricture, although none were identified in this population. Dogs in whom survival time could not be determined were also excluded.
Medical Records Review
Medical records were retrospectively reviewed by a single author (S.M.P.). History, signalment, concurrent disease, medication history, body weight, results of initial physical examination, clinicopathologic data, ultrasonographic findings, surgical versus medical management, presence of systemic inflammatory response syndrome (SIRS), and survival data were recorded. SIRS was defined as the presence of two or more of the following: body temperature <100.6°F or >102.6°F; heart rate >120 beats per min; respiratory rate >20 breaths per min; white blood cells count <6000/µL or >16,000/µL; or >3% bands.22 When available, serial clinicopathologic data were reviewed. Ultrasonographic reports that had been finalized by a radiologist who is board certified by the American College of Veterinary Radiology reports were reviewed to confirm ultrasonographic evidence of pancreatitis and suspected EHBDO. Whereas CBD dilation was based on a measured lumen diameter of >3 mm,1 gallbladder distention and cystic duct dilation were based on the radiologist’s subjective interpretation. Referring veterinarians and/or clients were contacted when necessary for information regarding survival.
Statistical Analysis
Descriptive statistics were calculated. Normally distributed continuous variables were expressed with mean and standard deviations, whereas non-normally distributed continuous variables were expressed with median and range. Categorical variables were expressed as frequencies and proportions. Median survival times were determined by the use of the Kaplan-Meier product limit method. Cox multivariable survival methods were employed to determine which factors were associated with survival time. Two-way interactions among the main effects were investigated. An interaction term was retained based on a P value <.05. If not effect modifiers, univariate analysis was performed initially and factors with P value <.20 were tested in the model. Factors were retained in the model based on P value <.05 or if found to be a confounder (changing model coefficients by >15%). Variables evaluated included dog characteristics (age, body weight, sex), clinical signs (duration, vomiting, diarrhea, inappetence), concurrent disease, physical exam findings at presentation (temperature, heart rate, blood pressure, abdominal pain, jaundice, hydration status), clinical chemistry (liver and kidney variables, including changes in total bilirubin), complete blood count, ultrasound findings (gallbladder, CBD, cystic duct, intrahepatic duct distention/dilation), surgical versus medical management, and the presence of SIRS. Hemostatic variables (prothrombin time [PT], activated partial thromboplastin time [aPTT], D-dimers) were evaluated on univariate analysis but not included in the multivariable model because only a subset of dogs (<28) had these variables included in the data set. Continuous variables were centered prior to analysis. The assumption of proportional hazards was tested via graphs and a likelihood ratio test. All analyses were performed using commercially available statistical softwarea.
Results
Study Population
Forty-five dogs met the inclusion criteria. All dogs were diagnosed between January 2002 and June 2016. These included 20 castrated males and 25 spayed females. The mean age was 8.6 ± 3.7 yr. The mean weight was 17.8 ± 12.2 kg. The population of dogs included mixed breeds (n = 14), Labrador retrievers (n = 3), pugs (n = 3), Boston terriers (n = 2), and a variety of other pure breeds (n = 23).
Clinical Signs, Physical Examination Findings, and Concurrent Diseases
Vomiting was the most commonly reported clinical sign, present in 43 of 45 dogs (95%). Other commonly reported clinical signs included inappetence (41/45 dogs, 91%) and diarrhea (11/45 dogs, 24%). Duration of clinical signs ranged from 1 to 44 days (median 8 days). Duration of clinical signs was ≤14 days in 38 dogs, >14 days in 5 dogs, and unknown in 2 dogs. Physical examination at the time of admission revealed abdominal pain in 31 of 45 dogs (69%), icterus in 29 of 45 dogs (64%), dehydration in 20 of 45 dogs (44%), and an elevated rectal temperature (≥102.5°F) in 10 of 45 dogs (22%).
Twenty-eight of 42 dogs (67%) had concurrent diseases. Six dogs had endocrine disease, including diabetes mellitus (n = 3), hyperadrenocorticism (n = 1), combined diabetes mellitus and hyperadrenocorticism (n = 1), and hypothyroidism (n = 1). One dog had an adrenal mass at the time of diagnosis; however, it was undetermined if the mass was functional. Four dogs had systolic heart murmurs, with two of the dogs having known mitral valve disease. One of these dogs had congestive heart failure 3 yr prior to presentation. One dog had historical sick sinus syndrome. Nine dogs had neurologic disease, including idiopathic epilepsy (n = 5), vestibular disease (n = 2), intervertebral disc disease (n = 1), and multifocal central nervous system disease (n = 1). Eight dogs had dermatologic disease, including allergies with otitis externa (n = 6), vulvar dermatitis (n = 1), and aural hematoma (n = 1). Five dogs had renal or lower urinary tract disease, including cystic calculi and/or crystalluria (n = 2), proteinuria (n = 1), leiomyoma of the bladder wall (n = 1), and cystitis (n = 1). Two dogs tested positive for antibodies against the C6 peptide of Borrelia burgdorferib. Other concurrent diseases included gastritis (n = 1), recent exploratory laparotomy (n = 1), and tracheal collapse (n = 1). One dog had both hepatocellular carcinoma and a gastrointestinal stromal cell tumor on biopsies performed at the time of surgery.
Clinicopathologic Data
Complete Blood Count
Complete blood count results were available from 40 dogs at the time of presentation. Eighteen of 40 dogs (45%) had neutrophilia, with 9 of these 18 dogs having band neutrophilia. Fourteen of 40 dogs (35%) were anemic. Five of 40 dogs (12.5%) were thrombocytopenic, whereas 8 of 40 dogs (20%) had thrombocytosis. Selected complete blood count data for dogs who survived to discharge and those who did not are included in Table 1.
Serum Chemistry
Serum total bilirubin was increased in all dogs for whom a value was available (n = 42), ranging from 0.6 to 28.3 mg/dL. Other abnormalities included increases in serum alkaline phosphatase in 41 of 42 dogs (98%), alanine aminotransferase in 39 of 42 dogs (93%), and cholesterol in 37 of 40 dogs (93%). Seven of 40 dogs (17.5%) were azotemic at admission, as defined by blood urea nitrogen (BUN) greater than 30 mg/dL and/or creatinine greater than 1.6 mg/dL. Urine specific gravity was not measured in any of these seven dogs prior to initiation of fluid therapy. Selected serum chemistry data for dogs who survived to discharge and those who did not are included in Table 1.
Coagulation Parameters
Thirty dogs had PT and aPTT recorded. The mean PT was 8.4 ± 1.68 s (reference range 6.8–10.2 s). The mean aPTT was 16.6 ± 21.1 s (reference range 10.7–16.4 s). One dog had prolonged PT, whereas two dogs had prolonged aPTT. D-dimers were increased in 11 of 17 dogs (65%; mean 0.59 ± 0.59 µg/mL; reference range <0.2 µg/mL).
SIRS
Sufficient clinical data were available to assess criteria of SIRS in 35 dogs. SIRS was present in 13 of 28 dogs (46.4%) who survived to discharge and in 3 of 7 dogs (42.9%) who did not survive to discharge. SIRS was present in 6 of 12 dogs (50%) who were treated surgically and in 10 of 23 dogs (43.5%) who were treated medically. There was no statistically significant difference in the frequency of SIRS in survivors versus nonsurvivors or in surgically versus medically treated dogs.
Ultrasonography
All ultrasonographic examinations were performed or reviewed by a radiologist who is board certified by the American College of Veterinary Radiology. This included 10 different radiologists over the study period.
Pancreas
Abdominal ultrasonography was performed in 45 dogs. The pancreas was diffusely or segmentally hypoechoic and surrounded by hyperechoic mesentery in all dogs. The pancreas was subjectively enlarged or thickened in 32 of 45 dogs (71%), with irregular or lobular margins in 15 dogs (33%). Hypoechoic, dissecting lesions were noted in one dog, and hyperechoic streaking was present in three dogs. Fifteen dogs (33%) had more severe changes in the right limb, whereas two dogs (4%) had more severe changes in the left limb. Two dogs had hypoechoic changes extending into the periphery of the pancreas. Cranial abdominal lymph node enlargement was noted in 19 dogs, with the hepatic (n = 11) and pancreaticoduodenal (n = 11) lymph nodes most commonly affected. Hypoechogenicity was noted in the mesenteric (n = 3), gastric (n = 2), jejunal (n = 1), medial iliac (n = 1), and hypogastric (n = 1) lymph nodes. Abdominal effusion was present in 21 dogs (46%). The quantity of effusion was described as scant in 6 dogs, mild in 10 dogs, and moderate in 2 dogs. No estimate of quantity was reported in three dogs. Cytologic evaluation of effusion by a clinical pathologist was performed in one dog, which was consistent with bile peritonitis. One other sample was examined briefly by the admitting emergency clinician to rule out septic peritonitis but was not characterized further.
Biliary System
Forty-three dogs had CBD dilation at admission, whereas the remaining two dogs developed CBD dilation on subsequent ultrasound examinations during hospitalization. The mean CBD diameter was 8.0 ± 3.5 mm. CBD diameter in dogs who survived to discharge and in those who did not are included in Table 1. Cystic duct appearance was reported in 37 dogs, with 27 of 37 dogs (73%) showing cystic duct dilation. Cystic duct diameter was measured in 10 of these 27 dogs, with a mean of 7.05 ± 3.57 mm. Intrahepatic bile ducts were described in 32 dogs and were reported as dilated in 11 dogs (34%). Similarly, extrahepatic bile ducts were described in 28 dogs, with 11 dogs (39%) having evidence of dilation. Subjective gallbladder distension was noted in 40 of 45 dogs (88%). Data regarding frequency of subjective gallbladder distension and cystic duct dilation in dogs who survived to discharge and in those who did not are included in Table 2.
Treatment
Fifteen dogs (33%) were managed surgically, whereas 30 dogs (67%) were managed medically. Surgical management was evenly distributed throughout the study period, with eight dogs having surgery in the first half of the study period (2002–2009) and seven dogs having surgery in the second half of the study period (2010–2016). Surgical management included stent placement in most dogs (n = 11). Of the four remaining dogs, one had a cholecystojejunostomy, one had a tube cholecystostomy, and two had an exploratory laparotomy performed. One exploratory laparotomy was performed when bile peritonitis was identified after the dog had been medically managed for 14 days with failure to improve clinically. Surgery revealed bile leakage from a hepatic duct. A liver lobectomy was performed; however, there was evidence of continuous bile leakage following surgery, and the patient was euthanized 36 hr later. The surgical report for this case did not mention evaluation of the CBD or describe correction of EHBDO, although the report described the pancreas as enlarged and inflamed. A liver biopsy performed during surgery was consistent with the presence of EHBDO (canalicular bile stasis with periductular fibrosis and lymphoplasmacytic inflammation). A case summary provided by the surgeon stated that the hepatic duct rupture was suspected to be due to a combination of severe regional inflammation and EHBDO secondary to pancreatitis. A second patient had an exploratory laparotomy performed by their primary care veterinarian prior to referral; however, no surgical report was available. This case was included as a surgically treated case for the purpose of risk factor assessment. Thirteen of 14 available surgical reports confirmed the gross appearance of pancreatitis and EHBDO suspected to be caused by regional inflammation associated with pancreatitis. Pancreatic biopsy was attempted in 7 of 15 dogs, but only 3 of 7 obtained pancreatic tissue (n = 1 with chronic pancreatitis, n = 2 with no significant lesions). The remaining four biopsies were peripancreatic tissue (n = 3) or peripancreatic lymph node (n = 1), all showing regional inflammation. Liver biopsies were performed in 10 of 15 dogs. Eight of the 10 had histologic changes consistent with the presence of EHBDO including periductular fibrosis or fibroplasia (n = 6), canalicular bile stasis (n = 5), lymphoplasmacytic periductular inflammation (n = 2), and/or cholangitis (n = 2 chronic neutrophilic, n = 1 lymphoplasmacytic). The two dogs lacking liver biopsy evidence of EHBDO had focal mass lesions biopsied: one was a nodule of hyperplasia; one was hepatocellular carcinoma.
Medical management consisted of a variety of fluid resuscitation protocols, blood products, antibiotics, antiemetics, antacids, and analgesics. The most commonly used antibiotics were enrofloxacin (29 of 45 dogs, 64%), metronidazole (28 of 45 dogs, 62%), and ampicillin (27 of 45 dogs, 60%). Other antibiotics administered included cefazolin (6 of 45 dogs, 13%), clindamycin (2 of 45 dogs, 4%), cephalexin (1 of 45 dogs, 2%), and cefoxitin (1 of 45 dogs, 2%). Antiemetics administered included metoclopramide (32 of 45 dogs, 71%), ondansetron or dolasetron (24 of 45 dogs, 53%), and maropitant (21 of 45 dogs, 47%). Antacids included famotidine (25 of 45 dogs, 56%), omeprazole or pantoprazole (15 of 45 dogs, 33%), and ranitidine (8 of 45 dogs, 18%). Eight of 45 dogs (18%) received sucralfate. Six of 45 dogs (13%) received fresh frozen plasma transfusions, whereas 2 of 45 dogs (4%) each received a packed red blood cell transfusion. One dog received packed red blood cells and fresh frozen plasma. Nutritional management included jejunostomy tubes (4 of 45 dogs, 9%), total parenteral nutrition (4 of 45 dogs, 9%), total parenteral nutrition in combination with a jejunostomy, gastrostomy or nasoesophageal tube (3 of 45 dogs, 7%), esophagostomy tubes (2 of 45 dogs, 4%), nasoesophageal tubes (2 of 45 dogs, 4%), and a nasogastric tube (1 of 45 dogs, 2%). Decompressive cholecystocentesis was performed in one dog on two occasions, but that dog eventually had cholecystojejunostomy performed and was included in the surgical group.
Survival
The median survival time of all dogs was 241 days (95% confidence interval [CI] 25–631; Figure 1). Thirty-four of 45 dogs (76%) survived to discharge. The median survival time of dogs who survived to discharge was 603 days (95% CI 177–981). Thirteen of 15 dogs (87%) who were treated surgically survived to discharge, whereas 21 of 30 dogs (70%) treated medically survived to discharge. For dogs surviving to discharge, there was no statistically significant difference in median survival time between those treated surgically (279 days, 95% CI 25–900) and those treated medically (775 days, 95% CI 177–1396; P = .75). Of nonsurvivors, four were euthanized and seven died. Twenty-nine of 45 dogs (64%) were known to be alive 30 days following hospital discharge, including 19 dogs who were treated medically and 10 dogs who were treated surgically. Selected continuous (Table 1) and categorical (Table 2) variables in dogs who survived to discharge and in those who did not are presented in table format.
Citation: Journal of the American Animal Hospital Association 56, 5; 10.5326/JAAHA-MS-6985
Cox univariate survival analysis identified 11 variables with P < .20 (age, diarrhea, inappetence, body temperature, alkaline phosphatase, cholesterol, azotemia, bands, hematocrit, cystic duct dilation, and gall bladder distention). Multivariable survival methods identified four factors significantly associated with survival. Controlling for body temperature at admission and the presence of subjective gallbladder distension on ultrasound, dogs 9 yr of age or older with azotemia at admission had a 9.9 greater hazard for death (95% CI 2.5–38.1; P = .001) compared with dogs younger than 9 yr of age without azotemia at admission, whereas dogs 9 yr of age or older without azotemia at admission had a 3.2 greater hazard for death (95% CI 1.3–8.1; P = .013) compared with dogs younger than 9 yr of age without azotemia at admission (Figures 2–4). Controlling for age, the presence of azotemia at admission, and body temperature at admission, dogs without subjective gallbladder distension on ultrasound had a 4.4 greater hazard for death (95% CI 1.3–15.4; P = .018) compared with dogs with subjective gallbladder distension on ultrasound. Controlling for age, the presence of azotemia at admission, and the presence of subjective gallbladder distension on ultrasound, dogs with a body temperature ≥102.5°F at admission had a 3.1 greater hazard for death (95% CI 1.3–7.7; P = .013) than dogs with a body temperature of <102.5°F at admission.
Citation: Journal of the American Animal Hospital Association 56, 5; 10.5326/JAAHA-MS-6985
Citation: Journal of the American Animal Hospital Association 56, 5; 10.5326/JAAHA-MS-6985
Citation: Journal of the American Animal Hospital Association 56, 5; 10.5326/JAAHA-MS-6985
No significant association was detected between survival and magnitude of serum liver enzyme activity, degree of hyperbilirubinemia, degree of CBD dilation, presence of SIRS, or surgical versus medical management. Similarly, no significant difference was detected in complete blood count results between survivors and nonsurvivors. The presence of concurrent disease did not show an association with survival.
Discussion
Although clinical findings and outcome of dogs with EHBDO secondary to pancreatitis have been described in a small number of cases with surgical6,15–18 and medical management,19 this study provides such information in a cohort of dogs with a specific focus on clinically suspected EHBDO secondary to pancreatitis. Thirty-four of 45 dogs (76%) survived to discharge. The median survival time of all dogs was 241 days (95% CI 25–631). In this population of dogs with clinically suspected EHBDO secondary to pancreatitis, age ≥9 yr, azotemia at admission, body temperature ≥102.5°F at admission, and a lack of subjective gallbladder distension were associated with an increased risk of death or euthanasia.
Seventy-six percent of dogs in our study with clinically suspected EHBDO secondary to pancreatitis survived to discharge. This is similar to a published survival rate in dogs undergoing surgery for EHBDO secondary to acute pancreatitis,6 and similar to most reported survival rates for dogs with acute pancreatitis, although there is a wide range of survival from 32 to 81%.6–10 A wide range of survival (10–90%) has also been reported for humans with pancreatitis, depending on the severity of disease and complications.23 Clinical severity indices for acute pancreatitis in dogs have been described,9,24 but we did not have sufficient data for most patients in this retrospective study to determine a clinical severity index score. It would be interesting in the future to evaluate whether a clinical severity index would correlate to outcome in a population of dogs with EHBDO secondary to pancreatitis. We were able to assess criteria for SIRS in the majority of dogs in our population, although the presence of SIRS was not associated with risk of death or euthanasia in this population.
Prior studies have evaluated risk factors for the development of pancreatitis in dogs.8,20,21 Most of these factors, including sex, breed, history of epilepsy, and history of endocrinopathy, were not associated with outcome in the dogs of our study. Because these factors have been associated with development of pancreatitis and not with outcome, the lack of association with outcome in our population of dogs is not surprising. We were unable to assess overweight body condition, history of dietary indiscretion, or history of prior gastrointestinal disease, other known risk factors for the development of pancreatitis,20,21 owing to inconsistent reporting in the medical records. Similar to the dogs of our study, the presence of azotemia has been associated with a poor outcome in dogs undergoing biliary surgery.16 However, other negative prognostic indicators for dogs undergoing biliary surgery15,16 including clinicopathologic parameters relating to liver function and coagulopathy, did not affect outcome in our study, although few dogs in our study had evidence of coagulopathy. We were unable to evaluate postoperative hypotension and dyspnea, reported negative prognostic indicators for dogs undergoing biliary surgery,15,16 owing to insufficient detail in the medical records. Only one dog in our population had a cholecystoenterostomy procedure, and none had septic bile peritonitis, so these reported negative prognostic indicators could not be evaluated. Differences between prognostic factors for dogs in our study and for the previously reported populations of dogs undergoing biliary surgery are to be expected, as only a small percentage of the dogs undergoing biliary surgery in prior studies had EHBDO secondary to pancreatitis. Furthermore, our population included dogs who were treated both surgically and medically. To the authors’ knowledge, this study is the first to evaluate prognostic factors in dogs with clinically suspected EHBDO secondary to pancreatitis. Dogs in our study who were younger than 9 yr of age without azotemia had a better prognosis for survival compared with dogs 9 yr of age or older with or without azotemia. Increasing age has been associated with a worse outcome in dogs undergoing biliary surgery and in humans undergoing medical treatment for acute pancreatitis.15,23 The impact of increasing age on the outcome in our study could be related to the higher likelihood that older dogs have concurrent disease. However, concurrent disease was not associated with outcome in this population. Chronic pancreatitis could be associated with advanced age. This can be difficult to diagnose and may have been present in some dogs of our study, potentially contributing to the effect of age on prognosis. Only one dog in our population was a breed reportedly predisposed to chronic pancreatitis (a Cavalier King Charles spaniel), but chronic pancreatitis may occur in dogs of any breed. Age may impact an owner’s decision regarding pursuit of aggressive treatment. Azotemia has been documented as a poor prognostic indicator for humans and dogs undergoing extrahepatic biliary surgery16,25 as well as for dogs receiving medical management for acute pancreatitis.26 Azotemia in dogs with EHBDO secondary to pancreatitis may be prerenal (secondary to dehydration) or renal in origin. Experimental mouse models show that the absence of bile salts in the gastrointestinal lumen leads to absorption of endotoxin, which may play a role in acute kidney injury via vasoconstriction, decreased glomerular filtration rate, and tubular degeneration.27 It is unknown whether these effects may occur in dogs. Renal azotemia may also occur in this population secondary to hypovolemia and decreased renal perfusion as a result of SIRS. We were unable to determine the origin of azotemia in the dogs of our study, as urine specific gravity results prior to the initiation of fluid therapy were not available. Additional data that may have helped differentiate prerenal from renal azotemia, such as improvement with fluid therapy or other indicators to suggest dehydration or hypovolemia (packed cell volume/total solids, lactate), were not collected. This represents a limitation of the assessment of azotemia in the dogs of our study. It is also important to note that we defined azotemia as an increase in BUN, creatinine, or both. The previous study identifying azotemia as a poor prognostic indicator in dogs undergoing extrahepatic biliary surgery defined azotemia as increased creatinine.16 Creatinine is less affected by nonrenal factors than is BUN. Increased BUN could be caused by gastrointestinal bleeding or high protein intake. Increased BUN and increased creatinine did not impact survival when evaluated as independent variables. Azotemia in our population may be a poor prognostic indicator because it reflects increased severity of disease, although it is possible that the presence of azotemia impacted owners’ decisions regarding pursuit of treatment.
We found that an elevation in body temperature ≥102.5°F at admission was associated with a worse prognosis. Elevated body temperature at admission could indicate increased severity of disease such as severe inflammation, necrosis, or abscessation, resulting in a worse outcome. Elevated body temperature is often present in dogs with sepsis or necrotizing cholecystitis.1,2,28 It has previously been reported as a negative prognostic indicator for humans undergoing biliary surgery.29 However, some reports have failed to show prognostic significance of an elevated body temperature on survival in dogs treated medically for pancreatitis26 or those undergoing biliary surgery.14–16 Evaluation of an elevated body temperature has limitations given the retrospective nature of our study. Physiologic causes of elevated body temperature such as patient stress and ambient temperature cannot be assessed with our study design. Persistence of elevated body temperature may have provided more evidence for true fever versus transient hyperthermia; however, resolution could also be related to therapeutic intervention. Therefore, we only assessed the effect of the body temperature at the time of admission. Assessment of body temperature only at admission also limits the reliability of our evaluation of SIRS. All variables used for determination of SIRS were recorded at admission. This could result in both false-positive characterizations (for example, physiologic hyperthermia or tachycardia associated with stress) and false-negative characterizations (for example, in patients who developed SIRS during hospitalization). Serial measurement of variables would have been ideal, but this information was not available for many dogs.
A difficult result to explain was that a lack of subjective gallbladder distension was associated with a worse outcome. One theory is that a lack of gallbladder distension in patients with EHBDO may have been due to a small rupture in the biliary system, resulting in bile peritonitis. Previous studies have associated septic and sterile bile peritonitis with a poor clinical outcome.14,30 However, the results of our study do not support this theory, as only three of five dogs lacking subjective gallbladder distension had abdominal effusion and none of the three with effusion had their effusion analyzed to confirm bile peritonitis. Other possible explanations for the lack of gallbladder distension in some dogs could include chronic cholecystitis with fibrosis and lack of distensibility, prior CBD obstruction, or ductal plate anomaly as the cause for CBD dilation. Determination of gallbladder distension was based on the subjective assessment of the radiologist performing the study. Because 10 different radiologists performed the studies over the study period, this subjective interpretation may have differed between radiologists. Gallbladder measurements were not evaluated to obtain objective data regarding gallbladder volume. Therefore, the consistency of identification and reporting of gallbladder distension is questionable. The reason for the noted association between a lack of subjective gallbladder distension and outcome is unknown.
Survival was not impacted by whether dogs were treated medically or surgically. Although surgical management and outcome of dogs with EHBDO secondary to pancreatitis has been fairly well described,6,15–18 information regarding medical management of this condition is limited.19 Although treatment type did not impact outcome, the study design is not able to answer the question of whether surgical or medical management is superior. The retrospective design does not allow us to determine reasons for which dogs were taken to surgery or not. The detail that we are able to extract from medical record review is limited. Furthermore, the number of cases in each group may leave us underpowered to detect a significant difference. It is also important to recognize that most cases taken to surgery were done so after a failure of medical management. This likely biases the surgical group to more severely affected cases and may result in an apparently worse outcome. Moreover, two of the surgical cases did not have surgical confirmation of EHBDO (one case with a surgical report that does not mention CBD patency, one case lacking a surgical report) or have documented surgical correction of EHBDO. We elected to include these cases in the study as they provided much data for assessment of outcome and prognostic factors. All cases in which surgery was performed were included in the surgically treated group when assessing the variable “surgical versus medical management.” Based on our results, it is clear that some dogs with clinically suspected EHBDO secondary to pancreatitis can survive without surgical intervention. However, failing to relieve a biliary obstruction in a timely manner may result in longer hospitalization time and systemic complications such as SIRS, sepsis, and death.8,18,30 The decision to pursue medical or surgical management should be made on a case-by-case basis considering all of the clinical information.
The study design provides limited detail regarding pancreatitis in the dogs of our study. Diagnosis of pancreatitis was based on clinical signs and ultrasonographic findings, as pancreatic biopsy samples were evaluated in only three dogs. This prevented us from determining the severity of pancreatitis and the presence of necrotizing pancreatitis, which may have affected prognosis. We were unable to determine if dogs had acute, chronic, or acute-on-chronic pancreatitis. Although most dogs had a short duration of clinical signs, several had clinical signs of >2 wk in duration, which would suggest a more chronic disease process. One pancreatic biopsy showed evidence of chronic pancreatitis. It is likely that some of the dogs had chronic pancreatitis. Although uncommon, chronic pancreatitis has been reported to cause EHBDO.15,17,31 Ultrasonographic findings in dogs with chronic pancreatitis vary widely but may include pancreatic enlargement and hypoechogenicity,31,32 which meet our inclusion criteria. It is unknown how the proportion of dogs with chronic pancreatitis may affect the results of our study, but dogs with chronic pancreatitis tend to have a better prognosis than those with acute pancreatitis, with death typically related to another disease process.31,32
In addition to those noted in the previous paragraphs, this study had several other important limitations. A sample size calculation for any prespecified variables was not performed prior to data collection; therefore, there is the possibility of type II error for any nonsignificant variables. Included cases were managed over a wide time range (2002–2016). It is possible that changes in management strategies over time contributed to differences in outcome. Owing to missing data for some dogs, multivariate analysis was not performed for each variable, which may have caused us to miss information that is associated with outcome. No dog younger than 9 yr of age was azotemic, precluding the comparison of azotemic and nonazotemic dogs in this age group. Ultrasonographic diagnoses of pancreatitis and EHBDO were based on radiology reports that had been finalized by a radiologist board certified by the American College of Veterinary Radiology. This included 10 different radiologists over the 14 yr time period during which cases were evaluated. Review of ultrasonographic images by a single radiologist would have strengthened the study by confirming information present in reports. Although all surgical reports described gross findings consistent with EHBDO and pancreatitis, we lacked this confirmation in medically treated dogs. Other conditions causing EHBDO may have been missed in medically managed dogs, or the dogs may have had another reason for bile duct dilation. However, clinical and ultrasonographic findings are relied upon in a clinical setting, so our results are generalizable to a true clinical population. The retrospective nature of the study also resulted in limited information regarding the reason for euthanasia, making it challenging to discern whether the decision was based on medical or financial factors.
Conclusion
In this population of dogs with clinically suspected EHBDO secondary to pancreatitis, the median survival time was 241 days (95% CI 25–631). Thirty-four of 45 dogs (76%) survived to discharge. Age ≥9 yr, azotemia at admission, body temperature ≥102.5°F at admission, and a lack of subjective gallbladder distension were significantly associated with an increased risk of death or euthanasia. These results should help clinicians provide owners with information regarding prognosis and potential outcome for dogs with this condition.
Survival curve for 45 dogs with clinical suspicion of extrahepatic bile duct obstruction and pancreatitis.
Survival curve for dogs <9 yr old without azotemia at admission, with or without subjective GB distension on ultrasound and temp ≥102.5°F at admission. GB, gallbladder; temp, body temperature.
Survival curve for dogs ≥9 yr old without azotemia at admission, with or without subjective GB distension on ultrasound and temp ≥102.5°F at admission. GB, gallbladder; temp, body temperature.
Survival curve for dogs ≥9 yr old with azotemia at admission, with or without subjective GB distension on ultrasound and temp ≥102.5°F at admission. GB, gallbladder; temp, body temperature.
Contributor Notes
aPTT (activated partial thromboplastin time); BUN (blood urea nitrogen); CBD (common bile duct); CI (confidence interval); EHBDO (extrahepatic bile duct obstruction); PT (prothrombin time); SIRS (systemic inflammatory response syndrome)
