Pathological evidence of pancreatitis in 43 horses (1986–2011)

Authors


Summary

Reasons for performing study

Definitive ante mortem diagnosis of pancreatitis in horses is difficult. Reports summarising the most common clinical signs, clinicopathological features and concurrent disorders in horses with a definitive diagnosis of pancreatitis that may aid in the recognition of disease are lacking.

Objectives

To describe case details, clinical signs, clinicopathological data and necropsy findings in horses with a definitive diagnosis of pancreatitis.

Methods

This was a retrospective study (1986–2011) and inclusion criteria consisted of horses with a definitive diagnosis of pancreatitis. A medical records database search was performed and data extracted included case details, clinical signs, clinical laboratory data and post mortem findings. Pancreatitis was defined as acute, active chronic or chronic and presumed primary or secondary, based on post mortem findings.

Results

Pancreatitis was diagnosed in 43 horses (acute pancreatitis in 34, active chronic in 4 and chronic in 5). A presumed diagnosis of primary pancreatitis was made in 6 horses. Pancreatitis was associated with gastrointestinal disorders in 28 horses (14 large colon, 10 small intestine and 4 gastric ruptures) and primary hepatic disease in 3 horses. Six horses had pancreatitis associated with other disorders: multiple endocrine neoplasia syndrome (one horse), strychnine toxicosis (one horse) and compromised immune system (4 horses).

Conclusion

Pancreatitis is an uncommon disorder that can occur as a primary problem or secondary to gastrointestinal, hepatic or immunocompromising disorders, and when it occurs it affects adult horses more commonly.

Potential relevance

Unexplained abdominal pain, gastric dilation or rupture, peritonitis and/or the presence of white fibrinous plaques and fat necrosis in the peritoneum and mesentery or mass-like structures in the root of the mesentery during an exploratory celiotomy should raise a suspicious of pancreatitis.

Introduction

Pancreatitis in the horse is uncommonly reported and primarily described as occurring in adult horses [1-8]. In 1937, Higginson reported 6 adult horses with partial or complete obstruction of the pancreatic duct that had pancreatitis as the main pathological finding [8]. Depending on the severity and duration of the problem, clinical signs of pancreatitis vary among horses. Horses with acute pancreatitis are presented with marked abdominal pain, gastrointestinal reflux and hypovolaemic shock [1, 5]. The main gross pathological findings in these cases are an enlarged pancreas and distended stomach [1, 5]. Chronic pancreatitis can cause weight loss, inappetance, lethargy and mild or recurrent signs of colic [9]. Migration of Strongylus equinus and Parascaris equorum to the pancreas can produce pancreatic tissue destruction and extensive fibrosis [4, 6]. Pancreatitis has also been documented in foals at post mortem examination [10-12]. Acute pancreatic necrosis in a foal was associated with recumbency, gastrointestinal reflux and hypovolaemic shock [12], and subacute to chronic pancreatitis has been associated with duodenitis, duodenal ulcers and possible vasculitis in foals [10, 11].

An ante mortem diagnosis of pancreatitis is difficult owing to the presence of nonspecific clinical signs and the lack of a definitive diagnostic test. Reference values for serum and peritoneal fluid amylase and lipase activities in horses have previously been published [13] but their accuracy for the diagnosis of pancreatitis in horses has not been evaluated [5, 12]. Serum amylase is not specific to pancreatic tissue. Damage to or inflammation of the small intestine and primary renal failure can also lead to increased amylase activity in horses [14]. Pancreatic injury as evidenced by changes in serum trypsin activity in horses with colic has been evaluated [15]. However, trypsin is produced in low amounts in the equine pancreas [16] and originates in other tissues in rodents and man [17]. The apparent lack of a specific site of enzyme production (amylase, lipase, trypsin) in the various tissues, nonspecific clinical signs and laboratory findings make the clinical diagnosis of pancreatitis in live animals challenging.

To the authors' knowledge, there are no reported descriptions of a larger population of horses with a definitive diagnosis of pancreatitis. The purpose of this study was to document case details, clinical presentations and concurrent pathological findings in horses with a confirmed diagnosis of pancreatitis at necropsy.

Materials and methods

The selection criteria for this study included horses of any breed, gender and age with a definitive diagnosis of pancreatitis based on post mortem findings. Evidence of pancreatitis included peripancreatic fat necrosis, pancreatic haemorrhage, oedema, necrosis, fibrosis, atrophy or inflammatory cell infiltration. The computerised medical records database from the University of California, Davis was searched between the years of 1986–2011 under the visit, presumptive diagnosis and pathological fields using the word: pancreatitis, pancreatic, pancreas disease. The following data were collected: case details, history, clinical signs, clinical pathological data and results from other diagnostic modalities, if available, and post mortem findings. Based on post mortem examination findings, pancreatitis was defined as acute, active chronic or chronic. This standardised classification scheme taken from human medicine has been adapted previously to small animal medicine [18]. Evidence of acute pancreatitis included the presence of peripancreatic fat necrosis, oedema, haemorrhage, pancreatic acinal cell necrosis or neutrophilic infiltration. Evidence of chronic pancreatitis included lymphoplasmacytic inflammation with permanent histopathological changes such as fibrosis and/or acinal atrophy. Chronic active pancreatitis was characterised by the concurrent presence of fibrosis and pancreatic cell necrosis or peripancreatic fat necrosis. In addition, horses were divided into 2 main groups, presumed primary pancreatitis and secondary pancreatitis. Horses with a presumed primary pancreatitis included horses that had pancreatic lesions as the main or only pathological finding that would explain their clinical signs. Horses with other mild pathological lesions that would not explain clinical signs were also included in this group. Secondary pancreatitis included horses with pancreatic lesions associated with or presumed to be secondary to other disorders.

Results

Forty-three horses met the inclusion criteria based on post mortem findings. During the same study period, 132,735 horses (264,407 visits) were admitted to our teaching hospital and, of those, 11,150 horses (14,122 cases) were admitted with a presenting complaint of colic. The median age of horses with pancreatitis was 11 years (range 2 months to 28 years). Only 2 horses were less than 1 year old, a 2-month-old Arabian colt and a 2-month-old Thoroughbred filly. Excluding the 2 foals, the median age was 11.5 years (range 1–28 years). The affected horses included 23 females and 20 males (15 geldings and 5 intact animals). Breeds identified included 17 Quarter Horses and related breeds (10 Quarter Horses, 4 Appaloosas, 3 Paints), 7 Thoroughbreds, 3 Arabians, 3 Morgans, 2 Warmbloods, 2 Draught, one Standardbred, one Mustang, one Fox Trotter, one pony and 5 other mixed breeds of horses. Presumed primary pancreatitis was diagnosed in 6 horses, and secondary or associated pancreatitis was diagnosed in the remaining 37 horses. Nineteen of 26 horses for which information was available in the medical records had gastric reflux on presentation to the hospital. Secondary pancreatitis was associated with gastrointestinal disorders in 28 horses (14 large colon, 10 small intestine and 4 gastric rupture), hepatic diseases in 3 and other disorders in 6 horses (multiple endocrine neoplasia-like syndrome, one horse; strychnine toxicosis, one horse and altered immune system function, 4 horses). Horses' age and physical examination findings on presentation (rectal temperature and heart and respiratory rates) by groups are presented in Table 1. Results from blood (complete blood count [CBC] and biochemistry panel) and peritoneal fluid analyses by groups are presented in Tables 2 and 3, respectively. The number of horses with acute, active chronic and chronic pancreatitis, peripancreatic fat necrosis, periportal hepatitis and cholangiohepatitis-cholelithiasis are presented in Table 4. Representative histological images of a normal pancreas and the pancreas of horses diagnosed with acute and chronic pancreatitis are presented in Figures 1-3, respectively

Figure 1.

Normal exocrine pancreas in a horse. Haematoxylin and eosin, bar = 25 μm.

Figure 2.

Acute exocrine pancreatitis in a horse. Large numbers of neutrophils infiltrate the pancreatic parenchyma and percolate between the intralobular septae and acini. The peripancreatic fat (inset) has evidence of necrosis. The arrow indicates saponification–chelation of calcium with fatty acids liberated by pancreatic enzymes (saponification). Haematoxylin and eosin, bar = 25 μm.

Figure 3.

Chronic pancreatitis in a horse. The arrows indicate the marked fibrosis between acinar lobuli (A) and surrounding the pancreatic ducts (D). There is an interstitial mononuclear cell infiltrate (asterisks). Haematoxylin and eosin, bar = 100 μm.

Table 1. Age, temperature and heart and respiratory rates of horses with a definitive pathological diagnosis of pancreatitis at presentation
GroupAgeTemperatureHeart rateRespiratory rate
No. horses (n = 43)(years)(°C)(beats/min)(breaths/min)
  1. Horses were divided based on presumed primary and secondary pancreatitis. Horses with secondary pancreatitis included horses with pancreatic lesions associated or presumed to be secondary to abnormalities affecting other organs. Values represent median (range), n = number of horses.
Presumed primary13.5 (7–19)38.2 (37.5–39.4)52 (44–80)38 (20–48)
(n = 6)n = 6n = 6n = 6n = 6
Large colon7 (1–22)38.1 (37.2–39.6)48 (32–96)21 (8–36)
(n = 14)n = 14n = 12n = 11n = 10
Small intestine9.5 (0.17–23)37.8 (37.3–38.6)77 (36–136)31 (20–48)
(n = 10)n = 10n = 7n = 10n = 8
Hepatic12 (10–13)37.4 (36.7–38.1)60 (60–60)20
(n = 3)n = 3n = 2n = 2n = 1
Gastric rupture7.5 (1–14)38.3 (37.2–38.6)60 (48–84)31 (14–48)
(n = 4)n = 4n = 3n = 4n = 4
Other12 (0.17–28)37.7 (37.2–38.9)66 (60–72)45 (14–75)
(n = 6)n = 5n = 3n = 2n = 2
Table 2. Results from the CBC and serum biochemistry analysis of horses with a definitive pathological diagnosis of pancreatitis
GroupPCVTPWBCNeutBands/toxicGlucoseGGTSDHASTBile acidsLipaseTotal bilir
No. of horses (n = 43)%g/dln/uln/uln/ulmg/dliu/liu/lIU/lμmol/liu/lmg/dl
  1. Cases were grouped as presumed primary or secondary pancreatitis. Values represent median (range), n = number of cases. PCV = pack cell volume, TP = total protein, WBC = white blood cells, Neut = neutrophils, GGT = gamma-glutamyltransferase, SDH = succinate dehydrogenase, AST = aspartate aminotransferase, total bilir = total bilirubin.
Reference range30–465.8–8.75000–11,6002600–6800Rare59–1228–220–8138–4090–2023–870.5–2.3
Presumed primary38 (35–51)7.4 (6.5–8.2)7100 (2400–12,175)4492 (1392–6896)269 (0–570)135 (114–199)54 (22–139)37 (29–89)813 (480–1734)7711,8235.8 (5.3–8.9)
(n = 6)n = 6n = 6n = 6n = 4n = 4n = 4n = 4n = 4n = 4n = 1n = 1n = 4
Large colon39.3 (30–67)6.5 (3.9–8.5)5975 (4000–12,500)3950 (2244–11,250)0 (0–1989)134 (115–251)15 (8–169)17 (3–132)437 (187–1199)866363.3 (1.2–8.7)
(n = 14)n = 12n = 12n = 11n = 7n = 7n = 7n = 7n = 7n = 7n = 1n = 1n = 7
Small intestine49 (28–71)7.2 (4.2–9.5)8065 (3200–15,100)7078 (5775–12,835)0 (0–2921)155 (107–437)25 (12–63)11 (2–45)424 (213–570)3.5 (1–10.4)
(n = 10)n = 10n = 9n = 8n = 5n = 4n = 6n = 6n = 6n = 6  n = 6
Hepatic (n = 3)47.5 (44–51)7.7 (7.3–8.1)11,450 (5600–17,300)8957 (4592–13,321)56 (0–112)115 (80–149)264 (64–464)35.5 (35–36)748 (539–957)115.5 (28–203)12.4 (12–12.8)
 n = 2n = 2n = 2n = 2n = 2n = 2n = 2n = 2n = 2n = 2 n = 2
Gastric rupture48.5 (46–51)8.15 (7.5–8.8)19,100 (18,500–19,700)17,139985392208479  7.8
(n = 4)n = 2n = 2n = 2n = 1n = 1n = 1n = 1n = 1n = 1n = 1
Other54 (22–60) 15,400         
(n = 6)n = 3n = 1
Table 3. Results of peritoneal fluid analysis of horses with a definitive pathological diagnosis of pancreatitis
GroupTNCCTotal proteinLipase
No. of horses (n = 43)n/ulmg/dliu/l
  1. Cases were grouped as presumed primary or secondary pancreatitis. Values represent median (range), n = number of cases, TNCC = total nucleated cell count.
Reference range≤2500<2.5 0–36
Presumed primary (n = 6)16,525 (975–161,000)2.8 (1.2–4.4) 
n = 6n = 6
Large colon (n = 14)1,725 (100–29,400)2 (0.7–3.4) 
n = 7n = 7
Small intestine (n = 10)2500 (1000–44,000)4.8 (2.7–7.4) 
n = 8n = 8
Hepatic (n = 3)29,250 (22,500–36,000)2.6 (1.5–3.6)66,000
n = 2n = 2n = 1
Gastric rupture (n = 4)21,540 (3080–40,000)2.85 (2.5–3.2)87,000
n = 2n = 2n = 1 
Other (n = 6)18,5502.9 
n = 1n = 1
Table 4. Table summarising the classification of pancreatitis and presence of periportal hepatitis, cholangiohepatitis/cholelithiasis and peripancreatic fat necrosis in 43 horses with a definitive pathological diagnosis of pancreatitis
GroupAcuteChronicActive-chronicPeriportal hepatitisCholangio hepatitis – cholelithiasisPeripancreatic fat necrosis – steatitis
Total (n = 43)345416723
  1. Cases were grouped as presumed primary or secondary pancreatitis.
Presumed primary (n = 6)600105
Large colon (n = 14)743537
Small intestine (n = 10)901516
Hepatic (n = 3)210122
Gastric rupture (n = 4)400201
Other (n = 6)600211

Presumed primary pancreatitis (6 horses)

All 6 horses with a presumed diagnosis of primary pancreatitis had acute pancreatitis and presented with severe abdominal pain unresponsive to analgesics. Five of these 6 horses had gastric reflux on presentation. Two of these 6 horses underwent exploratory abdominal midline celiotomy, during which a mass-like structure of undetermined aetiology proximal to the duodenum was palpated (determined at necropsy to be the pancreas). Both horses were subjected to euthanasia during surgery and pancreatitis was not suspected. Two of the 6 horses were subjected to euthanasia on presentation owing to owners' financial constraints and the presence of unmanageable pain. A fifth horse that was subjected to euthanasia during hospitalisation presented with signs of colic and gastrointestinal reflux. This horse developed fever (38.8°C) and continued to produce gastrointestinal reflux during hospitalisation. Blood work revealed increased serum liver enzymes (gamma-glutamyltransferase [GGT] 100 iu/l, reference range 8–22 iu/l and succinate dehydrogenase [SDH] 70 iu/l, reference range 0–8 iu/l). An ultrasound guided liver biopsy revealed normal liver tissue with mild evidence of cholestasis. Diagnostic laparoscopy was performed and multiple white plaques were identified on the peritoneal serosa. The horse was subjected to euthanasia owing to increased discomfort despite medical treatment. Serum lipase collected prior to euthanasia was 11,823 iu/l (reference range 23–87 iu/l). At necropsy, in addition to acute necrotising pancreatitis with diffuse and extensive fat necrosis, mild congestion in one of the liver lobes was detected. The sixth horse presented with a chronic history of weight loss and acute onset of colic. Rectal examination identified a gelatinous-like nonmoveable mass in the right paralumbar fossa that extended cranially. Ultrasound examination revealed a 10 cm fluid-filled mass that extended cranially from the right kidney to the level of the liver. An exploratory laparoscopy allowed sampling of fluid and ruled out a possible ruptured ureter. The horse was subjected to euthanasia for progressive colic signs. On necropsy the mass was identified to be the necrosed pancreas and peripancreatic tissue, the right kidney and the duodenum.

Secondary pancreatitis (37 horses)

Twenty-eight horses with primary gastrointestinal disease (14 large colon, 10 small intestine and 4 gastric rupture) were presented with moderate to severe abdominal pain.

Pancreatitis associated with large colon abnormalities (14 horses)

Nine of the 14 horses had a large colon volvulus, one had a right dorsal displacement of the large colon, 3 a large colon impaction and one enterotyphlocolitis. One of the 9 horses with large colon volvulus had a 10–15 cm semi-firm mass palpated, but not seen, at the root of the mesentery during surgery. An abscess was suspected owing to the suppurative-like material that was observed. Fibrinous plaques were noted along the mesentery of the small intestine. Severe engorgement of blood and lymphatic mesenteric vessels was also noted. The horse was subjected to euthanasia owing to poor access to the mass, compromised bowel and a guarded prognosis. On necropsy the mass was identified to be pancreatic necrosis with severe peripancreatic fat necrosis. The horse with enterotyphlocolitis had Salmonella type E identified by faecal culture.

Pancreatitis associated with small intestinal abnormalities (10 horses)

Eight of 10 horses were diagnosed with a small intestinal strangulating obstruction and 2 had enteritis. Three of 8 horses with small intestinal strangulation underwent surgery and the remaining 5 were subjected to euthanasia on admission. One horse underwent 2 surgeries 9 days apart. On the second surgery, a severely necrotic fatty mass at the root of the mesentery was palpated and resected. This horse developed post operative ileus, fever and peritonitis and was subjected to euthanasia. The mass was confirmed to be severe peripancreatic fat necrosis on histological examination, and severe acute necrotising pancreatitis was identified on post mortem histological evaluation. The second horse was subjected to euthanasia at a repeat laparotomy 10 days after the initial surgery during which a jejunocaecostomy was performed. The horse developed adhesions that fused to an abscess surrounding the necrotising ileal stump. A severe acute necrotising pancreatitis was also evident on necropsy. The third horse developed post operative ileus after jejunocaecostomy and was subjected to euthanasia on Day 4 after failure to respond to treatment. Acute neutrophilic pancreatitis was diagnosed on post mortem examination.

Two horses were diagnosed with enteritis unresponsive to medical treatment and were consequently subjected to euthanasia. One horse was determined to have eosinophilic enteritis based on histological examination of biopsy tissue collected during surgery. On necropsy, the horse had active chronic necrotising pancreatitis and marked peripancreatic fat necrosis. The other animal was a foal presented with ptyalism, bruxism and abdominal pain. The foal was subjected to euthanasia, and severe duodenal ulceration and diffuse neutrophilic pancreatitis were diagnosed at necropsy.

Pancreatitis associated with gastric rupture (4 horses)

Four horses had both gastric rupture and acute pancreatitis. Three of the 4 horses were diagnosed with gastrointestinal rupture at presentation and were subjected to euthanasia. One horse was treated medically for acute colic: the horse refluxed continuously and was subjected to euthanasia after 24 h. The horse had increased peritoneal fluid lipase activity (87,984 iu/l; reference range 0–36 iu/l). This horse had a partial thickness gastric rupture, extensive peripancreatic fat necrosis and an enlarged oedematous pancreas that resulted in duodenal mural constriction and dilation of the stomach and proximal duodenum. One colt had a large mesenteric lymph node abscess (Streptococcus equi equi) with adhesions causing partial duodenal obstruction that resulted in gastric rupture. The other 2 horses had no gross or histological alterations to explain an association between pancreatitis and gastric rupture.

Pancreatitis associated with liver abnormalities (3 horses)

One horse presented for acute onset of colic, and a mass on the right side of the abdomen along the root of the mesentery was palpated on rectal examination. The horse underwent a ventral celiotomy and multiple friable masses were palpated near the duodenum. The horse was subjected to euthanasia during surgery owing to duodenal rupture that occurred during manipulation of the masses. Chronic intermittent obstruction of the common bile duct by choleliths was diagnosed at necropsy. The pancreas was firm, fibrotic, enlarged, and attached to the gastric serosa and right dorsal colon. A chronic fibrosing pancreatitis was diagnosed on gross and histopathological examination.

The second horse presented for evaluation of colic with no evidence of gastrointestinal reflux. The horse had increased serum liver enzymes (GGT 64 iu/l, SDH 36 iu/l) and abdominal fluid lipase activity during hospitalisation (66,000 iu/l). Abnormalities in the liver identified by abdominal ultrasonography and biopsy raised suspicion of a bile duct obstruction. The horse was subjected to euthanasia owing to failure to respond to supportive medical therapy. On necropsy, severe portal oedema and periportal necrosis were identified. No gross evidence of obstruction was noted and a functional obstruction of the common bile duct was suspected. Acute pancreatitis with severe diffuse peripancreatic fat necrosis was identified at necropsy.

The third horse, which presented with a 2 week history of weight loss, was icteric and ataxic on admission. Increased serum liver enzymes (GGT 464 iu/l, SDH 35 iu/l) findings were consistent with hepatic dysfunction, and abdominocentesis yielded a copious amount of serosanguinous fluid. At necropsy, a lymphoplasmacytic chronic inflammation of the duodenum and jejunum was noted and a mixed population of bacteria was isolated from the bile duct by bacterial culture. The horse was diagnosed with severe chronic cholelithiasis and cholangiohepatitis and chronic enteritis with acute pancreatitis and peripancreatic fat necrosis at necropsy.

Pancreatitis associated with other disorders (6 horses)

Six of 43 horses had pancreatitis associated with other primary disorders. The first horse had a history of transient anorexia and presented for acute onset of gastrointestinal reflux and severe colic signs. This horse did not respond to medical treatment and was subjected to euthanasia. Based on post mortem findings, a diagnosis of multiple endocrine neoplasia-like syndrome was made, with marked chronic eosinophilic enteritis and acute pancreatitis. The second horse died suddenly after a short episode of stiffness, tremors and convulsions. This horse was diagnosed with acute strychnine toxicosis; histopathological evidence of acute necrotic and multifocal haemorrhagic pancreatitis, peripancreatic fat necrosis and mild eosinophilic gastroenteritis was present. High levels of strychnine were identified in the stomach contents at necropsy.

The remaining 4 of 6 horses had an altered immune system function. These included a 2-month-old Arabian colt that presented with bronchopneumonia and a history of recurrent infections. The colt was diagnosed with severe combined immunodeficiency of Arabian foals, and had acute suppurative pancreatitis on necropsy with evidence of adenovirus infection in the pancreas and lungs. Another case was an adult horse with a 2 week history of high fever, lethargy and oedema of the pelvic limbs that did not respond to aggressive medical therapy with anti-inflammatory drugs, antibiotics and corticosteroids. This horse was subjected to euthanasia and a severe diffuse generalised mononuclear inflammation of undetermined cause affecting multiple organs, including the pancreas, was identified. The results of a variety of diagnostic tests, including those for equine infectious anaemia, were negative. The 2 remaining adult horses had pituitary pars intermedia dysfunction (PPID). The first horse was admitted for acute onset of pneumonia, was subjected to euthanasia 48 h after presentation, and a diagnosis of severe suppurative necrotising pneumonia was made. Nocardia asteroides was identified on histological stains of lung tissue. A pituitary pars intermedia adenoma and adrenal cortical hyperplasia were also identified during necropsy. Peripancreatic fat necrosis was also evident, along with necrotic suppurative inflammation adjacent to the pancreas. Nocardia was not identified on histological sections of the pancreas; however, dissemination of Nocardia to the vicinity of the pancreas was suspected based on the pronounced level of inflammation. The second horse with PPID died suddenly, and severe chronic active septic pancreatitis with secondary peritonitis due to Streptococcus equi equi and Escherichia coli and concurrent mild chronic cholangiohepatitis were identified during post mortem examination.

Discussion

This study highlights key aspects of pancreatitis in horses. There was no predisposition of breed, gender or age in affected horses compared to our hospital population; however, pancreatitis appeared to be rare in foals. Moderate to severe abdominal pain was the most common clinical sign in horses with either presumed primary or secondary pancreatitis. Pancreatitis was commonly associated with gastrointestinal and liver disease and occasionally with compromised immune system function. Similar to previous case reports, we identified horses with gastric distension and/or rupture associated with acute pancreatitis [1, 5].

Four horses with acute pancreatitis also had gastric rupture. Gastric dilation has previously been associated with acute necrotising pancreatitis in horses [5]. It has been postulated that pain arising from acute necrotising pancreatitis could reduce gastric motility and produce pyloric spasms [5]. In addition, mechanical obstruction of the duodenum due to an inflamed pancreas might be responsible for the gastric distension and rupture in some cases. Owing to the retrospective nature of the study, however, it was not possible to determine whether the acute pancreatitis was the primary problem or was secondary to peritonitis as the result of gastric rupture.

In this study, 28 horses had primary gastrointestinal lesions and 12 also had mild to moderate periportal hepatitis. In addition, 3 horses had a severe primary liver disease. It has previously been hypothesised that horses with enteritis or strangulated small intestinal lesions may develop acute pancreatitis and hepatitis as the result of ascending influx of intestinal fluid through the pancreatic and bile ducts, with subsequent activation of pancreatic enzymes [7, 8, 11]. Horses with large colon abnormalities, especially those with volvulus and displacements, could have reduced blood flow to the pancreas. As occurs commonly in pancreatitis in man, it has been hypothesised that pancreatitis in horses may be due to obstruction of the common bile and pancreatic ducts. In our study, 7 horses had cholangiohepatitis and/or cholelithiasis, which allowed reflux of bile into the pancreatic duct with activation of pancreatic enzymes and the subsequent development of pancreatitis [8, 19]. Once activated, these enzymes are responsible for autodigestion of pancreatic tissue, resulting in necrosis of the acini and pancreatic islets with interstitial fat necrosis and necrotising vasculitis [19]. The release of pancreatic enzymes stimulates the production of inflammatory cytokines, thereby triggering an inflammatory cascade which leads to a systemic inflammatory response syndrome [19]. A similar mechanism might be responsible for pancreatitis in some of our horses, as many had lesions that compromised the pancreatic duct or could result in stasis and distension of the proximal region of the small intestine.

Four horses had a mass palpated at the root of the mesentery during surgery. In these cases, the masses had an abscess-like structure with suppurative-like material leaking from the base of the mesentery. Upon post mortem evaluation, these masses were confirmed to be the pancreas with peripancreatic fat necrosis. White fibrinous plaques along the mesentery and peripancreatic or diffused peritoneal fat necrosis were present in many horses with acute pancreatitis at necropsy, which should therefore alert the surgeon to possible pancreatitis if observed during exploratory surgery. In addition, 2 horses had a mass on the dorsal right side of the abdomen palpated during rectal examination, and in one the mass was seen on laparoscopy.

Six horses presented with pancreatitis presumed to be associated with other abnormalities. Both horses with strychnine toxicosis, as well as the horse with multiple endocrine neoplasia-like syndrome, had acute pancreatitis with eosinophilic gastroenteritis. Eosinophilic gastroenteritis with associated acute pancreatitis has previously been associated with Strongylus sp. larvae migration [4]. There were no post mortem findings consistent with larval migration. Acute pancreatitis has also been reported in a nonfatal strychnine poisoning case involving a young woman [20]. The proposed mechanisms in that case included direct chemical inflammation to the pancreas or damage secondary to a microcirculatory disturbance or spasm of the pancreatic duct [20].

Two horses previously diagnosed with PPID had septic necrotising pancreatitis. Although there have been no previous reports of septic pancreatitis in horses with PPID, there is a case report of acute pancreatitis in a horse with PPID [3]. Horses with PPID could be predisposed to recurrent infections owing to compromised immune system function [21]. Consequently, the septic pancreatitis in the horse in the present study may have been associated with compromised immune system function. Predisposition to infection with adenovirus and pancreatitis has been reported in immune compromised people, Arabian foals with severe combined immunodeficiency and Fell ponies with immunodeficiency syndrome [22-24]. Recently, a mare with insulin-dependent diabetes mellitus, associated with presumed autoimmune polyendocrine syndrome resulting in chronic lymphocytic thyroiditis, adrenalitis and pancreatitis, was reported [25].

In conclusion, pancreatitis in horses can be a primary problem or associated with other disorders, commonly those involving the gastrointestinal tract or liver. Pancreatitis may be underdiagnosed owing to the presence of nonspecific abdominal pain, on which other more common causes must be considered. However, pancreatitis should be considered in horses with unexplained moderate to severe abdominal pain with or without gastric reflux, gastric distension or rupture and peritonitis. Based on the results of this retrospective study, veterinary surgeons should explore the root of the mesentery in all horses undergoing abdominal surgery and consider pancreatitis if white fibrinous plaques are evident on mesenteric and serosal surfaces or mass-like structures and/or fat necrosis are palpated at the root of the mesentery and adjacent tissues. Although rare, horses with a history of recurrent infections or compromised immune system function along with signs of gastrointestinal disease may have concurrent pancreatitis.

Based on the information from this study, it appears that pancreatitis is uncommonly diagnosed in horses and rare in foals. However, this study documents confirmed cases of pancreatitis in which the severity of disease and prognosis resulted in euthanasia or death. These findings raise the possibility that milder cases of pancreatitis may be overlooked. In addition, in some cases the pancreatic lesions could have been coincidental findings and unrelated to the reason for euthanasia. Results from this study emphasise the importance of a thorough macroscopic and histological evaluation of the pancreas in horses with a history of abdominal pain and gastrointestinal and liver disease. Further consideration should be given to the measurement of pancreatic enzymes in serum and abdominal fluid, ultrasonographic evaluation of the pancreas, and exploratory laparoscopy/laparotomy with biopsy in addition to routine physical and blood work evaluation of horses presenting with the clinical signs reported here.

Authors' declaration of interests

No competing interests have been declared.

Sources of funding

None.

Authorship

Yamout: data collection, study execution, data analysis and interpretation, preparation of manuscript; Nieto: study design, data collection, study execution, data analysis/interpretation, preparation of manuscript; Anderson: data collection, study execution, data analysis/interpretation, preparation of manuscript; de Cock: data analysis and interpretation, preparation of manuscript; Vapniarsky: data analysis and interpretation, preparation of manuscript; Aleman: study design, data collection, study execution, data analysis/interpretation, preparation of manuscript.

Ancillary