Introduction
Peritoneopericardial diaphragmatic hernias (PPDHs) in dogs and cats are uncommon congenital malformations, with an estimated prevalence of 0.02% to 0.6% in dogs and cats presented to veterinary specialty hospitals.1,2 Reports of single cases and case series of approximately 114 cats and 199 dogs characterize the clinical findings, diagnostic features, and surgical corrective measures of PPDH.1–14 A PPDH may be discovered after an unexpected death or may be associated with respiratory signs, cardiac tamponade, pleural effusion, entrapped herniated viscera, gallbladder torsion, or distress during parturition.1,2,4–6,9,10,14 Radiographic features include an unusually prominent or dense cardiac silhouette, gas-filled viscera overlying the cardiac silhouette, or an indistinct margin between the cardiac silhouette and diaphragm.1,4,10–12 Prevalence of PPDH is high in Weimaraners and domestic medium- and longhair cats (eg, Himalayans and Persians).1–4,9,13,14 More than 50% of animals with PPDH have ≥ 1 liver lobe and occasionally the gallbladder herniated into the pericardial sac.4,9 Sometimes cystic pericardial lesions are attached to the herniated liver or pedunculated stalk of remnant biliary epithelium.6,8 Notably, histologic features of herniated and nonherniated liver lobes from animals with PPDH are infrequently described in the veterinary literature.5,7,9,14 Congenital central diaphragmatic hernias are also rare in dogs and cats and may be associated with congenital hiatal hernias.15 Also, affected cats may have concurrent hepatobiliary disorders such as cholangitis and cholelithiasis.15
During the last 20 years, one of the authors (SAC) has recognized dogs and cats with PPDH and congenital central diaphragmatic hernia (CCDH) that have concurrent congenital fibropolycystic liver or kidney malformations such as ductal plate malformations (DPMs). Ductal plate malformations represent embryonic abnormalities secondary to dysfunction of the primary cilia that result in defective tubulogenesis and affect the formation of bile ducts. In people, DPMs are classified as ciliopathies, causally linked with impaired function or structure of primary cilia.16–21 Because primary cilia are present in the liver on only cholangiocytes, not hepatocytes, DPMs are more appropriately termed cholangiociliopathies.22 Primary cilia are nonmotile solitary antenna-like organelles projecting from the surface of most eukaryotic cells, and they orchestrate directional orientation, cellular proliferation, involution, or apoptosis that is critical for normal biliary and renal tubulogenesis.16–18 These cilia also transduce signaling molecules essential for cross talk between numerous essential metabolic and developmental pathways.16 Depending on the causal mutation, dysfunctional or malformed primary cilia can lead to several DPM phenotypes.
Proliferative-like DPM is characterized by abundant malformative bile duct profiles embedded in an exuberant extracellular matrix; these dimensionally expand portal tracts, thus forming fibroductal trabeculae.19,23–25 Periportal fibrosis evolves in the absence of obvious duct-centric inflammation. However, DPM is nearly always associated with minor mixed scattered inflammatory portal infiltrates (lymphocytes, macrophages, and fewer neutrophils and plasma cells), thought to reflect enhanced risk for bacterial cholangitis. Hepatocyte cords maintain normal orientation without disruption that is common to necroinflammatory liver injury. However, liver samples with severe portal-to-portal bridging of fibroductal trabeculae may be misclassified as cirrhosis despite the absence of regenerative nodules or disorganized hepatocyte cords.20,26 Occasionally, islands of hepatocytes are entrapped by proliferative fibroductal trabeculae in DPM. A sentinel feature of proliferative-like DPM is the direct intersection of hepatocytes with proliferative-like cholangiocytes at or extending across the limiting plate, in the absence of remodeling, inflammation, or necrosis.20 If necessary, cell proliferation may be confirmed with immunohistochemical staining for Ki-67 to rule out an acquired ductular reaction that might be confused with DPM.20 In DPM, ductal proliferation is not evident and consequently Ki-67 is not expressed.
Caroli DPM phenotype is distinguished by malformative medium-to-large bile ducts with irregularly distended or variably sacculated silhouettes, evaginating diverticular buds, and occasional scattered circumferential satellite bile duct profiles, similar to the embryologic ductal plate.23 As in people, ruling out antecedent mechanical cholestasis is essential before diagnosing Caroli DPM. A subset of animals with proliferative-like or Caroli DPM evolve to a congenital hepatic fibrosis phenotype. In this subset, presinusoidal portal hypertension develops secondary to noncompliant portal fibrosis.20 Animals with congenital hepatic fibrosis may then develop acquired portosystemic shunts (APSS) and abdominal effusion.20,21 However, as in people, clinical features of congenital hepatic fibrosis are variable in animals.26
Choledochal cyst DPM involves diverticulum, sacculation, or an appendix-like evagination forming a cystic structure anywhere along the common bile duct or cystic duct. More often observed in cats than dogs, most choledochal cysts present as cystic malformations at the junction of the common bile duct and duodenal papilla.20,21
Expansive cystadenoma DPMs are also more often encountered in cats.21,25 These lesions are parenchymal-effacing complex microcystic ductal malformations that are embedded in excessive extracellular matrix. In cats, cystadenomas often involve the right liver lobes and frequently abut the gallbladder or cystic duct. Biliary cystadenomas also may affiliate with proliferative-like or Caroli DPM in adjacent or distant hepatic parenchyma.21 In cats and dogs, DPM may coexist with gallbladder agenesis or hypoplasia, agenesis or hypoplasia of ≥ 1 liver lobe, and, rarely, a congenital portosystemic shunt.21,22,27 Many forms of polycystic kidney disease in animals (and in people) also evolve from tubular dysgenesis associated with primary ciliary dysfunction.
Some phenotypes are linked with progressive hepatic fibrosis that culminates in the clinical syndrome of congenital hepatic fibrosis.16–19 Although DPMs in some animals have been linked with mutations influencing the function of primary cilia, dysfunction of primary cilia has not been linked with diaphragmatic malformations in any species, to the authors’ knowledge.28–33 Interestingly, coexistent congenital abnormalities including umbilical hernias, body wall defects, cardiac malformations, and skeletal anomalies are also reported in dogs and cats with PPDH or other congenital diaphragmatic hernias.1,3,4,9,12,14 However, comparatively few reports5,7,9,14,33 include a description of concurrent histologic features consistent with DPM. The presence of a DPM increases the lifelong risk for bacterial cholangitis and cholelithiasis, may permit evolution to a phenotype of congenital hepatic fibrosis, and may coexist with polycystic kidney disease (another DPM phenotype) that may cause progressive kidney dysfunction. Consequently, the presence of DPM should increase vigilance for the detection and management of these associated risks.
The objectives of the retrospective study reported here were to determine the association between PPDH or CCDH and DPM in dogs and cats, categorize the phenotypes of DPM, and determine whether immunohistochemical staining of ductal elements with cytokeratin-19 (CK-19) and assessment of cell proliferation (Ki-67 expression) can assist with distinguishing between DPM and injuries provoking ductular reactions (ie, traumatically herniated, torsed, or ischemic liver lobes).
Materials and Methods
Animals
Cats and dogs with PPDH or CCDH that had liver tissue submitted for histologic examination were identified from case reviews archived by one of the authors (SAC) and through an electronic keyword search of hospital and anatomic pathology databases at the Cornell University College of Veterinary Medicine from 1981 to 2019. Medical records were reviewed to confirm the diagnosis of PPDH or CCDH. During the database search, animals that could serve as a control group were also identified; these animals had various diagnoses of liver disease. Liver samples needed to be available for histologic examination for study inclusion of the animals. Liver samples from healthy animals that were available from unrelated studies were used as a control group for the immunohistochemical staining portion of the study. Animals were determined to be healthy on the basis of the results of physical examination, CBC, serum biochemical analysis, urinalysis, pre- and postprandial serum total bile acid concentrations, and histologic examination of the liver.
Pertinent case material was transcribed from the medical records to document that the study population was representative of that for previously reported1–4,9 animals with PPDH (and other congenital diaphragmatic hernias). Pertinent case material included the following: signalment, clinical signs (fever, lethargy, hyporexia or anorexia, weight loss, vomiting, diarrhea, dyspnea or tachypnea, distended abdomen [abdominal effusion], and neurologic signs consistent with hepatic encephalopathy), results of clinicopathologic diagnostic tests (CBC, serum biochemical analysis, and urinalysis), diagnostic imaging (radiographic, ultrasonographic, or CT) findings, date of definitive diagnosis, date of surgical intervention, surgical reports, survival time (days from the date of definitive diagnosis to the time of death) and age (at the time of death), cause of death (related or unrelated to PPDH or CCDH), occurrence of unexpected death, and results of necropsy examinations. If needed, survival details were verified by means of contacting the primary care veterinarians or pet caretakers. Reports of diagnostic imaging findings, surgery, and necropsy were used to confirm the diagnosis of PPDH or CCDH and the presence or absence of liver herniation.
Histologic examination of liver biopsy samples
Liver samples were fixed in neutral-buffered 10% formalin, embedded in paraffin blocks, sectioned at 5 µm, and stained with H&E and Masson trichrome stains, the latter of which to characterize the presence and extent of fibrillar collagen deposition, at the authors’ diagnostic histology laboratory. Liver samples from clinical cases were initially interpreted by a board-certified anatomic pathologist and an anatomic pathology resident-in-training. Additionally, all samples were independently examined by board-certified anatomic pathologists (ADM, EC) and a board-certified small animal internist with expertise in histologic examination of liver tissue (SAC). Liver samples were inspected for changes consistent with mechanical bile duct occlusion (gallbladder entrapment in the hernia or torsed liver lobes that might lead to mechanical cholestasis), cholangitis or cholangiohepatitis (possible causes of bile duct hyperplasia), impaired transhepatic perfusion (severe sinusoidal congestion, ischemia, and parenchymal extinction), portal venous hypoperfusion (which commonly accompanies DPMs in dogs), and DPM.18–22,28,33–41
Classification of DPM was based on characteristic gross and histologic features that manifested as these distinct phenotypes, as follows: proliferative-like, Caroli, choledochal cyst, and expansive cystadenoma malformation with proliferative-like or Caroli malformations.18,19 Some animals with DPM manifest > 1 phenotype, and many also have evidence of portal venous hypoperfusion with reduced profiles of portal veins (similar to that reported in people).23,42 Ruling out antecedent mechanical cholestasis before diagnosing Caroli DPM was done for each animal in the present study. Because many forms of polycystic kidney disease evolve from tubular dysgenesis associated with primary ciliary dysfunction, animals that had polycystic kidney lesions were also classified as having DPM.
Specific histologic features of each liver sample (ie, evidence of mechanical cholestasis, cholangitis or cholangiohepatitis, impaired transhepatic perfusion, portal venous hypoperfusion, and DPM) were designated as present or absent, and the phenotype of DPM was characterized. Histologic features in herniated and nonherniated liver samples from animals with PPDH or CCDH were compared side by side when possible. Histologic features of samples of herniated liver lobes from animals that had PPDH or CCDH with and without DPM were compared with histologic features of samples of traumatically herniated, torsed, and ischemic liver lobes from dogs and cats of the control group. The histologic features of these samples were compared to determine whether the histologic features in herniated liver lobes could be used to distinguish samples with and without DPM despite changes attributable to liver incarceration or ischemia.
Immunohistochemical staining of liver biopsy samples
Samples from herniated and nonherniated liver lobes from 11 animals with PPDH or CCDH (7 dogs and 4 cats), 9 pathological control animals (4 dogs and 1 cat with traumatic diaphragmatic hernia, 2 dogs and 1 cat with liver lobe torsion, and 1 dog with thromboembolic hepatic ischemia), and 23 healthy control animals (8 dogs and 15 cats) underwent immunohistochemical staining for CK-19 (Leica Microsystems Inc), a biomarker for the detection of biliary epithelium, and for Ki-67 (Agilent Technologies), a biomarker of cellular proliferation.43–45 Positive and negative controls (species-specific tissue arrays constructed from healthy dogs and cats) had been validated for specificity of antibody isotypes and positive and negative antigen signals. The cross-sectioned positive CK-19–stained bile duct profiles per portal tract (BD-PTs) were manually enumerated in 20 operator-designated fields of interest on digitally scanned slides at 200× magnification (Aperio CS2 microscope scanner, Leica Biosystems Inc).
In addition to permitting identification of proliferative cholangiocytes, Ki-67 expression was also used to determine the hepatocyte proliferative index (HPI), which represented the percentage of hepatocytes with positively stained nuclei. This analysis queried the presence or absence of a panlobular proliferative influence in animals with PPDH and CCDH. The HPI was determined by means of enumerating the number of positively stained cells among > 5,000 hepatocytes/biopsy sample in 10 operator-designated fields of interest on digitally scanned slides at 200× magnification. The numbers of hepatocytes, hepatocytes that expressed Ki-67, and sinusoidal nuclei per square millimeter of liver tissue were counted manually and with an automated-assisted method reliant on a proprietary algorithm (Aperio nuclear algorithm, Leica Biosystems Inc). Automated-assisted counting was optimized for nuclear size, shape (roundness, elongation, and compactness), and staining hue.
Statistical analysis
The number of animals that had PPDH or CCDH with and without DPM and the various DPM phenotypes were enumerated for each species. Distributions of age, sex, and survival time (days from the date of definitive diagnosis to death) and age (at the time of death) of animals that had PPDH or CCDH with and without DPM were examined by use of histograms and the Kolmogorov-Smirnov test. Because most data were nonparametric, descriptive statistics were expressed as median (range) and 95% CI, unless otherwise specified. Sex distribution (male vs female) for each species was evaluated for animals with and without DPM by use of 2 × 2 tables through comparison to an evenly distributed population. Differences in ages between animals that had PPDH or CCDH with and without DPM were evaluated with the Wilcoxon rank sum test. The numbers of animals with clinicopathologic features within, above, or below reference intervals were enumerated to assess similarities between these animals and those with PPDH reported in previous publications. Kaplan Meier survival graphs were created to visualize survival curves in dogs and cats with PPDH or CCDH. The Gehan-Wilcoxon and log-rank tests were used to compare survival differences, and the 50th percentile survival time (95% CI) was reported. Meaningful comparisons of median survival time between those animals that underwent surgery and those that did not and comparisons of median survival time for those that underwent surgery and did or did not have DPM could not be completed because of lack of statistical power. The Wilcoxon rank sum test was used to compare the BD-PTs and HPIs between healthy dogs and cats and between animals in the control group and animals with PPDH or CCDH. Statistical analyses were performed with commercial software (Statistix version 10; Analytical Software). Values of P ≤ 0.05 were considered significant.
Results
Animals
A total of 36 animals met the case definition for study enrollment. Of these animals, 31 were diagnosed with PPDH (dogs, n = 14; cats, 17) and 5 were diagnosed with CCDH (dogs, 4; cats, 1). The electronic keyword searches also identified 14 animals that had histologic lesions of the liver but without concurrent PPDH or CCDH and were subsequently assigned to the pathologic control group. Lesions were as follows: traumatic diaphragmatic hernia (dogs, n = 5; cats, 1), liver lobe torsion (dogs, 4; cats, 2), and ischemic hepatopathy (hepatic thromboembolic venous injury; dogs, 2; cats, 0). Twenty-three healthy animals (dogs, n = 8; cats, 15) were used as histologic controls for the immunohistochemical staining portion of the study.
For the 18 dogs with PPDH or CCDH, median (range; 95% CI) age was 1.7 years (0.01 to 11 years; 1.0 to 4.5 years), 13 dogs were male (sexually intact, n = 8; neutered, 5) and 5 were female (sexually intact, 3; spayed, 2) and included the following breeds: Labrador Retriever (2), German Shepherd Dog (2), Weimaraner (2), mixed-breed dog (2), and 1 each of Beagle, English Bulldog, Collie, Keeshond, Miniature Schnauzer, Plott Hound, Rottweiler, Shar Pei, Vizsla, and Yorkshire Terrier. Among these 18 dogs, 13 had hepatic DPM; 3 of those 13 also had renal DPM. One puppy with PPDH but without DPM had gestational exposure to doxycycline and prednisone. Of 13 dogs with DPM, median age was 1.7 years (0.2 to 10 years; 0.3 to 4.4 years), 9 were male (sexually intact, n = 6; neutered, 3) and 4 were female (sexually intact, 2; spayed, 2), and breeds were Labrador Retriever (2), Weimaraner (2), and 1 each of English Bulldog, Collie, German Shepherd Dog, Miniature Schnauzer, Plott Hound, Rottweiler, Shar Pei, Vizsla, and Yorkshire Terrier. Median age and sex of dogs with PPDH or CCDH did not significantly differ between those with or without DPM, and the characteristics of the study population were similar to those of dogs with PPDH described in a previous report.4
For the 18 cats with PPDH or CCDH, median (range; 95% CI) age was 4.5 years (0.2 to 15 years; 3.4 to 8.9 years), 10 were male (sexually intact, n = 2; neutered, 8) and 8 were female (sexually intact, 3; spayed, 5), and types were domestic shorthair (8), domestic longhair (4), Persian (3), and 1 each of Maine Coon, Ragdoll, and Siamese. Among these 18 cats, 13 had hepatic DPM (1 with renal DPM) and 2 had only renal DPM. Of the 15 cats with DPM, median age was 4.0 years (0.2 to 15 years; 2.4 to 9.1 years), 9 were male (sexually intact, n = 1; neutered, 8) and 6 were female (sexually intact, 2; spayed, 4), and types were domestic shorthair (8), domestic longhair (3), Persian (3), and Siamese (1). Median age and sex of cats with PPDH or CCDH did not significantly differ between those with or without DPM, and the characteristics of the study population were similar to those of cats with PPDH described in a previous report.4
Clinical features for the dogs with PPDH or CCDH included lethargy (15/18), vomiting or inappetence (12/18), tachypnea or dyspnea (8/18), abdominal effusion (7/18), and previously documented increased liver enzyme activities (5/14). Dogs with DPM had clinical signs similar to those of dogs without DPM. Peritoneopericardial diaphragmatic hernia or CCDH was confirmed in 5 dogs via thoracic radiography, and diaphragmatic hernia was confirmed in 10 dogs and cystic kidney lesions in 2 dogs via abdominal ultrasonography completed by board-certified veterinary radiologists. Eight dogs did not undergo abdominal ultrasonographic examination because this diagnostic modality was not routinely available at the time of presentation or because dogs presented at the point of death. Some of these had radiographic evidence of a diaphragmatic hernia. Some dogs unexpectedly died during or immediately after routine anesthesia for surgical or dental procedures without an opportunity for diagnostic imaging.
Clinical features for the cats with PPDH or CCDH included tachypnea or dyspnea (12/18), lethargy (11/18), vomiting or inappetence (8/18), previously documented increased liver enzyme activities (2/14), and abdominal effusion (1/18). Cats with DPM had clinical signs similar to those of cats without DPM. No cats had signs consistent with hepatic encephalopathy. Peritoneopericardial diaphragmatic hernia or CCDH was confirmed in 6 cats via thoracic radiography, and diaphragmatic hernia was suggested in 7 cats and cystic kidney lesions were confirmed in 2 cats via abdominal ultrasonography. Eleven cats did not undergo abdominal ultrasonographic examination because this diagnostic modality was not routinely available at the time of presentation or because cats presented at the point of death. Some of these had radiographic evidence of a diaphragmatic hernia. Some cats also had unexpectedly died during or immediately after routine anesthesia for surgical or dental procedures without an opportunity for diagnostic imaging. Of the 10 cats that were necropsied, cystic renal tubular malformations were confirmed in 3 cats and pancreatic ductal malformations in 2 cats.
Clinicopathologic details were available for 13 dogs, including 10 of the 13 with DPM. Hematologic and serum biochemistry results did not distinguish dogs with DPM from those without DPM and were consistent with the variable results previously reported4 for dogs with PPDH. Pertinent results included Hct within the reference interval (41% to 58%) for all dogs, except 2 that had nonregenerative anemia (Hct, 22% and 31%). The values for mean corpuscular volume were within the reference interval (64 to 76 fL), except for breed-related microcytosis (61 fL) in a Shar Pei. White blood cell counts were within the reference interval (5.7 × 103 to 14.2 × 103 cells/µL) for all dogs, except 4 that had neutrophilic leukocytosis; none of these dogs had a left shift. Differential WBC counts were variable, with monocytosis noted for 3 dogs (range, 1.7 × 103 to 3.3 × 103 cells/µL; reference interval, 0.1 × 103 to 1.3 × 103 cells/µL). Platelet counts were within the reference interval (186 × 103 to 585 × 103 platelets/µL), except for 1 dog that had mild thrombocytopenia (144 × 103 platelets/µL). Mild to moderate hypoalbuminemia was noted for 7 dogs (range, 2.2 to 2.7 g/dL; reference interval, 2.9 to 4.2 g/dL). One dog (Shar Pei) had breed-related renal amyloidosis with pathological proteinuria and was euthanized shortly after surgery. Among the other hypoalbuminemic dogs, 1 died from congenital hepatic fibrosis DPM and APSS, 1 died from acute kidney injury and rhabdomyolysis that evolved from status epilepticus and malignant hyperthermia, 1 died during general anesthesia for routine neutering, and 1 had pericardial tamponade. Hypoalbuminemia resolved in the 3 dogs that survived following diaphragmatic herniorrhaphy, including the dog with pericardial tamponade; thus, transient hypoalbuminemia was likely attributed to a negative acute phase effect or distributional factors in these dogs. One dog had mild hypoglobulinemia (1.7 g/dL; reference interval, 1.9 to 3.9 g/dL), 1 dog was hyperglobulinemic (4.8 g/dL), and 1 dog had azotemia (BUN, 157 mg/dL [reference interval, 9 to 23 mg/dL]; creatinine, 5.3 mg/dL [reference interval, 0.6 to 1.4 mg/dL]) that was unrelated to DPM (caused by renal amyloidosis). Two other dogs had prerenal azotemia (characterized by BUN of 27 and 39 mg/dL and creatinine concentrations within the reference interval), which was resolved with fluid therapy. Five dogs were hypocholesterolemic (reference interval, 150 to 332 mg/dL). Of these dogs, 4 had pericardial tamponade caused by PPDH, and serum cholesterol concentration normalized after surgical intervention. One hypocholesterolemic dog with congenital hepatic fibrosis and APSS succumbed to these conditions postoperatively. One dog with congenital hepatic fibrosis and APSS was hypercholesterolemic (585 mg/dL), but cholesterol normalized after long-term treatment with levothyroxine. Some dogs with DPM had liver enzyme activities ≥ 2-fold the upper limit of reference intervals (alanine aminotransferase activity, 3.0- to 16.7-fold, and alkaline phosphatase, 4.8- to 13.0-fold [n = 3]; aspartate aminotransferase activity, 2.4-fold [1]; γ-glutamyltransferase activity, 2.8-fold [1]). Mild hyperbilirubinemia was noted for 1 dog (0.7 mg/dL; reference interval, 0 to 0.6 mg/dL). Historical illnesses or clinicopathologic or histologic features consistent with extrahepatic bile duct obstruction (EHBDO) were not evident among dogs with DPM. Urine from 5 dogs had specific gravities that ranged from 1.011 to 1.055 and pH that ranged from 6.0 to 7.5. Glucosuria, ketonuria, pyuria, bacteriuria, and cylindruria were not noted, but 2 dogs had ammonium biurate crystalluria.
Clinicopathologic details were available for 12 cats, 8 of which had DPM. Hematologic and serum biochemistry results did not distinguish between cats with and without DPM and were consistent with the variable results previously reported3,4 for cats with PPDH. Hematologic findings included Hct within the reference interval (31% to 48%) for all cats, except for 3 that had mild nonregenerative anemia. The values of mean corpuscular volume were within the reference interval. The WBC counts were within the reference interval (5.1 × 103 to 16.2 × 103 WBCs/µL) for all cats, except for 2 that had neutrophilic leukocytosis and 1 that had leukopenia. One of the 2 cats with leukocytosis had a left shift with 0.5 × 103 band neutrophils/µL (reference interval, ≤ 0.3 × 103 band neutrophils/µL) and toxic changes to the mature neutrophils. The differential WBC count was variable among cats, with lymphocytosis in 4 young cats, lymphopenia in 2 cats, monocytosis in 3 cats, and eosinopenia in 2 cats. Platelet counts were within the reference interval (195 × 103 to 624 × 103 platelets/µL), except for 2 cats with mild thrombocytosis. No cat had serum biochemistry results consistent with synthetic liver failure (ie, marked hypoalbuminemia, hypocholesterolemia, and subnormal BUN concentration).
Two cats were hyperbilirubinemic; one of the cats that had a 2-fold increase in serum total bilirubin concentration (relative to the upper reference interval) had Caroli DPM associated with neutrophilic hepatitis and feline infectious peritonitis, whereas the other cat had a 27-fold increase in total bilirubin concentration that was associated with combined Caroli and choledochal cyst DPM, cholelithiasis, and suppurative bacterial cholangiohepatitis. This cat had clinical signs and clinicopathologic changes consistent with intermittent bile duct occlusion caused by the choledochal cyst and cholelithiasis. In other cats, bile duct obstruction was not detected through ultrasonographic examination, surgical observations, or necropsy. Considering all cats with DPM, liver enzyme activities were ≥ 2-fold the upper reference interval in the 2 hyperbilirubinemic cats; they had 2.1- and 5.0-fold increases in alanine aminotransferase activity. The cat with obstructive cholangiopathy also had an 8.6-fold increase in aspartate aminotransferase activity, a 5.3-fold increase in alkaline phosphatase activity, and a 4.0-fold increase in γ-glutamyltransferase activity. Urinalyses for 6 cats indicated specific gravities that ranged from 1.008 to 1.065, pH between 6.0 and 7.0, and no glucosuria, ketonuria, pyuria, bacteriuria, or cylindruria.
Histologic examination of liver biopsy samples
Thirty-six animals with PPDH or CCDH had liver tissue samples that were collected as biopsies antemortem (dogs, n = 10; cats, 8) and at necropsy (dogs, 8; cats, 10). For the 13 dogs with DPM, phenotypes included isolated proliferative-like DPM (n = 8); combined proliferative-like and Caroli DPM (2); combined proliferative-like, Caroli, and renal DPM (2); and combined Caroli and renal DPM (1). Among dogs with isolated proliferative-like DPM, 3 had congenital hepatic fibrosis and APSS; 2 of these dogs manifested signs of hepatic encephalopathy (ie, cognitive impairment, ataxia, and somnolence), 1 had ammonium biurate crystalluria, and 1 had hyperammonemia. One dog with DPM had a single congenital extrahepatic portosystemic shunt. One of 11 dogs that underwent a necropsy had confirmed Caroli DPM and renal and pancreatic cystic tubular malformations. For the 15 cats with DPM, phenotypes included proliferative-like DPM (n = 5), Caroli DPM (3), combined proliferative-like and Caroli DPM (2), proliferative-like and Caroli DPM with a large cystadenoma (1), Caroli DPM and a choledochal cyst complicated by cholelithiasis (1), Caroli and proliferative-like DPM with polycystic kidneys (1), and polycystic kidneys without hepatic DPM (2). Two cats had the congenital hepatic fibrosis phenotype. All cats with polycystic kidneys were Persians.
Classic histologic features of proliferative-like DPM were noted in 12 of the 13 dogs with DPM in the nonherniated liver lobes (Figure 1). In 5 of these 12 dogs with proliferative-like DPM, small islands of hepatic parenchyma were entrapped by proliferative-like bile duct profiles (Figure 2). The sentinel interfacing of hepatocytes and proliferative-like cholangiocytes in quiescent parenchyma was confirmed in each dog. Histologic features of the nonherniated liver lobes in dogs with PPDH or CCDH and DPM included mild portal tract inflammation (5/13), a predominance of lymphoplasmacytic infiltrates (4/13), and additional neutrophilic infiltrates (1/13). Three of the 5 dogs with Caroli DPM had occasional embryonic ductal plate configurations with satellite bile duct profiles circumferentially oriented around an irregularly sacculated interseptal duct. Features consistent with portal venous hypoperfusion were verified in 10 of 13 dogs with DPM; in 3 of these dogs, portal vein silhouettes were unidentifiable. Necrotic hepatocytes were rarely identified in any dog.
Photomicrographs of liver samples from a dog with peritoneopericardial diaphragmatic hernia (PPDH) and proliferative-like ductal plate malformation (DPM) stained with H&E (A) and Masson trichrome (B) stains. In both panels, the liver sample on the left is from a herniated lobe and the liver sample on the right is from a nonherniated lobe. Note the damaged parenchyma (lacy, collapsed appearance) in the herniated lobe and fibroductal bridging trabeculae (pale blue staining cholangiocyte nuclei [A] and blue staining fibrillar collagen [B]) in the nonherniated lobe. In panel B, also note the increased fibrillar collagen (stained blue) in the sample of the herniated lobe. Bar = 4 mm.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Photomicrograph of liver sample stained with H&E from a nonherniated lobe from a dog with PPDH and proliferative-like DPM. Malformative bile duct profiles are embedded in a pale pink staining extracellular matrix. Illustrated are multifocal mixed inflammatory infiltrates within the fibroductal trabecula that are not targeting the bile ducts and a mixed WBC population fluxing across hepatic sinusoids. Also evident are numerous small, binucleated hepatocytes, consistent with portal venous hypoperfusion, and small islands of hepatocytes (arrows) entrapped within the fibroductal trabecula. Bar = 200 µm.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Classic histologic features of proliferative-like DPM were confirmed in 9 of the 15 cats with DPM in the nonherniated liver lobes, 4 of these combined with the Caroli DPM phenotype. The sentinel interfacing of hepatocytes and proliferative-like cholangiocytes in quiescent parenchyma was confirmed in each cat with a proliferative-like DPM. Small islands of hepatic parenchyma entrapped by proliferative fibroductal partitions were verified in 2 cats. Histologic features in the nonherniated liver lobes from 7 cats with PPDH or CCDH and DPM included mild (n = 5) or moderate (2) mixed lymphoplasmacytic and minor neutrophilic portal infiltrates. The liver sample of the cat with suppurative bacterial cholangitis associated with cholelithiasis and a choledochal cyst that caused mechanical cholestasis had prominent neutrophilic infiltrates and evidence of bile duct obstruction. Sinusoidal flux of WBCs in 5 cats was similar in composition to those at the portal regions. Thick-walled muscular serpiginous arterials were noted in the portal tracts of 11 cats and were inconsistently associated with other stereotypic features of portal venous hypoperfusion, including lobular atrophy and unidentifiable portal vein silhouettes. Stereotypic features of portal venous hypoperfusion were less evident in cats with DPM, compared with dogs with DPM. Eight cats with Caroli DPM had occasional embryonic ductal plate configurations with satellite bile duct profiles circumferentially oriented around an irregularly distended or sacculated interseptal duct (Figure 3). Necrotic hepatocytes were rarely identified in any cat.
Photomicrographs of serial liver sections (A to D) from a cat with PPDH and Caroli DPM and a liver sample from a cat with PPDH and combined proliferative-like and Caroli DPM (E). A—Irregularly distended medium-sized bile duct surrounded by numerous malformative satellite bile duct profiles (asterisk), several thick-walled muscular arteries (a), and distended lymphatic vessels (L). Stain = H&E stain. B—Malformative bile duct profiles (the larger bile duct [asterisk] with irregular budding protrusions and surrounding scattered satellite bile duct profiles) and bile duct structures that are embedded in excessive extracellular matrix (fibrillar collagen staining blue) expand this portal region. Also note the distended lymphatic vessels (white spaces), thick-walled arteries (a), and margin of the portal vein (pv). Stain = Masson trichrome stain. C— Immunohistochemical staining for cytokeratin-19 (CK-19) with red signal detection illustrates malformative bile duct profiles. D—Immunohistochemical staining for Ki-67 with red signal detection indicates the absence of cholangiocyte proliferation (no Ki-67 staining). The nuclei of rare sinusoidal cells are immunoreactive (arrows), a normal phenomenon. E—Immunohistochemical stain for CK-19 with red signal detection illustrates malformative bile duct profiles. Note embryologic DPM patterning (continuum of small proliferative-like ducts circumferentially marginating a more central dominant bile duct; ductal elements red stain positive). These cholangiocytes were Ki-67 negative, ruling out cholangiocyte proliferation and a ductular reaction. The more dorsal duct displays an irregular DPM profile. Bar = 300 µm.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
In all animals with hepatic DPM, cellular infiltrates were adjacent to cholangiocytes but did not target biliary epithelium (Figures 2 and 4). Histologic lesions that involved other acinar components (eg, central veins and hepatic cords) were not seen except for occasional sequestration of hepatocytes by fibroductal trabeculae. No dogs and cats with DPM had regenerative liver nodules; isolation of hepatocytes that formed parenchymal pseudonodules was most evident in animals with congenital hepatic fibrosis (Figure 4).
Photomicrographs of liver samples stained with Masson trichrome stain from a dog with PPDH and proliferative-like DPM with congenital hepatic fibrosis phenotype that also had presinusoidal portal hypertension, acquired portosystemic shunts, and abdominal effusion. A—The liver sample on the left is from a herniated lobe, and the liver sample on the right is from a nonherniated lobe. Fibrillar collagen is stained blue. Bar = 7 mm. B—Higher magnification of herniated liver shown in panel A. Note the diffuse fibrosis (blue), absence of hepatocytes, extreme numbers of cross-sectional malformative bile duct profiles, and mixed scattered inflammatory infiltrates not targeting the ductal elements. Bar = 200 µm. C—Higher magnification of nonherniated liver shown in panel A. Note the numerous bile duct profiles that populate the bridging fibroductal trabeculae. Bar = 200 µm. D—Higher magnification of nonherniated liver shown in panel A illustrating thick fibroductal trabeculae isolating hepatic parenchyma that retains normal hepatocyte cord structure. Bar = 600 µm.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Immunohistochemical staining of liver biopsy samples
The numbers (median; range; 95% CI) of BD-PTs that were indicated by immunoreactivity to CK-19 in dogs that had PPDH or CCDH with or without DPM (5.0; 1.4 to 100.8; 1.7 to 28.5) and in dogs that had PPDH or CCDH with DPM (10.5; 2.4 to 100.8; 1.9 to 40.3) were significantly (both P < 0.001) higher than in healthy dogs (1.4; 1.2 to 1.6; 1.3 to 1.6). The lowest number of BD-PTs was seen in a dog with Caroli DPM and polycystic kidneys. The number of BD-PTs in 5 dogs with PPDH or CCDH without DPM (1.7; 1.4 to 4.1; 1.0 to 3.6) was not significantly different from that in healthy dogs but was significantly (P = 0.004) lower than in dogs with DPM. The numbers of BD-PTs in cats that had PPDH or CCDH with or without DPM (6.6; 1.9 to 11.0; 4.4 to 7.8) and in cats that had PPDH or CCDH with DPM (7.1; 1.9 to 11.0; 4.5 to 8.9) were significantly (P < 0.001 and P = 0.003, respectively) higher than in healthy cats (2.3; 1.7 to 2.6; 2.0 to 2.3). Among cats with DPM, the lowest number of BD-PTs was seen in those with only Caroli DPM or polycystic kidneys. The number of BD-PTs in 4 cats that had PPDH or CCDH without DPM (2.2; 2.0 to 2.8; 1.7 to 2.9) was not significantly different from that in healthy cats but was significantly (P = 0.044) lower than in cats with DPM. The number of BD-PTs in healthy cats was significantly (P < 0.001) higher than in healthy dogs.
Hepatocyte proliferative index (median; range; 95% CI) did not significantly differ between the tissue samples of nonherniated liver lobes from dogs that had PPDH or CCDH with DPM (1.1%; 0.3% to 2.9%; 0.5% to 2.2%) and the liver lobes from healthy dogs (0.52%; 0.2% to 0.9%; 0.3% to 0.8%). Three dogs that had PPDH or CCDH without DPM had a median HPI of 2.0% (1.1% to 2.9%; 0.4% to 3.6%); this group was too small for relevant statistical comparisons. No mitotically active hepatocytes were identified on histologic examination of the liver tissue samples from healthy dogs and dogs that had PPDH or CCDH with or without DPM.
Hepatocyte proliferative index (median; range; 95% CI) did not significantly differ between the tissue samples of nonherniated liver lobes from cats that had PPDH or CCDH with DPM (0.5%; 0.3% to 2.0%; 0.4% to 2.0%) and the liver lobes from healthy cats (0.5%; 0.4% to 1.1%; 0.4% to 0.9%). The HPI in 2 cats that had PPDH without DPM was 0.4% and 0.7%; this group was too small for relevant statistical comparisons. No mitotically active hepatocytes were identified in either group.
Immunoreactivity to Ki-67 was not noted in the cholangiocytes of healthy dogs or dogs that had PPDH or CCDH with or without DPM. No lobular tropism of Ki-67 immunoreactive single or clustered cells was observed adjacent to portal tracts (Figure 5); thus, the absence of a ductular reaction was verified. However, the cholangiocytes in the herniated or torsed liver lobes of several dogs in the pathological control group that were investigated for Ki-67 immunoreactivity were immunoreactive for Ki-67, signifying a ductular reaction (Figure 6). Numerous Ki-67 immunoreactive inflammatory cells were also usually observed in the incarcerated or ischemic liver tissue.
Photomicrographs of liver samples from the same dog in Figure 4. A—Immunohistochemical staining for CK-19 with red signal detection of liver samples from a herniated lobe (left image) and from a nonherniated lobe (right image). Bar = 6 mm. B—Note the myriads of CK-19 immunoreactive ductal elements in the sample of the herniated liver in panel A. Bar = 200 µm. C—Note the numerous CK-19 immunoreactive ductal elements that populate the fibroductal trabeculae in the sample of the nonherniated liver in panel A. Bar = 300 µm. D—Immunohistochemical staining for Ki-67 with red signal detection of a liver sample from a herniated lobe illustrates only immunoreactive nuclei of scattered inflammatory infiltrates and not cholangiocytes. Bar = 200 µm. E—Immunohistochemical staining for Ki-67 of a liver sample from a nonherniated lobe confirms the absence of cholangiocyte proliferation (ie, absence of bile duct staining). Bar = 300 µm.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Immunoreactivity to Ki-67 was not noted in the cholangiocytes of healthy cats or cats that had PPDH or CCDH with or without DPM. Lobular tropism of stained or clustered cells adjacent to portal tracts was also not seen; thus, the absence of a ductular reaction was verified.
The number of Ki-67 immunoreactive sinusoidal cells was variable among dogs and cats with PPDH or CCDH, reflecting individual differences in the flux of WBCs across hepatic sinusoids and the systemic response to incarcerated tissue (Figure 2). The number (median; range; 95% CI) of Ki-67 immunoreactive sinusoidal cells per square millimeter of liver tissue in healthy dogs (15.8; 5.7 to 46.1; 9.1 to 33.1) was significantly lower than in dogs with PPDH or CCDH (76.4; 45.0 to 310; 33.0 to 176.9; P < 0.001) and pathological controls (40.7; 24.9 to 205.7; 2.2 to 155.9; P = 0.03). The number of Ki-67 immunoreactive sinusoidal cells per square millimeter of liver tissue in healthy cats (15.3; 1.7 to 96.6; 9.8 to 45.7) was also significantly (P = 0.03) lower than in cats with PPDH or CCDH (119.0; 8.0 to 453.0; 20.0 to 485.0). Differences in HPI and the number of Ki-67 immunoreactive sinusoidal cells per square millimeter of liver tissue were not significant between healthy dogs and cats.
Histologic comparison of herniated and nonherniated liver lobes in animals that had PPDH or CCDH and DPM confirmed the preservation of malformative proliferative-like bile duct profiles in herniated tissue without Ki-67 immunoreactivity (Figures 3 and 5). Because herniated liver lobes may have undergone changes secondary to incarceration within the pericardial sac (ie, lobe torsion or infarction), herniated lobes were compared with those from pathological controls (ie, lobes damaged by traumatic herniation, torsion, or infarction). In contrast to DPM-affected lobes, lobes damaged by traumatic herniation, torsion, or ischemic injury had panlobular parenchymal extinction with nearly a global absence of bile duct profiles and preservation of thick tortuous arterials (likely because of vascular accommodation to sustain tissue perfusion although they occasionally had intraluminal thrombi) or had small, scattered foci of proliferative cholangiocytes with immunoreactivity to Ki-67 (Figure 6). In animals with peracute presentation for liver lobe torsion, the dominant histologic feature was centrilobular or panlobular congestion with early centrilobular hepatocyte vacuolation and degeneration.
Photomicrographs of liver samples from a dog with monolobular ischemic liver injury (thrombosis of hepatic vein) that shows a sample of unaffected liver (A; bar = 300 µm) and a sample of affected (ischemic) liver (B to D; bars = 200 µm). A—Unaffected liver (PT = portal tract, CV = central vein) with normal lobular architecture. Stain = H&E stain. B—Liver with ischemic injury that has disrupted lobular architecture with extensive fibroplasia, reactive bile duct elements, numerous scattered inflammatory infiltrates, and congested vasculature. Stain = H&E stain. C and D—Serial sections of liver with ischemic injury documenting ductal elements with CK-19 (red signal detection [C]) and proliferative cholangiocytes with Ki-67 (red signal detection [D]). Features define a ductular reaction not seen in DPM. Also note the immunoreactive nuclei of multifocal inflammatory infiltrates that are Ki-67 immunoreactive, a normal feature.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Survival analyses
Of the 18 dogs with PPDH or CCDH, 7 unexpectedly died and 2 were immediately euthanized because of profound clinical illness. Four of the 7 dogs that died unexpectedly did so at home after acute onset of dyspnea and collapse, 2 died during or immediately after uncomplicated ovariohysterectomy, and 1 died from acute kidney injury and status epilepticus that subsequently caused malignant hyperthermia and rhabdomyolysis. Of 3 dogs with a congenital hepatic fibrosis DPM phenotype, age at diagnosis was 1, 2, and 3 years with survival times of 548, 20, and 850 days, respectively. The DPM phenotype overtly contributed to the death of 3 dogs that had undergone successful PPDH or CCDH surgery. Two of these dogs had congenital hepatic fibrosis, APSS, ascites, signs of hepatic encephalopathy, and severe loss of body condition; one dog died 20 days after surgery and the other was euthanized 2.3 years later. The third dog succumbed to polycystic kidney disease 2.6 years after successful PPDH surgery. Two dogs were eventually euthanized because of chronic illness unrelated to DPM; one was euthanized because of pulmonary sarcoma 4 years after PPDH surgery and the other for Shar Pei breed–related renal amyloidosis 2 months after PPDH surgery. Four dogs were still alive 0.2 to 1.5 years after their diagnoses. Kaplan Meier all-cause mortality analyses for dogs with PPDH or CCDH indicated a 50th percentile survival time of 7.0 days (95% CI, 1 to 1,460 days) and age at death of 2.0 years (0.8 to 9.4 years).
Survival time significantly (Gehan-Wilcoxon test, P = 0.004; log-rank test, P = 0.03) differed between dogs with (n = 13) and without (5) DPM; however, 10 of the 13 dogs with DPM underwent surgical intervention, whereas none of the dogs without DPM had surgery (often because of sudden death). Median (range; 95% CI) survival time for the dogs with DPM that underwent surgical intervention was 730 days (7.0 to 2,373 days; 315 to 1,558 days) with median survival age of 5.4 years (0.2 to 13.6 years; 3.6 to 9.8 years). Median survival time for all 13 dogs with DPM was 445 days (1 to 2,373 days; 189 to 1,216 days), and median survival age was 4.7 years (0.2 to 13.6 years; 2.3 to 7.9 years). Median survival time for the 5 dogs without DPM (and without surgical intervention) was 1 day (1 to 1,460 days; 1 to 1,103 days), and median survival age was 0.7 years (0.01 to 7.4 years; 0.01 to 5.9 years). The uneven distribution of dogs with DPM that underwent surgical intervention and the small number of dogs with DPM that did not undergo surgical intervention (n = 3) precluded meaningful statistical comparisons.
Of the 18 cats with PPDH or CCDH, 11 died before or simultaneous with the diagnosis of congenital diaphragmatic hernia. Among these 11 cats, 7 unexpectedly died; 1 pregnant cat that was at full term suddenly became severely dyspneic, 2 cats were discovered dead at home, 3 cats died during or immediately after elective surgical or dental procedures, and 1 cat died at home 2 days after PPDH corrective surgery. The remaining 4 cats were euthanized; 1 cat each for failure to thrive, feline infectious peritonitis, polycystic kidney disease and subsequent chronic kidney failure, or pleural effusion. Five of the cats with longer survival times had been euthanized because of recurrent bacterial cholangitis associated with Caroli DPM (n = 2), cardiac arrhythmia that required a pacemaker (1), metastatic renal carcinoma (1), and a large mineralized lung mass (1). Two cats with a congenital hepatic fibrosis DPM phenotype were diagnosed at 5.2 and 15.0 years of age and survived for 180 days and 1 day, respectively. One cat with PPDH and DPM survived 2.5 years and 1 cat with PPDH but without DPM survived 3.5 years after their diagnoses. Kaplan Meier all-cause mortality analyses for cats with PPDH or CCDH indicated a 50th percentile survival time of 1 day (95% CI, 1 to 1,660 days) and age at death of 5.3 years (1.0 to 13.8 years).
Of the 6 cats that underwent surgical correction of PPDH or CCDH, median (range; 95% CI) survival time for 5 with DPM was 660 days (1 to 2,008 days; 1 to 1,739 days) and median survival age was 7.1 years (3.0 to 13.8 years; 2.6 to 13.2 years). Survival time for the 1 cat without DPM was 1,278 days, and its age at the time of death was 9.1 years. Of the 12 cats without surgical intervention for PPDH or CCDH, 10 with DPM had a median survival time of 1 day (1 to 4,380 days; 1 to 1,429 days) and a median age at the time of death of 1.4 years (0.2 to 16.2 years; 1.2 to 6.2 years). The remaining 2 cats survived 1 and 45 days after PPDH or CCDH diagnosis with age at the time of death of 1.9 and 12.0 years, respectively. Survival times and ages at the time of death between cats with and without DPM and between cats with and without surgical intervention were not compared because of the small sample sizes. Survival times and ages did not significantly differ between dogs and cats with PPDH or CCDH (Figure 7).
Kaplan Meier survival curves for 18 dogs (dashed line) and 18 cats (solid line) with PPDH or congenital central diaphragmatic hernia, for which 13 dogs and 15 cats had concurrent DPMs. Survival time was not significantly different between species.
Citation: Journal of the American Veterinary Medical Association 259, 9; 10.2460/javma.259.9.1009
Discussion
Although PPDH in dogs and cats has been extensively reviewed in the veterinary literature, findings of the present study verified a previously unrecognized association between congenital diaphragmatic hernias and DPM. Greater than 70% of dogs and cats with PPDH or CCDH also had hepatic DPM, renal DPM, or both. Prior studies1–4,9,11 of PPDH likely did not include the possibility of concurrent DPM because of the lack of discerning physical and clinicopathologic features of DPM, the lack of histologic evaluation of nonherniated liver tissue, and pathologists’ unfamiliarity with the histologic features of DPM. Indeed, inspection of clinicopathologic details of dogs and cats with PPDH or CCDH in the present study did not discern differences between animals with and without DPM, with the exception of animals with a congenital hepatic fibrosis phenotype that had APSS. The findings for 3 dogs and 3 cats with renal DPM in the present study were not unprecedented because > 60% of people with Caroli DPM also have kidney involvement,46,47 and concurrent liver and kidney polycystic malformations have previously been reported29,48–50 in dogs and cats (although not specifically designated as DPM). The frequency of DPM was possibly underestimated in the present study because kidney tissue was not routinely examined.
Ductal plate malformation represents an end point disruption of tubulogenesis and is used as an umbrella term for embryologic dysgenesis of the biliary ducts, renal tubules, and, rarely, pancreatic ducts.51 Involvement of pancreatic ductal elements was recognized in the present study and has previously been described50 in cats with congenital biliary malformations. The commonality among the numerous genetic mutations associated with DPM in people is disruption in the function or structure of the primary cilia that are critically involved in embryogenesis of biliary ducts and renal tubules.22,24,52–54 During embryogenesis, ductal plate elements not progressing to mature bile ducts ordinarily undergo involution or become periportal hepatocytes.55,56 However, defective orchestration of this process can lead to retention of embryonic ductal structures or proliferative responses and excessive extracellular matrix that geometrically expands portal tracts.20,54
The wide spectrum of genetic causes and epigenetic modifiers of primary cilia function reconciles with the clinical diversity among DPM syndromes, with some that are clinically silent and others that progress to organ dysfunction.54,56–59 Insufficient involution of embryonic ductal elements at the level of segmental bile ducts (first branch of the hepatic duct) can evolve to large sacculated ducts typical of Caroli DPM.25,42 This phenotype is aggravated by an associated degradation of structural components (eg, laminin and type IV collagen) that are essential for tensile strength of large ductal structures.48,60 Insufficient involution of ductal elements at the level of interlobular or intralobular bile ducts leads to the development of large irregular sacculated ducts or myriads of tiny ductal structures amalgamated in an abundant extracellular matrix (fibrosis). Both proliferative-like and Caroli DPM phenotypes may evolve to congenital hepatic fibrosis because of an apparent lifelong accrual of extracellular matrix.20,21,25,42 Isolated failure of ductal plate resorption at the level of interlobular bile ducts can evolve to microscopic von Meyenburg complexes, which are usually found at the liver lobe margins.61 These malformations were not considered for classification of DPM in the present study.
Ductal plate malformations complicated by congenital hepatic fibrosis are clinically the most severe and were characterized in 3 dogs and 2 cats in the present study. Ages at the time of death for the affected dogs were 2.1, 2.5, and 5.3 years, with congenital hepatic fibrosis as the cause of death for 2 of the 3 dogs, and for the affected cats were 5.2 and 15.0 years, with congenital hepatic fibrosis as the cause of death for 1 cat. The pathogenesis of fibrosis in congenital hepatic fibrosis differs from that of fibrosis in necroinflammatory liver disease that is driven by cytokine-provoked transformation of perisinusoidal hepatic stellate cells (Ito cells) into myofibroblasts.62 In chronic hepatitis, collagen deposition in the space of Disse (the space between the sinusoidal endothelium and hepatocyte surface) predominates. In DPM, fibrosis is limited to the portal regions, and when it is severe, it extends within bridging fibroductal trabeculae that span between portal triads. The pathogenesis of fibrosis in DPM involves reactive but not proliferative cholangiocytes, recruited macrophages, and the process of biliary epithelial-to-mesenchymal transitioning (to myofibroblasts).62–69 Bile duct dysgenesis in DPM seemingly imparts a tendency for cholangiocyte activation and production of chemokines and cytokines.63 Accumulation of fibrogenic signaling molecules (collagen stimulatory growth factor, transforming growth factor-β2 [with upregulated transforming growth factor-β2 receptors], and tumor necrosis factor-α) in the poorly vascularized extracellular matrix typical of DPM is thought to amplify and sustain collagen accumulation.63–69 Different pathogenic factors drive fibrogenesis in DPM, compared with necroinflammatory liver injury, which clarifies why anti-inflammatory immunomodulation does not curtail collagen accrual in DPM. Indeed, the ultimate treatment for people with congenital hepatic fibrosis is liver transplantation.
Many factors that impact the liver can provoke hepatocyte and cholangiocyte proliferation, with the latter, a ductular reaction, a more common response.44,45,68,69 Activated cholangiocytes are differentiated from hepatic progenitor cells or transdifferentiated from hepatocytes. Typically, ductular reactions assume an unorganized clustering or linear configuration, lacking defined ductular structure (ie, no visible lumen) and extending into the adjacent parenchyma and along the margins of portal regions.19,70,71 These responses may critically rejuvenate hepatic mass when the proliferative ability of mature hepatocytes is compromised.19,70–72 Definitive diagnosis of proliferative-like DPM requires distinction from a ductular reaction.25,68 Activated cholangiocytes in ductular reactions secrete cytokines, chemokines, and growth and angiogenic factors; express a rich repertoire of receptors; and recruit inflammatory cells, myofibroblasts, and endothelial cells.68,71 Thus, the histologic appearance of a ductular reaction is in stark contrast with the comparatively bland histologic appearance of a proliferative-like DPM.19,73 A sentinel feature of proliferative-like DPM is the direct juxtaposition of differentiated biliary epithelium against hepatocytes at or beyond the limiting plate of the portal tract. Although this interface can sometimes be confused with a ductular reaction, inflammation or remodeling is absent in DPM.20 In cases with equivocal histologic features, proliferative cholangiocytes can be phenotyped by use of the cell proliferation marker Ki-67; cholangiocytes in DPM are not immunoreactive to Ki-67.44,74 This strategy was used to prove the persistence of proliferative-like DPM in the herniated liver lobes from animals with DPM, whereas the liver lobes from pathological controls (unaffected by DPM) displayed proliferative cholangiocytes that reflected an acquired ductular reaction.
Acquired cholangiopathies such as EHBDO and cholangitis initially provoke ductular reactions, typically associate with increased cholestatic liver enzyme activity with or without variable hyperbilirubinemia, and histologically demonstrate duct-centric inflammation, ductal injury, and often periportal hepatitis.75 Compared with fibrosis that regionally expands the portal tracts in DPM, obstructive cholangiopathies are less extensive, often laminate ductule structures with concentric fibrosis, and do not bridge portal regions unless obstruction is complete and chronic (for a duration of months).20,75 Exclusion of a previous EHBDO before considering a diagnosis of Caroli DPM is important because distended tortuous bile ducts may persist after chronic EHBDO despite resolution of its other features. Clinicopathologic features of EHBDO usually do not associate with Caroli DPM, unless a Caroli phenotype is associated with symptomatic bacterial cholangitis, choledochitis, or cholelithiasis. In 1 cat in the present study, historic EHBDO secondary to cholelithiasis was provoked by Caroli and choledochal cyst DPM phenotypes. In animals with DPM and concurrent bacterial cholangitis, vacillating liver enzyme activities without hyperbilirubinemia are often manifest. This creates a diagnostic conundrum if a pathologist who is unfamiliar with the histologic appearance of DPM only describes regional inflammation. Liver tissue from cats with bacterial cholangitis usually displays cholangiocyte proliferation intimate to inflammatory foci. Comparatively, however, this proliferative response is subtle, considering the breadth of malformative ductal elements in DPM.
The HPI did not significantly differ between healthy dogs and cats and those with PPDH or CCDH. This finding confirmed the absence of hepatocyte proliferation in the affected animals and was expected because of the lack of evidence that suggested hepatocyte proliferation contributes to DPM morphology. A difference in HPI was not seen between healthy dogs and cats. The fact that animals with PPDH or CCDH and proliferative-type DPM had significantly higher BD-PT, compared with healthy animals, was expected. However, the significantly higher BD-PT in healthy cats, compared with healthy dogs, was unanticipated and likely represented an anatomic difference between these species.
The strong relationship between PPDH or CCDH and DPM in the present study suggested an influence of primary cilia on the development of the diaphragm, especially embryogenesis of the septum transversum. Seemingly, this has not been previously suggested or experimentally explored. Interestingly, people with congenital diaphragmatic hernias commonly have other malformations.76 Although these other malformations predominantly involve cardiac structures, liver lobe hypoplasia or atresia and gallbladder agenesis have also been documented.74 Curiously, these hepatobiliary anomalies are also variably recognized in people and animals with DPM.20,21,74 Embryologically, the diaphragm evolves from several structural components.77,78 Initially, the septum transversum forms an incomplete partition between the pericardial and peritoneal cavities, eventually evolving the central tendon. Pleuroperitoneal membranes thereafter fuse with the dorsal esophageal mesentery and septum transversum, separating the thoracic and abdominal cavities and forming the crura of the diaphragm. Fusion defects in the pleuroperitoneal membranes during embryogenesis can cause congenital diaphragmatic hernias, rarely including hiatal hernias.77 Interrupted remodeling of the septum transversum into the liver and sinus venosus may contribute to PPDH.77 Because the diaphragm forms part of the pericardial sac in people, PPDH may evolve secondary to trauma (ie, tearing of the septum transversum or pleuroperitoneal folds).2,79–81 However, trauma is an unlikely cause of PPDH in dogs and cats because these species lack similar continuity between the abdominal, pleural, and pericardial cavities.81
A previous report33 documented complex midline malformations in 5 Golden Retriever littermates that were circumstantially linked with midgestational administration of doxycycline, prednisone, and tramadol. Malformations in that litter included PPDH in all 5 puppies along with variable malformations of the heart, liver, kidneys, and skeleton; cryptorchidism; and abdominal wall hernias. Histologic evaluation of liver tissue from 1 puppy was consistent with the congenital hepatic fibrosis DPM phenotype. This dog also had gallbladder agenesis and congenital hepatic fibrosis associated with presinusoidal portal hypertension, APSS, and ascites.33 Another puppy had a hypoplastic or an atretic liver lobe, which occasionally accompanies DPM in dogs and cats.20,21 Interestingly, 1 dog with PPDH but without DPM in the present study had gestational exposure to prednisone and doxycycline. In a retrospective case series of dogs and cats with PPDH, a young Weimaraner with stunted body condition, poor surgical recovery, signs of hepatic encephalopathy, ascites, and a cirrhotic liver was described.2 That dog likely had a congenital hepatic fibrosis DPM phenotype, which is commonly mistaken for cirrhosis.20,26 A Himalayan cat with hepatic cysts entrapped in a PPDH is described in one report,5 and cats with PPDH and polycystic kidney disease, which is known to associate with hepatic DPM, were described in other reports.3,7 Another publication14 documents PPDH in a kindred of Persian cats (1 queen and 4 kittens), a breed with a high risk for polycystic kidney disease because of a mutation in the polycystin-1 precursor gene (functionality essential for primary cilia).22,28 Findings in the present study corroborated an association between hepatic and renal DPM; renal DPM lesions were documented in 3 dogs with PPDH that also had hepatic DPM and in 3 cats with PPDH (1 with hepatic DPM and 2 with only polycystic kidneys).
Previous reviews1,3 of 98 cats with PPDH reported overrepresentation of longhaired breeds; in the present study, 50% were longhaired cats. Considering these findings, predisposition to PPDH may reflect mutation of the polycystin-1 precursor gene that more commonly affects Persian and other longhaired cats. This finding also supported the supposition that diaphragmatic embryogenesis may be influenced by the primary cilia. In dogs with PPDH, a predisposition for Weimaraners has been reported.1,2 Histologic evaluation of liver tissue from these dogs has rarely been pursued, although 1 Weimaraner with PPDH reportedly2 also had hepatic cirrhosis. Each of the 2 Weimaraners in the present study had proliferative-like DPM, with 1 qualifying for congenital hepatic fibrosis that led to death. On the basis of these observations, the liver tissue from Weimaraners with PPDH should be further evaluated histologically to determine whether DPM associates with PPDH in this breed.
Finding DPM lesions in the liver, kidneys, or both of dogs or cats with PPDH or CCDH is more than an anatomic or physiologic curiosity. Animals with hepatic DPM have an increased risk for bile-borne bacterial infections, as reported in people.20,21,49 Caroli-type malformations impose additional increased risk for intra- or extrahepatic cholelithiasis and choledochitis.20,82 Also, severe proliferative-like and Caroli DPM malformations may evolve to congenital hepatic fibrosis. Presence of renal DPM heightens the risk for progressive decline in kidney function secondary to cyst expansion that compressively injures adjacent nephron segments, as shown in cats.29 In the present study, the impact of DPM on survival time in animals with PPDH or CCDH could not be determined because most (> 70%) animals had DPM. Because all animals with PPDH or CCDH, even those with sudden death, were included, the impact of DPM on sudden death also could not be determined. The disproportionate distribution of only dogs with DPM to a group treated by surgical intervention prevented assessment of DPM on postoperative survival time. Likewise, the small number of cats with and without DPM that underwent surgical intervention prevented the evaluation of differences in survival times between these groups.
A shortcoming of this retrospective study was that ≥ 2 nonherniated liver lobes or kidneys were not biopsied for all animals. Consequently, the number of animals with hepatic or renal DPM phenotypes may have been underestimated. Phenotypes of DPM may differentially affect the liver lobes in a patient, with some lobes lacking DPM and others classically affected, hypoplastic, or atretic.20,21,83 Histologic comparison of herniated liver lobes from animals with PPDH and DPM and from animals with other liver lobe changes impacted by other conditions, including traumatic herniation, torsion, or ischemia (pathological controls), supported the supposition that clustered proliferative-like bile duct DPM persists in herniated liver lobes. Thus, DPM can be histologically defined in animals with PPDH, even when only herniated liver samples are available for evaluation. In this circumstance, determining Ki-67 immunoreactivity is recommended to confirm the absence of proliferative bile ducts (ie, ductular reaction), but basing the diagnosis of DPM solely on examination of liver tissue from a herniated liver lobe is not a recommended strategy. The most straightforward and confident diagnosis of DPM is based on the histologic evaluation of biopsies collected from ≥ 2 nonherniated liver lobes.
In conclusion, a strong association between DPM and PPDH or CCDH was demonstrated in dogs and cats. Recognition of gallbladder agenesis or hypoplasia, atresia of nonherniated liver lobes, or polycystic kidney lesions (determined by diagnostic imaging studies, surgery, or necropsy) should trigger consideration of DPM as a differential diagnosis. The best strategy for appraisal of DPM status in dogs and cats with PPDH is histologic and immunohistochemical evaluation of numerous liver tissue samples from several herniated and nonherniated liver lobes. Biopsy of several nonherniated liver lobes is advised because of the potential for variation of DPM involvement among lobes. Caution is advised when a diagnosis of cirrhosis or chronic cholangiohepatitis is made for some animals with PPDH or CCDH because DPM may be misidentified, particularly the congenital hepatic fibrosis phenotype. Thus, discordance between clinical and clinicopathologic features (ie, lack of jaundice and absence of increased serum cholestatic liver enzyme activities) and histologic interpretation of cirrhosis or advanced chronic cholangiohepatitis should instigate a request for rereview. Diagnosis of DPM warrants lifelong vigilance for complicating bacterial cholangitis, cholecystitis, and cholelithiasis. Progression to congenital hepatic fibrosis should be considered in animals that develop ascites and APSS.
References
- 1. ↑
Banz AC, Gottfried SD. Peritoneopericardial diaphragmatic hernia: a retrospective study of 31 cats and eight dogs. J Am Anim Hosp Assoc 2010;46(6):398–404.
- 2. ↑
Evans SM, Biery DN. Congenital peritoneopericardial diaphragmatic hernia in the dog and cat: a literature review and 17 additional case histories. Vet Radiol Ultrasound 1980;21(3):108–116.
- 3. ↑
Reimer SB, Kyles AE, Filipowicz DE, et al. Long-term outcome of cats treated conservatively or surgically for peritoneopericardial diaphragmatic hernia: 66 cases (1987–2002). J Am Vet Med Assoc 2004;224(5):728–732.
- 4. ↑
Burns CG, Bergh MS, McLoughlin MA. Surgical and nonsurgical treatment of peritoneopericardial diaphragmatic hernia in dogs and cats: 58 cases (1999–2008). J Am Vet Med Assoc 2013;242(5):643–650.
- 5. ↑
Liptak JM, Bissett SA, Allan GS, et al. Hepatic cysts incarcerated in a peritoneopericardial diaphragmatic hernia. J Feline Med Surg 2002;4(2):123–125.
- 6. ↑
Less RD, Bright JM, Orton EC. Intrapericardial cyst causing cardiac tamponade in a cat. J Am Anim Hosp Assoc 2000;36(2):115–119.
- 7. ↑
Rendano VT, Parker RB. Polycystic kidneys and peritoneopericardial diaphragmatic hernia in the cat: a case report. J Small Anim Pract 1976;17(7):479–485.
- 8. ↑
Scruggs SM, Bright JM. Chronic cardiac tamponade in a cat caused by an intrapericardial biliary cyst. J Feline Med Surg 2010;12(4):338–340.
- 9. ↑
Morgan KRS, Singh A, Giuffrida MA, et al. Outcome after surgical and conservative treatments of canine peritoneopericardial diaphragmatic hernia: a multi-institutional study of 128 dogs. Vet Surg 2020;49(1):138–145.
- 10. ↑
Lamb CR, Mason GD, Wallace MK. Ultrasonographic diagnosis of peritoneopericardial diaphragmatic hernia in a Persian cat. Vet Rec 1989;125(8):186.
- 11. ↑
Hay WH, Woodfield JA, Moon MA. Clinical, echocardiographic, and radiographic findings of peritoneopericardial diaphragmatic hernia in two dogs and a cat. J Am Vet Med Assoc 1989;195(9):1245–1248.
- 12. ↑
Bellah JR, Whitton DL, Ellison GW, et al. The surgical correction of concomitant cranioventral abdominal wall, caudal sternal, diaphragmatic, and pericardial defects in dogs. J Am Vet Med Assoc 1989;195:1722–1726.
- 13. ↑
Neiger R. Peritoneopericardial diaphragmatic hernia in cats. Compend Contin Educ Pract Vet 1996;18(5):461–468.
- 14. ↑
Margolis C, Zakošek Pipan M, Demchur J, et al. Congenital peritoneopericardial diaphragmatic hernia in a family of Persian cats. JFMS Open Rep 2018;4(2):2055116918804305.
- 15. ↑
Phillips H, Corrie J, Engel DM, et al. Clinical findings, diagnostic test results, and treatment outcome in cats with hiatal hernia: 31 cases (1995–2018). J Vet Intern Med 2019;33(5):1970–1976.
- 16. ↑
Gunay-Aygun M. Liver and kidney disease in ciliopathies. Am J Med Genet C Semin Med Genet 2009;151C(4):296–306.
- 17.
Madhivanan K, Aguilar RC. Ciliopathies: the trafficking connection. Traffic 2014;15(10):1031–1056.
- 19. ↑
Fabris L, Fiorotto R, Spirli C, et al. Pathobiology of inherited biliary diseases: a roadmap to understand acquired liver diseases. Nat Rev Gastroenterol Hepatol 2019;16(8):497–511.
- 20. ↑
Pillai S, Center SA, McDonough SP, et al. Ductal plate malformation in the liver of boxer dogs: clinical and histological features. Vet Pathol 2016;53:602–613.
- 21. ↑
Center SA. Ductal plate malformations. In: Proceedings of the American College of Veterinary Internal Medicine Forum. American College of Veterinary Internal Medicine; 2017:792–797.
- 22. ↑
Masyuk AI, Masyuk TV, LaRusso NF. Cholangiocyte primary cilia in liver health and disease. Dev Dyn 2008;237(8):2007–2012.
- 23. ↑
Kawasaki T, Carmichael FJ, Saldivia V, et al. Relationship between portal venous and hepatic arterial blood flows: spectrum of response. Am J Physiol 1990;259(6 pt 1):G1010–G1018.
- 24. ↑
Gunay-Aygun M, Gahl WA, Heller T. Congenital hepatic fibrosis overview. In: Adam MP, Ardinger HH, Pagon RA, et al., eds. GeneReviews. University of Washington; 1993–2021:1–41.
- 25. ↑
Desmet VJ. Congenital diseases of intrahepatic bile ducts: variations on the theme “ductal plate malformation”. Hepatology 1992;16(4):1069–1083.
- 26. ↑
Alsomali MI, Yearsley MM, Levin DM. Diagnosis of congenital hepatic fibrosis in adulthood fibrosis. Am J Clin Pathol 2020;153(1):119–125.
- 27. ↑
Sato K, Sakai M, Hayakawa S, et al. Gallbladder agenesis in 17 dogs: 2006–2016. J Vet Intern Med 2018;32(1):188–194.
- 28. ↑
Lyons LA, Biller DS, Erdman CA, et al. Feline polycystic kidney disease mutation identified in PKD1. J Am Soc Nephrol 2004;15(10):2548–2555.
- 29. ↑
Eaton KA, Biller DS, DiBartola SP, et al. Autosomal dominant polycystic kidney disease in Persian and Persian-cross cats. Vet Pathol 1997;34(2):117–126.
- 30.
Drögemüller M, Jagannathan V, Welle MM, et al. Congenital hepatic fibrosis in the Franches-Montagnes horse is associated with the polycystic kidney and hepatic disease 1 (PKHD1) gene. PLoS One. 2014;9(10):e110125.
- 31.
Stayner C, Poole CA, McGlashan SR, et al. An ovine hepatorenal fibrocystic model of a Meckel-like syndrome associated with dysmorphic primary cilia and TMEM67 mutations. Sci Rep. 2017;7(1):1601.
- 32.
Dillard KJ, Hytönen MK, Fischer D, et al. A splice site variant in INPP5E causes diffuse cystic renal dysplasia and hepatic fibrosis in dogs. PLoS One 2018;13(9):e0204073.
- 33. ↑
Kaplan JL, Gunther-Harrington CT, Sutton JS, et al. Multiple midline defects identified in a litter of Golden Retrievers following gestational administration of prednisone and doxycycline: a case series. BMC Vet Res 2018;14(1):86–95.
- 34.
Eipel C, Abshagen K, Vollmar B. Regulation of hepatic blood flow: the hepatic arterial buffer response revisited. World J Gastroenterol 2010;16(48):6046–6057.
- 35.
Lanier VC Jr, Buchanan RD, Foster JH. Hepatic morphologic changes following end-to-side portocaval shunt in dogs. Am Surg 1968;34(3):185–195.
- 36.
Schermerhorn T, Center SA, Dykes NL, et al. Characterization of hepatoportal microvascular dysplasia in a kindred of cairn terriers. J Vet Intern Med 1996;10(4):219–230.
- 37.
Baade S, Aupperle H, Grevel V, et al. Histopathological and immunohistochemical investigations of hepatic lesions associated with congenital portosystemic shunt in dogs. J Comp Pathol 2006;134(1):80–90.
- 38.
Lee KC, Winstanley A, House JV, et al. Association between hepatic histopathologic lesions and clinical findings in dogs undergoing surgical attenuation of a congenital portosystemic shunt: 38 cases (2000–2004). J Am Vet Med Assoc 2011;239(5):638–645.
- 39.
Sobczak-Filipiak M, Szarek J, Badurek I, et al. Retrospective liver histomorphological analysis in dogs in instances of clinical suspicion of congenital portosystemic shunt. J Vet Res 2019;63(2):243–249.
- 40.
Lautt WW. Mechanism and role of intrinsic regulation of hepatic arterial blood flow: hepatic arterial buffer response. Am J Physiol 1985;249(5 pt 1):G549–G556.
- 41. ↑
Terada T, Ishida F, Nakanuma Y. Vascular plexus around intrahepatic bile ducts in normal livers and portal hypertension. J Hepatol 1989;8(2):139–149.
- 42. ↑
Desmet VJ. Ludwig symposium on biliary disorders—part I. Pathogenesis of ductal plate abnormalities. Mayo Clin Proc 1998;73(1):80–89.
- 43. ↑
West AB, Chatila R. Differential diagnosis of bile duct injury and ductopenia. Semin Diagn Pathol 1998;15(4):270–284.
- 44. ↑
Schotanus BA, van den Ingh TSGAM, Penning LC, et al. Cross-species immunohistochemical investigation of the activation of the liver progenitor cell niche in different types of liver disease. Liver Int 2009;29(8):1241–1252.
- 45. ↑
Ijzer J, Schotanus BA, Vander Borght S, et al. Characterisation of the hepatic progenitor cell compartment in normal liver and in hepatitis: an immunohistochemical comparison between dog and man. Vet J 2010;184(3):308–314.
- 46. ↑
Yonem O, Bayraktar Y. Clinical characteristics of Caroli’s syndrome. World J Gastroenterol 2007;13(13):1934–1937.
- 47. ↑
Suchy FJ. Caroli disease. UpToDate. Accessed June 16, 2020. https://www.uptodate.com/contents/caroli-disease.
- 48. ↑
McKenna SC, Carpenter JL. Polycystic disease of the kidney and liver in the Cairn Terrier. Vet Pathol 1980;17(4):436–442.
- 49. ↑
McAloose D, Casal M, Patterson DF, et al. Polycystic kidney and liver disease in two related West Highland White Terrier litters. Vet Pathol 1998;35(1):77–81.
- 50. ↑
Bosje JT, van den Ingh TS, van der Linde-Sipman JS. Polycystic kidney and liver disease in cats. Vet Q 1998;20(4):136–139.
- 51. ↑
Raynaud P, Tate J, Callens C, et al. A classification of ductal plate malformations based on distinct pathogenic mechanisms of biliary dysmorphogenesis. Hepatology 2011;53(6):1959–1966.
- 52. ↑
Drenth JP, Chrispijn M, Bergmann C. Congenital fibrocystic liver diseases. Best Pract Res Clin Gastroenterol 2010;24(5):573–584.
- 53.
Kerkar N, Norton K, Suchy FJ. The hepatic fibrocystic diseases. Clin Liver Dis 2006;10(1):55–71.
- 54. ↑
Veland IR, Awan A, Pedersen LB. Primary cilia and signaling pathways in mammalian development, health and disease. Nephron Physiol 2009;111(3):39–53.
- 55. ↑
Wills ES, Roepman R, Drenth JPH. Polycystic liver disease: ductal plate malformation and the primary cilium. Trends Mol Med 2014;20(5):261–270.
- 56. ↑
Carpentier R, Suñer RE, van Hul N, et al. Embryonic ductal plate cells give rise to cholangiocytes, periportal hepatocytes, and adult liver progenitor cells. Gastroenterology 2011;141(4):1432–1438.
- 57.
Lee-Law PY, van de Laarschot LFM, Banales JM, et al. Genetics of polycystic liver diseases. Curr Opin Gastroenterol. 2019;35(2):65–72.
- 58.
Terada T, Nakanuma Y. Detection of apoptosis and expression of apoptosis-related proteins during human intrahepatic bile duct development. Am J Pathol 1995;146(1):67–74.
- 59. ↑
Gunay-Aygun M, Tuchman M, Font-Montgomery E, et al. PKHD1 sequence variations in 78 children and adults with autosomal recessive polycystic kidney disease and congenital hepatic fibrosis. Mol Genet Metab 2010;99(2):160–173.
- 60. ↑
Yasoshima M, Sato Y, Furubo S, et al. Matrix proteins of basement membrane of intrahepatic bile ducts are degraded in congenital hepatic fibrosis and Caroli’s disease. J Pathol 2009;217(3):442–451.
- 61. ↑
Wong MY, McCaughan GW, Strasser SI. An update on the pathophysiology and management of polycystic liver disease. Expert Rev Gastroenterol Hepatol 2017;11(6):569–581.
- 62. ↑
Xu J, Liu X, Koyama Y, et al. The types of hepatic myofibroblasts contributing to liver fibrosis of different etiologies. Front Pharmacol 2014;5:167.
- 63. ↑
Locatelli L, Cadamuro M, Spirlì C, et al. Macrophage recruitment by fibrocystin-defective biliary epithelial cells promotes portal fibrosis in congenital hepatic fibrosis. Hepatology 2016;63(3):965–982.
- 64.
Tsunoda T, Kakinuma S, Miyoshi M, et al. Loss of fibrocystin promotes interleukin-8-dependent proliferation and CTGF production of biliary epithelium. J Hepatol 2019;71(1):143–152.
- 65.
Harada K, Sato Y, Ikeda H, et al. Epithelial-mesenchymal transition induced by biliary innate immunity contributes to the sclerosing cholangiopathy of biliary atresia. J Pathol 2009;217(5):654–664.
- 66.
Liu Y. Epithelial to mesenchymal transition in renal fibrogenesis: pathologic significance, molecular mechanism, and therapeutic intervention. J Am Soc Nephrol 2004;15(1):1–12.
- 67.
Narkewicz MR, Kasaragod A, Lucia MS, et al. Connective tissue growth factor expression is increased in biliary epithelial cells in biliary atresia. J Pediatr Surg 2005;40(11):1721–1725.
- 68. ↑
Sun T, Annunziato S, Tchorz JS. Hepatic ductular reaction: a double-edged sword. Aging (Albany NY) 2019;11(21):9223–9224.
- 69. ↑
Fabris L, Brivio S, Cadamuro M, et al. Revisiting epithelial-to-mesenchymal transition in liver fibrosis: clues for a better understanding of the “reactive” biliary epithelial phenotype. Stem Cells Int 2016;2016:2953727.
- 70. ↑
Desmet VJ. Ductal plates in hepatic ductular reactions. Hypothesis and implications. I. Types of ductular reaction reconsidered. Virchows Arch 2011;458(3):251–259.
- 71. ↑
Sato K, Marzioni M, Meng F, et al. Ductular reaction in liver diseases: pathological mechanisms and translational significances. Hepatology 2019;69(1):420–430.
- 72. ↑
Turányi E, Dezsö K, Csomor J. Immunohistochemical classification of ductular reactions in human liver. Histopathology 2010;57(4):607–614.
- 73. ↑
Strazzabosco M, Fabris L. Development of the bile ducts: essentials for the clinical hepatologist. J Hepatol 2012;56(5):1159–1170.
- 74. ↑
Funaki N, Sasano H, Shizawa S, et al. Apoptosis and cell proliferation in biliary atresia. J Pathol 1998;186(4):429–433.
- 75. ↑
Aller M-A, Arias J-L, García-Domínguez, et al. Experimental obstructive cholestasis: the wound-like inflammatory liver response. Fibrogenesis Tissue Repair 2008;1(1):6.
- 76. ↑
Fauza DO, Wilson JM. Congenital diaphragmatic hernia and associated anomalies: their incidence, identification, and impact on prognosis. J Pediatr Surg 1994;29(8):1113–1117.
- 77. ↑
Noden DM, de Lahunta A. Malformations of the diaphragm. In: Noden DM, de Lahunta A, eds. The Embryology of Domestic Animals: Developmental Mechanisms and Malformations. The Williams & Wilkins Co; 1985:288–290.
- 78. ↑
Swain JM, Klaus A, Achem SR, et al. Congenital diaphragmatic hernia in adults. Semin Laparosc Surg 2001;8(4):246–255.
- 79. ↑
Bolton GR, Ettinger S, Roush JC II. Congenital peritoneopericardial diaphragmatic hernia in a dog. J Am Vet Med Assoc 1969;155(5):723–730.
- 80.
Clinton JM. A case of congenital periocardio-peritoneal diaphragmatic communication in a dog. J Am Vet Radiol Soc 1967;8:57–60.
- 81. ↑
Baker GJ, Williams CS. Diaphragmatic pericardial hernia in the dog. Vet Rec 1966;78(17):578–583.
- 82. ↑
Reed CA. Pericardio-peritoneal hernia in mammals, with description of a case in the domestic cat. Anat Rec 1951;110(1):113–119.
