Portosystemic Vascular Anomalies & Hepatic MVD: Evidence of Common Genetics in Small Dogs
ACVIM 2008
Sharon A. Center, DVM, DACVIM
Ithaca, NY, USA

Introduction

Extrahepatic portosystemic vascular anomalies (PSVA) in small breed dogs, originally described in the 1970s, have received notorious attention in the veterinary literature. We estimate that this congenital malformation comprises approximately 0.6 to 2.0% of a specialty hospital patient population. However, microvascular dysplasia (MVD), a genetically related hepatic vascular disorder is far more common; 15 to 30:1 compared to the prevalence of PSVA in affected kindreds based on investigations by the author in an expansive genotyping initiative. Unfortunately, discovery of high serum bile acid concentrations in dogs with MVD often leads to expensive and invasive diagnostic assessments for PSVA in pet dogs. MVD dogs are typically asymptomatic, the lesion is non-progressive, and affected dogs usually live a normal life span. Dogs affected only with MVD are typically discovered after finding high serum bile acid concentrations either after screening kindreds for PSVA or serendipitously at the time of non-hepatic illness. Finding the genetic cause of the PSVA/MVD trait is important and our laboratory has been conducting such a genotyping initiative in small breed dogs during the past 4 years. Pedigree analyses in several small pure breed dogs (Tibetan Spaniels, Cairn Terriers, Maltese, Yorkshire Terriers, Miniature Schnauzers, Havanese, Norfolk Terriers, Norwich Terriers, Shih Tzu) and linkage analyses supports that PSVA and MVD are inherited and genetically related disorders.

While PSVA affected dogs represent the most severe vascular malformations and usually manifest clinical signs, we estimate that approximately 15-20% of these are asymptomatic seemingly reflecting their smaller degree of shunting. We have identified breeding bitches with PSVA in 3 investigated breeds. These bitches successfully whelped several small litters before their PSVA status was clarified. Similarly, we have identified breed influential asymptomatic PSVA stud dogs in several breeds. However, the PSVA/MVD trait has largely been perpetuated by the innocent selection of asymptomatic MVD dogs as foundation breeding stock. We are confident that prior publications addressing the genetic basis of PSVA have widely under-estimated trait incidence in Cairn Terriers owing to the failure to acknowledge the relationship between PSVA and MVD as well as the use of ammonia rather than total serum bile acids (TSBA) as a method of trait phenotyping. Ammonia determinations do not identify dogs with MVD and the test is simply unreliable in practical application. Our studies substantiate a wide variation in the prevalence of the PSVA/MVD trait among different dog breeds and among kindreds of dogs within a breed. Data describe prevalence rates ranging from 30% to > 80% in afflicted kindreds. Further studies in our Tibetan Spaniel kindred with an initial affectation rate of 32% has shown that the prevalence vacillates as high as 50% with kindred expansion (more breeding, new combinations of parents, importation of new unrelated founders from other countries), suggesting that this is a breed-wide trait rather than a kindred-limited disorder. Our clinical observations suggest that PSVA may be a lethal trait in full expression or penetrance explaining, in part, the lower number of dogs with PSVA compared to MVD. In some kindreds, the high frequency of the trait seemingly has lead to embryonic deaths (fetal resorption) and small litter sizes (e.g., certain Maltese and Yorkshire Terrier kindreds).

Vasculogenesis/Angiogenesis: Considerations1,2

The vascular system derives from two fundamental processes: vasculogenesis and angiogenesis. Vasculogenesis is the formation of blood vessels in the embryo (differentiation of endothelial precursor cells [angioblasts] into endothelial cells in concert with formation of a primitive vascular network). Steps essential to embryonic vasculogenesis include establishment of angioblasts from mesoderm, assembly of angioblasts into vascular structures, formation of vascular lumens, and the organization of a primitive continuous vascular network. Mechanisms essential to vasculogenesis are dynamic and are modulated by interactions between cells and the extracellular matrix (ECM) in the presence of growth factors and morphogens (e.g., differentiation of mesodermal cells to angioblasts, migration of angioblasts, formation of blood islands, and primitive intra-embryonic vascular network. Vascular precursors generating the primary vascular plexus assemble at the location of developing vessels. Migration to the region of vascular development occurs under the regulation of a large group of soluble (e.g., growth factors, regulatory proteins) and interaction molecules. A fine balance of complex signals impose overlapping attractive and repulsive gradients that guide cell migration (e.g., soluble factors from regions needing vascularization diffuse to endothelial precursors, attracting them and keeping them in their migratory phenotype state until they reach their target). The process proceeds during embryogenesis as angioblasts migrate into developing organs and assemble to form primordial blood vessels where newly differentiated endothelial cells provide the structure of the primary capillary plexus. Angiogenesis, defined as the growth of new capillaries from pre-existing blood vessels, thereafter occurs. This process involves 4 distinct phases again under the influence of complex and interacting signals: 1) sprouting of capillaries from pre-existing vessels, 2) non-sprouting angiogenesis resulting in enlargement, fusion or splitting of pre-existing vessels by transcapillary pillars, 3) pruning: loss of certain endothelial tubes and cells, and 4) maturation that involves the recruitment of pericytes/smooth muscle cells associated with circulatory and metabolic demands. While non-perfused primordial vessels regress, others differentiate under the influence of circulating factors and shear stress that induce modifications in cell-to-cell and cell-matrix interactions. Once mature, the vessel is relatively stable with low endothelial cells turnover. There are a large number of potential candidate genes that may influence vasculogenesis/angiogenesis and liver development. We have determined the location of a large number of candidate genes in the annotated canine genome that have potential to influence vasculogenesis/angiogenesis relative to our common region of linkage in breeds with inheritable PSVA/MVD; none are within 5 Mb distance.

PSVA/MVD Trait is Complex

Dogs with PSVA have distinct microscopic hepatic abnormalities reflecting portal hypoperfusion as well as other vascular abnormalities. Lesions include: increased hepatic arteriole (small arteries) cross sections, the presence of non-perfused vessels (lymphatics), small, juvenile or under-developed portal triads (very small structures) randomly located in the hepatic parenchyma, hepatic lobular atrophy, an increased number of small binucleated hepatocytes, and in some dogs, multifocal lipogranulomas and lipogranulomas perivascular to the hepatic venule that in some, impose a veno-occlusive effect . Microscopic features affiliated with PSVA are indistinguishable from those observed in dogs with MVD. Arterioles increase in cross sections (often referred histopathologically as arteriolar duplication) as a result of their tortuous or coiled conformation, confirmed by arteriograms of dogs with portal hypoperfusion. Cross sections of such vessels yields an apparent greater number of arterioles within portal triads. Such arteriolar change is associated with the hepatic arteriolar buffer response that compensates for portal hypoperfusion (increased flow and pressure due to acquired dominance of the arterial circulation). Yet it remains unclear whether these arteriolar changes reflect de novo neovascularization or remodeling of pre-existent arterioles. On careful inspection of large hepatic biopsy specimens from some dogs with MVD, some hepatic venules are inappropriately located adjacent to or within portal triads, enabling direct intrahepatic shunting. Most PSVA and MVD affected dogs have hepatic venules with thickened or prominent throttling musculature. In some dogs, outflow occlusion at the level of the hepatic venule is frankly evident. Surgical wedge and laparoscopic liver biopsies collected from multiple liver lobes in dogs with MVD and PSVA confirm that microscopic lesions are inconsistent among liver lobes. Our studies have confirmed that needle biopsy and single liver lobe biopsies can miss lesions in MVD affected dogs. Portovenograms, colorectal scintigraphy, and MRI contrast studies in dogs with MVD corroborate histologic evidence that liver lobes have variable portal perfusion. This variability in liver lobe portal perfusion in dogs with MVD likely explains the apparent "portal streamlining" described in one study where colorectal scintigraphy was used to differentiate PSVA dogs among suspected patients.3 Considered collectively, histologic and imaging features corroborates that MVD and PSVA represent complex disorders of hepatic angiogenesis / vasculogenesis and that the PSVA/MVD trait expresses with a spectrum of severities. 1) It is certain that MVD occurs in dogs lacking extrahepatic macroscopic shunting vasculature and that it often co-exists with PSVA: PSVA dogs with MVD do not undergo bile acid normalization on full ligation. 2) It is certain that some dogs with PSVA can be cured by surgical ligation: they lack MVD. 3) It is certain that a small subset of dogs with PSVA have true severe hypoplasia (atresia) of the extrahepatic portion of the portal vein resulting in the formation of multiple acquired portosystemic shunts. 4) It is certain that a small subset of dogs with PSVA have true severe hypoplasia (atresia) of the intrahepatic portal vasculature that results in the formation of multiple acquired portosystemic shunts. 5) It is apparent that some dogs from kindreds affected with the PSVA/MVD trait may have vascular malformations not influencing hepatic portal perfusion (splenic vein to vena cava shunts). What is confusing and unappreciated in the veterinary community is that any cause of extrahepatic shunting at any age causes hepatic histologic features consistent with portal "hypoplasia" as has been described in PSVA/MVD.4 This has been proven by several research groups working with normal dogs with surgically created Eck fistulas (end to side portal vein to vena cava anastomosis). Within 4 months of Eck fistula formation, hepatic arteriolarization becomes well established such that histologic features cannot be differentiated from those associated with PSVA/MVD. Thus, the most prominent histologic lesions in PSVA/MVD reflect portal hypoperfusion. Considering evidence accumulated from pedigree and linkage analyses and careful histologic and imaging studies of affected dogs, microscopic features associated with the PSVA/MVD trait cannot be accurately described by the terminology of portal hypoplasia. Hypoplasia denotes lack of embryologic vessel formation which unfortunately cannot be discerned on liver biopsy. Rather, liver biopsy identifies lesions reflecting portal hypoperfusion. The wide variety of data accumulated to date recommends that the terminology of portal hypoplasia as a histologic diagnosis be terminated and replaced with portal hypoperfusion. The complexity and variability of histologic lesions in dogs with MVD argues that this malformation syndrome is characterized by microscopic vascular abnormalities not simply isolated to the portal vein. While a syndrome of primary portal hypoplasia has been described associated with fibrosis, this lesion is exceedingly uncommon, and not typical of the PSVA/MVD trait.5 Variation in the type of vascular malformations have been notable among breeds with some Havanese dogs presenting with complex malformations. Popular breeds afflicted with PSVA/MVD (Yorkshire Terriers, Maltese) have a wide spectrum of vascular malformations ranging from dogs with PSVA lacking clinical signs to dogs with multiple vascular malformations having severe clinical signs evident within the first few weeks of life (Consult table at end of notes)

Genetic Evidence for a Common Inherited PSVA/MVD Trait

We have investigated the inheritance of the PSVA/MVD trait using total serum bile acid concentrations (enzymatic linked methodology determining total serum bile acid concentrations, paired pre-meal and 2-hour post-meal bile acid samples) for phenotype affectation status (values > 25 uM/L = affected, < 25 uM/L = unaffected). Medical records for dogs with historical illnesses have been reviewed to remove phenocopy errors. Generally, puppies included in our genotyping studies have serum bile acid concentrations measured when they are < 6 months of age; this avoids inadvertent phenocopy errors. Deceased elderly founders have had liver tissue collected at death to verify their affectation status; in no case have phenotyping errors been detected. We initially investigated the potential informativeness of a kindred of Tibetan Spaniels with a long standing history of a probable inheritable PSVA and MVD trait in which affected dogs had been clinicopathologically characterized, imaging studies completed, and liver biopsies evaluated. We used a simulation program to determine the maximum achievable LOD score with and without linkage between microsatellite markers and the trait.The kindred had a 31% trait frequency and included 11 founders and 47 progeny. After simulation predicted adequate pedigree informativeness, we completed a genome-wide survey for linkage with the PSVA/MVD trait, with affectation status designated by serum bile acid concentrations, using the microsatellite screening set-2 (MSS-2). Two-point linkage analysis with assigned liability scores dependent on the total serum bile acid concentrations, identified one significantly linked region with a peak LOD score of 3.4 (a LOD score of > 3.0 is considered significant at a genome wide level). Multipoint linkage analysis confirmed linkage to the region encompassing the peak LOD score, worthy of saturation microsatellite genotyping. The genetic model best fitting our data is trait transmission as an autosomal dominant with incomplete penetrance. To better resolve the linkage region, we undertook mapping with additional microsatellite markers, developed specifically for this region. This study narrowed the area of interest to 9 Mb. We expanded our linkage investigation initially to a second breed (Cairn Terriers) commonly afflicted with PSVA/MVD (5 families, 11 founders, n= 90 dogs), and then to 3 further afflicted breeds (Maltese [8 families, n=80 dogs], and a smaller kindred of Havanese and Shih Tzu dogs. Studies confirmed linkage between the PSVA/MVD trait and markers in this region in all breeds with a cumulative peak LOD score > 6.0. This established that the PSVA/MVD trait in each of these breeds is at least allelic, and may represent an ancient founder mutation predating the separation of the breeds into isolate populations. Careful examination of microsatellite haplotypes in each breed revealed informative recombinations that have further narrowed our region of interest. We are currently undertaking dense SNP mapping in our region of interest in breeds with confirmed linkage with the PSVA/MVD trait. Several additional breeds are undergoing pedigree and linkage studies for inclusion in SNP haplotyping.

Since we propose an autosomal dominant mode of inheritance with incomplete penetrance it remains possible that one or more contributing loci may influence the PSVA/MVD trait. Alternatively, since we have identified a few dogs with vascular malformations not influencing hepatic perfusion, we acknowledge that serum bile acids are unable to correctly phenotype all dogs carrying the PSVA/MVD mutation. Along the same lines, the complex dynamic physiologic variables influencing the endogenous bile acid challenge could further complicate correct phenotype designation. That serum bile acid concentrations are unable to accurately identify every dog carrying the PSVA/MVD trait is consistent with the dismal success of attempted trait elimination when bile acid concentrations are used for selection of breeding stock. Currently, breeders are advised to avoid using dogs with total serum bile acid concentrations > 25 uM/L as foundation stock, and to appraise the bile acid status of all puppies produced to identify parents silently carrying the trait. It is obvious that it is essential to continue our genotyping initiative to identify a disease risk haplotype or mutation associated with the PSVA/MVD trait. Hopefully we will be able to develop a test or tests that can accurately identify dogs carrying the PSVA/MVD determinants. We have on hand, extracted DNA samples from > 1,000 dogs phenotyped with bile acids from large informative pure breed kindreds as well as several hundred samples from PSVA dogs awaiting validation of a disease-linked risk allele or SNP haplotypes test.

Types (arbitrary designation) of PSVA / MVD syndrome malformations according to vascular manifestations.

 

Single congenital extrahepatic portosystemic communication

Two congenital extra-hepatic porto-systemic shunts

MVD:
attenuated intra-hepatic portal tributaries (variably reduced flow to full atresia)

MVD:
Abnormal location and flow through hepatic venule: lobule outflow obstruction

Poorly developed extra-hepatic portal vein (porta hepatis): extra-hepatic portal hypoplasia

No extra-hepatic portal vein in porta hepatis: portal atresia

Acquired porto-systemic shunts

Ductus venosus, right or left divisional branch

Projected outcome: with successful surgical ligation*

Type 1

+

+/-

-

-

-

-

-

-

1

Type 2

+

+/-

+

-

-

-

- (possible with ligation)

-

2 or 3

Type 3

+

+/-

+

+

-

-

+/-
(+ with ligation)

-

2, 3, or 4

Type 4

+

+/-

+

-

+

-

+/-
(+ with ligation)

-

3,4,5

Type 5

+

+/-

+/-

-

-

+

+

-

4,5

Type 6

-

-

-

-

-

-

-

+

1

Type 7

-

-

+

-

-

-

-

+

2 or 3

Type 8

-

-

+

-

-

-

+

+

4,5

* "Cured" =1
Improved but retains high SBA=2
Requires medical Rx, retains high SBA=3
No Change = 4
Worsened =5

References

1.  Schmidt A. Circ Res 2007:101;125.

2.  Brouillard P. Clin Genet 2003;63:340.

3.  Daniels GB. Vet Radiol Ultrasound 2004:45:78-84.

4.  WSAVA Standards for Clinical and Histological Diagnosis of Canine and Feline Liver Diseases.

5.  Spee B. Comparative Hepatol 2005;4:7 doi:10.1186/1476-5926-4-7

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Sharon Center, DVM, DACVIM
Cornell University
Ithaca, NY


MAIN : Liver Study Group : Portosystemic Vascular Anomalies
Powered By VIN
SAID=27