Inheritance of Renal Amyloidosis in Shar-Pei

posted: by: Wurtsboro Veterinary Clinic Tags: "Clinic Specials" "News" 

Journal of Heredity 1993; 84:438-442.

Inheritance of Renal Amyloidosis in Chinese Shar-Pei dogs
Ariel L. Rivas (1),  Linda Tintle (2),  Vicki Meyers-Wallen (3,4), Janet M. Scarlett (3), Curtis P. van Tassell (5) and Fred W. Quimby (1)

1 Department of Pathology. New York State College of Veterinary Medicine. Cornell University. Ithaca, NY 14850
2 Wurtsboro Veterinary Clinic. Wurtsboro, NY 12790
3 Department of Clinical Sciences. New York State College of Veterinary Medicine. Cornell University. Ithaca, NY 14850
4 J.A. Baker Institute. New York State College of Veterinary Medicine. Cornell University. Ithaca, NY 14850
5 Department of Animal  Sciences.  New York State College of Agriculture. Cornell University. Ithaca, NY 14850

The authors thank Drs. Kim Doane, James Dixon, John Dwyer, Clifford Glade, Mark Goodell, Edward Gschrey, Beth Jamison, John Ouellette, Pamela Schott, James Tompsett and Adawna Windom for providing clinical cases and Lillian Kletter for the pedigree analysis.


Renal amyloidosis (RA) and recurrent fever of unknown origin (RFUO) are characteristics of Familial Mediterranean Fever (FMF), a human disorder inherited as an autosomal-recessive trait.  Although no animal model has been established for FMF, a similar syndrome of RFUO and RA has been reported in Chinese Shar-Pei (CSP) dogs. This report addressed two questions: 1) is RA inherited in CSP dogs and 2) if so, is it possible to hypothesize the type of inheritance involved? Two studies were conducted to answer these questions. A historical cross-sectional comparison, which included CSP and non-CSP dogs with RA; and a prospective study that included CSP dogs with RA, RFUO or both. The cross-sectional comparison resulted in an odds ratio of 10 for RA in CSP dogs under 7 years of age.  A prospective study of 28 dogs with RA or RFUO identified 20 that had RFUO and RA, three with RA alone, and five with RFUO alone. RFUO preceded RA in all cases with both conditions. RFUO/RA were observed in both sexes.  Four dogs with RFUO with/without RA were born to parents that either were alive at age 7 or died because of conditions other than kidney failure/RA. When one parent was known to express one of these conditions, prevalence of RA was between 25% and 50% among littermates.  The results of both cross-sectional and prospective studies support the hypothesis that RA is inherited in CSP dogs in association with RFUO, that they develop RA at a younger age than non-CSP dogs, and that the acquisition of this trait is compatible with autosomal recessive inheritance.  However, further studies are needed to discriminate between simple (single-gene) and polygenic (multiple-gene) autosomal recessive inheritance of this condition.  In sum, these finding support previous reports that indicate that CSP dogs with combined RFUO/RA may provide an animal model for human FMF.


Renal amyloidosis (RA), a condition in which amyloid fibrils are deposited extracellularly in the kidney, has been described as a primary, inherited disorder and as a disorder secondary to chronic  inflammatory diseases  (i.e.,  tuberculosis,  rheumatoid arthritis). Frequent episodes of unexplained fever, inflammation of the serosa and joints, and a predisposition to renal amyloidosis are observed in Familial Mediterranean Fever (FMF) (Heller et al. 1955). It has been reported that FMF is a genetic disease, with autosomal recessive inheritance (Sohar et al. 1967; Rogers et al. 1989). This disease occurs predominantly in human ethnic groups from eastern Mediterranean regions  (Arab,  Armenian and Jews) (Matzner 1989).   Non-Ashkenazi Jewish, Armenian, Turkish and Arab populations are primarily affected by FMF (Sohar et al.  1967; Schwabe and Peters; 1974; Ozdemir and Sokmen; 1969 and Barakat et al. 1986).  The frequency of the gene for this disorder is high among some groups of these populations, reaching 1 in 22 among Jews in North Africa and 1 in 14 among Armenians in Los Angeles (Rogers et al. 1989).

No animal model has yet been established for this disease, although a similar condition has been described in CSP dogs (DiBartola et al. 1990; Rivas et al., 1992). RA has been reported in CSP dogs in association with recurrent fever of unknown origin (RFUO), elevated serum interleukin-6, hypergammaglobulinemia and normal or increased lymphocyte proliferation in response to mitogen stimulation in vitro (Rivas et al. 1992). However, the inheritance of RA/RFUO in CSP dogs has not been characterized. The purpose of this study was to assess whether there is an inherited predisposition to RA/RFUO in CSP dogs and, if so, to develop a hypothesis regarding its mode of inheritance. 

Material and Methods

Cross-sectional study.

Dogs with RA that were diagnosed histopathologically, as described below, between 1980 and 1990 at colleges participating in the Veterinary Medical Data Program (VMDP) at Purdue University, were selected to estimate the relative risk of RA among CSP. The VMDP includes data from veterinary teaching hospitals in the United States and Canada. Relative risk relates the occurrence of disease in a test group compared to a control group.  Estimates of relative risk were calculated by determination of the odds ratio, defined as the odds that a case (dog with RA) is a CSP, divided by the odds that a control (dog without RA) is a CSP. Therefore, an odds ratio > 1, implies that the frequency of RA is greater in CSP than in non-CSP. Cases seen several times during the study period were counted only once, and subdivided by sex and age.

Prospective studies.

Clinicians at Wurtsboro Veterinary Clinic and in private hospitals from the United States and Canada (see Acknowledgements) were asked to identify cases in CSP dogs with renal disease or RFUO. Animals with renal failure (blood urea nitrogen  >  28  mg/ml,  and/or  creatinine  >  1.8  mg/ml)  and histopathological confirmation of RA; or with >3 episodes/year of rectal temperatures > 41 C of unknown origin (RFUO); were selected and followed between 1989 and 1991. Cases of RA were identified histopathologically by demonstration of secondary amyloidosis type A (AA) by Congo red staining (with stain removal using KMn04 elution), and green birefringence under polarized light   (van Rijswijk and van Heusden 1979). The cases of RFUO/RA identified in CSP were described in detail elsewhere by clinical, pathological, immunological and epidemiological observations (Rivas et al. 1992). Pedigrees of individual RFUO/RA cases were determined from pedigree certificates issued by the Chinese Shar-pei Club of America.  Inbreeding coefficients (F) were calculated by the tabular method (Wright 1922; van Vieck et al. 1987). Dogs with F values higher than 0.03 were considered as inbred. F values below 0.03 were regarded as computational background and hence, disregarded.

Statistical analysis.

Odds ratios were used to estimate the strength of associations between breed and disease (Everett 1977). Prevalence of expected and observed cases (in progenies of > 6 year-old dogs), were compared using the chi square test. The relationship between age of onset and the cumulative frequency of cases was  compared by regression analysis.  Unless  indicated otherwise, the level of significance was p < 0.05.


Cross-sectional study.

Cross-sectional analysis of VMDP records indicated that CSP dogs, regardless of age, had a higher risk of renal amyloidosis than did non-CSP dogs (odds ratio: 4.5; p< .001, table 1). Among dogs 7 years of age or younger, the odds ratio for CSP dogs to have renal amyloidosis was 10.  Renal amyloidosis was diagnosed at an earlier age in CSP than in other breeds. All affected CSP dogs were diagnosed before age seven, with mean age of diagnosis in CSP dogs of 4 years (2-6 years, 95  confidence interval). In contrast, 68% of non-CSP dogs were diagnosed after 7 years of age (table 1). No retrospective data were obtained about RFUO since recurrent fever of unknown origin is not catalogued in VMDP records.

Prospective study.

This study involved more CSP dogs with histopathological confirmation of renal amyloidosis than were found through the VMDP records. Twenty eight CSP dogs having RFUO, RA or both conditions were identified (table 2).

Individual pedigrees were examined from these 28 cases. Twenty affected dogs (71%) had both RFUO and RA, while 8 cases (29%) had only one of these conditions. Onset of RFUO preceded that of RA in cases showing both conditions. Three dogs had RA and died without showing RFUO. Five dogs with RFUO since one year of age (numbers 16, 17, 18, 26 and 28) did not have evidence of amyloidosis two years after onset. RA was observed in both sexes. Mean age at diagnosis was 5.0 years (95%  confidence interval: 4-6 years, n=23), which closely approximates the results of the cross-sectional study (table 2). The age range at diagnosis of RA was relatively constant, as indicated by an apparently linear relationship between age at onset and the cumulative frequency of cases (r: .94) (figure 1).

A composite pedigree of 24 dogs with RA/RFUO is shown in figure 2 (dogs number 3, 4, 13 and 18 are not shown in the diagram because they did not have relatives in common with the other dogs within 4 generations). Seventeen of these 24 dogs could be regarded as belonging  to  3  familial  clusters.  These clusters were characterized by having a common ancestor within 3 generations. Cluster 1 included 5 dogs (numbers 1, 2, 5, 17 and 25). Cluster 2 included 9 dogs (numbers 7, 8, 14, 15, 19, 20, 22, 26 and 28).  Cluster 3 included 3 dogs (numbers 9, 11 and 12).  Three dogs showing at least RA, were born to parents known to be free of kidney failure/renal amyloidosis at age 5 (dogs number 10, 15 and 27).  Complete pedigree information was obtained from four sibling groups (clusters A, B, D and F) in which one parent was known to have RA or at least, kidney failure, and the other parent not affected at > 6 years of age (figure 2).  When the entire progeny from these groups was examined, the proportion of dogs with at least, RA, was: 18% (cluster A); 50% (cluster B); 25% (cluster D) and 50%  (cluster F). The comparison of the observed proportion (median of four littermate groups: 29%) of dogs with RA (number of affected/total) was significantly different from the proportion predicted (> 50%) by the hypothesis of simple dominant inheritance (P< .03, chi square). Dogs affected with at least RFUO had the same prevalence in those groups: 18% (cluster A), 50% (cluster B), 50% (Cluster D), 25% (cluster F).  Although inbreeding was observed in some animals with RFUO/RA (i.e., dog 7), inbreeding coefficients revealed no significant differences between affected and clinically normal CSP dogs.

Recurrent fever of unknown origin appeared to develop early in life. When both RFUO and RA occurred in the same individual dog, 2 years elapsed (median difference) between the expression of RUFO and onset of renal disease. All cases expressing RFUO alone began to show this condition prior to 2 years of age.


Inbreeding and close line breeding of stock are common practices among purebred dog breeders. Consequently, diseases due to recessive traits are more frequently observed in purebred dogs than in most human populations where inbreeding is usually avoided (Patterson et al. 1989)

The cross-sectional study indicated a predisposition of CSP for RA. CSP dogs developed RA at a greater frequency and at an earlier age than did non-CSP dogs. Unfortunately, there are no population based data to estimate true prevalence of this disease among CSP dogs. The lack of VMDP records on RFUO prevented us from assessing the predisposition of CSP for RFUO.

Prospective studies delineated some characteristics of the CSP predisposition for RA. An association between the Expression of RFUO and RA was suggested as 20 of 28 dogs had both conditions, and in the majority of these, RFUO preceded RA by 2 or more years.  Hence, RFUO in these CSP dogs is likely to be an early phenotypic expression of the genotype responsible for RA, as well as a clinical predictor for the progression to the more severe phenotypic expression to amyloid deposition in the kidney (RA). This is supported by the fact that dogs with RUFO alone (n=5) were young (< 2 years of age) when febrile episodes were first observed. However, there was no association between the number and severity of febrile episodes and the subsequent age of onset of RA (data not shown). Three dogs in which febrile episodes were not observed, experienced RA. This parallels what is described as phenotype II of FMF, which is speculated to occur as a result of different manifestations of a pleiotropic gene (Zemer et al. 1986). In addition, previous studies showed that dysregulation or serum IL-6 (a cytokine that induces the febrile response and serum amyloid A) was observed in CSP dogs afflicted with RFUO and/or RA (Rivas et al. 1992).  FMF patients have a similar association between the expression of RFUO and onset of renal Disease and RA (Sohar et al. 1967).

The hypothesis that RA at least, is an inherited disease in purebred CSP dogs was assessed in the prospective study.  Criteria suggestive of inherited disease include: 1) higher relative risk within a breed or family, and 2) a relatively constant range of age at onset of the disease.  Further, the expression of the condition in both sexes is a hallmark of autosomal traits, whereas the appearance of normal parents and siblings of affected animal characterizes recessive conditions (Patterson et al. 1989).  The finding that familiar clusters of RA were observed in CSP dogs reared in different geographical areas from the United States and Canada argues against environmental factors playing a significant role in expression of the condition. Amyloidosis was diagnosed in a relatively constant age range in this breed, which results in a statistically significant linear relationship between age at diagnosis and cumulative frequency of cases (Table 2 and Figure 2).  Diseases presented in a relatively constant age range are compatible with the hypothesis of pre-programmed activation of the molecular events leading to the expression of the condition (i.e., genetic diseases).  In contrast, diseases due to environmental factors alone usually result in non-linear relationships between age at diagnosis and cumulative frequency of cases (Glickman et al. 1987). Hence genetic, and not environmental factors, are more likely to be relevant in the development of this disease.

The expression of these conditions in dogs of both sexes was not compatible with the hypothesis of sex-linked inheritance.  Simple dominant inheritance of RA was not supported by these findings. When one of the parents was known to have one of these conditions, the observed prevalence of either RA or RFUO among littermates of 4 clusters or families was statistically different from the prevalence expected (50% or more) under the hypothesis of simple dominant inheritance. In addition, it was observed that dogs with RA were born to parents free of RA after five years of age.  RFUO was associated with RA (71% of the affected dogs had both RA and RUFO). However, the lack of data about RUFO in the cross- sectional study precludes the conclusion of a genetic predisposition to this condition.

The occasional absence of affected animals in consecutive generations is also compatible with the hypothesis of autosomal recessive inheritance (Nicholas 1987). Familial clusters of affected animals born to unaffected parents can be expected when a recessive gene is present in high frequency, as occurs with the founder effect.  When an ancestor which is a carrier of the trait is related to most animals (i.e. founder), there is a high frequency of carriers within the population (Diamond and Rotter 1987). Under such circumstances there is a high probability of matings between carriers of the trait, and the disorder may appear in consecutive generations.  This possibility could explain the expression of RA observed in two consecutive generations (numbers 8, 22, and 24). The occurrence of cases in two consecutive generations has also been reported in human families with RA (Rogers et al. 1989).  However, when a carrier is mated to a noncarrier, the disorder may be absent in one generation, but reappears in the next generation when carriers are again bred by chance.  This possibility is compatible with the observed expression of RA in most of the dogs under study.

In conclusion, the prospective studies support the hypothesis that RUFO and RA are associated signs of the same disorder in dogs of the CSP breed.  The combined data from cross-sectional and prospective studies suggest that the predisposition for RA is an inherited trait in CSP dogs.  All data to date are compatible with autosomal recessive inheritance.  However, mating/breeding experiments are necessary to distinguish polygenic from single-gene autosomal recessive inheritance. Furthermore, these studies lend support to the hypothesis that affected families of CSP dogs may represent a spontaneous model for renal amyloidosis and, possibly, FMF.


Barakat MH, Karnik AM, Majeed HWA, el-Sobki NI, and Fenech FF, 1986. Familial Mediterranean fever (recurrent hereditary polyserositis) in Arabs. Quart J Med 60:837-847.

Diamond JM and Rotter JI, 1987. Observing the founder effect in human evolution. Nature 329:105-106.

Di Bartola SP, Tarr MJ, Webb Dm, and Giger U, 1990 Familial renal amyloidosis in Chinese Shar-pei dogs. JAVMA 197(4):483-487

Everett BS, 1977. The analysis of contingency tables. Boston: Chapman Hall.

Glickman LT, Lee LA, and Cohen ND, 1987. Logarithmic normal distribution of incubation times for canine diseases (Sartwell’s model) AM J Vet Res 48(5):842-847.

Heller H, Sohar E, Kariv I, and Sherf L, 1955. Familial Mediterranean fever. Harefuah 45:91.

Matzner Y, 1989 Familial Mediterranean Fever. Isr J Med Sci 25:547-549.

Nicholas FW, 1987 Veterinary genetics. Oxford U.K.: Clarendon Press.

Ozdemir AI and Sokmen C, 1969. Familial Mediterranean fever among the Turkish people. Am J Gastroenterol 51:311-316.

Patterson DF, Aguirre GA, Fyfe JC, Giger U, Green PL, Haskins ME, Jerzyk PF, and Meyers-Wallen VN, 1989. Is this a genetic disease? J Small Anim Pract 30:127-139.

Rivas AL, Tintle L, Kimball E, Scarlett J and Quimby F, 1992. A canine febrile disorder associated with elevated interleukin-6 (IL-6). Clinic Immunol Immunopathol 64:35-45.

Rogers DB, Shohat M, Petersen GM, Bickal J, Congleton J, Schwabe AD, and Rotter J, 1989. Familial Mediterranean fever in Armenians: autosomal recessive inheritance with high gene frequency. Am J Med Genet 34:168-172.

Schwabe AD and Peters RS, 1974. Familial Mediterranean fever in Armenians: analysis of 100 cases. Medicine 53:453-462. 1974.

Sohar E, Gafni J, Pras M, and Heller H, 1967. Familial Mediterranean fever. Am J Med 235 43:277-253.

van Rijswijk MH and van Heusden CWCJ, 1979. The potassium permanganate method. Am J Path 97(1): 43-58.

Van Vleck LD, Pollak EJ, and Branford Oltenacu EA, 1987. Genetics for the animal sciences. New York: W. E. Freeman and Co.

Wright S, 1922.  Coefficients of inbreeding and relationship. Am Nat 56:330-338.

Zemer D, Pras M, Sohar E, Modan M Cabili S and Gafni J, 1986. Colchicine in the prevention and treatment of the amyloidosis of Familial Mediterranean fever. N Engl J Med 314(16):1001-1005.