Feline Panleukopenia Virus Is Not Associated With Myocarditis or Endomyocardial Restrictive Cardiomyopathy in Cats

Abstract

Canine parvovirus-2 (CPV-2) is nearly indistinguishable from feline panleukopenia virus (FPV) and is a well-known cause of viral myocarditis in young puppies; however, it is not known whether either FPV or CPV-2 naturally infects feline cardiomyocytes and causes myocarditis. Endomyocarditis (EMC) and left ventricular endomyocardial fibrosis (LVEF), clinically known as “endomyocardial restrictive cardiomyopathy,” are important feline heart diseases suspected to have an infectious etiology. A continuum is suggested with EMC representing the acute reaction to an unknown infectious agent and LVEF the chronic manifestation of repair. The purpose of this study was to determine (1) whether there is natural parvovirus infection of the feline myocardium and (2) whether parvoviral infection is associated with feline EMC and/or LVEF. In a retrospective study, polymerase chain reaction and sequencing for the parvovirus VP1/2 gene was performed on archived heart tissue from cats with endomyocardial disease and controls. Similar methods were used prospectively on myocardial tissues from shelter-source kittens. Although 8 of 36 (22%; 95% confidence interval [CI], 11%–40%) shelter kittens had parvoviral DNA in myocardial tissue, VP1/2 DNA was not detected in 33 adult cases or 34 controls (95% CI, 0% to ~11%). These findings were confirmed by in situ hybridization: adult cats did not have detectable parvovirus DNA, although rare intranuclear signal was confirmed in 7 of 8 shelter-source kittens. In kittens, parvovirus was not significantly associated with myocarditis, and in situ hybridization signal did not colocalize with inflammation. Although infection of cardiomyocytes was demonstrated in kittens, these data do not support a role for parvovirus in EMC-LVEF.

Although some types of feline heart disease have an underlying genetic (hypertrophic cardiomyopathy) or nutritional component (certain types of dilated cardiomyopathy), similar predisposing factors have not been observed for endomyocarditis (EMC) and left ventricular endomyocardial fibrosis (LVEF), conditions that are known clinically as “endomyocardial restrictive cardiomyopathy.”^13,14,20,31 Because of age relationships and shared lesion distribution, a continuum between EMC and LVEF is speculated, where EMC represents the acute myocardial reaction to a putative infectious agent and LVEF represents the chronic manifestation of subsequent myocardial repair.^14,31

EMC-LVEF is an important disease of cats. Previous retrospective studies from a university pathology service identified EMC-LVEF as the cause of death in >4% of cats compared with 2.3% and 2.2% for hypertrophic cardiomyopathy and dilated cardiomyopathy, respectively.³¹ Clinical findings are nonspecific and similar to other types of feline heart disease.⁸ The varying degree of inflammation and repair within EMC-LVEF lesions suggests an infectious etiology. EMC is characterized by neutrophilic, lymphocytic, or mixed inflammation of the endocardium and myocardium, with granulation tissue observed in some cases.^13,20 Interstitial pneumonia is frequent in cats with EMC.³¹ With LVEF, there is marked fibrosis of the left ventricular endocardium with lesser involvement of the myocardium.^13,19,31 Typically, fibrotic scar tissue spans the left ventricular apex and outflow tract involving both the left ventricular free wall and interventricular septum. Less commonly, there is diffuse endomyocardial fibrosis as well as variable cardiomyocyte degeneration, infarction, and inflammation. Biventricular involvement is rare.¹³ Previous studies have suggested a role for parvovirus in feline heart disease as DNA was detected within heart tissues of cats with cardiomyopathy.²³ Parvoviruses cause myocarditis in various species; however, the role of parvovirus in EMC-LVEF remains unknown.^6,7,15,21,28

The parvoviruses are small, single-stranded DNA viruses that are important causes of infectious disease in animals, especially in dogs and cats.^11,27 Feline panleukopenia virus (FPV) and canine parvovirus-2 (CPV-2) are very closely related and cause similar diseases, which can be particularly devastating in shelter situations.^17,27,30,33 CPV-2 was not described until the 1970s and is presumed to be derived from FPV or another closely related parvovirus.^17,30,33 Both dogs and cats present with signs associated with infection of rapidly dividing cells or without premonitory signs (subacute form).^22,24,27 Occasionally, there is generalized disease in neonates.²⁷ A notable difference between FPV and CPV-2 is the occurrence of necrotizing parvoviral myocarditis in puppies born to naive bitches and infected with CPV-2 within 2 weeks of birth or late in utero.^2,15,21,28 More rarely, myocarditis is observed in pups receiving inadequate colostrum.²¹ Puppies with parvoviral myocarditis typically present acutely with cardiac failure or sudden death at 3 to 8 weeks of age, but delayed presentation of cardiac signs is also recognized.^2,21 These infections are characterized by myocardial degeneration and necrosis, cardiomyocyte viral inclusions, and colocalization of viral antigen to cardiomyocytes. In natural and experimental parvoviral myocarditis, viral inclusions and immunostaining for viral antigen (in heart and other tissues) are temporally limited as detection of viral antigen dissipates after acute infection.^21,22 In neonatal puppies experimentally infected with CPV-2, viral inclusions are absent and immunostaining for viral antigen (in heart and other tissues) is negative in puppiess surviving to 8 weeks.²¹

Despite high frequency of exposure to parvoviruses, it is not known whether they naturally infect the feline heart and cause myocarditis. Previous retrospective studies amplified FPV DNA from archived sections of myocardium from cats with various types of cardiomyopathy (including restrictive cardiomyopathy) but not from cats having normal hearts.²³ Until recently, the significance of the association between parvovirus and feline cardiac disease has not been investigated further.¹⁸ Because the heart’s response to infection and injury is nonspecific, we hypothesized that early life parvoviral infection of the feline heart causes EMC and, subsequently, LVEF. The goal of this study was to determine if parvoviral detection was associated with feline endomyocardial disease by investigating the spectrum of EMC and LVEF.

Materials and Methods

Case Selection

Retrospectively, formalin-fixed, paraffin-embedded (FFPE) tissue from the New York Animal Health Diagnostic Center archives was used to evaluate the role of parvovirus in the development of EMC and LVEF. A search of the archives between June 2007 and November 2014 was performed, and cases limited to cats at least 1 year of age were identified using the keywords feline or cat and endomyocardial fibrosis, endocardial fibrosis, endocardial scar, endomyocarditis, or restrictive cardiomyopathy. We excluded cases having keywords hypertrophic and dilated. Control cases were identified using keywords describing acute trauma, neoplasia, or other noncardiac causes of sudden death. A similar age distribution of control cases was selected from the same time period and source.

To describe the prevalence of parvovirus in the myocardium of kittens, a prospective study was performed enrolling shelter-source kittens and juvenile cats up to 11 months old, with the further criterion: (1) euthanized due to a positive SNAP Parvo test (IDEXX, Westbrook, ME) and/or clinical signs consistent with FPV infection (including diarrhea, vomiting, inappetence, weight loss, dehydration, bone marrow suppression [secondary bacterial infections], or cerebellar ataxia) or (2) kittens dying suddenly without premonitory signs. Oversampling during disease outbreaks and sampling of multiple kittens from the same litter were avoided. A set of standardized myocardial tissues was saved in 10% neutral buffered formalin, including the right atrial and ventricular free wall, left atrial and ventricular free wall, and a section of septum including the septal leaflet of the right atrioventricular valve and a section of the left ventricular outflow tract.^13,31 Tissues were also frozen for molecular analysis.

Histopathology Evaluation

Histopathology was evaluated by a veterinary pathologist blinded to disease status. Sections of heart stained with hematoxylin and eosin (HE) were scored for each of myocardial infiltrating inflammatory cells (“myocarditis”), cardiomyocyte changes (“degeneration and/or necrosis”), and endomyocardial or myocardial fibrosis (“fibrosis”) on a scale (for each) of 0 to 3 (0 = none, 1 = mild, 2 = moderate, 3 = severe); a score ≥1 was considered to indicate the presence of type of pathology.

Polymerase Chain Reaction

For each archived case, two 15-μm sections were cut from FFPE tissue blocks. If additional organs were included in the tissue block, the noncardiac tissues were discarded. DNA was extracted using the QIAamp DNA FFPE Tissue kit (Qiagen, Valencia, CA), and the amount of nucleic acid obtained was quantified using a NanoDrop (Thermo Scientific, Wilmington, DE) spectrophotometer. Polymerase chain reaction (PCR) for the feline actin gamma 1 gene (ACTG1; Suppl. Table S1) was performed to evaluate the quality of the genomic DNA. Cases were excluded if there was poor amplification of this housekeeping gene. PCR was then performed to detect parvoviral VP1/2 DNA (Suppl. Tables S1 and S2). PCR primers amplifying a larger 638-nucleotide fragment of VP1/2 were used to detect parvovirus in fresh-frozen kitten tissues (Suppl. Table S1). Extracted DNA (100–400 ng) was combined with primers and PCR Master Mix (DreamTaq Green PCR Master Mix; Thermo Scientific, Foster City, CA). Following PCR, the sample reaction was gel-purified in 2% agarose.

Bands located at the expected amplicon length were extracted from the gel (MicroElute Gel Extraction Kit; Omega, Norcross, GA), sequenced by the Cornell University Biotechnology Resource Center, and then analyzed using an online BLAST search of GenBank (blast.ncbi.nlm.nih.gov/). For all PCR experiments, appropriate controls were performed, including no template reaction and purified FPV plasmid DNA as a positive control (gift of C. Parrish).

In Situ Hybridization

In situ hybridization (ISH) was performed on all archived cases (n = 7; 4 EMC-LVEF cases and 3 controls) that had an amplicon <200 bp (using PCR primers for parvoviral VP1/2 DNA) that did not match parvovirus sequences in GenBank. ISH also was performed on all shelter-source kittens in which sequencing of the PCR amplicon matched to that of a known parvovirus (n = 8). ISH was performed using >20 double “Z” oligo pair probes designed to hybridize with CPV-2/FPV DNA spanning VP1/2 (nucleotides 2379–3712) using a commercially available kit (ACD RNAscope, Hayward, CA).³⁴ In brief, slides were deparaffinized followed by target retrieval in boiling Target Retrieval Reagent (ACD RNAscope). Endogenous peroxidases were blocked with hydrogen peroxide followed by protease digestion (Protease Plus; ACD RNAscope) in a humidity-controlled oven (HybEZ Oven; ACD RNAscope). Probe hybridization, signal amplification, and DAB staining were performed according to the manufacturer’s instructions. Appropriate controls were present in every run, including feline intestinal tissue with FPV infection (confirmed by immunohistochemistry and virus isolation) and feline myocardial tissue probed with a Negative Control Probe-DapB (ACD RNAscope).

Immunohistochemistry

Immunohistochemistry (IHC) was performed on all shelter-source kittens in which sequencing of the PCR amplicon matched to that of a known parvovirus (n = 8) by the New York Animal Health Diagnostic Center Histopathology laboratory using a rabbit polyclonal antibody (provided by C. Parrish). Feline intestinal tissue with confirmed FPV enteritis and sections incubated without primary antibody served as controls.

Statistical Analysis

Agent prevalences between cases and controls were compared with two 2-sided Fisher exact tests; the agent prevalences in subsets of animals were described using the Wilson score, continuity-corrected confidence intervals (CIs), or, if those could not be calculated because of totals of zero, Fisher exact CIs. The overall α was 5%. Bonferroni adjustments were used to correct for multiplicity for the primary comparisons of agent prevalences. Association between myocardial lesions and the presence of myocardial parvoviral DNA in shelter-source kittens was similarly compared and described with the same procedures and α = 5%. Statistical software programs were used.^1,32

Results

Case Demographics

A search of the New York State Animal Health Diagnostic Center archives identified 1056 feline necropsy cases from the search period between June 2007 and November 2014. Thirty-eight feline necropsy cases (3.60%; 95% CI, 2.6%–4.9%) were identified using the database search terms for EMC-LVEF. A cohort of 33 EMC-LVEF (Suppl. Table S3) cases was identified based on tissue availability and amplification of a housekeeping gene. The median (minimum, maximum) age of the 28 EMC-LVEF cats with a recorded age was 4.8 (1, 17) years. There were 13 castrated males, 5 intact males, 8 spayed females, 6 intact females, and 1 cat of unknown sex. Most cats were domestic shorthairs (Suppl. Table S3). The recorded causes of death of the EMC-LVEF group are summarized in Table 1.

Table 1.

Recorded Causes of Death of Adult Cats With EMC-LVEF and of Control Adult Cats.

Cause of Death	EMC-LVEF (n = 33), No. (% Total; 95% CI)	Control (n = 34), No. (% Total; 95% CI)
Myocarditis	11 (33; 19–52)	0
Restrictive/fibrosing cardiomyopathy	8 (24; 12–43)	0
Unspecified cardiomyopathy	2 (6; 2–20)	0
Congestive heart failure	1 (3; 0.2–18)	0
Noncardiac neoplasia	2 (6; 2–20)	6 (18; 7–35)
Trauma	3 (9; 2–25)	19 (56; 38–72)
Unknown	2 (6; 2–20)	0
Cardiac necrosis	1 (3; 0.2–18)	0
Gastrointestinal foreign bodies	0	3 (9; 2–25)
Other noncardiac conditions	3 (9; 2–25)	6 (18; 7–35)

CI, confidence interval; EMC-LEVF, endomyocarditis–left ventricular endocardial fibrosis.

Thirty-four control cases of similar ages to the EMC-LVEF cases were selected (Suppl. Table S3). The median (minimum, maximum) age of the 34 control cats with a recorded age was 5 (1, 16) years. There were 12 castrated males, 4 intact males, 10 spayed females, 7 intact females, and 1 of unknown sex. The recorded causes of death of controls are presented in Table 1.

We analyzed tissues from 36 shelter-source kittens (Suppl. Table S5). The median (minimum, maximum) age of the kittens was 5 (1, 43.5) weeks. Sexes included 1 castrated male, 19 intact males, 15 intact females, and 1 of unknown sex. Represented breeds included 29 domestic shorthairs, 4 domestic mediumhairs, 2 domestic longhairs, and 1 unknown.

Histopathology Evaluation: EMC-LVEF Cases and Controls

The myocardium from the cohort of 33 EMC-LVEF cats had variable histologic lesions of inflammation, degeneration, and fibrosis (Suppl. Table S4, summarized in Table 2). Evidence of myocarditis (score ≥1) was present in 16 of 33 EMC-LVEF cases. Six of the 16 cats with myocarditis (38%; 95% CI, 16%–64%) had moderate to severe myocarditis (Fig. 1). For the 5 cats with moderate to severe myocarditis with age data available, the median age (minimum, maximum) was 2.2 (1, 6) years. Half of the cases (8/16) had both endocardial and myocardial inflammation, 4 of 16 had inflammation limited to the endocardium, and 4 of 16 had primarily myocardial inflammation.

Table 2.

Histopathologic Findings in 33 Adult Cats With EMC-LVEF.

	Myocarditis (Score ≥ 1)	Fibrosis	Cardiomyocyte Degeneration	Hybrid
No. (%; 95% CI)	16 (48; 31–66)	20 (61; 42–77)	19 (58; 39–74)	5 (21; 7–31)
Distribution, No.
Endocardial	4	2	4	5
Endomyocardial	8	18	13	0
Myocardial	4	0	2	0

CI, confidence interval; EMC-LEVF, endomyocarditis–left ventricular endocardial fibrosis.

Figures 1–4.

Heart, cat. Figure 1. Necrotizing myocarditis; case No. C7. The endomyocardium of the left papillary muscle is disrupted and expanded by lymphocytes, macrophages, neutrophils, and plasma cells mixed with fibrin and eosinophilic necrotic debris. Hematoxylin and eosin (HE). Figure 2. Left ventricular endocardial fibrosis; case No. C8. The endocardium is markedly thickened by a layer of connective tissue that is >1 mm thick (arrowheads). Inset: The thickened endocardium has areas of matrix alteration that individualizes cells (chondroid metaplasia). HE. Figure 3. Left ventricular endocardial fibrosis; case No. C8. The endocardium is markedly thickened by collagenous connective tissue. Masson’s trichrome. Figure 4. Hybrid endomyocarditis and endocardial fibrosis; case No. C32. The endocardium is expanded by variable loose fibrous proliferation accompanied by macrophages and lymphocytes and lined by plump activated endothelial cells (inset). The endothelial lining is focally disrupted and covered by a mat of amorphous eosinophilic fibrin. HE.

Endomyocardial inflammatory infiltrate varied in adult EMC-LVEF cases (Fig. 1). Many cases had lymphocytic and histiocytic infiltrates (5/16); 3 of 16 had histiocytic, neutrophilic, and necrotizing inflammation; 3 of 16 had lymphocytic, histiocytic, and necrotizing inflammation; 2 of 16 had histiocytic inflammation; and 1 case each had neutrophilic (1/16) or lymphocytic inflammation (1/16).

Twenty cases (Table 2, Suppl. Table S4) showed evidence of endocardial and/or myocardial fibrosis (Figs. 2, 3). Of these, the 18 with age data had a median (minimum, maximum) age of 6 (1, 17) years. Twelve of these 20 cases (36%) had moderate to severe endocardial fibrosis (Figs. 2, 3); these cats had a median age (minimum, maximum) of 9 (2, 17) years. One case with marked endocardial fibrosis included areas of chondroid metaplasia (Fig. 2). A different case demonstrated severe myocardial fibrosis lacking involvement of endocardium.

Five cases (Table 2) were considered to have hybrid lesions having both mild to severe inflammation and mild to severe fibrosis with endocardial involvement (Fig. 4). Of the 5 cases with hybrid lesions, 2 had both moderate to severe myocarditis and fibrosis (Fig. 4). Cardiomyocyte changes of mild degeneration and necrosis were present in 19 of 33 cases and usually were associated with myocardial inflammation (Table 2, Suppl. Table S4).

Of the adult control cats, 5 of 34 (15%; 95% CI, 6%–32%) had histologic lesions of scattered minimal cardiomyocyte changes or focal degeneration in the papillary muscle, although none had myocarditis. Evidence of mild multifocal myocardial (nonendocardial) fibrosis was present in 3 of 34 control cases (9%; 95% CI, 2%–25%).

Histopathology Evaluation: Shelter-Source Kittens

Myocarditis was present in 5 of 36 of the shelter-source kittens (14%; 95% CI, 5%–30%; Suppl. Table S5). The median (minimum, maximum) age of kittens with myocarditis was 8.5 (4, 26) weeks; the median age of kittens without myocarditis was 5 (1, 44) weeks. Most affected kittens had moderate myocarditis (4/5), and the fifth kitten had mild myocarditis. Multifocal histiocytic, lymphocytic, and necrotizing myocarditis (Fig. 5) was observed in all affected kittens, and associated fibrosis was not observed.

Figures 5–6.

Heart, kitten. Figure 5. Myocarditis, case No. K32. There are multifocal to coalescing aggregates of lymphocytes, macrophages, and plasma cells within the myocardium. Inset: Higher magnification. Hematoxylin and eosin. Figure 6. Case No. K23. Labeling of cardiomyocyte nuclei (arrow) for parvoviral nucleic acid in the heart of a kitten without myocarditis. In situ hybridization.

PCR

Amplicons <200 bp were detected in 4 of 33 adult cat EMC-LVEF cases and 3 of 34 controls by PCR testing of heart tissue using parvoviral primers. However, the sequence of these amplicons did not match FPV or CPV-2 sequences in GenBank.

Amplicons were detected in 8 of 36 shelter-source kittens (22%; 95% CI, 11%–40%; Suppl. Table S5) by PCR testing of heart tissue using parvoviral primers. All 8 sequences matched FPV capsid protein gene sequences in GenBank (Suppl. Table S6). The median (minimum, maximum) age of sequence-positive kittens was 7.5 (1, 26) weeks, while that of sequence-negative kittens was 5 (1, 43) weeks. Six sequence-positive kittens had been vaccinated for feline parvovirus within 2 weeks of death (average 8.2 α 3.4 days). Five amplicon sequences matched Felocell virus strain (GenBank EU498681.1), FPV virus strain 97/06-1 (EU498714.1), and FPV strain 41/02 (EU498683.1) capsid protein genes with high identity. One kitten (K33) having a similar high identity sequence had not been vaccinated. For the other kittens in which PCR amplicons were detected using primers for parvoviral VP1/2 DNA (K11, K21, K24), the sequences of the amplified products did not match vaccine sequences in GenBank, and the top matches are given in Supplemental Table S6.

Detection of Parvovirus in the Myocardium

ISH for CPV-2/FPV DNA was performed on 4 adult cat EMC-LVEF cases and 3 controls in which a PCR amplicon was detected (although the sequence of the amplicon did not match parvovirus sequences in GenBank). Corresponding to the PCR results, parvoviral DNA signal was not detected by ISH in myocardial tissue.

ISH signal for parvovirus was detected in the myocardium of 7 of 8 (88%; CI, 47%–99%) shelter-source kittens that were positive for parvovirus by PCR (Suppl. Table S5). The labeling was sparse and randomly distributed, with punctate signal localized to the nucleus of cardiomyocytes and occasionally within endothelial cells and interstitial inflammatory cells (Fig. 6). Similarly, by IHC, there was rare punctate immunoreactivity localized to the nucleus of cardiomyocytes (Suppl. Fig. S1) in all 8 kittens positive for parvovirus by PCR (Suppl. Table S5).

Statistical Analysis

There was no statistical difference between the EMC-LVEF cats and the control cats in detection of myocardial FPV (P > .99; Suppl. Table S7) or CPV-2 DNA (P > .99; Suppl. Table S7).

In the shelter-source kittens, there was no significant association between myocardial lesions and the presence of myocardial parvoviral DNA as detected by PCR (P = .30; Suppl. Table S7).

Discussion

We hypothesized that early life parvoviral infection of the feline heart causes EMC and, subsequently, LVEF. In contrast with a previous retrospective study documenting FPV DNA in FFPE cardiac tissue of cats with idiopathic hypertrophic, dilated, and restrictive cardiomyopathy but not in control cats, our study focused on cats with endomyocarditis and endomyocardial fibrosis.²³ Similar to recent studies screening for viral genomes in a subset of adult cats with fully developed endomyocardial fibrosis, we did not identify myocardial parvoviral DNA in adult cats and furthermore did not identify parvovirus-associated myocarditis in shelter-source kittens by PCR or ISH.¹⁸ Our results do not support the hypothesis that parvovirus is associated with EMC-LVEF/endomyocardial restrictive cardiomyopathy.

PCR-based approaches are limited by the difficultly of establishing causation, particularly in suspected cases of myocarditis. Despite a propensity for parvoviral DNA persistence in tissues, it was uniformly absent in the myocardium of adult cats with histologic evidence of EMC or endomyocardial fibrosis; our sample sizes were adequate to rule out prevalences >~11% (detection of parvoviral DNA in myocardium of adult cats in previous studies) with 95% confidence.²³ In a variety of species, including the cat, lengthy persistence of parvoviral DNA in tissues is recognized.^3,16,25 After virus inactivation by the host immune response, residual parvoviral DNA persists in long-lived cells that have limited turnover, including the cardiomyocyte.^4,16,25 The finding of viral DNA in tissue is not necessarily associated with disease of that tissue; however, it represents an archive of exposure of the individual to certain infectious agents (the Bioportfolio), particularly parvoviruses.²⁵ In our study, there was no difference in the presence of myocardial parvoviral DNA between EMC-LVEF and control groups.

We observed myocarditis in 14% of our shelter-source kittens that had been euthanized or died. We evaluated the heart tissue of these shelter-source kittens for parvovirus as the continuation of cardiac development into the early postnatal period may correspond to a window of cardiac parvoviral susceptibility similar to neonatal puppies.²¹ Parvoviruses infect cells during the S phase of the cell cycle via transferrin receptors.^26,33 In newborn mice, there is an initial high-frequency (>50%) DNA synthesis in cardiomyocytes with subsequent age-related decline; furthermore, postnatal cardiomyocytes are known to undergo DNA replication without completion of the cell cycle.²⁹ Parvoviral DNA was amplified from the myocardium of shelter-source kittens; however, the detection of parvovirus was not associated with myocardial lesions. Parvoviral signal (ISH > IHC) was sparse and primarily present in the nuclei of cardiomyocytes and not restricted to areas of myocardial inflammation in shelter-source kittens. Autolysis or prolonged fixation may have contributed to the lack of ISH signal in one kitten positive for parvovirus by PCR.

Most parvovirus PCR+ and ISH+ kittens had a history of parvoviral vaccination within 2 weeks of death; however, not all of amplicon sequences matched parvovirus vaccine sequences, and parvovirus was also detected in the hearts of unvaccinated kittens. All FPV strain sequences share high sequence identity; current vaccine strains share highest identity with historical FPV strains. Differences in the Felocell VP2 sequence from most of the recent field FPV strains occur at nucleotides 693 and 1684.¹² The extent to which attenuated live vaccine virus can be detected by PCR after vaccination is not well characterized, although vaccination 5 to 12 days prior can be detected by available tests.³³ Nonetheless, our data suggest that parvovirus can infect feline cardiomyocytes but that infection is not associated with necrotizing myocarditis resembling CPV-2 infection of puppies. This is consistent with previous reports of intranuclear inclusions and ultrastructural evidence of viral particles in cardiomyocytes of 13-day-old kittens with disseminated parvovirus infection and the occurrence of inclusions in subendocardial and left ventricular cardiomyocytes in neonatal kittens experimentally infected with FPV.^5,9,10

In cats as in humans, restrictive cardiomyopathy is a non-specific classification of idiopathic fibrotic myocardial disease that manifests clinically as diastolic dysfunction with relative preservation of systolic function. Unfortunately, recognition of the entity is complicated by vague clinical signs, heterogeneity in the pathologic phenotypes, nonspecific terminology, and inadequate sampling.^8,31 Our retrospective study focused on the pathologic phenotype of endomyocardial fibrosis, characterized by a thick layer of fibrosis expanding the left ventricular endocardium, and endomyocardial inflammation. While characterization of included cases was likely limited by the available nonstandardized archived samples, our series included 12 cases with moderate to severe endocardial fibrosis, 12 cases with endomyocarditis or endocarditis-myocarditis, and 5 cases with hybrid lesions. In our study, age distributions of EMC and LVEF were consistent with previous descriptions.³¹ Cats with moderate to severe endomyocarditis (median, 2.2 years) were much younger (by inspection) relative to cats with moderate to severe endocardial fibrosis (median, 6 years).

Our study did not identify PV DNA by PCR and sequencing within cardiac tissue of adult cats with histologic evidence of EMC-LVEF endomyocardial disease, and this result was confirmed by ISH. Although PV infection of cardiomyocytes was demonstrated in kittens, PV was not associated with myocarditis or myocardial damage. EMC-LVEF may be caused by an infectious etiology; however, we conclude there is no association between EMC-LVEF and FPV or CPV-2 infection.