Valvular endocarditis in the horse: 20 cases (1993–2020)

Abstract

Medical records of 20 horses with a confirmed diagnosis of valvular endocarditis at the Ontario Veterinary College between January 1, 1993 and February 3, 2020 were reviewed. The diagnosis was based on physical examination findings, complete blood (cell) count (CBC), serum biochemistry, echocardiography, blood culture, and post-mortem findings. Common presenting signs included tachycardia, pyrexia, weight loss, lameness/joint distension, and a heart murmur. Clinicopathological findings included leukocytosis, anemia, hypoalbuminemia, hyperglobulinemia, and elevated inflammatory markers. Culture from 5 horses yielded Actinobacillus equuli in 2 cases and Actinobacillus suis in 1 case. Of the 20 horses included in this study, 17 were euthanized and 3 were treated. Only 1 case had follow-up more than 1 year after discharge.

Introduction

Valvular endocarditis (VE) is an uncommon, life-threatening cardiac disorder in horses (1–3). This condition results from a microbial infection of the endothelial surface of the heart characterized by bacterial or fungal invasion of the valves (valvular endocarditis) or wall (mural endocarditis). Infection results in fibrinous clots or vegetations which impede normal cardiac function. A variety of bacterial species have been isolated from blood or valvular lesions; however, Actinobacillus equuli, Streptococcus spp., and Staphylococcus spp. are the most common bacterial agents isolated in horses (4). The aortic and mitral valves are most commonly affected, followed by the tricuspid and pulmonary artery valves (5,6). Acute and chronic clinical manifestations have been described (7). Chronic endocarditis occurs when a previously damaged valve provides a portal of entry, while in the acute form, bacteria adhere to undamaged valvular tissue (3,6,8). The inciting cause or associated risk factors in these cases are rarely determined; however, septic jugular vein thrombophlebitis is reported in most cases of tricuspid valve bacterial endocarditis (1,8–10). In cattle, foot abscesses, rumenitis, reticular abscess, or other septic processes can result in sustained or recurrent bacteremia, predisposing the animal to the development of bacterial endocarditis (5).

In humans, there is an increased prevalence of endocarditis with prosthetic heart valves or hospital acquired infections (11), skin or lung infections, dental surgery (i.e., extraction), and IV drug abuse (11). Musculoskeletal signs are a common presenting complaint in human cases of valvular endocarditis, with 25% to 44% of cases listing joint or limb pain as the reason they presented for health care (12).

No breed or sex predilection has been reported in horses; however, young animals appear to be more commonly affected (1,2,13). Clinical signs of VE are usually non-specific. Commonly described signs include persistent pyrexia (> 38°C), chronic inflammation, detection of a heart murmur(s), and sepsis or septic emboli of unknown origin (2,14). In addition, tachycardia, lameness, or joint effusion, and weight loss has been reported (1,14). Intrinsic heart disease, such as cardiac arrhythmias are less frequent (2). The diagnostic work-up includes history, clinical signs, complete blood (cell) count (CBC), serum biochemistry profile, blood culture, echocardiography, and/or post-mortem examination (1–3). Typical clinicopathological abnormalities include leukocytosis with a mature neutrophilia, hypoalbuminemia, hyperglobulinemia, hyperfibrinogenemia, and non-regenerative anemia (2). Treatment consists of long-term broad-spectrum antimicrobials or selected antimicrobial therapy based on bacteriological culture and sensitivity (3,15,16). Response to treatment can be monitored through repeated CBCs, serological markers of inflammation, and echocardiography. The prognosis is guarded to poor, particularly when there is evidence of valvular insufficiency (1). Prompt diagnosis is challenging due to the insidious clinical signs (2,8).

The objective of this case series was to summarize the clinical and laboratory findings in 20 confirmed cases of VE.

Materials and methods

Case selection criteria

Medical records from all horses that were admitted to the Large Animal Veterinary Teaching Hospital, Ontario Veterinary College between January 1, 1993 and February 3, 2020 with a final diagnosis of VE were reviewed. All horses included in the study had a suspected diagnosis of VE based on history, physical examination, and clinicopathologic evidence of inflammatory disease. Confirmation was based on echocardiographic and post-mortem findings. The following variables were extracted from the medical records: signalment, presenting complaint, and clinical signs (rectal temperature, respiratory rate, heart rate, and auscultation of heart and lungs). Detailed description of the heart murmurs (intensity grade, location, timing during the cardiac cycle, shape, and tonal quality) were correlated with echocardiographic findings. The presence or absence of any lameness or joint effusion, and weight loss were also included in the database.

Data analysis

Mean ± standard deviation (SD), median and range values were calculated for age, hematocrit, white blood cell count, total protein, albumin, globulin, serum amyloid A, and plasma fibrinogen. Blood culture was available for 5 cases and postmortem findings were available for 17 of the 20 cases. Short-term survival data were available for 18/20 cases.

Results

Twenty horses with a confirmed diagnosis of VE were included in this case series; 12 were males (6 stallions; 6 geldings), and 8 were female. The median age at presentation was 3.8 y (mean: 7.3 y; range: 0.9 to 26.5 y). The breeds affected were Thoroughbreds 8/20; Standardbreds 4/20; Quarter Horses 5/20 and 1 each of grade, Trakehner, and miniature. Eighteen horses had 1 heart valve affected, while 2 horses had 2 valves affected. Of the 18 with 1 valve affected, 7 were aortic (35%), 6 were tricuspid (30%), 4 were mitral (20%), and 1 was pulmonic (5%). Of the 2 horses with 2 valves affected, the combinations of valves affected were aortic + mitral, and aortic + tricuspid.

The most common presenting clinical signs were pyrexia of unknown origin (13/20), weight loss (11/20), and presence of a heart murmur (11/20). The prevalence of tachycardia, tachypnea, and fever amongst the total case population and individual valves are shown in Table 1. Heart murmurs were present in 17/20 cases and were graded between 3 to 6/6. There was no association between the valve affected and the presence of a murmur. On a scale of 0 to 6, the cardiac murmurs were graded between 3 and 6 for horses with aortic VE, between 3 and 4, for horses with mitral and tricuspid VE, and 2 for the pulmonary VE case. A summary of the grade and side of auscultation of the murmur according to the affected valve is shown in Table 2.

Table 1.

Affected valves and clinical parameters at presentation of horses with valvular endocarditis.

Valve affected	Heart rate > 40 bpm	Respiratory rate > 24 breaths/min	Temperature (> 39°C)
Tricuspid	5/6	5/6	5/6
Aortic	3/7a	3/7	4/7
Mitral	3/4	3/4	3/4
Pulmonic	1/1	1/1	1/1
Aortic + tricuspid	1/1b	1/1	0/1
Aortic + mitral	0/1	0/1	0/1
Total cases affected	13/20	13/20	13/20

^aOne horse with aortic endocarditis was presented with ventricular tachycardia.

^bHorse presented with atrial fibrillation. bpm — beats/min.

Table 2.

Clinical (grade and site of auscultated murmur) and location of infarcted organs on post-mortem examination of horses with valvular endocarditis.

Valve affected	Murmur grade (0–6)	Side of auscultated murmur	Site of infarct
Aortic	4/6	L	Renal
	5/6	R	Cardiac, renal, GIT
	5/6	L + R	—
	6/6	L+ R	—
	3/6	L	Cardiac, renal
	4/6	L	Renal
	4/6	L + R	Renal, GIT
Mitral	4/6	L	—
	—	—	Renal
	3/6	L	—
	4/6	L	—
Tricuspid	3/6	R	—
	3/6	R	—
	—	—	—
	4/6	R	Renal, lung
	3/6	R	Lung, GIT
	—	—	Renal, limb, nasal conchae
Pulmonic	2/6	L	—
Aortic + Tricuspid	3/6 L	L + R	—
	5/6 R
Aortic + Mitral	6/6	L	Renal, limb, cardiac, liver, CNS

L — left; R — right; GIT — gastrointestinal tract; CNS — central nervous system; — no data available.

Hyperfibrinogenemia (> 3 g/L) occurred in 11/13 (85%) cases, leukocytosis (10 to 20 × 10⁹/L) in 12/18 (66%) and leukopenia (2.4 × 10⁹/L) in 1 case. An albumin/globulin ratio < 1.0 was measured in 10/12 (83%) cases and anemia (hematocrit < 0.29 L/L) was present in 7/17 cases (41%). Serum amyloid A (SAA) values were available for 3/20 cases and were markedly elevated in all 3 (> 2000 mg/L).

Blood culture results were available for 5/20 cases. Actinobacillus equuli (2/5), Actinobacillus suis (1/5), Enterococcus casseliflavus (1/5) and a Staphylococcus spp. (1/5) were isolated. There was a discrepancy in 2/5 cases when comparing pre-mortem blood culture results to post-mortem culture of the endocardium.

Echocardiographic findings were available for 14 of the 20 cases. Evidence of ventricular dilation (8/14) and abnormal fractional shortening (9/14) were the most common findings. Valvular lesions ranged from valves/chordae tendineae thickening to small masses on the valve leaflets to large pedunculated masses extending from the mass into a major vessel (aorta). Valvular insufficiency with turbulent blood flow/jet formation was recorded in all cases (Figure 1A). Effusion (pleural/peritoneal) was noted in 7 cases (3 aortic, 1 mitral, 1 tricuspid, 1 pulmonary and the mixed aortic/tricuspid case). Two horses were presented in atrial fibrillation (mitral and the aortic/tricuspid valves) and 1 horse with aortic VE was presented as a cardiac emergency with ventricular tachycardia.

Figure 1.

A — Ultrasound image from the right parasternal, long-axis view at the 5th intercostal space. The arrowhead highlights a markedly thickened aortic valve cusp due to vegetative endocarditis. B and C — Gross appearance of the affected valves at necropsy. Note the markedly thickened valve cusps (white arrowheads) with vegetative lesions typical of valvular endocarditis.

Only 3 of the 20 horses survived and were discharged from the hospital, the other 17 were euthanized. Short-term survival data were available for only 1 horse, which survived to a minimum of 1 year after discharge (20-year-old Quarter Horse mare with aortic VE). Two horses were discharged from the hospital but were lost to follow-up. Euthanasia was recommended in both of the latter cases based on clinicopathological and echocardiographic findings.

Post-mortem examination was performed in 17/20 cases (7 females, 5 intact males, and 5 geldings). Multifocal infarcts were present in 13/17 cases. Aortic (5/6) and tricuspid (4/6) VE cases were most likely to have multiorgan infarctions at post-mortem examination (Table 2). One horse with aortic and mitral VE had renal infarcts and infarcts were also present within the cardiac parenchyma and gastrointestinal system. Horses with tricuspid VE had lesions in the lung parenchyma with evidence of embolic disease also reported in the kidney, gastrointestinal system, and distal limbs. The horse with pulmonic VE had no evidence of embolic disease. The most severe case of embolic disease was found in the horse with mitral/aortic VE; infarcts were present in the kidney, left and right hind limbs, cardiac parenchyma, liver, and brain. Postmortem findings were consistent with the echocardiographic findings including valvular thickening and pedunculated mass formation leading to valvular dysfunction (Figure 1B, C).

Discussion

This retrospective study compiled the clinical, clinicopathological, diagnostic imaging, and post-mortem investigation data for horses with confirmed valvular endocarditis. As previously reported, these cases may present a diagnostic challenge for the equine clinician, and the data presented here may add additional information to assist in identifying such cases. Valvular endocarditis was commonly found in young adult horses, which is similar to the findings in previous studies (1,2,13). The reason for this observation is unknown but an increased occurrence of bacteremia in younger animals, in addition to lower immune function and more frequent worm infestation as predisposing factors maybe a plausible explanation (2). Other potential risk factors may include the common practice of intravenous injections in racehorses, and high level of stress potentially affecting the immune system function; however, a potential bias in this observation due to other factors not accounted for must be considered. Differences in sex distribution in horses with VE have been reported, and similarly, there were slightly more female horses in this case series (1,13). Human and canine studies have demonstrated a strong sex correlation, with males more likely (2.5×) to develop VE (11,17). In horses, the role of sex as a risk factor in the development of equine VE remains to be determined.

The distribution of valvular lesions showed a predilection towards the left-sided heart valves. This is in agreement with previous studies, which also reported a higher incidence of endocardial lesions on the left side (mitral and aortic valves), suggesting that the increased incidence of left-sided VE may be related to the pressure differences between left and right chambers (16,18). The aortic valve was the most commonly affected (7/20) in the present study, which is similar to previous studies (1,13). The pulmonic valve was the least affected site (1/20) and this is also in agreement with previous reports (5).

Clinical signs of bacteremia in horses include recurrent fever, tachypnea, tachycardia, anorexia, and weight loss (1,19). Septic emboli associated with VE may cause signs depending on which body systems are affected. Signs reported in the literature include lameness, joint distension, coughing, pneumonia, hematuria, pyuria, mastitis, petechiae, peripheral edema, diarrhea, jugular vein thrombosis, colic, poor growth, laminitis, and more generally, any chronic disease with an associated murmur (1,14). The most common presenting complaints in this case series were pyrexia of unknown origin, weight loss, and lameness (or joint distension), which are commonly reported in horses with VE.

Signs of VE are variable and may be correlated with the side of the heart affected leading to septic emboli being shed to the body (left-sided lesions) or lungs (right-sided lesions) (6). Endocarditis of the left heart valves often manifests as arthritis, osteomyelitis, vasculitis, or nephrosis, whereas endocarditis of the right heart valves leads to pulmonary thromboembolism, lung infarction, or lung abscesses (14). Shifting limb lameness is a common finding in horses and dogs with bacterial endocarditis (19). Myocardial infarcts are more likely in horses with left-sided endocarditis and may contribute to the development of congestive heart failure (1). In this study, 2 horses with left-sided valvular endocarditis had decompensated heart failure and were presented with primarily cardiac clinical signs. Murmurs were identified in most of these cases and ranged from grade 3/6 to 6/6. Murmurs may be absent in some VE cases, however, in particular in those with right-sided endocarditis because of the lower blood pressure compared with the left-side and hence, slower blood flow and less severe turbulence (3,20).

The wide range of systemic manifestations of VE often occurs secondary to hematogenous septic emboli from the endocarditis lesion or production of antibodies causing immune-mediated synovitis/arthritis (3,21). Causative organisms within the valvular lesion are protected from the host immune system by fibrin and platelet clumps, which cover the site of infection. The combination of these factors, in addition to proliferative granulation tissue, ultimately results in valvular dysfunction (3).

Changes in CBC, serum biochemical parameters and inflammatory markers (fibrinogen and SAA) are commonly found in valvular endocarditis cases. Serum amyloid A is a sensitive acute-phase protein in the horse (22) and has been established as a valuable inflammatory biomarker in bovine endocarditis (23). However, the usefulness of SAA as a marker of chronic inflammation is unknown for horses and serial testing of SAA may provide more accurate information regarding response to treatment and prognosis (24).

The clinical use of cardiac-specific biomarkers has been investigated in the horse. Cardiac troponin I (cTnI) has been established as a sensitive and specific biomarker of myocardial injury in human medicine and is used to provide diagnostic and prognostic information in patients with primary and secondary cardiac disease (25). A reference range for cTn1 in clinically normal adult Thoroughbreds has been established (26). Because cTn1 has a short half-life, it can provide valuable “real time” information when interpreted alongside echocardiographic findings and other serum biochemical values (25,27). Recently, a direct comparison of cTnI and a high-sensitive cardiac troponin T (hs-cTnT) assay has been conducted in horses to aid in detecting and quantifying myocardial damage (28). cTnT has been validated in human medicine as an accurate and preferred biomarker of cardiac myocardial damage (29). While further investigation is required in equine patients, cTnT has been validated for equine patients with standardized reference ranges and kinetics established (30,31). cTnT may provide a more sensitive alternative assay to aid in determining the extent of disease, particularly in horses presenting with pyrexia of unknown origin without a murmur on auscultation (28).

Valvular endocarditis is often difficult to diagnose in the early stages of the disease. The vague and nonspecific clinical signs may persist without warranting an intensive cardiac investigation. The information presented in this case series provides equine practitioners with clinical, laboratory, and diagnostic imaging data that may be useful while investigating horses with a similar clinical presentation.