Cardiac Troponin I Concentrations in Ponies Challenged with Equine Influenza Virus
The animal experiments were conducted at Fort Collins, CSU, and the cTnI assays were conducted at the University of Pennsylvania, Kennett Square, PA. This research was presented as an abstract at ACVIM Forum, San Antonio, TX, June 2008.
Corresponding author: Mary M. Durando, Department of Clinical Studies, New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, Kennett Square, PA 19348; e-mail: firstname.lastname@example.org.
Background: Myocarditis is thought to occur secondary to equine influenza virus (EIV) infections in horses, but there is a lack of published evidence.
Hypothesis/Objectives: We proposed that EIV challenge infection in ponies would cause myocardial damage, detectable by increases in plasma cardiac troponin I (cTnI) concentrations.
Animals: Twenty-nine influenza-naïve yearling ponies: 23 were part of an influenza vaccine study (11 unvaccinated and 12 vaccinated), and were challenged with 108 EID50 EIV A/eq/Kentucky/91 6 months after vaccination. Six age-matched healthy and unvaccinated ponies concurrently housed in a separate facility not exposed to influenza served as controls.
Methods: Heparinized blood was collected before and over 28 days after infection and cTnI determined. Repeated measures analysis of variance, chi-square, or clustered regression analyses were used to identify relationships between each group and cTnI.
Results: All EIV-infected ponies developed clinical signs and viral shedding, with the unvaccinated group displaying severe signs. One vaccinated pony and 2 unvaccinated ponies had cTnI greater than the reference range at 1 time point. At all other times, cTnI was <0.05 ng/mL. All control ponies had normal cTnI. There were no significant associations between cTnI and either clinical signs or experimental groups. When separated into abnormal versus normal cTnI, there were no significant differences among groups.
Conclusions and Clinical Importance: This study demonstrated no evidence of severe myocardial necrosis secondary to EIV challenge with 108 EID50 EIV A/eq/Kentucky/91 in these sedentary ponies, but transient increases in cTnI suggest that mild myocardial damage may occur.
clinically healthy, age-matched ponies not exposed to influenza
cardiac troponin I
cardiac troponin T
equine influenza virus
unvaccinated, influenza-naïve ponies challenged with EIV via nebulizer
influenza-naïve ponies that received a primary vaccination series 6 months before challenge with EIV via nebulizer
Viral respiratory disease occurs commonly in horses used for athletic competition; transmission is facilitated by large congregations stabled in close proximity. Horses in active work (ie, racehorses and other performance horses) frequently travel to competitions, increasing exposure to various pathogens to which they may be immunologically naïve.1 Equine influenza virus (EIV) is a common virus implicated in outbreaks of respiratory disease.2,3 Myocarditis or histopathologic evidence of myocardial damage is thought to occur secondary to certain viral infections in horses, including EIV, equine herpes virus 1, African horse sickness virus, equine viral arteritis, and equine infectious anemia.2,4–6 However, there is a lack of published information confirming a direct, causal relationship between myocarditis and viral agents in horses. Some horses that present with myocarditis may have a history of previous upper respiratory tract infection. In these cases, a viral agent is thought to be associated with the development of cardiac disease, based on a clinical association between prior disease and cardiac disease. This conclusion is extrapolated from what occurs in people, in whom viral infections are a common cause of myocarditis.7–10
There are hindrances to the identification of a virus as a causative agent of myocarditis. One difficulty has been a lack of available, practical diagnostic tests. Definitive proof of viral myocarditis in people requires a positive endomyocardial biopsy specimen. However, biopsies are invasive, and require multiple samples, pathologists, and immunologists trained in evaluating cardiac biopsy specimens and several different laboratory tests to document viral infection, an inflammatory response secondary to viral infection, or both.9,11 Although biopsy specimen is the gold standard for diagnosis in people, its sensitivity is questionable, and only a small percentage of people with clinically suspected viral myocarditis have the diagnosis confirmed.12 A technique for acquiring right atrial biopsy specimens in horses recently has been described,13 but still it is considered an invasive, experimental procedure that is not readily available.
Recently, more attention has been given to biochemical markers in the diagnosis of cardiac diseases. Cardiac troponins are myocardial regulatory proteins that are specific to the myocardium and released with cardiomyocyte damage.14,15 With the development of assays to measure cardiac troponin concentrations in the blood, the ability to diagnose myocardial injury has improved. Cardiac troponin I (cTnI) concentrations are increased in a number of human cardiac diseases, including viral myocarditis.12,16,17 Both cTnI and cardiac troponin T (cTnT) have been used to monitor viral myocarditis in people,12,17 and cTnT has been shown to correlate well with viral infection in the heart, and to precede the development of histologically confirmed inflammation in a mouse model of myocarditis.18 In this experimental model, cTnT concentrations increased by day 3 after infection, persisted at day 7, and had returned to normal by day 14. In horses, assay of cTnI has been validated,19 has been documented to increase in conjunction with certain cardiac diseases,20–22 and has been evaluated in a wider variety of equine diseases than has cTnT. In addition, cTnI is considered to be more sensitive and more specific for cardiac injury than other markers such as the cardiac isoforms of creatine kinase and lactate dehydrogenase.12–15,23
The purpose of this study was to determine the effect of acute EIV infection on cTnI in ponies. Increases in cTnI in ponies acutely infected with EIV would support its role as an etiologic agent in myocarditis, and could be the basis for more elaborate confirmatory and mechanistic studies. We proposed the hypothesis that influenza-challenged ponies would have evidence of myocardial damage, as shown by increases in cTnI. In addition, we sought to determine the variability of cTnI over time in healthy individual animals. Our 2nd hypothesis was that the day-to-day variability of cTnI in the healthy, control ponies would be insignificant.
Material and Methods
Experimental Animals and Study Design
This study was part of a larger study evaluating efficacy and duration of immunity to EIV after vaccination with a live recombinant canary pox vaccine expressing influenza virus hemagglutinin genes.a A detailed description of this vaccine trial is presented elsewhere.24 Briefly, the study was conducted as a blinded, randomized-controlled challenge trial. Twenty-three influenza naïve, 6-month-old male ponies were randomly assigned to either a vaccination (VAC: n = 12) or an unvaccinated (UNVAC: n = 11) group. The vaccine used was a Recombitek vaccinea containing 106.5 50% fluorescent antibody infectious dose of 2 live recombinant canary pox viruses, the first expressing the HA of A/eq/Kentucky/94 (American lineage) and a second expressing the HA of A/eq/Newmarket/2/93 (Eurasian lineage). The vaccine was administered IM in the neck immediately after reconstitution with 2 mL of diluent. The VAC group received 2 doses of vaccine at a 5-week interval, and both groups were subjected to a challenge infection 6 months after administration of the second vaccine. The individuals who performed the vaccinations were not involved with the remainder of the study, and individuals performing challenge and postchallenge clinical observations were blinded to group assignments. The challenge infection was performed by a 5-minute nebulization of challenge virus into a face mask (108 EID50 EIV A/eq/Kentucky/91). Throughout the study, the VAC and UNVAC ponies were housed together in a dry lot, open-air enclosure with a concrete apron with access to shelter and feeding area, but kept separate from all other horses. During this EIV challenge trial, 6 additional healthy and unvaccinated influenza naïve six 1-month-old ponies (5 females, 1 male) served as unchallenged controls (CON). These ponies were housed at a separate facility, approximately 10 miles away from the experimental ponies, but in a similar open-air, dry lot facility. For this reason, investigators were not blinded to the identity of ponies in this control group. All ponies were fed twice a day with a diet of hay and pelleted concentrate, and received mineral supplementation. Sampling and handling was performed by 1 individual, who sampled the CON ponies first, on the same days as the challenge ponies. This individual followed strict biosecurity hygiene measures, and showered out of the challenge facility to prevent cross-contamination. The maintenance and experimental protocols followed the animal care guidelines of the Animal Care and Use Committee, Colorado State University.
Detailed descriptions of physical examinations, viral sample collection, and serum antibody immunoassays are presented elsewhere.24 Physical examinations and rectal temperature measurements were conducted throughout the experiment after each vaccination, daily for 21 days postchallenge infection, and at all sample collection times. For evaluation of clinical disease, a clinical score consisting of the frequency of coughing, type of nasal discharge, and whether or not dyspnea was present was established and used to grade severity of challenge infection. A score of 0 was assigned if no cough was present, 1 if ponies coughed once, and 2 if there were 2 or more coughs. No nasal discharge was assigned 0, serous discharge 1, mucopurulent discharge 2, and profuse mucopurulent discharge 3. No dyspnea (respiratory rate <36 breaths/min) was 0, mild dyspnea (respiratory rate >36 breaths/min) was 2, and severe dyspnea (respiratory rate >36 breaths/min with additional clinical signs of respiratory distress) was assigned 4. These individual scores were added together to create the total clinical score for each pony on each day. The maximal daily score obtainable was 9. Clinical disease was considered to be severe if the daily clinical score was >4. Ponies also were monitored for clinical signs consistent with secondary bacterial bronchopneumonia using the following diagnostic criteria: persistent (>5 days) signs of pyrexia, severe mucopurulent nasal discharge, abnormal lung sounds, anorexia, or weight loss. If these criteria were met, affected animals were treated with antibiotics, anti-inflammatory medication, or both. If ponies were unresponsive to therapy and developed persistent and severe respiratory distress, they were euthanized.
Nasal swabs for detection of EIV shedding were collected before challenge, and for 21 consecutive days postchallenge with Dacron swabs.b Swabs were stored in 1 mL of virus transport medium at −80°C, and analyzed by a previously described real-time reverse transcriptase 1-step PCR assay.24 Blood was collected by jugular venipuncture at various intervals before and after vaccination and challenge for antibody assays, and plasma was stored at −20°C.
Blood for cTnI assay was collected from VAC and UNVAC ponies by jugular venipuncture into lithium heparin blood tubesc at the following time points: 1 day before challenge infection with EIV and on postchallenge days 2, 3, 5, 7, 10, 14, 17, 21, 24, and 28. The CON ponies were sampled on each of the same days, except for day 10. Blood samples were centrifuged and plasma obtained immediately, and aliquots of samples were frozen on the same day as collection, and stored at −70°C until analysis was performed. Analysis was performed with the Dade-Behring Stratus CS analyzer.d The range of detection for this analyzer is 0–50 ng/mL. Reference ranges for this laboratory have been published previously.25,26 Values <0.1 ng/mL were considered normal.
Changes in cTnI over time were compared with determine the effect of influenza infection. A 2-way repeated measures analysis of variancee was used to detect effects of group (VAC, UNVAC, CON) and time on cTnI. The differences among groups for cTnI status (abnormal versus normal) were compared by chi-square analysis. A cTnI <0.1 ng/mL was considered normal, based on our laboratory reference range. Associations between clinical parameters and cTnI were examined by clustered linear regression. The difference between UNVAC and VAC for frequency of severe clinical disease (clinical score >4) was compared by Fisher's exact test. Significance differences were determined to be at a P-value <.05.
Clinical Disease and Viral Shedding
A detailed description of the outcome of challenge infection has been presented elsewhere.24 Challenge infection with EIV in UNVAC ponies led to severe clinical signs of disease, including pyrexia, mucopurulent nasal discharge, dyspnea, and weight loss. In UNVAC ponies, the severity of clinical signs required therapeutic intervention in 8 of 11 ponies with nonsteroidal anti-inflammatory drugs and antibiotics, and 1 pony was euthanized on day 9 because its respiratory disease and dyspnea were unresponsive to therapy. On necropsy, this pony exhibited severe, diffuse, subacute, fibrinonecrotic bronchointerstitial pneumonia, but no gross or histopathologic abnormalities were observed in the heart. Clinical signs of disease were limited in VAC ponies, and were significantly less severe than those in UNVAC ponies (lower incidence of clinical score >4) throughout the postchallenge period (P < .01). None of the VAC ponies required medical treatment postchallenge infection. The CON ponies remained clinically normal throughout the study period.
All animals were negative for viral shedding before challenge infection. All ponies in the UNVAC and VAC groups tested positive for ≥1 day postchallenge. The duration of virus shedding was up to 8 days, and was not different among the challenged pony groups. However, the amount of daily postchallenge viral shedding was higher in the UNVAC group as compared with the VAC group (P < .05); this difference was typically 1–2 log-fold greater in UNVAC.
All CON ponies had normal cTnI at all time points (mean ± SD, 0.00 ± 0.01 ng/mL; Table 1). No significant effect of sample day was observed. No significant effect of time or treatment was observed in the EIV-challenged groups. Two UNVAC ponies had cTnI above the reference range (0.13 ng/mL on day 14 and 0.46 ng/mL on day 28), and 1 VAC pony had a cTnI above the reference range (0.32 ng/mL on day 5). All other samples were <0.05 ng/mL (Table 1). The pony that was euthanized had normal cTnI at each of the sample times before euthanasia. No significant associations were seen between cTnI and temperature (P= .93), body weight (P= .34), or clinical event score (P= .13). There were no significant differences in mean cTnI among groups (CON 0.00 ± 0.01, VAC 0.01 ± 0.03, UNVAC 0.01 ± 0.05, P= .16). The number of ponies with abnormal cTnI also was not different among groups (0/6 CON, 1/12 VAC, 2/11 UNVAC; P= .48). The relationship between antibody responses and cTnI was examined, with no apparent relationship observed (data not shown).
Table 1. cTnI concentrations by group for the postchallenge observation period.
|CON||0.00 ± 0.00||0.00 ± 0.00||0.01 ± 0.02||0.00 ± 0.00||0.00 ± 0.00||NA||0.00 ± 0.00||0.00 ± 0.00||0.00 ± 0.00||0.00 ± 0.00||0.00 ± 0.00|
|UNVAC||0.00 ± 0.00||0.01 ± 0.00||0.01 ± 0.02||0.00 ± 0.01||0.01 ± 0.01||0.01 ± 0.01||0.01 ± 0.04||0.01 ± 0.01||0.01 ± 0.01||0.01 ± 0.01||0.05 ± 0.14|
|VAC||0.01 ± 0.01||0.01 ± 0.01||0.01 ± 0.01||0.03 ± 0.09||0.00 ± 0.01||0.01 ± 0.01||0.00 ± 0.01||0.01 ± 0.01||0.01 ± 0.01||0.00 ± 0.01||0.00 ± 0.01|
Ours is the 1st controlled study that attempted to demonstrate a direct relationship between equine influenza infection and myocardial damage with cTnI as an indicator of myocardial cell necrosis. Although certain viruses have been shown to result in cardiac damage,4–6 no studies have documented a cause and effect relationship between EIV infection and myocardial damage antemortem. Only 3 of 23 challenge-infected ponies had increased cTnI, and in 2 ponies, these increases were of short duration (<3 days). Although no lasting biochemical evidence of myocarditis was present, these transient increases in cTnI suggested that myocardial injury did occur, because none of the CON ponies had increases in cTnI over the duration of the study. All UNVAC ponies challenged with influenza virus developed clinically relevant disease, with intervention required in the majority of ponies. The young age of the ponies made it more likely they would be at greater risk for developing more severe clinical disease than mature ponies.27,28 In people, viral myocarditis is more likely to occur or be more severe in children than in adults.29–31 Despite the severity of challenge and age of the ponies, none of the ponies had persistently or markedly increased cTnI, although there were transient increases in cTnI.
Samples were collected over a 4-week period because the timing of possible cardiac insult relative to infection is unknown in horses. In 3 ponies that had increases, they occurred at various time points (5 days postchallenge for the VAC pony, and 14 and 28 days postchallenge for the UNVAC ponies). In people, timing of illness may be dependent on the mechanism causing the disease. Both direct viral damage and host inflammatory response to the virus can contribute to acute and subacute disease, seen within a few days to a few weeks after infection.10,32,33 The associated clinical signs can range from subclinical or mild disease, which may be self-limiting, to fulminant myocarditis, congestive heart failure, and death.34 More chronic sequelae, such as progressive cardiac dysfunction and dilated cardiomyopathy, also can occur and may be a result of immune-mediated mechanisms.8,10,32,35 These complications may not be manifest until weeks to months after exposure.36 The mechanism of cardiac insult in these ponies is unknown, but there was no apparent relationship between the occurrence of increased cTnI and antibody titer.
Although we observed only minor biochemical evidence of cardiac insult in these ponies, the effects in active performance horses may differ from those observed in these sedentary ponies. Active training, insufficient rest, or both may exacerbate mild cardiac damage or may decrease the immune response, decreasing the horse's ability to clear virus.37 A study by Gross et al28 in horses challenged with influenza virus showed that horses exercised postchallenge had more severe clinical signs of disease than those rested postchallenge. These horses also had a higher rate of weight loss and loss of body condition, suggesting that protein catabolism may have contributed to a loss in lean body mass. In mice infected with influenza virus, decreased protein synthesis and increased protein degradation may contribute to the observed myocardial muscle wasting.38
Clinically healthy control ponies were used to determine variability in cTnI over time. To our knowledge, ours is the 1st study to examine changes in cTnI over time in healthy ponies. All CON ponies had undetectable cTnI concentrations (0 or 0.01 ng/mL) at all time points, except for 1 pony that had a cTnI of 0.06 ng/mL (within our reference range of <0.1 ng/mL) at 1 time point. These ponies had cTnI within the reference range already determined for adult horses of different breeds for the analyzer used.25,26
Echocardiograms and electrocardiograms to document that the ponies had normal cardiovascular systems before challenge, and that the ponies with increased cTnI had concurrent abnormalities in cardiac function, could not be not performed in this study. Therefore, the functional effects of the increases in cTnI are unknown. However, because each pony had a normal prechallenge cTnI, and the CON ponies did not have variations in cTnI over the same duration of time, the transient increases suggested that they were associated with viral challenge. Definitive proof of cause and effect would have required endomyocardial biopsies, but cardiac troponins have been used to follow both clinical and experimentally induced viral myocarditis in people and laboratory animals.12,17,18
Financial constraints prevented daily sampling or sampling beyond 28 days. Although cTnI increases in some ponies may have been missed by not collecting samples daily, the sampling frequency was sufficient to determine that none of the ponies had prolonged or severe increases in cTnI up to 1 month postchallenge. In people, cardiac troponins are released into systemic circulation soon after cardiomyocyte injury (within 1–8 hours) and can persist for 4–7 days, depending on the extent of injury.39 Persistence of cTnI in the circulation is attributed to the gradual breakdown of the cardiomyocyte after infarction, and the slow release of cTnI that is complexed to the thin filaments. Approximately 95% of cTnI is structurally bound, but a small percentage (approximately 5%) is found free in the cytoplasm, and may account for small, transient increases in circulating cTnI.15 cTnI may be released from the cytoplasm after myocardial cell membrane damage and leakage of unbound cTnI, without complete necrosis and irreversible damage. cTnI from this source may not persist as long as that released during cell necrosis or cell death, and the increases may be more modest.39,40 In horses, cTnI is released within 3–4 hours of insult, although the half life of cTnI is not known.
Additional research is needed to determine the effect of influenza virus infection on cTnI concentrations in horses with naturally occurring disease and in horses infected with different strains of influenza virus, which may have different tropisms for cardiac muscle. We used a virus type that was highly pathogenic based on previous experience, but differences may exist in effect on cardiac muscle between strains. In addition, the effect of influenza virus infection on horses in active training should be evaluated, as well as correlation of any changes in cTnI with measures of cardiac function, such as electrocardiography and echocardiography.
a Recombitek, Merial, Duluth, GA
b Baxter Healthcare Corporation, McGaw Park, IL
c Becton Dickinson Company, Franklin Lake, NJ
d Dade Behring, Bear, DE
e JMP 4.0.4, SAS Institute, Cary, NC
We thank Donna Teleis for assistance with cTnI sample analysis.
The work was supported by the Raymond Firestone Trust Research Grant, University of Pennsylvania.