Introduction

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an important cause of syncope and sudden death in people and boxer dogs.^1–5 In the boxer breed, the disease is typically manifested by malignant right ventricular (RV) origin arrhythmias, left ventricular (LV) myocardial dysfunction, or both.^5,6 Genetic studies in the boxer have linked both ARVC and phenotypic dilated cardiomyopathy (DCM) to a deletion in the gene encoding for the desmosomal protein striatin, implying that DCM in the boxer is likely a continuum of the same disease process.^7,8 Although the disease has been best characterized in the boxer, isolated cases of confirmed or suspected ARVC have been described in other breeds, and recent reports and clinical observations suggest that the related English bulldog is also affected by a form of the disease.^5–17

Arrhythmogenic right ventricular cardiomyopathy was first described in an English bulldog with sustained RV-origin ventricular tachycardia (VT) by Santilli et al. in 2009.¹⁶ Postmortem histopathology of the heart of this dog documented fatty infiltration of the RV outflow tract, thus confirming the ARVC diagnosis. Another report by Santilli et al. in 2011 described the electrophysiologic characteristics and echocardiographic findings in five English bulldogs that also presented with clinical signs consistent with ARVC.¹⁷ Studies of the disease in this breed have thus far been limited to these reports and it is unknown whether the clinical manifestations and genetic basis of the disease are the same as those in boxers.

The purpose of this retrospective study was to examine the clinical features, survival characteristics, and prognostic factors in a population of English bulldogs with presumed ARVC. Particular attention was given to the presenting signalment and clinical signs, electrocardiogram (ECG) characteristics, medical treatments, and clinical outcome of these dogs. Findings from this study may help to increase awareness and understanding of the disease in English bulldogs, allowing for enhanced clinical suspicion and diagnosis of the disease in this breed.

Materials and Methods

English bulldogs with a clinical diagnosis of ARVC were identified by searching the medical record databases of Foster Hospital for Small Animals at Tufts University Cummings School of Veterinary Medicine and the New England Veterinary Cardiology practice. The databases were searched to identify all English bulldogs examined between 2001 and 2013 in which a clinical diagnosis of ARVC was made by the attending cardiologist. Dogs found to have frequent RV origin (left bundle branch block [LBBB] morphology) ventricular premature complexes (VPCs; greater than one VPC per 3 min diagnostic ECG or 100 VPC per 24 hr on telemetry or Holter monitoring) or runs of VT were included in the study. Dogs with clinically significant congenital heart disease, or a systemic disease process that could explain the observed arrhythmias, were excluded.

Information collected from the medical records included signalment, presenting signs, physical examination, diagnostic ECG and in-hospital telemetry findings, echocardiographic measurements, administered cardiac medications, and clinical outcome, including survival time and cause of death. Additional information collected if available included 24-hr ambulatory electrocardiographic (Holter) monitor data, radiographic findings, and selected hematologic parameters (packed cell volume, total solids, blood urea nitrogen, creatinine, alanine aminotransferase, and aspartate aminotransferase). Arrhythmias were classified as follows: ventricular ectopy (VE) Class 1—single isolated VPCs; VE Class 2—couplets/bigeminy/ trigeminy; and VE Class 3—VT. Congestive heart failure (CHF) was diagnosed if cardiogenic pulmonary edema or cavitary effusions were present. Dogs were classified into one of three clinical cohorts based on their clinical signs and diagnostic findings at the time of initial diagnosis: Group 1—subclinical arrhythmia; Group 2—symptomatic arrhythmia; and Group 3—CHF. Following the medical record review, referring veterinarians and/or owners of dogs were contacted for those dogs for whom the outcome could not be determined from the medical record. Follow-up information collected included the animal’s current clinical status, or the date and cause of death (i.e., euthanasia versus natural death; sudden cardiac death or death due to refractory CHF).

Descriptive statistics were used to characterize clinical manifestations and survival times for dogs in each clinical group. Data were graphically inspected to evaluate for normality; however, due to small sample sizes, nonparametric methods were used for all analyses. Categorical data were compared by use of χ² tests, whereas continuous data were analyzed using Kruskal-Wallis one-way ANOVA initially, followed by Mann-Whitney U tests as appropriate to determine pair-wise significance levels. Potential associations between relevant clinical data and survival times were evaluated using Spearman’s rank-correlation coefficient. Survival function and associated survival curves were calculated using Kaplan-Meier Product Limit estimates and survival curves were compared using the log-rank test. Dogs still alive at the end of the study were censored in the survival analysis; dogs that were lost to follow-up were included in the survival analysis until the last date they were known to be alive and censored thereafter. Statistical tests were considered to show a significant difference with a P ≤ .05. Statistical analyses were performed using commercially available statistical software^a.

Results

Thirty-one dogs met the inclusion criteria for the study. Twenty-three dogs were male (22 castrated) and eight were spayed females. The male to female ratio was 2.9:1. The mean age was 9.2 ± 1.6 yr (range 7–13 yr). At the time of diagnosis, mean body weight was 26.9 ± 6.5 kg (range 18.0–45.3 kg) and median body condition score was 6/9 (range 5/9–9/9; n = 23). Fifteen dogs (48%) had a systolic heart murmur (grade I/VI in five dogs; grade II/VI in seven dogs; and grade III/VI in three dogs). Twenty-four dogs (77%) had an arrhythmia detected during physical examination. The average physical exam heart rate (HR) was 162 ± 45 bpm (range 100–280 bpm). The mean systolic Doppler blood pressure was 115 ± 27 mm Hg (n = 14; range 84–158 mm Hg). Twenty-one dogs (68%) had thoracic radiographs taken. Sixteen (76%) of those dogs had radiographic generalized cardiomegaly, eight dogs (38%) had pulmonary edema, and five dogs (24%) had pleural effusion.

At the time of initial presentation, 5 dogs (16%) were classified as Group 1 (subclinical arrhythmia); 10 dogs (32%) were classified as Group 2 (signs referable to arrhythmia); and 16 dogs (52%) were classified as Group 3 (overt signs of CHF; Table 1). All of the dogs in Group 1 had arrhythmia incidentally discovered after presentation for an unrelated complaint; one of these dogs presented for evaluation of a thoracic (extracardiac) mass, one dog presented for ongoing treatment of hypothyroidism, one dog had a presurgical evaluation for ophthalmologic surgery, one presented for a head tilt, and one presented for gastrointestinal distress. Nine dogs in Group 2 exhibited syncope and one dog presented with lethargy and gastrointestinal signs presumed secondary to sustained VT. Dogs with CHF were younger than dogs with symptomatic arrhythmia (P = .02), more likely to exhibit supraventricular arrhythmia (SVA; P = .04) and cough (P = .04), and less likely to present with VT (P = .01). Six dogs in Group 3 (38%) had predominant signs of left-sided CHF with radiographic evidence of pulmonary edema; seven dogs (33%) had right-sided CHF, and three dogs (19%) had biventricular CHF. Six dogs in Group 3 (38%) also had a history of syncope or collapse, 11 dogs (69%) exhibited dyspnea, and 7 dogs (44%) exhibited cough. Compared with dogs in Group 1, dogs with CHF were more likely to present with dyspnea (P = .01). There were no significant differences with regard to body weight, body condition score, HR, respiratory rate, or selected laboratory values between clinical groups, apart from total solids, which was higher in Group 1 (P = .04).

TABLE 1 Clinical and Laboratory Findings in 31 English Bulldogs with Arrhythmogenic Right Ventricular Cardiomyopathy Grouped on the Basis of Clinical Manifestations at the Time of Initial Presentation

View Fullscreen

Thirteen dogs (42%) had evidence of VT documented on in-hospital diagnostic ECG or telemetry, and eight of those dogs demonstrated “R-on-T” phenomenon. Thirteen dogs (42 %) were assigned a VE Class of 1, 5 dogs (16%) were VE Class 2, and 13 dogs (42%) were VE Class 3. Twenty-nine dogs had exclusively RV origin VPCs (LBBB morphology QRS complexes positive in leads I, II, III, and aVF with deep S waves in leads aVR and aVL), and three dogs had concurrent right bundle branch block morphology VPCs (positive in aVR and aVL with deep S waves in leads I, II, III, and aVF). Seven dogs had concurrent SVAs (two with atrial fibrillation and five with atrial premature contractions and paroxysms of supraventricular tachycardia); all of these dogs had concurrent CHF. No differences in ECG VE Class or incidence of VT during hospitalization were noted between clinical groups (Table 2).

TABLE 2 ECG and Echocardiographic Findings in 31 English Bulldogs with Presumed ARVC

View Fullscreen

Twenty dogs (65%) were monitored via telemetry during hospitalization and 10 dogs (31%) had 24-hr Holter monitor recordings. Three dogs had Holter monitoring performed prior to starting any antiarrhythmic medications; the remainder of Holter analyses were performed following initiation of antiarrhythmic drugs.

All dogs had an echocardiogram performed. Nineteen dogs (59%) exhibited subjective RV dilation based on the right parasternal short axis view, and only three dogs (9%) had subjective pulmonary artery enlargement (pulmonary artery diameter larger than the aorta on right parasternal short axis view at the level of the heart base). Twenty dogs (63%) had echocardiographic evidence of left atrial enlargement as defined by an M-mode derived left atrial to aortic root ratio obtained on a right parasternal short axis view of greater than 1.3:1. The LV fractional shortening was greater in the dogs with subclinical arrhythmia compared with the dogs with symptomatic arrhythmia (P = .01) or CHF (P = .04; Table 2); however, no differences in LV internal dimension in diastole or LV internal dimension in systole were noted between groups. Eighteen dogs (58%) had mitral regurgitation, and 20 dogs (65%) had tricuspid regurgitation. The mean tricuspid regurgitation velocity was 2.5 ± 0.4 m/sec (range 1.8–3.1 m/sec) in the 10 dogs in whom it could be measured. Four dogs (13%) had a concurrent heart base mass found on echo that was judged to be incidental (not causing pericardial effusion or compression of surrounding structures).

Twenty-seven dogs (87%) were started on oral antiarrhythmic therapy. Eight (26%) animals were started on sotalol alone, 3 (10%) were on mexiletine alone, and 13 (42%) were on both mexiletine and sotalol simultaneously. Twelve dogs (39%) required IV antiarrhythmic therapy for VE while in the hospital. Eleven dogs received IV lidocaine boluses (median 1 bolus of 2 mg/kg, range 1–6 mg/kg) and four dogs were placed on a lidocaine continuous rate infusion (median 50 mcg/kg/min, range 50–75 mcg/kg/min). Four animals were administered IV procainamide (median 1 bolus of 10 mg/kg, range 4–17 mg/kg) and two of them were maintained on a procainamide continuous rate infusion (median 30 mcg/kg/min, range 30–38 mcg/kg/min). No obvious associations were seen between survival times and the specific antiarrhythmic protocol used. Nine animals received IV furosemide (median 3 mg/kg, range 2–6 mg/kg). A variety of other cardiac medications were administered chronically in varying combinations (Table 3).

TABLE 3 Chronic Oral Cardiac Medications Used in the Treatment of 31 English Bulldogs with Presumed ARVC

View Fullscreen

At last contact, 18 (58%) dogs were dead, 10 (32%) were alive, and 3 (10%) were lost to follow-up. Of the animals that died, 4 dogs (22%) died suddenly, 2 dogs (11%) died of refractory CHF, 10 (56%) were euthanized for cardiac issues, and 2 (11%) died or were euthanized for noncardiac issues. The presence of syncope (r = −.40; P = .03), a higher examination HR (r = −.43; P = .04), and a higher serum potassium level (r = −.60; P = .002) were all weakly correlated with shorter survival. Figure 1 shows the Kaplan-Meier survival curves for the three clinical groups. The median survival time (MST) of all dogs was 8.3 mo. Dogs with subclinical disease had an MST of 12.4 mo, dogs exhibiting syncope lived 4.9 mo, and dogs with CHF lived 8.3 mo. The log-rank test of the Kaplan-Meier survival curves did not show a significant difference in survival between the three groups.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

English bulldogs develop a form of cardiomyopathy that is similar in many ways to ARVC in boxer dogs; however, this study also suggests some important potential differences in the manifestations of presumed ARVC in these two related breeds. Notably, the majority of English bulldogs described in this study presented with signs of CHF. Although a majority of boxers with cardiomyopathy in the United Kingdom appear to present with signs of myocardial dysfunction and a high incidence of CHF, the majority of North American boxers with ARVC have overtly normal LV function and CHF is an uncommon manifestation of the disease in this country.^6,9 Further prospective studies involving active screening for ARVC in bulldogs may help determine whether CHF accompanying ARVC is truly more common in this breed, or whether the high prevalence of CHF in the dogs of this study rather reflects the increased disease severity of dogs who are treated at referral institutions.

Harpster first described three clinical categories of cardiomyopathy in the boxer.¹⁸ The first category includes subclinical dogs who have VE detectable on ECG or Holter monitoring but exhibit no outward signs of cardiac disease, much like the English bulldogs in Group 1 of our study. The second category includes dogs who present with ventricular tachyarrhythmia resulting in clinical signs of exercise intolerance, weakness, or syncope, similar to the dogs in Group 2 of this study. The final category includes dogs with myocardial dysfunction and CHF, most similar to the dogs in Group 3 of this study. Dogs in the latter group also frequently exhibit arrhythmias of both ventricular and supraventricular origin.^11,18 The English bulldogs in this study were grouped according to presenting clinical signs rather than echocardiographic measurements of systolic function, and it is interesting to note that the LV fractional shortening in the CHF group was not significantly different from that of dogs with symptomatic arrhythmia without CHF. However, both of these groups had a lower LV fractional shortening compared with the subclinical group. This may indicate that dogs in Group 1 were at an earlier stage of disease relative to those in Groups 2 or 3. Longitudinal analysis of echocardiographic parameters and monitoring of disease progression would be required to further clarify this finding. In boxers, the same genetic mutation in the striatin gene that was originally described in dogs with ARVC was recently found to be commonly associated with phenotypic DCM, implying that DCM in the boxer is likely a continuum, or end result, of the same disease process.⁸ Given the retrospective nature of the study, genetic testing was not carried out to evaluate whether any of these dogs possessed the striatin mutation found in some boxers with ARVC. The genetic underpinnings of ARVC in English bulldogs and the relationship between the arrhythmic and structural forms of the disease have yet to be determined but merit future study.

A strong male predisposition for ARVC was noted in this study, with male bulldogs outnumbering females by a factor of 2.9:1. This is consistent with some prior studies in boxers, and in young people in the second to fourth decades of life, that also show a male predilection for ARVC.^9,11,19–21 Gender did not appear to have a significant impact on survival of bulldogs in this study. This is consistent with the results of several other studies failing to demonstrate a gender effect on the prognosis or survival of boxers with ARVC.^9,19,21 However, in a recent study by Motskula et al., male boxers had an increased incidence of myocardial dysfunction and increased cardiac mortality relative to their female counterparts.²¹ Dogs in our study were older than most boxers first presenting with ARVC, with a mean age at the time of presentation of 9.2 yr. The mean age of boxers with ARVC has ranged from 5.7–6 yr in dogs with myocardial dysfunction to 7.6–7.7 yr in dogs with arrhythmia alone.^9,11,19 In the present study, bulldogs with CHF were younger than dogs in the other subgroups, possibly suggesting that these dogs had a more malignant or rapidly progressive form of the disease.

Supraventricular arrhythmias were relatively uncommon in this study, present in 23% of the dogs, and seen only in dogs with concurrent CHF. Supraventricular arrhythmias have also been described in people and boxers with ARVC. Harpster reported a 13% incidence of atrial fibrillation or supraventricular tachycardia in his initial description of cardiomyopathy in the breed, whereas 20 out of 48 boxers with myocardial dysfunction (42%) reported by Baumwart et al. exhibited supraventricular tachyarrhythmias.^11,18 Atrial arrhythmias have been described in 14% of people with ARVC and are associated with a higher incidence of death, heart failure, and left atrial enlargement on echocardiography.²² In this study, the presence of SVA was correlated with CHF, and all dogs with SVA were in Group 3. However, no associations were seen between presence or absence of SVA and survival time or left atrial size in this study.

All dogs in the present study had ventricular arrhythmia because this was a criterion of study inclusion. Most dogs had VPCs that were exclusively of presumed RV origin (LBBB morphology) with positive deflections in leads I, II, III, and aVF, as described previously in boxers and bulldogs with ARVC.^5,6,16–18 However, three dogs had concurrent right bundle branch block morphology VPCs (negative in leads I, II, III, and aVF). The latter may have originated from the RV side of the interventricular septum or outflow tract, or from the left ventricle, implying concurrent LV involvement, as has been seen in both boxers and people with ARVC.^6,8,11,23,24

Incidental heart base masses are common in older bulldogs and were seen on echocardiography in 4 of 31 dogs (13%) in this study. Most masses found at this location in brachycephalic breeds are aortic body tumors (also called chemodectomas). These masses tend to be slow growing and locally invasive, eventually causing clinical signs from either pericardial effusion or compression of surrounding cardiovascular structures, neither of which accompanied the heart base masses seen in the dogs of this study. Arrhythmias are an uncommon manifestation of heart base masses in dogs, seen in 4 of 33 dogs with histologically confirmed aortic body tumors in one study.²⁵ Although one of the dogs in that study had atrial fibrillation preoperatively, the three dogs with ventricular arrhythmias all had arrhythmias limited to the postoperative period that resolved with standard supportive therapies.²⁵

Prior studies have shown that syncope, myocardial dysfunction, worsening arrhythmia class, and increasing age are significant negative prognostic indicators for boxers with ARVC.^9,19 In this study, the MST of the subclinical group was numerically longer than that of the symptomatic arrhythmia or CHF group, and dogs with symptomatic arrhythmia (mostly syncopal dogs) appeared to fare worse than dogs with active CHF. Although the survival times of the three groups were not significantly different in Kaplan-Meier analysis, these results should be interpreted with caution given the small sample sizes and low event rates, particularly in Group 1. In boxers, the MST of dogs with myocardial dysfunction has been reported to be only 17 wk, whereas bulldogs with overt CHF in the present study had an MST of 8.3 mo.¹⁰ Conversely, prior studies of boxers with asymptomatic ARVC have demonstrated MSTs ranging from 23 to 70 mo, compared with 12.4 mo for subclinical bulldogs in this study.^9,19,20 The latter could suggest that bulldogs develop a more malignant form of ARVC compared with their boxer counterparts; however, it more likely reflects increased clinical suspicion and active screening for ARVC in boxers, resulting in enhanced detection of the disease at earlier stages. Prospective studies involving widespread Holter screening for ARVC in bulldogs may help to elucidate the prevalence of the disease and provide a more accurate picture of prognosis and survival times afforded by early detection.

The bulldogs in this study were on a variety of cardiac medications, tailored to their clinical signs. The most commonly used antiarrhythmics in boxer dogs are sotalol or a combination of mexiletine and either sotalol or atenolol^a.²⁶ Both sotalol and a combination of atenolol and mexiletine have been shown to reduce arrhythmia frequency in boxers, and the combination of sotalol and mexiletine may be more effective in reducing arrhythmia frequency in boxers than sotalol alone.^26,27 However, to date no antiarrhythmic has been shown to be superior in the prevention of sudden death or prolongation of survival, and people with ARVC are treated with placement of internal cardioverter defibrillators.²¹ In the present study, the most common antiarrhythmic protocol used was a combination of sotalol with mexiletine, with variable use of enalapril, pimobendan, and diuretics for dogs with CHF. Given the retrospective nature of this study and the lack of standardized medication regimens, it was not possible to determine the independent effects of individual medications on survival time. However, future prospective studies may help evaluate the efficacy of these medications in English bulldogs with ARVC.

There were several important limitations to this retrospective study. The only way to definitively diagnose ARVC is via histopathologic confirmation of fatty/fibrofatty infiltration on postmortem examination or endomyocardial biopsy. Although the dogs in this study fit the clinical description of ARVC, this was not confirmed with histopathology. In order to increase the size of our study population, data were collected from dogs evaluated at two different practices in the New England region over a 12-yr period. Consequently, interpretation of echocardiographic data was carried out individually by the cardiologist at each hospital, and several two-dimensional and Doppler echocardiographic parameters now routinely measured were not collected or not available for review. Furthermore, the relatively small number of dogs in this study and nonstandardized diagnostic and medication protocols significantly limit the utility and interpretation of the survival data, including the ability to calculate stable estimates.

Conclusion

This study describes the clinical characteristics of a larger group of English bulldogs with presumed ARVC than previously described. Compared with boxers with ARVC in the United States, the English bulldogs in this study appeared to have a higher incidence of CHF. The overall survival time of the dogs in this study was 8.3 mo from the time of diagnosis, and no differences in survival were detected with respect to clinical grouping. The information gleaned from this analysis may raise clinical awareness of ARVC in English bulldogs. Further prospective studies are needed to better determine the incidence, prognosis, and genetic basis of the disease, and to gain more information on optimal treatment strategies. The English bulldog may also serve as another viable animal model for studying this disease in humans.