• Heart failure;
  • 24 hours Holter antiarrhythmic;
  • Rate control


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

Background: Atrial fibrillation (AF) with excessively high ventricular rates (VR) occurs in dogs with advanced heart disease. Rate control improves clinical signs in these patients. Optimal drug therapy and target VR remain poorly defined.

Hypothesis: Digoxin-diltiazem combination therapy reduces VR more than either drug alone in dogs with high VR AF.

Animals: Eighteen client-owned dogs (>15 kg) with advanced heart disease, AF, and average VR on 24-hour Holter > 140 beats per minute (bpm).

Methods: After baseline Holter recording, dogs were randomized to digoxin or diltiazem monotherapy, or combination therapy. Repeat Holter evaluation was obtained after 2 weeks; dogs were then crossed over to the other arm (monotherapy or combination therapy) for 2 weeks and a third Holter was acquired. Twenty-four hour average VR, absolute and relative VR changes from baseline, and percent time spent within prespecified VR ranges (>140, 100–140, and <100 bpm) were compared. Correlations between serum drug concentrations and VR were examined.

Results: Digoxin (median, 164 bpm) and diltiazem (median, 158 bpm) decreased VR from baseline (median, 194 bpm) less than the digoxin-diltiazem combination (median, 126 bpm) (P < .008 for each comparison). With digoxin-diltiazem, VR remained <140 bpm for 85% of the recording period, but remained >140 bpm for 88% of the recording period with either monotherapy. Serum drug concentrations did not correlate with VR.

Conclusions and Clinical Importance: At the dosages used in this study, digoxin-diltiazem combination therapy provided a greater rate control than either drug alone in dogs with AF.


angiotensin-converting enzyme


atrial fibrillation


24-hour average heart rate from Holter at baseline


congestive heart failure


thoracic radiograph




combination therapy of digoxin and diltiazem




heart rate


24-hour average heart rate from Holter recordings


absolute change in heart rate from baseline


percent change in heart rate from baseline


New York Heart Association


ventricular rate


24-hour average heart rate from Holter after 2 weeks of therapy

Atrial fibrillation (AF) with a rapid ventricular rate (VR) commonly occurs in dogs with advanced heart disease, and results in substantial morbidity, including worsening of congestive heart failure (CHF) and decreased survival times.1–3 Most dogs with AF and rapid VR fail electrocardioversion or do not maintain sinus rhythm, and improvement in clinical signs usually depends on rate control.4 Additionally, in humans with AF and CHF, VR control may provide clinical outcomes that are not different from those achieved with rhythm control after cardioversion.5

Clinicians have traditionally used digoxin for VR control in dogs with AF, but digoxin can inadequately control excessively rapid VR in dogs.1 Furthermore, digoxin inadequately controls VR in human patients with AF during exercise or CHF, because increased sympathetic tone overrides the vagal effects of digoxin resulting in “break-through.”6,7 Diltiazem also slows VR acutely in dogs with experimentally induced AF,8 but little is published on diltiazem in dogs with spontaneous AF. Additionally, some veterinary clinicians hesitate to use diltiazem because of potential negative inotropic effects. In humans, diltiazem typically fails to control VR adequately.7,9 Conversely, satisfactory VR control is often achieved in human patients with combinations of digoxin and either calcium channel blockers or β-blockers.7,10–12 Unfortunately, virtually no information is available about either diltiazem monotherapy or digoxin combination therapy in dogs with AF.a,1 Assessment of VR control in veterinary medicine also remains poorly defined. Several studies have demonstrated that continuous heart rate information collected for 24 hours by ambulatory electrocardiographic (Holter) monitoring provides more accurate VR information than short ECG “snapshots” obtained in a hospital setting,13,14 and thus is considered closer to ideal.15–17 Additionally, more complex analysis of VR control is possible with Holter monitoring.

Optimal or adequate VR control in dogs with AF remains similarly undefined. Limited dataa,4 measured by ECG or auscultation rather than 24-hour Holter suggest a target VR of <160 beats per minute (bpm) in dogs with AF secondary to CHF. No studies, however, have demonstrated that these target rates are either adequate or optimal in terms of improved management of CHF, enhanced quality of life, or prolonged survival. Recommendations for human patients suggest that target VR approximate the “situation-appropriate” sinus rates (ie, the VR of a patient with AF at rest or during a 6-minute walk test should be similar to what the sinus rate would be during that particular activity if the same patient were not in AF).18,19 Thus, more aggressive VR control in dogs with rapid VR might be appropriate.

We sought to determine which treatment produces the greatest decrease in VR without adverse effects. We compared the effect of DG, DT, and a DGDT combination in decreasing VR in large-breed dogs with sustained, naturally occurring AF with rapid VR and severe cardiac disease, as assessed by 24-hour ambulatory ECG analysis.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

Study Design and Patient Selection

In this prospective, randomized crossover, clinical trial, we examined the ability of 3 pharmacological protocols—DG or DT monotherapy, or DGDT—to control VR in large-breed dogs with severe cardiac disease exhibiting AF and high VR.

Candidate dogs, presented to Cornell University Hospital for Animals, underwent complete cardiac evaluation by a cardiologist, including physical examination, ECG (to confirm AF), echocardiogram, CXR, and a baseline Holter recording. To qualify for enrollment, the 24-hour average heart rate (HRHOLTER) recorded at baseline (before starting any antiarrhythmic drugs) had to exceed 140 bpm, and body weight had to be >15 kg. We chose these criteria because dogs with underlying heart disease and HRHOLTER≤ 140 bpm might not warrant VR control, and large-breed dogs are most commonly afflicted by AF secondary to severe heart disease. Qualifying patients were then randomly assigned to a specific drug treatment order, established by a random number generatorb before starting the trial. The study design and subject allocation are illustrated in Figure 1.


Figure 1.  Study design. Examination schedule and order of drug administration is shown. Numbers within circles represent the number of dogs in a particular phase of the study. PE, physical examination; ECG, electrocardiogram; echo, echocardiogram; CXR, thoracic radiograph; DG, digoxin phase; DT, diltiazem phase; DGDT, combination therapy (digoxin and diltiazem); 2 wk, 2 weeks; BL, baseline; AF, atrial fibrillation; [digoxin], serum digoxin concentration measurement; HR, heart rate; BW, body weight.

Download figure to PowerPoint

CHF, when present, had to be controlled medically before the 1st treatment (phase 2 of 3) Holter recordings; in the dogs with CHF, the baseline Holters were acquired while CHF was still inadequately controlled. Dogs in CHF received standard medication for management of CHF as required (eg, diuretics, ACE inhibitors, and pimobendan), excluding DG. If patients were already receiving DG at presentation, DG therapy had to be discontinued for at least 1 week before acquiring the baseline Holter recording. Dogs were treated with DGc at approximately 0.005 mg/kg PO q12h, DTd at approximately 3 mg/kg PO q12h, or DGDT at these dosages. The dogs received each treatment for 2 weeks (to ensure drugs reached steady-state) before reexamination. During examination, a 1-minute ECG was acquired to document AF, serum samples were obtained for assessment of DG (6–8 hours post-pill) or DT (4–6 hours post-pill) concentrations, and a Holter recording was obtained. After removing the Holter recorder, each dog received the alternate scheduled treatment for 2 weeks, after which a 3rd examination was performed. Serum DG concentration was measured in-house by a fluorescent polarization immunoassay.e Serum DT concentration was measured by high-performance liquid chromatography at a commercial laboratory.f

Qualifying pet owners signed a consent form upon enrollment. The study protocol was approved by the Cornell University Institutional Animal Care and Use Committee.

Data Acquisition and Analysis

Holter monitoring was performed using 3-channel digital recorders, in an orthogonal lead arrangement.f A small area of hair was clipped for each electrode (2 per channel and 1 ground) and ECG electrodes were connected to the Holter recorder, which was fitted into a custom-made vest. Routine activities for the dogs were encouraged and an activity diary was kept by the owner.

We acquired Holter data at a sampling frequency of either 200 or 400 Hz, stored it on 350-MB removable PC flashcards, and transferred it to a hard drive for automated analysis by proprietary software.g We tabulated hourly VR averages and calculated the 24-hour average heart rate (HRHOLTER). To ensure the fidelity of automated data analysis, a trained technician blinded to the treatment (SAH) performed a thorough manual inspection and corrective editing of each Holter.

We evaluated the longitudinal dynamics of VR control over 24 hours by assessing the hourly average VR. The Holter recordings were not initiated at the same time of day for all dogs in the study. Thus, we examined all the hourly average VR both in real time (to examine the effects of daytime and nighttime activity on VR) and after normalizing to time of Holter application (to examine the effects of hospital visit on VR). We then examined the ability of each treatment to maintain VR within 3 arbitrarily defined VR ranges; (1) high VR (>140 bpm), (2) target range VR (100–140 bpm), and (3) slow VR (<100 bpm), and compared the percentage of time in each category for each treatment group.

The minimum hourly average VR of each Holter was evaluated to identify the presence of inappropriate bradycardia (heart rate < 40 bpm) induced by any of the treatment regimens. We also examined, but did not statistically analyze, pauses, maximum heart rates, and ventricular tachyarrhythmias (singles, couplets, and runs of ventricular tachycardia).

Statistical Analysis

Data were initially summarized by calculating descriptive statistics using commercial and community-based software.h,i All data were assessed for normality by the Shapiro-Wilk normality tests and subsequently analyzed by nonparametric tests. Confidence intervals for median values were calculated as described previously.20 The median, ranges, and 95% confidence intervals for median values are reported for HRHOLTER at baseline (BL-HRHOLTER) and after 2 weeks of treatment (2WK-HRHOLTER) with each of the 3 treatments. Because heart rate responses in dogs receiving a single drug first followed by combination therapy were not appreciably different from responses in dogs receiving combination therapy first followed by single drug, we concluded that no time/treatment interaction was present, and pooled the results for these groups. To determine whether any treatment provided better VR control than any other treatments, we compared the absolute change in heart rate from baseline (ΔHR = 2WK-HRHOLTER−BL-HRHOLTER) and the % change in heart rate from baseline (%ΔHR =ΔHR/BL-HRHOLTER× 100) between DGDT and DG or DGDT and DT with Wilcoxon's signed rank tests and between diltiazem and digoxin with a rank sum test.

We used Wilcoxon's signed rank tests to examine whether differences in VR control between monotherapy and combination therapy could be due to differences in either oral dose (mg/kg) or serum drug concentrations (ng/ml). The median and range for drug dosages and serum concentrations are reported. We used Spearman's rank correlation to examine whether VR control achieved by either drug was associated with serum drug concentration.

We used Bonferroni-corrected Wilcoxon's signed rank tests to compare the ability of each treatment to maintain VR within the prespecified VR ranges, except when comparing DG and DT groups, where we used Bonferroni-corrected rank sum tests.

In all analyses, to conserve the experiment-wise error rate (defined as the overall likelihood of committing a type-I error in a study by performing multiple independent statistical tests), a P value of <.008 was considered significant.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

Patient Characteristics and Study Design

Twenty-four large-breed dogs were initially recruited; 4 dogs were excluded because BL-HRHOLTER was <140 bpm, 1 because body weight was <15 kg, and 1 dog died suddenly during the 1st phase of the study (after completing the baseline examination). The remaining 18 dogs all received DGDT, 11 dogs received DG, and 9 dogs received DT. Two dogs received all 3 treatments (Fig 1). These breeds were represented as Doberman Pinscher (4), German Shorthair Pointer (3), Great Dane and Greyhound (3), and 1 of each of the following: Newfoundland, Bull Mastiff, Rottweiler, St Bernard, German Shepherd mix breed, Shar Pei, and American Bulldog. Ten dogs were diagnosed with dilated cardiomyopathy, 7 dogs with degenerative valve disease, and 1 dog with AV valve dysplasia.

Fifteen dogs were in uncontrolled or poorly controlled CHF at presentation: 6/15 were not on any CHF medications at baseline examination, 2/15 were receiving furosemide alone, and 7/15 dogs were receiving combination diuretic and ACE inhibitor therapy. After acquiring the baseline Holter, all 15 dogs with CHF were stabilized and maintained on CHF therapy. All 15 dogs received furosemide and ACE inhibitors; additionally, 4 of these 15 received spironolactone, whereas 4 of them received pimobendan, and 6 of the 15 received both spironolactone and pimobendan. Three dogs had severe cardiac disease and AF, but no CHF at baseline.

Drug Efficacy Assessed by Holter Recordings

Each treatment significantly decreased 2WK-HRHOLTER from BL-HRHOLTER (DG P < .005, DT P < .005, DGDT P < .0001; Table 1, Fig 2). However, DGDT resulted in a greater reduction in heart rate (60 bpm) than obtained with either DG (24 bpm) or DT (28 bpm) alone (P < .008 for each comparison; Table 1). Additionally, DGDT produced a greater %ΔHR (30%) than either DG (14%) or DT alone (20%) (P < .008 for each comparison; Table 1). The %ΔHR did not differ between DG and DT (P= .54; Table 1). When comparing within-dog responses, DGDT decreased HRHOLTER by an additional 28 bpm over DG (median; range, 7–70 bpm) and by 22 bpm over DT (median; range, 2–49 bpm). All 18 dogs receiving DGDT achieved HRHOLTER < 160 bpm and 12/18 achieved HRHOLTER < 140 bpm. Conversely, only 4/9 dogs receiving DT and 5/11 dogs receiving DG achieved HRHOLTER <160 bpm; 3/9 and 2/11 in these groups, respectively, achieved HRHOLTER < 140 bpm.

Table 1.   Effects of 3 different treatments on VR of dogs with atrial fibrillation (AF).
Treatment (Number of Cases)HRHOLTER (bpm)%ΔHR Median (Range) 95% CI
Baseline Median (Range) 95% CI2-Week Median (Range) 95% CIΔHR Median (Range) 95% CI
  • Different from baseline (P < .001).

  • 95%CI, 95% confidence intervals for median values; HRHOLTER, 24 hours average heart rate determined by Holter monitoring; bpm, beats per minute; ΔHR, change in heart rate from baseline (bpm); %ΔHR, % change in heart rate from baseline. Note that all heart rate differences (ΔHR and %ΔHR) are reported as negative values because each treatment resulted in a reduction of the median heart rates as compared with baseline.

  • Median values within columns with different superscripts (a or b) are significantly different (P < .008).

Digoxin (11)195 (142–215) 172–199164 (119–194)a 144–180−24 (−79, 11)a −35, −12−14 (−40, 8)a −18, −8
Diltiazem (9)185 (142–234) 148–202158 (114–178)a 119–177−28 (−69, −4)a −52, −7−20 (−29, −2)a −27, −4
Digoxin-diltiazem (18)194 (142–234) 162–199126 (93–158)b 112–146−60 (−100, −29)b −69, −44−30 (−47, −16)b −36, −28

Figure 2.  Real-time (A) and normalized (B) hourly ventricular rate (VR) profiles of the 24-hour Holter recordings in dogs with atrial fibrillation. Median values of hourly average VR are shown, and the 4 curves represent baseline ♦ (BL, n = 18), digoxin ▪ (DG, n = 11), diltiazem ▴ (DT, n = 9), or digoxin and diltiazem • (DGDT, n = 18) therapy. (A) The unshaded area represents the period in the home environment for nearly all dogs. During the periods defined by the shaded portions, patients are in various environments (ie, hospital, travelling from/to hospital), and variable numbers of dogs are included during these periods (see Appendix 1 for details). (B) Shaded portions show a decrement (left) or increment (right) in VR, most likely associated with stress of the hospital visit. The unshaded area most likely represents VR in home environments during normal activity. All treatments reduced VR from baseline (P < .008), and DGDT reduced VR the most (P < .008). Additional descriptive statistics have been omitted from the figures for clarity, but are detailed in Appendix 1.

Download figure to PowerPoint

Neither the DG dosage (0.005 mg/kg PO q12h; range, 0.0035–0.0071 mg/kg) nor the serum DG concentration (0.8 ng/mL; range, 0.27–1.57 ng/mL) during DG monotherapy differed from the digoxin dosage (0.005 mg/kg PO q12h; range, 0.0039–0.0058 mg/kg) or serum DG concentration (1.27 ng/mL; range, 0.49–2.9 ng/mL) in the same dogs during DGDT therapy (P= .7 and .11, respectively). Serum concentrations remained within the therapeutic range (0.5–2 ng/mL) in all but 1 dog, and none of the dogs in the study showed clinical signs of DG toxicity. Similarly, the DT dosage during DT monotherapy (3.21 mg/kg PO q12h; range, 2.67–3.50 mg/kg) did not differ from the DT dosage in the same dogs during DGDT therapy (3.29 mg/kg PO q12h; range, 2.4–5.65 mg/kg; P= .8). Technical problems (substance interference reported by the laboratory) restricted analysis of DT concentrations to 9/18 dogs. The DT concentration in these 9 dogs was 32 ng/mL (range, 10–134 ng/mL) while they received the DGDT therapy.

VR Control over Time

Figure 2 displays the median hourly average VR both in real time (Fig 2A) and normalized to the time of Holter application (Fig 2B) for each treatment group over the 24-hour period. To preserve figure clarity, specific descriptive statistics for these data are presented in Appendix 1. The unshaded portion in Figure 2A (6:00 pm to 12 noon the following day) is highlighted because nearly all dogs in all treatment phases had Holter data during this period and all dogs were in their home environment and at least 1 hour of Holter recording had been acquired by this time. Thus, this period most likely represents the VR control that might be anticipated in a home environment. All 4 groups showed similar profiles, which were vertically separated. An initial steep decrement in VR was apparent in the normalized data, with an increment in the terminal hours in the baseline and DG groups (Fig 2B, shaded regions). We initiated >70% of Holter recordings after 14 hours; thus, the unshaded period represents the period from early evening to the following noon for most dogs. In agreement with the median HRHOLTER reported in Table 1, the normalized hourly baseline VR was consistently >180 bpm and showed relatively little hour-to-hour variation after the initial decrease. Normalized DG and DT monotherapy profiles indicated more erratic variation in VR over the 24-hour period, whereas DGDT therapy had a relatively smooth profile after the initial decrease (Fig 2B).

Figure 3 displays the percentage of time VR remained in each category with each treatment over the 24 hours of recording. Within each VR range, DGDT therapy differed significantly from baseline (P= .0002, .003, and .003). Additionally, dogs receiving DGDT therapy spent significantly less time at VR > 140 bpm than dogs receiving either DG or DT monotherapy (P= .005 and .008).


Figure 3.  Box-and-whisker diagram demonstrating the ability of each treatment to maintain ventricular rate (VR) in specific ranges. Data are displayed as median percentage of time that the VR spends in each range (black line); box limits represent the interquartile range, and whiskers represent the maximum and minimum values. Note that in the VR range <100 bpm, the baseline and digoxin groups have median and interquartile values of 0%. BL, baseline (n = 18); DG, digoxin (n = 11); DT, diltiazem (n = 9); DGDT, digoxin-diltiazem (n = 18) therapy. Horizontal bars inline image connect groups that are significantly different (P < .008).

Download figure to PowerPoint

The median minimum hourly VR at baseline was 90 bpm (range, 54–150 bpm) and all treatment groups experienced decreased median minimum hourly average VR (DG: 68 bpm, 52–90 bpm; DT: 67 bpm, 53–95 bpm; DGDT: 53 bpm, 40–104 bpm). The minimum hourly VR were all recorded between midnight and 6:00 am (“24” and “6,” respectively, in Fig 2A), and were considered appropriate for sleep-associated bradycardia.

Relationship of DG or DT Concentrations with VR Control

During both DG monotherapy and DGDT therapy, we could find no correlation between serum DG concentration and HRHOLTER. (P= .9 and .67).

Similarly, during DGDT therapy, we could find no correlation between serum DT concentration and HRHOLTER. (P= .76). In the DT monotherapy group, the DT concentrations were not available because of technical errors at the commercial laboratory and thus could not be correlated to HRHOLTER.

Adverse Effects

No dogs developed signs of DG toxicity and no clinically relevant adverse effects associated with DT (eg, weakness, collapse, syncope, and hypotension) were documented. One dog (a Doberman Pinscher) exhibited occasional pauses of up to 4.4 seconds followed by ventricular escape beats while sleeping during baseline recording. These pauses were prolonged with DT monotherapy (longest pause of 5.7 seconds); however, the remaining VR of this dog was comparable to all dogs in the study and no clinical signs of weakness or syncope were reported. No adverse effects were noted in dogs receiving DGDT combination therapy. Most dogs in this study had clinically inconsequential ventricular arrhythmias but no dogs in the study required antiarrhythmic medication for the ventricular arrhythmias.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

Our study demonstrates that in dogs with underlying heart disease and AF with rapid VR, DGDT for 2 weeks results in greater VR reduction than that achieved with either DT or DG alone. Although both drugs given individually produced significant reductions in VR as compared with pretreatment, substantial additional VR reduction was achieved in our patients with combination therapy.

DGDT therapy decreased VR most effectively (Table 1), and maintained VR < 140 bpm for approximately 60% of the time, whereas both monotherapies achieved this level of VR control only 12% of the time. In no patient was VR higher with DGDT therapy than with monotherapy. In humans, this improvement is attributed to an enhanced control of heart rate during exercise.7,9–12 The sustained VR control during periods of regular activity in the home environment (Fig 2A) and the relatively flat normalized hourly VR profile compared with monotherapy (Fig 2B) suggest that VR control during daily activity in dogs may be improved with combination therapy. Calcium channel blockers in canine AF likely complement the vagomimetic effects of DG, consistent with studies in human patients in sinus rhythm.21 Although Bonagura and Ware1 reported favorable effects of co-administration of DG with propranolol in dogs with AF, our study provides evidence that DT might provide a suitable substitute for propranolol.

Although we improved VR control with DGDT therapy, all treatments showed a VR decrement in the hours immediately after Holter application (Fig 2B). This most likely represents a decrease from the high level of stress experienced during the hospital visit, and highlights potential concerns of measuring treatment efficacy in a hospital environment. Thus, excluding this period from analysis might provide a more physiologically representative assessment of VR control. We highlighted the period spent in the home environment in Figure 2A (6:00 pm to 12 noon), although our analysis was based on 24-hour averages, because 24-hour averages are more easily obtained with routine clinical Holter analysis, and would be more relevant to clinicians. Additionally, the hourly profiles demonstrate that differences between treatments, assessed by HRHOLTER, were not because of marked transient changes in VR in the DGDT group that might have skewed the HRHOLTER data enough to produce a difference; rather, these effects persisted over the entire recording period.

Although the dogs in this study received standard DG dosages (0.005 mg/kg q12h), and maintained serum concentrations within the therapeutic range (0.9 ng/mL), DG therapy decreased heart rate significantly, but only modestly. All our patients receiving DG monotherapy had CHF, and their VR remained >140 bpm for almost 90% of time with DG therapy. Less than 50% of dogs achieved the traditionally accepted satisfactory (in-hospital ECG-derived) VR control (140–160 bpm) with DG therapy alone.4 This is consistent with findings in human patients with AF and CHF in whom markedly increased sympathetic tone overrides the vagomimetic effects of DG.7,22 A lack of correlation between serum DG concentration and HRHOLTER argues against the hypothesis that higher serum concentrations might have improved VR control, consistent with findings in humans, where even toxic concentrations fail to achieve appropriate VR control.6,10

DT monotherapy decreased the VR comparable to DG monotherapy. This parallels observations in humans with AF in whom moderate-dose DT monotherapy inadequately controlled rate,7,9 but attenuated exercise-induced VR “break-through” events better than DG.7,10 Similar to DG monotherapy, DT monotherapy resulted in VR > 140 bpm for almost 90% of the time, and both monotherapies differed from DGDT therapy. The wide VR range for the DT group shown in Figure 3 was because of 3 dogs in this group that did not have CHF. When these were excluded from analysis, the range resembled the group receiving DG (data not shown).

Whether higher doses of DT would safely improve VR control remains to be determined. Based on preliminary observations of the extended-release DT used in this study, we selected and achieved a well-tolerated target dosage of 3 mg/kg. We found no correlation between DT serum concentrations and HRHOLTER with DGDT therapy. In normal anesthetized Beagle dogs with experimentally induced AF, acute oral administration of DT at 5 mg/kg (but not 2.5 mg/kg) decreased VR to pre-AF values.8 However, because pharmacokinetics and pharmacodynamics of various DT preparations differ, direct dose comparisons are inappropriate. Nevertheless, “high-dose” DT therapy was superior to “low-dose” (360 versus 240 mg/day) as assessed by ECGs at rest and during exercise in human patients with AF.9

Similar to previous reports in dogs,8 DT concentrations measured in our study would be considered subtherapeutic by human standards (50–200 ng/mL), possibly due to extensive first-pass clearance of oral DT in dogs.23 Interestingly, the dog with the greatest VR control with DGDT therapy also had the highest DT dose and serum concentration, whereas the dog with the least VR control had the second-lowest DT concentration, but a near-median oral dose. Most importantly, however, with DGDT therapy, the DT serum concentrations we measured exerted a significant effect on HRHOLTER. These serum concentrations, therefore, represent effective therapeutic concentrations for dogs with severe cardiac disease, when used with DG. Although DT dose titration might have resulted in even greater VR control, the study design limited our ability to manipulate the DT dose.

The optimal target heart rate for dogs with high-rate AF secondary to severe heart disease and its assessment are poorly defined. Ideally, the lowest VR achievable without hemodynamic compromise is desired. Recent guidelines in humans recommend a target VR similar to a “situation-appropriate” sinus rate: 60–80 bpm at rest and 90–115 bpm during mild exercise18,19 in NYHA I and II patients. Surprisingly, target rates in NYHA III or IV human patients are poorly defined, but are generally tailored to performance on 6-minute walk tests, and approximate VR of similar patients in sinus rhythm. Our study shows that HRHOLTER < 140 bpm is easily achieved in this population of dogs without increased morbidity. Indeed, the median HRHOLTER on DGDT therapy was 126 bpm (and 128 bpm for the CHF dogs), and only 6/18 dogs (5/15 CHF dogs) had HRHOLTER > 140 bpm. Thus, our target VR may have actually been conservative, and more aggressive VR control might be warranted in these patients. However, this ultimately needs to be demonstrated by improvements in both hemodynamic variables (eg, cardiac output, atrial pressures, and oxygen delivery) and clinical variables (eg, decreased morbidity and mortality), which our study did not address.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

The study involved only 18 dogs in a somewhat complex crossover design. Only 2 dogs received all treatments, but all remaining dogs received DG or DT and DGDT therapy in a randomized order. Thus, as some of our data illustrate, some variables had large variation. Although this strengthens the differences we observed, some differences might have been missed because of relatively low power. Nevertheless, our data provide clinically important and relevant findings on VR control in AF.

The study results are applicable only to large-breed dogs with high-rate AF, which constitute a clinically relevant population. Because we have no reason to suspect differences in pharmacodynamics of DT or DG, in dogs of different sizes, we would expect similar outcomes in all dogs that have high-rate AF secondary to severe heart disease.

Baseline Holter recordings were acquired in most dogs while CHF was present, which might contribute to excessively high VR in those dogs. It is likely that stabilization of uncontrolled CHF (with diuretics, ACE-inhibitors, and pimobendan) may have contributed to the effect on VR observed at the first-treatment Holter. However, randomly assigning dogs to each treatment arm would control for this factor. Indeed, as Figure 1 illustrates, similar numbers of dogs received each of the treatment options as their first treatment, suggesting that any effect of CHF medication on VR would be equally distributed.

Drug doses were predetermined, rather than being optimized on a per-patient basis, to simplify study design. The doses selected represented clinically relevant doses, and we feel that the comparisons are valid. It is possible that greater VR reductions could have been achieved with dosages more specifically tailored to each patient. Nevertheless, the VR reductions with combination therapy demonstrated that our a priori standardized approach results in responses that have been considered “clinically appropriate” as suggested previously.4

The lack of correlations between serum concentrations and HRHOLTER should be viewed cautiously, because they are based on small sample populations and relatively stringent statistical criteria (P < .008). Larger pharmacodynamic studies are required to substantiate our initial observations.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

We have shown that the combination of DG and DT produces a greater reduction in VR in dogs with high-rate AF secondary to severe heart disease than either DG or DT administered individually. Combination therapy consistently (but not universally) decreased VR to <140 bpm as determined by 24-hour average heart rate. Although a decrease in VR is easily measured, it remains a surrogate end-point. The definitive endpoint of effective VR control should be an improvement in clinical signs, quality of life, and mortality. Future clinical studies that investigate how VR control prolongs survival and improves clinical signs in dogs with spontaneous AF are warranted.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

a Hamlin R. Clinical use of Diltiazem for dogs with atrial fibrillation. Proc Am Coll Vet Intern Med 1988:759

b Random Number Generator for Excel (add-in for Microsoft Excel 2007), Microsoft Office 2003, Microsoft Corporations, Bellevue, WA

c Lanoxin; Glaxo Smithkline, Research Triangle Park, NC

d Dilt-XR 240 mg (diltiazem hydrochloride extended release), Apotex Inc, Toronto, ON, distributed by Apotex Corp, Weston, FL

e Abbott Laboratories, North Chicago, IL

f Medtox Laboratories Inc, St Paul, MN

g Spacelabs Burdick Inc, Deerfield, WI

h Statistix 8.0, Tallahassee, FL


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix

The authors thank Hollis N. Erb for invaluable statistical consultation and for offering comments on the manuscript. This study was supported in part by grants from the Doberman Pinscher Foundation and Cornell University.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix
  • 1
    Bonagura JD, Ware WA. Atrial fibrillation in the dog: Clinical findings in 81 cases. J Am Anim Hosp Assoc 1986;22:111120.
  • 2
    Calvert CA, Pickus CW, Jacobs GJ, et al. Signalment, survival, and prognostic factors in Doberman Pinschers with end-stage cardiomyopathy. J Vet Intern Med 1997;11:323326.
  • 3
    Vollmar AC. The prevalence of cardiomyopathy in the Irish Wolfhound: A clinical study of 500 dogs. J Am Anim Hosp Assoc 2000;36:125132.
  • 4
    Kittleson KD, Kienle RD. Small Animal Cardiovascular Medicine. St Louis, MO: Mosby Inc; 1998:473.
  • 5
    Cain ME, Curtis AB. Rhythm control in atrial fibrillation—one setback after another. N Engl J Med 2008;358:27252727.
  • 6
    Goldman S, Probst P, Selzer A, et al. Inefficacy of “therapeutic” serum levels of digoxin in controlling the ventricular rate in atrial fibrillation. Am J Cardiol 1975;35:651655.
  • 7
    Farshi R, Kistner D, Sarma JS, et al. Ventricular rate control in chronic atrial fibrillation during daily activity and programmed exercise: A crossover open-label study of five drug regimens. J Am Coll Cardiol 1999;33:304310.
  • 8
    Miyamoto M, Nishijima Y, Nakayama T, et al. Acute cardiovascular effects of diltiazem in anesthetized dogs with induced atrial fibrillation. J Vet Intern Med 2001;15:559563.
  • 9
    Roth A, Harrison E, Mitani G, et al. Efficacy and safety of medium- and high-dose diltiazem alone and in combination with digoxin for control of heart rate at rest and during exercise in patients with chronic atrial fibrillation. Circulation 1986;73:316324.
  • 10
    Maragno I, Santostasi G, Gaion RM, et al. Low- and medium-dose diltiazem in chronic atrial fibrillation: Comparison with digoxin and correlation with drug plasma levels. Am Heart J 1988;116:385392.
  • 11
    Koh KK, Kwon KS, Park HB, et al. Efficacy and safety of digoxin alone and in combination with low-dose diltiazem or betaxolol to control ventricular rate in chronic atrial fibrillation. Am J Cardiol 1995;75:8890.
  • 12
    Koh KK, Song JH, Kwon KS, et al. Comparative study of efficacy and safety of low-dose diltiazem or betaxolol in combination with digoxin to control ventricular rate in chronic atrial fibrillation: Randomized crossover study. Int J Cardiol 1995;52:167174.
  • 13
    Marino DJ, Matthiesen DT, Fox PR, et al. Ventricular arrhythmias in dogs undergoing splenectomy: A prospective study. Vet Surg 1994;23:101106.
  • 14
    Miller RH, Lehmkuhl LB, Bonagura JD, et al. Retrospective analysis of the clinical utility of ambulatory electrocardiographic (Holter) recordings in syncopal dogs: 44 cases (1991–1995). J Vet Intern Med 1999;13:111122.
  • 15
    Mason JW. A comparison of electrophysiologic testing with Holter monitoring to predict antiarrhythmic-drug efficacy for ventricular tachyarrhythmias. Electrophysiologic Study versus Electrocardiographic Monitoring Investigators. N Engl J Med 1993;329:445451.
  • 16
    Morganroth J. Evaluation of antiarrhythmic therapy using Holter monitoring. Am J Cardiol 1988;62:18H23H.
  • 17
    Knoebel SB, Williams SV, Achord JL, et al. Clinical competence in ambulatory electrocardiography. A statement for physicians from the AHA/ACC/ACP Task Force on Clinical Privileges in Cardiology. Circulation 1993;88:337341.
  • 18
    Saxonhouse SJ, Curtis AB. Risk and benefits of rate control versus maintenance of sinus rhythm. Am J Cardiol 2003;91:27D32D.
  • 19
    Fuster V, Rydén LE, Cannom DS, et al. Task Force on Practice Guidelines, American College of Cardiology/American Heart Association; Committee for Practice Guidelines, European Society of Cardiology; European Heart Rhythm Association; Heart Rhythm Society. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation-executive summary: A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients with Atrial Fibrillation). J Am Coll Cardiol 2006;48:e149e246.
  • 20
    Bland M. Estimation. In: BlandM, ed. An Introduction to Medical Statistics. New York: Oxford Universitiy Press; 2000:122136.
  • 21
    Mitchell LB, Jutzy KR, Lewis SJ, et al. Intracardiac electrophysiologic study of intravenous diltiazem and combined diltiazem-digoxin in patients. Am Heart J 1982;103:5766.
  • 22
    Tse HF, Lam YM, Lau CP, et al. Comparison of digoxin versus low-dose amiodarone for ventricular rate control in patients with chronic atrial fibrillation. Clin Exp Pharmacol Physiol 2001;28:446450.
  • 23
    Kohno K, Takeuchi Y, Etoh A, et al. Pharmacokinetics and bioavailability of diltiazem (CRD-401) in dog. Arzneimittelforschung 1977;27:14241428.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. Footnotes
  9. Acknowledgments
  10. References
  11. Appendix
Table A1.   Descriptive statistics for hourly average ventricular rates illustrated in Figure 2.
TreatmentClock TimeHolter Time
Time (h)Sample (n)Min25thMedian75thMaxTime (h)Sample (n)Min25thMedian75thMax