Journal Pre-proof

Comparison of single versus seven-day Holter analysis for the identification of dilated
cardiomyopathy predictive criteria in apparently healthy Doberman Pinscher dogs

Tamilselvam Gunasekaran, BVSc & AH, Nicholas B. Olivier, PhD, Robert A. Sanders,
MS

PII: S1760-2734(20)30003-5
DOI: https://doi.org/10.1016/j.jvc.2020.01.003
Reference: JVC 606

To appear in: Journal of Veterinary Cardiology

Received Date: 12 March 2019

Revised Date: 13 January 2020
Accepted Date: 22 January 2020

Please cite this article as: Gunasekaran T, Olivier NB, Sanders RA, Comparison of single versus
seven-day Holter analysis for the identification of dilated cardiomyopathy predictive criteria in
apparently healthy Doberman Pinscher dogs, Journal of Veterinary Cardiology, https://doi.org/10.1016/
j.jvc.2020.01.003.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2020 Elsevier B.V. All rights reserved.

1 Comparison of single versus seven-day Holter analysis for the identification of

2 dilated cardiomyopathy predictive criteria in apparently healthy Doberman

3 Pinscher dogs.

4 Tamilselvam Gunasekaran BVSc & AH; Nicholas B Olivier PhD; Robert A Sanders MS.

5 Department of Small Animal Clinical Sciences, College of Veterinary Medicine,

6 Veterinary Medical Center, 736 Wilson Rd, East Lansing, Michigan 48824, United

7 States of America.

9 Short running title “Seven-day Holter monitoring in asymptomatic Doberman Pinscher

10 dogs”

11

12 Address correspondence to Dr. Robert Sanders at ras@msu.edu.

13

14 Acknowledgements

15 This study was funded by Doberman Pinscher Health Foundation of America and

16 George Ward Fund for pure breed research – College of Veterinary Medicine, Michigan

17 State University Endowed Research Funds.

18

19

20

1
21 Abstract

22 Introduction: The primary objective of this study was to test whether seven-day Holter

23 recording improves the sensitivity of detecting dilated cardiomyopathy (DCM) predictive

24 criteria (DCMp) compared to 24-hour Holter in asymptomatic Dobermans Pinschers

25 (DP).

26 Animals: 28 asymptomatic DP with normal echocardiographic examinations.

27 Methods: Dogs with normal echocardiographic examinations underwent seven-day

28 Holter monitoring. The presence of ≥ 50 ventricular premature complexes and or ≥ one

29 couplet / one triplet / one episode of ventricular tachycardia per 24-hour period was

30 considered positive for DCMp.

31 Results: Five dogs were positive on the first day and an additional six dogs tested

32 positive from days two to seven of the Holter recording. The number of positive dogs

33 detected by four days was significantly different (p = 0.031) compared to the first day

34 Holter.

35 Conclusions: Seven-day Holter recording detected significantly more dogs with DCMp

36 compared to the first day Holter. Follow up studies are warranted to evaluate the long-

37 term accuracy of multiple-day Holter analysis in predicting the development of DCM in

38 DP.

39 Keywords

40 electrocardiography; arrhythmia; screening

41

2
42 Abbreviation Table

DCM dilated cardiomyopathy

DCMp dilated cardiomyopathy predictive criteria

DP Doberman Pinschers

ECG electrocardiography

IQR interquartile range

NT-proBNP N-terminal pro B-type natriuretic peptide

VA ventricular arrhythmia

VPC ventricular premature complex

VT ventricular tachycardia

43

44

45

46

47

48

49

50

51

52

53
3
54 Introduction

55 Dilated cardiomyopathy (DCM) is the most common primary myocardial disease

56 affecting Doberman Pinscher dogs (DP) [1, 2]. It is an inherited, slowly progressive

57 disease with reported prevalence as high as 58.8% and 63.2% in European and North

58 Americana DP dogs respectively [3]. The disease has a long preclinical phase during

59 which most dogs remain asymptomatic. Once clinical signs develop, it usually

60 progresses very quickly, and affected dogs die of congestive heart failure or experience

61 sudden cardiac death [1-3]. Sudden cardiac death, presumably due to ventricular

62 fibrillation, has been reported in about 25-30% of the dogs with DCM irrespective of the

63 echocardiographic findings [4]. Early detection of echocardiographic features of

64 preclinical DCM is important as therapies initiated during this stage can prolong the

65 symptom free survival duration in DP dogs [5,6]. A placebo controlled, multi-center

66 study in DP dogs has shown that therapy with pimobendanb during the preclinical DCM

67 stage can delay the onset of symptoms [5]. Treatment with the angiotensin converting

68 enzyme inhibitor, benazepril has also been shown to delay the progression of preclinical

69 DCM to overt clinical stage [6]. Additionally, early identification of affected dogs may

70 also help in modification of breeding programs [7]. Given the heritable nature of the

71 disease and positive response to drug therapy initiated before the onset of clinical signs,

72 early diagnosis during the preclinical stage of DCM is desirable, so that therapeutic and

73 breeding decisions can be made.

74 Various screening tools have been evaluated, aimed at early diagnosis during the

75 preclinical stage of DCM. A genetic test to detect a mutation of the pyruvate

76 dehydrogenase 4 gene, encoding for a mitochondrial protein involved in energy

4
77 metabolism has been developed in DP with DCM [8]. Dogs that are homozygous or

78 heterozygous for this gene mutation are considered to be at increased risk for

79 developing DCM [8]. However, the association between pyruvate dehydrogenase 4

80 gene mutation and DCM could not be replicated in a larger study performed in

81 European DP dogs [9]. Recently, a variant in the titin gene (DCM2) was reported to be

82 highly associated with the development of spontaneous model of canine DCM [10]. This

83 test is also commercially available for screening DP dogsc. However, due to variable

84 expression and incomplete penetrance of these genes, not all the dogs that have these

85 mutations will eventually develop DCM [7]. Additionally, having a negative genetic test

86 result does not preclude a dog from developing DCM in the future [7]. Biomarkers such

87 as cardiac troponin I and N-terminal pro B-type natriuretic peptide (NT-proBNP) have

88 been evaluated for screening healthy DP for the diagnosis of preclinical DCM [11-13].

89 These studies evaluated the diagnostic accuracy of biomarker levels to detect

90 preclinical DCM based on a pre-specified echocardiographic and or 24-hour ambulatory

91 ECG (Holter) cut-off criteria. Cardiac troponin I and NT-proBNP levels had a sensitivity

92 of 86.6% and 76.5% respectively to detect echocardiographic abnormalities during

93 preclinical DCM stage. But the sensitivity to detect abnormalities on 24-hour Holter

94 recordings was much lower (45.2% for NT-proBNP and 70.5% for cardiac troponin I),

95 limiting the use of these biomarkers as stand-alone screening tools for early diagnosis

96 of DCM [11-13]. Also, NT-proBNP and cardiac troponin I levels are not specific for DCM

97 as they can be elevated in association with other cardiac diseases and in various

98 systemic diseases [13].

99

5
100 Screening for preclinical DCM is also performed using trans-thoracic echocardiography.

101 Various cut off values have been proposed for M-mode obtained left ventricular

102 chamber dimensions for the diagnosis of echocardiographic features of preclinical DCM

103 [5-6]. However, volumetric measurements of the left ventricle obtained using Simpsons

104 method of discs was reported to have better sensitivity compared to M-mode

105 measurements for the diagnosis of echocardiographic features of preclinical DCM in DP

106 dogs [14]. Unfortunately, often the diagnostic yield of stand-alone echocardiography to

107 detect preclinical stage of DCM is low since not all dogs during this stage will have

108 echocardiographic abnormalities [3]. Various studies have shown that ventricular

109 arrhythmias (VA) can occur prior to development of echocardiographic changes during

110 preclinical DCM [1-3]. A study evaluating the prevalence of DCM in DP dogs of different

111 age groups reported that 13% of dogs have isolated echocardiographic abnormalities

112 without VA [3]. Twenty nine percent of dogs have both echocardiographic changes and

113 VA, while 37% of dogs only have VA (detected via 24-hour Holter monitoring) during the

114 preclinical stage of DCM [3]. Considering that a large proportion of dogs have VA during

115 the preclinical stage of DCM, screening for the VAs in an asymptomatic DP provides the

116 best opportunity for early diagnosis [15]. A study in DP with DCM concluded that five-

117 minute ECGs had high specificity to predict occult DCM, but they were highly insensitive

118 to detect ventricular premature complexes (VPCs) compared to a 24-hour Holter

119 monitoring [16]. Advanced high frequency short- term ECG recording may be helpful to

120 diagnose occult DCMd. This type of evaluation requires high fidelity ECG recorders and

121 custom software for high frequency ECG analysis that is not routinely available in the

122 clinical setting. Currently, yearly screening with a 24-hour Holter monitor in addition to

6
123 transthoracic echocardiography is recommended as the gold standard to screen during

124 preclinical stage of DCM [15]. A longitudinal study using 24-hour Holter monitoring in

125 healthy DP without echocardiographic evidence of DCM reported that 100% of dogs

126 with ≥ 50 VPCs in a 24-hour period and 94% of dogs with ≥ one couplet or one triplet or

127 one episode of ventricular tachycardia (VT) in a 24-hour period subsequently developed

128 DCM [17].

129 Despite having better sensitivity compared to short-term ECG to detect occult VA, the

130 diagnostic accuracy of a single 24-hour Holter could be affected by spontaneous day-to-

131 day variation in the frequency and complexity of VA [18,19]. A study in people has

132 demonstrated that seven-day Holter monitoring increased the sensitivity of detecting

133 non-sustained VT episodes compared to a 24-hour Holter recording due to spontaneous

134 day-to-day variability in VA [20]. A seven-day Holter study in boxer dogs diagnosed with

135 arrhythmogenic right ventricular cardiomyopathy reported similar spontaneous day-to-

136 day variability in VPC frequency and complexity [19]. In this study, Boxer dogs with

137 lower VPC frequencies had higher degree of spontaneous day-to-day variability as

138 compared to dogs with higher VPC frequencies [19]. Currently, no information is

139 available in DP dogs regarding the degree of spontaneous day-to-day variability of VA

140 during any stage of DCM. If spontaneous day-to-day variability in VPC frequency and

141 complexity exists in DP during preclinical stage of DCM, it could significantly impact the

142 diagnostic accuracy of a single 24-hour Holter recording. The objective of this study is to

143 evaluate the degree of day-to-day variation in DCM predictive criteria (DCMp) of

144 asymptomatic DP dogs by comparing the number of positive dogs identified by the first

145 24-hour period of the Holter recording to the entire seven-day Holter monitoring.

7
146 Animals, Materials and Methods

147 Client owned DP between 3 - 8 years of age were prospectively evaluated at College of

148 Veterinary Medicine, Veterinary Medical Center, Michigan State University, between the

149 years 2014 - 2017. After thorough review of medical history, dogs that did not have any

150 reported history of collapse, weakness, or other clinical signs suggestive of

151 cardiovascular disease were considered for enrollment. Dogs were excluded if they

152 were receiving any cardiac medications. Dogs were also excluded if they were receiving

153 fish oil, taurine, or L-carnitine on the possibility that these might affect VA frequency.

154 However, dogs were enrolled if a minimum of six weeks wash out period was followed

155 after stopping the above-mentioned supplementation. Lactating, pregnant and animals

156 in estrus were also excluded to avoid any potential effect of reproductive status on the

157 frequency and or complexity of VA [21]. After initial screening through history, dogs

158 underwent routine cardiology evaluation including physical examination, transthoracic

159 echocardiographic evaluation, and seven-day Holter monitoring.

160 Transthoracic echocardiographye was performed by a cardiology resident (TG) under

161 the supervision of a board-certified veterinary cardiologist (RAS) using standard views

162 as previously recommended for screening of DCM in DP dogs [15]. Left ventricular

163 internal end‐diastolic diameter ≥ 47mm and left ventricular internal end‐systolic diameter

164 ≥ 38mm were considered to be abnormal [15]. Left ventricular end systolic and end

165 diastolic volume indices were calculated in dogs where appropriate echocardiographic

166 images were available for calculation using biplane Simpson’s method of discs [14].

167 Dogs that had normal echocardiographic measurements underwent seven-day Holter

168 monitoring.
8
169 Holter monitors were applied to the dogs using previously described techniques [22].

170 The Holter recordings were analyzed using an automated computer softwaref with

171 analyses verified by the cardiology resident (TG) under the supervision of a board-

172 certified veterinary cardiologist (RAS). This verification process involved detection of

173 VPCs that were identified as normal complexes by the automated system and vice

174 versa. For each 24-hour period of the Holter recording, mean heart rate per 24-hour

175 period, the frequency of VPCs, couplets, triplets, and number of VT episodes were

176 tabulated for each dog. For classification of ventricular origin complexes, ventricular

177 complexes occurring within ≤ 100% of prevailing RR interval were classified as VPCs

178 and ventricular complexes occurring at ≥ 150% of the prevailing RR interval were

179 considered as ventricular escape complexes. Ventricular complexes occurring between

180 100 - 150% of the prevailing RR interval were classified as non-premature ventricular

181 complexes. Non-premature ventricular complexes were not included in the analysis of

182 DCMp. Total VPC count per day included both isolated VPCs and VPCs in complex

183 forms. Ventricular tachycardia was defined as an episode of four or more consecutive

184 VPCs occurring at a rate of ≥ 160 beats per minute. Ventricular couplets and triplets

185 were defined as pairs and triples of VPCs that were initiated by a ventricular complex

186 occurring within ≤ 100% of prevailing RR interval respectively. Ventricular couplets and

187 triplets that were initiated by a ventricular complex occurring more than ≥100% of

188 prevailing RR interval were not included in couplet or triplet count. Dogs that had ≥ 50

189 VPCs, and or ≥ one couplet and or ≥ one triplet and or ≥ one VT in a 24-hour period

190 were considered as having positive test for DCMp [14]. Dogs that had < 50 VPCs, no

191 couplets, no triplets or no episodes of VT in a 24-hour period were considered as having

9
192 negative test for DCMp for each day of the Holter recording. For each dog the number of

193 positive DCMp days out of entire seven days were tabulated. Dogs that had positive

194 tests for DCMp on the first 24-hour period of the Holter recording were designated first-

195 day Holter positive. Dogs that had negative test on the first 24-hour period of the

196 recording but had positive test on any other day of the recording were designated as

197 seven-day Holter positive. Dogs that had negative tests for DCMp on all seven days of

198 the Holter recording were classified as seven-day Holter negative.

199 Statistical analysis was performed using a commercial software programg. Statistical

200 significance was set at alpha < 0.05. Normality was tested using Shapiro-Wilk test and

201 by visual inspection of histograms. Normally distributed continuous variables were

202 expressed as mean ± standard deviation. Non-parametric continuous variables were

203 expressed as median with range and or interquartile range (IQR; 25th percentile-75th

204 percentile) as applicable. Categorical variables were reported as frequency and

205 percentages. Wilcoxon rank sum test was used to compare body weight, age, left

206 ventricular internal diameter in diastole, left ventricular internal diameter in systole,

207 fractional shortening and mean heart rate / 24-hour period between the positive and

208 negative dogs for DCMp. The proportion of positive results between the first day and

209 seven-day Holter recording was compared using one tailed McNemar’s exact test.

210 Results

211 Signalment: Flow chart of dogs from study enrolment to Holter outcome in this study is

212 depicted in figure 1. A total of 28 dogs were initially recruited to the study. Two dogs

213 were excluded due to abnormal echocardiographic left ventricular chamber

214 measurements. Two additional dogs were excluded since the QRS morphology during
10
215 suspected ectopic beats could not be clearly differentiated between supraventricular

216 and ventricular origin beats. Twenty-four dogs completed the seven-day Holter study, of

217 which 19 were females (11 spayed) and five were males (one neutered). The median

218 age was 5.12 years (IQR = 4.4 - 6.2 years). The median body weight was 32.7 kg (IQR

219 = 29.75 - 39.5 kg).

220 Holter results: Ventricular premature complexes were detected in 17 of the 24 dogs

221 (71%) during the entire seven-day Holter recording. The maximum VPC frequency per

222 24-hour period ranged from 2 - 2862 (median = 6 VPCs / 24-hour period; IQR = 2 - 47

223 VPCs / 24-hour period). The percentage of dogs and their respective frequency of

224 maximum VPCs per 24-hour period is depicted in table 1. Ten out of 24 (42%) dogs and

225 seven out of 24 dogs (29%) dogs had at least one couplet and at least one triplet per

226 24-hour period of the recording respectively. The fastest instantaneous heart rate of

227 couplet / triplet for each dog ranged from 200 - 300 beats per minute. The median

228 number of couplets / 24 -hour period was three couplets (range = 1 - 225 couplets / 24-

229 hour period; IQR = 2 - 103 couplets / 24-hour period) and the median number of triplets

230 / 24-hour period was two triplets (range = 1 - 91 triplets / 24-hour period; IQR = 1 - 4

231 triplets / 24-hour period) in dogs that had couplets and triplets respectively. Only two

232 dogs had episodes of VT with one dog having one episode and another dog having five

233 episodes of VT within 24-hour period.

234 Thirteen out of 24 dogs (54%) had negative tests for DCMp on all seven days of the

235 Holter recording. Of these 13 dogs, six dogs (46%) had VPCs and their maximum VPC

236 frequency per 24-hour period ranged from 2 - 46 VPCs (median = 5 VPCs / 24-hour

237 period; IQR = 2 - 13 VPCs / 24-hour period). Seven out of these 13 dogs (54%) did not

11
238 have any VPCs during the entire seven-day Holter recording. Eleven dogs had positive

239 tests for DCMp on at least one day of the seven-day Holter recording. Five out of 11

240 dogs (45%) were first day Holter positive. Of the 11 positive dogs, four dogs (36%) had

241 positive tests based on ≥ 50 VPCs / 24-hour period criteria. Seven dogs (64%) had

242 positive tests solely based on ≥ one couplet / triplet / VT criteria. These dogs did not

243 have ≥ 50 VPCs on any day of the seven-day Holter recording. All of the dogs that were

244 positive based on ≥ 50 VPC criteria also had couplets and triplets on at least one day of

245 the seven-day Holter recording. None of the dogs had positive tests based on VT

246 criteria alone. The relationship between the number of DCMp positive dogs and the

247 number of days positive for DCMp out of seven days of the Holter recording is depicted

248 in figure 2. Note that only one dog had positive test for DCMp criteria on all seven days

249 of the Holter recording. Table 2 provides a 2 x 2 contingency table with the number of

250 DP dogs meeting the DCMp criteria according to the results of single versus seven-day

251 Holter recording. Seven-day Holter recording detected significantly more dogs positive

252 for DCMp (p = 0.016) compared to the first day Holter recording. The probability of

253 obtaining a positive result for DCMp improves with longer duration of Holter recording

254 (Fig. 3). Four days of Holter recording detects significantly a greater number of positive

255 dogs for DCMp criteria compared to the first day of the Holter recording (p = 0.031) (Fig.

256 3). The number of dogs that had positive tests for DCMp criteria was not significantly

257 different between four days and seven days of Holter recording (Fig. 3). Supplementary

258 table A (table available in Supplementary Material online) provides the baseline data

259 comparison between the dogs that are positive and negative for DCMp criteria. Of all the

260 baseline variables compared, only left ventricular internal diameter in diastole (p = 0.04)

12
261 was significantly higher in dogs that were positive for DCMp compared to negative dogs

262 (see supplementary table A). Percent day to day spontaneous variability in VPC

263 frequency was calculated using the formula previously reported for Boxer dogs with

264 arrhythmogenic right ventricular cardiomyopathy [19]. Table 3 provides the percent day

265 to day variation in the number of VPCs for each dog in the study.

266 Seven out of 11 positive dogs (64%) had ventricular escape complexes (range = 0 - 154

267 / 24-hour period; median = 2 / 24-hour period; IQR = 0 -15 / 24-hour period) and ten out

268 of 11 positive dogs (91%) had non-premature ventricular ectopic complexes (range = 0 -

269 467 / 24-hour period; median = 3 / 24-hour period = IQR: 0 - 152 / 24-hour period). Out

270 of the 13 dogs that were negative for DCMp, one dog had ventricular escape complexes

271 (two total) and two dogs had non-premature ventricular complexes (three and eight

272 beats / 24-hour period). Addition of these late coupled ventricular complexes did not

273 change the total number of dogs positive for DCMp. However, in one dog the number of

274 positive DCM days increased from five to six days.

275 Discussion

276 Our results show that seven-day Holter recording detects a greater number of dogs with

277 DCMp criteria than a single 24-hour Holter assessment. This finding is consistent with

278 several human studies that showed longer duration Holter monitoring detects more

279 patients with arrhythmic events [20, 23]. The increased sensitivity of seven-day Holter

280 recoding is due to very high spontaneous day-to-day variation of VPC frequency and

281 complexity noted in the study dogs. For example, only one dog (one out of 11 dogs

282 positive for DCMp) had positive test for DCMp on all seven days of the Holter recording.

283 If a routine, single 24-hour Holter recording is performed, only the above-mentioned dog
13
284 will have a 100% probability of testing positive for DCMp, irrespective of the day when

285 the Holter recording is performed. The rest of the positive dogs (ten out 11 dogs positive

286 for DCMp) will have varying probabilities of testing falsely negative for DCMp. Also, note

287 that there is nearly 100% spontaneous day to day variation in VPC frequency even in

288 dogs with high VPC frequency / 24-hour Holter period.

289 Recently published screening guidelines for DCM in DP dogs classified dogs that have

290 50 - 300 VPCs / 24-hour Holter period as DCM suspects and recommended a recheck

291 Holter evaluation within one year of initial evaluation [15]. This recommendation was

292 based on a study that showed improved specificity of diagnosing preclinical DCM with

293 two Holter examination within one year showing ≥ 50 VPCs on a 24-hour Holter

294 recordingi. In this study even a single 24-Holter showing ≥ 50 VPCs had 100%

295 sensitivity for diagnosing preclinical DCM but the specificity was slightly lower (88.1%)

296 compared to two positive Holter examinations (98% specificity). However, the follow up

297 period was only one year, and it is possible that with longer follow up period the

298 specificity of a single positive 24-Holter (≥ 50 VPCs) may improve. The DCMp criteria

299 used in this study was based on a previous, longer duration (two years) longitudinal

300 study that evaluated the association between the initial 24-hour Holter findings and

301 development of overt DCM in healthy DP dogs [17]. In the aforementioned study, 100%

302 of dogs with ≥ 50 VPCs and 94% of dogs with at least one couplet or triplet developed

303 echocardiographic criteria for DCM [4]. It is noteworthy that, 42% of dogs (36 out of 89

304 dogs with known DCM status) in the above-mentioned study did not have any VPCs on

305 their initial Holter recording but they subsequently developed echocardiographic

306 features of DCM [4]. There are various possible explanations why these dogs did not

14
307 have VPCs on their initial Holter recording. One possibility is that some of these dogs

308 had isolated echocardiographic abnormalities during their initial Holter examination and

309 could have subsequently developed VA on follow up evaluations. It is also possible that

310 in some of these dogs, the 24-hour Holter recording was not sensitive enough to detect

311 VA and multiple day Holter screening could have improved the sensitivity of detecting

312 dogs with VA.

313 Our findings have important clinical implications for screening asymptomatic DP dogs

314 during the preclinical stage of DCM. Due to the presence of spontaneous variation in

315 VPC frequency and complexity, classifying DCMp status based on a single 24-hour

316 Holter recording may not actually represent the true arrhythmia status in these dogs.

317 This can impact therapy and breeding decisions. Our study also found that four days of

318 Holter recording detected significantly more positive dogs compared to the first 24-hour

319 period of the Holter recording. Further increase in the recording duration (from day four

320 to seven days) did not significantly increase the number of positive dogs for DCMp.

321 Considering the cost associated with longer duration Holter screening a total of four

322 days might be a more practical and economical option for owners screening DP dogs

323 for occult DCM.

324 In our study, ventricular escape complexes and non-premature ventricular complexes

325 were not included in the VPC frequency count. These complexes were predominantly

326 noted during times of slow heart rate periods. The classification of ventricular

327 complexes based on the prevailing RR interval in dogs could be affected by the

328 presence of significant sinus arrhythmia. Therefore, some of the non-premature

329 ventricular complexes could actually represent VPCs especially late diastolic VPCs.

15
330 Interestingly, these non-premature ventricular complexes and ventricular escape

331 complexes were noted almost exclusively in dogs that had positive tests for DCMp. Only

332 two of the negative dogs for DCMp had these non-premature ventricular complexes.

333 Further long-term follow-up studies are needed to evaluate the prognostic significance

334 of these non-premature ventricular complexes and their association to various stages of

335 DCM.

336 In our study, of the 11 dogs that were positive for DCMp, only four dogs tested positive

337 based on ≥ 50 VPCs / 24-hour period criteria. As previously reported 100% of dogs with

338 ≥ 50 VPCs / 24-hour period on a Holter recording will go onto develop overt DCM [17].

339 However, this predictive value only applies to a 24-hour Holter recording and cannot be

340 directly extrapolated to a seven-day Holter recording. It is possible that not all the dogs

341 that had positive tests based on ≥ 50 VPCs on a seven Holter recording will go on to

342 develop DCM in the future. Additionally, seven dogs tested positive solely based on

343 couplet / triplet and VT criteria. These dogs did not have ≥ 50 VPCs on any day of the

344 seven-day Holter recording. The couplet / triplet / VT criteria have been shown to have a

345 slightly lower predictive value (94%) compared to the ≥ 50 VPCs criteria (100%

346 predictive value) [14]. Considering the lower predictive value for the couplet / triplet / VT

347 criteria, seven-day Holter monitoring could be overly sensitive in detecting VA

348 complexity, resulting in false positive results. However, recently published screening

349 guidelines for preclinical DCM in DP dogs recommended a recheck evaluation in any

350 dogs that show complexity in VA (couplet, triplet, or VT at high instantaneous rate)

351 irrespective of the number of VPCs, highlighting the importance of VA complexity in

352 occult DCM diagnosis [13]. At this time, it is unknown if all the dogs that tested positive
16
353 for DCMp on a seven-day Holter recording will go on to develop overt DCM, but it is

354 reasonable to consider more aggressive follow up of these dogs. Also, some the dogs

355 that tested negative for DCMp criteria in this study could have been evaluated prior to

356 development of arrhythmogenic or echocardiographic manifestations of DCM.

357 Therefore, long term follow-up studies are needed to risk stratify asymptomatic DP dogs

358 based on their DCMp status and number of positive DCMp days on a seven-day Holter

359 recording. In our study comprehensive diagnostic workup to screen for any systemic

360 cause for VA was not performed. Therefore, non-cardiac cause for these VAs due to

361 subclinical systemic disease is not completely ruled out in some dogs.

362 Conclusions

363 Multiple day Holter recording detects a greater proportion of DCMp positive dogs than

364 single 24-hour Holter assessment. The probability of obtaining a positive result for

365 DCMp is significantly improved with four days of Holter recording. Further long-term

366 follow-up studies are warranted to evaluate the accuracy of a multiple day Holter

367 recording in predicting the development of DCM in the future.

368 Conflicts of interest statement

369 The authors have no conflicts of interest to declare.

370

371

372

373

17
374 Footnotes

375 a. O'Grady M.R., Horne R. The prevalence of dilated cardiomyopathy in Doberman

376 Pinschers: A 4.5-year follow-up (abstract). J Vet Intern Med 1998; 12:199.

377 b. Pimobendan, Boehringer Ingelheim, St. Joseph, Minnesota, United States

378 c. Veterinary Genetics Laboratory at North Carolina State University webpage and

379 webinar on dilated cardiomyopathy: https://cvm.ncsu.edu/genetics/doberman-pinscher-

380 dilated-cardiomyopathy.

381 d. Spiljak MS, Petric DA, Wiberg M, et al. Advanced electrocardiography can identify

382 occult cardiomyopathy in Doberman pinschers. September 2011, 21st ECVIM-CA

383 Congress, Unpublished conference paper, 2011. Seville, Spain. Available online

384 through Veterinary Information Network at http://www.vin.com/doc/?id=5072308

385 e. Vivid E90, General Electric Medical System, Waukesha, Wisconsin, United States

386 f. Pathfinder SL, Spacelab’s Healthcare, Snoqualmie, Wisconsin, United States

387 g. Stata, StataCorp LP, College Station, Texas, United States

388 i. Geraghty N, Wess G. Vergleich verschiedener Holterkriterien zur Diagnose des

389 arrhythmischen Stadiums der dilatativen Kardiomyopathie beim Dobermann.

390 Tierärztliche, Facultät der LMU München; 2011. p. 1-107.

391 References

392 [1] O'Grady MR, O'Sullivan ML. Dilated cardiomyopathy: an update. Vet Clin North Am

393 Small Anim Pract 2004;34:1187–207.

18
394 [2] Tidholm A, Haggstrom J, Borgarelli M, Tarducci A. Canine idiopathic dilated

395 cardiomyopathy. Part I: aetiology, clinical characteristics, epidemiology, and pathology.

396 Vet J 2001;162:92–107.

397 [3] Wess G, Schulze A, Butz V, Simak J, Killich M, Keller LJ, Maeurer J, Harmann K.

398 Prevalence of dilated cardiomyopathy in Doberman pinschers in various age groups. J

399 Vet Intern Med 2010;24:533–8.

400 [4] Calvert CA, Jacobs G, Pickus CW, Smith DD. Results of ambulatory

401 electrocardiography in overtly healthy Doberman pinschers with echocardiographic

402 abnormalities. J Am Vet Med Assoc 2000;217:1328–32.

403 [5] Summerfield NJ, Boswood A, O'Grady MR, Gordon SG, Dukes-McEwan J, Oyama

404 MA, Smith S, Patteson M, French AT, Culshaw GJ, Braz-Ruvivo L, Estrada A,

405 O’Sullivan ML, Loureiro J, Willis R, Watson P. Efficacy of Pimobendan in the Prevention

406 of Congestive Heart Failure or Sudden Death in Doberman Pinschers with Preclinical

407 Dilated Cardiomyopathy (The PROTECT Study). J Vet Intern Med 2012;26:1337–49.

408 [6] O'Grady MR, O'Sullivan ML, Minors SL, Horne R. Efficacy of benazepril

409 hydrochloride to delay the progression of occult dilated cardiomyopathy in Doberman

410 Pinschers. J Vet Intern Med 2009;23:977–83.

411 [7] Dutton E, Lopez-Alvarez J. An update on canine cardiomyopathies - is it all in the

412 genes? J Small Anim Pract 2018;59:455-64.

413 [8] Meurs KM, Lahmers S, Keene BW, White SN, Oyama MA, Mauceli E, Lindblad-Toh

414 K. A splice site mutation in a gene encoding for PDK4, a mitochondrial protein, is

19
415 associated with the development of dilated cardiomyopathy in the Doberman pinscher.

416 Hum Genet 2012;131:1319-25.

417 [9] Owczarek‐Lipska M, Mausberg T, Stephenson H, Dukes‐McEwan J, Wess G, Leeb

418 T. A 16‐bp deletion in the canine PDK4 gene is not associated with dilated

419 cardiomyopathy in a European cohort of Doberman Pinschers. Anim Genet 2013:44;

420 239-49.

421 [10] Meurs KM, Friedenberg SG, Kolb J, Saripalli C, Tonino P, Woodruff K, Olby NJ,

422 Keene BW, Adin DB, Yost OL, DeFrancesco TC, Lahmers S, Tou S, Shelton GD,

423 Granzier H. A missense variant in the titin gene in Doberman pinscher dogs with familial

424 dilated cardiomyopathy and sudden cardiac death. Hum Genet 2019:138:515-24.

425 [11] Wess G, Butz V, Mahling M, Hartmann K. Evaluation of N-terminal pro-B-type

426 natriuretic peptide as a diagnostic marker of various stages of cardiomyopathy in

427 Doberman Pinschers. Am J Vet Res 2011;72:642-9.

428 [12] Singletary GE, Morris NA, O'Sullivan LM, Gordon SG, Oyama MA. Prospective

429 evaluation of NT-proBNP assay to detect occult dilated cardiomyopathy and predict

430 survival in Doberman Pinschers. J Vet Intern Med 2012;26:1330-6.

431 [13] Wess G, Simak J, Mahling M, Hartmann K. Cardiac troponin I in Doberman

432 Pinschers with cardiomyopathy. J Vet Intern Med 2010;24:843-9.

433 [14] Wess G, Maurer J, Simak J, Hartmann K. Use of Simpson's method of disc to

434 detect early echocardiographic changes in Doberman Pinschers with dilated

435 cardiomyopathy. J Vet Intern Med 2010;24:1069-76.

20
436 [15] Wess, O. Domenech, J. Dukes-McEwan, Haggstrom J, Gordon S. European

437 Society of Veterinary Cardiology screening guidelines for dilated cardiomyopathy in

438 Doberman Pinschers. J Vet Cardiol 2017;19:405-15.

439 [16] Wess G, Schulze A, Geraghty N, Hartmann K. Ability of a 5-Minute

440 Electrocardiography (ECG) for Predicting Arrhythmias in Doberman Pinschers with

441 Cardiomyopathy in Comparison with a 24-Hour Ambulatory ECG. J Vet Intern Med

442 2010;24:333–71.

443 [17] Calvert CA, Jacobs GJ, Smith DD, Rathbun SL, Pickus CW. Association between

444 results of ambulatory electrocardiography and development of cardiomyopathy during

445 long-term follow up of Doberman pinschers. J Am Vet Med Assoc 2000;216:34-9.

446 [18] Toivonen L. Spontaneous variability in the frequency of ventricular premature

447 complexes over prolonged intervals and implications for antiarrhythmic treatment. Am J

448 Cardiol 1987;60:608-12.

449 [19] Spier AW, Meurs KM. Evaluation of spontaneous variability in the frequency of

450 ventricular arrhythmias in Boxers with arrhythmogenic right ventricular

451 cardiomyopathy. J Am Vet Med Assoc 2004;224:538–41.

452 [20] Pastor-Perez FJ, Manzano-Fernandez S, Goya-Esteban R, Pascual-Figal DA,

453 Barquero-Perez O, Rojo-Alvarez JL, Martinez-Espejo MD, Chavarri MV, García-

454 Alberola A. Comparison of detection of arrhythmias in patients with chronic heart failure

455 secondary to non-ischemic versus ischemic cardiomyopathy by 1 versus 7-day Holter

456 monitoring. Am J Cardiol 2010;106:677-81

21
457 [21] Cordina R, McGuire MA. Maternal cardiac arrhythmias during pregnancy and

458 lactation. Obstet Med 2010;3:8-16.

459 [22] Sanders RA, Kurasowa TA, Sist MD. Ambulatory electrocardiographic evaluation

460 of the occurrence of arrhythmias in healthy Salukis. J Am Vet Med Assoc 2018;252:966-

461 9.

462 [23] Dagres N, Kottkamp H, Piorkowski C, Weis S, Arya A, Sommer P, Bode K, Gerds-

463 Li JH, Kremastinos DT, Hindricks G. Influence of the duration of Holter monitoring on

464 the detection of arrhythmia recurrences after catheter ablation of atrial fibrillation:

465 Implications for patient follow-up. Int J Cardiol 2010;139:305–6.

466

467

468

469

470

471

472

473

474

475

476
22
477 Figure 1

478 Flow chart of seven-day Holter study in healthy Doberman Pinschers with normal

479 echocardiographic parameters. DCM: dilated cardiomyopathy; VPC: ventricular

480 premature complex; DCMp: ≥ 50 VPCs and or ≥ 1 couplet and or ≥ 1 triplet and or ≥ 1

481 VT /24-hour periods of Holter recording; VT: ventricular tachycardia.

482 Figure 2

483 Depicts the number of dogs that are positive for dilated cardiomyopathy criteria based

484 on number of days that they are positive out of seven days of the Holter recording. Note

485 that only one dog is positive on all seven days of the Holter recording. DCM: dilated

486 cardiomyopathy.

487 Figure 3

488 Effect of increasing duration of Holter recording on obtaining positive dilated

489 cardiomyopathy predictive criteria in healthy Doberman Pinscher dogs with normal

490 echocardiographic examinations. Note the increasing probability of obtaining a positive

491 result with increasing duration of the Holter recording. * denotes statistical significance.

23
Table 1

Maximum VPC frequency

Number of dogs Percentage of dogs %
per 24-hour period

0 7 29

1-10 9 38

11-25 2 8

25-49 2 8

≥50 4 17

Maximum frequency of ventricular premature complexes per 24-hour period of the

seven-day Holter recording in healthy Doberman Pinscher dogs with normal

echocardiographic examinations. VPC: ventricular premature complex.

1
Table 2

Seven-day Holter positive Seven-day Holter negative Totals

First day Holter positive 5 0 5

First day Holter negative 6 13 19

Totals 11 13 24

p = 0.016; McNemar’s exact test.

Proportion of positive and negative Doberman Pinscher dogs for dilated cardiomyopathy

predictive criteria as detected by single versus seven-day Holter recording.

1
Table 3

Dog Max VPC Min VPC Median VPC Variability %*

DCMp positive dogs
1 296 0 1 100
2 21 0 11 100
3 2231 37 156 98
4 5 0 1 100
5 2873 610 1262 79
6 47 9 12.5 81
7 2 0 0 100
8 2 0 0 100
9 2 0 0 100
10 819 0 138 100
11 4 0 0 100
DCMp negative dogs
12 0 0 0 -
13 0 0 0 -
14 0 0 0 -
15 0 0 0 -
16 0 0 0 -
17 0 0 0 -
18 0 0 0 -
19 4 0 0 100
20 46 0 0 100
21 2 0 0 100
22 2 0 0 100
23 6 0 0 100
24 13 0 0 100

Spontaneous day to day variation in ventricular premature complex frequency in

apparently healthy Doberman Pinscher dogs with normal echocardiographic

measurements. DCMp: dilated cardiomyopathy predictive criteria; VPC = ventricular

premature complex; Max VPC = maximum number of VPCs per 24 hours on a seven-

1
day Holter; Min VPC = minimum number of VPCs per 24 hours on a seven-day Holter;

Median VPC = median number of VPCs per 24 hours on a seven-day Holter recording.

*variability % = Maximum VPCs/24 hours – Minimum VPCs/24 hours

-----------------------------------------------------------------------

Maximum VPCs/24 hours

2
1
1
1