SEARCH

SEARCH BY CITATION

Keywords:

  • Asymmetric dimethylarginine;
  • Dog;
  • Endothelial dysfunction

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Background

Cardiovascular disease is associated with endothelial dysfunction in humans and studies of plasma biomarkers suggest that dogs with myxomatous mitral valve disease (MMVD) might also have endothelial dysfunction.

Hypothesis

That progression of mitral regurgitation (MR) is associated with development of endothelial dysfunction.

Animals

Forty-three Cavalier King Charles Spaniels (CKCS) with MR of varying severity.

Methods

Privately owned CKCS were prospectively recruited and divided in 4 groups: (1) 12 CKCS with minimal MR; (2) 9 CKCS with mild MR; (3) 11 CKCS with moderate-severe MR; and (4) 11 CKCS with moderate-severe MR and clinical signs compatible with heart failure. Dogs underwent blood sampling, echocardiography, blood pressure (BP) recordings, and flow-mediated vasodilation (FMD) measurements. The effect of progressive MR on FMD was determined by multivariate analyses.

Results

Flow-mediated vasodilation decreased with progression of MR. Group 4 (4.79 ± 3.22%) had significantly lower FMD than groups 1 (10.40 ± 4.58%) and 2 (10.14 ± 3.67%) (< .005) and group 3 (6.79 ± 3.98%) had a significantly lower FMD than group 1 (= .03). Increasing left ventricular end-diastolic diameter (= .0004, R2 = 0.27) and the combination of age (= .01) and body weight (= .002) (R2 = 0.31) were significantly associated with reduced FMD. FMD did not correlate with sex, BP, or plasma markers.

Conclusions and Clinical Importance

Reduced FMD indicates that increased disease severity in CKCS with MMVD is associated with development of endothelial dysfunction which might be a future therapeutic and/or diagnostic target.

Abbreviations
ACE

angiotensin converting enzyme

ADMA

asymmetric dimethylarginine

BP

blood pressure

CKCS

Cavalier King Charles Spaniels

CV

coefficient of variation

CHF

congestive heart failure

FMD

flow-mediated vasodilation

FVI

flow velocity integral

HR

heart rate

LA/Ao

left atrial to aortic root ratio

LVIDd

left ventricular end-diastolic internal dimension

LVIDs

left ventricular end-systolic internal dimension

MAP

mean arterial pressure

MVP

mitral valve prolapse

MR

mitral regurgitation

MMVD

myxomatous mitral valve disease

NOx

nitrate and nitrite

NO

nitric oxide

ROI

region of interest

SDMA

symmetric dimethylarginine

Myxomatous mitral valve disease (MMVD) is a common heart valve disease in dogs,[1, 2] accounting for approximately 7.5% of all dogs presented for veterinary investigation.[3] MMVD has a long preclinical period with increasing severity of mitral regurgitation (MR) before progression to congestive heart failure (CHF). However, the duration of the preclinical period is unpredictable and when CHF occurs survival time varies between 6 and 14 months.[4] Therefore, studies of the pathogenesis are important to improve diagnosis, treatment, and prognosis of this common disease of dogs.

The endothelium maintains normal blood flow by regulating a baseline vasodilatory, antithrombotic, and anti-inflammatory state, and nitric oxide (NO) is an important endothelial mediator involved in these processes.[5] In several human diseases[6-9] including heart failure and valvular disease[10, 11] this homeostatic balance normally maintained by the endothelium could be shifted toward vasoconstriction, coagulation, inflammation, and remodeling of vessels. Endothelial dysfunction can develop in asymptomatic patients[12] and is a strong independent predictor of outcome in CHF patients.[13] Experimentally induced CHF in dogs causes endothelial dysfunction and reduces vasodilation caused by decreased NO availability,[14-16] and endothelial dysfunction could potentially play a role in several spontaneous diseases in dogs.

Measuring NO directly is difficult, and therefore several other biomarkers related to NO generation are often used as markers of endothelial function and NO bioavailability. A few studies measuring plasma biomarkers of endothelial function in relation to the progression of MR in dogs have shown conflicting results,[17-20] possibly because of variation in age, breed, diet, exercise level, kidney function, disease severity, and medication between studies.[21] To clarify the relation between endothelial function and MMVD, a more direct noninvasive technique would be preferable. The standard technique for measuring endothelial dysfunction in human patients is assessment of flow-mediated vasodilation (FMD).[22] The vasodilation occurs as a result of increased vascular shear stress, produced by increased velocity of blood, stimulating endothelial cells to release NO and other vascular mediators.[22] The increased velocity of blood can be assessed as the flow velocity integral (FVI). FMD has only recently been applied and shown to be feasible in healthy dogs.[23, 24] So far no studies of dogs have connected direct measurements of endothelial function with indirect measurements such as plasma markers, which are widely used in human studies.[8, 25] Asymmetric dimethylarginine (ADMA) is an endogenous inhibitor of NO synthase, and several studies in human patients have shown correlations between ADMA or the l-arginine/ADMA ratio and FMD.[6-9] The l-arginine/ADMA ratio decreased in dogs with increasing degree of MMVD; however, this decrease was mainly caused by the effect of increasing age.[18]

The Cavalier King Charles Spaniel (CKCS) is genetically predisposed to developing MR at an early age[2, 26] and approximately 50% of CKCS have a heart murmur caused by MR at the age of 5–6 years.[27] It is a docile breed and has been suggested as a good spontaneous model of MMVD.[28] Endothelial dysfunction might be associated with the development of MMVD and if this is the case it could be used as a future risk assessment tool or to document the benefit of therapeutic interventions.

The hypothesis of the present study was that progression of MR is associated with development of endothelial dysfunction. The aim of the study was to evaluate FMD in CKCS with a range of severity of MMVD and investigate the effects of age, body weight, sex, creatinine, blood pressure (BP), plasma concentrations of ADMA, symmetric dimethylarginine (SDMA) and l-arginine, and echocardiographic parameters of MMVD progression on FMD.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Study Population

The prospective study included 80 privately owned CKCS recruited from a database of breeders. Dogs were examined at the Department of Basic Animal and Veterinary Sciences, Faculty of Life Sciences, University of Copenhagen (N = 72) or at a veterinary cardiology referral practice: Din Veterinaer, Helsingborg, Sweden (N = 8). Dogs with clinically systemic disease other than heart disease, as determined by physical examination, hematology, and serum biochemistry, were excluded from the study. Other than cardiac medication for dogs in the heart failure group, no dog received other medications or was vaccinated in the month before study. Dogs were ≥4 years of age, not pregnant or lactating, and food was withheld for 12–18 hours before study. Dogs were categorized according to their degree of MR (jet size, defined as the maximum area of the mitral regurgitant jet as a percentage of the total area of the left atrium) evaluated echocardiographically (color flow mapping in 2 orthogonal projections including left apical 4 chamber view and right parasternal long axis 4 chamber view)[29] and the absence or presence of clinical signs of CHF and subsequently divided into 4 groups: (1) CKCS with no clinical signs and no or minimal (≤15%) MR, (2) CKCS with no clinical signs and mild (>15% and ≤50%) MR, (3) CKCS with no clinical signs and moderate-to-severe (>50%) MR, and (4) CKCS with moderate-to-severe (>50%) MR and clinical signs compatible with CHF (cough, exercise intolerance, restlessness at night, and/or dyspnea) responsive to furosemide treatment. All owners gave written consent and the study was approved by the Danish Animal Welfare Division.

Examination Procedures

The examination comprised, in order of occurrence, an interview with the owner, collection of blood samples, physical examination, echocardiography, BP measurement, and finally an FMD measurement. None of the dogs were sedated for the examination and the owners were present to calm the dogs. Blood was taken by jugular venipuncture and centrifuged within 30 minutes of collection. Plasma samples were stored at −80°C until analysis.[21] Echocardiography1 was performed by 1 experienced echocardiographer (LHO). Blinded echocardiographic analysis included estimation of degree of MR (%) and left atrial diameter measured by use of 2-D short-axis view for calculation of left atrial to aortic root ratio (LA/Ao).[30] The degree of mitral valve prolapse (MVP) was evaluated in the right parasternal long axis 4-chamber view.[31] Left ventricular dimensions (left ventricular end-diastolic and end-systolic internal dimensions [LVIDd and LVIDs]) were measured from short-axis view by use of 2-D guided M-mode.[32] BP (systolic, diastolic, and mean arterial pressure [MAP]) was measured on the tail artery with the high definition oscillometric technique.2[33]

FMD Image Aquisition

A protocol similar to that described previously was followed.[24] An ultrasound unit1 with a 13 MHz linear array transducer was used to acquire 2-D ultrasonographic images of the right brachial artery. All images were acquired by the same FMD sonographer (SGM) with an experience of approximately 50 FMD scans before the study. Continuous 2-D image recording of all phases of the cardiac cycle for off-line automated edge detection and luminal diameter measurement were performed in defined time periods at constant settings so that a constant number of frames (12 frames per second) were recorded for all dogs. Pulsed-wave spectral Doppler ultrasonography was used to record blood flow in the brachial artery (3 mm sample volume). The mean angle correction used was 67° ± 3.7°.

For each dog, 2-D longitudinal ultrasonographic images of the artery were obtained during a 30-second period, followed by a 10-second recording of arterial blood flow velocity waveforms obtained via pulsed-wave Doppler ultrasonography before cuff inflation. For induction of reactive hyperemia, the brachial artery was occluded for 5 minutes by inflating a blood pressure cuff above 200 mmHg as described previously.[24] A cuff with a width of 5 cm approximately equivalent to the circumference of the distal portion of the antebrachium was used. An additional 10-second recording of spectral Doppler ultrasonographic blood flow velocity waveforms was recorded immediately after cuff deflation, and no later than 15 seconds after cuff deflation 2-D longitudinal ultrasonographic images of the artery were obtained sequentially during the following 3-minute period.

Data were analyzed off-line with commercially available software3 by either 1 of 2 observers (21 analyses by SGM and 22 analyses by CK) as described previously.[24, 34] A region of interest (ROI) was defined manually for each image frame. By use of automated border detecting software, a mean luminal diameter measurement along a segment of the artery was calculated for each frame. Once analysis was complete, the numeric output was exported to a spreadsheet. Figure 1 illustrates the total raw data of 1 dog in the study including baseline and post-occlusion recordings.

image

Figure 1. (A) Data were analyzed off-line using commercially available software. A region of interest was defined manually for each image frame. By use of automated border detecting software, a mean luminal diameter measurement along a segment of artery was calculated for each frame. (B) The total raw data of 1 dog in the study including baseline and post-occlusion recordings. For each dog, mean baseline diameter was calculated from the luminal diameter measurements obtained prior to cuff inflation. The maximum luminal diameter following cuff deflation was defined as the mean of the peak 10 frames (maximum peak ±5 frames) of the post-inflation curve.

Download figure to PowerPoint

Calculation of FMD

For each dog, mean baseline diameter was calculated from the luminal diameter measurements obtained before cuff inflation. The maximum luminal diameter after cuff deflation was defined as the mean of the peak 10 frames (maximum peak ±5 frames) of the post-occlusion curve. The FMD for each dog was calculated as follows: FMD (%) = ([Maximum luminal diameter − mean baseline diameter] / mean baseline diameter) ×100. Time to peak luminal diameter was also determined from the generated post-occlusion curve (Fig 1B).

Calculation of FVI

The FVI (ie, the area under the spectral pulsed-wave Doppler time-velocity curves) was measured in 5 consecutive arterial spectral Doppler ultrasonographic blood flow velocity envelopes throughout the cardiac cycle before cuff inflation and immediately after cuff deflation by use of commercially available software.4 A ROI was defined manually and the FVI was calculated automatically. Percentage change in FVI associated with cuff deflation was calculated as follows: Increase in FVI (%) = ([mean post-occlusion FVI − mean pre-occlusion FVI] / mean pre-occlusion FVI) × 100.

Plasma Analyses

Plasma concentrations of l-arginine, ADMA, and SDMA were determined simultaneously by high-performance liquid chromatography as described previously.[21, 35] The l-arginine, ADMA, and SDMA had an analytical recovery of 98–102% and the interassay coefficient of variation (CV) was <3% for all compounds.

Statistical Analyses

All statistical calculations were performed by statistical software.5 Differences in descriptive data, echocardiographic parameters, ultrasonographic, and plasma markers of endothelial function were tested by a one-way analysis of variance. Body weight, age, sex, heart rate (HR), and serum creatinine were included as covariates when analyzing ultrasonographic and plasma markers of endothelial function. Associations between the measured response variable (FMD) and the explanatory variables including echocardiographic variables (dog group, LA/Ao, MVP, LVIDd, LVIDs), plasma endothelial markers (plasma ADMA, SDMA, and l-arginine concentrations and l-arginine/ADMA ratio), BP measurements (systolic BP, diastolic BP, and MAP), body weight, age, sex, heart rate (HR), and serum creatinine were analyzed by use of multivariate analyses. BP, plasma markers, and echocardiographic variables were entered separately in the models because of high covariance among these variables. If dog group was statistically significant, posthoc groupwise comparisons (Students t-tests) were performed. Reductions were performed in a backward, stepwise manner until only statistically significant effects remained. For each model, the residuals were tested for normality and homogeneity of variation and increase in FVI, time to peak diameter and SDMA concentration were log-transformed to obtain normality of the residuals. Results are shown as mean ± SD or as median values and interquartile intervals when the raw data were not normally distributed. The level of significance was chosen as < .05.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Descriptive Data

Of the 80 CKCS, 4 dogs were excluded from the study because of evidence of clinically important systemic disease other than heart disease, 1 CHF dog was excluded because of lack of furosemide treatment, and 32 dogs were excluded because they were unable to lie still during the FMD ultrasonography. The cardiac medication of the CHF dogs on the day of examination included furosemide (9 dogs), angiotensin converting enzyme (ACE) inhibitors (6 dogs), pimobendan (5 dogs), and spironolactone (3 dogs) either given separately or in combination. Group 4 was significantly older than the other groups and group 3 was significantly older than group 1 (Table 1). The echocardiographic variables LA/Ao, MVP, LVIDd, and LVIDs were significantly elevated in groups 3 and 4 (Table 1).

Table 1. Descriptive and echocardiographic characteristics of the four groups of dogs with MR.
 Group 1Group 2Group 3Group 4
CKCS, No or Minimal MRCKCS, Mild MRCKCS, Moderate-to-Severe MRCKCS in Heart Failure
  1. CKCS, Cavalier King Charles Spaniel; MR, mitral regurgitation; BP, blood pressure; MAP, mean arterial pressure; LA/Ao, Left atrial to aortic root ratio; LVIDd, left ventricular end diastolic diameter; LVIDs, left ventricular end systolic diameter.

  2. Except for sex, data are expressed as mean ± standard deviation.

  3. a

    P < .05 compared to all other groups.

  4. b

    P < .05 compared to groups 1 and 2.

  5. c

    P < .05 compared to groups 1.

n1291111
Sex (m/f)4/82/77/47/4
Age (months)59 ± 1782 ± 3584 ± 28c123 ± 21a
Body weight (kg)9.05 ± 1.429.41 ± 1.9210.16 ± 1.7010.60 ± 1.47
Heart rate (beats per minute)119 ± 17114 ± 13125 ± 18136 ± 13b
Systolic BP (mmHg)155 ± 12151 ± 18149 ± 16159 ± 9
Diastolic BP (mmHg)82 ± 986 ± 883 ± 889 ± 17
MAP (mmHg)107 ± 10107 ± 11106 ± 9104 ± 8
LA/Ao (ratio)1.24 ± 0.131.28 ± 0.101.55 ± 0.16b2.02 ± 0.44a
Mitral valve prolapse (mm)2.36 ± 1.634.44 ± 2.198.27 ± 4.73b10.18 ± 2.71b
LVIDd (mm)28.40 ± 2.0729.18 ± 3.9633.16 ± 3.81b42.39 ± 4.42a
LVIDs (mm)20.26 ± 2.5321.67 ± 2.0022.22 ± 2.03c24.18 ± 2.50b

Markers of Endothelial Function

The baseline luminal diameter was significantly associated with body weight (< .0001, R2 = 0.33, Table 2). Following cuff release, the overall median time to peak luminal diameter was 56 seconds (45–64 seconds). Absolute change in luminal diameter was significantly lower in group 4 compared to groups 1 and 2 (< .01) (Table 2). FMD was significantly lower in group 4 than groups 1 and 2 (< .005) and group 3 was lower than group 1 (P = .03) (Table 2; Fig 2). Increasing LVIDd (= .0004, R2 = 0.27) was also associated with reduced FMD. In analyses including the effects of other echocardiographic variables (LVIDs, LA/Ao, and MVP), baseline diameter and blood pressure only the combined effects of increased age (= .01) and body weight (= .002) were significantly associated with reduced FMD (R2 = 0.31). There was no difference among groups in the plasma l-arginine, ADMA, and SDMA concentration or the l-arginine/ADMA ratio (Table 2) and the plasma markers were not associated with any of the ultrasonographic markers of endothelial function.

image

Figure 2. Flow-mediated vasodilation (FMD) in dogs with increasing severity of myxomatous mitral valve disease (MMVD) measured as an increasing percentage of mitral regurgitation (MR) and clinical signs compatible with congestive heart failure (CHF). CKCS with CHF had a significantly reduced FMD compared to CKCS with no, minimal or mild MR. CKCS with severe MR had a significantly reduced FMD compared to CKCS with no to minimal MR. Bars represent mean ± SD. *< .005 compared to CKCS with no, minimal or mild MR, #= .03 compared to CKCS with no or minimal MR.

Download figure to PowerPoint

Table 2. Characteristics of FMD and plasma marker concentrations in the 4 groups of dogs with mitral regurgitation.
 Group 1Group 2Group 3Group 4
CKCS, No or Minimal MRCKCS, Mild MRCKCS, Moderate-to-Severe MRCKCS in Heart Failure
  1. CKCS, Cavalier King Charles Spaniel; MR, mitral regurgitation; FVI, flow velocity integral; FMD, flow-mediated vasodilation; ADMA, asymmetric dimethylarginine; SDMA, symmetric dimethylarginine.

  2. Data are expressed as mean ± standard deviation or as medians and 25th to 75th percentiles.

  3. a

    P < .01 compared to groups 1 and 2.

  4. b

    P < .05 compared to groups 1.

  5. c

    Body weight was a significant covariate (P < .0001).

Baseline luminal diameter  (mm)1.59 ± 0.201.64 ± 0.171.72 ± 0.251.87 ± 0.14c
Increase in FVI (%)243 (110–382)247 (174–331)156 (149–530)325 (200–461)
Time to peak luminal  diameter (seconds)62 (42–88)50 (42–61)53 (48–62)54 (49–66)
Absolute change in  luminal diameter (mm)0.17 ± 0.060.16 ± 0.060.12 ± 0.070.09 ± 0.06a
FMD (%)10.40 ± 4.5810.14 ± 3.676.79 ± 3.98b4.79 ± 3.22a
ADMA (μmol/L)1.44 ± 0.261.53 ± 0.301.49 ± 0.261.39 ± 0.20
SDMA (μmol/L)0.35 (0.33–0.36)0.37 (0.33–0.41)0.40 (0.34–0.41)0.38 (0.33–0.54)
l-Arginine (μmol/L)96.8 ± 15.1103.6 ± 27.4106.4 ± 24.584.9 ± 19.2
l-Arginine/ADMA ratio69.5 ± 16.469.5 ± 20.971.4 ± 7.561.1 ± 10.0

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

MMVD and Endothelial Dysfunction

The present study showed that FMD decreased with progression of MR. The study also showed an association between the combination of increasing age (= .01) and body weight (= .002) (R2 = 0.31) and decreasing FMD although this was not significant when analyzing the effects of increasing LVIDd on FMD (= .0004, R2 = 0.27).

Human patients with mitral valve disease and age-matched healthy controls have peripheral vasodilatory responses (measured as forearm blood flow) mediated by endothelium-dependent and endothelium-independent vasodilators that are significantly impaired in patients with symptomatic chronic heart failure because of valvular heart disease.[11] In the present study, FMD was associated with the echocardiographic variable LVIDd and CKCS with clinical signs compatible with CHF and moderate-to-severe MR had a reduced FMD compared to CKCS with no or minimal MR (Fig 2; Table 2). The CHF dogs were in cardiac treatment and as both ACE inhibitors and pimobendan have vasodilatory effects,[36-38] the decreased FMD, greatest in the CHF group, might have been attenuated by the cardiac treatment which could have increased FMD and thus camouflaged an even larger FMD reduction because of CHF. The increase in FVI which gives an indication of microvascular function[39] did not differ between groups in the present study. The angle of correction that was needed to align with the vessel lumen in this study was higher than what is considered acceptable (≤60 degrees) to obtain accurate Doppler velocities.[40] FVI results should thus be interpreted accordingly and are only reported to demonstrate the successful achievement of reactive hyperemia.

Several studies of dogs have measured plasma markers as an indirect measurement of endothelial function in relation to MMVD with conflicting results. A diet-controlled study of CKCS with asymptomatic stages of MMVD showed that plasma nitrate and nitrite (NOx) concentration decreases with increasing MR.[19] Although the study did not include very old dogs with clinical signs of disease, the results support the findings of the present study, ie, that endothelial dysfunction might be associated with MR progression. Another study showed no difference in plasma NOx concentration between healthy dogs and dogs in CHF because of MMVD and dilated cardiomyopathy. These dogs were of different breeds and age and without controlled diet.[17] The plasma ADMA concentration and the l-arginine/ADMA ratio might be better markers of endothelial function as they are not influenced by diet and water intake and have been shown to be affected in dogs with CHF because of MMVD. However, these plasma markers are also affected by age, kidney status, breed, and exercise.[18, 21] In the present study, there were no differences between the groups in the plasma ADMA, SDMA, and l-arginine concentration or the l-arginine/ADMA ratio (Table 2). None of these markers of NO production were associated with FMD, which may be because of a different association of l-arginine, ADMA, or both with vascular NO production in dogs as compared to humans. Consequently, plasma concentrations of these markers might therefore not adequately reflect endothelial function in dogs.

FMD Measurements in Dogs

The present study confirmed the feasibility of applying FMD measurement in dogs as an in vivo estimate of endothelial function. The technique requires the dog to lie still for 10–15 minutes without sedation, which in the present study was possible in approximately 60% of the examined dogs. Thus, there has been a selection in the examined population of CKCS because the stressed and impatient dogs could not be included. It is a concern that especially dogs in CHF find it uncomfortable to lie still on the side for 10–15 minutes and that the included part of this group mainly represent dogs that were stable on treatment. Fortunately, the excluded dogs in the present study were spread fairly equal between all 4 groups. Alternatively, dogs would have to be sedated and because sedative drugs might also affect endothelial function this would have to be further studied. The FMD technique and software used were very similar to that reported in a previous canine study.[24] The main difference between these studies is that the present study timed the recordings rather than using ECG guiding to have exactly the same time recording from all dogs because not all dogs had the same heart rate. In a healthy dog population consisting of several different breeds, body weight was the only independent contributing factor for FMD.[24] In the present study, only 1 breed was studied and therefore dogs were of similar body weight (Table 1). However, even small changes in body weight within the same breed had a significant effect on both baseline diameter and on FMD. From human studies it is suggested that obesity is associated with endothelial dysfunction;[41] however, scaling procedures to correct FMD measurements for body composition have not been recommended.[42] In the present study, body composition was only subjectively assessed on a rough scale, and it is therefore not possible to conclude whether the effect of body weight on FMD is caused by obesity or larger body size, and these effects need to be studied further. MMVD is known to progress with age and it is very frequent in the CKCS breed.[2] Therefore, it is difficult to find old CKCS without MR. A previous study of FMD in healthy dogs has shown that FMD is reduced in dogs older than 6 years,[24] and several studies in human patients have also shown that FMD and NO availability decreases with age.[25, 43]

The FMD technique is challenging, and both human and canine studies have shown a large variation in FMD measurements.[23, 24, 44] However, it is encouraging that the measurements (baseline, increase in FVI, time to peak luminal diameter and FMD) in the healthy small dog group examined previously[24] are very comparable to the results of the CKCS with no to minimal MR (group 1) in the present study (Table 2).

Future Studies

The present study suggests that the progression of MR is associated with development of endothelial dysfunction. This should be further studied and might be of importance in improving future treatment of the disease. Future studies are required to study the effect of other naturally occurring conditions such as obesity, hypertension, kidney disease, inflammation, and cardiomyopathy on FMD measurements in the canine population. However, the present results underline the importance of using age and weight or size matched groups.

Limitations

Unfortunately, 40% of the dogs were excluded from the study, which was mainly caused by limb movements and thus loss of recordings. This resulted in small group sizes, limiting the power of the study. The fact that the groups were not age matched is a limitation of the study. It is important to standardize the age of the dog groups included in future studies, because increasing age has a significant effect on FMD. To avoid breed and body weight differences only 1 breed was studied; however, a study with multiple weight matched breeds including scaling of body composition is needed to show whether these findings apply to other breeds as well. Thoracic radiographs were not part of the standard protocol. In 6 of 11 dogs with clinical signs of CHF radiographs were taken prior to the study to rule out respiratory diseases. All dogs in group 4 had clinical signs compatible with CHF, echocardiographic evidence of severe mitral valve insufficiency as well as enlargement of the left ventricle and atrium and were responsive to furosemide treatment and it is therefore unlikely that these clinical signs were caused by other diseases than CHF. The heart failure dogs received medication on the day of examination. Furthermore, several different medications were given to different dogs. This is a limitation of the study as we cannot estimate whether the given medication of the dogs affected FMD. Ideally, dogs in heart failure should have been examined before treatment was initiated; however, this was not possible because of practical and ethical reasons. A future study might consider including dogs diagnosed with CHF and only receiving furosemide treatment. It is recommended that ultrasonographers should have experience of 100 FMD scans,[22] but in this study the ultrasonographer had only performed 50 scans before the study, and it is likely that repeatability of the technique improves with further experience. Finally, 2 persons performed the off-line FMD analyses and interobserver variation may have occurred. However, because joint training and discussion of cases was performed throughout the study this variation is suggested to be small.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The authors thank Christina T. Kjempf, Dennis S. Jensen, and Hanne L.P. Carlsson, Department of Basic Animal and Veterinary Sciences, Faculty of Life Sciences, University of Copenhagen, Denmark, and Sigrid de Jong, Department of Clinical Chemistry, VU University Medical Center, The Netherlands for their excellent technical assistance. Dr Beate Egner, Babenhausen, Germany is acknowledged for lending out the high-definition oscillometric equipment used for blood pressure measurements.

The work was supported by the Danish Council of Independent Research | Medical Sciences (Project no 271-07-0784 and project no 271-08-0998), The Augustinus foundation, and the Danish Medical Association.

Footnotes
  1. 1

    Vivid i echocardiograph, GE-Medical, Milwaukee, WI

  2. 2

    VET Memodiagnostic HDO monitor, S+B medVET, Babenhausen, Germany

  3. 3

    Brachial Analyzer for Research, Vascular Research Tools, Medical Imaging Applications LLC, Coralville, IA

  4. 4

    EchoPAC PC. Version 108.x.x., GE VINGMED ULTRASOUND AS, Horten, Norway

  5. 5

    SAS statistical software, version 9.1, SAS Institute, Cary, NC

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References