Methods

The electronic medical record system of IDEXX Telemedicine Consultants was searched for YTs, Pugs, POMs, and BTs undergoing thoracic radiography between January 1, 2023, and March 31, 2023. The radiographs had been previously submitted to IDEXX Telemedicine Consultants for telemedicine review. Dogs were included if they had 2- or 3-view thoracic radiographs performed (at least 1 right lateral thoracic radiograph and 1 ventrodorsal or dorsoventral view) and no evidence of cardiopulmonary or systemic disease was detected. All dogs with incomplete radiographic studies and/or known extracardiac disease were excluded (ie, abnormalities on thoracic radiographs such as pleural effusion, lymphadenopathy, pneumonia, or neoplasia). Radiographic studies with poor positioning that limited the accuracy of VHS and VLAS measurements as judged by the investigating cardiologist were also excluded. All dogs included in the study must have had a normal cardiac auscultation as noted by the original veterinarian performing the physical exam, which was documented in the telemedicine consultation form. All dogs with a reported heart murmur were excluded. If a dog had no reported heart murmur but had subjective cardiomegaly reported by the initial IDEXX radiologist or cardiologist, was receiving a cardiac medication that could affect cardiac size (ie, pimobendan or diuretics), had a history of receiving a grain-free diet, or had a history of an elevated N-terminal pro-brain natriuretic peptide, the dog was termed “suspected cardiac disease” and also excluded.

Data collected from the patient record and radiographs included age, weight, sex, VHS, and VLAS. The VHS and VLAS measurements were performed by the same board-certified cardiologist in all breeds. As the radiographs were all in a digital format, measurements were performed using a digital caliper and were performed on the right lateral radiograph. The measurement of VHS was performed using the technique initially described by Buchanan and Bücheler¹ where the long axis of the heart was measured from the center of the carina to the most distal contour of the ventral apex of the heart. The carina was defined as the radiolucent circular structure within the trachea, which represents the bifurcation of the left and right mainstem bronchi. The short axis of the heart was measured in the central third region of the heart, perpendicular to the long axis. Both axis measurements were then positioned over the thoracic vertebral bodies beginning at the cranial edge of the fourth thoracic vertebra. The sum of both axes was used to determine the number of vertebral units to the nearest 0.1 vertebra (Supplementary Figure S1). All measurements were performed on the right lateral thoracic radiograph in all cases to maintain consistency.

For the VLAS measurement, a line was drawn from the center of the ventral aspect of the carina to the caudal aspect of the left atrium where it intersected with the dorsal border of the caudal vena cava. The measurement was then positioned over the fourth thoracic vertebra as before, and the total vertebral units were used to calculate the VLAS (Supplementary Figure S2).¹⁴

Statistical analysis

Continuous variables were described using median values and the IQRs, while categorical variables were presented in terms of proportions. As recommended by the National Conference of Standards Laboratories International, a nonparametric method was selected due to its robustness to heavy skewness observed in the distribution of VHS and VLAS values across breeds. Methodologic adherence to the Clinical and Laboratory Standards Institute's EP28-A3 guidelines was maintained by employing a 95% nonparametric bootstrap method with a 90% CI for calculations. All statistical analyses were conducted using R software, version 4.3.4 (R Foundation for Statistical Computing).

Results

During the study period, a total of 1,776 YTs, 1,431 POMs, 945 Pugs, and 711 BTs were identified. Of these, 1,293 YTs, 1,051 POMs, 554 Pugs, and 515 BTs were excluded. Reasons for exclusion for each breed are provided (Supplementary Table S1). Consequently, there were a total of 483 YTs, 380 POMs, 391 Pugs, and 196 BTs available for analysis. There were 89 (18.4%) intact female YTs, 149 (30.9%) spayed female YTs, 63 (13.0%) male intact YTs, 164 (34.0%) male castrated YTs, and 18 (3.7%) YTs with no sex reported. There were 70 (18.4%) female intact POMs, 98 (25.8%) female spayed POMs, 70 (18.4%) male intact POMs, 128 (33.7%) male castrated POMs, and 14 (3.7%) POMs with no sex reported. There were 56 (14.3%) female intact Pugs, 118 (30.2%) spayed female Pugs, 66 (16.9%) male intact Pugs, 137 (35.0%) male castrated Pugs, and 14 (3.6%) Pugs with no sex reported. There were 8 (4.1%) female intact BTs, 78 (39.8%) female spayed BTs, 14 (7.1%) male intact BTs, 85 (43.4%) male castrated BTs, and 11 (5.6%) BTs with no sex reported. Age and weight statistics are reported (Table 1). Median VHS, VLAS, and reference intervals for all breeds are provided (Tables 2 and 3; Figures 1 and 2).

View Full Size

Vertebral heart size (VHS) histogram for all 4 breeds. Solid lines represent the upper and lower reference limits, and the area between these represents the reference interval. The dashed lines represent the CI for those reference limits.

Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.24.10.0320

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Vertebral left atrial size (VLAS) histogram for all 4 breeds. Solid lines represent the upper and lower reference limits, and the area between these represents the reference interval. The dashed lines represent the CI for those reference limits.

Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.24.10.0320

• Download Figure
• Download figure as PowerPoint slide

Discussion

The study here reports similar reference intervals for YTs and POMs but narrower reference intervals for Pugs and BTs compared to a previous study.⁷ The difference in the results may relate to the difference in study methodologies. The previous study⁷ excluded dogs < 1.5 years of age whereas our study did not have any age limitations. The investigators of this study also excluded dogs with reported systemic disease; there was no report of excluding subjects with systemic disease in the previous study.⁷ However, the investigators suspect the primary reason for the difference in the results is related to difference in sample size. The study population of each breed in this study is at least 10 times greater than the initial study,⁷ which is likely the primary reason for the difference in results.

Interestingly, in our study, there was a high frequency of Pugs and BTs with a VHS of 12 vertebrae. The significance of these findings is unclear. It may reflect inadvertently including dogs with underlying cardiac disease despite our exclusion criteria. However, if the investigators included multiple dogs with occult cardiac disease, we would expect our VHS reference interval to be higher than the previous study⁷ due to dogs with cardiac disease having higher VHS values.⁴ However, the upper limits of our reference interval for both Pugs and BTs are actually lower compared to the previous study.⁷ Another possibility is the inadvertent inclusion of dogs with pathology along the heart base such as heart base tumors or lymph node enlargement. Brachycephalic breeds such as BTs are predisposed to chemodectomas,¹⁵ which may explain the findings here. The investigators did exclude dogs with suspected heart base tumors although a study found thoracic radiographs had low sensitivity at identifying dogs with heart base tumors.¹⁶ Consequently, without an echocardiogram, the investigators are unable to definitively state all dogs with heart base tumors were excluded, which may also contribute to the high frequency of Pugs and BTs with a VHS of 12 vertebrae. Further evaluation into VHS reference intervals Pugs and BTs comparing radiographic to echocardiographic findings may be warranted.

In this study, only the right lateral radiograph was used to measure the VHS and VLAS values, as this was the radiographic view used to derive VHS values in the initial study.⁷ If there was no right lateral radiograph provided, the study was excluded from analysis. Previous studies¹⁷ have reported significant differences between VHS values when measured from a right or left lateral radiograph in the same dog. As such it is very plausible VHS and VLAS values will be different between right and left lateral radiographs for these breeds as well. Consequently, the reference intervals reported here should only be utilized when measuring VHS and VLAS on the right lateral thoracic radiograph.

The study has several limitations. There was a large amount of data from telemedicine consultations, and in some cases, the demographic data were incomplete (ie, age, weight, and/or sex). The study included cases throughout the country, and therefore, radiographic technique and positioning were not standardized. Some studies⁹ have shown higher body conditions were associated with higher VHS scores but were not evaluated here. The dogs included did not have a reported heart murmur or significant systemic disease, but detection of underlying cardiac disease or systemic disease is dependent on the clinical proficiency of submitting veterinarians. The authors are unable to control for phase of respiration, which may have a mild effect on the accuracy of VHS measurements¹⁸; further studies evaluating VHS and VLAS measurements in different phases of respiration may be warranted. The heartworm status was not known for all dogs, and heartworm disease causes VHS to increase in affected patients although all dogs that were reported to be heartworm positive were also excluded.¹⁹ Other diseases that can affect VHS measurement such as pericardial effusion²⁰ cannot be completely ruled out although is considered less likely given the absence of clinical signs necessary for inclusion in the study population. Dogs that were reported to receive grain-free diets were excluded given the association with dilated cardiomyopathy.²¹ However, diet was not universally reported in all cases, and therefore, it is possible a small population of dogs receiving a grain-free diet were inadvertently included in the study population as well. Finally, it also would have been ideal to have 2 cardiologists blinded to each other's results measuring VHS and VLAS in the study population and to compare to results to look for evidence of interobserver variability, but this was not performed here. Including multiple observers would enhance the reliability of the VHS and VLAS measurements and provide broader validation of these reference intervals as well as contribute to a more comprehensive understanding of how these values may vary among veterinarians.

In conclusion, the authors report the reference intervals for VHS and VLAS in YTs, POMs, Pugs, and BTs. Clinicians should be aware of these results to avoid erroneously diagnosing cardiomegaly in these breeds. Continued investigation into breed-specific VHS measurements in other canine breeds appears warranted.

Supplementary Materials

Supplementary materials are posted online at the journal website: avmajournals.avma.org.

Disclosures

The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.

Funding

The authors have nothing to disclose.