Radiographic evaluation of cardiac size and shape for potential eccentric cardiac enlargement is of diagnostic value in dogs and cats with suspected cardiac disease,1–4 and echocardiography provides a reliable assessment of the heart's internal structures.3,5 Generalized cardiomegaly, pericardial effusion, and some chamber enlargements (most likely left atrial enlargement) are the most common abnormalities evaluated with radiography.1,2 Although trained personnel can often recognize cardiomegaly evident on thoracic radiographic images, inexperienced veterinarians may benefit from a quantitative procedure to assess the size and shape of the cardiac silhouette. Radiographic procedures to determine the cardiac size by calculating the VHS1 or cardiothoracic ratios6,7 have been reported; however, these methods rely on evaluation of the overall size of the cardiac silhouette without a separate assessment of each cardiac dimension and therefore incorporate all cardiac chambers. The techniques that rely on calculating the cardiothoracic ratios are complicated and time consuming and require a specific software program, thus impeding their use in clinical practice.1,8,9 Furthermore, the cardiothoracic ratios may be of little diagnostic value in dogs because of marked variation in thoracic cavity conformation among dog breeds and may have questionable usefulness in dogs with certain cardiovascular or pulmonary diseases.1
Veterinary practitioners and researchers have used the VHS as an objective method for assessing cardiomegaly in dogs for the past 2 decades.1,10,11 However, disparities in thoracic conformation and thoracic vertebral characteristics owing to breed variation, congenital anomalies, shortened vertebral segments, or intervertebral disk space narrowing may result in higher values of VHS and subsequent misin-terpretation of VHS results as cardiomegaly.4,10–13 The accuracy of the VHS method is also influenced by individual variations among dogs in regard to actual heart size and thoracic vertebral length.13 These variations, in addition to interobserver variability,13 could contribute to the weak correlation recently identified between the VHS and the length of each corresponding cardiac dimension measured on thoracic radiographic images of dogs.9 These combined factors hinder the use of reference ranges of VHS to distinguish between radiographically normal and enlarged cardiac silhouettes in dogs.13 Alternatively, the MHS relies on the length of a single easily identified reference, the manubrium. The ML strongly correlates with the corresponding cardiac axes, and thus MHS was alternatively established and used to determine cardiac short-, long-, and overall-MHSs in healthy large- and small-breed dogs with radiographically normal cardiopulmonary structures.9 However, to our knowledge, the CSI and MHSs have not yet been assessed in dogs with confirmed cardiac disease.
The purpose of the present study was to evaluate the usefulness of MHSs in distinguishing between dogs with and without cardiac disease. Our study aimed to evaluate the CSIs and MHSs measured on right lateral thoracic radiographic images of large- and small-breed dogs with echocardiographically confirmed cardiac disease and cardiomegaly and to compare those values with values for healthy dogs. We hypothesized that the CSIs and MHSs would be higher in dogs with confirmed cardiac disease and would vary with the side of the heart affected (ie, the right side of the heart [right-sided cardiac disease] vs the left side of the heart [left-sided cardiac disease]). We also hypothesized that different reference values for MHSs may be required for radiographically evaluating cardiac dimensions in the different breed-size groups (large-breed dogs or small-breed dogs).
Materials and Methods
The study protocol was designed and performed in the Small Animal Hospital at the University of Florida College of Veterinary Medicine. No ethical approval was required because the retrospective study was based solely on reviews of medical records, including radiographic and echocardiographic images, generated during routine veterinary care of the dogs included in the study.
Animals
Medical records at the University of Florida Small Animal Hospital were searched to identify client-owned large-breed dogs (≥ 16.0 kg [35.2 lb]) and small-breed dogs (≤ 12.0 kg [26.4 lb]) that had undergone thoracic radiography and had echocardiographically confirmed moderate to severe right- or left-sided cardiac disease with cardiomegaly (case dogs) between January 1, 2012, and December 31, 2015. Dogs with confirmed pericardial effusion were excluded from the study, as were dogs with radiographic evidence of vertebral anomalies or deformed manubria.
For the control group, we retrieved medical records, including thoracic radiographic images, of client-owned large-breed dogs (≥ 16.0 kg) and small-breed dogs (≤ 12.0 kg) that had been included in a previous study9 conducted by 2 of the authors (AAM and CRB). These dogs had no history or clinical signs of cardiovascular disorders (eg, no record of a heart murmur or gallop detected during thoracic auscultation) or respiratory diseases and had no abnormalities evident on thoracic radiography. Dogs with a VHS < 10.9 were considered clinically normal as previously described14 and included as control dogs. Echocardiography was not required for control dogs.
Data collection
For each dog, data collected from the medical record included age, body weight, sex, and breed. For case dogs, characteristics of their cardiac disease, including the side of the heart affected, were also recorded.
Radiographic measurements
Radiographic procedures were standardized by use of standard exposure factors and source-to-image receptor distance, and digitized radiographic images were retrieved with an image-archiving communication systema and a diagnostic workstation. All thoracic radiographic images included in the study were obtained at the time of peak inspiration of the imaged dog, with no sedation or anesthesia. Radiographic images were approved in terms of quality and positioning by a board-certified radiologist (CRB) and analyzed by an investigator (KEP) unaware of the echocardiographic findings. All measurements were performed on right lateral thoracic radiographic images of each dog. The cSAL, cLAL, and ML were measured as previously described.9 Pleural and pericardial fat opacity was excluded during measurements of the cSAL and cLAL. The cLAL was defined as the length of a line extending between the ventral border of the carina and the most caudoventral margin of the cardiac apex. The cSAL was defined as the length of a line drawn perpendicular to the cardiac long axis at the widest point of the cardiac silhouette.
The CSIs, MHSs, and VHSs were calculated for all dogs. The CSI was the cSAL:cLAL ratio, calculated to objectively evaluate the roundness of the cardiac silhouette9 and modified from the lateral sphericity index (ie, the cLAL:cSAL ratio).4 For the MHSs, the ML was used to standardize the cardiac dimensions in that the short-MHS was the cSAL:ML ratio, the long-MHS was the cLAL:ML ratio, and the overall-MHS was calculated with the formula (cLAL + cSAL)/ML.9 The VHS was determined by counting the number of vertebrae (beginning at the cranial end plate of T4) that represented the same length as the sum of the lengths of the cLAL + cSAL as previously described.1,9
Statistical analysis
Results for variables were assessed with and passed a D'Agostino-Pearson omnibus test of normality before further analysis. Data were reported as mean ± SD and with 95% CIs for the mean. An unpaired, 2-tailed t test was used to compare mean results for CSIs, MHSs, and VHSs across groups: all control dogs versus all case dogs, control dogs versus case dogs grouped by breed size (large- or small-breed dogs), control dogs versus case dogs grouped by breed size and the side of the heart affected (right-sided or left-sided cardiac disease), and case dogs grouped by breed size with right-sided versus left-sided cardiac disease. Data analyses were performed with commercially available statistical software,b and values of P < 0.05 were considered significant.
Results
Animals
The control group consisted of 120 healthy dogs (60 large-breed dogs and 60 small-breed dogs) that were evaluated for reasons other than cardiovascular or respiratory disorders (eg, routine evaluation for pulmonary metastases) and had no abnormalities detected on thoracic radiography. Of the 60 large-breed control dogs, 25 were castrated males, 7 were sexually intact males, 20 were spayed females, and 8 were sexually intact females. Of the 60 small-breed control dogs, 24 were castrated males, 3 were sexually intact males, 27 were spayed females, and 6 were sexually intact females. Mixed-breed dogs (15/60 [25%]) were most common for the large-breed control dogs, whereas Dachshunds (6/60 [10%]) were most common for the small-breed control dogs (Table 1).
Distributions of breeds or types of client-owned large-breed dogs (≥ 16.0 kg [35.2 lb]) and small-breed dogs (≤ 12.0 kg [26.4 lb]) that either had echocardiographically confirmed moderate to severe cardiac disease with cardiomegaly (case dogs; n = 64) or were healthy and without evidence of cardiac disease (control dogs; 120) evaluated at the University of Florida Small Animal Hospital between January 1, 2012, and December 31, 2015, and included in the study.
| Breeds or types | No. (%)* of control dogs | No. (%)* of case dogs |
|---|---|---|
| Large-breed dogs | 60 (100) | 23 (100) |
| Mixed breed | 15 (25) | 4 (17) |
| Labrador Retriever | 9 (15) | 1 (4) |
| Golden Retriever | 9 (15) | 1 (4) |
| German Shepherd Dog | 5 (8) | 3 (13) |
| Boxer | 3 (5) | 2 (9) |
| Border Collie | 3 (5) | 0 |
| Weimaraner | 2 (3) | 0 |
| Greyhound | 2 (3) | 1 (4) |
| Australian Shepherd | 1 (2) | 1 (4) |
| Doberman Pinscher | 1 (2) | 5 (22) |
| Newfoundland | 1 (2) | 0 |
| Bernese Mountain | 1 (2) | 0 |
| Great Pyrenees | 1 (2) | 0 |
| Swiss Mountain | 1 (2) | 0 |
| Rhodesian Ridgeback | 1 (2) | 0 |
| Dalmatian | 1 (2) | 0 |
| American Pit Bull Terrier | 1 (2) | 0 |
| Whippet | 1 (2) | 0 |
| Belgian Shepherd | 1 (2) | 0 |
| Great Dane | 1 (2) | 2 (9) |
| Staffordshire Bull Terrier | 0 | 1 (4) |
| English Springer Spaniel | 0 | 1 (4) |
| Siberian Husky | 0 | 1 (4) |
| Small-breed dogs | 60 (100) | 41 (100) |
| Dachshund | 6 (10) | 2 (5) |
| Mixed breed | 5 (8) | 13 (32) |
| Yorkshire Terrier | 5 (8) | 2 (5) |
| Miniature Schnauzer | 5 (8) | 0 |
| Miniature Poodle | 4 (7) | 0 |
| Cocker Spaniel | 3 (5) | 3 (7) |
| Chihuahua | 3 (5) | 3 (7) |
| Shih Tzu | 3 (5) | 1 (2) |
| Pembroke Welsh Corgi | 3 (5) | 1 (2) |
| Pekingese | 3 (5) | 1 (2) |
| Beagle | 2 (3) | 4 (10) |
| Miniature Pinscher | 2 (3) | 1 (2) |
| Bichon Frise | 2 (3) | 0 |
| Jack Russell Terrier | 2 (3) | 0 |
| Lhasa Apso | 2 (3) | 0 |
| Scottish Terrier | 2 (3) | 0 |
| French Bulldog | 1 (2) | 0 |
| Italian Greyhound | 1 (2) | 0 |
| Boston Terrier | 1 (2) | 0 |
| Chinese Crested | 1 (2) | 0 |
| Pomeranian | 1 (2) | 2 (5) |
| Papillon | 1 (2) | 1 (2) |
| Shiba Inu | 1 (2) | 0 |
| Cavalier King Charles Spaniel | 1 (2) | 4 (10) |
| West Highland White Terrier | 0 | 1 (2) |
| Rat Terrier (Feist) | 0 | 1 (2) |
| Maltese | 0 | 1 (2) |
The sums of percentages in breed-size groups may not be equal to 100% because of rounding.
The case group consisted of 64 dogs (23 large-breed dogs and 41 small-breed dogs) that had cardiomegaly confirmed with echocardiography and had undergone concurrent thoracic radiography. Of the 23 large-breed case dogs, 11 were castrated males, 8 were sexual intact males, 3 were spayed females, and 1 was a sexually intact female. Of the 41 small-breed case dogs, 11 were castrated males, 6 were sexually intact males, 22 were spayed females, and 2 were sexually intact females. Doberman Pinschers (5/23 [22%]) were most common for the large-breed case dogs, whereas mixed-breed dogs (13/41 [32%]) were most common for the small-breed case dogs (Table 1). Among the 23 large-breed case dogs, 6 had congenital cardiac disease and 17 had acquired cardiac disease. The congenital cardiac diseases included subaortic stenosis (n = 4) and patent ductus arteriosus (left-to-right shunt [1] or right-to-left shunt [1]). The acquired cardiac diseases included dilated cardiomyopathy (n = 9), heartworm disease with pulmonary hypertension (6), aortic and mitral valves endocarditis (1), and mitral valve degenerative disease (1). When grouped by the side of the heart affected, the 23 large-breed case dogs consisted of 7 with right-sided cardiac disease and 16 with left-sided cardiac disease. Among the 41 small-breed case dogs, 12 had congenital cardiac disease and 29 had acquired cardiac disease. The congenital cardiac diseases included pulmonic stenosis (n = 8), left-to-right patent ductus arteriosus (1), mitral valve dysplasia (1), double outlet right ventricle (1), and pulmonary artery hypoplasia with pulmonary hypertension (1). The acquired cardiac diseases included mitral valve degenerative disease (n = 16), heartworm disease with pulmonary hypertension (9), mitral and tricuspid valve degenerative disease (1), tricuspid valve degenerative disease (1), hypertrophic cardiomyopathy (1), and dilated cardiomyopathy (1). When grouped by the side of the heart affected, the 41 small-breed case dogs consisted of 20 with right-sided cardiac disease and 21 with left-sided cardiac disease.
None of the large-breed dogs had radiographic evidence of vertebral anomalies or deformed manubrium. However, 3 of the small-breed control dogs were excluded from the control group because of radiographic evidence of a deformed manubrium, and thus 3 additional small-breed control dogs were enrolled as replacements. None of the small-breed case dogs had radiographic evidence of vertebral or manubrium deformities.
The mean ± SD age was significantly (P = 0.001) older for large-breed control dogs (7.9 ± 3.5 years), compared with large-breed case dogs (5.0 ± 3.2 years; Table 2); however, there was no substantial difference in mean age for small-breed control dogs versus small-breed case dogs. Similarly, the mean body weight did not differ substantially for case dogs versus control dogs in either breed-size group (large-breed dogs or small-breed dogs).
Descriptive statistics for the dogs (controls vs cases) described in Table 1, stratified by breed size.
| Variables | Large-breed dogs | Small-breed dogs | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Controls (n = 60) | Cases (n = 23) | Controls (n = 60) | Cases (n = 41) | |||||||
| Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value | |
| Age (y) | 7.9 ± 3.5 | 7.0–8.8 | 5.0 ± 3.2 | 3.6–6.4 | 0.001 | 9.1 ± 3.7 | 8.2–10.1 | 7.6 ± 4.5 | 6.2–9.1 | 0.075 |
| Body weight (kg) | 31.1 ± 10.6 | 28.3–33.8 | 34.9 ± 10.6 | 30.2–39.6 | 0.157 | 7.7 ± 3.7 | 6.7–8.6 | 7.8 ± 5.3 | 6.1–9.5 | 0.918 |
Radiographic measurements
Control dogs versus case dogs overall—Measurements (Figure 1) of cSAL, cLAL, cSAL + cLAL, and ML and results for CSI, VHS, and short-, long-, and overall-MHS were compiled. When results for all 64 case dogs were compared with results for all 120 control dogs, the mean ± SD values of CSI, short-MHS, long-MHS, overall-MHS, and VHS were significantly (P ≤ 0.001) higher for case dogs (85.0 ± 9.0%, 2.6 ± 0.4, 3.1 ± 0.5, 5.7 ± 0.8, and 11.3 ± 1.2, respectively), compared with control dogs (79.0 ± 6.0%, 2.2 ± 0.3, 2.8 ± 0.4, 5.1 ± 0.7, and 10.5 ± 0.7, respectively; Figure 2; Table 3). Similarly, the mean values for these variables were also significantly (P ≤ 0.001) higher for large-breed case dogs versus their controls (Figure 3; Table 4) and for small-breed case dogs versus their controls (Figure 4; Table 5).
Representative right lateral thoracic radiographic images of client-owned dogs with echocardiographically confirmed moderate to severe cardiac disease with cardiomegaly evaluated at the University of Florida Small Animal Hospital between January 1, 2012, and December 31, 2015, and showing the measurements for cLAL, cSAL, and ML and the resulting CSI, VHS, and short-, long-, and overall-MHS. A—A 2-year-old 30.1-kg (66.2-lb) castrated male German Shepherd Dog with dilated cardiomyopathy and associated left-sided cardiomegaly (cLAL = 163.5 mm; cSAL = 130.3 mm; ML = 48.8 mm; CSI = 80%; short-MHS = 2.7; long-MHS = 3.4; overall-MHS = 6.1; VHS = 12.1). B—A 5-year-old 9.0-kg (19.8-lb) female Dachshund with heartworm disease and associated right-sided cardiomegaly (cLAL = 83.2 mm; cSAL = 72.4 mm; ML = 29.2 mm; CSI = 87%; short-MHS = 2.5; long-MHS = 2.9; overall-MHS = 5.4; VHS = 11.0).
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.888
Box-and-whisker plots of CSI (A), short-MHS (B), long-MHS (C), overall-MHS (D), and VHS (E) for large-breed dogs (≥ 16.0 kg [35.2 lb]) and small-breed dogs (≤ 12.0 kg [26.4 lb]) that either had echocardiographically confirmed moderate to severe cardiac disease with cardiomegaly (cases; n = 64) or were healthy and without evidence of cardiac disease (controls; 120) evaluated at the University of Florida Small Animal Hospital between January 1, 2012, and December 31, 2015. For each plot, the boxes represent the 25th to 75th percentiles; the line and cross within each box represent the median and mean, respectively; and the whiskers represent the range.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.888
Cardiac variables (CSI; short-, long-, and overall-MHS; and VHS) calculated from measurements obtained from right lateral thoracic radiographic images of the dogs (controls [n = 120] vs cases [64]) described in Table 1.
| Controls | Cases | ||||
|---|---|---|---|---|---|
| Cardiac variables | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value |
| CSI (%) | 79.0 ± 6.0 | 78.0–80.0 | 85.0 ± 9.0 | 83.0–87.0 | < 0.001 |
| Short-MHS | 2.2 ± 0.3 | 2.2–2.3 | 2.6 ± 0.4 | 2.5–2.7 | < 0.001 |
| Long-MHS | 2.8 ± 0.4 | 2.8–2.9 | 3.1 ± 0.5 | 3.0–3.2 | < 0.001 |
| Overall-MHS | 5.1 ± 0.7 | 4.9–5.2 | 5.7 ± 0.8 | 5.5–5.9 | < 0.001 |
| VHS (vertebrae) | 10.5 ± 0.7 | 10.4–10.6 | 11.3 ± 1.2 | 11.0–11.6 | < 0.001 |
Box-and-whisker plots of CSI (A), short-MHS (B), long-MHS (C), overall-MHS (D), and VHS (E) for the large-breed dogs (60 control dogs and 23 case dogs [7 with right-sided cardiac disease and 16 with left-sided cardiac disease]) described in Figure 2. Cases L = Case dogs with left-sided cardiac disease. Cases R = Case dogs with right-sided cardiac disease. See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.888
Cardiac variables (cSAL; cLAL; cSAL + cLAL; ML; CSI; short-, long-, and overall-MHS; and VHS) for the large-breed control dogs (n = 60) versus the large-breed case dogs collectively (23) and further subgrouped according to the side of the heart affected (right-sided [7] or left-sided [16] cardiac disease).
| Large-breed control dogs (n = 60) | Large-breed case dogs (n = 23) | Large-breed case dogs with right-sided cardiac disease (n = 7) | Large-breed case dogs with left-sided cardiac disease (n = 16) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Variables | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value* | Mean ± SD | 95% CI | P value* | Mean ± SD | 95% CI | P value* |
| cSAL (mm) | 107.2 ± 12.3 | 104.1–110.4 | 122.2 ± 13.3 | 116.4–128.1 | < 0.001 | 122.2 ± 8.8 | 113.0–131.5 | 0.005 | 122.3 ± 14.9 | 114.3–130.2 | < 0.001 |
| cLAL (mm) | 139.6 ± 17.7 | 135.0–144.2 | 149.5 ± 19.5 | 140.8–158.1 | 0.032 | 138.6 ± 11.2 | 126.8–150.4 | 0.894 | 153.6 ± 20.6 | 142.6–164.5 | 0.008 |
| cSAL + cLAL (mm) | 246.8 ± 29.0 | 239.3–254.3 | 271.7 ± 30.1 | 258.4–285.1 | 0.001 | 260.8 ± 18.3 | 241.7–280.0 | 0.252 | 275.8 ± 33.0 | 258.2–293.4 | 0.001 |
| ML (mm) | 5I.54 ± 8.5 | 49.4–53.7 | 50.1 ± 8.0 | 46.6–53.7 | 0.499 | 50.5 ± 5.2 | 45.0–55.9 | 0.763 | 50.0 ± 9.0 | 45.2–54.8 | 0.525 |
| CSI (%) | 77.0 ± 5.0 | 76.0–78.0 | 82.0 ± 8.0 | 79.0–86.0 | 0.001 | 88.0 ± 6.0 | 82.0–94.0 | < 0.001 | 80.0 ± 8.0 | 76.0–85.0 | 0.056 |
| Short-MHS | 2.1 ± 0.2 | 2.0–2.2 | 2.5 ± 0.3 | 2.3–2.6 | < 0.001 | 2.4 ± 0.2 | 2.2–2.7 | 0.001 | 2.5 ± 0.3 | 2.3–2.7 | < 0.001 |
| Long-MHS | 2.7 ± 0.3 | 2.7–2.8 | 3.0 ± 0.4 | 2.9–3.2 | < 0.001 | 2.8 ± 0.2 | 2.5–3.0 | 0.834 | 3.1 ± 0.4 | 2.9–3.3 | < 0.001 |
| Overall-MHS | 4.8 ± 0.5 | 4.7–5.0 | 5.5 ± 0.6 | 5.2–5.8 | < 0.001 | 5.2 ± 0.4 | 4.7–5.7 | 0.084 | 5.6 ± 0.7 | 5.2–6.0 | < 0.001 |
| VHS (vertebrae) | 10.7 ± 0.5 | 10.6–10.9 | 11.2 ± 0.9 | 10.7–11.6 | 0.013 | 10.9 ± 0.6 | 10.3–11.5 | 0.517 | 11.3 ± 1.0 | 10.7–11.8 | 0.007 |
Values of P < 0.05 indicated a significant difference in results for control dogs versus the respective grouping of case dogs.
Box-and-whisker plots of CSI (A), short-MHS (B), long-MHS (C), overall-MHS (D), and VHS (E) for the small-breed dogs (60 control dogs and 4I case dogs [20 with right-sided cardiac disease and 2I with left-sided cardiac disease]) described in Figure 2. See Figure 3 for key.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.888
Cardiac variables (cSAL; cLAL; cSAL + cLAL; ML; CSI; short-, long-, and overall-MHS; and VHS) for the small-breed control dogs (n = 60) versus the small-breed case dogs collectively (41) and further subgrouped according to the side of the heart affected (right-sided [20] or left-sided [21] cardiac disease).
| Small-breed control dogs (n = 60) | Small-breed case dogs (n = 41) | Small-breed case dogs with right-sided cardiac disease (n = 20) | Small-breed case dogs with left-sided cardiac disease (n = 21) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Variables | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value* | Mean ± SD | 95% CI | P value* | Mean ± SD | 95% CI | P value* |
| cSAL (mm) | 64.3 ± 10.5 | 61.6–67.1 | 73.7 ± 19.2 | 67.6–79.7 | 0.002 | 70.5 ± 20.4 | 60.9–80.1 | 0.084 | 76.7 ± 17.9 | 68.5–84.8 | < 0.001 |
| cLAL (mm) | 79.6 ± 13.4 | 76.1–83.1 | 85.9 ± 23.2 | 78.6–93.2 | 0.089 | 77.3 ± 23.6 | 66.2–88.4 | 0.59 | 94.1 ± 20.1 | 84.9–103.2 | < 0.001 |
| cSAL + cLAL (mm) | 143.9 ± 23.4 | 1 37.9–150.0 | 159.5 ± 41.8 | 146.3–172.7 | 0.018 | 147.8 ± 43.7 | 127.4–168.2 | 0.616 | 170.7 ± 37.6 | 153.6–187.8 | < 0.001 |
| ML (mm) | 28.0 ± 6.5 | 26.3–29.7 | 28.8 ± 9.2 | 25.8–31.7 | 0.617 | 27.9 ± 11.1 | 22.6–33.3 | 0.976 | 29.6 ± 7.4 | 26.2–32.9 | 0.365 |
| CSI (%) | 8I.0 ± 6.0 | 80.0–83.0 | 86.0 ± 8.0 | 84.0–89.0 | < 0.001 | 92.0 ± 8.0 | 88.0–95.0 | < 0.001 | 8I.0 ± 6.0 | 79.0–84.0 | 0.910 |
| Short-MHS | 2.4 ± 0.4 | 2.3–2.5 | 2.7 ± 0.4 | 2.5–2.8 | < 0.001 | 2.7 ± 0.5 | 2.5–2.9 | 0.002 | 2.6 ± 0.4 | 2.5–2.8 | 0.005 |
| Long-MHS | 2.9 ± 0.5 | 2.8–3.0 | 3.1 ± 0.6 | 2.9–3.3 | 0.08I | 2.9 ± 0.5 | 2.7–3.2 | 0.89I | 3.3 ± 0.5 | 3.0–3.5 | 0.009 |
| Overall-MHS | 5.3 ± 0.8 | 5.1–5.5 | 5.8 ± 0.9 | 5.5–6.1 | 0.008 | 5.6 ± I.0 | 5.2–6.1 | 0.131 | 5.9 ± 0.9 | 5.5–6.3 | 0.006 |
| VHS (vertebrae) | 10.3 ± 0.8 | 10.1–10.5 | 11.3 ± 1.3 | 10.9–11.7 | < 0.001 | 10.6 ± 0.7 | 10.3–11.0 | 0.095 | 12.0 ± 1.4 | 11.3–12.6 | < 0.001 |
Values of P < 0.05 indicated a significant difference in results for control dogs versus the respective grouping of case dogs.
Control dogs versus case dogs grouped by breed size and side of the heart affected—When results for case dogs grouped by breed size (large-breed or small-breed dogs) and side of the heart affected (left-sided or right-sided cardiac disease) were compared with results for control dogs grouped by breed size, important differences were identified. In large-breed dogs, the mean ± SD CSI was significantly (P < 0.001) greater in case dogs with right-sided cardiac disease (88.0 ± 6.0%) versus control dogs (77.0 ± 5.0%), but did not differ significantly (P = 0.056) for case dogs with left-sided cardiac disease (80.0 ± 8.0%) versus control dogs (Figure 3; Table 4). The mean ± SD short-MHSs in large-breed case dogs with right-sided cardiac disease (2.4 ± 0.2) or left-sided cardiac disease (2.5 ± 0.3) were significantly (P = 0.001 and P < 0.001) higher than that in large-breed control dogs (2.1 ± 0.2). Further, in large-breed case dogs with left-sided cardiac disease, the mean ± SD long-MHS (3.1 ± 0.4), overall-MHS (5.6 ± 0.7), and VHS (11.3 ± 1.0) were significantly (P < 0.001, P < 0.001, and P = 0.007) greater than those in large-breed control dogs (2.7 ± 0.3, 4.8 ± 0.5, and 10.7 ± 0.5).
The mean ± SD CSI and short-MHS were significantly (P < 0.001 and P = 0.002) greater for small-breed case dogs with right-sided cardiac disease (92.0 ± 8.0% and 2.7 ± 0.5) versus small-breed control dogs (81.0 ± 6.0% and 2.4 ± 0.4; Figure 4; Table 5). However, small-breed case dogs with left-sided cardiac disease had significantly greater mean ± SD short-MHS (2.6 ± 0.4; P = 0.005), long-MHS (3.3 ± 0.5; P = 0.009), overall-MHS (5.9 ± 0.9; P = 0.006), and VHS (12.0 ± 1.4; P < 0.001), compared with small-breed control dogs (2.4 ± 0.4, 2.9 ± 0.5, 5.3 ± 0.8, and 10.3 ± 0.8, respectively).
On the basis of these findings and the 95% CI of the means for these variables in large- and small-breed control dogs, the combination of a CSI > 78.0% and > 83.0%, respectively, and a short-MHS ≥ 2.2 and ≥ 2.5, respectively, but with long- and overall-MHSs not meaningfully different than in control dogs, suggested the presence of right-sided cardiac disease in case dogs. However, in large- and small-breed dogs, the combination of a short-MHS ≥ 2.2 and ≥ 2.5, respectively; a long-MHS > 2.8 and ≥ 3.0, respectively; and an overall-MHS > 5.0 and ≥ 5.5, respectively, suggested the presence of left-sided cardiac disease in case dogs.
Case dogs grouped by breed size with right-sided versus left-sided cardiac disease—In large-breed case dogs, the mean ± SD CSI was significantly (P = 0.040; Figure 3; Table 6) higher in those with right-sided cardiac disease (88.0 ± 6.0%), compared with left-sided cardiac disease (80.0 ± 8.0%). However, the mean ± SD long-MHS was significantly (P = 0.048) higher in large-breed case dogs with left-sided cardiac disease (3.1 ± 0.4), compared with those with right-sided cardiac disease (2.8 ± 0.2) cardiac disease. No substantial differences were detected in the mean short-MHS, overall-MHS, VHS, age, or body weight between large-breed dogs with right-sided versus left-sided cardiac disease.
Cardiac variables (cSAL; cLAL; cSAL + cLAL; ML; CSI; short-, long-, and overall-MHS; and VHS) for large- and small-breed case dogs with right-sided versus left-sided cardiac disease.
| Large-breed case dogs (n = 23) | Small-breed case dogs (n = 41) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Right-sided cardiac disease (n = 7) | Left-sided cardiac disease (n = 16) | Right-sided cardiac disease (n = 20) | Left-sided cardiac disease (n = 21) | |||||||
| Variables | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value* | Mean ± SD | 95% CI | Mean ± SD | 95% CI | P value* |
| cSAL (mm) | 122.2 ± 8.8 | 113.0–131.5 | 122.3 ± 14.9 | 114.3–130.2 | 0.995 | 70.5 ± 20.4 | 60.9–80.1 | 76.7 ± 17.9 | 68.5–84.8 | 0.309 |
| cLAL (mm) | 138.6 ± 11.2 | 126.8–150.4 | 153.6 ± 20.6 | 142.6–164.5 | 0.111 | 77.3 ± 23.6 | 66.2–88.4 | 94.1 ± 20.1 | 84.9–103.2 | 0.019 |
| cSAL + cLAL (mm) | 260.8 ± 18.3 | 241.7–280.0 | 275.8 ± 33.0 | 258.2–293.4 | 0.3I | 147.8 ± 43.7 | 127.4–168.2 | 170.7 ± 37.6 | 53.6–187.8 | 0.079 |
| ML (mm) | 50.5 ± 5.2 | 45.0–55.9 | 50.0 ± 9.0 | 45.2–54.8 | 0.906 | 27.9 ± 1 1.1 | 22.6–33.3 | 29.6 ± 7.4 | 26.2–32.9 | 0.585 |
| CSI (%) | 88.0 ± 6.0 | 82.0–94.0 | 80.0 ± 8.0 | 76.0–85.0 | 0.04 | 92.0 ± 8.0 | 88.0–95.0 | 8I.0 ± 6.0 | 79.0–84.0 | < 0.001 |
| Short-MHS | 2.4 ± 0.2 | 2.2–2.7 | 2.5 ± 0.3 | 2.3–2.7 | 0.729 | 2.7 ± 0.5 | 2.5–2.9 | 2.6 ± 0.4 | 2.5–2.8 | 0.666 |
| Long-MHS | 2.8 ± 0.2 | 2.5–3.0 | 3.1 ± 0.4 | 2.9–3.3 | 0.048 | 2.9 ± 0.5 | 2.7–3.2 | 3.3 ± 0.5 | 3.0–3.5 | 0.07 |
| Overall-MHS | 5.2 ± 0.4 | 4.7–5.7 | 5.6 ± 0.7 | 5.2–6.0 | 0.186 | 5.6 ± I.0 | 5.2–6.1 | 5.9 ± 0.9 | 5.5–6.3 | 0.388 |
| VHS (vertebrae) | 10.9 ± 0.6 | 10.3–11.5 | 11.3 ± 1.0 | 10.7–11.8 | 0.42 | 10.6 ± 0.7 | 10.3–11.0 | 12.0 ± 1.4 | 11.3–12.6 | 0.001 |
Values of P < 0.05 indicated a significant difference between results for case dogs with left-sided versus right-sided cardiac disease in the respective breed-size group.
In small-breed case dogs, the mean ± SD CSI was significantly (P < 0.001; Figure 4; Table 6) higher in those with right-sided (92.0 ± 8.0%) versus left-sided (81.0 ± 6.0%) cardiac disease. Although there were no substantial differences in the short-, long-, or overall-MHSs in small-breed case dogs with left-sided versus right-sided cardiac disease, the mean ± SD VHS was significantly (P = 0.001) higher in case dogs with left-sided cardiac disease (12.0 ± 1.4), compared with those with right-sided cardiac disease (10.6 ± 0.7). In addition, small-breed case dogs with left-sided cardiac disease were significantly (P = 0.003) older (mean ± SD age, 9.6 ± 3.5 years; range, 1.6 to 15.6 years), compared with those with right-sided cardiac disease (mean ± SD age, 5.6 ± 4.7 years; range, 0.2 to 14.1 years). The mean body weight did not differ between small-breed case dogs with left-sided versus right-sided cardiac disease.
Discussion
In a recent study9 on healthy dogs, the cSAL and cLAL were expressed relative to each other in the cSAL:cLAL ratio (which is the CSI) and relative to the corresponding ML (yielding the MHSs) to mitigate differences in measured dimensions of radiographic cardiac silhouettes related to interbreed variation. The present study focused specifically on quantitative assessment of the sphericity (CSI) and dimensions (MHSs) of the cardiac silhouette on right lateral thoracic radiographic images for large- and small-breed dogs with and without moderate to severe right- or left-sided heart disease. Results of the present study indicated that the mean CSIs and MHSs differed between the control and case groups, suggesting that these values could be useful quantitative, objective measures to help identify cardiac disease in dogs. For instance, in the present study, the mean CSI was significantly (P < 0.001) greater in case dogs with right-sided cardiac disease, compared with control dogs, and was significantly (P = 0.040 and P < 0.001, respectively) greater in large- and small-breed case dogs with right-sided cardiac disease, compared with those with left-sided cardiac disease. Further, the mean long-MHS was significantly (P = 0.048) greater in large-breed case dogs with left-sided cardiac disease, compared with those with right-sided cardiac disease.
For large- and small-breed dogs in the present study, results with the combination of a CSI > 78.0% and > 83.0%, respectively, and a short-MHS ≥ 2.2 and ≥ 2.5, respectively, but with long- and overall-MHSs below the upper limit of the 95% CI for control dogs, was suggestive of the presence of right-sided cardiac disease with cardiomegaly affecting the right side of the heart. Therefore, the cardiac silhouette appeared rounder in case dogs with right-sided cardiomegaly, compared with that in control dogs or in case dogs with left-sided cardiac disease.
Our findings also indicated that the mean short-, long-, and overall-MHS and VHS were greater in case dogs with left-sided cardiac disease, compared with control dogs. In large- and small-breed dogs of the present study, the combination of a short-MHS ≥ 2.2 and ≥ 2.5, respectively, a long-MHS > 2.8 and ≥ 3.0, respectively, and an overall-MHS > 5.0 and ≥ 5.5, respectively, was suggestive of the presence of left-sided cardiac disease with cardiomegaly affecting the left side of the heart. These findings supported our hypothesis that the CSIs and MHSs would be higher in dogs with confirmed cardiac disease and would vary with the side of the heart affected. In addition, these findings also supported our hypothesis that different reference values for MHSs may be required for radiographically evaluating cardiac dimensions in the different breed-size groups (large-breed dogs or small-breed dogs).
The nonsignificant difference in the CSI between control dogs and case dogs with left-sided cardiac disease in the present study may have been because of greater lengths of cSAL and cLAL in individual case dogs. When both of these cardiac axes increase proportionally, their ratio (the CSI) would not be expected to change. Therefore, the CSI should not be used alone to evaluate cardiac size in dogs, but instead to obtain an estimate of the shape of the cardiac silhouette. Additionally, the lower mean CSI in case dogs with left-sided cardiac disease, compared with right-sided cardiac disease, could have been explained by the latter having had higher mean long-MHS and cLAL in the present study.
The cLAL spans the left atrium and left ventricle, whereas the cSAL spans the atria (most likely at the level of the coronary groove and atrioventricular valves) to include portions of the right and left atria.1,9,10,15,16 Therefore, the greater cSAL and cLAL in case dogs with left-sided cardiac disease in the present study could have been attributed to possible left atrial enlargement and associated left-sided cardiomegaly. Unlike large-breed case dogs with left-sided cardiac disease, small-breed case dogs with left-sided cardiac disease had significantly (P = 0.001) greater VHS, but no substantial difference in mean short-, long-, or overall-MHSs, compared with case dogs that had right-sided cardiac disease in the present study. This discrepancy in the results for VHS between large- and small-breed case dogs with left-sided cardiac disease could have been attributed to possible thoracic vertebral anomalies and intervertebral disk space narrowing that are commonly encountered in small-breed dogs.9,13 Furthermore, a recently identified weak association between VHS and the lengths of corresponding cardiac axes9 may have been another factor that contributed to this discrepancy.
Determining VHS and MHSs may not be useful in the diagnosis of cardiac diseases that develop without eccentric cardiac enlargement, such as in diseases with concentric hypertrophy of the left ventricle.1,9,17,18 However, to our knowledge, concentric hypertrophy of the left ventricle may still change the shape and size of the cardiac silhouette, and hence, echocardiography remains the foremost diagnostic tool to confirm heart diseases and associated cardiac structural abnormalities in dogs.10,19,20
In a previous report,13 determining VHS did not rely on the experience of investigators but rather relied on the individual selection of the reference points for measurements of cSAL and cLAL and then the transformation of these cardiac dimensions into VHS. The lack of testing of inter- and intraobserver variability associated with the MHS procedure is therefore considered a limitation of the present study. Further research is required to test the repeatability of the MHS procedure in dogs. An additional limitation was that dogs with mild cardiac disease (the most challenging level of severity of cardiac disease to diagnose with radiography) were not investigated in the present study. We suggest that future research involve a larger population of dogs with various cardiac diseases divided into 3 groups on the basis of disease severity (mild, moderate, and severe) as determined with echocardiography, for which results of CSIs and MHSs for each group would then be compared with results for control dogs. Such an investigation could evaluate for associations between severity of cardiac disease and calculated scores (CSIs, MHSs, and VHSs) for cardiac enlargement evident on thoracic radiography. Additionally, the CSI and MHSs may also be determined for dogs grouped by specific cardiac disease diagnosed with echocardiography.
The lack of echocardiography to rule out cardiac disease in control dogs in the present study was a limitation; however, including in control groups clinically normal or healthy animals that may not have undergone specific diagnostic procedures is often a drawback of a retrospective study design. Another limitation was that all measurements were performed only on the right lateral thoracic radiographic image of each dog and were not assessed on the corresponding ventrodorsal image. The left lateral thoracic image, however, was not used in the present study because of the typically rounder cardiac silhouette on the left lateral versus right lateral radiographic projection. Additionally, right versus left lateral recumbency does not significantly influence radiographic assessment of VHS in dogs.1
Another limitation was the lack of assessment of interbreed variation within each breed-size group; however, we expected that standardization of each cardiac axis by use of the corresponding ML would be beneficial in assessment of cardiac size in each dog breed within each breed-size group. Nonetheless, future evaluation of MHSs in specific dog breeds may still be required.
Physiologic variations in cardiac cycle and respiratory phase influence the radiographic assessment of cardiac size in healthy dogs and cats.21–24 These variations could make it difficult to distinguish between radiographically normal and enlarged cardiac silhouettes.21–23 In the present study, all radiographic images were assumed to have been taken at end-inspiratory tidal volume; however, the influences of cardiac cycle and volume overload on the radiographic appearance of the cardiac silhouette could have been another limitation of the present study and should therefore be considered when cardiac axes are evaluated by use of MHSs. Although a study11 shows that body condition score affected VHS in Lhaso Apsos, other studies1,14,25–28 show that age, sex, body weight, body condition score, and thoracic conformation (eg, wide vs deep) did not influence the mean VHSs for large-, medium-, and small-breed dogs. Therefore, the potential influences of these variables on the radiographic assessments of the cardiac silhouette were not analyzed in the present study.
Results indicated that MHSs could be useful, objective radiographic values to help assess dogs for potential heart disease, and we recommend that MHSs be added to the diagnostic tools used by veterinarians when screening for heart disease in dogs. In the present study, large- and small-breed case dogs with wide and long cardiac silhouettes and subsequently high MHSs generally had left-sided cardiomegaly, whereas those with only wide cardiac silhouettes and subsequently only high short-MHS generally had right-sided cardiomegaly and a relatively round heart. When evaluating cardiac silhouettes in right lateral thoracic radiographic images, practitioners should assess cardiac size and calculate the CSI to evaluate the sphericity of the heart. We recommend that the overall-, short-, and long-MHSs also be calculated to further assess overall cardiac size objectively. Nonetheless, echocardiography remains the noninvasive gold standard for assessment of cardiomegaly in dogs.
Acknowledgments
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CI | Confidence interval |
| cLAL | Cardiac long-axis length |
| cSAL | Cardiac short-axis length |
| CSI | Cardiac sphericity index |
| MHS | Manubrium heart score |
| ML | Manubrium length |
| VHS | Vertebral heart score |
Footnotes
Merge PACS, Merge Healthcare Inc, Chicago, Ill.
Prism, version 7, GraphPad Software Inc, La Jolla, Calif.
References
1. Buchanan JW, Bücheler J. Vertebral scale system to measure canine heart size in radiographs. J Am Vet Med Assoc 1995;206:194–199.
2. Root CR, Bahr RJ. The heart and great vessels. In: Thrall DE, ed. Textbook of diagnostic veterinary radiology. 4th ed. Philadelphia: WB Saunders Co, 2002;402–419.
3. Schober KE, Maerz I, Ludewig E, et al. Diagnostic accuracy of electrocardiography and thoracic radiography in the assessment of left atrial size in cats: comparison with trans-thoracic 2-dimensional echocardiography. J Vet Intern Med 2007;21:709–718.
4. Guglielmini C, Diana A, Santarelli G, et al. Accuracy of radiographic vertebral heart score and sphericity index in the detection of pericardial effusion in dogs. J Am Vet Med Assoc 2012;241:1048–1055.
5. Lamb CR, Boswood A. Role of survey radiography in diagnosing canine cardiac disease. Compend Contin Educ Pract Vet 2002;24:316–326.
6. Empel W. Beurteilung der Grosse und Gestalt des Herzschattens des Deutschen Schaferhundes. Kleintierpraxis 1974;19:7–13.
7. Castro MG, Tôrres RCS, Araújo RB, et al. Radiographic evaluation of the cardiac silhouette in clinically normal Yorkshire Terrier dogs through the vertebral heart size method. Arq Bras Med Vet Zootec 2011;63:850–857.
8. Spasojevic-Kosic L, Krstic N, Trailovic RD. Comparison of three methods of measuring vertebral heart size in German Shepherd Dogs. Acta Vet (Beogr) 2007;57:133–141.
9. Mostafa AA, Berry CR. Radiographic assessment of the cardiac silhouette in clinically normal large- and small-breed dogs. Am J Vet Res 2017;78:168–177.
10. Gugjoo MB, Hoque M, Saxena AC, et al. Vertebral scale system to measure heart size in dogs in thoracic radiographs. Adv Anim Vet Sci 2013;1:1–4.
11. Jepsen-Grant K, Pollard RE, Johnson LR. Vertebral heart scores in eight dog breeds. Vet Radiol Ultrasound 2013;54:3–8.
12. Lamb CR, Wikeley H, Boswood A, et al. Use of breed-specific ranges for vertebral heart scale as an aid to the radiographic diagnosis of cardiac disease in dogs. Vet Rec 2001;148:707–711.
13. Hansson K, Haggstrom J, Kvart C, et al. Interobserver variability of vertebral heart size measurements in dogs with normal and enlarged hearts. Vet Radiol Ultrasound 2005;46:122–130.
14. Gulanber EG, Gonenci R, Kaya U, et al. Vertebral scale system to measure heart size in thoracic radiographs of Turkish Shepherd (Kangal) dogs. Turk J Vet Anim Sci 2005;29:723–726.
15. Saida Y, Tanaka R, Yamane Y, et al. Relationships between vertebral heart size (VHS) and echocardiographic parameters in dogs with mitral regurgitation. Adv Anim Cardiol (Jpn) 2006;39:55–63.
16. Srivastava M, Pandey RP, Srivastava A, et al. Vertebral heart scale score of mongrel dogs. Indian J Canine Pract 2014;6:5–8.
17. Lamb CR, Tyler M, Boswood A, et al. Assessment of the value of the vertebral heart scale in the radiographic diagnosis of cardiac disease in dogs. Vet Rec 2000;146:687–690.
18. Buchanan JW. Vertebral scale system to measure heart size in radiographs. Vet Clin North Am Small Anim Pract 2000;30:379–393.
19. Kayar A, Gonul R, Or ME, et al. M-mode echocardiographic parameters and indices in the normal German Shepherd Dog. Vet Radiol Ultrasound 2006;47:482–486.
20. Muzzi RA, Muzzi LA, de Arau'jo RB, et al. Echocardiographic indices in normal German Shepherd Dog. J Vet Sci 2006;7:193–198.
21. Silverman S, Suter PF. Influence of inspiration and expiration on canine thoracic radiographs. J Am Vet Med Assoc 1975;166:502–510.
22. Toal RL, Losonsky JM, Coulter DB, et al. Influence of cardiac cycle on the radiographic appearance of the feline heart. Vet Radiol Ultrasound 1985;26:63–69.
23. Toombs JP, Ogburn PN. Evaluating canine cardiovascular silhouettes: radiographic methods and normal radiographic anatomy. Compend Contin Educ Pract Vet 1985;7:579–587.
24. Olive J, Javard R, Specchi S, et al. Effect of cardiac and respiratory cycles on vertebral heart score measured on fluoroscopic images of healthy dogs. J Am Vet Med Assoc 2015;246:1091–1097.
25. Bavegems V, Van Caelenberg AV, Duchateau L, et al. Vertebral heart size ranges specific for Whippets. Vet Radiol Ultrasound 2005;46:400–403.
26. Greco A, Meomartino L, Raiano V, et al. Effect of left vs right recumbency on the vertebral heart score in normal dogs. Vet Radiol Ultrasound 2008;49:454–455.
27. Webster N, Adams V, Dennis R. The effect of manual lung inflation vs spontaneous inspiration on the cardiac silhouette in anesthetized dogs. Vet Radiol Ultrasound 2009;50:172–177.
28. Ghadiri A, Avizeh R, Fazli G. Vertebral heart scale of common large breeds of dogs in Iran. Int J Vet Res 2010;4:107–111.
