Caudal pulmonary artery to vein ratio on radiography can predict pulmonary hypertension in dogs with mitral regurgitation

Pulmonary hypertension (PH) is an important predictor of poor outcomes in dogs with mitral regurgitation (MR). The feasibility of radiography to predict PH in dogs with MR is unknown. This retrospective, observational, and analytic study aimed to identify a radiographic parameter to predict PH in dogs with MR. A total of 302 dogs diagnosed with MR on echocardiography were enrolled. Medical record and radiographic findings such as the size of the main pulmonary artery, left atrium, left ventricle, and right chamber, and cranial and caudal pulmonary arteries and veins were evaluated according to the presence of PH. The diameters of the cranial and caudal pulmonary vessels were compared to the fourth rib and the ninth rib, respectively, and the ratio of the pulmonary artery to the corresponding vein (CdPA/CdPV) was calculated. Pulmonary hypertension was diagnosed in 77 dogs (25.5%) and the prevalence of PH increased with MR grade. The CdPA/CdPV was significantly higher (P < 0.001) in the presence of PH. Multivariate analysis showed that the CdPA/CdPV was the only independent radiographic parameter that had a significant association with PH in dogs with MR (P = 0.028). The cut-off value of the CdPA/CdPV = 1.10 showed 90.6% specificity and 31.1% sensitivity for detecting PH in dogs with MR. In dogs with MR, PH can be predicted with high specificity when the caudal pulmonary artery is 1.1 times larger than the corresponding vein on radiographs.     

ACCURACY OF A COMPUTED TOMOGRAPHY BRONCHIAL WALL THICKNESS TO PULMONARY ARTERY DIAMETER RATIO FOR ASSESSING BRONCHIAL WALL THICKENING IN DOGS

Computed tomography is increasingly being used in veterinary medicine to evaluate animals with pulmonary signs such as coughing, tachypnea, and exercise intolerance, however, a quantitative measure of bronchial wall thickening has yet to be validated in veterinary medicine. Canine chronic bronchitis is a disease that is characterized histologically by thickening of the bronchial walls. Thoracic CT images of 16 dogs with chronic bronchitis and 72 dogs presenting for conditions unrelated to cough were evaluated. A ratio comparing the bronchial wall thickness to the adjacent pulmonary artery diameter was obtained in the right and left cranial and caudal lung lobes. There was no significant difference in dogs with chronic bronchitis or unaffected dogs between the left and right hemithorax, patient weight, patient age, image slice thickness, or CT machine used. Dogs with chronic bronchitis were found to have a significantly greater ratio than unaffected dogs (P < 0.001). The ratios in the cranial lung lobes were found to be significantly greater than the caudal lung lobes in both chronic bronchitis and unaffected dogs (P < 0.001). A receiver operating characteristic curve of the ratios in the cranial lung lobes had an area under the curve of 0.912, indicating high accuracy in predicting for bronchial wall thickening. A ratio of ≥0.6 in the cranial lung lobes was found to have a sensitivity of 77% and specificity of 100% in predicting for the presence of chronic bronchitis, and we propose using this cut‐off as supportive of bronchial wall thickening on CT.     

PULMONARY HYPERTENSION IN DOGS WITH MITRAL REGURGITATION ATTRIBUTABLE TO MYXOMATOUS VALVE DISEASE

Pulmonary hypertension has been associated with mitral insufficiency caused by chronic degenerative valve disease in dogs. Our aim was to search for associations between left atrial to aortic root ratio, end‐systolic and end‐diastolic volume indices, and changes in the right ventricular to right atrial pressure gradient as estimated by the peak velocity of tricuspid regurgitation in dogs with chronic degenerative valve disease and different classes of heart failure. Dogs, for which follow‐up was available were evaluated for changes in the right ventricular to right atrial systolic pressure gradient over time. Three hundred and forty‐four dogs were studied; 51 in the International Small Animal Cardiac Health Council class 1a, 75 in class 1b, 113 in class 2, 97 in class 3a, and 8 in class 3b. The mean values for right ventricular to right atrial systolic pressure gradient, end‐systolic volume index, end‐diastolic volume index, and left atrial to aortic ratio were 49.2 ± 17.1 mmHg, 149.12 ± 60.8 and 37.7 ± 21.6 ml/m², and 1.9 ± 0.5, respectively. A weak positive correlation was found between the right ventricular to right atrial systolic pressure gradient and the left atrial to aorta ratio (r=0.242, P<0.0001), end‐diastolic volume index (r=0.242, P<0.0001), and end‐systolic volume index (r=0.129, P<0.001). Follow up was available for 49 dogs. Of these, 18 had an increased, 12 a decreased, and 19 a stable right ventricular to right atrial systolic pressure gradient despite therapy. The equivalence point between the sensitivity and specificity curves of about 80% in the coincident point corresponded to a right ventricular to right atrial systolic pressure gradient of 48 mmHg. Our results suggest an association between the progressive nature of chronic degenerative mitral valve disease and pulmonary hypertension. It is of clinical interest that, with a right ventricular to right atrial systolic pressure gradient pressure gradient at or above 48 mmHg, pulmonary hypertension does not appear to improve despite therapy targeted at lowering the left atrial load.     

NONINVASIVE ESTIMATION OF CENTRAL VENOUS PRESSURE IN ANESTHETIZED DOGS BY MEASUREMENT OF HEPATIC VENOUS BLOOD FLOW VELOCITY AND ABDOMINAL VENOUS DIAMETER

Determination of central venous pressure (CVP) is relevant to patients with right heart disease, hypovolemia, and following intravenous fluid therapy. We hypothesized that changes in CVP in dogs could be predicted by measurements of hepatic vein diameter, caudal vena cava (CVC) diameter, and hepatic venous flow velocities. Nine healthy American Foxhounds were anesthetized. Following baseline recordings, intravenous fluids were administered to increase CVP. Volume administration created treatment periods with CVP ranges of 5, 10, 15, 20, and 25 mm Hg. Flow velocities in the right medial hepatic vein were recorded using pulsed wave Doppler ultrasound. Hepatic vein, CVC, and aorta diameters were determined with B‐mode ultrasound. Variables were compared across the treatment periods by ANOVA for repeated measures. Relationships between CVP, Doppler, and B‐mode variables were evaluated using Spearman's rank correlations, multiple linear regression, and repeated measures linear regression. The a‐, S‐ and v‐wave velocities were augmented significantly with volume loading. The best part (semipartial) correlation coefficients predicting increasing CVP were identified with v‐wave velocity (0.823), S‐wave velocity (?0.800), CVC diameter (0.855), and hepatic vein diameter (0.815). Multiple linear regression indicated that CVP in this study could be predicted best by a combination of CVC and hepatic vein diameter and the v‐wave velocity (r=0.928). Ultrasound imaging identified gallbladder and pancreatic edema consistently, likely related to acute volume loading. These findings may be applicable in the assessment of volume status, dogs with right heart disease, and during serial monitoring of dogs receiving fluid or diuretic therapy.     

RADIOGRAPHIC EVALUATION OF CAUDAL VENA CAVA SIZE IN DOGS

Dilation of the caudal vena cava (CVC) on lateral thoracic radiographs is often interpreted as suggestive of right-sided congestive heart failure. To quantitate the clinical utility of evaluating CVC size as an indicator of right-sided heart disease, we compared the ratio of the diameter of the CVC as measured on a left lateral thoracic radiograph to the descending aorta (Ao), length of the thoracic vertebra above the tracheal bifurcation (VL), and width of the right fourth rib (R4) in 35 dogs with right heart disease and 35 control dogs. Each CVC ratio (CVC/Ao, CVC/VL, CVC/R4) was statistically larger in dogs with right heart disease. Response operating characteristic curves and likelihood ratios were used to determine ratios helpful in identifying dogs with right heart disease. A CVC/Ao > 1.50, CVC/VL > 1.30, or CVC/R4 > 3.50 are strongly suggestive of a right-sided heart abnormality in a patent.     

THORACIC RADIOGRAPHIC ABNORMALITIES IN 200 DOGS WITH SPONTANEOUS HEARTWORM INFESTATION

Thoracic radiographs of 200 dogs with spontaneously occurring heartworm disease were reviewed. Radiographs of 28 dogs (14%) were normal. In the remaining 172 dogs various combinations of cardiopulmonary abnormalties were found. The most frequently observed combination, noted in 61 of 200 dogs, was right ventricular, main pulmonary, and right cranial lobar pulmonary artery enlargemetn. Dogs with severe increse of one parameter generally had severe increase of the other two parameters also. One hundred five of the 200 dogs had a right cranial lobar pulmonary artery of normal size. Thus, right cranial lobar pulmonary artery size, when normal, is not a sensitive indicator of the absence of heartworm disease. There was a statistically significant positive relation between the size of the caudal vena cava and that of the right ventricle.     

COMPUTED TOMOGRAPHY BRONCHIAL LUMEN TO PULMONARY ARTERY DIAMETER RATIO IN DOGS WITHOUT CLINICAL PULMONARY DISEASE

Bronchiectasis is diagnosed in humans using multiple computed tomography (CT) criteria, the most important being dilatation of the bronchi. The most widely used criterion for detection of bronchial dilatation is a bronchial lumen to pulmonary artery diameter (bronchoarterial [BA]) ratio >1.0. No studies have been performed to determine the BA ratio in normal dogs. Thoracic CT images of 24 dogs without clinical pulmonary disease were reviewed. The BA ratio of the lobar bronchi of the left cranial (cranial and caudal parts), right cranial, right middle, left caudal, and right caudal lung lobes was measured. The mean of the mean BA ratio for all dogs was 1.45±0.21 (99% confidence interval [CI]=1.34–1.56). The mean of the mean BA ratio as determined by lung lobe was 1.45±0.04 (99% CI=1.41–1.49). The range of individual BA ratios was 0.8–2.0. There was no significant difference in mean BA ratios as a function of lung lobe ( P =0.60). The BA ratio in these clinically normal dogs was consistent and may be a useful tool in evaluating for bronchiectasis on CT images. BA ratios >2.0 were not identified in this population, suggesting a threshold to differentiate normal from abnormal bronchi.     

PULMONARY ANGIOGRAPHY WITH 64‐MULTIDETECTOR‐ROW COMPUTED TOMOGRAPHY IN NORMAL DOGS

Pulmonary angiography using 64‐multidetector‐row computed tomography (MDCT) was used to evaluate pulmonary artery anatomy, and determine the sensitivity of pulmonary artery segment visualization in four Beagle dogs using images reconstructed to 0.625 mm and retro‐reconstructed to 1.25 and 2.5 mm slice thickness. Morphologically, characteristic features included a focal narrowing in the right cranial pulmonary artery in all dogs, which should not be mistaken as stenosis. While the right cranial pulmonary artery divided into two equally sized branches that were tracked into the periphery of the lung lobe in all dogs, only a single left cranial (cranial portion) lobar artery was present. Compared with 1.25 and 2.5 mm retro‐reconstructions, 0.625 mm reconstructions allowed for detection of significantly (P≤0.05) more pulmonary artery segments and sharper depiction of vessel margins. Clinical applications such as prevalence and significance of diameter changes, and detection of pulmonary arterial thrombembolism on lobar and sublobar level, using pulmonary angiography with 64‐MDCT applying 0.625 mm reconstruction slice thickness remain to be established.     

Short-Term Hemodynamic Evaluation of Circumferential Mitral Annuloplasty for Correction of Mitral Valve Regurgitation in Dogs

Objective —To determine the short-term hemodynamic effects associated with circumferential mitral annuloplasty (CMA) in dogs with mitral regurgitation.
Study Design —Prospective experimental study.
Animals —Seven healthy adult mongrel dogs.
Methods —Mitral regurgitation was surgically induced, and annular dilation occurred. Echocardiography and cardiac catheterization were used to determine forward ejection fraction (FEF), regurgitant fraction (RF), pulmonary capillary wedge pressure (PCWP), and annular diameter before and immediately after CMA in five dogs. FEF and RF were also evaluated 7 days after annuloplasty.
Results —Mean annular diameter and PCWP were significantly reduced immediately after CMA. Significant increases in FEF of 19% and 22% were shown immediately and 7 days after CMA. Significant reductions in RF of 19% and 22% were also shown immediately and 7 days after annuloplasty.
Conclusions —Sustained hemodynamic benefits and a reduction in annular diameter were achieved by CMA in a canine model of mitral regurgitation.
Clinical Implications—CMA may be a suitable treatment for heart failure because of mitral regurgitation when early signs of cardiovascular decompensation persist despite appropriate medical management.     

Determination of heart dimensions in cattle via 2-D-mode echocardiography

The goal of this study was to determine via echocardiography the size of the left and right cardiac ventricles and the width of the interventricular septum and the left free ventricular wall in 51 healthy cows. The heart regions were examined in standing cows using a 3.0 Mhz sector transducer in 2-D-Mode. The dimensions of the heart were measured in the caudal long and short axes on the right side and in the caudal and cranial long axes on the left. The diameter of the ventricles was determined in a plane immediately beneath the mitral or tricuspid valves and that of the aorta and pulmonary artery in a plane immediately above the aortic and pulmonary valves, respectively. At the end of the study, all of the cows were slaughtered, the hearts were removed and the same parameters were determined using a tape measure. Results of in vivo and in vitro measurements were compared. In the right caudal long axis, the diameters of the left ventricle during both diastole (x +/- s = 7.0 +/- 0.73 cm) and systole (4.5 +/- 0.69 cm) were larger than those of the right ventricle during diastole (4.1 +/- 1.02 cm) and systole (3.6 +/- 0.98 cm). The diameter of the ventricles during diastole was larger than that during systole. Analogous results were obtained in both other axes. The diameter of the right ventricle during systole was larger when measured in the right caudal long axis (3.6 +/- 0.98 cm) than in the right caudal short axis (3.2 +/- 1.15 cm). This was also true for measurements obtained during diastole. The interventricular septum and the left ventricular wall were thicker during systole than during diastole. The diameter of the pulmonary artery was larger during diastole (5.6 +/- 0.82 cm) than systole (5.2 +/- 0.84 cm). The diameter of the aorta was smaller than that of the pulmonary artery and did not change significantly during diastole (4.9 +/- 0.92 cm) and systole (4.8 +/- 0.80 cm). The diameters of both ventricles measured at post mortem were smaller than those measured in vivo during diastole and larger than those measured during systole. There were no significant differences between the measurements performed twice, three days apart, in 11 of the cows.     

Standard planes for ultrasonographic examination of the portal system in dogs

A scanning protocol for the systematic ultrasonographic examination of the portal system in dogs was developed. Seven planes were used to image the portal system. With the dogs in left lateral recumbency, 3 transverse planes obtained via the right intercostal spaces were used to visualize the portal vein and right portal branch, and a longitudinal plane obtained with the transducer caudal to the last right rib was used to visualize the portal bifurcation. With the dogs in dorsal recumbency, a longitudinal plane was used as an alternative method of visualizing the portal vein and its bifurcation. Finally, with the dogs in right lateral recumbency, longitudinal planes obtained with the transducer in the left flank were used to visualize the hepatic artery, the left renal vein, and the left testicular or ovarian vein. To diagnose or rule out portosystemic shunting, the right portal branch, the left testicular or ovarian vein, the portal vein immediately caudal to the portal bifurcation, and the portal vein at the level of the celiac artery should be examined with this scanning protocol.     

COMPUTED TOMOGRAPHIC AND RADIOGRAPHIC CHARACTERISTICS OF AORTIC LESIONS IN 42 DOGS WITH SPIROCERCOSIS

Spirocerca lupi is a common cause of vomiting, regurgitation, and sudden death in dogs that live in tropical or subtropical regions. Sudden death due to aortic rupture may occur with no preceding clinical signs. The purpose of this prospective study was to compare radiographic and computed tomographic (CT) characteristics of aortic lesions in a cohort of 42 dogs with endoscopically confirmed spirocercosis. Dorsoventral and right lateral recumbent thoracic radiographic findings were compared with pre‐ and postcontrast thoracic CT findings. Aortic mineralization was detected using CT in 18/42 dogs (43%). Three dogs had faint diffuse aortic wall mineralization. Using CT as the reference standard, radiographs had a sensitivity and specificity of 6% and 96%, respectively, for detecting aortic mineralization. A total of 20 aortic aneurysms were detected using CT in 15/42 dogs (36%). Using CT as the reference standard, radiographs had a sensitivity and specificity of 86% and 56%, respectively, for detecting aortic aneurysms. Respiratory motion, aortic displacement by esophageal masses and Spirocerca nodules adjacent to the aorta mimicked aneurysm formation on radiographs. Aortic thrombi were seen in two dogs in postcontrast CT images. Findings from this study indicated that aortic mineralization and aneurysm formation are common in dogs with spirocercosis. Findings also supported the use of pre‐ and postcontrast CT as effective methods for detecting and characterizing these lesions.     

Evaluation of collagenase-induced mitral valve regurgitation in dogs

OBJECTIVE: To investigate the hemodynamic changes induced by injecting collagenase into the mitral valve to induce mitral valve regurgitation (MVR) in dogs. ANIMALS: 9 healthy Beagles. PROCEDURE: Dogs were randomly assigned to 3 groups: control (saline [0.9% NaCl] solution; n = 3), single collagenase injection (C1; 3), and 2 collagenase injections (C2; 3). Open-heart surgery was performed, and saline or collagenase solutions were injected into the mitral valve. Before and weekly for 11 weeks after surgery, radiography, echocardiography, and phonocardiography were performed. Mean pulmonary arterial pressure and mean pulmonary arterial wedge pressure (mPAWP) were measured before and 11 weeks after surgery. Postmortem examinations were performed after dogs were euthanatized. RESULTS: No changes were detected in the control group during the 11-week follow-up period. A systolic murmur and MVR developed 1 week after surgery in groups C1 and C2. The murmur changed from a protosystolic to a pansystolic murmur, and left atrial diameter and the left atrial-to-aortic root diameter ratio increased with time. Mean pulmonary arterial pressure and mPAWP were greater 11 weeks after surgery in groups C1 and C2, compared with presurgery values. During necropsy, tissue loss was detected in the mitral valve at the site of collagenase injection. Degree of regurgitation corresponded to lesion size. CONCLUSIONS AND CLINICAL RELEVANCE: Injection of collagenase into the mitral valve of healthy dogs induced MVR, and dogs with MVR developed progressive hemodynamic changes without acute overload. Collagenase-induced MVR may be an appropriate model for evaluation of prognostic markers of idiopathic MVR in dogs.     

IMAGING DIAGNOSIS—COMPUTED TOMOGRAPHIC ANGIOGRAPHY CHARACTERISTICS OF MULTIPLE VASCULAR ANOMALIES IN A SENIOR DOG WITH LATE‐ONSET REGURGITATION

A 10‐year‐old dog weighing 3.4 kg presented with intermittent regurgitation. Esophagography revealed that the thoracic esophagus was compressed dorsally at the region of the fourth intercostal space and segmentally dilated from the second to third intercostal region. Three‐dimensional computed tomographic (CT) angiography confirmed a suspected vascular ring anomaly and also revealed multiple other vascular anomalies. These included aberrant right subclavian artery, absence of bilateral external jugular veins, right‐gastric caval shunt, and a completely duplicated caudal vena cava. Findings supported the use of thoracic CT angiography to rule out additional vascular malformations in dogs with suspected vascular ring anomaly.     

Comparison of a computed tomographic pulmonary trunk to aorta diameter ratio with echocardiographic indices of pulmonary hypertension in dogs

There are limited criteria for the detection of pulmonary hypertension in dogs undergoing computed tomography (CT) for pulmonary disease. This retrospective analytical exploratory study compared a CT pulmonary trunk to aorta ratio with echocardiographic estimates of pulmonary hypertension. Dogs having both a contrast thoracic CT and echocardiogram were selected and maximal pulmonary trunk and descending aorta diameters were measured by two observers on a single transverse CT image. Computed tomographic diameter ratios were compared with the echocardiographic parameters of tricuspid regurgitation gradient, right ventricular acceleration time‐to‐ejection time ratio, pulmonary insufficiency gradient, and pulmonary artery to aorta diameter. A total of 78 dogs were sampled, with 44 dogs having one or more finding suggestive of pulmonary hypertension. A moderate positive correlation was shown between tricuspid regurgitation gradient and CT pulmonary trunk to aorta ratio (r = 0.61, P‐value < 0.0001). Mean CT pulmonary trunk to aorta ratio of dogs with moderate (P = 0.0132) and severe (P < 0.0003) pulmonary hypertension were significantly higher than normal dogs. There was no significant difference in mean CT pulmonary trunk to aorta ratio between normal and mild pulmonary hypertension dogs (P = 0.4244). The intraclass correlation coefficient (0.72) showed good reproducibility of the ratio. Findings indicated that CT pulmonary trunk to aorta ratio is a reproducible and potentially useful method to predict moderate and severe pulmonary hypertension in dogs, but not mild pulmonary hypertension. In dogs undergoing thoracic CT for pulmonary disease, an increased ratio should prompt follow up echocardiography.     

Diagnostic value of echo-Doppler and tissue Doppler imaging in dogs with pulmonary arterial hypertension

BACKGROUND: Diagnosis of pulmonary arterial hypertension (PAH) relies on Doppler measurement of pulmonic and tricuspid regurgitation (TR). However, these are not always detectable. HYPOTHESIS: Tissue Doppler imaging (TDI), a novel noninvasive ultrasound technique, provides indirect but sensitive and specific assessment of elevated systolic pulmonary artery pressure (SPAP) in dogs. ANIMALS: One hundred and five dogs with TR. METHODS: Prospective observational study. Dogs were categorized as presenting normal (group 1, n = 45), mildly increased (group 2, n = 19), or moderately to severely increased (group 3, n = 41) SPAP, based on TR peak velocities (< 2.5, 2.5-3.0, and > 3.0 m/s, respectively). Ten quantitative echo-Doppler- and TDI-derived variables were assessed, including the main pulmonary arterial diameter to aortic diameter ratio, pulmonary flow acceleration time, and acceleration-to-ejection time ratio, the Tei index of right ventricular function, and 6 longitudinal basal right ventricular TDI variables. RESULTS: A significant correlation was observed between SPAP and each of the 10 tested variables (P < .05). Conventional echo-Doppler variables were less discriminating than the TDI for predicting increased SPAP. The combined systolic and diastolic right TDI index had the highest sensitivity and specificity (89% and 93% respectively, for a cutoff of 11.8 cm/s) and could discriminate between dogs in group 1 from dogs in group 2. CONCLUSIONS AND CLINICAL IMPORTANCE: TDI provided effective predictors of systolic PAH and demonstrated that both alterations in right-sided systolic and diastolic myocardial function can occur with mild increases in SPAP.     

EFFECTS OF INTRAVENOUS DEXMEDETOMIDINE ON CARDIAC CHARACTERISTICS MEASURED USING RADIOGRAPHY AND ECHOCARDIOGRAPHY IN SIX HEALTHY DOGS

Dexmedetomidine is a highly specific and selective α2‐adrenergic receptor agonist widely used in dogs for sedation or analgesia. We hypothesized that dexmedetomidine may cause significant changes in radiographic and echocardiographic measurements. The objective of this prospective cross‐sectional study was to test this hypothesis in a sample of six healthy dogs. Staff‐owned dogs were recruited and received a single dose of dexmedetomidine 250 μg/m² intravenously. Thoracic radiography and echocardiography were performed 1 h before treatment, and repeated 10 and 30 min after treatment, respectively. One observer recorded cardiac measurements from radiographs and another observer recorded echocardiographic measurements. Vertebral heart score and cardiac size to thorax ratio on the ventrodorsal projection increased from 9.8 ± 0.6 v to 10.3 ± 0.7 v (P = 0.0007) and 0.61 ± 0.04 to 0.68 ± 0.03 (P = 0.0109), respectively. E point‐to‐septal separation and left ventricle internal diameter in diastole and systole increased from 2.4 ± 1.1 to 6.6 ± 1.9 mm, 32.3 ± 8.1 to 35.5 ± 8.8 mm, and 19.4 ± 6 to 27.0 ± 7.2 mm, respectively (P < 0.05). Fractional shortening and sphericity index decreased from 40.7 ± 5.8 to 24.4 ± 2.9%, and 1.81 ± 0.07 to 1.58 ± 0.04, respectively (P < 0.05). Moderate‐to‐severe mitral regurgitation and mild pulmonic regurgitation occurred in all dogs after dexmedetomidine administration. Findings indicated that dexmedetomidine could cause false‐positive diagnoses of valvular regurgitation and cardiomegaly in dogs undergoing thoracic radiography and echocardiography.     

Associations between thoracic radiographic changes and severity of pulmonary arterial hypertension diagnosed in 60 dogs via Doppler echocardiography: A retrospective study

Doppler echocardiography is a noninvasive method for estimating and grading pulmonary arterial hypertension. No current literature associates significance of radiographic findings with severity of pulmonary arterial hypertension. We hypothesized that the number and conspicuity of radiographic findings suggestive of pulmonary arterial hypertension would be greater based on the severity of pulmonary arterial hypertension. Dogs with pulmonary arterial hypertension and normal control dogs were included in this retrospective, case control study. Three radiologists blinded to echocardiographic results scored thoracic radiographs for right ventricular and main pulmonary artery enlargement and pulmonary lobar artery enlargement, tortuosity, and blunting by multiple methods. Presence or absence of each finding was scored in an additive fashion and averaged for each grade of pulmonary arterial hypertension severity. Seventy‐one dogs (60 dogs with pulmonary arterial hypertension and 11 control dogs) of which some had multiple studies were included: 20 mild, 21 moderate, 25 severe, and 11 absent pulmonary arterial hypertension. The following radiographic findings were significantly associated with increasing pulmonary arterial hypertension severity: right ventricular enlargement by “reverse D” and “3/5–2/5 cardiac ratio” methods, main pulmonary artery enlargement, and caudal lobar artery enlargement by the “3rd rib” method. Mean scores for severe pulmonary arterial hypertension and normal dogs were significantly different (P‐value < 0.0001). Mean scores between different pulmonary arterial hypertension grades increased with severity but were not statistically significant. Individually and in combination, radiographic findings performed poorly in differentiating severity of pulmonary arterial hypertension. Findings indicated that thoracic radiographs should be utilized in conjunction with Doppler echocardiography in a complete diagnostic work‐up for dogs with suspected pulmonary arterial hypertension.