Esophageal Varices due to a Probable Arteriovenous Communication in a Dog

A 6-year-old 40-kg castrated male Samoyed dog was presented for evaluation of chronic remittent lameness of the right forelimb. The dog had a history of polyuria-polydipsia (PU/PD) and lethargy over the previous year. Physical examination was unremarkable except for a grade II systolic murmur at the apex of the heart on the left side. A 6-lead ECG was within normal limits. Arterial blood gas analysis indicated mild respiratory alkalosis (pH, 7.457, reference range, 7.370– 7.450; pCO2, 29.2 mm Hg, reference range, 30.6– 39.1 mm Hg; pO2, 90 mm Hg, reference range, 81–105 mm Hg; HCO3, 20.9 mmol/L; reference range, 19.1– 25 mmol/L). No clinically relevant abnormalities were found on CBC, serum biochemistry, serum protein electrophoresis, or hemostasis profile. Urinalysis revealed markedly reduced urine osmolality (142 mOsm/ kg; reference range, 600–2400 mOsm/kg), consistent with the PU/PD, reported by the owner. Both the urinary cortisol/creatinine ratio and urinary bile acid concentration were within normal limits, excluding hyperadrenocorticism and hepatic failure as underlying causes of the PU/PD. Thoracic radiographs disclosed a generalized increase in the size of the cardiac silhouette and pulmonary vascular enlargement, suggesting pulmonary overcirculation. An ELISA for Dirofilaria immitis antigen was negative. Two-dimensional, M-mode echocardiography (transducer frequency, 2.0–3.0 MHz) revealed left atrial enlargement, left ventricular eccentric hypertrophy, and impaired systolic function (end-diastolic dimension, 65.7 mm; end-systolic dimension, 44.2 mm; shortening fraction, 32.7%) with normal valves. Spectral and colorflow Doppler examination disclosed mild mitral, aortic, and pulmonic valve insufficiency. Both the tricuspid and the telediastolic pulmonic valvular peak regurgitant jet velocities were increased as follows: 3.26 m/s (normal, #2.5 m/s) and 2.44 m/s (normal, #2.0 m/s), respectively. According to the Bernoulli’s equation modification, the systolic pulmonary artery pressure was estimated to be 42.5 mm Hg and the diastolic pulmonary artery pressure was estimated to be 23.9 mm Hg, values consistent with mild pulmonary hypertension. The dog was anesthetized and subjected to radiography of the right forelimb and total body multidetector computed tomography (MDCT). The radiographs were negative for abnormalities, and synovial fluid examination of the shoulder and stifle joints did not indicate evidence of any inflammatory pathology. MDCT scans of the brain, thorax, and abdomen were obtained. For the thoracic and abdominal scans, the dog was positioned in dorsal recumbency, and we employed the following parameters: helical modality, 120 kV, 200 mA, 0.7-second rotation tube, 0.526 pitch, and 1.2-mm slice thickness. For the brain scan, the dog was positioned in sternal recumbency and the scanner parameters were as follows: axial modality, 120 kV, 310 mA, 2-second rotation tube, 0.625 slice thickness, and 10-mm intervals. For an enhanced series, 2-mL/kg iodixanol 320 mg I/mL was injected via a 22-gauge catheter into the right cephalic vein at a 3 mL/second infusion rate, through a computed tomography injector system. The brain and abdominal MDCT scans were normal. However, MDCT of the neck and chest revealed 12 pairs of ribs, an enlarged heart, and enlarged pulmonary vessels. Both of the bronchoesophageal arteries were enlarged and connected with an enormous network of homogeneously enhancing serpentine structures involving the thoracic esophagus (esophageal and paraesophageal varices). The bronchoesophageal vein was extremely dilated (Figs 1, 2). The right azygous and hemizygous veins were normal. The cranial vena cava was dilated as was the cervical vertebral venous system, which protruded into the vertebral canal. These findings were consistent with an arteriovenous communication (single or multiple fistulas, possibly between the thoracic aorta and the azygous system) with resultant venous distension and esophageal varicosity formation. To further examine for suspected esophageal varices, we subjected the dog to an immediate videoendoscopic examination while still anesthetized. The proximal esophagus was of normal diameter and had normal mucosal features. However, numerous tortuous submucosal structures protruding into the esophageal lumen consistent with esophageal varices were encountered in the distal 3rd of the esophagus. Varix diameter ranged from 1 to 4 mm, and varices are submucosal esophageal vessels. Endoscopic criteria predictive factor for variceal bleeding include the size of esophageal varices as well as the ‘‘red signs’’ on the mucosa overlying esophageal varices. In our case, the characteristics of the mucosa overlying the varices were normal, From ‘‘San Marco’’ Private Veterinary Clinic, Padova, Italy (Bertolini, De Lorenzi, Ledda); and ‘‘San Marco’’ Private Veterinary Laboratory, Padova, Italy (Caldin). Reprint requests: Giovanna Bertolini, DVM, ‘‘San Marco’’ Private Veterinary Clinic, via Sorio 114/C-35141 Padova, Italy; e-mail: bertolini@sanmarcovet.it. Submitted March 15, 2007; Revised May 4, 2007; Accepted June 22, 2007. Copyright E 2007 by the American College of Veterinary Internal Medicine 0891-6640/07/2106-0031/$3.00/0 Case Report