Techniques

Patient selection—Dogs with clinical signs of severe and progressive disease (severe exercise intolerance, cyanosis at rest, and increasing PCV over time), with echocardiographic or angiocardiographic confirmation of TOF, and without evidence of pulmonary artery hypoplasia or atresia were considered for this procedure (Figure 1). All dogs had been evaluated on multiple occasions by cardiologists who had determined that medical or conservative treatment was no longer beneficial. Owners gave full informed consent for the mBT shunt procedure to be carried out on their dogs.

View Full Size

Digital subtraction cardiac angiogram of a dog with TOF (dog 6). The catheter (C) has been passed down the cranial vena cava and across the septal defect; the tip is in the left ventricle (LV). The contrast agent has simultaneously outlined the aortic arch (Ao) and the pulmonary artery (PA). LSA = Left subclavian artery. BCT = Brachycephalic trunk.

Citation: Journal of the American Veterinary Medical Association 231, 5; 10.2460/javma.231.5.721

• Download Figure
• Download figure as PowerPoint slide

Anesthesia—The anesthetic protocol used in each institution varied according to drug availability, disposition of the dogs, and personal preference of the anesthesiologists involved. All dogs received either an opioid alone (meperidine, 3 mg/kg [1.36 mg/lb]) or an opioid combined with an anticholinergic and a benzodiazepine (hydromorphone, 0.2 mg/kg [0.09 mg/lb]; glycopyrrolate, 0.01 mg/kg [0.005 mg/lb]; and midazolam, 2.5 mg/kg [1.14 mg/lb]) via IM injection. Anesthesia was induced by use of a combination of fentanyl (5 μg/kg [2.27 μg/lb]), etomidate (2.5 mg/kg), lidocaine (20 mg/kg [9.1 mg/lb]), and midazolam (2.5 mg/kg) administered via IV injection. Anesthesia was maintained via continuous rate infusion of fentanyl (0.7 μg/kg/min [0.32 μg/lb/min]) alone or in combination with isoflurane in oxygen delivered via a cuffed endotracheal tube. During anesthesia, catheters were placed to permit central venous pressure and direct arterial pressure monitoring and quantification of urine production. Perioperative antibiosis was provided via IV administration of cefazolin^b (20 mg/kg, q 2 h) or cefuroxime^c (20 mg/kg, q 2 h) throughout the procedure.

mBT shunt technique—For each dog, a left lateral fourth intercostal or fifth rib resection thoracotomy was performed. Heparin (30 to 50 U/kg [13.6 to 22.7 U/lb]) was administered IV. The left subclavian artery was identified, and an approximately 4-cm length close to the aorta was dissected free from surrounding mediastinal tissues (Figure 2). The pericardium was opened to expose the lateral aspect of the left main pulmonary artery. A segment of 6-mm-diameter expanded PTFE vascular graft or previously harvested autologous carotid artery was cut to extend from the intended site of anastomosis on the subclavian artery to the left main pulmonary artery; the ends were cut at an angle so that the graft would lie flat once it was sutured in place. A Cooley or Codman vascular clamp was used to isolate a 2-cm length of the subclavian artery. A No. 11 Bard-Parker scalpel blade was used to make an initial incision in the left lateral aspect of the subclavian artery; this incision was extended to the desired length (measured against the cut end of the graft) by use of 60° Potts scissors. The graft was then sutured to the edges of the incision in the subclavian artery in an end-to-side manner with 5-0 or 6-0 polypropylene^d or PTFE^e suture in a simple continuous manner. The clamp was removed from the subclavian artery, and the graft was allowed to fill with blood. Then, either the clamp was replaced on the subclavian artery or a Hopkins bulldog clamp was placed on the graft itself to facilitate hemostasis. The pulmonary artery was opened between the jaws of a Cooley or Codman vascular clamp, and the graft was sutured to the pulmonary artery in the same manner as described for the left subclavian artery anastomosis. The hemostatic clamp on the graft or subclavian artery was released prior to tightening the suture on the pulmonary artery anastomosis to allow expulsion of any residual air. After the suture was tied, all remaining clamps were removed (Figure 3). The thoracic cavity was lavaged with warm saline (0.9% NaCl) solution, a thoracostomy tube was placed, and the thoracic wall closed in a routine manner. The shunt anastomoses in dogs 1 and 2 were performed by an experienced pediatric cardiac surgeon (JWG) assisted by an experienced veterinary surgeon (DJB); in dogs 3 and 4, anastomoses were performed by an experienced veterinary surgeon (DEH) under the supervision of an experienced cardiac surgeon (TT), and in dogs 5 and 6, anastomoses were performed by an experienced veterinary surgeon (DJB).

View Full Size

Photograph of the heart and great vessels of the dog in Figure 1 exposed through a left fifth intercostal thoracostomy. The left cranial lung lobe has been reflected caudally. The pulmonary artery trunk (PA) has been exposed via incision of the overlying pericardium. The LSA has been isolated and is encircled by a silicone vessel loop. Aorta = Ao. See Figure 1 for remainder of key.

Citation: Journal of the American Veterinary Medical Association 231, 5; 10.2460/javma.231.5.721

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photograph obtained during the mBT shunt procedure in the dog in Figure 1 after the graft has been sutured in place. Notice that the diameter of the graft is greater than the diameter of the LSA.

Citation: Journal of the American Veterinary Medical Association 231, 5; 10.2460/javma.231.5.721

• Download Figure
• Download figure as PowerPoint slide

Postoperative care—All dogs were monitored in an intensive care unit following surgery. Analgesia was provided via IM administration of morphine or hydromorphone (0.3 to 0.5 mg/kg [0.14 to 0.23 mg/lb], q 4 h); local analgesia was provided via administration of bupivacaine or ropivacaine (1.5 mg/kg [0.68 mg/lb]) administered into the pleural space via the thoracostomy tube. Supplemental oxygen and IV fluid therapy were administered on an as-needed basis. Thoracostomy tubes were removed as soon as no fluid was obtained when aspirated. In addition, acetylsalicylic acid was administered PO (5 to 10 mg/kg [2.27 to 4.5 mg/lb]) every other day for 3 months after surgery.

All animals that survived (n = 5) were reevaluated at either institution approximately 1 month after surgery and underwent full cardiac investigation including a CBC, ECG, and echocardiography. Dogs were reevaluated at the referral institution on an ad hoc basis thereafter.

Data collection—Data regarding the immediate preoperative period collected from the medical records of these dogs included breed, age, sex, body weight, PCV, serum total solids concentration, and PaO₂ when the dog was conscious and breathing room air (when available). At the 1-month reevaluation, PCV and serum total solids concentration were recorded. The last recorded PCV and serum total solids concentration, the time at which those variables were evaluated in relation to the procedure, and the last echocardiographic evidence of shunt patency were recorded. Additional follow-up information (regardless of whether the dog was alive or dead) was determined via telephone conversations with the owners and their veterinarian.

Results

Six dogs underwent the mBT shunt procedure (Table 1). Four dogs were evaluated and treated at the Matthew Ryan Veterinary Hospital of the University of Pennsylvania (dogs 1 through 4); 2 dogs were evaluated and treated at the Queen Mother Hospital of The Royal Veterinary College, London (dogs 5 and 6). The smallest dog (dog 4) weighed 2.9 kg (6.38 lb); all other dogs weighed > 10 kg (22 lb). In dog 4, a carotid artery autograft was used to create the shunt, whereas the remaining 5 dogs received a synthetic (expanded PTFE) graft (6 mm in diameter). All the dogs survived surgery. Dog 4 developed anemia during the first 12 hours after surgery and required a whole blood transfusion; it also developed respiratory failure 24 hours after surgery and required mechanical ventilation for a period of 36 hours. That dog was weaned from ventilator support but shortly after weaning underwent a cardiac arrest and died. Postmortem examination was not permitted. Four of the 5 surviving dogs had a sustained decrease in PCV, compared with the presurgical value, and all of the surviving dogs had PCV values that remained stable over time. In all the dogs for which data was available, there was an increase in PaO₂ during the immediate postoperative period, compared with the PaO₂ prior to surgery. Two of the dogs that survived to discharge did eventually die. According to the local veterinarians' records, dog 1 died at the age of 11.5 years (6 years after the mBT shunt procedure) as a result of congestive heart failure, but no further details are available; dog 2 died suddenly at home at the age of 7 years (6 years after the mBT shunt procedure), but similarly no further medical details were available. All other dogs were alive at the last known follow-up (ie, 30 to 54 months after surgery).

Discussion

To the authors' knowledge, the 6 dogs with TOF-associated clinical signs in the present report represent the largest group to have undergone surgical palliation described in the veterinary medical literature. Five of the 6 dogs reported here survived the surgery and had an uneventful postoperative recovery. Two of these dogs subsequently died, but both survived for 6 years after the mBT shunt procedure. Three other dogs were alive at 36, 48, and 18 months after the procedure. The dog that died shortly after surgery in the present study differed from the other dogs in 2 important aspects. First, it was the smallest dog (weight, 2.9 kg; all other dogs weighed > 10 kg), and second, it had a section of native carotid artery used to create the shunt rather than a synthetic material. The decision to use carotid artery rather than 3- or 4-mm-diameter expanded PTFE tubing was made on the basis of experience in human neonates, in which inadequate palliation following use of 4-mm synthetic shunts was evident in 23 of 34 (68%) of patients because of partial or complete shunt occlusion 2 years after surgery.¹³ In the dog that died, size of the vessel and physical size of the dog made the surgical procedure more challenging. The procedure appeared to go well, and the dog's recovery was proceeding well initially. However, the dog died as a result of respiratory failure and subsequent cardiac arrest; because a necropsy was not permitted, the causes of those complications were not established, nor was patency of the shunt evaluated after surgery or death. The risk of surgery-related complications, such as suture occlusion or thrombus formation at the anastomotic site, might be higher in physically smaller dogs, and such complications could, in part, explain the postoperative events in that particular dog in the present study. Unfortunately, comments regarding the cause of death in that dog are speculative, and no firm conclusions or recommendations can be made on the basis of the experience with that patient. The poor outcome in that small-sized dog concurs with the subjective assessment of the procedure in small dogs (< 10 kg) by Eyster.²² However, the fact that successful palliation of TOF via microvascular anastomosis of the internal thoracic artery to the pulmonary artery has been reported in a cat²⁴ suggests that a technique that is different from that described in this report might be more appropriate for small dogs. The only data available for the 2 dogs that died 6 years after surgery was the owners' assessment of quality of life just before death. Cause of death would be important in evaluating the procedure reported here but unfortunately was not available. Even if both dogs had shunt-related complications (occlusion or excessive flow) that caused their death, their treatment could still be considered successful because they both lived 6 years after the procedure. In addition, both these dogs enjoyed a better quality of life after surgery, according to their owners. When compared with other palliative procedures reported in the veterinary literature, dogs that have undergone an mBT shunt procedure for palliation of TOF have appeared to have improved shortand long-term survival.

In the present study, objective data collected to assess the shunt success were limited to arterial PaO₂ before and immediately after surgery, echocardiographic evidence of shunt patency, and serial PCV evaluations. Comparison of immediate preoperative PaO₂ with immediate postoperative values revealed only a modest increase in PaO₂ following shunt procedure in 4 of the treated dogs. Preoperative values were derived from blood samples collected from conscious dogs, and factors such as the stress associated with arterial puncture could have influenced these results. In addition, postoperative samples were collected the day after surgery while the dogs were in an intensive care unit and could have been influenced by postoperative thoracic surgical factors such as pain, lung atelectasis, and analgesic drug treatment. All dogs in which this comparison could be made did, however, have improvement in PaO₂ as a result of the surgical procedure. In all of the dogs evaluated echocardiographically after surgery, turbulent blood flow in the pulmonary artery at the site of entry of the shunt was evident, and frequently the shunt itself could be identified. Such evaluations were done intermittently for each dog, but from the last known record of shunt patency for those dogs in which it was assessed, the interval of patency ranged from 8 to 15 months after surgery.

Indirect evidence of shunt patency can be derived from PCV evaluations. Hypoxia is a potent stimulus for hematopoiesis, and secondary polycythemia is a classic feature of untreated TOF. All the dogs in the study reported here had either high absolute PCV or high PCV relative to their age prior to surgery. All the dogs had a similar or lower PCV value 1 month after surgery, compared with the value before surgery. The PCV values were monitored long term on a more random basis, but all appear to have remained stable at periods from 8 to 36 months after surgery. The PCV in all dogs remained stable over the period in which it was measured. Clearly, PCV can be influenced by a myriad of factors such as dehydration and many disease states. In dogs that were otherwise in good health, a stable PCV could be considered good evidence of a functional shunt and may be more reliable than PaO₂ assessed in awake dogs, which is a measurement that could be easily influenced by stress during sample collection. Although the objective data collection was not standardized, there is reasonable evidence to suggest that the shunts in the dogs of this report remained patent and functioned well for an extended period of time after placement.

The mBT shunt, as reported here, has several advantages over shunting operations that have been previously reported in dogs. First, none of the dogs that survived surgery in the present study developed congestive heart failure as a short-term complication of the procedure. This fact suggests that use of the left subclavian orifice to regulate blood flow protected these dogs against excessive pulmonary overcirculation. In contrast, congestive heart failure was reported⁷ in 1 of 4 dogs that underwent a Potts aorticopulmonary anastomosis. Second, early occlusion of the shunt was not evident in any of the dogs that survived to discharge from the hospital, and evidence of shunt patency lasting years was available for several dogs. In contrast, thrombotic complications associated with the classic BT or Potts procedures have been reported.^7,21,23 In the mBT shunt procedure described in this report, a shunt of greater diameter than the native left subclavian artery was used, and this probably helped to reduce the risk of thrombotic complications. In addition, the shunt material (expanded PTFE) and the antiplatelet treatment (aspirin) perhaps minimized the risk of thrombosis as a cause of shunt failure. Suture material and suturing technique can also influence the formation of a thrombus at an anastomosis site. In the first 2 procedures, the shunt anastomoses were performed by an experienced cardiac surgeon, but subsequent surgeries were performed by trained veterinary surgeons either alone or with an experienced assistant. As with any surgical technique, results improve as experience with a procedure is gained. The authors conclude that this surgery is within the capability of any trained surgeon with vascular surgical experience. The third advantage of this procedure is that, should complications develop at the left subclavian anastomosis site, the vessel can safely be ligated.^22,23 In contrast, if technical difficulties such as tearing of the aorta occur during the Potts or any of the central aorta–to–pulmonary artery anastomoses, massive hemorrhage can rapidly become life-threatening. Of course, if similar complications develop at the pulmonary artery anastomoses site, hemorrhage can be just as serious a problem. Because these procedures were performed on either adult- or nearly adult-sized dogs, distortion of the pulmonary artery at the site of shunt anastomosis secondary to growth of the animal would seem less likely, although this was not studied specifically; however, there was no echocardiographic evidence that this had developed in any of the dogs. As reported in humans,^13,14 the mBT shunt procedure appears to be a superior procedure for palliation of the clinical signs of TOF in dogs.

Complete repair of TOF would be necessary to give an affected dog what would be considered a normal quality of life. Herrtage et al¹⁷ reported successful surgical correction of a dog with TOF but that dog died suddenly after 3 years of an apparently normal life. In another report,¹⁸ complete repair of TOF in an acyanotic dog was successful, and the dog was free of clinical signs for at least 3 years. Orton et al⁸ performed surgical correction of TOF in 2 dogs that remained free of clinical signs 2 and 3 years after surgery. It is likely that, as in humans, successful surgical repair will provide better quality of life and improved long-term results than shunt placement in TOF-affected dogs. At present, open heart repair is only available at a limited number of veterinary medical centers around the world, and generally the financial cost to the owner and the relative risk to the patient are both high. Whether open heart repair of TOF in dogs actually does provide a better quality of life and increase life expectancy, compared with other treatments, has not been rigorously investigated. Indeed, because of the small number of dogs that receive surgical treatment for this condition worldwide, it may never be possible to determine that one treatment is superior to another by direct comparison. Extrapolating from data available on treatment of TOF in humans, it is logical for veterinary surgeons to continue to pursue corrective surgery as a goal for affected dogs. Palliative surgery involving the mBT shunt procedure may, however, represent a relatively lower-risk and lower-cost procedure that will improve both quality of life and life expectancy for dogs affected with TOF.