Introduction

Congenital portosystemic shunts are anomalous vascular connections between the portal and systemic venous systems that lead to hepatic dysfunction secondary to diversion of blood away from the liver. Portosystemic shunts are classified as intrahepatic or extrahepatic, depending on whether the shunting vessel is located within or outside of the liver parenchyma. Extrahepatic shunts are most common in dogs, with small-breed dogs being overrepresented, whereas intrahepatic shunts occur most often in medium- and large-breed dogs.^1–3 Congenital intrahepatic portosystemic shunts (IHPSSs) occur infrequently in small- and toy-breed dogs.^1–4

Attenuation is the preferred treatment for IHPSSs because medical management has been associated with poor survival times owing to progressive liver dysfunction and associated clinical signs.^5,6 Most commonly, only partial attenuation of IHPSSs is possible because of clinically unacceptable portal hypertension following complete attenuation.^7,8 Reported surgical techniques for partial attenuation of IHPSS include suture ligation, cellophane banding, and placement of ameroid constrictors or hydraulic occluders.^9–13 However, open surgical techniques have historically been associated with high perioperative complication and mortality rates, with reported rates up to 77% and 28%, respectively.^11,14,15

More recently, percutaneous transvenous coil embolization (PTCE) has been evaluated for the treatment of IHPSSs and has been shown to have perioperative complication and mortality rates similar to or better than rates reported for open surgical techniques.^16–18 Although small- and toy-breed dogs are rarely presented with IHPSS, these cases may be more difficult to manage surgically owing to their small size and may therefore experience different outcomes than previously reported for predominantly large-breed dogs. Conversely, PTCE may offer a safe, minimally invasive alternative for treatment of these unusual cases.

To the authors’ knowledge, no studies have evaluated the outcome of PTCE specifically for treatment of IHPSSs in small- and toy-breed dogs. However, a small case series of 4 cats undergoing PTCE for treatment of IHPSSs reported no major procedural complications and good outcomes, with all cats experiencing resolution of their clinical signs.¹⁹

The objective of the case series reported here was to evaluate the outcome of small- and toy-breed dogs with IHPSSs that were treated with PTCE. We hypothesized that PTCE would be a safe and effective treatment option for these dogs, with complication and mortality rates similar to rates reported for large-breed dogs.

Materials and Methods

Percutaneous transvenous coil embolization

Percutaneous transvenous coil embolization was generally performed as previously described.^16–18 A stab incision was made over the right external jugular vein, and an 18- or 22-gauge over-the-needle catheter was placed in the jugular vein. When a 22-gauge over-the-needle catheter was used for access to the jugular vein, a 0.018-inch guide wire was introduced into the catheter, and a transition dilator system was placed (Mini-Access Kit; Merit Medical). The guidewire and inner cannula were removed, and a 0.035-inch hydrophilic guidewire (Weasel Wire; Infiniti Medical) was introduced into the jugular vein. A 5F dilator was used to dilate the tract into the jugular vein, and an 11F vascular access sheath (Super Sheath; Boston Scientific) with dilator was introduced into the jugular vein over the guidewire and secured to the skin with nylon suture. A 9F introducer sheath was used in 5 dogs weighing 2.6 to 5.4 kg. In the 2 smallest dogs (1.5 and 2.8 kg), a 6F introducer sheath was used.

A 4F hooked angiographic catheter (Cobra; Infinity Medical) was advanced into the introducer sheath and used to select the IHPSS, and a 5F marker angiographic catheter (Marker-Straight; Infiniti Medical) was placed next to the 4F catheter in the caudal vena cava, beyond the shunt ostium. Both catheters were connected to a transducer, and resting portal vein and vena caval pressures were obtained. Combined digital subtraction portocavography was then performed with iodinated contrast medium (Omnipaque; 300 mg of L/mL) diluted 1:1 in sterile saline (0.9% NaCl) solution. Contrast medium was administered at a dose of < 2 mL/kg while performing a manual breath-hold (Figure 1). For patients in which a 6F introducer sheath was used, only a 4F hooked angiographic catheter was placed through the sheath, and separate portograms and cavograms were obtained. Measurements obtained from previous CT images for stent selection and placement were confirmed.

View Full Size

Transverse (A) and dorsal (B) portal-phase, postcontrast CT images of a 6-month-old Miniature Australian Shepherd with a right-divisional intrahepatic portosystemic shunt (IHPSS; asterisks), digital subtraction angiograms showing the right divisional IHPSS (asterisks) prior to percutaneous transvenous coil embolization (C) and following placement of a stent in the caudal vena cava and coils in the IHPSS (D), and a ventrodorsal radiographic image following placement of a 20 X 80-mm vena caval stent and 7 thrombogenic coils.

Citation: Journal of the American Veterinary Medical Association 260, 12; 10.2460/javma.22.03.0143

• Download Figure
• Download figure as PowerPoint slide

A straight-tip exchange guidewire (Fixed Core Wire Guide; Cook Medical; Bentson Wire; Infiniti Medical) was placed through the 5F marker catheter, and both the marker and hook catheters were removed. A self-expanding nitinol stent (Vet Stent-Cava; Infiniti Medical) of predetermined length and diameter, based on measurements from previous CT images, was advanced over the guidewire and deployed in the caudal vena cava across the opening of the shunt ostium (Figure 1). For the dogs in which a 9F introducer sheath was used, a urethral stent (Vet Stent-Urethra; Infiniti Medical) was placed instead of a vena caval stent owing to patient size. For dogs that required a 6F introducer sheath, an open-cell, self-expanding vascular stent (Zilver 635) was used. Following stent placement, the guidewire was advanced through the stent interstices and into the portal vein via the IHPSS, and a hook catheter was placed over the wire into the portal vein. The catheter was attached to a transducer for monitoring of portal vein pressures during coil placement.

A second hook catheter was placed in the IHPSS over a guidewire. A PTFE guidewire (Benston; Cook Medical or Infiniti Medical) was passed through the catheter for delivery of pushable thrombogenic coils (MReye embolization coils; Cook Medical or Infiniti Medical) into the IHPSS (Figure 1). In the 2 dogs in which a 6F introducer sheath was used, a 4F angled catheter (Berenstein; Infiniti Medical) was introduced into the portal vein via the stent interstices. A microcatheter (Renegade STC-18; Boston Scientific) over a microwire (Glidewire GT; Terumo Medical Corp) was placed through the angled catheter, and the microwire was used to push tornado-configured microcoils (VortX; Boston Scientific) into the shunt. After delivery of each coil, portal vein pressure was measured. After delivery of coils was complete, the catheters were removed over a guidewire, and a 7F triple-lumen catheter (Arrow; Teleflex) was placed in the jugular vein over the guidewire (University of California-Davis) or the catheters were removed, the entry site was closed, and strong digital pressure was maintained over the right jugular vein for 20 to 30 minutes to achieve hemostasis (University of Florida).

Results

Twenty dogs that underwent PTCE for treatment of an IHPSS met the inclusion criteria. Of these, 6 were treated at the University of Florida and 14 were treated at the University of California-Davis. Seven dogs were of mixed breeding; breeds of the remaining dogs were French Bulldog (n = 5), Miniature Australian Shepherd (2), Corgi (2), Chihuahua (1), Cavalier King Charles Spaniel (1), Shiba Inu (1), and Toy Poodle (1). Five dogs were spayed females, 4 were sexually intact females, 5 were castrated males, and 6 were sexually intact males. Body weight ranged from 1.5 to 10.0 kg (mean ± SD, 6.32 ± 2.57 kg). Median age at the time of PTCE was 8.1 months (IQR, 6.9 to 14.4 months).

Clinical signs reported included lethargy (n = 9), vomiting (8), anorexia or hyporexia (8), ptyalism (4), head pressing (3), ataxia (3), twitching or tremoring (2), stargazing (2), abdominal distension (2), diarrhea (2), regurgitation (1), weight loss (1), pawing the ground (1), chewing unusual objects (1), anxiety (1), hematochezia (1), stranguria (1), hematuria (1), hiding (1), circling (1), and restlessness (1). Two dogs had no clinical signs, and the shunt was discovered because of abnormally high hepatic enzyme activities on routine bloodwork. All dogs were receiving medical management for their shunt at the time of the procedure; medications consisted of levetiracetam (12.5 to 33.3 mg/kg, PO, q 8 h; n = 20), lactulose (0.2 to 0.8 mL/kg, PO, q 8 to 12 h; 6), metronidazole (5 mg/kg, PO, q 12 h; 3), omeprazole (0.8 to 1 mg/kg, PO, q 12 h; 3), famotidine (0.5 mg/kg, PO, q 12 h; 2), amoxicillin–clavulanic acid (12.9 to 15.8 mg/kg, PO, q 12 h; 2), lantoprazole (0.5 mg/kg, PO, q 12 h; 1), and neomycin (21.2 mg/kg, PO, q 12 h; 1). All dogs were being fed a protein-restricted diet.

All dogs underwent both CT angiography and fluoroscopy for evaluation of their shunt. Nine dogs also underwent abdominal ultrasonography prior to CT and PTCE. In all dogs, a single IHPSS was identified. The IHPSS was classified as right divisional in 12 (60%) dogs, left divisional in 6 (30%), and central divisional in 2 (10%). One dog had concurrent intrahepatic portal hypoplasia. Nephrolithiasis was diagnosed in 9 (45%) dogs, mineral sediment in the urinary bladder was identified in 7 (35%) dogs, and cystoliths were diagnosed in 1 (5%) dog. Eight (40%) dogs had no uroliths or mineral debris on imaging. Pertinent preoperative clinicopathologic values are summarized (Table 1).

Mean baseline portal vein pressure was 8 mm Hg (range, 1 to 16 mm Hg; SD, 3.51 mm Hg), mean baseline vena caval pressure was 7 mm Hg (range, 0 to 15 mm Hg; SD, 3.46 mm Hg), mean post-PTCE portal vein pressure was 12 mm Hg (range, 3 to 19 mm Hg; SD, 4.06 mm Hg), and mean post-PTCE vena caval pressure was 7 mm Hg (range, 1 to 15 mm Hg; SD, 3.6 mm Hg). Mean initial gradient between portal vein and vena caval pressures was 1 mm Hg (range, 1 to 4 mm Hg; SD, 1.2 mm Hg), and mean post-PTCE gradient was 6 mm Hg (range, 0 to 11 mm Hg; SD, 2.7 mm Hg). The most commonly used stent size was 18 X 80 mm (n = 7), followed by 14 X 60 mm (4), 16 X 80 mm (3), 12 X 60 mm (2), 20 X 80 mm (1), 22 X 80 mm (1), 12 X 40 mm (1), and 10 X 20 mm (1). Coil diameter ranged from 3 to 8 mm, and a mean of 6 coils were placed (range, 1 to 14 coils; SD, 3.3 coils). Mean procedure time was 102 minutes (range, 68 to 145 minutes; SD, 22.2 minutes). Concurrent surgical procedures were performed in only 1 dog; this was also the dog with the longest procedure time (145 minutes). A small, ventral celiotomy was performed for catheterization of a mesenteric vein to measure portal vein pressures in this dog.

Intraoperative complications occurred in 4 of the 20 (20%) dogs; > 1 complication occurred in all 4 dogs, for a total of 9 intraoperative complications. All complications were either grade 1 (5/9) or 2 (4/9) and consisted of hypotension (n = 2), hypothermia (2), bradycardia (2), ventricular premature contractions (2), and hypercapnia (1). A grade 1 postoperative complication occurred in 1 dog, which regurgitated at the time of extubation. Median hospitalization time was 3 days (IQR, 2 to 3 days).

Follow-up time for all dogs ranged from 36 to 1,705 days (median, 413.5 days; IQR, 224 to 804 days). Clinical signs resolved in 19 of the 20 (95%) dogs, and median time to resolution of clinical signs was 21 days (IQR, 0 to 36 days). No dogs underwent additional shunt attenuation procedures. At the time of the last available follow-up, all dogs were receiving some form of medical management. Medications being administered at this time consisted of omeprazole (n = 15), famotidine (3), lactulose (1), lantoprazole (1), and neomycin (1). Three dogs were still being fed a protein-restricted diet.

Median time to the last available follow-up clinicopathologic test results was 205.5 days (IQR, 149.5 to 384.5 days). Results of follow-up clinicopathologic testing were not available for 4 dogs. Median BUN concentration at the time of the last available follow-up testing (6.5 mg/dL; IQR, 4 to 10 mg/dL) was significantly (P = .002) higher than median preoperative BUN concentration (4 mg/dL; IQR, 3 to 5 mg/dL). However, median albumin concentration at the time of the last available follow-up testing (2.6 g/dL; IQR, 2.2 to 3.1 g/dL) was not significantly (P = .24) different from median preoperative albumin concentration (2.5 g/dL; IQR, 2.1 to 2.9 g/dL). Follow-up protein C concentrations were available for only 2 dogs and were 48% (preoperative protein C, 34%) and 86% (preoperative protein C, 33%).

At the time of data collection, 3 of the 20 (15%) dogs were dead, with 2 of these dogs dying of causes related to the IHPSS. One- and 2-year survival rates were 92% (Figure 2). Longer survival rates could be calculated owing to a lack of follow-up information beyond 2 years for most patients. A postmortem examination was not performed on any of the dogs that died. One dog was euthanized 1,090 days after undergoing PTCE because of a duodenal mass suspected to be neoplastic that was discovered on abdominal ultrasonography and was causing mechanical obstruction. Another dog was euthanized 267 days after undergoing PTCE because of a suspected gastrointestinal ulcer and secondary septic shock following ovariohysterectomy performed by the referring veterinarian and oral administration of NSAIDs. The third dog was euthanized 1,178 days after undergoing PTCE by the referring veterinarian because of a suspected blood clot at the region of the caval stent and secondary infection.

View Full Size

Product-limit survival curve for 20 small- and toy-breed dogs with a congenital IHPSS treated with PTCE.

Citation: Journal of the American Veterinary Medical Association 260, 12; 10.2460/javma.22.03.0143

• Download Figure
• Download figure as PowerPoint slide

Discussion

In the present study, PTCE resulted in favorable outcomes for small- and toy-breed dogs with an IHPSS. Clinical signs resolved in all but 1 dog by 36 days after the procedure. At the time of the final follow-up, most dogs (80%) were only receiving medical management in the form of an antacid, which has been recommended for life-long treatment of dogs with IHPSS owing to the high incidence of gastrointestinal hemorrhage or ulceration.¹⁶ One dog was receiving neomycin and lactulose and was receiving a protein-restricted diet; this was the same dog in which clinical signs did not resolve. Three dogs were being fed a protein-restricted diet in addition to receiving an antacid, despite recommendations that the diet could be discontinued.

One of the perceived difficulties when performing PTCE in small-breed dogs was the size of the equipment and implants used. Percutaneous transvenous coil embolization requires a vascular access site that is large enough to facilitate placement of stents and coils of sufficient size and diameter. Previous reports^16–18 of PTCE in large-breed dogs typically describe the use of a 10F to 12F vascular introducer sheath to accommodate the vena caval stent delivery system. In 5 of the cases included in the present study, urethral stents were used instead of vena caval stents to account for the difference in diameter of the vena cava. In the 2 dogs in which a 6F introducer was used, open-cell vascular stents were placed instead because they can be delivered through a smaller vascular access site and are more malleable than closed-cell stents. Furthermore, smaller patients are likely to have smaller-diameter shunts. Thrombogenic coils used for PTCE of IHPSSs are typically 8 mm in diameter,¹⁶ which may be too large for small- and toy-breed dogs or cause dramatic increases in portal vein pressure with placement. In a previous report by Culp et al,¹⁸ a single 0.035-inch, 8-mm-diameter, 5-cm-long coil was placed in 2 dogs weighing < 10 kg (Miniature Poodle, 1.6 kg; Miniature Schnauzer, 6.7 kg) owing to a dramatic change in portal vein pressure with coil placement. Case et al¹⁷ reported a similar finding, with a Miniature Pinscher developing peracute portal hypertension after placement of a single 0.035-inch, 8-mm-diameter, 5-cm-long coil because of complete occlusion of the shunt. The coil was extracted, which resolved the portal hypertension, and smaller 0.035-inch, 3 X 4-mm tornado coils were used instead without complications. In a report¹⁹ of PTCE in cats with IHPSSs, smaller coils (0.018 and 0.035 inches; 4 to 5 mm) were used to account for the smaller vascular anatomy of the patients and were well tolerated in all 4 cats. In the present study, 0.035-inch, 8-mm-diameter, 5-cm-long coils were placed in 9 dogs with no complications. Portal vein pressure was measured after each coil was placed to ensure there was no dramatic change in pressure that would lead to portal hypertension, and as many as fourteen 8-mm coils (range, 2 to 14 coils) were placed in some patients. For some dogs, tornado-configured coils or a microcatheter system for placement of microcoils was used. The tornado-configured coils have a wide base and narrow tip, which gives them a narrow profile and may be ideal in cases in which a smaller landing zone is available. Additionally, the microcoils and smaller vascular stents can be deployed via a smaller introducer, which may be beneficial in very small dogs. The dogs that benefited from these devices tended to be the smallest of the dogs included in this report.

Only minor (grade 1 and 2) intraoperative and postoperative complications were reported, at rates (20% and 5%, respectively) similar to or better than those reported previously in studies evaluating PTCE and open surgical techniques for treatment of IHPSSs.^{11,12,14,16,17,22} Minor intraoperative complications that were experienced consisted of hypotension, hypothermia, bradycardia, ventricular premature contractions, and hypercapnia. One dog experienced a minor postoperative complication (regurgitation on extubation). This dog did not develop any further issues (eg, aspiration pneumonia, esophageal stricture, or esophagitis) and did not require any additional management. It is difficult to ascertain whether these complications were a direct result of the PTCE procedure, as they were all complications commonly experienced with general anesthesia.

No major (ie, > grade 2) intraoperative or postoperative complications occurred in this cohort of dogs; however, this may have been due to the small number of dogs in the present study. Previously reported major intraoperative and postoperative complication rates associated with PTCE of IHPSSs in dogs range from 3% to 8% and from 8% to 13%, respectively.^16,17 Reported major intraoperative complications include substantial portal hypertension, severe acute gastrointestinal hemorrhage, and coil migration; major postoperative complications described include seizures and hepatoencephalopathy, respiratory arrest secondary to suspected disseminated intravascular coagulopathy, cardiac arrest, hemorrhage from the jugular access site requiring transfusion, pneumonia, suspected portal hypertension, and acute death of unknown cause.^16–18 Owners should be appropriately counseled regarding the possibility of these complications occurring.

Median hospitalization time following the PTCE procedure in the presents study was 3 days (IQR, 2 to 3 days), which is likely due to surgeon preference for a 3-day hospitalization period at the University of California-Davis, where most of the cases were treated. In contrast, the other surgeon in this report generally discharged patients after 24 to 48 hours. This was similar to previous reports of PTCE, in which dogs undergoing endovascular procedures had a significantly shorter median hospitalization time, compared with those undergoing cellophane banding (1 vs 3 days, respectively).¹⁷

One- and 2-year survival rates were both 92% in the present study, which were similar to previously reported survival rates following PTCE (87% and 80%, respectively).¹⁷ No perioperative or early postoperative deaths occurred in this cohort, compared with rates of up to 28% following open surgical techniques. The mortality rate related to the IHPSS in this cohort was 10% (2/20). Of the 2 dogs that died of causes related to the IHPSS, 1 was euthanized owing to a suspected clot that developed at the region of the caval stent 1,178 days after undergoing PTCE. The second dog was euthanized 267 days after undergoing PTCE because of suspected gastrointestinal ulceration and secondary septic shock. This was the same dog that did not experience a resolution of clinical signs. It is possible that this dog had persistent shunting through the IHPSS or developed additional IHPSSs postoperatively and may have benefited from an additional PTCE procedure. In a study by Weisse et al,¹⁶ 16% of dogs undergoing PTCE of an IHPSS underwent additional PTCE for placement of more coils because of a recurrence of clinical signs. Currently, there are no standard recommendations regarding when to perform additional PTCE procedures, and this decision is largely based on severity of clinical signs, quality of life, and clinicopathologic and imaging results as well as owner and clinician discretion. Further evaluation of the clinical benefit of additional PTCE procedures as well as specific indications for doing so are warranted.

The second dog that was euthanized was no longer receiving antacid medications and underwent ovariohysterectomy. Postoperatively, the dog was given meloxicam for analgesia. Two days later, the dog was presented to an emergency clinic because of profuse melena and collapse. Results of clinicopathologic testing were consistent with anemia secondary to gastrointestinal hemorrhage, acute kidney injury, and liver failure. Gastroduodenal ulceration is a known adverse effect of treatment with NSAIDs, and administration of NSAIDs has been shown to increase the risk for gastrointestinal ulceration or erosion by a factor of 6.3.²³ Owing to the risk of gastrointestinal hemorrhage and ulceration and the possibility of persistent liver insufficiency in dogs with IHPSS, the authors believe NSAIDs should not be used in these patients even after PTCE.¹⁶

Distribution of single IHPSS morphology in this group was similar to what has been previously reported, with right- and left-divisional shunts occurring more frequently than central-divisional shunts.^{16,17,22,24,25} However, no cases of multiple IHPSSs were identified in our cohort. The incidence of multiple IHPSSs has been reported to be as high as 10%.¹⁶ In a previous study,¹⁶ the authors theorized that the true incidence of multiple IHPSSs may be underestimated because of their subtle nature and the possibility that they can go undetected on imaging. Additionally, use of direct angiography in the same study led to the detection of more shunts than were seen with cross-sectional imaging, indicating that direct angiography may be a more sensitive imaging modality for the diagnosis of multiple IHPSSs. It is possible that cases of multiple IHPSSs were missed in the present study; however, all dogs underwent both CT angiography and direct angiography, making this less likely.

After mixed-breed dogs, French Bulldogs were the most common dogs in this cohort, accounting for 25% of the cases included. To the authors’ knowledge, no previous predisposition or hereditary component for IHPSSs has been identified in French Bulldogs. Owing to the limited sample size of this study, it is impossible to determine whether this breed is truly predisposed to developing IHPSS. It is possible that they were overrepresented in this cohort secondary to the geographic locations of the referral institutions included, breeding pools in those regions, or other unforeseen confounding factors. A larger, more representative sample is necessary to establish a true prevalence.

Limitations of the present study were largely inherent to its retrospective nature and included a lack of standardization of preoperative diagnostic testing, medical management, and postoperative follow-up; a lack of long-term follow-up information for recent cases; a lack of postmortem examinations; and a lack of complete data. Additionally, the small sample size of our cohort limited our ability to draw conclusions, and differences in definitions of complications between studies make comparisons difficult. To the authors’ knowledge, this represents the largest report of small- and toy-breed dogs with IHPSSs. For many patients, standard PTCE techniques were used without complications. In 2 of the smallest patients, use of microcatheters, microcoils, and tornado-configured coils facilitated the procedure. Our results indicate that PTCE offers an effective and safe treatment option for these cases, with similar to better complication and mortality rates, compared with rates for open surgical techniques.