Outcome of four cats with congenital intrahepatic portosystemic shunts treated by a modified percutaneous transvenous coil embolization technique

Abstract

Case series summary

Simultaneous portogram/cavogram and continuous portal pressure monitoring during coil deployment are standard practices during percutaneous transvenous coil embolization (PTCE) for the interventional treatment of congenital intrahepatic portosystemic shunts (IHPSSs). In cats, vascular access is more challenging because of their smaller size. The minimum vascular sheath size required to accommodate standard instrumentation for dual-catheter PTCE is 9 Fr, which can be challenging to place in the feline jugular vein. The objective of this case series was to retrospectively describe a single-catheter modification of the PTCE technique in cats to avoid complications associated with large vascular sheaths and the outcomes of these cats. Four cats from two different institutions underwent single-catheter PTCE for left divisional IHPSSs between January 2024 and May 2025. Data collected from medical records included patient demographic data and peri- and post-procedural data. Single-catheter PTCE was performed in all cats through a 6 Fr vascular sheath percutaneously placed in the jugular vein. Isolated portogram and cavogram were performed to localize the shunt, measure caval diameter and determine caval stent positioning using a 4 Fr Berenstein catheter. A laser-cut stent was selected based on CT and cavogram measurements and positioned to span the shunt ostium. The same Berenstein catheter was used for portal pressure measurement and coil deployment by alternating between the shunt and portal vein locations. No complications associated with venous trauma or portal hypertension were reported.

Relevance and novel information

The case series results suggest the modified PTCE technique is a viable alternative that was associated with reasonable procedural duration and avoided vascular complications that may be associated with large vascular sheaths in cats.

Introduction

Congenital portosystemic shunts in cats are uncommon, with intrahepatic portosystemic shunts (IHPSSs) diagnosed rarely.^{1 –7} IHPSSs based on the anatomical locations are divided into left, right and central divisional shunts. Surgical attenuation of IHPSSs is generally considered more challenging compared with extrahepatic portosystemic shunts (EHPSSs)^8,9 because of challenges inherent to anatomical differences. Previously described surgical treatment options include surgical dissection and suture ligation,^10,11 occlusion with an ameroid constrictor, ⁸ minimally invasive percutaneous transvenous coil embolization (PTCE) with or without stent placement,^12,13 and a hybrid combined open and interventional approach. ¹⁴

In dogs, PTCE has become a routine minimally invasive strategy for treating IHPSSs,^13,15 although the same approach is challenging in cats because of their smaller size relative to the required catheters. Previously reported mean vein diameters in healthy cats are 2.55 ± 0.38 mm ¹⁶ for the femoral veins, 3.74 ± 0.73 mm for the right and 3.76 ± 0.70 mm for the left external jugular veins, respectively, ¹⁷ which may limit vascular sheath options. Simultaneous portogram/cavogram and continuous portal pressure (PP) monitoring during coil deployment are standard PTCE techniques. The minimum vascular sheath size required to accommodate standard instrumentation with two catheters for PTCE is 9 Fr, which can be challenging to place in a feline jugular vein.

In the first case of this series, standard vascular access using a modified Seldinger technique and placement of a 9 Fr vascular sheath was performed in the right jugular vein without complication. However, this large sheath likely occluded venous flow into the cranial vena cava, resulting in vessel collapse after catheter and wire retraction and inability to replace the wires or catheters despite multiple attempts. A venogram performed through the right jugular vascular sheath showed extravasation of contrast into the mediastinum, raising concern for vascular trauma. The procedure was aborted and the right jugular vein ligated. The challenges associated with a large vascular sheath in this first case prompted the need for consideration for procedural alternatives that would allow for use of a smaller vascular sheath, such as a 6 Fr vascular sheath, and single catheter for venous contrast studies, pressure measurements and coil placement.

The objective of this case series was to retrospectively describe a modification of the PTCE technique in cats to avoid complications associated with large vascular sheaths and to report the clinical outcomes.

Case series description

This was a multicenter retrospective descriptive case series. Medical records were collected and reviewed from one university teaching hospital and one private practice between January 2024 and May 2025. Cats were included if they had a diagnosis of IHPSS and subsequently underwent PTCE with a single-catheter technique. Diagnosis of IHPSS was based on a combination of blood work and CT angiography (CTA). Data recorded included signalment (age, sex, breed, body weight), history, presenting complaint (clinical signs) and diagnostic results (biochemical and clinicopathologic findings, CTA). Medical treatments before and after PTCE, such as use of antibiotics, proton-pump inhibitor, lactulose, anticonvulsant and low-protein diet, were also recorded.

PTCE procedure details and outcome were recorded into a spreadsheet (Excel version 1906; Microsoft).

Patient characteristics

Four cats met the inclusion criteria. Median age at the time of the presentation for PTCE was 12 months (range 8–30) and median body weight was 3 kg (range 2.7–3.8). All the cats were domestic shorthairs, comprising two spayed females, one intact male and one castrated male. All cats had copper-colored irises. Three had a history of neurologic signs, including ataxia, aggression, seizures and head tremors. Two cats had ptyalism and lethargy. All cats received standard pre-procedure medical management including levetiracetam (median dose 19.3 mg/kg PO q8h [range 18.5–20]), omeprazole (median dose 1 mg/kg PO q12h [range 1–1.1]), lactulose (median dose 1.25 ml PO q8h [range 0.7–2.5]) and metronidazole (median dose 7.4 mg/kg PO q12h [range 6.6–8.5]) or amoxicillin-clavulanic acid (13.8 mg/kg PO q12h). Clinicopathologic and biochemical findings are included in Tables 1 and 2.

Table 1.

Clinicopathologic findings of relevant complete blood count (CBC) parameters (hematocrit [HCT], mean cell volume [MCV], mean cell hemoglobin concentration [MCHC] and platelets [PLT]) in cats with intrahepatic shunts before percutaneous transvenous coil embolization (PTCE) (n = 4) and 6 weeks after PTCE (n = 3)

CBC parameters	HCT (RI 31–48%)	MCV (RI 40–52 µm3)	MCHC (RI 32–35 µm3)	PLT (RI 195.00–624.00 × 103/µl)
Before PTCE	30 (27.2–35.6)	36.6 (35.4–38.1)	34.55 (33–36.5)	190.00 (68.00–285.00)
After PTCE	31.3 (30.7–32.1)	38 (33.2–42.4)	33.7 (32.6–34)	214.00 (153.00–285.00)

Data are median (range), with laboratory reported reference intervals (RIs)

Table 2.

Clinicopathologic findings of relevant biochemistry parameters (blood urea nitrogen [BUN], creatinine [CREA], glucose [GLU], albumin [ALB], alanine aminotransferase [ALT], alkaline phosphatase [ALP], cholesterol) and ammonia (NH₃) level in cats before percutaneous transvenous coil embolization (PTCE) (n = 4) and 6 weeks after PTCE (n = 3)

Biochemistry parameters	BUN (RI 15–32 mg/dl)	CREA (RI 1–2 mg/dl)	GLU (RI 67–168 mg/dl)	ALB (RI 2.4–3.8 g/dl)	ALT (RI 33–152 U/l)	ALP (RI 27–87 U/l)	Cholesterol (RI 96.0–248.0 mg/dl)	NH3 (RI 11–35 µmol/l)
Before PTCE	16 (10–18)	1 (0.6–1.1)	91 (81–102)	2.95 (2.6–3.2)	42.5 (32–118)	73 (59–115)	147 (114–181)	291.5 (142–350)
After PTCE	22 (13–30)	1.1 (0.8–1.1)	92 (75–121)	3.1 (2.9–3.4)	49 (41–63)	36 (36–48)	161 (123–223)	N/A

Data are median (range), with laboratory reported reference intervals (RIs)

N/A = not available

CTA confirmed a left divisional IHPSS in all four cats, with associated microhepatica in three and renomegaly in all cases. In one cat, mineral attenuating sediment was present in the urinary bladder; however, no discrete uroliths were identified in the bladder or kidneys. The remainder of the abdominal findings were unremarkable in all cats.

Surgical procedure

A single-catheter PTCE was performed in all cats. The patients were placed in dorsal recumbency, and a radiopaque ruler was placed beneath the patients for reference during stent deployment during PTCE (Figure 1a). A 22 G over-the-needle catheter was placed into the left (n = 1/4) or right jugular (n = 3/4) veins. In 3/4 cats, a 4 Fr Mini-Access Kit (MAK; Merit Medical) dilator/sheath was used over the 0.018 inch × 40 cm stainless-steel wire provided in the kit to facilitate placement of a 0.035 inch × 150 cm hydrophilic wire (Infiniti Medical). A 6 Fr × 5 cm vascular sheath (Cordis Advanti) was placed over the wire and secured to the skin with 3-0 Nylon (Ethicon). In one case, the vascular access sheath (Prelude Ideal Hydrophilic Sheath Introducer; Merit Medical) was used after initial jugular access with a 22 G over-the-needle catheter. The 6 Fr vascular sheath dilator in this kit is tapered to a 0.018 inch wire (supplied as part of the kit), which allows for a single wire to be placed through the 22 G over-the-needle catheter and to facilitate the dilator/6 Fr sheath placement in the jugular vein.

Figure 1.

Fluoroscopic images of an 11-month-old domestic shorthair (cat 3) undergoing percutaneous transvenous coil embolization of a left divisional intrahepatic portosystemic shunt (IHPSS). (a) The cat in dorsal recumbency with radiopaque ruler positioned under the patient. (b) Angiographic catheter that had been passed through the shunt and into the portal vein. (c) Isolated portogram under digital subtraction angiography (DSA). (d) Isolated cavogram with caval diameter under DSA. (e) Deployed caval stent in the vena cava spanning the shunt ostium and avoiding the right atrium. (f) One catheter in the IHPSS during coil deployment. (g) The same catheter was used for portal pressure measurement and coil deployment by alternating between the shunt and portal vein locations. (h) Portogram after three coils showing shunt attenuation was performed, which showed the shunt occlusion. (i) Lateral projection of caval stent, thrombogenic coils and triple lumen catheter (arrow)

A 4 Fr Berenstein catheter (Merit OEM) was advanced over the 0.035 inch wire, into the shunt and to the portal vein to determine baseline PP (Figure 1b). Central venous pressure (CVP) was measured from the vascular access sheath and initial gradient between the two was calculated.

Isolated portogram (Figure 1c) and cavogram were performed with a 50:50 mixture of iodinated contrast (Omnipaque 350 mg I/ml; GE Healthcare) and 0.9% NaCl to determine anatomic landmarks. The caval diameter obtained from the CTA was confirmed with a cavogram using a 5 Fr straight marker catheter (Merit Medical) (Figure 1d). The marker catheter was removed over the hydrophilic wire and exchanged for an appropriate size laser-cut self-expanding stent. The stent was deployed in the vena cava spanning the shunt ostium and avoiding the right atrium (Figure 1e), using the radiopaque ruler beneath the cat for position landmarks. The stent delivery system was removed over the wire and replaced with the 4 Fr Berenstein catheter. The Berenstein catheter (Merit OEM) was advanced into the portal vein and post-stent PPs and CVPs were obtained. The same Berenstein catheter was withdrawn into the shunt adjacent to the caval stent and 5 mm × 5 cm platinum embolization coils (Cook Medical) were deployed (Figure 1f). The same catheter was used for PP measurement and coil deployment by alternating between the shunt and portal vein locations until the desired PP and caval pressure gradient was achieved (Figure 1g). Once the desired pressure gradient was achieved, before removal of the Berenstein catheter, the authors routinely performed a final portogram to confirm degree of shunt occlusion (Figure 1h). The vascular sheath was removed over a 0.018 inch guidewire and exchanged for a 5 Fr × 13 cm triple lumen catheter (Figure 1i). No intraoperative complications were noted. The coil size was selected based on the size of the shunt lumen, to ensure adequate packing within the vessel, to promote stability and thrombosis, and available inventory.

The median procedure time was 125 mins (range 98–170). The median initial PP was 5 mmHg (range 3–11), median initial CVP was 0.5 mmHg (range 0–6) and the median baseline gradient was 4 mmHg (range 1–7). The median final pressure gradient was 9 mmHg (range 8–10). The median number of used embolization coils was 2 (range 0–3). The data reported per cat are recorded in Table 3 and the steps of the procedure in Table 4.

Table 3.

Percutaneous transvenous coil embolization (PTCE) procedure details for four cats

Cat	Baseline PP (mmHg)	Baseline CVP (mmHg)	Baseline gradient (mmHg)	Stent size (mm)	Number of coils	Final PP (mmHg)	Final CVP (mmHg)	Final gradient (mmHg)
1	3	1	2	10 × 60	3	12	2	10
2	3	0	3	8 × 40	2	7	−1	8
3	11	6	5	10 × 60	2	3	11	8
4	7	0	7	8 × 40	0*	10	0	10

^*Cat 4 initially had 0 coils placed and then one coil placed at its second PTCE procedure

CVP = central venous pressure; PP = portal pressure

Table 4.

Comparison of percutaneous transvenous coil embolization (PTCE) techniques

Steps	Standard PTCE	Modified PTCE
1. Vascular access	18 g IVC 0.035 inch angled hydrophilic wire	22 g IVC 0.018 inch wire (hydrophilic or Cope wire)
2. Vascular sheath	10 Fr – dogs 9 Fr – cats	6 Fr
3. Shunt catheterization, pressure gradient	CVP – vascular sheath PP – 4 Fr angiographic catheter	CVP – vascular sheath PP – 4 Fr angiographic catheter
4. Localization of anatomical landmarks	Simultaneous portogram/cavogram	Isolated portogram and cavogram
5. Coil deployment and PP measurement	Two angiographic catheters – Coil deployment – Continuous PP monitoring	Single 4 Fr angiographic catheter – Coil deployment and intermittent PP via same catheter

CVP = central venous pressure; IVC = inferior vena cava; PP = portal pressure

Post-treatment and long-term follow-up

Median duration of hospitalization was 2.5 days (range 2–3). One cat (25 %) developed post-attenuation seizures (PAS) 72 h postoperatively (24 h after discharge) and was euthanized because of progressive neurologic signs 48 h later.

One cat needed additional intervention 2.5 months later to add more coils based on clinical signs and blood work results. During this cat’s initial procedure, a pressure gradient between PP and CVP of 10 mmHg developed after caval stent deployment, and no additional coils were placed. To our knowledge, this acute increase of the PP after caval stent deployment has not been previously reported, although it has been suggested the stent alone may increase the pressure gradient. ¹⁵ This rise may be explained by a resistance and flow disturbance at the stent site due to stent wall compliance.

Medical management was discontinued 6 weeks after PTCE. The patient’s clinical signs returned and blood work confirmed persistent derangements, for which re-catheterization of the shunt was recommended. During the second PTCE, the baseline PP was 9 mmHg and baseline CVP 2 mmHg. After one coil (5 mm × 5 cm) was placed, the PP increased to 15 mmHg and CVP decreased to 1 mmHg, leading to a final gradient of 14 mmHg. No signs of postoperative portal hypertension or other complications developed, and the cat had a routine recovery to discharge. The surviving three cats recovered well, with a median follow-up of 18 months (range 9–19). Medical management, including lactulose, antibiotics and anticonvulsant, was gradually weaned based on blood work (Tables 1 and 2) and clinical status by a median of 2 months (range 1–3). One cat was transitioned from a low-protein diet to a moderate-protein diet 12 months postoperatively; dietary information for the remaining surviving cats was not reported. Proton-pump inhibitor therapy was continued in all the patients.

Discussion

The use of a smaller vascular sheath and single catheter for individual venous mapping and alternating coil deployment and PP monitoring enabled successful treatment of IHPSSs in four cats without complications such as vascular trauma, stent malpositioning or portal hypertension. In humans, the rate of central venous catheter placement complications is in the range of 5–19.5%, including major complications such as arterial puncture, hemothorax, hematoma, ongoing hemorrhage, pneumothorax, thoracic duct damage, tearing injury to vessels, cardiac tamponade and aortic puncture.^{18 –20}

Complications associated with central venous access in veterinary medicine are poorly documented, and many of the severe complications documented in humans have not been reported in cats and dogs. Complications reported for central venous catheterization in dogs and cats include inability to puncture the vessels, hematoma, malposition and cardiac dysrhythmia. Smaller size, lower body weight and thinner body condition score are risk factors for multiple catheterization attempts, with no significant differences between species. Failed venipuncture attempts occurred more frequently in patients with lower Doppler blood pressure; however, this finding did not reach statistical significance, as reported by Reminga et al. ²¹ Lower body weight was related to an increased occurrence of dysrhythmias (premature ventricular complexes) during the procedure in both cats and dogs. ²¹ The standard PTCE technique (Table 4) reported in dogs uses a minimum 10 Fr vascular sheath, which is necessary to accommodate the 10 Fr delivery system for commercially made closed cell laser-cut caval stents (Vet Stent-Cava; Infiniti Medical) and the two angiographic catheters that are used for continuous PP monitoring during coil delivery (two 4 Fr) and simultaneous portogram and cavogram (4 Fr and 5 Fr). In the previous studies of PTCE in cats, a 9 Fr vascular sheath with dilator was used without any complications.^12,13 In cats, the 9 Fr vascular sheath can facilitate the use of two angiographic catheters for simultaneous cavogram/portogram and continuous PP monitoring using the same instrumentation described for dogs. However, the self-expanding open cell laser-cut nitinol stent (Vet-Stent Urethra; Infiniti Medical) used is mounted on a 6 Fr delivery system, requiring only a 6 Fr vascular sheath. In one cat in the same study by Culp et al, ¹³ a 9 Fr vascular access sheath was used and simultaneous cavogram/portogram was performed; however, the PP was measured intermittently using only one angiographic microcatheter.

In the largest cohort study of PTCE for IHPSSs in small and toy breed dogs, separate portogram/cavogram with intermittent PP was described with a 6 Fr vascular sheath. ²² In addition, a single feline case report described PTCE for EHPSS using a 6 Fr vascular sheath with one angiographic catheter. ²³ Previously reported PTCE techniques in cats include procedures performed with or without stent placement,^12,13 depending on the anatomy of the PSS and the risk factor of coil migration to the pulmonary artery. Since an open cell laser-cut stent could be placed in a smaller vascular sheath than the closed laser-cut stent (6 Fr vs 10 Fr) that was reasonably sized for the cat jugular, and caval stenting eliminates the risk of coil migration, the decision was made to place a caval stent in all cats, especially since pulmonary artery migration of a coil in a cat could be fatal without retrieval. Given the diameters of the feline shunts, a 3 mm coil could also be considered to decrease the risk of an abrupt increase in PP. However, caution must be exercised as it is possible that a small coil could migrate through the open cell stent design pending how the stent opened. It was for this reason that 5 mm coils were used in the cats in this case series.

Feline patient size can be a limiting factor regarding vascular access. The smallest introducer is the 9 Fr (3.63 mm outer diameter and 3.21 mm internal diameter), which can accommodate two 4 Fr catheters (1.33 mm outer diameter and 1.07 mm internal diameter) or a 4 Fr and 5 Fr (1.67 mm outer diameter and 1.17 mm internal diameter) catheter for simultaneous pressure monitoring and coil deployment. In the authors’ experience, such large vascular access sheaths can be associated with complications such as venous disruption or tearing and may prolong the procedure, or prevent successful treatment, warranting procedural modification as described. For this reason, this technique could be considered in any sized patient if there are reasons to avoid a larger vascular access sheath, especially when larger diameter open cell laser-cut stents become available.

The modified single-catheter PTCE was performed in all cases after jugular venous collapse and trauma associated with a 9 Fr vascular sheath during standard PTCE in case 1. The 4 Fr MAK (Merit Medical) dilator is tapered to an 0.018 inch wire, allowing the use of a 22 G over-the-needle catheter to obtain vascular access, advancement of a 0.018 inch wire (0.018 inch × 40 cm stainless-steel wire provided in the MAK or 0.018 inch × 50 cm angled hydrophilic wire) (Infiniti Medical), and placement of the 4 Fr vascular sheath, which facilitates placement of a 0.035 inch × 150 cm hydrophilic wire. The MAK allows for the use of a small catheter to gain initial vascular access and dilation to a sheath that will accommodate a standard 0.035 inch wire and subsequent placement of a 6 Fr vascular sheath with a hemostasis valve over the 0.035 inch wire (Cordis Advanti). In one case, the hydrophilic access sheath (Prelude Ideal Hydrophilic Sheath Introducer; Merit Medical) was used, which offers a single transition from 0.018 inch wire to 6 Fr vascular sheath because the dilator is tapered to a 0.018 inch wire instead of a 0.035 inch wire. The authors found this to be a useful method of avoiding multiple catheter and wire exchanges and facilitated more rapid vascular access.

Alternating catheter placement between the shunt and the portal vein may be time consuming, given the relatively small size of portosystemic shunts in cats and the low number of coils required to achieve appropriate attenuation,^13,21 this approach appears to result in a reasonable procedural time (mean 125 mins [range 98–170]) compared with a previous report (mean 130.5 mins [range 95–141]) ¹³ and may reduce the risk of vascular complications.

Of the four cats, one (25%) developed PAS 72 h postoperatively. In the largest retrospective study of 50 cats with portosystemic shunts, 31 (62%) experienced post-attenuation neurologic signs. Of these 31 cases, PAS occurred in 11 (35%) cats. ²⁴ These findings highlight the importance of discussing the potential postoperative neurologic complications with the owners.

Alternating catheter placement between the shunt and the portal vein may be time consuming and of more limited utility in cases where access to the portal vein via the shunt is challenging because of the shunt morphology, which needs to be considered when assessing preoperative CT and planning for PTCE.

Limitations of the current case series include small sample size and retrospective design.

Conclusions

The case series results suggest the modified single-catheter and small vascular sheath PTCE technique is a reasonable alternative to prevent vascular complications in cats. As a result of the small size of portocaval shunts and the need for few coils, this modification does not prolong the procedure time compared with what has previously been reported in cats and dogs.