SUPRAHEPATIC VENACAVAPLASTY (CAVAPLASTY) WITH RETROHEPATIC CAVA EXTENSION IN LIVER TRANSPLANTATION: EXPERIENCE WITH FIRST 115 CASES

Abstract

Background.  

We first introduced the orthotopic liver transplantation utilizing cavaplasty technique in 1994. This paper describes the surgical technique and assesses the outcome of the cavaplasty OLT.

Methods.  

The cavaplasty procedure was used in 115 consecutive orthotopic liver transplantations, including six left lateral and two right lobe transplantations, between November 1994 and September 2000. Fifty-three (66.3%) transplantations required femoro-axillary veno-venous bypass in the initial 4 years, whereas only eight (22.9%) needed VB in the subsequent 2 years. Conversion to piggyback or standard technique was not necessary in any patient.

Results.  

Median results are as follows: operative time 4.5 hr, warm ischemia time 25 min, and blood transfused (packed red blood cells) 6 units. These findings did not differ between first transplantation and retransplantation. There were no perioperative deaths related to the cavaplasty technique. No hepatic venous outflow obstruction was observed, including living-related OLTs. No patient required postoperative hemodialysis for acute renal failure. The median intensive care and hospital stays were 2 days and 10 days, respectively.

Conclusions.  

The cavaplasty technique requires no retrocaval, hepatic vein, or short hepatic vein dissection, and the inferior vena cava can be preserved, which provides advantages for hepatectomy and easy hemostasis, especially during retransplantation. The wide-open triangular caval anastomosis is easy to perform, allowing short implantation time and size matching and avoiding outflow obstruction. The short implantation time reduces the need for veno-venous bypass. Our experience indicates that the cavaplasty technique can be applied to all patients and is justified by minimal technical complications.

INTRODUCTION

The technique of orthotopic liver transplantation (OLT) has evolved significantly since its introduction by Starzl et al. in 1963 ⁽¹⁾. Their widely applied codified procedure involves the removal of the retrohepatic vena cava (RVC) as part of the recipient hepatectomy, which may result in substantial hemodynamic instability during the anhepatic phase ⁽²⁾. The stress of caval cross-clamping and removal was subsequently minimized by the development of a femoro-porto-axillary veno-venous bypass (VB), as described by Shaw et al. in 1984 ⁽³⁾. The piggyback (PB) technique, with preservation of the native inferior vena cava (IVC) and direct anastomosis between the donor’s vena cava and the recipient’s hepatic veins, was described by Calne and Williams ⁽⁴⁾, Starzl ⁽⁵⁾, and Tzakis et al. ⁽⁶⁾. This strategy avoids retrocaval dissection and offers the advantage of better hemodynamics during the anhepatic phase and therefore reduces the need for VB ^(7,8). In the last decade, the PB OLT technique has become accepted worldwide.

However, controversy remains regarding PB OLT technique. To preserve caval flow throughout the procedure, the donor cava can be sewn to the middle and left hepatic veins, but this leaves the recipient at risk for outflow problems, which can be severe enough to require retransplantation ^(9,10). Although modifying the PB technique using three suprahepatic vein anastomoses or side-to-side cavocaval anastomosis decreases the risk of outflow obstruction ^(11–13), the risk cannot be totally eliminated ^(9,10). In addition, there is partial caval occlusion with the use of these modified PB techniques. Likewise, increased hepatectomy time and hemorrhage are associated with the preservation of the IVC for the PB technique ⁽⁹⁾. Inadequate graft size, anatomical abnormalities of the RVC, and difficult hepatectomies may further limit its use ^(8–10). In fact, Busque et al. reported that 33 of 131 (25%) consecutive adult PB OLTs were converted to the standard technique because of anatomical problems, tumor, or other difficulties ⁽⁸⁾.

Since 1994, we have designed a novel OLT technique (cavaplasty) for easy hepatectomy, cavocaval anastomosis, and elimination of hepatic venous outflow obstruction ^(14,15). Shortened implantation time reduces the need for VB. This publication details our surgical technique and summarizes our results and complications with the cavaplasty technique over a 6-year period.

MATERIAL AND METHODS

This is a retrospective review of 115 consecutive OLTs performed by Y.W. at the University of Iowa Hospitals and Clinics from November 1994 to September 2000. There were 101 first transplantations and 14 retransplantations in 111 patients. The mean age of the donors was 41.7±16.8 (range 18–84). Recipient pretransplantation data, including demographics, previous upper abdominal operations, cause of liver disease, Child-Pugh classification of hepatocellular function, and United Network for Organ Sharing (UNOS) status at the time of transplantation are shown in Table 1. We analyzed the cavaplasty technique in terms of its feasibility, use of VB, operative time, warm ischemia time, intraoperative blood requirements, surgical complications, postoperative course, intensive care unit (ICU) and hospital length of stay, and patient/graft status. Donors were procured using the standard technique and liver grafts were preserved with University of Wisconsin solution. Posttransplantation immunosuppression consisted of tacrolimus and prednisone. Acute rejection episodes were treated with high-dose steroid therapy. Steroid-resistant rejection episodes were treated with antithymocyte globulin (ATG/ALG) or muromonab-CD3 (OKT3). The median duration of posttransplantation follow-up was 22.6 months (range from 2 days to 69.9 months). No patients were lost to follow-up after transplantation.

Table 1: Baseline demographics

Surgical Technique

The recipient’s hilar structures were dissected, ligated, and divided as described by Starzl ⁽¹⁾. The left triangular and gastrohepatic ligaments were divided. The right triangular ligament was then divided and the right lobe was moved from the diaphragm to the IVC. Dissection was always minimized during the hepatectomy. The portal vein and IVC were subsequently cross-clamped. If the patient could not tolerate this cross-clamping with the use of vasopressor, a single femoro-axillary VB was used. VB was preferred for patients with limited cardiovascular reserve. After transecting the portal vein, two large vascular clamps were placed straight down the RVC, incorporating the suprahepatic and infrahepatic vena cava. A German vascular clamp was applied to the superior vena cava, while the other vascular clamp was placed obliquely on the IVC above the adrenal vein (Fig. 1). The liver was removed with care to preserve the IVC and the hepatic veins. The orifices of the three hepatic veins were connected and then the anterior wall of the cava was opened with longitudinal incision to fashion a wide-open triangular cuff (Fig. 2). The posterior wall of the donor IVC was also incised to fashion a wide-open triangular cuff sized to match the IVC opening in the recipient (Fig. 2). The wide donor and recipient triangular caval openings were anastomosed using three 4-O Prolene sutures continuously (Fig. 1). After flushing the graft with 400 to 800 ml cold Lactate Ringer’s solution, the donor’s infrahepatic vena cava was ligated, and portal vein anastomosis performed. The cross-clamps were then removed and the liver was reperfused (Fig. 2). Lastly, the remaining hepatic artery and bile duct anastomoses were completed.

Figure 1: Cavaplasty orthotopic liver transplantation vena cava anastomosis.

Figure 2: Regular cavaplasty orthotopic liver transplantation. (2a) Lines of incision on the recipient hepatic veins and inferior vena cava. (2b) Wide-open triangular cuff on the recipient inferior vena cava. (2c) Donor cavatomy from dorsal aspect. (2d) Finished cava anastomosis.

For split, left-lateral and right-lobe living-related liver transplantations, the free surface of the donor hepatic vein was opened longitudinally for an appropriate distance to match the opening of the recipient vena cava. This often required transection of a thin layer of hepatic tissue covering the vein proximally. Then triangular openings were created in both the vena cava and hepatic veins (Fig. 3 and Fig. 4 for left-lateral and right-lobe living-related transplant, respectively). A continuous, wide-open triangular caval anastomosis was then performed by connecting the three walls of the triangular openings.

Figure 3: Living-related cavaplasty orthotopic liver transplantation using left lateral segment of liver graft. (3a) Wide-open triangular cuff on the recipient inferior vena cava. (3b) Left lateral segment graft from living donor. (3c) Donor left lateral segment hepatic vasotomy. (3d) Finished cava anastomosis.

Figure 4: Adult living-related cavaplasty orthotopic liver transplantation using right lobe of liver graft. (4a) Lines of incision on the recipient hepatic veins and inferior vena cava and closed left hepatic vein. (4b) Recipient cavatomy (right and middle hepatic vein). (4c) Donor right lobe hepatic vasotomy. (4d) Finished cava anastomosis using right lobe of liver.

Statistical Analysis

Normally distributed variables were presented as mean±SD. Skewed variables were characterized by the median value. The Mann-Whitney U test was used to compare quantitative differences. A P-value <0.05 was considered significant. The survival data were computed using the Kaplan-Meier method. SPSS Version 9.0 software (SPSS, Inc., Chicago, IL) was used for the statistical analysis.

RESULTS

The cavaplasty technique was successfully performed in all 115 OLTs. No patients required conversion to standard or PB OLT technique, including 6 pediatric living-related liver transplants using the left lateral segment and two adult living-related liver transplants using the right lobe. Femoro-axillary VB was used in 61 (53.0%) transplantations. However, with the accumulation of experience by the transplant team, the use of VB dropped from 66.3% in the initial 4 years of the study to 22.9% in the subsequent 2 years. No patients were placed on VB in most the recent year (Fig. 5). VB was not necessary in any of the living-related OLTs. Five of 6 patients with fulminant hepatic failure tolerated cavaplasty OLT without VB despite the absence of portosystemic collaterals. However, femoro-axillary VB was used to maintain hemodynamic stability in 8 of the 10 patients with preexisting coronary artery disease. No patient required porto-axillary VB.

Figure 5: Femoro-axillary bypass versus no bypass over time.

Median results are as follows: operative time 4 hr and 30 min (range 2 hr 6 min to 9 hr 20 min), warm ischemia time 25 min (range 19–30 min), and intraoperative red blood cell transfusion (packed red blood cells (PRBCs)) 6 units (range 0–49 units). The preoperative and maximum serum creatinine in the first 5 postoperative days were 1.5±1.1 and 1.9±1.2 mg/dl, respectively. The difference between pretransplantation and maximum posttransplantation serum creatinine did not reach statistical significance between patients with or without VB (median increase 0.1 versus 0.2 mg/dl, respectively, P =0.830). No patient required postoperative hemodialysis for acute renal failure. There was no difference between first transplantation and retransplantation in operative time (4 hr 25 min vs. 5 hr 25 min, P =0.491), warm ischemia time (24 vs. 29 min, P =0.069), or PRBC usage (6 vs. 7.5 PRBC units, P =0.249). Likewise, the intraoperative blood requirement did not differ significantly between patients with or without prior upper abdominal surgery (6 vs. 6 PRBC units, P =0.739); or between patients with or without chronic viral hepatitis (8 vs. 6 PRBC units, P =0.265).

There were no anatomic limitations to the use of cavaplasty technique in these 115 consecutive OLTs, including 2 patients with chronic Budd-Chiari syndrome and 6 patients with hepatocellular carcinoma (HCC). There were 2 patients with stage I, II, and III tumors each, according to the primary tumor, the regional lymph nodes, and metastatic disease (TNM classification) as outlined by the American Joint Committee on Cancer. Currently, four of six HCC patients are well, without tumor recurrence after follow-ups of 6.6, 16.3, 23.1, and 35.4 months, respectively. One patient died with tumor recurrence 10.7 months after transplantation, with occult vascular invasion of tumor revealed in explant pathology. The second died from sepsis secondary to a diabetic foot infection at 23.4 months after transplantation, without evidence of tumor recurrence.

There was no perioperative mortality related to this technique. Other posttransplantation surgical complications are listed in Table 2. Retransplantation was required in four patients: immediate hepatic artery thrombosis (1), delayed (17 months later) hepatic artery thrombosis (1), primary nonfunction (1) and recurrent hepatitis C (1). Postoperative course and outcomes are presented in Table 3. One-year actuarial patient and graft survival for the 101 first transplants was 91.2% and 87.0%, respectively. One-year actuarial patient and graft survival for the 14 retransplantations was 78.6% and 78.6%, respectively. Eleven patients died within 1 year of OLT. The causes of death were fungal infection (2), early recurrent hepatitis C (2), recurrent HCC (1), pancreatic carcinoma (1), posttransplantation lymphoproliferative disorder (1), brain death from fulminant hepatic failure (1), spontaneous intracranial hemorrhage (1), hemorrhagic pancreatitis after endoscopic retrograde cholangiopancreatography (1), and sudden cardiac arrest (1).

Table 2: Surgical complications

Table 3: Postoperative course and outcomes

DISCUSSION

The standard technique of OLT is a well-established surgical procedure that involves removal of the retrohepatic vena cava along with the diseased liver ⁽¹⁾. Both the IVC and portal vein are cross-clamped during the anhepatic phase and, therefore, VB is used to help maintain hemodynamic stability and decompress the portal system. This has contributed to a decrease in morbidity and early mortality ⁽³⁾. However, VB is associated with complications such as thromboembolism, air embolism, and hypothermia ⁽¹⁶⁾.

Hepatectomy with IVC preservation or PB procedure can maintain hemodynamic stability during the anhepatic phase without the use of VB; therefore, it is particularly suitable for patients with marginal cardiovascular reserve ⁽⁸⁾. Other advantages of the PB technique, such as fewer hemorrhagic complications because of the absence of retrocaval dissection, make it an attractive alternative to the use of the standard technique and VB. However, there are disadvantages to the PB, such as acute Budd-Chiari syndrome from hepatic venous outflow obstruction, prolonged hepatectomy time, and difficult hepatectomy, which may cause hemorrhage ^(8–10). Different technical modifications of the PB technique such as three suprahepatic vein anastomoses or side-to-side cavocaval anastomosis have been described to minimize these potential complications ^(11–13). Because of its advantages, the PB OLT technique has become accepted worldwide during the past decade. Also, results of OLT have been improving over the past 2 decades ⁽¹⁷⁾. However, the surgical procedure still remains a major undertaking, particularly the critical period surrounding hepatectomy and implantation.

In the standard OLT technique, the hepatectomy requires a retrocaval dissection with complete RVC and IVC clamping. The cavaplasty technique minimizes these difficulties during hepatectomy in retrocaval dissection. The lower caval vascular clamp is placed obliquely above the adrenal vein to avoid right adrenal vein dissection and potential adrenal injury. Compared with PB hepatectomy, cavaplasty obviates the need for short hepatic vein or hepatic vein dissection. Therefore, there is no extra dissection time and less potential for hemorrhage during the hepatectomy. In our experience, this hepatectomy technique (cavaplasty) reduced our operative time and blood transfusion requirement. It is especially advantageous for retransplantation. By placing the two large vascular clamps straight down during the hepatectomy, it is unnecessary to perform superior vena cava or IVC dissection. Furthermore, the dissection of scar tissue around the posterior wall of RVC or IVC can be avoided during retransplantation. Our data demonstrate that the operative time, warm ischemia time, and intraoperative blood requirement did not differ significantly between first transplantation and retransplantation. Likewise, our procedure is feasible for patients with chronic hepatitis B or C, despite the presence of dense inflammatory adhesions, and for patients with Budd-Chiari syndrome complicated by a hypertrophied caudate lobe wrapping around the vena cava. Compared with PB, cavaplasty technique might provide adequate surgical margins for patients with stage I or II HCC through removal of the diseased liver in conjunction with the anterior wall of IVC, while leaving enough cuff for the caval anastomosis. Similar to the standard technique, the IVC is also cross-clamped during hepatectomy by the cavaplasty technique. This cross-clamping decreases renal perfusion. However, no patients required postoperative hemodialysis for acute renal failure. The difference between preoperative and maximum creatinine in the first 5 postoperative days did not reach statistical significance between patients with and without VB. Our results presented here indicate that, with easy hepatectomy and anastomosis, this cavaplasty technique can minimize the cross-clamping time and thereby preserve the renal function.

The cavaplasty technique involves opening all three hepatic veins and the anterior wall of the vena cava to fashion a wide-open triangular opening in the recipient. This technique has several advantages during implantation. First, it offers a broad-based and adjustable caval anastomosis, for the greatest possible venous outflow. Indeed, no hepatic venous outflow obstruction was observed in these 115 cavaplasty OLTs. This wide-open triangular caval anastomosis also eliminates anastomotic kinking resulting from an event such as the rotation of a small liver graft in a large abdominal cavity, which might result in outflow obstruction. In fact in one case, a liver graft from an 8-year-old donor was perfectly matched for an 82-kg adult recipient. This broad-based and adjustable cavocaval anastomosis poses a special advantage for right-lobe living-related liver transplantation by offering the greatest possible venous outflow. In this patient series, no reconstruction of accessory hepatic veins was necessary to prevent limited outflow. Second, this single wide-open caval anastomosis is easy to perform. Compared with other published data, our warm ischemia time was short (Table 4), and biliary complications were low ⁽¹⁸⁾. Third, it allows easy size matching. Size discrepancies between the donor and recipient will not hinder the use of cavaplasty, because the recipient vena cava triangular opening is adjustable to accommodate the donor vena cava.

Table 4: Comparison of perioperative and postoperative data between different techniques for cava anastomosis in liver transplantation

Compared with standard technique, the PB technique reduces the use of VB. However, selected use of the VB is still indicated in the PB technique ⁽¹⁹⁾. The portal vein and vena cava are also cross-clamped in the cavaplasty technique. Most patients tolerated this clamping well without VB by the application of vasopressor. The use of vasopressor can avoid large-volume fluid infusion. Our data show that there were no complications related to the use of vasopressor, which might be related to the short clamping time, as reflected in our short warm ischemia time. Accumulation of experience by the transplant team led to diminished use of VB over time, from 66.3% in the initial 4 years of the study to 22.9% in the latter 2 years. No patients were placed on VB after January 2000. Our results also demonstrated that porto-axillary VB was not necessary to maintain hemodynamic stability in any patient. After the liver was revascularized, no reperfusion syndrome was seen in these 115 OLTs, perhaps attributable to the short anhepatic phase and warm ischemia time. In addition to short clamping time, short hepatectomy time and less bleeding make cavaplasty our preferred technique for patients with marginal cardiovascular reserve. We did not observe any detrimental effects in our 10 patients with preexisting coronary artery disease, including 3 who underwent coronary stent or bypass surgery before transplantation. In our practice, femoro-axillary VB will be indicated for patients intolerant of portal and vena cava cross-clamping and for those with marginal cardiovascular reserve.

A further randomized comparison between the cavaplasty and other classical techniques is necessary. However, analysis of relevant data reported by other transplantation experts up to 2000 permits an assessment of our surgical results and demonstrates a positive impact of this cavaplasty procedure on perioperative and postoperative events, which in turn may contribute to the short ICU and hospital stays experienced by our patients (Table 4) ^{(3,6–8,13,19–27)}. In addition, the number of transplantations performed is an important characteristic to consider when evaluating the feasibility and outcomes of surgical technique.

In summary, this cavaplasty technique was feasible for all 115 OLTs with few technical complications. No retrocaval dissection was required and IVC was preserved. It also preserved the right adrenal vein and avoided risk of injury to the adrenal gland. It allowed short and easy hepatectomy, thus posing an advantage for retransplantation. The wide-open triangular caval anastomosis is easy to perform, allowing for size matching and avoidance of outflow obstruction. When only a single donor hepatic vein is available in living-related right-lobe transplantations, this large, broad-based triangular anastomosis eliminates the risk of anastomotic kinking and outflow obstruction by allowing for the maximum possible venous drainage. No reconstruction of the accessory hepatic veins is needed. Femoro-axillary VB will be necessary only for patients intolerant of portal and caval clamping with vasopressors and for patients with limited cardiovascular reserve. We believe this technique offers advantages over both the standard and PB OLT techniques. The real value will be determined by more clinical practice in the liver transplantation field. With respect to these intraoperative and postoperative considerations, cavaplasty is a technical step forward for OLT.