Auxilliary Liver Transplantation According to the RAPID Procedure in Noncirrhotic Patients

Technical Aspects and Early Outcomes

Abstract

Objective: 

To present technical details and short-term experiences of liver transplantation as a 2-stage procedure using small for size grafts in a multicenter cohort study.

Background: 

Two-stage liver transplantation using small for size grafts should be a feasible procedure with lower morbidity and mortality rates. Retrospective cohort study between 2015 and 2022 with multicenter experience. Twenty-three resection and partial liver transplantation with delayed total hepatectomy procedures for noncirrhotic indications were performed in 6 European centers (20 with grafts from living donors and 3 after deceased donation). Procedure’s feasibility, graft volumetric changes, morbidity, and mortality of donor and recipient were explored.

Results: 

There was a low donor morbidity (4.3%) in our cohort. Hypertrophy of the graft was rapid (mean graft volume increases 107% between both stages) and offered the opportunity for remnant hepatectomy after a median of 14 days. In all cases, portomesenteric flow was routed to the graft by right remnant portal vein ligation. Portal vein inflow modulation to alleviate transient harmful portal hypertension was not needed in any case. Early postoperative mortality (4.3%) of the recipients were low. Ten patients suffered from complications ≥IIIb according to the Clavien-Dindo classification.

Conclusions: 

Two-stage liver transplantation is a feasible option for noncirrhotic patients allowing the safe use of small for size grafts and could possibly be extended with caution to liver diseases with portal hypertension and cirrhosis. The resection and partial liver transplantation with delayed total hepatectomy technique might be a viable option for expanding the donor pool given the current organ shortage especially for low–model of end stage liver disease patients.

Two-stage liver transplantation was established by Rudolf Pichlmayr and Henry Bismuth in the late 1980s as a bridging therapy for patients with acute liver failure. For this purpose, postmortem left split livers were used. After regeneration of the recipients’ own liver, immunosuppression was slowly discontinued.^1–3 The Kyoto group used a 2-stage procedure with living donor partial livers to treat metabolic diseases.⁴ In this scenario, the original remnant liver was removed in a second stage after a certain time interval for regeneration of the transplanted split liver in the recipient to prevent a small for size syndrome.⁵ A large variety of indications were used for such a 2-stage transplantation during this time. The Oslo group reported on the RAPID concept (Resection And Partial Liver Transplantation With Delayed Total Hepatectomy) for patients with nonresectable colorectal liver metastases utilizing left lateral segment splits from deceased donors,⁶ and this was further followed up with the same transplant indication by using living donor graft by the Tübingen group.⁷ In these patients, normal liver function without portal hypertension is of special advantage for the success of the procedure.

The technique offers major advantages: it minimizes the donor morbidity through the minor resection in the case of living donation or the usage of extended right lobe for another adult recipient in case of a splitting procedure after deceased donation. It also confers a hepatic metabolic backup to the recipient in case of graft dysfunction. Despite this, the number of patients treated with a 2-stage procedure is small. Technical complexity and its potential impact on early outcome may represent 1 element of concern. A further common concern in this context is that the interactions of the 2 partial livers during regeneration process between stage 1 and 2 were and are not fully understood.

Herein, we present the experiences with auxiliary, 2-stage liver transplantation according to the RAPID concept for noncirrhotic indications using partial grafts from living and deceased donors with special consideration on technical aspects and early outcomes.

PATIENTS AND METHODS

We retrospectively collected the data from the transplant centers of Jena, Padova, Brussels, Oslo, Tübingen, and Munich who have performed 2-stage liver transplantation for patients with noncirrhotic liver diseases in the recent years.

The recorded parameters were the following:

Indication for transplantation; donor type (deceased or living); the segment-orientated type of resection according to the Brisbane classification⁸ and The “NewWorld” Terminology for hepatectomies⁹; technique of portal branch ligation and measurements/results of portal pressure measurement; donor liver segments used according to the Couinaud classification; type of graft venous outflow reconstruction, the portal venous, and arterial inflow; usage of a graft for any vascular reconstruction; type of biliary anastomosis; interval between the 2 stages (in days); technique of remnant hepatectomy; graft to body weight ratio (GBWR) at stages 1 and 2; graft volume at stage 1 and 2; performance of quantitative liver function test and results at stages 1 and 2; donor and recipient morbidity; graft function; recipient’s outcome and duration of follow-up (we focused on the first 12 months and mainly focused on the postoperative results within 90 days).

Transplantation Technique

During the first stage, a left hemihepatectomy was performed in the recipients (segments 1–4) except for 1 case where a left trisectionectomy was performed because of 1 large metastasis along the Cantlie line. In place of the recipient’s left lobe, the left (-lateral) liver lobe was transplanted. After reperfusion of the graft, redirection of portal flow to the graft was performed by total/subtotal ligation of the remnant right portal branch or resection of the portal bifurcation to induce hypertrophy. Portal flow redirection was done with or without the guidance of flow and pressure measurements of graft inflow. After a variable waiting period (see below), whereby the liver function was continuously monitored, the removal of recipient’s remnant right liver lobe was performed. An illustration of the technique is shown in Figure 1.

FIGURE 1: Technical principle of the 2-stage transplantation (figures from Rauchfuß et al¹⁰): 1A, Recipient’s deceased liver. 1B, Left-sided hepatectomy (segments 1–4) of the deceased liver. 1C, Transplantation of the left (-lateral) liver lobe and ligation of the right portal branch. 1D, Removal of the deceased right liver lobe after a median waiting time of 14.5 days.

The living donor operation was performed in all cases as an open approach. The liver transection was performed using a Cavitron Ultrasonic Surgical Aspirator (Integra LifeScience, Princeton).

Calculation of Kinetic Growth Rate

For the calculation of a kinetic growth rate, the following formula was used:

$\frac{\left(\mathrm{Volume}\mathrm{in}\mathrm{ml}\mathrm{at}\mathrm{stage}\mathrm{2}-\mathrm{Volume}\mathrm{in}\mathrm{at}\mathrm{stage}\mathrm{1}\right)}{\mathrm{Interval}\mathrm{between}\mathrm{both}\mathrm{stages}(\mathrm{days})}=\mathrm{KGR}\left(\frac{\mathrm{mL}}{\mathrm{d}}\right)$

Statistical Analysis

All analyses were performed using IBM SPSS Statistics 25 (IBM, Armonk, NY). First, line charts were performed to illustrate the increase of absolute liver volume and graft to body weight ratio between the 2 steps (Fig. 2). Subsequently, the significance of this volume increment was tested by means of Wilcoxon signed-rank test for dependent samples. Nonparametric method was chosen because of low numbers, and that normal distribution could not be predicted. Probability levels of P<0.05 were considered statistically significant.

FIGURE 2: Increase of graft volume (A) and GBWR (B) between stage 1 and 2.

RESULTS

Twenty-three noncirrhotic patients were transplanted in the 6 centers according to the RAPID procedure. Twenty transplantations were performed with grafts from living donors, 3 were splits grafts from deceased donors. The distribution of performed transplantation was Jena n=10; Brussels n=4; Oslo and Padova n=3; Munich n=2; and Tübingen n=1. In the majority of cases (n=16), the left lateral lobe (segments 2 and 3) was used (H23 according to the “NewWorld” terminology); in 7 cases, the left lobe (segments 2–4) was transplanted (H234 according to the “NewWorld” terminology). All but 1 were blood group compatible. The only incompatible recipient was transplanted after the application of the local desensitization protocol. The median follow-up was 671 days (range: 20–2675 days), whereby the majority of cases were performed between 2020 and 2022.

Indications

In the majority of cases, the indication for transplantation was nonresectable liver metastases from colorectal carcinoma (n=21); in 1 case the indication was nonresectable liver metastases of neuroendocrine tumor and in another case the patient suffered from a β-catenin–mutated irresectable liver adenomatosis.

Living Donors

The living donors had the following relationships to the recipients: siblings (n=5), children (n=9), spouses (n=3), nephew (n=1), brother in law (n=1), and friend (n=1). The median donor age was 43 years (range: 19–63), the median operation time was 285 minutes (range: 180–642), and the median hospital stay was 8 days (6–12 days).

Technical Aspects

Left Recipient Hepatectomy

Segment 1 was removed en bloc with left liver in all patients to completely free the residual right liver from the vena cava. This offers more place to the left graft and allows a better venous outflow for the graft by using the common ostium of left hepatic vein and middle hepatic vein. This procedure aimed at facilitating the completion hepatectomy 15 days later.

In 3 cases, the transection plane was pretreated with Microwave ablation to reduce the risk of encountering viable tumor cells in the transection phase.

Reconstruction of the Venous Outflow

In 9 patients, the venous outflow was reconstructed directly to the stump of the left hepatic vein, in 11 patients to the ostium of the middle and the left hepatic vein, and in 3 cases to the ostium of the middle and the left hepatic vein using a ring of autologous saphenous vein implanted during the bench graft preparation.

Reconstruction of the Portal Venous Inflow and Techniques of the Right Portal Inflow Modulation

In 16 patients, the graft’s portal vein was anastomosed end-to-end to the stump of the left portal vein. In 4 patients, the portal bifurcation was resected and the graft’s portal vein was reconstructed end-to-end with the main portal trunk. In 3 patients, the portal inflow was established end-to-side to the main portal trunk of the recipient.

The portal venous pressure was measured intraoperatively in 52.2% of the patients. In general, an increase of the portal pressure up to 20 mm Hg was tolerated and a complete ligation of the portal branch was performed. Subtotal ligation, means in this context, that the ligation suture was slowly tightened up to a portal pressure of 20 mm Hg and left in this position. Therefore, the technique for diverting portal flow to the graft differed widely: in 4 patients, right portal vein ligation was only performed subtotally, in 11 patients totally, in 2 patients totally but 2 or 3 days, respectively, after the stage 1 procedure; in 4 patients the portal bifurcation was instead resected and in 2 patients the right portal branch was transected. The highest measured portal venous pressure after right portal vein clamping was 30 mm Hg. This was 1 of the 4 cases, where only a subtotal ligation of the right portal branch was performed.

Reconstruction of the Arterial Inflow

The graft arterial anastomosis was performed to the stump of the left hepatic artery of the recipient in 10 cases, whereas in 2 interposition graft (in both cases recipient’s saphenous vein) was used.

In 10 patients, the graft artery was implanted end-to-side on to the common hepatic artery. An interposition graft was used in 6 of these reconstructions, in 5 the recipient’s saphenous vein and in 1 case an allogenic iliac artery from a deceased donor.

In 2 cases, the graft artery was anastomosed end-to end to a dominant left gastric artery of the recipient.

Biliary Duct Anastomosis

In 16 patients, the biliary drainage was established by a bilioenteric anastomosis. In 7 patients, a duct-to-duct anastomosis to the stump of the recipients left bile duct stump was performed. The latter was performed in an end-to-end technique suturing the stump of the graft to the left bile duct of the recipient using a 6/0-resorbable suture with single stitches. Key point of this step is a maintained perfusion especially of the recipient’s bile duct. In 1 case, the duct–duct anastomosis was switched to a bilioenteric anastomosis for a persistent leakage of the initial anastomosis.

A silicon sheet was positioned between the graft and the right liver lobe in 4 patients to avoid adhesions.

Second Stage: Removal of Remnant Native Liver

The removal of remnant native liver was performed as step 2 of the operation after adequate volumetric and functional regeneration of the liver graft. The median time interval between the 2 stages was 14 days, ranging from 10 to 60 days. Volumetric and functional assessment of the graft was requested before this second surgical step. A delayed second-stage hepatectomy was adopted because of a portal vein thrombosis with impaired graft function at Tc99 mebrofenin scintigraphy in 1 case and the blood-group incompatible transplantation in another patient.

In one of the centers, all the patients (3 cases) underwent remnant hepatectomy by means of a fully videolaparoscopic approach with liver extraction by a Pfannenstiel or a 10-cm incision. In the rest of the present series, remnant hepatectomy was performed by conventional laparotomy.

The techniques in the individual patients are summarized in Table 1.

TABLE 1 - Operative Details of the Individual Patients

Patient Number	Donor Type	Venous Outflow	Portal Venous Inflow	Arterial Inflow	Use of Graft	Biliary Drainage
1	Living	Stump LHV	E-E left portal stump	Stump left hepatic artery	No	Bilioenteric
2	Living	Stump LHV	E-E main portal trunk (resection of bifurcation)	Stump left hepatic artery	No	Bilioenteric
3	Living	Stump LHV	E-E main portal trunk (resection of bifurcation)	Stump left hepatic artery	No	Bilioenteric
4	Living	Stump LHV	E-E main portal trunk (resection of bifurcation)	Stump left hepatic artery	No	Bilioenteric
5	Living	Stump LHV	E-E main portal trunk (resection of bifurcation)	Stump left hepatic artery	No	Bilioenteric
6	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (VSM)	Bilioenteric
7	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (VSM)	Bilioenteric
8	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (VSM)	Bilioenteric
9	Living	Ostium MHV/LHV	E-E left portal stump	Direct in common hepatic artery	No	Duct-duct
10	Living	Ostium MHV/LHV	E-E left portal stump	Stump left hepatic artery	No	Duct-duct
11	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (VSM)	Duct-duct
12	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (Iliac artery after deceased donation)	Duct-duct
13	Living	Ostium MHV/LHV	E-E left portal stump	Common hepatic artery	Yes (VSM)	Duct-duct
14	Living	Ostium MHV/LHV	E-E left portal stump	Stump left hepatic artery	Yes (VSM)	Duct-duct
15	Living	Ostium MHV/LHV	E-E left portal stump	Stump left hepatic artery	Yes (VSM)	Duct-duct
16	Living	Ostium MHV/LHV with autologous saphenous ring	E-E left portal stump	E-E with left gastric artery	No	Bilioenteric
17	Living	Ostium MHV/LHV with autologous saphenous ring	E-E left portal stump	E-E with left gastric artery	No	Bilioenteric
18	Living	Ostium MHV/LHV with autologous saphenous ring	E-E left portal stump	Stump left hepatic artery	No	Bilioenteric
19	Living	Ostium MHV/LHV	E-E left portal stump	Stump left hepatic artery	No	Bilioenteric
20	Deceased	Stump LHV	Side-End main portal trunk	Direct in common hepatic artery	No	Bilioenteric
21	Deceased	Stump LHV	Side-End main portal trunk	Direct in common hepatic artery	No	Bilioenteric
22	Deceased	Stump LHV	Side-end main portal trunk	Direct in common hepatic artery	No	Bilioenteric

LHV indicates left hepatic vein; MHV, middle hepatic vein; VSM, great saphenous vein.

Volumetric Changes and Liver Function

The median graft to body weight ratio was 0.48% prior stage 1 (range: 0.28–0.74). The highest GBWR of 0.74 was a calculation bias, as the donor had multiple liver cysts, which could not be subtracted in the volumetric calculation. Before performing stage 2, GBWR was 0.99% (range: 0.58–1.39). The median graft volume increased from 355 mL (prior stage 1) to 735 mL (prior stage 2). The increment in GBWR and graft volume was statistically significant (P<0.0001) and is depicted in Figure 2. The median kinetic growth rate was 25 mL/d (range: 8–64).

Donor and Recipient Morbidity

There was no donor mortality in the postoperative course. In 1 case, a bilioenteric anastomosis had to be performed because of an injury of the right bile duct during the donor procedure.

Seven recipients suffered from a Clavien-Dindo IIIb complication, mainly biliary leakages. Three patients had Clavien-Dindo IVa complications (2 patients with persistent biliary leakages and 1 with a compartment syndrome of the legs because of an excess of compression by deep vein thrombosis prevention measures. This patient subsequently needed hemodialysis).

In the early postoperative course, 1 patient died. The 90 days mortality of the recipients was 4.3% (Clavien-Dindo V: multiorgan failure after hepatic artery thrombosis on POD 44).

Follow-up

The median follow-up was 696 days. The 1-year-survival is shown in Figure 3. Patients and graft survival are identical, no recipient was retransplanted.

FIGURE 3: One-year survival curve of the recipients.

The recurrence free survival of the patients with colorectal liver metastases is 66.6% (7 recipients suffered from recurrence within the first year after transplantation).

The extended right lobes of the deceased donors were transplanted successfully in adult recipients. All 3 patients are in good clinical condition.

DISCUSSION

We described the first multicentric European experience with the RAPID technique in adult noncirrhotic recipients. To the best of our knowledge, this is the largest series since the publication of the Kyoto group in the middle 1990s.^11,12 After the introduction of right lobe grafts from living donors for adults, the 2-stage transplantation was only performed in occasional cases or very small case series.¹³ It is nowadays an innovative approach using left (-lateral) splits after deceased donation or after left (-lateral) living donation.

Our short-term experiences are good. Importantly, there is no donor mortality accompanied with a low donor morbidity (4.3%). The 90-day recipient morbidity and mortality level were also low (4.3%) and even better than the ones reported from the ALPPS (Association Liver Partition and Portal vein ligation for staged hepatectomy) registry that represents a similar surgical strategy for the same indication in technically resectable patients. Assuming the dogma that the incidence of morbidity of the living donors depends on the mass of parenchyma resected, we have chosen this “double equipoise”¹⁴ concept shifting the risk from the donor to the recipient. Moreover, the relevantly low living donor risk of a left lateral donation, associated to the absence of harm on the waiting list, maximize the overall transplant benefit of RAPID setting (Fig. 4). In this context, when planning the individual protocols in the centers, all were inspired by the recent publications of the RAPID procedure from the Oslo group⁶ (the patients of the RAPID trial are also included in this series) and from Scatton et al^16,17 and applied this method of the 2-stage liver transplantation for noncirrhotic indications.

FIGURE 4: Equipoise in liver transplantation: RAPID technique optimizes the ratio between harm (donor and waiting list) and benefit. Illustration according to Vitale et al. LDLT indicates living donor liver transplantation; LT, liver transplantation; RL, right lobe.¹⁵

A 2-stage transplantation is a safe procedure for patients with normal liver function and no portal hypertension, even after a left-sided resection. Despite the comparable outcomes of oncologic transplantation for unresectable colorectal metastases with a range of conventional widely accepted indications, these patients fall through the current raster of the allocation system. In this situation, 2-stage transplantation opens up the possibility of a curative treatment option for affected patients. First, the current practice of partial livers from living donors does not burden the organ pool, and even if postmortem left lateral partial livers were used, it would scarcely do so. The 3 additional patients having been transplanted with the extended right lobe after deceased donation showed a good outcome. Second, left lateral donation is significantly safer than right partial liver donation, which is usually required in adults. Third, in the event of a split-liver transplant failure, the technique theoretically allows for pullback without the need for a second emergency liver transplantation. This technique allows the expansion to otherwise nonestablished transplantation indications like colorectal liver metastases for the following reasons: (1) in case of living donation, the donor morbidity and risk is minimized, as only 2 segments are used for transplantation, (2) even in donors, who are not suitable for right lobe donation (eg, due to anatomical factors), this method is applicable in most cases, (3) in case of deceased donation, the donor pool is not burdened, as the extended right lobe is still available for another (adult) recipient. The latter argument cannot be applied without restrictions in countries like Germany where left-lateral lobes after deceased donation are mostly allocated to children and the split itself is not common because of the donor quality.¹⁸ Even in such context, the diffuse adoption of donor age cutoffs for children transplantation (eg, 50 years) potentially leaves rooms for donor resource to be dedicated to RAPID procedures.

There are several technical issues, which should be addressed when performing this special type of transplantation: finding a proper transection plane was possible in all recipients to create enough space for liver vein anastomosis and placement of the graft. In 1 case, a left trisectionectomy was mandatory: a large colorectal liver metastasis was located along the Cantlie line with invasion of median hepatic vein. Finding the correct transection line is a crucial part of the preoperative planning to minimize the theoretical risk of tumor spread and avoiding a R1-situation during the first step of the procedure. Whether a microwave pretreatment of the transection plane increases the chances for avoiding tumoral cell spread has to be explored with a dedicated study. Left hepatic vein reconstruction still remains a crucial technical phase, as a correct outflow is fundamental for graft function and regeneration. Even in the presence of an adequately wide anastomotic ring, the relevant and fast regeneration has the potential to create a kinking of the anastomosis that needs to be prevented through, a wide anastomosis, correct graft location, and fixation of the graft. Suturing a saphenous ring to the left hepatic vein ostium during the bench phase as proposed by Lee in his dual graft experience may represent an additional advantage.

Portal anastomosis was done all times to the recipients left branch or main stump in all living donation grafts, whereas it was performed end-to-side to the main portal trunk. On the basis of the center’s preference, the right portal branch was either (subtotally) ligated (in 1 case even 3 days after stage 1) or transected. Performing “only” a ligation has the theoretical advantage to reestablish the flow in the right portal systems easily through resolution of the ligation in case of any graft complication requiring graft removal. However, limited portal hyperperfusion is crucial for hypertrophy induction.¹² This is yet not completely understood but known from transplant setting and from ALPPS procedure.¹⁹ In a few cases, 2-stage transplantations in recipients with portal hypertension were performed without any portal inflow modulation in the first stage, but an inflow increase in the graft was observed after stage 1.¹⁷ In contrast, in noncirrhotic recipients without inflow modulation no graft growth was observed.²⁰ Especially, in noncirrhotic recipients, the portal pressure should be measured routinely after reperfusion. The cutoff of the portal pressure should be <20 mm Hg. In a few cases, where the portal pressure increased considerably, we maintained low flow (subtotal ligation) to the right lobe. This is in accordance with the original RAPID protocol that prescribes portal “banding” instead of closure or alternatively hemi in cases of high portal pressure.

Arterial reconstruction is performed either using saphenous vein from the common hepatic artery in side to end fashion, or directly on the stump of the recipient’s left artery so that the bile duct and its arterial supply up to the hilum is not hampered.

The “Achilles heel” also in 2-stage transplantation is the bile duct anastomosis, especially due to the potential trauma during stage 2. In this series, bilioenteric and duct-to-duct anastomosis were performed. The complication rate is comparable.

One might be concerned about the fact that a liver lobe with tumor tissue (metastases) is left in an immunosuppressed patient. We did not find any evidence that a tumor progression is provoked because of this fact. The situation is similar to the ALPPS approach where a tumor spread is not a question of debate during the waiting time between both steps, acknowledging that ALPPS patients are not immunosuppressed. If the time interval between both steps is planned as short as possible, we do not see an exaggerated risk of tumor spread. Furthermore, the data from Grut et al²¹ showed that there is no increased growth of lung metastases in immunosuppressed patients. That supports our thesis that the effect of immunosuppression is less important during the short phase between the 2 steps. Studies on circulating tumor cells will better elucidate the pathophysiology of this intermediate phase. In the meantime, accurate patient selection, pretransplant chemotherapeutic pressure and transplant timing during a disease regression phase have a probably crucial role in reducing the risk of tumor cell spread.

In all cases, a rapid growth of the graft comparable to the in ALPPS data²² was observed allowing stage 2 in almost all cases after a median of 14 days. In 1 case, the time interval was disproportionally long (60 days). However, this was the only 1 ABO-incompatible transplantation, which mandated a delay of the second stage until the removal of the right liver lobe could be performed safely.

Feasibility of stage 2 by means of laparoscopic approach (never described before in a transplant setting) has been shown in 3 cases. The rationale for such a choice is in the will to minimize the second step related trauma favoring an early recovery and a reduced hospitalization. Full division of right portal vein during step 1, when indicated, is suggested to reduce hilar dissection during the subsequent laparoscopic intervention.

We believe that this method has the potential to be extended to other indications especially for adult recipients with chronic liver insufficiency and cirrhosis given the severe organ shortage in most countries worldwide. However, the liver function should be considered and this technique might only be feasible in Child-Pugh A cirrhosis and/or patients with a low MELD score. Another possible indication using this special kind of transplantation might be in recipients suffering from metabolic disorders.

To summarize, the short-term follow-up of the series presented here is positive. The 2-stage procedure was well tolerated from all patients with low morbidity. The measurement of graft portal vein pressure, as well as graft portal and arterial flow should be performed as routine to gain more data. However, the long-term results in the group of colorectal patients are needed to assess the value of the procedure more accurate since liver transplantation for colorectal liver metastases are not yet accepted as standard in most European countries.