Living Donor Liver Transplantation with Small Left Lobe Grafts: Prospective Validation of Utility of Splenectomy in Selected Recipients

Introduction

Living donor liver transplantation (LDLT) is now used worldwide to overcome the shortage of cadaveric donor organs and to reduce waitlist mortality by providing timely transplantation opportunities [1–3]. Despite its advantage in terms of high-quality grafts from healthy donors, the major obstacle in LDLT is the limited size of the graft that can be safely harvested from a living donor. Although the use of left lobe grafts, which are usually smaller than right lobe grafts, can enhance donor safety and expand the donor pool, right lobe grafts are still widely selected as the first choice for LDLT because of concern regarding small-for-size syndrome (SFSS), especially in Western countries [4]. By contrast, left lobe grafts have gained broader acceptance in Asian countries, where favorable graft outcomes have been achieved. Reports suggest that the graft-to-recipient weight ratio (GRWR) can be safely reduced to 0.6% with the use of portal inflow modulation (PIM) techniques [5,6]. Many PIM techniques can reportedly decrease the portal vein pressure and improve graft vascular compliance, resulting in a reduced risk of SFSS; such techniques include portosystemic shunt creation, splenic artery ligation, splenic artery embolization, and splenectomy [7–11]. Consequently, there is a growing trend toward incorporating PIM techniques in LDLT with left lobe grafts. However, there is no consensus on the necessity of or indications for PIM techniques in LDLT. Whereas some studies report excellent long-term outcomes with left lobe grafts even without PIM techniques [12,13], others highlight the risks associated with splenectomy, such as postoperative haemorrhage and life-threatening infections [14]. Thus, no consensus has been reached on their utility and indications in LDLT.

At our institution, left lobe grafts with the middle hepatic vein (H234-MHV or H1234-MHV) have been used as our first choice in adult-to-adult LDLT [15]. We recently reported our experience performing LDLT involving small liver grafts and found risk factors for SFSS and graft survival during an era when PIM techniques were not routinely performed [16]. Based on these findings, in April 2021, we began performing splenectomy with the intention of PIM in selected patients according to our risk assessment for SFSS. Thereafter, we prospectively accumulated experience performing LDLT with splenectomy as a PIM because of the lack of prospective studies evaluating the effects of splenectomy in LDLT. In the present study, we aimed to verify the effectiveness of our current protocol for splenectomy in selected recipients undergoing LDLT using left lobe grafts. Furthermore, to evaluate the extent to which splenectomy during LDLT with small grafts improves graft outcomes, the outcomes of these patients were compared with those of patients who underwent LDLT in the era when PIM was not indicated. In this study, we demonstrate that the strategic use of splenectomy based on risk assessment can mitigate the risk of SFSS and improve clinical outcomes in LDLT involving small-for-size left lobe grafts.

Material and Methods

STUDY POPULATION AND VARIABLES:

A total of 299 consecutive patients underwent LDLT from January 2005 to January 2024 at Nagasaki University Hospital. Among these, 178 adult patients who received left lobe grafts (H234-MHV or H1234-MHV) were identified from a prospectively maintained database. Five patients who underwent re-transplantation were excluded from the analysis. The study was approved by the Ethics Committee of Nagasaki University Hospital (approval No. 20012022-2) and conducted in accordance with the Declaration of Helsinki. The collected data included recipients’ sex, age, body mass index, liver diseases, Model for End-stage Liver disease (MELD) score at the time of LDLT, Child-Pugh grade, medical condition at the time of LDLT, presence of portal vein thrombosis, comorbidities, donor age, ABO incompatibility, GRWR, graft weight to standard liver volume ratio (GW/SLV) [17], and operative factors, including operative time, blood loss, blood transfusion, warm and cold ischemia times, and use of splenectomy.

GRAFT SELECTION AND ASSESSED OUTCOMES:

We selected a left lobe graft with the middle hepatic vein (H234-MHV) as our first choice when the estimated GV/SLV was >30%. If the estimated GV/SLV for an H234-MHV graft was <30%, a left lobe graft including the S1 region (H1234-MHV) was used, provided its estimated GV/SLV was >30%. Before March 2021, splenectomy during LDLT was indicated only when the platelet count at the time of transplantation was <50 000/μL, due to hypersplenism, because thrombocytopenia was a contraindication for initiating interferon and ribavirin therapy in patients with hepatitis C virus cirrhosis (previous policy). In our previous work, we found that a MELD score of ≥20 and donor age of ≥50 years were risk factors for SFSS in patients receiving small grafts with a GRWR of <0.6% [16]. Accordingly, in April 2021, we began performing splenectomy with the intention of PIM, based on graft size, donor age, MELD score, and portal venous pressure (PVP) after graft implantation, in addition to thrombocytopenia (current policy) (Figure 1). PVP measurement was performed when the GW/SLV was <40% or the GRWR was <0.8%, particularly in the presence of other risk factors, such as donor age of ≥50 years and MELD score of ≥20. GV/SLV and GRWR were calculated using the actual graft weight measured on the back table. PVP was measured after arterial reconstruction. As we described previously [18], a 6Fr intravenous catheter (HAKKO Co, Ltd, Osaka, Japan) was inserted into a mesenteric vein of the jejunum 100 cm distal to the ligament of Treitz. According to a previous report, we set the PVP cutoff value at 20 cmH2O, which is almost equal to the cutoff of 15 mmHg used as an indicator to avoid SFSS when performing splenectomy [18]. Splenectomy was performed when the PVP after arterial reconstruction exceeded 20 cmH2O.

To evaluate the utility of our current policy for splenectomy, the outcomes of recipients who underwent LDLT using left lobe grafts after April 2021 were compared with those of recipients who underwent LDLT before March 2021. Given the significant improvements in surgical outcomes at our institution after 2015 [15], the study period before March 2021 was further separated into 2 eras: era 1 (January 2005–December 2014) and era 2 (January 2015–March 2021). Era 3 was defined as the transplant era using the current policy for splenectomy after April 2021. Because previous studies have highlighted the negative impact of SFSS on short-term survival outcomes [19,20], the primary outcomes were the 1-year graft and patient survival rates in this study. The secondary outcomes were surgical outcomes, including the incidence rates of SFSS, early allograft dysfunction (EAD), and surgical complications. Biliary complications were defined as reported previously [21]. The occurrence of SFSS was determined according to previous reports [22,23]. EAD was defined as described previously [24]. Fifteen patients in era 3 were analyzed using a descriptive method. The final follow-up date was June 10, 2024. Furthermore, we performed subgroup analyses on high-risk patients receiving small-for-size grafts (GW/SLV of <40% and/or GRWR of <0.8%) and presenting additional risk factors, such as donor age of ≥50 years or a MELD score of ≥20.

STATISTICAL ANALYSIS:

Differences in categorical variables were evaluated using the Pearson χ² test. When any expected value was ≤5, the Fisher exact probability test was used as an alternative. Graft and patient survival rates were estimated using the Kaplan-Meier method and a log-rank test. All statistical analyses were performed using JMP Pro version 16 software (SAS Institute Inc, Cary, NC, USA). Values of P<0.05 were considered statistically significant.

Results

PATIENTS’ CHARACTERISTICS ACROSS ERAS:

In total, 173 adult-to-adult LDLTs using left lobe grafts were included in the analyses. The rates of left lobe graft use for adult-to-adult LDLT were 62.5% (100/160) in era 1, 62.4% (58/93) in era 2, and 32.6% (15/46) in era 3. Patients’ characteristics according to eras are detailed in Table 1. With regard to etiology, the rates of hepatitis B and C were significantly lower and the rates of alcohol- or metabolic dysfunction-associated steatohepatitis were higher in era 3. The rate of ABO incompatibility was higher in era 3. Although splenectomy was performed more frequently in era 3, intraoperative blood loss and transfusion of red blood cells were comparable between the eras.

COMPARISONS OF SURVIVAL OUTCOMES ACROSS ERAS:

The 1-year graft survival rates in era 3 tended to be superior to those in eras 1 and 2 (P=0.057 and P=0.143, respectively) (Figure 2A). Likewise, the 1-year patient survival rates in era 3 tended to be superior to those in eras 1 and 2 (P=0.063 and P=0.173, respectively; Figure 2B).

INCIDENCE RATES OF EAD, SFSS, AND SURGICAL COMPLICATIONS ACROSS ERAS:

The proportion of high-risk patients was highest in era 3 (53.3%) and lowest in era 2 (29.3%) (Table 2). The incidence rates of EAD in eras 1, 2, and 3 were 54.0%, 13.8%, and 13.3%, respectively. The incidence rate of EAD with the previous policy (eras 1 and 2) was 39.2%, whereas that in the era with the current policy (era 3) was 13.3% (P=0.047). The incidence rates of SFSS in eras 1, 2, and 3 were 41.0%, 27.6%, and 20.0%, respectively. The incidence rate of SFSS with the previous policy was 36.1%, whereas that with the current policy was 20.0% (P=0.211). These findings suggest that the current splenectomy policy can reduce the risk of EAD in LDLT using left lobe grafts. Surgical complication rates, including hepatic artery thrombosis, portal vein thrombosis, and biliary complications, were comparable between the current and previous policy eras (Table 2).

SUBGROUP ANALYSES OF HIGH-RISK RECIPIENTS:

In the subgroup analyses of high-risk recipients, the rates of splenectomy use were higher in era 3 (75.0%) than in era 1 (51.2%) and era 2 (35.3%) (Table 3). The incidence of EAD in the current policy group (era 3) tended to be lower than that in the previous policy group (25.0% vs 60.3%, P=0.125). Similarly, the incidence of SFSS in the current policy group also showed a trend toward being lower than in the previous policy group (25.0% vs 55.2%, P=0.143). The survival outcomes for high-risk recipients across eras are presented in Figure 3. Survival outcomes in era 2 were similar to those in era 1. However, graft and patient survival rates in era 3 tended to be better than those in eras 1 and 2.

DETAILED OUTCOMES OF PATIENTS RECEIVING LEFT LOBE GRAFTS IN ERA 3:

The outcomes of 15 patients who underwent LDLT using left lobe grafts in era 3 are summarized in Table 4. The median age of the recipients and donors were 52 (37–68) years and 40 (23–60) years, respectively. The median MELD score was 17 (11–35). The median GRWR was 0.65 (0.47–1.11), and the median GW/SLV was 33.1 (27.3–48.0). Splenectomy was performed in 9 (60%) patients, and the indications for splenectomy were PVP of >20 cmH2O in 4 patients and a platelet count of <50 000/μL in 5 patients. No patients developed splenectomy-associated complications. Three patients developed SFSS and 2 patients developed EAD. At a median follow-up of 20.5 months after LDLT, all patients were alive and doing well.

Discussion

In this study, we evaluated patient’s clinical outcomes after we had prospectively accumulated experience with the current policy for performing splenectomy to avoid SFSS-related graft failure in selected patients, according to our risk assessment. The findings indicated that the current policy era was associated with improved graft and patient survival outcomes, compared with the previous policy eras. Notably, the incidence of EAD was significantly lower under the current policy. Subgroup analyses of high-risk patients further supported these findings, demonstrating better survival outcomes in the current policy era. Despite the aggressive use of small left lobe grafts, with a median GRWR of 0.65%, all patients survived without requiring re-transplantation in the era with the current policy. These findings suggest that our current splenectomy strategies might be useful to minimize the risk of graft loss due to SFSS.

In some transplant centers, splenectomy during LDLT is contraindicated because of concern regarding an increased risk of intraoperative massive bleeding, post-transplant infection, and thrombotic complications [25–27]. Therefore, although no patients in our cohort experienced graft loss due to splenectomy-related complications, the surgical risk of splenectomy should not be underestimated. Consistent with previous reports [25–27], longer operation times and a higher incidence of portal vein thrombosis were observed in LDLT recipients who underwent splenectomy than in those who did not (Table 5). To avoid unnecessary splenectomy during LDLT, we believe that the decision to perform splenectomy should be based on appropriate risk assessments. Of 15 patients in the era with the current policy, 5 patients did not undergo splenectomy, because of the lack of risk factors (except for graft size). Notably, of those 5 patients, 4 (80%) did not develop SFSS even with their limited graft size. Therefore, the indications for splenectomy should not be determined based on graft size alone.

In most transplant centers, splenectomy during LDLT is indicated based on graft size, disease severity, and portal hypertension, including the presence of splenomegaly, oesophagogastric varices, portosystemic shunts, graft quality, and graft hemodynamics [9,28–31]. According to the existing literature, PIM techniques are considered in cases of portal hypertension (PVP of >15–20 mm Hg and/or portal vein flow (PVF) of >300 mL/100 g of graft weight), severe decompensation due to liver disease (MELD score of >20), small-for-size grafts (GRWR of <0.7–0.8), and advanced donor age (>45–50 years) [32]. However, because of patient and graft hemodynamic heterogeneity and the multifactorial nature of SFSS, the application of these indications varies widely among transplant centers, and strong recommendations for PIM indications are lacking in the current literature. Because only retrospective studies evaluating the beneficial effects of splenectomy in LDLT exist in the literature, the findings of this study that strictly adhered to the indications for splenectomy based on the MELD score, donor age, and PVP might be helpful for recommending splenectomy in highly selected patients.

Graft hemodynamics are assessed using various methods at different centers. The volume of PVF can be measured by direct puncture of the portal vein or using intraoperative Doppler ultrasonography with probes specifically designed for the vessel size [28,33], and a significant reduction in PVF following splenectomy has been reported [19]. At our center, PVP is used as a surrogate marker for portal hyperperfusion, as supported by previous studies [30,34]. Our technique, which involves catheter insertion through the jejunal vein, is technically straightforward and does not require specialized materials. Prior research has shown that a significant reduction in PVP after splenectomy can predict graft outcomes [35,36]. Therefore, PVP is considered a reliable indicator of portal hyperperfusion, comparable to PVF.

Some transplant centers perform preperfusion splenectomy to prevent graft damage caused by initial hyperperfusion [9,31]. These centers base their decisions to perform splenectomy on pretransplant factors, such as graft size, donor age, MELD score, and the degree of portal hypertension, rather than on graft hemodynamics. By contrast, we consider PVP after reperfusion to be a critical factor in deciding whether to perform PIM techniques. Consequently, splenectomy at our center is performed after reperfusion. Notably, 2 patients (cases 2 and 11 in Table 3) did not require splenectomy, because their PVP after reperfusion was <20 cmH2O, and they subsequently developed neither EAD nor SFSS. Recent studies have identified clinical indicators that may predict post-reperfusion graft hemodynamics [28,37]. Prospective studies evaluating the utility of preperfusion splenectomy based on these indicators are warranted in the future.

It has been reported that splenic artery ligation can decrease the portal inflow volume in liver transplantation [8,33] and this was confirmed in the LDLT setting [38]. In 1 patient (case 12), splenic artery ligation was chosen as an alternative to splenectomy because of concerns about massive bleeding from collateral vessels around the spleen. In this patient, because SFSS occurred and refractory ascites continued for 3 months, splenectomy was performed after transplantation. Thereafter, ascites and graft dysfunction were promptly resolved, and the patient was discharged. A group from India reported that while splenic artery ligation was effective in preventing SFSS in recipients with mildly elevated PVP, they preferred to create a hemiportocaval shunt for recipients with higher PVP [39]. Furthermore, the most recent randomized trial to date showed that although splenic artery ligation significantly reduces both portal inflow and PVP, it cannot decrease the incidence rates of early graft dysfunction in adult LDLT [40]. Therefore, splenic artery ligation alone might not be sufficient to modulate portal hemodynamics and avoid graft damage from excessive portal inflow in LDLT recipients with severe portal hypertension.

We should acknowledge the 3 main limitations of this study. First, considering the small number of patients included in the study, it is difficult to generalize our findings to other populations. Second, although we prospectively accumulated experience with splenectomy, the comparative analyses were conducted retrospectively with a historical cohort. Therefore, randomized prospective studies are needed to mitigate selection bias and validate our conclusions. A third limitation is that, because this retrospective study included LDLT recipients over several decades, the improved survival outcomes observed in the recent era 3 might partially reflect advancements in surgical techniques, perioperative management, and immunosuppressive therapies. Interestingly, in the subgroup analyses of high-risk patients, no significant improvement in survival rates was observed between eras 1 and 2, whereas survival outcomes in era 3 were better than those in eras 1 and 2. This finding suggests that the improved outcomes in era 3 were primarily attributable to the policy change implementing splenectomy as a PIM technique for high-risk patients.

Conclusions

This study demonstrated favorable survival outcomes in LDLT involving left lobe grafts when splenectomy was performed based on identified risk factors for SFSS. These findings suggest the utility of performing splenectomy based on risk assessments, including disease severity, donor age, and portal hypertension in LDLT using left lobe grafts. Given the limited sample size and the retrospective nature of this study, further prospective studies are warranted to validate these findings.