Living donor liver transplantation (LDLT) is widely recognized as a viable strategy to address the shortage of organs from deceased donors globally [1]. Ensuring the safety of both donors and recipients is paramount in planning LDLT, with graft size being a crucial determinant of success for both parties involved. Utilizing a small graft in LDLT elevates the risk of small-for-size syndrome (SFSS), making it imperative to secure a graft volume that fulfills the recipient’s metabolic needs without compromising donor safety [2].

Historically, a graft-to-recipient weight ratio (GRWR) of 0.8 has been the benchmark for safe LDLT [3]. However, numerous Asian and Western centers have shared their experiences with grafts having GRWRs less than 0.8, revealing that favorable outcomes can be achieved with GRWRs ranging from 0.6 to 0.8 through meticulous recipient and donor selection and effective portal flow modulation (PFM) [4-6]. Recently, Reddy et al. [7] reported a 1-year graft survival rate of 65.6% even in LDLT with GRWR less than 0.6 in an international multicenter study.

Despite these insights, the evidence supporting the safety of GRWR<0.8 grafts is primarily derived from single-center studies, which often suffer from small sample sizes and selection biases, such as lower model for end-stage liver disease (MELD) scores in the GRWR<0.8 group. Moreover, risk factors for graft loss using GRWR<0.8 grafts, in comparison to larger grafts, remain ambiguous. Recently, large volume studies in Korea showed that recipient age older than 60 years, pretransplant MELD score greater than 15, and male donor were risk factors for GRWR<0.8 compared to GRWR≥0.8 [8]. This study aimed to validate those risk factors in patients who underwent LDLT using GRWR<0.8 within a single center cohort.

Study Materials

We performed a retrospective analysis of 56 adult LDLT conducted at Severance Hospital, Korea, from July 2007 to December 2022. All available information regarding recipients, donors, and LDLT surgeries was obtained from the institutional LT data base. Underlying liver diseases were classified as viral, alcoholic, or other. Graft types were categorized as right, left, and posterior lobe. Graft steatosis was assessed during donor surgery and categorized into two groups (>10% or ≤10%) based on pathological examination. The GRWR was calculated using the formula: GRWR=(graft weight [g]/recipient weight [g])×100, with graft weight measured immediately before implantation.

The study was conducted in accordance with the Declaration of Helsinki, as revised in 2013. The study was approved by the Institutional Review Board of Severance Hospital (4-2024-0315), and individual consent for this retrospective analysis was waived because of its retrospective design.

Definition and Outcomes

Patients were classified according to the number of risk factors for GRWR<0.8, as identified in a previous Korean multicenter study [8]. The risk factors were recipient age ≥60 years, MELD score ≥15, and male donor. Based on these, patients were divided into three groups: those with no risk factors (Risk 0 group), those with one risk factor (Risk 1 group), and those with two or more risk factors (Risk ≥2 group). These groups were then compared in terms of baseline characteristics and graft survival. The primary outcome of the study was graft loss, which was defined as either the need for retransplantation or patient death. Patients were followed until graft loss, up to five years post-transplantation, or until December 2022, whichever came first.

Statistical Methods

Data were presented as mean±standard deviation, median (interquartile range [IQR]) for continuous variables and as number (proportion) for categorical variables, where appropriate. Comparisons of continuous variables were conducted using Student’s t-test, while categorical variables were compared using the chi-square test, as appropriate. Graft survival among the risk factor groups was evaluated using Kaplan–Meier curves and the log-rank test. Due to the single-center nature of this study, which resulted in a smaller sample size and event number for each group, only unadjusted analyses were applied. All statistical analyses were performed using the R statistical package, version 4.4.1 for macOS (http://cran.r-project.org/), with a significance threshold set at p<0.05.

Distribution of GRWR

Among 56 LDLT recipients who used GRWR<0.8 grafts, distribution of GRWR was depicted in Fig. 1. Median GRWR was 0.74 with IQR 0.69–0.78, and minimum value was 0.49. Four was GRWR<0.6 (7.1%), 11 was GRWR 0.6–0.7 (19.6%) and 41 was GRWR 0.7–0.8 (73.2%).

Figure 1. Distribution of graft-recipient weight ratio (GRWR).

Baseline Characteristics

As shown in Table 1, the average age was 53.6±8.4 years. Most recipients were male (82.1%). The average body mass index (BMI) was 26.1±3.5. Hypertension was present in 16 patients (28.6%), diabetes mellitus in 23 patients (41.1%), and cardiovascular disease in 4 patients (7.1%). Regarding underlying diseases, 67.9% of the patients had viral-related liver disease, 17.9% had alcoholic liver disease, and 14.3% had other causes. Hepatocellular carcinoma was present in 33 patients (58.9%). The median pretransplant MELD score was 12 (IQR 9–16). The median operation time was 609 minutes (IQR 501–700), and the median cold ischemic time was 120 minutes (IQR 101–147). The median number of transfused red blood cell packs was 5 (IQR 2–8). PFM was performed in 16 patients (28.6%). Donor characteristics showed a median age of 35 years (IQR 24–44), with 50.0% of donors being male. The average donor BMI was 22.6±3.2. Regarding graft type, 36 patients (64.3%) received right lobe grafts, 14 patients (25.0%) received left lobe grafts, and 6 patients (10.7%) received right posterior lobe grafts. ABO incompatibility was noted in 16 patients (28.6%). Macrovesicular steatosis ≥10% was observed in 7 patients (12.5%).

Table 1 . Baseline characteristics of LDLT recipients received GRWR<0.8 graft.

Variable	LDLT recipient (n=56)
GRWR
GRWR<0.6	4 (7.1)
GRWR 0.6–0.7	11 (19.6)
GRWR 0.7–0.8	41 (73.2)
Age (yr)	53.6±8.4
Male sex	46 (82.1)
BMI	26.1±3.5
Hypertension	16 (28.6)
Diabetes mellitus	23 (41.1)
Cardiovascular disease	4 (7.1)
Underlying group
Viral	38 (67.9)
Alcoholic	10 (17.9)
Others	8 (14.3)
Hepatocellular carcinoma	33 (58.9)
Pretransplant MELD	12 (9–16)
Operation time (min)	609 (501–700)
Cold ischemic time (min)	120 (101–147)
Transfusion RBC (pack)	5 (2–8)
Portal flow modulation	16 (28.6)
Donor age (yr)	35 (24–44)
Male donor	28 (50.0)
Donor BMI	22.6±3.2
Graft type
Right	36 (64.3)
Left	14 (25.0)
Right posterior	6 (10.7)
ABO incompatibility	16 (28.6)
Macrovesicular steatosis ≥10%	7 (12.5)

Values are presented as number (%), mean±standard deviation, or median (interquartile range)..

LDLT, living donor liver transplantation; GRWR, graft-recipient weight ratio; BMI, body mass index; MELD, model for end-stage liver disease; RBC, red blood cell..

When categorized into three groups according to the number of risk factors, the proportion of male patients decreased as the number of risk factors increased (p=0.007, Table 2). Donor BMI tended to be higher in higher risk factor groups (p<0.001). For graft types, almost all patients (94.4%) in the Risk 0 group used the right lobe, except for one who used the left lobe. In the Risk 1 group, 70.0% used the right lobe, 15.0% used the left lobe, and 15.0% used the right posterior lobe. In the Risk ≥2 group, 61.1% used the right lobe, while 11.1% used the left lobe, and 27.8% used the right posterior lobe.

Table 2 . Baseline characteristics of patients, according to the number of risk factors for GRWR<0.8.

Variable	Risk 0 (n=18)	Risk 1 (n=20)	Risk≥2 (n=18)	p-value
GRWR				0.487
GRWR<0.6	0 (0)	2 (10.0)	2 (11.1)
GRWR 0.6–0.7	3 (16.7)	3 (15.0)	5 (27.8)
GRWR 0.7–0.8	15 (83.3)	15 (75.0)	11 (61.1)
Age (yr)	52 (49–55)	50 (49–54)	61 (52–64)	0.007
Male sex	17 (94.4)	18 (90.0)	11 (61.1)	0.017
BMI	27.1±3.4	25.6±3.1	25.6±4.1	0.341
Hypertension	4 (22.2)	7 (35.0)	5 (27.8)	0.682
Diabetes mellitus	10 (55.6)	7 (35.0)	6 (33.3)	0.315
Cardiovascular disease	0 (0)	3 (15.0)	1 (5.6)	0.191
Underlying group				0.821
Viral	13 (72.2)	12 (60.0)	13 (72.2)
Alcoholic	2 (11.1)	5 (25.0)	3 (16.7)
Others	3 (16.7)	3 (15.0)	2 (11.1)
Hepatocellular carcinoma	14 (77.8)	9 (45.0)	10 (55.6)	0.115
Pretransplant MELD	10 (8–11)	12 (8–15)	17 (13–27)	<0.001
Operation time (min)	614.1±122.2	589.3±103.1	646.1±195.8	0.486
Cold ischemic time (min)	127 (114–144)	114 (101–150)	117 (95–131)	0.635
Transfusion RBC (pack)	4 (0–7)	3 (2–8)	8 (5–11)	0.092
Portal flow modulation	2 (11.1)	8 (40.0)	6 (33.3)	0.124
Donor age (yr)	32.4±11.3	33.5±12.2	39.1±10.9	0.190
Male donor	0 (0)	11 (55.0)	17 (94.4)	<0.001
Donor BMI	20.4±2.6	22.6±2.2	24.7±3.3	<0.001
Graft type				<0.001
Right	17 (94.4)	14 (70.0)	11 (61.1)
Left	1 (5.6)	3 (15.0)	2 (11.1)
Right posterior	0 (0)	3 (15.0)	5 (27.8)
ABO incompatibility	3 (16.7)	9 (45.0)	4 (22.2)	0.119
Macrovesicular steatosis ≥10%	1 (5.6)	2 (10.0)	4 (22.2)	0.220

Values are presented as number (%), median (interquartile range), or mean±standard deviation..

GRWR, graft-recipient weight ratio; BMI, body mass index; MELD, model for end-stage liver disease; RBC, red blood cell..

Graft Survival Comparison according to the Number of Risk Factors for GRWR<0.8

During the follow-up period, graft survival significantly decreased as the number of risk factors increased (p=0.015, Fig. 2). The Risk 0 group showed 100% graft survival during the 5 years after LT, whereas graft survival was 80.0% at 1 year and 72.7% at 5 years in the Risk 1 group, and 77.8% at 1 year and 54.5% at 5 years in the Risk ≥2 group. When comparing subgroups according to graft type, the 5-year graft survival was 100%, 71.4%, and 75.0% for the Risk 0, Risk 1, and Risk ≥2 groups in patients who received right lobe grafts (p=0.065, Fig. 3). In patients who received grafts other than the right lobe (left and right posterior lobe), the 5-year graft survival was 100%, 66.7%, and 44.5% (p=0.42), respectively.

Figure 2. Graft survival according to the number of risk factors for graft-recipient weight ratio<0.8. LT, liver transplantation.

Figure 3. Graft survival according to the number of risk factors and graft types. (A) Right lobe graft, (B) other than right lobe graft. LT, liver transplantation.

When comparing donor age with a cutoff set at 45 years (about 75 percentile value), graft survival was not significantly different between the donor age ≥45 and <45 groups (78.9% vs. 69.2%, p=0.25, Fig. 4).

Figure 4. Graft survival according to donor age. LT, liver transplantation.

Previous single-center studies have reported that GRWR<0.8 grafts can be safely used in LDLT with proper PFM and appropriate donor-recipient selection [4-6,9]. However, a recent Korean multicenter study found a significant but small decrease in survival for GRWR<0.8 compared to larger grafts, suggesting three risk factors such as age≥60, MELD≥15, and male donor [8]. Current study demonstrated that outcomes of LDLT using GRWR<0.8 grafts were well stratified according to the number of risk factors identified in previous multicenter studies. Although the statistical significance was limited due to the sample size, similar trends were observed in the subgroups based on graft type. Therefore, it is crucial to accurately evaluate the risk factors associated with using GRWR<0.8, consider the patient's condition and expected survival, assess the availability of other donors, and take into account the regional deceased donor pool when deciding on LDLT.

PFM has been shown to effectively prevent SFSS by reducing portal inflow and pressure when using small grafts [10]. However, a recent international multicenter study by Reddy et al. [7] reported that PFM or portal hemodynamics are not independent predictors of SFSS when using GRWR<0.6 grafts in LDLT. Instead, the study highlighted that recipient factors, such as MELD score and pretransplant hospital status, are more critical in predicting SFSS when using small grafts. Although the KOTRY study could not demonstrate this due to the lack of PFM data, it did support Reddy et al.’s [7] findings by showing that two of the independent risk factors for GRWR<0.8 are related to the recipient. In our GRWR<0.8 cohort, only 16 out of 56 patients received PFM because a definite indication for PFM has not been established even in recent periods, limiting the evaluation of the effect of PFM.

It is well-documented that advanced donor age serves as a significant risk factor for graft loss following LDLT [11]. Additionally, multiple studies have indicated that an older donor age poses a risk particularly when utilizing small-sized grafts [12,13]. However, these earlier studies relied on data from single centers and did not compare grafts with a GRWR of less than 0.8 to those with a GRWR of 0.8 or greater. Previous KOTRY study showed donor age was not specific risk factor for GRWR<0.8 graft compared to larger graft in LDLT, although the sample size of recipients who received grafts from donors older than 60 years was limited [8]. The current study also indicated no significant interaction between graft size and donor age. Further research is required to evaluate graft loss in GRWR<0.8 grafts from much older living donors, such as those in their 70s.

Despite the small sample size, this study reaffirmed that recipient age ≥60, MELD≥15, and male donor were important risk factors for using GRWR<0.8 grafts in LDLT, as shown in previous multicenter studies. Identifying risks associated with small-sized grafts and adequately managing potential complications is essential for safe LDLT.

There was no funding related to this study.

Deok-Gie Kim is an editorial member of the journal but was not involved in the review process of this manuscript. Any other authors have no conflict of interest.

Conceptualization: YJY, DGK. Data curation: YJY, EKM, DGK. Formal analysis: YJY, DGK. Investigation: EKM, DGK, MK, HHK, JGL, DJJ, MSK. Methodology: YJY, DGK. Project administration: DGK. Resources: DGK. Software: YJY. Supervision: DGK. Validation: DGK. Visualization: DGK. Writing – original draft: YJY, DGK. Writing – review & editing: DGK.