Evaluation of human leukocyte antigen histocompatibility and its impact on liver transplant graft outcomes

Abstract

Background:

The impact of human leukocyte antigen (HLA) histocompatibility on graft survival is well established in kidney and haematopoietic stem cell transplantation; however, its role in liver transplantation (LT) remains uncertain. Prior studies have reported inconsistent associations between HLA mismatching and LT graft outcomes. Given emerging evidence linking HLA mismatch to post-transplant complications, this relationship warrants reassessment.

Objectives:

To investigate the effect of HLA Class I (A, B, C) and Class II (DRB1, DQA, DQB) mismatching on LT graft survival in a single-centre retrospective analysis.

Design:

A retrospective, single-centre cohort study.

Methods:

610 patients who underwent their first deceased donor LT between January 2010 and March 2021 were included. The primary outcome was graft failure. Donor and recipient HLA typing was available at HLA-A, HLA-B and HLA-DRB1 loci in 610 patients, HLA-C in 38 patients, HLA-DQA in 63 patients and HLA-DQB in 63 patients. Graft survival was assessed using Kaplan–Meier analysis and predictors identified using Cox regression.

Results:

HLA-C mismatching was associated with reduced graft survival on Kaplan–Meier analysis (p = 0.02); however, this was based on a small subgroup (n = 38), and the univariate Cox estimate was imprecise (HR 9.24, 95% CI 1.01–83.48, p = 0.047). HLA-DRB1 mismatching was associated with improved graft survival (p = 0.047) and reduced graft failure risk on multivariate Cox regression (aHR 0.34, 95% CI 0.14–0.87, p = 0.02), although limited by the small number of matched transplants (n = 9). There was no association between mismatches at other loci and graft survival. Hepatic artery thrombosis worsened graft survival (aHR 3.42, 95% CI 2.44–4.80, p < 0.001) but was not associated with HLA mismatch.

Conclusion:

These results suggest that the effect of HLA histocompatibility in LT may be locus-specific. Associations were observed for differential graft outcomes at the HLA-C and HLA-DRB1 loci; however, given the small sample sizes and imprecision of estimates, these findings should be considered hypothesis-generating and require validation in larger prospective cohorts.

Keywords

human leukocyte antigen histocompatibility liver allograft outcomes liver transplantation

Introduction

The importance of human leucocyte antigen (HLA) histocompatibility in enhancing graft survival is well established in certain organ transplantations, particularly in kidney and haematopoietic stem cell transplantation.^1–3 While HLA matching is not routinely performed for liver transplantation (LT), short- and long-term graft and patient survival remain outstanding.

Previous studies have demonstrated conflicting associations between HLA mismatching and graft survival in LT, including negative,^4–7 positive^8–14 or no^15–21 correlation. Some authors have proposed a ‘dualistic role’ of HLA, whereby HLA matching simultaneously decreases rejection-related graft failure while augmenting other immunological processes that may contribute to graft failure, such as cytomegalovirus infection,¹² recurrent autoimmune liver disease²² and viral hepatitis.^13,23 Nevertheless, the relationship between HLA histocompatibility and LT graft outcomes remains unclear.

Reassessment of the role of HLA histocompatibility in LT is necessary. A substantial portion of the existing literature has become outdated, and the role of HLA mismatching on alloimmune responses can now be examined with greater precision compared to previous studies, due to the development of high-resolution HLA typing technologies. Antibody-mediated rejection has emerged as an important cause of allograft loss in LT, with recent studies demonstrating an association between HLA mismatching and the development of donor-specific antibodies.^24,25 Furthermore, Bricogne et al.⁴ recently found an association between HLA-A mismatching and reduced liver graft survival, primarily attributed to early graft loss resulting from hepatic artery thrombosis (HAT) and sepsis. Thus, HLA mismatching may play an important role in early post-transplant complications and graft survival.

In light of these findings, we conducted a single-centre retrospective analysis to investigate the effect of HLA Class I (A, B, C) and Class II (DRB1, DQA, DQB) mismatching on graft survival among a cohort of deceased donor LT recipients.

Materials and methods

Patient selection

Consecutive adult (age ⩾ 18 years) patients who underwent their first deceased donor LT at the Victorian Liver Transplant Unit in Melbourne, Australia, between January 2010 and March 2021, were screened for inclusion. Exclusion criteria included re-transplantation, multiple organ transplantation and patients with missing donor and recipient HLA typing data. A flow diagram outlining patient selection is provided in Supplemental Figure 1. Patient demographic and clinical data was obtained from the Victorian Liver Transplant Unit electronic database, which is prospectively completed from the time of LT assessment and maintained post-operatively as part of routine care. For this study, patient data were collected until time of death, loss of follow-up or until the study censor date of April 1st, 2023. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.²⁶

HLA typing

Donor and recipient HLA Class I and II typing is routinely conducted by Australian Red Cross Lifeblood. A variety of HLA typing technologies were utilised over the LT period in this study, consistent with the clinical standard of care at the time. These were; inhouse sequence-specific oligonucleotide (SSO) and sequence-based typing (SBT) in 2010, inhouse SBT and LABType™ Luminex^® SSO (One Lambda, Thermofisher) between 2011 and 2013, Connexio SBT and LABType™ Luminex^® SSO (One Lambda, Thermofisher) between 2014 and 2016, LABType Luminex^® SSO (One Lambda, Thermofisher) between 2017 and 2018, and next generation sequencing using MiaFora™ (Immucor, Werfen), AllType (One Lambda, Thermofisher) and AllType FASTplex (One Lambda, Thermofisher) kits from 2019 to present day.

We sought to obtain donor and recipient HLA typing data across the HLA-A, -B, -C, -DRB1, -DQA and -DQB loci. The HLA-DRB1 gene encodes the beta chain of the polymorphic component of the HLA-DR heterodimer protein and is used in serotyping due to there being minimal variation in the DRA1 gene. The HLA-DQA and -DQB genes encode the two subunits of the HLA-DQ heterodimer protein. Prior to April 2022, HLA typing at all loci was performed for donors, however only HLA-A, -B and -DRB1 typing was performed for recipients. Only data containing both donor and recipient HLA typing were included to enable mismatching analysis. A distinction between missing data and allele non-expression was made. Donor and recipient HLA mismatches were manually identified and quantified at each locus at the allele level. Donor and recipient HLA typing was available at HLA-A, -B and -DRB1 loci in 610 patients, at HLA-C loci in 38 patients, at HLA-DQA loci in 63 patients and at HLA-DQB loci in 63 patients.

Clinical variables

The primary outcome was graft failure. The cause for graft failure was categorised as allograft loss due to patient death, HAT, primary non-function, non-anastomotic biliary strictures, treated biopsy-proven rejection, primary disease recurrence or other causes. Cases of graft failure categorised as secondary to non-anastomotic biliary strictures did not have preceding HAT as a risk factor. Both acute (diagnosed <30 days post-transplantation) and delayed (diagnosed >30 days post-transplantation) HAT were included for analysis.

Additional donor and recipient characteristics were obtained to identify independent predictors of graft failure. Recipient characteristics included age, gender, smoking status, primary indication for LT, hepatocellular carcinoma (HCC) on explant, cholangiocarcinoma on explant, Model for End-Stage Liver Disease Sodium score (MELD-Na) at time of transplantation and ABO blood group. The primary indication for LT was categorised as alcohol, metabolic-dysfunction associated steatotic liver disease (MASLD), viral hepatitis, immune-mediated and cholestatic liver disease (autoimmune hepatitis, primary sclerosing cholangitis, primary biliary cirrhosis) and other (which included HCC as a primary indication for transplantation but not HCC found incidentally on explant). Donor characteristics included age, gender, status (brainstem or circulatory death), cold ischaemic time, Eurotransplant donor risk index and ABO blood group.

Statistical analyses

Categorical variables were presented as frequencies and proportions, whereas continuous variables were presented as median and interquartile range (IQR). Differences between groups were assessed utilising Wilcoxon rank sum analysis for continuous variables and Fisher’s exact test for categorical variables, because of the uneven sample sizes and small counts within specific groups. Graft survival was determined using Kaplan–Meier survival analysis, with the impact of HLA mismatch on graft survival evaluated via log-rank analysis. The thresholds for stratifying broad HLA loci mismatching were determined based on a method that approximated the division of data into tertiles. This approach was chosen to ensure meaningful differentiation of data between groups, as the use of exact tertiles or quartiles resulted in overlapping values.

Predictors for graft survival and HAT were assessed using Cox regression analysis. Prespecified covariates based on clinical relevance were included in univariate analysis: recipient age, sex, smoking status, primary indication for LT, hepatocellular carcinoma on explant, cholangiocarcinoma on explant, MELD-Na score at time of transplantation, donor sex, donor age, donor status (brainstem vs circulatory death), cold ischaemic time, Eurotransplant donor risk index, ABO blood group compatibility, HAT, and HLA mismatch at each individual and broad locus group. Variables achieving statistical significance (p ⩽ 0.10) in univariate analysis were included in subsequent multivariate analysis using forward selection. Covariates in multivariate analysis were statistically significant if p value <0.05. As HLA-C typing data was available in only 38 patients (6%), the high proportion of missing data precluded its inclusion in the primary multivariate Cox regression model. Inclusion of HLA-C mismatch in the multivariate model was explored but produced unstable estimates, including infinite coefficients and failure of model convergence. HLA-C mismatch was therefore assessed using univariate Cox regression in the cohort of patients with available HLA-C typing data only. Univariate Cox regression analysis of HAT did not include cholangiocarcinoma and individual HLA loci mismatch variables, because the low number of HAT events caused separation, resulting in infinite coefficients. R Studio version 2023 was used for all analyses, and a p value <0.05 was considered statistically significant.

Results

A total of 610 LT recipients with a median follow-up of 6.72 years (IQR 4.44–9.47) were included in the final analysis. Baseline recipient and donor characteristics of the cohort are summarised in Table 1.

Table 1.

Baseline demographics.

Characteristic	Value
Total, n	610
Follow-up, years	6.72 (4.44–9.47)
Male sex	411 (67)
Age, years	57 (48–62)
Previous smoking	342 (56)
Primary indication for liver transplantation
Alcohol	77 (13)
Metabolic-associated steatotic liver disease	45 (7)
Viral hepatitis	110 (18)
Immune-mediated and cholestatic liver disease	124 (20)
Other	254 (42)
Hepatocellular carcinoma^a	265 (43)
Cholangiocarcinoma^b	13 (2)
MELD-Na	22 (14–29)
Donor characteristics
Male sex	342 (57)
Age, years	48 (33–59)
Donor status
Brainstem death	565 (93)
Circulatory death^c	45 (7)
Cold ischaemic time, minutes^d	365 (302–440)
Eurotransplant donor risk index	1.37 (0.97–1.54)

Categorical variables presented as n (%) and continuous variables presented as median (IQR).

Hepatocellular carcinoma on explant.

Cholangiocarcinoma on explant.

5 (11%) circulatory death donors underwent normothermic machine perfusion with the OrganOx metra^® device.

n = 609.

MELD-Na, model for end-stage liver disease sodium.

Overall graft and patient survival

The probability of graft survival in this cohort was 90% at 1 year and 70% at 10 years following LT (Supplemental Figure 2(A)). Graft failure occurred in 145/610 (24%) patients, secondary to patient death (111/145, 76%), HAT (12/145, 8%), primary non-function (10/145, 7%), disease recurrence (5/145, 3%), non-anastomotic biliary strictures (4/145, 3%), rejection (1/145, 1%), post-transplant lymphoproliferative disorder (1/145, 1%) and donor-derived neuroendocrine tumour involving the graft (1/145, 1%). In patients who developed graft failure, the median time to allograft loss was 14 months (IQR 2–48).

The probability of patient survival in this cohort was 94% at 1 year and 75% at 10 years following LT (Supplemental Figure 2(B)). A total of 111/610 (18%) patients died during the follow-up period. Causes for patient death included de novo non-HCC malignancy (27/111, 24%), recurrent HCC (26/111, 23%), sepsis (18/111, 16%), atherosclerotic cardiovascular disease (6/111, 5%) and other (24/111, 31%).

Broad HLA loci mismatch and graft survival

Graft survival was initially evaluated based on mismatching across broad categories of HLA loci. There was no significant difference in graft survival over time observed with mismatches across all HLA loci (HLA-A, -B, -C, -DRB1, -DQA and -DQB), HLA Class I loci (HLA-A, -B, -C), HLA Class II loci (HLA-DRB1, -DQA, -DQB) and a combination of HLA-A, -B and -DRB1 loci (Supplemental Figure 3(A)–(D)).

Specific HLA loci mismatch and graft survival

The impact of mismatches at individual HLA loci on graft survival using time-to-event analyses was subsequently performed.

The occurrence of at least one mismatch at the HLA-C locus was associated with reduced graft survival compared to transplants fully matched at HLA-C (p = 0.02; Figure 1). However, this observation was based on a small subgroup of recipients with available HLA-C typing data (n = 38), with a limited number of events. Seven cases of graft failure occurred in 25 HLA-C mismatched transplants and were secondary to non-anastomotic biliary strictures (2), HAT (1), primary non-function (1), primary disease recurrence (1), chronic rejection (1) and patient death due to subdural haemorrhage (1). Patients with HLA-C mismatch notably had a shorter duration of follow-up compared to those with HLA-C matching (3 years vs 8 years, p = 0.00; Table 2). Apart from a lower proportion of males in the HLA-C mismatch group (24% vs 69%, p = 0.01), there were no differences in recipient or donor characteristics between HLA-C mismatched and fully matched transplants (Table 2).

Figure 1.

Kaplan–Meier plot of HLA-C mismatch transplants and graft survival.

Table 2.

Donor and recipient characteristics based on HLA-C mismatch status.

Characteristic	C match	C mismatch	p Value
Total, n	13	25
Follow-up, years	8 (8–8)	3 (3–4)	0.00*
Male sex	9 (69)	6 (24)	0.01*
Age, years	57 (46–60)	52 (40–58)	0.27
Previous smoking	6 (46)	10 (40)	0.74
Primary indication for liver transplantation			0.18
Alcohol	0 (0)	1 (4)
MASLD	2 (15)	1 (4)
Viral hepatitis	5 (38)	3 (12)
Immune-mediated and cholestatic liver disease	2 (15)	6 (24)
Other	4 (31)	14 (56)
Hepatocellular carcinoma^a	5 (38)	10 (40)	1.00
Cholangiocarcinoma^b	0 (0)	0 (0)	1.00
MELD-Na	24 (20–34)	20 (15–26)	0.20
Donor characteristics
Male sex	8 (62)	16 (64)	1.00
Age, years	51 (32–62)	40 (36–64)	0.81
Donor status			1.00
Brainstem death	13 (100)	24 (96)
Circulatory death	0 (0)	1 (4)
Cold ischaemic time, minutes	404 (348–470)	373 (298–453)	0.34
Eurotransplant donor risk index	1.54 (0.97–1.61)	1.37 (0.97–1.61)	0.96

p < 0.05.

Hepatocellular carcinoma on explant.

Cholangiocarcinoma on explant.

MASLD, metabolic-dysfunction associated steatotic liver disease; MELD-Na, model for end-stage liver disease sodium.

A mismatch at the HLA-DRB1 locus was associated with improved graft survival (p = 0.04) compared to transplants fully matched at HLA-DRB1 (Figure 2). HLA-DRB1 mismatched transplants were associated with a lower recipient MELD-Na score at the time of transplant (22 vs 29, p = 0.04), but more donor risk factors including older donor age (48 years vs 35 years, p = 0.03) and longer cold ischaemic time (365 min vs 301 min, p = 0.03; Table 3). The number of fully matched HLA-DRB1 transplants in the cohort was small (n = 9). Graft loss occurred in five out of nine fully matched HLA-DRB1 transplants and were primarily due to non-immune causes, including primary non-function (1), patient death secondary to sepsis (1), intra-abdominal bleeding (1), de novo lung malignancy (1) and squamous cell carcinoma (1).

Figure 2.

Kaplan–Meier plot of HLA-DRB1 mismatch transplants and graft survival.

Table 3.

Donor and recipient characteristics based on HLA-DRB1 mismatch status.

Characteristic	DRB1 match	DRB1 mismatch	p Value
Total, n	9	601
Follow-up, years	8.17 (6.00–11.55)	6.70 (4.44–9.46)	0.22
Male sex	6 (67)	405 (67)	1.00
Age, years	54 (32–56)	57 (48–62)	0.06
Previous smoking	5 (56)	337 (56)	1.00
Primary indication for liver transplantation			0.67
Alcohol	0 (0)	77 (13)
MASLD	0 (0)	45 (7)
Viral hepatitis	3 (33)	107 (18)
Immune-mediated and cholestatic liver disease	2 (22)	122 (20)
Other	4 (44)	250 (42)
Hepatocellular carcinoma^a	3 (33)	262 (44)	0.74
Cholangiocarcinoma^b	1 (11)	12 (2)	0.18
MELD-Na	29 (21–36)	22 (14–28)	0.04*
Donor characteristics
Male sex	5 (56)	337 (56)	1.00
Age, years	35 (25–41)	48 (33–59)	0.03*
Donor status			1.00
Brainstem death	9 (100)	556 (93)
Circulatory death	0 (0)	45 (7)
Cold ischaemic time, minutes	301 (273–395)	365 (303–440)	0.03*
Eurotransplant donor risk index	0.97 (0.97–1.37)	1.37 (0.97–1.54)	0.05

Hepatocellular carcinoma on explant.

Cholangiocarcinoma on explant.

p < 0.05.

MASLD, metabolic-dysfunction associated steatotic liver disease; MELD-Na, model for end-stage liver disease sodium.

No significant differences in graft survival were found for mismatches at the HLA-A or -B loci, compared to transplants fully matched at each respective locus (Supplemental Figure 4). There was also no difference in graft survival between one or two mismatches at the HLA-DQA or -DQB loci (Supplemental Figure 4).

Assessment of HLA compatibility as a predictor of graft failure

Univariate and multivariate Cox regression analyses were conducted to assess the impact of HLA compatibility, along with other donor and recipient factors, on the risk of graft failure. The primary multivariate Cox regression analysis was performed in the full cohort of 610 recipients. As HLA-C typing data were available in only 6% of recipients, HLA-C mismatch could not be included in the multivariate model and was therefore assessed separately using univariate Cox regression in recipients with complete HLA-C typing data.

In univariate Cox regression analysis of the HLA-C typed subgroup (n = 38), HLA-C mismatch was associated with an increased risk of graft failure, although the estimate was imprecise (HR 9.24, 95% CI 1.02–83.48, p = 0.047; Table 4). In the primary multivariate analysis (n = 610), HLA-DRB1 mismatch was associated with a reduced risk of graft failure (aHR 0.34, 95% CI 0.14–0.87, p = 0.02); however, this finding should be interpreted in the context of the small number of fully matched HLA-DRB1 transplants, as previously described (Table 4 and Figure 3). Mismatches at other individual and broad HLA loci groups were not associated with graft failure risk over time (Table 4).

Table 4.

Univariate and multivariate Cox regression analysis of predictors of graft failure.

Characteristic	Univariate			Multivariate
Characteristic	HR	95% CI	p Value	aHR	95% CI	p Value
Male	0.40	0.73–1.46	0.88
Age, years	1.03	1.01–1.05	0.001*	1.03	1.01–1.05	<0.001**
Previous smoking	1.04	0.75–1.44	0.83
Primary indication for liver transplantation
Alcohol	1
MASLD	1.34	0.57–3.19	0.50
Viral hepatitis	0.87	0.42–1.82	0.72
Immune-mediated and cholestatic liver disease	1.57	0.81–3.05	0.18
Other	1.94	1.06–3.58	0.03*	1.10	0.46–2.58	0.84
Hepatocellular carcinoma^a	1.37	0.99–1.89	0.06*	1.17	0.83–1.65	0.37
Cholangiocarcinoma^b	3.51	1.72–7.17	0.001*	2.88	1.37–6.03	0.005**
MELD-Na	0.99	0.98–1.01	0.45
Donor characteristics
Male sex	1.10	0.79–1.53	0.56
Age, years	1.00	0.99–1.01	0.61
DCD (vs DBD)	1.33	0.75–2.35	0.33
Cold ischaemic time, minutes	1.00	1.00–1.00	0.59
Eurotransplant donor risk index	0.73	0.41–1.32	0.30
Non-identical ABO blood group	1.15	0.74–1.79	0.53
Donor A2 blood group	0.89	0.39–2.03	0.80
Hepatic artery thrombosis	11.08	6.68–18.38	0.001*	3.42	2.44–4.80	<0.001**
HLA-A
0 mismatches	1
⩾1 mismatch	2.11	0.67–6.62	0.20
HLA-B
0 mismatches	1
⩾1 mismatch	1.25	0.18–8.95	0.82
HLA-C
0 mismatches	1
⩾1 mismatch	9.24	1.02–83.48	0.047*
HLA-DRB1
0 mismatches	1
⩾1 mismatch	0.40	0.16–0.98	0.04*	0.34	0.14–0.87	0.02**
HLA-DQA
1 mismatch	1
2 mismatches	1.62	0.34–7.83	0.55
HLA-DQB
1 mismatch	1
2 mismatches	0.57	0.14–2.38	0.44
All HLA loci (-A, -B, -C, -DRB1, -DQA, -DQB)
⩽4 mismatches	1
5 mismatches	1.12	0.69–1.83	0.64
⩾6 mismatches	1.05	0.66–1.67	0.83
HLA Class I (-A, -B, -C)
⩽3 mismatches	1
4 mismatches	0.95	0.68–1.32	0.76
⩾5 mismatches	2.00	0.86–4.64	0.11
HLA Class II (-DRB1, -DQA, -DQB)
⩽1 mismatch	1
2 mismatch	1.13	0.74–1.72	0.58
⩾3 mismatches	1.06	0.51–2.22	0.87
HLA-DRB1, -DQA, -DQB
⩽4 mismatches	1
5 mismatch	1.01	0.64–1.60	0.96
6 mismatches	0.97	0.62–1.52	0.91

Hepatocellular carcinoma on explant.

Cholangiocarcinoma on explant.

p ⩽ 0.10 in univariate analysis.

p < 0.05 in multivariate analysis.

aHR, adjusted hazard ratio; DBD, deceased brainstem donor; DCD, deceased circulatory donor; HR, hazard ratio; MASLD, metabolic-associated steatotic liver disease; MELD-Na, model for end-stage liver disease sodium.

Figure 3.

Forest plot of multivariate Cox regression analysis of predictors of graft failure.

Other independent predictors of graft failure on multivariate analysis included recipient age, cholangiocarcinoma and HAT (Table 4 and Figure 3). HAT was associated with a significant increase in the risk of graft failure (aHR 3.42, 95% CI 2.44–4.80, p < 0.001; Table 4). Cholangiocarcinoma on explant was also associated with increased graft failure risk (aHR 2.88, 95% CI 1.37–6.03, p = 0.005; Table 4), whereby 75% of failed grafts were secondary to patient death from disease recurrence. Recipient age had a small effect on graft failure risk that is unlikely to be of clinical significance (aHR 1.03, 95% CI 1.01–1.05, p < 0.001; Table 4). Although HCC was associated with an increased risk of graft failure in univariate analysis (HR 1.37, 95% CI 0.99–1.89, p = 0.06), this association lost statistical significance in multivariate analysis.

HLA compatibility and hepatic artery thrombosis

Given the findings by Bricogne et al.⁴ of an association between HLA-A mismatch, HAT and early graft loss, we proceeded to perform an additional sub-analysis on recipients who developed HAT. In the cohort of 610 recipients, 20 patients (3%) developed HAT. Among these cases, 14 were acute (within 30 days post-transplant) and 6 were delayed (occurring more than 30 days post-transplant). Graft loss occurred in 15 of 20 (75%) cases of HAT.

We found no relationship between HLA mismatch and the development of HAT (Supplemental Tables 1 and 2). The occurrence of non-identical ABO blood group (7/20, 35% vs 92/590, 16%; p = 0.03) and A2 donor blood group transplants (4/20, 20% vs 27/590, 5%; p = 0.01) was notably higher in HAT cases compared to non-HAT cases (Supplemental Table 1). Among the 4 A2 donor transplants resulting in HAT, 2 (50%) were A2 to O, 1 (25%) A2B to B and 1 (25%) A2 to B transplants. In non-HAT cases, there were 27 A2 donor transplants, with 20 (74%) A2 to O, 5 (19%) A2B to B and 2 (7%) A2 to B transplants. However, A2 donor blood group transplants were not associated with an increased risk of HAT in multivariate regression analysis (Supplemental Table 2). HAT was also not associated with classical risk factors, including HCC, smoking status, hepatic artery reconstruction, cold ischaemic time and deceased circulatory donor grafts (Supplemental Tables 1 and 2).

Discussion

We investigated the impact of HLA mismatching on graft survival in a cohort of 610 deceased donor LT recipients. On Kaplan–Meier analysis, HLA-C mismatch was associated with reduced graft survival, with a similar association observed in univariate Cox regression, although the estimate was imprecise (HR 9.24, 95% CI 1.01–83.48, p = 0.047). In contrast, HLA-DRB1 mismatch was associated with a reduced risk of graft failure on multivariate analysis (aHR 0.34, 95% CI 0.14–0.87, p = 0.02). However, both findings were derived from small comparator groups at each locus, and the possibility of chance findings and residual confounding cannot be excluded. HLA mismatching was not associated with the development of HAT, suggesting that any relationship between HLA histocompatibility and graft survival may be mediated through alternative immunological pathways.

Our findings raise the possibility of an association between HLA-C mismatch and reduced graft survival in LT recipients. This contrasts with the few studies that have examined the relationship between HLA-C histocompatibility and liver graft outcomes. Both a study of 134 living donor LT recipients and a larger study of 1042 deceased donor LT recipients found no association between HLA-C mismatch and graft survival.^4,21 Although HLA-C incompatibility has been associated with an increased risk of acute rejection^27–30 and reduced recipient survival,³¹ other literature has reported no effect of HLA-C incompatibility on either acute^5,21,32–35 or chronic³⁶ rejection. Our study did not have the rejection data necessary to investigate this; however, the causes of graft failure in HLA-C mismatched transplants were not clearly immune-mediated. The current conflicting literature may be attributed to the complex role of HLA-C in natural killer (NK) cell function. HLA-C is the major inhibitory ligand for killer cell immunoglobulin-like receptors (KIRs) expressed on NK cells.³⁷ In allogeneic transplantation, recipient NK cell activation can occur through the presence of activating HLA-C/KIR ligands or the absence of inhibitory HLA-C/KIR ligands.³⁷ Variable effects on liver graft outcomes have been demonstrated with HLA-C/KIR ligand disparity^{29,35,36,38,39} and the presence of the HLA-C inhibitory genotype on donor allografts.^{32,35,38–40} Importantly, the wide confidence interval for the HLA-C hazard ratio in our study reflects limited precision, driven by the small number of recipients with available HLA-C typing data. In addition, shorter follow-up in the HLA-C mismatched group may have resulted in underrepresentation of late graft failure. Taken together, while a biological rationale for an association between HLA-C mismatch and graft survival is plausible, these findings are insufficient to draw firm conclusions and require validation in larger cohorts with complete HLA-C typing data.

A difference in graft survival was also observed according to HLA-DRB1 matching status, with HLA-DRB1-mismatched transplants associated with a reduced risk of graft failure. However, this effect is likely overestimated given the small number of recipients (n = 9) with HLA-DRB1 matching, with graft failures largely attributed to non-immune causes, raising the possibility of a Type I error. Previous studies have produced mixed results, demonstrating either an inverse^6,13,41,42 association or no^{4,11,15–17,20,21} association between HLA-DR or HLA-DRB1 matching and graft survival in LT recipients. In the context of a small comparator group and limited events, this finding should be interpreted with caution and considered as hypothesis-generating.

Bricogne et al.⁴ recently demonstrated reduced graft and patient survival in HLA-A mismatched LTs driven by early graft loss secondary to HAT. Although a trend towards improved graft survival in HLA-A-matched grafts was observed in our cohort, this did not reach statistical significance. Furthermore, mismatching of HLA at any locus was not associated with the development of HAT. Interestingly, the HAT group had a greater proportion of A2 donor blood group transplants compared to the non-HAT group; however, this was not found to increase the risk of HAT when adjusting for other covariates. The A2 phenotype of the A blood group has a lower cell surface expression of the A antigen, resulting in decreased alloreactivity with recipient anti-A antibodies.⁴³ A2 LTs have expanded the donor pool for blood group O candidates, with A2 to O LTs showing comparable rates of post-transplant complications (including vascular complications), graft survival and patient survival, to O to O LTs.^43–45 Although excellent graft and patient survival rates have been demonstrated in A2 to B renal transplants,^46,47 there is limited literature in the LT setting. Notably, a case of refractory antibody-mediated rejection post-A2 to B LT has been reported.⁴⁸ While our study did not reveal a significant impact of A2 donor blood group on graft failure, further studies are required to explore the potential relationship between A2 donor LT and the development of HAT.

Our study has several limitations that should be considered. Less than 10% of patients had available HLA-C, -DQA and -DQB histocompatibility data due to the utilisation of less sensitive HLA typing technologies in earlier transplant eras. The high proportion of missing HLA-C data precluded its inclusion in the primary multivariable model, thus analyses involving HLA-C should be interpreted cautiously. Although HLA-DRB1 histocompatibility data was available in all patients, the number of HLA-DRB1 matched transplants was also small. The relatively small sample sizes of these cohorts may have introduced imprecision and increased the risk of a Type I error. Furthermore, HLA-C typing data was predominantly available in more recent transplant eras, raising the possibility of confounding by temporal changes in clinical practice. Advances in surgical technique, immunosuppression, and peri-operative care across the study period may have independently influenced graft outcomes, and the contribution of the transplant era to the observed associations, particularly at the HLA-C locus, cannot be fully determined in this retrospective study. Although most graft failures in the overall cohort were largely non-immune related, the absence of detailed rejection phenotyping, including antibody-mediated rejection and donor-specific antibody data, limits mechanistic interpretation of the observed associations. Furthermore, the low deceased circulatory donor rate in this cohort may impact the conclusions drawn, as neoantigen exposure from allograft injury could promote HLA-specific responses, thereby influencing early allograft outcomes. Nevertheless, the utilisation of a large, comprehensive database of LT donors and recipients with extensive follow-up was a significant strength of our study. Patient and donor characteristics and outcomes were collected prospectively from the time of waitlisting, resulting in a robust and reliable database that effectively minimises recall bias.

Conclusion

The findings of this study suggest an uncertain association between HLA histocompatibility and graft outcomes in LT recipients, thus reaffirming the current clinical practice of using non-HLA-matched donors to expand the donor pool in the face of increasing donor shortages. In fact, the observed potential contrasting relationships between different HLA loci on graft survival may reflect the complexity of immunogenetic factors in organ transplantation. Given the limited sample size at certain HLA loci, these findings are hypothesis-generating and should be interpreted with caution. Further evaluation in larger, prospective multicentre cohorts utilising high-resolution molecular HLA typing at the eplet level, with adequate follow-up duration, is required to better define these relationships.

Supplemental Material

sj-docx-1-tag-10.1177_17562848261447157 – Supplemental material for Evaluation of human leukocyte antigen histocompatibility and its impact on liver transplant graft outcomes

Supplemental material, sj-docx-1-tag-10.1177_17562848261447157 for Evaluation of human leukocyte antigen histocompatibility and its impact on liver transplant graft outcomes by Dorothy Liu, Kate Collins, Marcos V. Perini, Graham Starkey, Michael Fink, Robert Jones, Adam Testro, Marie Sinclair and Avik Majumdar in Therapeutic Advances in Gastroenterology

Footnotes

Acknowledgements

We would like to acknowledge the assistance of the Australian Red Cross Lifeblood in obtaining donor and recipient HLA typing data for this study.

Declarations

ORCID iD

Dorothy Liu

Supplemental material

Supplemental material for this article is available online.

References

Williams

Opelz

McGarvey

, et al. The risk of transplant failure with HLA mismatch in first adult kidney allografts from deceased donors. Transplantation 2016; 100(5): 1094–1102.

Dehn

Spellman

Hurley

, et al. Selection of unrelated donors and cord blood units for hematopoietic cell transplantation: guidelines from the NMDP/CIBMTR. Blood 2019; 134(12): 924–934.

Sheldon

Poulton

HLA typing and its influence on organ transplantation. Methods Mol Biol 2006; 333: 157–174.

Bricogne

Halliday

Fernando

, et al. Donor-recipient human leukocyte antigen A mismatching is associated with hepatic artery thrombosis, sepsis, graft loss, and reduced survival after liver transplant. Liver Transpl 2022; 28(8): 1306–1320.

Forner

Liwski

Alwayn

Human leukocyte antigen, allele, and eplet mismatches in liver transplantation; observations from a small, single center cohort. Hum Immunol 2018; 79(3): 154–159.

Balan

Ruppert

Demetris

, et al. Long-term outcome of human leukocyte antigen mismatching in liver transplantation: results of the national institute of diabetes and digestive and kidney diseases liver transplantation database. Hepatology 2008; 48(3): 878–888.

Knechtle

Kalayolu

D’Alessandro

, et al. Histocompatibility and liver transplantation. Surgery 1993; 114(4): 667–671.

Francavilla

Hadzic

Underhill

, et al. Role of HLA compatibility in pediatric liver transplantation. Transplantation 1998; 66(1): 53–58.

Wong

Donaldson

Devlin

, et al. Repeat HLA-B and -DR loci mismatching at second liver transplantation improves patient survival. Transplantation 1996; 61(3): 440–444.

10.

Muro

López-Álvarez

Campillo

, et al. Influence of human leukocyte antigen mismatching on rejection development and allograft survival in liver transplantation: is the relevance of HLA-A locus matching being underestimated? Transpl Immunol 2012; 26(2–3): 88–93.

11.

Donaldson

Underhill

Doherty

, et al. Influence of human leukocyte antigen matching on liver allograft survival and rejection: “the dualistic effect”. Hepatology 1993; 17(6): 1008–1015.

12.

Mañez

White

Linden

, et al. The influence of HLA matching on cytomegalovirus hepatitis and chronic rejection after liver transplantation. Transplantation 1993; 55(5): 1067–1070.

13.

Shin

Kim

Kwon

, et al. Role of human leukocyte antigen compatibility in graft outcomes after living donor liver transplantation. Transplant Proc 2016; 48(4): 1123–1129.

14.

Markus

Duquesnoy

Blaheta

, et al. Role of HLA antigens in liver transplantation with special reference to cellular immune reactions. Langenbecks Arch Surg 1998; 383(1): 87–94.

15.

Poli

Scalamogna

Aniasi

, et al. A retrospective evaluation of HLA-A, B and -DRB1 matching in liver transplantation. Transpl Int 1998; 11(Suppl. 1): S347–S349.

16.

Navarro

Herrine

Katopes

, et al. The effect of HLA class I (A and B) and class II (DR) compatibility on liver transplantation outcomes: an analysis of the OPTN database. Liver Transpl 2006; 12(4): 652–658.

17.

Jakab

Navarro

Colombe

, et al. Human leukocyte antigen and adult living-donor liver transplantation outcomes: an analysis of the organ procurement and transplantation network database. Liver Transpl 2007; 13(10): 1405–1413.

18.

Neumann

Guckelberger

Langrehr

, et al. Impact of human leukocyte antigen matching in liver transplantation. Transplantation 2003; 75(1): 132–137.

19.

Lan

Zhang

, et al. Impact of human leukocyte antigen mismatching on outcomes of liver transplantation: a meta-analysis. World J Gastroenterol 2010; 16(27): 3457–3464.

20.

Muro

Marín

Torío

, et al. Effect of HLA matching on liver graft survival. Transplant Proc 1999; 31(6): 2477–2479.

21.

Badawy

Kaido

Yoshizawa

, et al. Human leukocyte antigen compatibility and lymphocyte cross-matching play no significant role in the current adult-to-adult living donor liver transplantation. Clin Transplant 2018; 32(4): e13234.

22.

Doran

Geczy

Painter

, et al. A large, single center investigation of the immunogenetic factors affecting liver transplantation. Transplantation 2000; 69(7): 1491–1498.

23.

Mañez

Mateo

Tabasco

, et al. The influence of HLA donor-recipient compatibility on the recurrence of HBV and HCV hepatitis after liver transplantation. Transplantation 1995; 59(4): 640–642.

24.

Kubal

Mangus

Ekser

, et al. Class II human leukocyte antigen epitope mismatch predicts de novo donor-specific antibody formation after liver transplantation. Liver Transpl 2018; 24(8): 1101–1108.

25.

Meszaros

Dubois

Niemann

, et al. Class II human leukocyte antigen epitope mismatch predicts de novo donor-specific antibody formation after liver transplantation. Liver Transpl 2019; 25(1): 184–185.

26.

von Elm

Altman

Egger

, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Int J Surg 2014; 12(12): 1495–1499.

27.

Kok

Ilcken

Houwen

RHJ

, et al. The effect of genetic HLA matching on liver transplantation outcome: a systematic review and meta-analysis. Ann Surg Open 2023; 4(3): e334.

28.

Guiral

Segundo

Irure

, et al. Number of antibody-verified eplet in HLA-C locus as an independent factor of T-cell-mediated rejection after liver transplantation. Transplantation 2020; 104(3): 562–567.

29.

López-Alvarez

Moya-Quiles

Minguela

, et al. HLA-C matching and liver transplants: donor-recipient genotypes influence early outcome and CD8+KIR2D+ T-cells recuperation. Transplantation 2009; 88(3 Suppl.): S54–S61.

30.

Moya-Quiles

Muro

Torío

, et al. Human leukocyte antigen-C in short- and long-term liver graft acceptance. Liver Transpl 2003; 9(3): 218–227.

31.

Tajima

Hata

Kusakabe

, et al. The impact of human leukocyte antigen mismatch on recipient outcomes in living-donor liver transplantation. Liver Transpl 2022; 28(10): 1588–1602.

32.

Chen

Zhen

, et al. Donor with HLA-C2 is associated with acute rejection following liver transplantation in Southern Chinese. HLA 2022; 100(2): 133–141.

33.

Ono

Ide

Tanaka

, et al. Molecular mismatch predicts T cell-mediated rejection and de novo donor-specific antibody formation after living donor liver transplantation. Liver Transpl 2021; 27(11): 1592–1602.

34.

Oertel

Berr

Schröder

, et al. Acute rejection of hepatic allografts from HLA-DR13 (allele DRB1*1301)-positive donors. Liver Transpl 2000; 6(6): 728–733.

35.

Lee

Park

, et al. Human leukocyte antigen-C genotype and killer immunoglobulin-like receptor-ligand matching in Korean living donor liver transplantation. Ann Lab Med 2017; 37(1): 45–52.

36.

Legaz

Bolarín

Campillo

, et al. Killer cell immunoglobulin-like receptors (KIR) and human leucocyte antigen C (HLA-C) increase the risk of long-term chronic liver graft rejection. Int J Mol Sci 2022; 23(20): 12155.

37.

Mendel

Chavin

Bratton

, et al. HLA-C and liver transplant outcomes: interpreting the facts. Am J Transplant 2009; 9(7): 1491–1492.

38.

Hanvesakul

Spencer

Cook

, et al. Donor HLA-C genotype has a profound impact on the clinical outcome following liver transplantation. Am J Transplant 2008; 8(9): 1931–1941.

39.

Moroso

van der Meer

Tilanus

, et al. Donor and recipient HLA/KIR genotypes do not predict liver transplantation outcome. Transpl Int 2011; 24(9): 932–942.

40.

Tran

Middleton

Döhler

, et al. Reassessing the impact of donor HLA-C genotype on long-term liver transplant survival. Am J Transplant 2009; 9(7): 1674–1688.

41.

Campos

Quijano

Franco

, et al. Beneficial effects of HLA class II incompatibility in living donor liver transplantation. Transplant Proc 2003; 35(5): 1888–1891.

42.

Sieders

Hepkema

Peeters

, et al. The effect of HLA mismatches, shared cross-reactive antigen groups, and shared HLA-DR antigens on the outcome after pediatric liver transplantation. Liver Transpl 2005; 11(12): 1541–1549.

43.

Roberts

Lai

JC.

A2 liver transplantation across the ABO barrier: increasing options in the donor pool?

Clin Liver Dis (Hoboken) 2017; 10(6): 135–138.

44.

Kluger

Guarrera

Olsen

, et al. Safety of blood group A2-to-O liver transplantation: an analysis of the united network of organ sharing database. Transplantation 2012; 94(5): 526–531.

45.

Skogsberg

Breimer

Friman

, et al. Successful ABO-incompatible liver transplantation using A2 donors. Transplant Proc 2006; 38(8): 2667–2670.

46.

Alkhunaizi

de Mattos

Barry

, et al. Renal transplantation across the ABO barrier using A2 kidneys. Transplantation 1999; 67(10): 1319–1324.

47.

Bryan

Winklhofer

Murillo

, et al. Improving access to kidney transplantation without decreasing graft survival: long-term outcomes of blood group A2/A2B deceased donor kidneys in B recipients. Transplantation 2005; 80(1): 75–80.

48.

Puri

Rammohan

Sachan

, et al. ABO-incompatible living donor liver transplant from a blood type A2 donor to a type B recipient: a note of caution. Exp Clin Transplant 2022; 20(1): 100–103.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.46 MB