Preformed donor-specific antibodies are associated with acute rejection and biliary complications after liver transplantation: A Swiss Transplant Cohort Study analysis

Abstract

The impact of preformed donor-specific antibodies (DSA) in liver transplantation (LT) remains controversial despite evidence linking their presence to an increased risk of early allograft damage, as well as antibody- and T-cell–mediated rejection. In this nationwide analysis, preformed DSA were assessed using single-antigen bead assays [positive if mean fluorescence intensity (MFI) ≥1000]. This study included all LT recipients enrolled in the Swiss Transplant Cohort Study (STCS) who underwent LT between 2014 and 2016. One-year post-LT outcomes, including cumulative allograft and patient survival, as well as the incidence of biliary, vascular, and infectious complications, were compared between DSA-positive (DSA+) and DSA-negative (DSA−) individuals. Among 321 LT performed in 306 patients, preformed DSA were detected in 92 (28.7%) and more frequently observed in patients with a history of prior transplantation (p=0.008) or autoimmune liver disease (p=0.036). Class I and II DSA were present in 48.9% and 71.1% of DSA+ cases, with concomitant class I and II DSA in 20.7%. The median (IQR) cumulative MFI (cMFI) of the preformed DSA was 3768 (1875–10,537), and 52.2% of DSA+ patients harbored multiple DSA. While overall patient survival did not differ between DSA+ and DSA− individuals, DSA+ patients with cMFI ≥5000 exhibited a higher incidence of allograft failure and biopsy-proven rejection. Multivariate analysis revealed that the presence of preformed DSA was independently associated with biliary complications (HR 2.26, 95% CI 1.17–4.37, p=0.02) but not with vascular or infectious complications. In summary, preformed DSA were associated with biliary complications, increased rejection, and reduced allograft survival. These findings suggest pre-transplant immunological risk assessment and the need for tailored immunosuppressive strategies in LT.

INTRODUCTION

Despite significant advancements in surgical techniques, perioperative care, and immunosuppressive therapy, allograft rejection remains one of the many challenges that impact patient outcomes and long-term allograft survival.^1–4 One of the critical factors contributing to allograft rejection, in particular to antibody-mediated rejection (AMR), is the presence of anti-human leukocyte antigen (HLA) donor-specific antibodies (DSA).^5–7 These antibodies, which are developed in response to previous exposures to donor antigens, can arise from prior transplants, blood transfusions, and/or pregnancies.^5–7 DSA can be preformed if they are already present in the recipient’s serum at the time of transplantation, or de novo, if they were not present at the time of the transplantation but develop thereafter.

The presence of preformed DSA are known to be associated with an increased risk of AMR and poorer allograft outcomes in kidney, but also lung and heart transplantation.^8–12 In contrast, the liver is usually considered more resistant to antibody-mediated damage due to its unique immunological environment, which includes a high tolerance for alloantigens and a capacity for immune regulation.^1,13 Studies have challenged this notion, suggesting that preformed DSA might have a more significant impact on liver transplantation (LT) outcomes than previously thought.^14–17 These studies have reported associations between preformed DSA and increased incidences of acute rejection and reduced allograft survival.^15,17 The exact mechanisms through which DSA contribute to these adverse outcomes in LT are not fully understood, but it is believed that they may involve complement activation, endothelial injury, and subsequent inflammatory responses.¹⁰

The Swiss Transplant Cohort Study (STCS) provides a unique opportunity to analyze the role of preformed DSA in LT within a well-defined, multicentric, and nationwide population. The STCS collects comprehensive data on transplant recipients across 6 centers in Switzerland, including immunological profiles, clinical outcomes, and detailed follow-up information. This study aimed to investigate the prevalence of preformed DSA in LT recipients and their association with allograft survival, rejection episodes, and post-transplant complications within the first year after LT, in the STCS.

METHODS

Study population

We carried out a multicentric retrospective study including all patients who underwent LT between January 1, 2014, and December 31, 2016, in Switzerland and consented to participate in the STCS. Before transplantation, participation is offered to all solid-organ transplant candidates in the 6 Swiss transplant centers (Lausanne, Geneva, Zürich, Bern, Basel, St-Gallen) and in particular to all LT candidates transplanted in Geneva-Lausanne, Zürich, and Bern. This study was approved by the STCS Scientific Committee (reference: FUP099) and by the local ethics committee (Commission cantonale d’éthique de la recherche sur l’être humain, CER-VD; reference: 2017-01048).

Key inclusion/exclusion criteria

All participants enrolled in the STCS within the period mentioned above, regardless of age, were included. Patients who did not consent to participate in the STCS were excluded. Similarly, participants without available or retrievable data on preformed anti-HLA donor-specific antibody or participants with an anti-HLA antibody test but insufficient donor HLA typing information for interpretation were excluded.

Preformed anti-HLA antibody characterization and HLA typing

Screening and specificity analysis for preformed donor-specific anti-HLA antibodies (DSA) were determined using LABScreen Mixed Bead and LABScreen Single Antigen Bead assay (OneLambda Inc., Canoga Park, CA, USA), in case of positivity of the former test, by means of the Luminex technology (Luminex Corporation, Austin, TX, USA). Anti-HLA antibody measurements were prospectively performed within 24 hours before LT in a subset of patients and collected at each of the liver transplant centers. If not available, anti-HLA antibody measurements were retrospectively performed in batches on stored frozen serum samples (also collected within 24 hours before LT) at the National Reference Laboratory for Histocompatibility in Geneva. Mean fluorescence intensity (MFI) ≥1000 was considered positive, and cumulative MFI (cMFI) was calculated for each participant with more than one DSA with an MFI ≥1000, by adding the single MFI value of each DSA. In the manuscript, the denominations “preformed DSA” and “DSA” are used invariably.

Class I (A, B, C) and class II (DR, DQ, DP) HLA typing data from donors and recipients were retrieved from local and/or STCS databases. Whenever unavailable or incomplete in recipients positive for anti-HLA antibodies, complementary class I (A, B, C) and/or class II (DR, DQ, DP) HLA typing was obtained from cryopreserved DNA using low-resolution molecular HLA typing via PCR-sequence specific oligonucleotide probe (SSOP) hybridization in combination with Luminex (Luminex Corporation, Ausin, TX, USA) technology. Predicted Indirectly Recognizable HLA Epitopes (PIRCHE-II; application version 3.3.69, database version 3.47) score was calculated to explore its association with patient and allograft survival. Missing HLA typings from the C and DQ loci, and low-resolution HLA typing data were extrapolated using a multiple imputation approach, integrating data from the National Marrow Donor Program database 2007 for European descent. Because PIRCHE-II and antibody formation are logarithmically correlated, PIRCHE-II scores were transformed by natural logarithm [ln(PIRCHE-II +1)] for further analyses.¹⁸ Patients and donor pairs without complete HLA typing data to compute the PIRCHE-II score were excluded.

Data collection

We collected baseline recipient and donor characteristics for every LT procedure from the STCS database. As explained above, anti-HLA antibodies and HLA typing data were collected at each of the centers and completed whenever necessary. With regard to outcomes, information on patient and allograft survival, respectively, death and allograft loss, was systematically retrieved from the STCS database. We also collected information on biopsy-proven and clinically-suspected T-cell–mediated rejection episodes, overall complications, as well as biliary (anastomotic stenosis, anastomotic leak, intrahepatic, hilar),¹⁹ vascular (hepatic artery, portal vein, and/or hepatic vein stenosis or thrombosis), and infectious (any bacterial, viral, or fungal infection) complications. Detailed data on biliary complications not included in the STCS database (type of anastomosis, localization of biliary complication, grading of biliary complication, cofactors for biliary complications) were retrospectively collected. T-cell–mediated rejection episodes were described according to the 2016 Banff criteria.²⁰ No data on antibody-mediated rejection is available in the STCS. Data on retrieval (standard recovery, super-rapid recovery) and preservation methods (including machine perfusion), warm ischemia time for the donor organs, as well as on surgical procedures, were retrospectively collected. The analyses of outcomes were restricted to the first year after LT.

Statistical analysis

Baseline characteristics included median and interquartile ratio (IQR) or mean and standard deviation (SD) for continuous variables, and frequencies and percentages for categorical variables. The characteristics were compared between the preformed DSA-positive (DSA+) and the DSA-negative (DSA−) participants using Chi-square tests or the Fisher exact test for categorical variables and t tests or non-parametric testing (Mann–Whitney or Kruskal–Wallis rank sum test) for continuous variables, as appropriate.

Cumulative incidence was calculated for death and liver allograft failure (graft loss), considering the first event. The cumulative incidence was then compared between the DSA+ and the DSA− participants by means of the Gray test. Kaplan–Meier survival analysis was performed to compare DSA+ and DSA− patients, with the use of the Fleming–Harrington test. Finally, Cox proportional hazards models, with death and allograft loss as competing events, were employed to evaluate the impact of DSA on rejection and overall as well as specific complications such as biliary, vascular, and infectious complications. Variables for inclusion in multivariable models were selected based on clinical relevance, evidence from prior literature, and parsimony guided by the number of events. A p-value <0.05 was considered statistically significant. No corrections for multiple comparisons were applied.

RESULTS

Baseline patients’ characteristics

A total of 355 LT were performed in 336 patients between January 2014 and December 2016 in Switzerland. After the exclusion of 24 LT from 23 non-consenting participants and 10 LT without interpretable or retrievable information on preformed DSA, 321 LT in 313 patients were included in the study (Figure 1). Patients were mainly male (70.4%), Caucasian (93.9%), and receiving a first and single LT in 87.5% and 93.9% of the cases, respectively (Table 1). The majority of recipients received an allograft from a deceased donor (96.1%), and most commonly from a donor after brain death (89.5%). The leading indications for LT were cirrhosis due to chronic viral hepatitis (23.5%) and alcohol-associated cirrhosis (20.6%). An additional 10.5% of patients presented with a combination of these two etiologies, with or without coexisting metabolic dysfunction–associated steatotic liver disease (MASLD). At baseline, 82.9% of the participants received an induction immunosuppressive treatment, with basiliximab administered in 96.6% of the cases. A calcineurin inhibitor-based immunosuppression regimen was initiated in 96.7% of the cohort.

FIGURE 1.

Study flowchart. Abbreviation: STCS, Swiss Transplant Cohort Study.

TABLE 1.

Baseline characteristics of liver transplant recipients according to the presence or absence of preformed donor-specific antibodies

Characteristics	Preformed DSA positive	Preformed DSA negative	Total	p a
Total cohort, n (%)	92 (28.7%)	229 (71.3%)	321 (100%)	NA
Recipient age at Tx (y)	48.2±18.5	50.8±16.1	50.1±16.8	0.28
Male recipients, n (%)	60 (65.2%)	166 (72.5%)	226 (70.4%)	0.22
Caucasian, n (%)	86 (93.5%)	215 (93.9%)	301 (93.8%)	1.00
BMI, kg/m2	27.9±9.0	25.7±5.3	25.5±5.3	0.26
MELD-score used for allocation	27.9±9.0	27.1±8.8	27.3±8.8	0.468
Type of Tx, n (%)				1.00
Liver	88 (95.7%)	213 (93.0%)	301 (93.8%)
Liver–kidney	4 (4.3%)	10 (4.4%)	14 (4.4%)
Other comb.	0 (0.0%)	6 (2.6%)	6 (1.8%)
Previous liver transplantation, n (%)	19 (20.7%)	21 (9.2%)	40 (12.5%)	0.01
Indication for Tx, n (%)	91b	224b	315b	0.04
Alcohol	11 (12.1%)	52 (24.1%)	65 (20.6%)
Hepatitis B/C	19 (20.9%)	53 (23.7%)	72 (23.5%)
PBC	7 (7.7%)	6 (2.7%)	13 (4.1%)
PSC	6 (6.6%)	6 (2.7%)	12 (3.8%)
AIH	3 (3.3%)	3 (1.3%)	6 (1.9%)
MASLD	4 (4.4%)	6 (2.7%)	10 (3.2%)
Other	41 (45.1%)	96 (42.9%)	137 (43.5%)
Presence of HCC, n (%)	39 (42.4%)	97 (42.4%)	136 (42.4%)	1.00
Donor age at donation (y)	51.0±19.9	50.3±19.6	50.5±19.7	0.77
Male donors, n (%)	52 (56.5%)	125 (54.6%)	177 (55.1%)	0.80
Type of donor				1.00
Deceased	89 (96%)	220 (96.1%)	309 (96.3%)
Living	3 (3.3%)	9 (3.9%)	12 (3.4%)
Data on deceased donor
DBD	82 (92.1%)	197 (89.5%)	279 (90.3%)	0.67
DCD	7 (7.9%)	23 (10.5%)	30 (9.7%)
Machine	Yes: 8 (9.0%)	Yes: 25 (11.4%)	Yes: 33 (10.7%)	0.54
perfusion	No: 81 (91.0%)	No: 195 (88.6%)	No: 276 (89.3%)
Retrieval	STD: 82 (92.1%)	STD: 197 (89.5%)	STD: 279 (90.3%)	0.52
Method	SRR: 7 (7.9%)	SRR: 23 (10.5%)	SRR: 30 (9.7%)
Biliary	DTD: 71 (81.6%)	DTD: 178 (81.7%)	DTD: 249 (81.6%)	0.99
anastomosis typec	BD: 16 (18.4%)	BD: 40 (18.3%)	BD: 56 (18.4%)
Warm ischemia time (min)d	40.0 (28.5–50.0)	37.0 (29.0–54.8)	38.0 (29.0–54.0)	0.98
HOPE duration (min)d	103.5 (78.8–121.3)	106.5 (78.8–180.8)	103.5 (78.8–161.3)	0.63
Cold ischemia time (h)	6.6±2.2	6.7±2.3	6.7±2.3	0.63
CMV risk constellation, n (%)				0.75
Low risk	55 (61.8%)	137 (60.6%)	192 (61.0%)
Intermediate	15 (16.9%)	46 (20.4%)	61 (19.4%)
High risk	19 (21.3%)	43 (19.0%)	62 (19.7%)
Missing info.	3 (0.3%)	3 (0.9%)	6 (0.7%)
Induction IS, n (%)	75 (81.5%)	191 (83.4%)	266 (82.9%)	0.74
Maintenance IS, n (%)				0.12
TAC-based	80 (95.2%)	216 (97.3%)	296 (96.7%)
CsA-based	2 (2.4%)	4 (1.8%)	6 (2.0%)
mTOR-based	0 (0.0%)	1 (0.4%)	1 (0.3%)
Others	2 (2.4%)	1 (0.5%)	3 (1.0%)
HLA-A mismatche, n (%)	0—5 (7.6%)	0—13 (7.3%)	0—18 (7.4%)	0.16
	1—32 (48.5%)	1—63 (35.6%)	1—95 (39.1%)
	2—29 (43.9%)	2—101 (57.1%)	2—130 (53.5%)
HLA-B mismatche, n (%)	0—0 (0.0%)	0—3 (1.7%)	0—3 (1.3%)	0.84
	1—21 (31.8%)	1—56 (31.6%)	1—77 (31.7%)
	2—45 (68.2%)	2—118 (66.7%)	2—163 (67.1%)
HLA-DR mismatche, n (%)	0—0 (0.0%)	0—10 (5.6%)	0—10 (4.1%)	0.08
	1—26 (40.0%)	1—78 (44.1%)	1—104 (43.0%)
	2—39 (60.0%)	2—89 (50.3%)	2—128 (52.9%)
PIRCHE-II scoree	90.36±37.89	93.71±43.36	92.80±41.89	0.77

Notes: Categorial data expressed as a number (percentage) and continuous data as mean±standard deviation. Median and interquartile range were used for warm ischemia time and HOPE duration.

^a p-Value obtained via the Fisher exact test or Kruskal–Wallis rank sum test.

^b Six participants had missing data for liver disease etiology leading to transplantation.

^c Two deceased donors with missing data.

^d Missing data for warm ischemia time and HOPE duration were observed in 176 and 277 deceased donors, respectively.

^e Missing data for HLA-A and HLA-B, and HLA-DR were observed in 78 and 79 participants, respectively.

Abbreviations: AIH, autoimmune hepatitis; BD, bilio-digestive anastomosis; BMI, body mass index; CMV, cytomegalovirus; CsA, cyclosporine A; DBD, donor after brain death; DCD, donor after cardio-circulatory death; DSA, donor-specific antibodies; DTD, duct-to-duct anastomosis; HCC, hepatocellular carcinoma; HLA, human leukocyte antigen; HOPE, hypothermic oxygenated perfusion; IS, immunosuppression; mTOR, mammalian target of rapamycin; MASLD, metabolic dysfunction–associated steatotic liver disease; NA, not applicable; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; STD, standard recovery; SRR, super-rapid recovery; TAC, tacrolimus; Tx, transplantation.

Prevalence of preformed DSA in liver transplantation in the Swiss Transplant Cohort Study

Preformed DSA were overall detected in 92 (28.7%) of the 321 liver transplant procedures (Table 1). Baseline characteristics between DSA+ and DSA− participants were similar, except for a history of previous transplantation (20.7% vs. 9.2%; p=0.01) and autoimmune liver disease (17.6% vs. 6.7%; p=0.04 for overall comparison), which were more prevalent in the DSA+ population. Preformed class I and class II DSA were present in 48.9% and 71.1% of the DSA+ participants, with concomitant class I and class II DSA in 19 (20.7%) of them (Supplemental Table S1, http://links.lww.com/LVT/B105). DSA+ patients were more likely to have more than one DSA (52.2%, as opposed to 47.8% with only 1 DSA), and the median (IQR) cMFI of the preformed DSA was 3768 (1875–10,537). Of note, the proportions of HLA-A, HLA-B, and HLA-DR mismatches, as well as PIRCHE-II scores, were similar in both populations (Table 1).

Preformed DSA were not associated with overall survival, but correlated with increased allograft loss within the first-year post-LT

A Kaplan–Meier survival analysis showed no difference in overall survival between the DSA+ and the DSA− participants (Figure 2). Patients’ survival was indeed 92.4% and 94.3% at 1-year post-transplantation in the DSA+ and the DSA− population, respectively (Table 2). The median time to the occurrence of death was similar in both groups. The cumulative incidence curves of death and graft failure, considered as competing events, showed that 24 (75.0%) of the 32 combined events occurred within the first 3 months post-transplantation (Figure 3A). Although 1-year graft survival was similar in overall DSA+ and DSA−- patients (89.1% vs. 90.4%, p=0.49) (Figure 3B), the incidence of graft failure was higher in DSA+ patients with cMFI ≥5000 (Figure 3C) and ≥10,000 (p=0.02 and p=0.02, respectively) (Figure 3D). Causes of allograft loss differed by DSA status: compared with DSA− recipients (n=17), DSA+ recipients (n=9) had no primary non-function (0.0% vs. 35.3%) and similar vascular/ischemic loss (33.3% vs. 29.4%), but a greater proportion attributed to rejection (22.2% vs. 5.9%), biliary complications (11.1% vs. 5.9%), and unknown causes (33.3% vs. 11.8%) (Supplemental Table S2, http://links.lww.com/LVT/B105).

FIGURE 2.

Overall survival curve of patients with (DSA+) and without (DSA−) preformed donor-specific antibodies. Abbreviations: DSA−, anti-HLA donor-specific antibodies negative; DSA+, anti-HLA donor-specific antibodies positive; TX, transplantation.

TABLE 2.

Patient and graft survival within the first year after liver transplantation (time-to-event analysis)

Characteristics	Preformed DSA+	Preformed DSA−	Total	p a
Total cohort, n (%)	92 (28.7%)	229 (71.3%)	321 (100%)	NA
Median time to death (d)	30.0 (10.0–67.0)	29.0 (10.0–112.0)	29.5 (8.5–80.5)	0.97
Death	7 (7.6%)	13 (5.7%)	20 (6.2%)	0.52
Alive	85 (92.4%)	216 (94.3%)	301 (93.8%)
Median time to graft loss or death (d)	46.5 (1.5–132.3)	28.5 (2.0–50.5)	29.5 (0.8–80.5)	0.57
Death	1 (1.1%)	5 (2.2%)	6 (1.9%)	0.64
Graft loss	9 (9.8%)	17 (7.4%)	26 (8.1%)
Alive	82 (89.1%)	207 (90.4%)	289 (90.0%)

Note: Categorial data expressed as a number (percentage) and continuous data as median (interquartile ratio).

^a p-Value obtained via the Fisher exact test or Kruskal–Wallis rank sum test.

Abbreviations: DSA−, donor-specific antibody negative; DSA+, donor-specific antibody positive; NA, not applicable.

FIGURE 3.

Probability of death and allograft failure within the first year post-liver transplantation according to increasing levels of preformed donor-specific antibodies. (A) Cumulative incidence of death and allograft failure (death and allograft failure are analyzed as competing events). (B, C, and D) Probability of death and graft failure according to donor-specific antibody status, respectively, cMFI≥1000, cMFI≥5000, and cMFI≥10,000. Abbreviations: DSA−, anti-HLA donor-specific antibodies negative; DSA+, anti-HLA donor-specific antibodies positive; cMFI, cumulative mean fluorescence intensity; HLA, human leukocyte antigen; TX, transplantation.

Preformed DSA were associated with increased incidence of biopsy-proven and treated T-cell–mediated rejection episodes

Within the first year after LT, 68 (21.2%) biopsy-proven rejection episodes occurred in our population, with further 13 (4.0%) clinically-suspected rejection episodes. Biopsy-proven rejection episodes were significantly more frequent in DSA+ than in DSA− patients [28 (30.4%) vs. 40 (17.5%); p=0.01] and were more likely to be treated in DSA+ patients (p=0.03) (Table 3). The severity of rejection [rejection activity index (RAI)], the number of rejection episodes, and the time from LT to the first occurrence of rejection were similar between the 2 groups (Table 3). Treatment for rejection consisted, in most of the occurrences, of the administration of intravenous or oral corticosteroids and/or the increase or modification in the immunosuppressive drug treatment [42 (76.4%) and 9 (16.4%) of the 55 treated rejection episodes, respectively].

TABLE 3.

Rejection episodes within the first year after liver transplantation (time-to-event analysis)

Characteristics	Preformed DSA+	Preformed DSA−	Total	p a
Total cohort, n (%)	92 (28.7%)	229 (71.3%)	321 (100%)	NA
General				0.01
Biopsy-proven rejection	28 (30.4%)	40 (17.5%)	68 (21.2%)
Clinically-suspected rejection	1 (1.1%)	12 (5.2%)	13 (4.0%)
Treatment				0.03
Biopsy-proven rejection
Treated	22 (23.9%)	26 (11.4%)	48 (15.0%)
Non-treated	6 (6.5%)	14 (6.1%)	20 (6.2%)
Clinically-suspected rejection
Treated	1 (1.1%)	6 (2.6%)	7 (2.2%)
Non-treated	0 (0.0%)	6 (2.6%)	6 (1.9%)
Severity				0.28
Rejection activity index	6 (4–6)	5 (4–5)	5 (4–6)
Number of rejection episodes				0.51
0	63 (68.5%)	177 (77.3%)	240 (74.8%)
1	22 (23.9%)	37 (16.2%)	59 (18.4%)
2	5 (5.4%)	12 (5.2%)	17 (5.3%)
3	1 (1.1%)	2 (0.9%)	3 (0.9%)
4	1 (1.1%)	1 (0.4%)	2 (0.6%)

Note: Categorial data expressed as a number (percentage) and continuous data as median (interquartile ratio).

^a p-Value obtained via the Fisher exact test or Kruskal–Wallis rank sum test.

Abbreviations: DSA−, donor-specific antibody negative; DSA+, donor-specific antibody positive; NA, not applicable.

A Kaplan–Meier analysis showed that the graft failure- and rejection-free survival probability was lower in DSA+ patients, as compared with DSA− patients, but this did not reach statistical significance (p=0.15) (Supplemental Figure S1, http://links.lww.com/LVT/B105). The cumulative incidence curves of death, graft failure, and rejection, considered as competing events, showed that 64 (76.2%) of the 84 combined events occurred within the first 3 months post-transplantation, with later events being mostly related to rejection episodes (Supplemental Figure S2A, http://links.lww.com/LVT/B105). The incidence of graft failure or rejection was higher in DSA+ patients with cMFI ≥1000 (Supplemental Figure S2B, http://links.lww.com/LVT/B105) and ≥5000 (Supplemental Figure S2C, http://links.lww.com/LVT/B105) (p=0.03 and p<0.01, respectively), but not with cMFI ≥10,000 (Supplemental Figure S2D, http://links.lww.com/LVT/B105) (p=0.18), although there was a trend for a higher incidence of these events in DSA+ patients. In multivariable ordinal logistic regression analysis, preformed DSA were associated with moderate-to-severe rejection episodes, independently of the recipient’s age and gender, and induction and maintenance immunosuppressive treatment regimens (Supplemental Table S3, http://links.lww.com/LVT/B105).

Preformed DSA were independently associated with an increased risk for biliary complications within the first year following liver transplantation

We first analyzed the rate of overall complications within the first year post-transplantation, taking into account death and graft loss as competing events. LT recipients with preformed DSA experienced higher rates of post-transplant complications within the first year, but this did not reach statistical significance (Table 4 and Supplemental Table S4, http://links.lww.com/LVT/B105). Second, we analyzed biliary complications, which were found to be more frequent in the DSA+ versus DSA− patients (28.3% vs. 19.2%, p=0.08). Although the univariate analysis was not statistically significant, the multivariate cause-specific Cox proportional hazards models showed that preformed DSA were significantly associated with biliary complications (HR 2.26, 95% CI 1.17–4.37, p=0.02), independently of age, gender, donor type, history of previous transplantation, HLA mismatch as assessed by PIRCHE-II score, cold ischemia, and pre-transplant MELD-score (Table 4 and Supplemental Table S5, http://links.lww.com/LVT/B105). We next stratified biliary events into anastomotic and non-anastomotic categories. Anastomotic biliary complications were more frequent in DSA+ in comparison to DSA− patients, although this did not reach statistical significance (92.3% vs. 82.1% of biliary complications, p=0.32). Both anastomotic biliary leaks and stenoses accounted for this observation, with 26.9% versus 22.7% for the former and 65.4% versus 61.4% for the latter. The presence of preformed DSA was independently associated with anastomotic biliary complications (HR 2.07, 95% CI 1.18–3.61, p=0.01) (Table 5 and Supplemental Table S6, http://links.lww.com/LVT/B105). In a sensitivity analysis integrating surgical variables in the subgroup of patients with a deceased-donor liver allograft, the association was still present but did not reach statistical significance (HR 1.73, 95% CI 0.97–3.09, p=0.07) (Supplemental Table S7, http://links.lww.com/LVT/B105). The number of non-anastomotic events was insufficient to support multivariable modeling. Third, we compared the vascular and infectious complications and found no difference between DSA+ and DSA− patients (Table 4 and Supplemental Tables S8, S9, http://links.lww.com/LVT/B105).

TABLE 4.

Overall, biliary, vascular, and infectious complications within the first year after liver transplantation (time-to-event analysis)

Characteristics	Preformed DSA+	Preformed DSA−	Total	p a
Total cohort, n (%)	92 (28.7%)	229 (71.3%)	321 (100%)	NA
Complications (overall), %				0.07
No	50 (54.3%)	149 (65.1%)	199 (62.0%)
Yes	42 (42.7%)	80 (34.9%)	122 (38.0%)
Median time (IQR) to complications (overall)	12.0 (3.0–22.8)	15.0 (1.0–83.3)	13.0 (2.3–70.3)	0.61
Biliary complications				0.08
No	66 (71.7%)	185 (80.8%)	251 (78.2%)
Yes	26 (28.3%)	44 (19.2%)	70 (21.8%)
Median time (IQR) to biliary complications	24.0 (10.0–104.5)	60.0 (12.5–145.5)	40.0 (11.5–128.5)	0.31
Vascular complications				0.20
No	68 (73.9%)	187 (80.3%)	252 (78.5%)
Yes	24 (26.1%)	45 (19.7%)	69 (21.5%)
Median time (IQR) to vascular complications	8.5 (1.0–16.5)	3.0 (0.0–30.0)	4.0 (1.0–27.0)	0.58
Infectious complications				0.08
No	81 (88.0%)	215 (93.9%)	296 (92.2%)
Yes	11 (12.0%)	14 (6.1%)	25 (7.8%)
Median time (IQR) to infectious complications	140.0 (33.5–237.5)	66.0 (23.0–146.8)	96.0 (33.0–202.0)	0.25

Note: Categorial data expressed as a number (percentage) and continuous data as median (interquartile ratio).

^a p-Value obtained via the Fisher exact test or Kruskal–Wallis rank sum test.

Abbreviations: DSA−, donor-specific antibody negative; DSA+, donor-specific antibody positive; IQR, interquartile ratio; NA, not applicable.

TABLE 5.

Description of the 70 biliary complications within the first year after liver transplantation

Characteristics	Preformed DSA+	Preformed DSA−	Total	p a
Type of biliary reconstruction				0.05
Duct-to-duct	20 (76.9%)	41 (93.2%)	61 (87.1%)
Bilio-digestive	6 (23.1%)	3 (6.8%)	9 (12.9%)
Localization of biliary complication				0.60
Intrahepatic	2 (7.7%)	7 (15.9%)	9 (12.9%)
Hilar	0 (0.0%)	0 (0.0%)	0 (0.0%)
Anastomotic stenosis	17 (65.4%)	27 (61.4%)	44 (62.9%)
Anastomotic leakage	7 (26.9%)	10 (22.7%)	17 (24.3%)
Grading of biliary complication (according to Esser et al.)19				0.47
1	1 (3.8%)	0 (0.0%)	1 (1.4%)
2	1 (3.8%)	1 (2.3%)	2 (2.9%)
3a	18 (69.2%)	31 (70.5%)	49 (70.0%)
3b	4 (15.4%)	7 (15.9%)	11 (15.7%)
4	1 (3.8%)	5 (11.4%)	7 (8.6%)
5	1 (3.8%)	0 (0.0%)	1 (1.6%)

Abbreviations: DSA−, donor-specific antibody negative; DSA+, donor-specific antibody positive.

^a p value obtained via Pearson's x₂.

PIRCHE-II score did not predict patient survival, allograft survival, or post-transplant complications

We computed the PIRCHE-II score for 245 patients and donor pairs with available HLA typing data and obtained a mean (SD) PIRCHE-II score of 92.8±41.9 (Supplemental Table S10, http://links.lww.com/LVT/B105). Patient and graft survival did not differ between quartiles of ln(PIRCHE-II) scores, nor did acute and chronic rejection. In univariate and multivariate analyses, the continuous ln(PIRCHE-II) scores were not associated with mortality, graft loss, acute T-cell–mediated rejection, and post-transplant complications.

DISCUSSION

We investigated the impact of preformed DSA on LT outcomes using data from the STCS. To do so, we studied all LT performed over 3 years in the 3 transplantation centers performing LT in Switzerland (Geneva, Bern, Zürich). Our findings indicate that preformed DSA are associated with an increased risk of allograft loss, biopsy-proven T-cell–mediated rejection, and biliary complications within the first year post-LT. These results reinforce the growing body of evidence suggesting that, despite the liver’s unique immunological and structural properties, DSA are also clinically relevant in LT.^{1,5–7,21–23}

Our results demonstrated no significant difference in overall survival between DSA+ and DSA− recipients. However, an increased incidence of allograft failure was observed in patients with preformed DSA and cMFI levels higher than 5000. This aligns with previous studies indicating that high-strength DSA, particularly those exceeding certain cMFI thresholds, can contribute to antibody-mediated damage as well as acute and/or chronic allograft dysfunction.^{15,16,24–26} While the liver has often been considered resistant to immune-mediated rejection due to its inherent tolerogenic properties, our data suggest that preformed DSA above a critical threshold may overcome these protective mechanisms, leading to inferior graft outcomes.^1,27

Preformed DSA were significantly associated with an increased incidence of biopsy-proven and treated T-cell–mediated rejection. Although the overall rate of rejection was 21.2% in our cohort, DSA+ recipients experienced a notably higher rate of rejection episodes compared with their DSA− counterparts. These results are consistent with prior reports demonstrating that preformed DSA can act as immune triggers, exacerbating cellular or humoral rejection responses in adult and pediatric LT recipients.^17,28,29 The mechanisms underlying this phenomenon are not fully understood, but one hypothesis is that DSA may prime the recipient’s immune system, leading to heightened alloimmune activation upon exposure to donor antigens.²⁷ Importantly, our findings indicate that DSA+ recipients were more likely to receive treatment for rejection, reinforcing the notion that DSA+ may contribute to clinically-significant rejection episodes that require intensified immunosuppressive management. The association between preformed DSA and T-cell–mediated was independent of indirect alloreactivity as assessed by PIRCHE-II score, underscoring the intrinsic immunologic impact of DSA.

Our study also provides evidence that preformed DSA are independently associated with biliary complications. A preliminary cross-sectional study by our group including 95 LT recipients suggested a potential association between DSA and biliary complications.³⁰ This finding was however contested by others.³¹ In the present study, multivariate analysis demonstrated a significant association between DSA and overall as well as anastomotic biliary complications, suggesting that DSA could play a role in the pathogenesis of biliary injury. Potential mechanisms include DSA-mediated activation of endothelial cells, complement activation, and microvascular injury in the peribiliary plexus, leading to ischemia of the bile ducts.^13,27 Microthrombotic lesions have also been described in association with DSA, which may predispose to bile duct leaks and strictures. These effects could be more pronounced in the context of ischemia–reperfusion injury, particularly in grafts from DCD donors. Given that biliary complications are a major cause of post-transplant morbidity and graft loss, our findings underscore the need for enhanced screening and possibly pre-emptive strategies in DSA+ LT recipients.

Our findings have several significant clinical implications. First, they highlight the need for systematic pre-transplant DSA screening in LT recipients, as DSA appear to influence both acute rejection risk and allograft function. Second, given the observed association between DSA and biliary complications, closer postoperative monitoring and tailored immunosuppressive strategies may be warranted for DSA+ recipients. It is also important to mention that the rate of deceased donors after cardio-circulatory death (DCD) is relatively low (<10%) in our cohort, reinforcing the relevance of the association of DSA and biliary complications. Third, similar to other organs, our results suggest that cMFI thresholds could be incorporated into risk-stratification models, guiding decision-making regarding immunosuppressive management and potential desensitization protocols.

Future research should aim to refine risk stratification by integrating DSA, cMFI levels, and additional immune profiling tools. Prospective studies evaluating desensitization strategies or modified immunosuppressive protocols for high-risk DSA+ recipients could help mitigate the negative impact of DSA on LT outcomes. Furthermore, mechanistic studies exploring the interplay between DSA, endothelial dysfunction, and bile duct pathology may provide deeper insights into the pathophysiological processes underlying these complications.

It should be acknowledged that this study has some limitations. First, while our cohort was derived from a unique and well-defined, comprehensive, multicenter database, the sample size remains relatively modest, limiting the statistical power to detect certain associations with smaller effect sizes. Second, we did not evaluate the impact of de novo DSA, which have been shown to contribute to chronic rejection and late allograft loss.^24,32,33 Third, histological data on humoral rejection were not available within the STCS, precluding an assessment of this aspect of liver alloimmune injury.²⁰ Finally, the association between preformed DSA and anastomotic biliary complications was sensitive to multivariable model specification and population. The sensitivity analysis incorporating surgical variables was restricted to anastomotic complications in deceased-donor recipients and was not fully comparable to the analysis in the full cohort. We therefore interpret the association between preformed DSA and biliary complications with caution and recommend that prospectively harmonized procurement and surgical data should be included in future studies to analyze biliary complications.

In summary, our study demonstrates that preformed DSAs in LT are associated with increased rejection rates, a higher risk of allograft failure in recipients with high cMFI values, and biliary complications. These findings suggest the need for pre-transplant immunological assessment and individualized post-transplant management strategies. Although overall survival was not affected by the presence of preformed DSA, their impact on early allograft function and complications warrants further investigation. Future studies should focus on optimized immunosuppressive regimens and novel strategies to mitigate DSA-related risks, with the goal of improving long-term outcomes of LT recipients.

MEMBERS OF THE SWISS TRANSPLANT COHORT STUDY

Patrizia Amico, Adrian Bachofner, Vanessa Banz, Sonja Beckmann, Guido Beldi, Christoph Berger, Ekaterine Berishvili, Annalisa Berzigotti, Françoise-Isabelle Binet, Pierre-Yves Bochud, Petra Borner, Sanda Branca, Anne Cairoli, Emmanuelle Catana, Yves Chalandon, Philippe Compagnon, Sabina De Geest, Sophie De Seigneux, Michael Dickenmann, Joëlle Lynn Dreifuss, Thomas Fehr, Sylvie Ferrari-Lacraz, Andreas Flammer, Jaromil Frossard, Déla Golshayan, Nicolas Goossens, Fadi Haidar, Jürg Halter, Christoph Hess, Sven Hillinger, Hans Hirsch, Patricia Hirt, Linard Hoessly, Uyen Huynh-Do, Franz Immer, Nina Khanna, Michael Koller, Angela Koutsokera, Andreas Kremer, Thorsten Krueger, Christian Kuhn, Arnaud L’Huillier, Bettina Laesser, Frédéric Lamoth, Roger Lehmann, Alexander Leichtle, Oriol Manuel, Hans-Peter Marti, Michele Martinelli, Valérie McLin, Katell Mellac, Aurélia Merçay, Karin Mettler, Sara Christina Meyer, Nicolas Müller, Jelena Müller, Ulrike Müller-Arndt, Mirjam Nägeli, Dionysios Neofytos, Jakob Nilsson, Manuel Pascual, Rosmarie Pazeller, David Reineke, Juliane Rick, Fabian Rössler, Silvia Rothlin, Thomas Schachtner, Stefan Schaub, Dominik Schneidawind, Macé Schuurmans, Simon Schwab, Thierry Sengstag, Daniel Sidler, Federico Simonetta, Jürg Steiger, Guido Stirnimann, Ueli Stürzinger, Christian Van Delden, Jean-Pierre Venetz, Jean Villard, Julien Vionnet, Laura Walti, Caroline Wehmeier, and Patrick Yerly.

Supplementary Material

lvt-32-991-s001.docx ^{(1.5MB, docx)}