Liver Transplantation in Recipients With a Positive Crossmatch: A Retrospective Single-Center Match-Pair Analysis

A positive crossmatch (XM+) is considered a contraindication to solid abdominal organ transplantation except liver transplantation (LT). Conflicting reports exist regarding the effects of XM+ on post-transplant outcomes. The goal of this retrospective single-center analysis is to evaluate the influence of XM+ on relevant outcome parameters such as survival, graft rejection, biliary and arterial complications. Forty-nine adult patients undergoing LT with a XM+ between 2002 and 2017 were included. XM+ LT recipients were matched 1:2 with crossmatch negative (XM−) LT recipients based on the balance of risk (BAR) score. Patient and graft survival were compared using Kaplan-Meier survival analysis and the log-rank test. Comparative analysis of clinical outcomes in XM+ and XM− groups were conducted. Patient and graft survival were similar in XM+ and XM− patients. Rejection episodes did not differ either. Recipients with a strong XM+ were more likely to develop a PCR+ CMV infection. A XM+ was not associated with a higher incidence of biliary or arterial complications. Donor age, cold ischemia time, PCR+ CMV infection and a rejection episode were associated with the occurrence of ischemic type biliary lesions. A XM+ has no effects on patient and graft survival or other relevant outcome parameters following LT.


INTRODUCTION
A positive crossmatch (XM+) is usually considered a contraindication to all solid abdominal organ transplantations except liver transplantation (LT) (1,2). Therefore, crossmatch testing is mandatory before pancreas, intestinal and kidney transplantation (3). However, in the context of LT the effect of a XM+ on post-transplant outcomes remains ill-defined and LT is commonly performed regardless of the crossmatch testing results, often even before these results become available (3)(4)(5)(6)(7).
Compared to other abdominal organs, the liver seems to be in a privileged immunological situation due to its dual afferent blood supply, its unique antigenic sinusoidal vasculature line by Kupffer cells and its ability to absorb preformed donor specific antibodies (DSAs) by secreting soluble antigens (8)(9)(10). Reports of combined liver and kidney transplantations in the presence of a XM+ in which the recipient became XM− within hours following transplantation underline the liver's impressive immunologic capabilities (9,11,12).
Still, some authors suggest a link between inferior patient and graft survival and a higher rate of postoperative complications following LT in the presence of a XM+ (8,(13)(14)(15)(16)(17)(18). Others, however, were not able to duplicate those findings (6,10,12,(19)(20)(21)(22)(23)(24)(25). Yet, focusing on a XM+ alone might not tell the full story as XM strength (26) and type (T cell vs. B cell) may play a significant role concerning posttransplant outcomes (3,5,17,18). Fittingly, a T cell but not B cell dependent XM+ was reported to be associated with impaired graft survival (3). Historically, LT was essentially an emergency surgical procedure in order to keep cold ischemia time (CIT) short. While it seemed unthinkable to postpone a LT until crossmatch testing results become available only a few years ago, the advent of machine perfusion has changed clinical practice (27). Machine preservation offers the possibility to optimize transplant conditions including immunologic risk stratification pre-transplant. Considering these implications, it seems worthwhile to explore whether a XM+ influences post-transplant outcomes. Previous studies on this subject were hampered by a small number of patients and mostly lacked adequate controls and comparisons (15,16,26,28,29).
The aim of this match-pair analysis is to evaluate the influence of a XM+, including XM strength and type, on relevant clinical outcome parameters such as patient and graft survival, rejection episodes, biliary and arterial complications.

Study Population and Study Design
At the Medical University of Innsbruck, crossmatch testing is routinely performed for LT recipients. All adult patients who underwent XM+ deceased donor LT from donation after brain dead (DBD) donors between 2002 and 2017 were included. A 1:2 match-pair analysis was conducted, with patients who underwent LT with a negative crossmatch (XM−) serving as controls. Matching was performed based on the balance of risk (BAR) score (30,31).
The study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board; protocol code 1034/2022. The results were reported according to the STROBE guidelines (32).
Immunosuppression and Postoperative Care levels 6-8 ng/mL, gradually decreased to 6 ng/mL at 6 months, and 4-6 ng/mL at 12 months) and either mycophenolate mofetil (MMF) (1,000 mg twice daily) or mycophenolic acid (MPA) (720 mg twice daily). Steroids were gradually tapered to 5 mg prednisolone per day as part of the maintenance therapy. Complete steroid withdrawal was considered on an individual basis considering the side effect profile as well as the patient's immunologic risk. Reasons to divert from our standard protocol were related to recipient factors. Conversion from Tac to cyclosporine A (CsA) was considered in case of long-QT syndrome, or tacrolimus associated neurotoxicity. MMF/MPA was switched to azathioprine (Aza) in case of gastrointestinal side effects or to avoid the teratogenic potential in female patients wishing to conceive.

Definitions
Crossmatch All recipient sera were tested for cytotoxic antibodies against donor lymphocytes (CDC crossmatch). For the XM to be deemed positive more than 15% cytolysis had to be present. Additionally, a XM was defined as weakly positive when cytolysis ranged between 15% and 50% and strongly positive when cytolysis exceeded 50%. Cytotoxic cross-matching activity was tested before and after treatment with dithiothreitol (DTT) which inactivates IgM antibodies (35,36). For XM strength analysis the post DTT treatment value was employed. In addition to XM strength, the XM type (T cell dependent vs. B cell dependent) was recorded.

Graft Loss and Graft Dysfunction
Graft loss was defined as patient death or the need for liver retransplantation. Primary non-function was defined as peak AST ≥3000 IU/L plus at least one of the following criteria: INR ≥2.5, serum lactate ≥4 mmol/L and total bilirubin ≥10 mg/dL (values measured on postoperative day 3, biliary obstruction being excluded). Early allograft dysfunction (EAD) was defined according to the Olthoff criteria (37).

Rejections
Acute rejection was defined as biopsy proven rejection which required steroid bolus treatment (38). Steroid bolus treatment consisted of an intravenous steroid pulse of 500 mg methylprednisolone for three consecutive days. Chronic rejection was defined based on persistent laboratory abnormalities and histological confirmation (38).

Biliary Complications
Biliary complications were classified as bile duct leaks, biliary cast syndrome, anastomotic stenosis (AS) and non-anastomotic stenosis (NAS). Ischemic type biliary lesions (ITBL) were defined as NAS with or without biliary cast formation in the absence of hepatic artery stenosis or thrombosis (39)(40)(41).

Extended Criteria Donors
ECDs were defined according to the Eurotransplant Manual, Chapter 9: The Donor (42).

Outcomes
The primary outcome was patient and graft survival. Secondary outcomes included incidence and risk factors for rejection episodes as well as incidence, risk factors and type of biliary and arterial complications.

Recipient Characteristics
Forty-nine patients undergoing LT with a XM+ were matched 1: 2 with XM− patients. Matching was performed based on the BAR score. The indications for LT and recipient demographics are presented in

DISCUSSION
This analysis comparing XM+ and XM− LT recipients over the course of a 16-year period demonstrated that a XM+ has no obvious effects on patient and graft survival and does not appear to influence any of the relevant clinical outcome parameters following LT such as rejection episodes, biliary or arterial complications. Furthermore, neither XM type nor strength had any influence on post-transplant outcomes. Known risk factors for XM+ are female recipient sex, previous LT as well as immunologic indications for LT such as autoimmune hepatitis (AIH) (6,14,24). In contrast to an analysis by Ruiz et al. (6), patients with AIH were not at risk for a XM+ in our study. Considering that only four patients in our cohort underwent LT for AIH this finding needs to be viewed cautiously. However, similar to results reported by Ruiz et al. and others (8,13,24,44,45), we found a higher number of female recipients and re-transplantations in the XM+ group; attributable to previous pregnancies, blood transfusions during or in the aftermath of the primary transplant operation and sensitization caused by the initial graft itself. We also found the recipient BMI to be lower in XM+ recipients, which is in accordance with the finding that the XM+ group encompassed more female recipients.
A high rate of antibody induction (61.9%) was observed in the study cohort. This can be explained by the fact that our center took part in two IL2 antibody induction studies (PROTECT (33) and DIAMOND (34)) during the study period. While the overall antibody induction rate did not differ between XM+ and XM− negative patients, XM+ patients were more likely to receive alemtuzumab (although the absolute number was small). Interestingly, XM strength did not correlate with the use of antibody induction. However, XM strength did correlate with subsequent PCR+ CMV infections.
Overall, the number of rejection episodes was similar between our XM+ and XM− recipients. Previous studies have reported higher rejection rates in XM+ recipients (13,14,17,46,47 (14,17,46,47), and most of these studies were conducted decades ago when IS regimens were less intensive with lower CsA and Tac target levels. While originally reporting a higher complication rate in recipients with a XM+ Takaya et al. showed, in a follow-up study, that comparable outcomes can be achieved with a more intense IS regimen (48). The more intense IS regimen used in the follow-up study constitutes the standard IS regimen today at most transplant centers including ours (45). This might explain why, in more recent studies with more intense IS regimens, a XM+ had no influence on the occurrence of rejection episodes, patient and graft survival as well as overall complications (24,25,45,49), which is in accordance with our observations. To the contrary: in a recent study by Ünlü et al. (50) LT recipients perceived to be at an increased immunologic risk received more intense IS leading to higher infectious complications without providing any graft or patient survival benefit. Considering the liver's privileged immunologic status, a more intense IS for XM+ recipients might be unnecessary and even harmful. Accordingly, when analyzing their 20-year experience with XM+ LT recipients Ruiz et al. (6) found no association between a XM+ and graft complications as well as patient and graft survival. Compared to previous studies (44,51,52), we were unable to find any association between a XM+, including XM strength and type, and the occurrence of biliary complications. Unsurprisingly, patients with biliary complications had a higher graft loss rate and patients with ITBL required re-transplantation more often. ITBL remain one of the most worrisome complications following LT. Immunologic factors have been implicated in the pathogenesis of ITBL in addition to ischemia reperfusion injury and bile salt toxicity (39). While a XM+ had no influence on ITBL development in our study, allograft rejection as well as a positive CMV PCR were associated with an increased risk for the development of ITBL in uni-and multivariate analysis; as were older donor age and prolonged CIT, both well known risk factors for the development of ITBL. Furthermore, XM strength was positively associated with subsequent PCR+ CMV infections. Previous clinical studies have shown acute rejection and active CMV replication to be immunologic risk factor for the development of biliary complications in the context of LT (53)(54)(55)(56).
Interestingly, a PCR+ CMV infection in immunocompromised HIV positive patients has been known to cause destruction in the biliary tree for a long time, a condition termed AIDS cholangiopathy (57). In a study examining the effects of a CMV infection on rat liver grafts Martelius et al. provided experimental data supporting the role of CMV in the pathogenesis of bile duct injury (58). CMV infection leads to upregulation of MHC antigens and expression of vascular adhesion molecules such as VCAM-1 and ICAM-1 through secretion of pro-inflammatory cytokines (58,59). Similarly, allograft rejection is thought to induce an inflammatory state at the local level leading to endothelial injury (60,61). Since viability of the biliary tree depends on the oxygen rich arterial blood supply, an immune-mediated micro-vasculopathy may result in ischemic type injury to the bile ducts, providing a possible pathophysiological explanation for our findings (52,62,63).

Strengths and Limitations
The study compared XM+ with XM− LT using a 1:2 match-pair design. Matching was performed based on the BAR score which has shown to correlate best with post-transplant outcomes compared to other published risk scores (30,31). Strengths of our study include the prospectively maintained LT database at our center, the match-pair analysis and the relatively long followup. Limitations of the present study include the retrospective design and a possible bias concerning the selection of participants beyond the data displayed in the demographics. Despite performing a match-pair analysis in order to guarantee a homogenous comparison group, differences in donor and recipient characteristics did exist between the XM+ and XM− group. The donor BMI was significantly lower, and anhepatic as well as WIT were significantly shorter in the XM+ group compared to the XM− group. This may introduce a bias as a lower donor BMI and shorter ischemia times could imply favorable outcomes. Furthermore, the recipients' MELD score was found to be higher in the XM+ group. However, the BAR score which, among other factors, includes the MELD score and correlates with relevant outcome parameters following LT was used for match-pair analysis to mitigate potential biases. None of these factors had any significant influence on patient or graft survival in our cohort when performing univariate Cox proportional hazards regression analysis (Supplementary Tables S1, S2) as well as when adjusting for these differences in baseline characteristics in a multivariate Cox regression model (Supplementary Tables  S3, S4). Also, University of Wisconsin (UW) solution was more commonly used than Histidine-Tryptophan-Ketoglutarate (HTK) solution as a preservation solution in the XM+ group. UW used to be the gold standard for static cold storage perfusion of liver grafts but preservation with HTK is reported to be clinically equivalent (64,65). Concerns regarding the higher viscosity of UW leading to an incomplete flush of the peribiliary glands and an increase in ITBL have been voiced. However, these concerns have not materialized (66). Moreover, the type of preservation solution had no significant influence on the development of ITBL in our recipients in univariate binary logistic regression analysis (Supplementary Table S5).

CONCLUSION
In the present era of LT, a XM+ has no effects on graft and patient survival as well as postoperative complications. Therefore, our center policy will not change, and we will continue to transplant patients without waiting for XM testing results despite the logistical possibilities offered by the advent of normothermic machine perfusion. A PCR+ CMV infection was more likely to occur in recipients with a strongly positive XM. Together with allograft rejection, donor age and CIT, a PCR+ CMV infection was among the strongest independent predictor for the development of ITBL. Patients with ITBL had higher retransplantation rates than patients without ITBL.

DATA AVAILABILITY STATEMENT
Data is available upon reasonable request form the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethikkommission der Medizinischen Universität Innsbruck. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
FJK: Study design, data acquisition, data analysis, interpretation of the data, drafting and revising the manuscript; MF: Data acquisition, interpretation of the data, drafting and revising the manuscript; FM: Interpretation of the data, drafting and revising the manuscript; AB: Data acquisition, interpretation of the data, drafting and revising the manuscript; AV: Data acquisition, interpretation of the data, drafting and revising the manuscript; BC: Interpretation of the data, drafting and revising the manuscript; TR: Interpretation of the data, drafting and revising the manuscript; MM: Interpretation of the data, drafting and revising the manuscript; CM: Interpretation of the data, drafting and revising the manuscript; MR: Study design, data analysis, interpretation of the data, drafting and revising the manuscript; HU: Study design, data analysis, interpretation of the data, drafting and revising the manuscript; DÖ: Interpretation of the data, drafting and revising the manuscript; RO: Interpretation of the data, drafting and revising the manuscript; SS: interpretation of the data, drafting and revising the manuscript; AW: Study design, data acquisition, data analysis, interpretation of the data, drafting and revising the manuscript.