Objectives: The influence of donor-recipient sex mismatches on long-term graft survival after liver transplant is controversial. In this study, our aim was to characterize the differences in long-term graft outcome after liver transplant in more than 2000 cases with special regard to sex match and mismatch.

Materials and Methods: In this retrospective, single center study of 2144 adult primary liver transplant recipients (median follow-up of 92 months), we analyzed specific long-term graft survival and the effect of different donor and recipient sex combinations (Kaplan-Meier, multivariate regression).

Results: In the 15-year follow-up, female recipients (58.6%) had significantly better graft survival than male recipients did (51.6%, P = .031). Matched and mismatched male-female combinations revealed significant differences (P = .003): a male donor-female recipient combination showed the best 15-year graft survival (61.1%), and a female donor-male recipient combination showed the worst graft survival (48.6%), whereas male-male (53.3%) and female-female combinations (55.6%) were not significantly different (P = .967). Donor age (P < .0001), body mass index
(P = .021), female sex (P = .015), Eurotransplant Donor Risk Index > 1.4 (P < .001), recipients’ age (P < .0001), indication for liver transplant (P < .0001), and kidney function (P = .003) significantly affected graft survival. In the multivariate analysis model, a Eurotransplant Donor Risk Index > 1.4 and impaired kidney function at liver transplant again emerged as significant negative predictors. Female donors and male recipients showed significantly more unfavorable characteristics con­cerning long-term graft survival.

Conclusions: The impressive long-term graft survival benefit of male donor-female recipient versus female donor-male recipient and of male donor-female recipient versus matched groups (male-male, female-female) in liver transplant may be caused by significant differences in donor quality and recipient charac­teristics and may not be related to sex itself.

Key words : Donor-recipient matching, Eurotransplant Donor Risk Index, Gender, Long-term outcome

Introduction

Liver transplant as the standard treatment for end-stage liver disease faces different burdens today versus those shown 2 decades ago. Increasing organ shortage has led to more frequent transplants of extended-criteria donor organs. In the near past, several scores to assess the individual risk of early graft failure have been developed.^1-4 One of these was suggested by Braat and associates to best fit the donor characteristics and resulting risks in the European transplant region.¹

In addition to various factors that were found to be significant predictors of graft outcome, some studies have revealed that donor-recipient sex mismatch (in particular female donor-male recipient) negatively affects graft survival.^5-7 However, other studies have reported either that male-female mismatch had no effect or attributed the outcome difference to other donor factors.^8,9 Specifically in 1 US registry study of more than 28 000 liver transplants, donor quality, as measured by the Donor Risk Index,¹⁰ was suggested to be the real cause of inferior graft outcome in female donor-male recipient combinations.⁸

On the other hand, Sanfey concluded the follo­wing in a concise review of sex aspects in liver and kidney transplant: “We need to be mindful of these differences in relation to gender-specific diseases, hormonal and immunological differences in designing clinical protocols and treatment pathways to improve outcomes in transplant.”¹¹

Most of the few publications on donor-recipient sex mismatch rely on registry data with short follow-up (eg, the above-mentioned United Network for Organ Sharing registry study provides a follow-up of 5 years). Thus, this topic stays controversial, and more observational data are needed. In this study, our aim was to characterize the effect of sex differences on long-term graft survival after liver transplant. This report analyzed more than 2000 cases at a single center with special regard to sex matching and mismatching and sex-specific differences regarding donor quality. A precise analysis of reasons for graft loss according to sex match is also provided.

Materials and Methods

This is a retrospective study of 2144 adult patients who underwent primary orthotopic liver transplant from September 1988 to December 2012 at our center. Median follow-up was 92 months (range, 0-304 mo), with mean follow-up of 108.3 ± 86.4 months. During the study, 2675 full liver transplants were performed. Of these, the following patients were excluded from analysis: 74 who had combined liver-kidney transplant, 11 who had multiorgan transplant, 294 who required repeat transplants, and 152 who were under 18 years old. Data were obtained from our prospectively collected liver transplant database (missing values were acquired by reviewing patient and/or donor medical records). In some cases, recipient or donors datasets were incomplete concerning the respective analysis and thus excluded. In Table 1 and Table 2, the numbers of patients who were included for the particular analyses are shown. Overall data completeness was > 95% for all analyses except those concerning Eurotransplant Donor Risk Index (81.8%). For the Eurotransplant Donor Risk Index calculation, the missing factor was in 89% of cases the latest gamma-glutamyl transpeptidase measurement of the donor, which is essential for calculating this index.

This study was approved by our Institutional Ethical Review Board.

Recipient characteristics
Recipient male-to-female ratio was 1.61:1 (1324 men, 820 women). At primary liver transplant, mean patient age was 49.9 ± 11.1 years. Mean laboratory Model for End-Stage Liver Disease (MELD) score was 17.5 ± 8.5, and median laboratory MELD score was 15 (range, 6-40). Patients were subdivided into several groups according to similar indication for transplant. Overall, alcoholic cirrhosis (23.7%), hepatocellular carcinoma (19.3%; comprising all patients with hepatocellular carcinoma despite the underlying disease), cholestatic or autoimmune diseases (15.0%; including all patients with primary or secondary sclerosing cholangitis, primary or secondary biliary cirrhosis, and autoimmune hepatitis), hepatitis C virus infection (12.9%), and hepatitis B virus infection (8.3%) were the most common indications for liver transplant in our patient cohort. Table 1 displays all recipient characteristics stratified by sex.

Laboratory Model for End-Stage Liver Disease calculation
Laboratory MELD scores were calculated using the pretransplant serum bilirubin level, serum creatinine level, and international normalized ratio according to Kamath and associates.¹² “Quick test” values were converted into the corresponding international normalized ratio using the laboratory’s charge number. Serum bilirubin, international normalized ratio, and serum creatinine values < 1.0 (mg/dL) were set to 1.0 to preclude negative scores. Serum creatinine level was capped at 4.0, and laboratory MELD scores were capped at 40.

Donor characteristics
Donor male-to-female ratio was 1.4:1 (1252 men, 892 women). Mean donor age was 44.4 ± 17.9 years, and median donor age was 45 years (range, 9-90 y). Mean Eurotransplant Donor Risk Index was 1.63 ± 0.43, with median of 1.58 (range, 1.03-5.65).

According to the German legislation, no donations after cardiac death were performed. Causes of brain death were subdivided into 4 groups according to the main categories in the Eurotransplant Donor Risk Index, with 30.7% of donors with trauma, 8.4% with anoxia, 57.8% with cerebrovascular accident, and 3.1% with other. Table 2 shows the main donor characteristics stratified by sex.

Eurotransplant Donor Risk Index calculation
Calculation of Eurotransplant Donor Risk Index was performed as published by Braat and associates1 as follows: Eurotransplant Donor Risk Index = exp {0.960[(0.154 if 40 y ≤ age < 50 y) + (0.274 if 50 y ≤ age < 60 y) + (0.424 if 60 y ≤ age < 70 y) + (0.501 if 70 y ≤ age) + (0.079 if cause of death is anoxia) + (0.145 if cause of death is cerebrovascular accident) + (0.184 if cause of death is other) + (0.411 if donation is after cardiac death) + (0.422 if partial/split) + (0.105 if regional share) + (0.244 if national share)] + [0.010 × (cold ischemia time – 8 h)] + 0.06[(latest gamma-glutamyl transpeptidase level (U/L) – 50)/100] + (0.180 if rescue offer)}.

For Kaplan-Meier estimates, donors were grouped as high and low Eurotransplant Donor Risk Index (≤ 1.4 and > 1.4), according to previous work16 of our group indicating this value as a discriminator of long-term graft survival.

Other variables
Overweight (high body mass index) was defined according to the World Health Organization as body mass index (weight/height²) above 25. The Modification of Diet in Renal Disease (MDRD) formula was used to calculate the estimated glomerular filtration rate. An estimated glomerular filtration rate between ≥ 30 and < 60 mL/min/1.73 m² was considered as moderately impaired renal function; rates < 30 mL/min/1.73 m² were defined severely impaired renal function.¹³

Statistical analyses
Continuous data are shown as median and range or mean ± standard deviation. Graft loss was defined as graft failure and/or patient death. Kaplan-Meier estimates were used to calculate survival curves. Differences in survival curves were compared using log-rank statistics. Multivariate logistic regression modeling with stepwise backward covariate selection was performed to determine the association between different variables and long-term graft survival. Variables that showed a significant effect in univariate analysis (P < .05) were entered in the regression models. Parametric data were compared by unpaired t test for continuous variables or one-way analysis of variance with post hoc intergroup differences tested by Holm-Sidak. Categorical variables were compared by the χ2 test. A P value < .05 was considered statistically significant. All statistical calculations were performed with SPSS software (SPSS: An IBM Company, version 22.0 for Windows, IBM Corporation, Chicago, IL).

Results

Patient and graft survival
Overall survival at 1, 5, 10, and 15 years was 88.2%, 78.0%, 67.7%, and 59.1% for patients and 83.3%, 72.1%, 62,6%, and 54.4% for grafts. Figure 1A and 1B show the sex-specific patient and graft survival results. Kaplan-Meier estimates with log-rank statistics revealed significant differences according to the sex of the recipient (P = .003 for patients; P = .031 for grafts).

Long-term graft survival and matched and mismatched sex combinations
Our matched and mismatched analyses included the following cohorts: 859 male donor to male recipient transplants, 393 male donor to female recipient transplants, 427 female donor to female recipient transplants, and 465 female donor to male recipient transplants. In these 4 groups, Kaplan-Meier estimates revealed a statistically significant difference concerning long-term graft survival (P = .003) (Figure 2). Graft survival at 1, 5, 10, and 15 years was best for male donor to female recipient (84.0%, 76.3%, 67.7%, and 61.1%) followed by female donor to female recipient (83.4%, 75.7%, 64.1%, and 55.6%) and male donor to male recipient (85.0%, 71.9%, 63.0%, and 53.3%). Patients in the female donor to male recipient group had the worst long-term graft survival (79.6%, 65.6%, 55.6%, and 48.6%). Intergroup analyses showed that gender-matched groups (male-male vs female-female) did not differ significantly (P = .967) concerning long-term graft survival, whereas female donor to female recipient versus female donor to male recipient (P = .021) and male donor to male recipient versus female donor to male recipient (P = .009) were significantly different.

Other factors significantly affecting long-term graft survival
Several donor and recipient factors other than sex of donor or recipient proved to affect significantly long-term graft survival. One factor was donor age (P < .0001) and recipient age (P < .0001), with comparisons made at < 30 years old, 30 to 55 years old, and > 55 years old, which demonstrated a highly significant influence on long-term graft survival. In addition, donor body mass index (with comparisons made at body mass index < 25, 25-35, and > 35; P = .021), recipient’s impaired kidney function at time of liver transplant (moderately impaired; P = .003), and indication for transplant of the recipient (P < .0001) were found to be negatively associated with long-term graft survival.

Differences in recipient and donor characteristics according to sex
All examined parameters stratified by recipient sex are shown in Table 1 and by donor sex in Table 2. Male and female recipients differed significantly in almost all examined characteristics. Male recipients were older, had a higher body mass index at transplant, and presented with a higher incidence of being overweight (had higher body mass index). However, female recipients had significantly higher laboratory MELD scores, had a lower estimated glomerular filtration rate, and consecutively also a higher incidence of moderately impaired and severely impaired renal function. Concerning indications for transplant, male recipients had a significantly higher incidence of hepatocellular carcinoma, hepatitis B virus infection, and alcohol-induced cirrhosis, whereas female recipients had significantly higher incidence of cholestatic or autoimmune diseases, acute liver failure, and cryptogenic cirrhosis.

On the other hand, female donors were sig­nificantly older (48.1 ± 17.6 y vs 41.6 ± 17.6 y) and cause of death by cerebrovascular accident was significantly more frequent and death by trauma significantly less often shown. This resulted in a significantly higher Eurotransplant Donor Risk Index score (1.72 ± 0.43 vs 1.56 ± 0.42) compared with male donors.

As shown in Table 3, the Eurotransplant Donor Risk Index scores in the 4 donor-recipient com­binations were also significantly different (P < .001, analyses of variance). Intergroup analyses again revealed statistically significant differences of Eurotransplant Donor Risk Index score for all combinations (P ≤ .001) except for the female donor-female recipient versus the female donor-male recipient combination (P = .836, post hoc Holm-Sidak analyses).

Multivariate analyses
In the multivariate model, Eurotransplant Donor Risk Index > 1.4 (P < .0001; hazard ratio of 1.658; 95% confidence interval, 1.403-1.959) and impaired kidney function at liver transplant (P = .005; hazard ratio of 1.321; 95% confidence interval, 1.089-1.602) emerged as significant negative predictors of long-term graft survival, whereas the other factors from the univariate analyses were not significantly associated with worse long-term graft survival.

Reasons for graft loss
Table 4 displays the reasons for graft loss according to donor-recipient sex match. Reasons for graft loss were assigned to 13 categories. The most frequent reason for graft loss (excluding category titled “others,” which comprised many different reasons with low frequency) in male-male liver transplant combinations was de novo malignancy, followed by infection and by hepatocellular carcinoma recurrence. In male donor-female recipient liver transplant, infection was most frequent, followed by cardio­vascular disease, and then by de novo malignancy. In female donor-female recipient liver transplant, again cardiovascular disease and infection were included in the 3 most frequent reasons for graft loss; however, interestingly, initial nonfunction was the third most frequent reason for graft loss. Finally in the female donor-male recipient liver transplant group cardio­vascular disease, followed by infection, and then hepatocellular carcinoma recurrence were the most frequent reasons for graft loss.

Another interesting result is the relatively high incidence of hepatic artery thromboses in the female donor-male recipient group.

Discussion

In the present study, we showed that female donors were significantly older, died significantly more frequently from cerebrovascular causes and less frequent by trauma, and thus had a significantly higher Eurotransplant Donor Risk Index score. Given this and keeping in mind the higher incidence of hepatocellular carcinoma in the male recipient group and the higher incidence of cholestatic or autoimmune diseases (patients with the best achievable long-term outcomes¹⁴) in the female recipients, the finding of better outcomes in male donor-female recipient versus female donor-male recipient transplants is not surprising. The better long-term graft survival of male donor-female recipient versus male donor-male recipient transplants may be explained by the significantly different indications and their effects on long-term graft (and patient) survival,¹⁴ because the differences in survival curves develop approximately 2 to 3 years after transplant when the early effects of donor quality have faded and the effect of the primary indication for liver transplant becomes more important. This is also supported by the analysis of reasons for graft loss. In male recipients (for both male donor-male recipient and female donor-male recipient combinations), hepatocellular carcinoma recurrence as reason for graft loss was 1 of the 3 most frequent reasons.

Interestingly, in the female donor-male recipient combination, 10.2% of all graft loss was caused by hepatic artery thromboses (compared with < 3 % in all other donor-recipient pairs). Whether the thromboses were caused by technical or immuno­logic problems cannot be answered by this study; however, this would be of interest for further analyses. If one assumes technical reasons and excludes these cases, the main reasons for graft loss do not change significantly (data not shown). On the other hand, acute or chronic rejection as reason for graft loss was more frequent in female recipients regardless of whether the donor was male or female.

This study has some strengths and limitations: Concerning the limitations, this is a single-center retrospective study that spans over many years when organ acceptance and utilization criteria have changed; therefore, some of the significance of the findings may be subject to bias. On the other hand, this is the largest single center study on this topic with the longest follow-up presented so far.

When we reviewed the literature on the studied topic, we found an impressive shortage of available publications. Most of the few papers available are at least 8 to 10 years old. In a more recent published paper by Hibi and associates about prognostic factors of simultaneous liver-kidney transplant, a female donor was identified as negative predictor of graft and patient survival. Unfortunately, no precise explanation was given.¹⁵

The findings of the presented study are different from some previous studies that reported female donor-male recipient mismatch as a significant risk factor for graft failure. One of the main reasons is that these reports did not consider the multiple other donor risk factors that are included in the Eurotransplant Donor Risk Index or Donor Risk Index.^6,7 On the other hand, Croome and associates suggested that female donor-male recipient hepatic transplant is associated with statistically significant poorer graft survival than other donor-recipient combinations despite adjust­ment for donor risks and recipient variables.⁵ In the present study, we attribute the inferior long-term graft survival in female donor-male recipient liver transplant to reduced female donor quality and unfavorable characteristics in male recipients. We thereby reconfirm some of the midterm (5 year) outcome registry data findings of Lai and associates8 in a large long-term single-center analysis and additionally provide reason for graft loss data.

In conclusion, it seems that the reason for better graft survival in male donor-female recipient versus female donor-male recipient and male-male and female-female matched liver transplant is not sex of the donor or recipient itself but an accumulation of negative factors. We can only speculate why trans­plant surgeons and physicians in this retrospective analysis tended to accept organs from female donors that had these negative factors and thus were of worse quality than organs from male donors. One reason might be that the Eurotransplant Donor Risk Index is a relatively new and not yet implemented tool in organ allocation. Further multicenter data should be acquired to elucidate this issue.

References:

1. Braat AE, Blok JJ, Putter H, et al. The Eurotransplant donor risk index in liver transplantation: ET-DRI. Am J Transplant. 2012;12(10):2789-2796.
   CrossRef - PubMed
2. Dutkowski P, Oberkofler CE, Slankamenac K, et al. Are there better guidelines for allocation in liver transplantation? A novel score targeting justice and utility in the model for end-stage liver disease era. Ann Surg. 2011;254(5):745-753.
   CrossRef - PubMed
3. Ioannou GN. Development and validation of a model predicting graft survival after liver transplantation. Liver Transpl. 2006;12(11):1594-1606.
   CrossRef - PubMed
4. Tekin K, Imber CJ, Atli M, et al. A simple scoring system to evaluate the effects of cold ischemia on marginal liver donors. Transplantation. 2004;77(3):411-416.
   CrossRef - PubMed
5. Croome KP, Segal D, Hernandez-Alejandro R, Adams PC, Thomson A, Chandok N. Female donor to male recipient gender discordance results in inferior graft survival: a prospective study of 1,042 liver transplants. J Hepatobiliary Pancreat Sci. 2014;21(4):269-274.
   CrossRef - PubMed
6. Marino IR, Doyle HR, Aldrighetti L, et al. Effect of donor age and sex on the outcome of liver transplantation. Hepatology. 1995;22(6):1754-1762.
   CrossRef - PubMed
7. Rustgi VK, Marino G, Halpern MT, Johnson LB, Umana WO, Tolleris C. Role of gender and race mismatch and graft failure in patients undergoing liver transplantation. Liver Transpl. 2002;8(6):514-518.
   CrossRef - PubMed
8. Lai JC, Feng S, Roberts JP, Terrault NA. Gender differences in liver donor quality are predictive of graft loss. Am J Transplant. 2011;11(2):296-302.
   CrossRef - PubMed
9. Zeier M, Dohler B, Opelz G, Ritz E. The effect of donor gender on graft survival. J Am Soc Nephrol. 2002;13(10):2570-2776.
   CrossRef - PubMed
10. Feng S, Goodrich NP, Bragg-Gresham JL, et al. Characteristics associated with liver graft failure: the concept of a donor risk index. Am J Transplant. 2006;6(4):783-790.
    CrossRef - PubMed
11. Sanfey H. Gender-specific issues in liver and kidney failure and transplantation: a review. J Womens Health (Larchmt). 2005;14(7):617-626.
    CrossRef - PubMed
12. Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33(2):464-470.
    CrossRef - PubMed
13. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Ann Int Med. 2003;139(2):137-147.
    CrossRef - PubMed
14. Schoening WN, Buescher N, Rademacher S, et al. Twenty-year longitudinal follow-up after orthotopic liver transplantation: a single-center experience of 313 consecutive cases. Am J Transplant. 2013;13(9):2384-2394.
    CrossRef - PubMed
15. Hibi T, Sageshima J, Molina E, et al. Predisposing factors of diminished survival in simultaneous liver/kidney transplantation. Am J Transplant. 2012;12(11):2966-2973.
    CrossRef - PubMed
16. Schoening W, Helbig M, Buescher N, et al. Eurotransplant Donor-Risk-Index and recipient factors: Influence on long-term outcome after Liver Transplantation. - A large single center experience. Clin Transplant. 2016 Feb 8. doi: 10.1111/ctr.12714. [Epub ahead of print].