CLINICAL
AND
TRANSLATIONAL RESEARCH
Low-Serum Testosterone Levels Pre-Liver Transplantation Are Associated With Reduced Rates of Early Acute Allograft Rejection in Men Marie Sinclair,1,2,4 Peter W. Angus,1,2 Paul J. Gow,1,2 Rudolf Hoermann,2 Tamara Mogilevski,1 and Mathis Grossmann2,3 Background. Low pretransplant serum testosterone has recently been associated with increased mortality in men awaiting liver transplantation, but the potential impact on rejection has not yet been investigated. Methods. Pretransplantation serum testosterone, SHBG, and other variables were collected on 190 consecutive men who received a liver transplant between 2007 and 2013. Rates of subsequent acute cellular rejection were recorded. Multivariable analysis was performed to define variables associated with rejection and other clinically important end points. Results. Thirty (16%) of 190 men experienced acute cellular rejection. Lower pretransplant testosterone was associated with lower rejection rates, j7% (95% confidence interval [CI], j2% to j12%) per nmol/L decrease in total testosterone and j4% (95% CI, j0.5% to j7%) per 10 pmol/L decrease in free testosterone. Total testosterone (correlation 0.29, P=0.04) and free testosterone (correlation 0.37, P=0.01) correlated significantly with the histological severity of rejection. Older age at transplant (+5% [95% CI, 9%Y2%]) per year, and nonautoimmune etiology of liver disease (OR, 1.0 for autoimmune, 0.22 [95% CI, 0.07Y0.73] for hepatitis C virus, and 0.58 [95% CI, 0.21Y1.71] for other etiologies) were also associated with decreased rejection risk. In a generalized linear model including the covariates testosterone, SHBG, age, etiology, and MELD, total testosterone remained a significant predictor of rejection (adjusted OR, 1.06; P=0.03), as did age at transplant (OR, 0.95; P=0.01). Conclusion. Low preliver transplant serum testosterone independently predicts a decreased risk of acute allograft rejection. Whether testosterone is a marker of disease-associated immune dysfunction or has immune-modulatory effects requires further study. Keywords: Liver transplantation, Rejection, Testosterone. (Transplantation 2014;98: 788Y792)
urvival rates after orthotopic liver transplantation currently exceed 85% at 12 months (1). Acute cellular rejection, however, remains a frequent problem, with approximately 25% of patients experiencing an episode within 3 months of transplant (2, 3). Prevention and treatment of rejection requires high doses of immunosuppressive medications that are associated with significant side effects. Thus, there is major clinical interest in identifying pretransplant
S
The authors declare no funding or conflicts of interest. 1 Victorian Liver Transplant Unit, Austin Health, Heidelberg, Victoria, Australia. 2 Department of Medicine, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia. 3 Endocrine Unit, Austin Health, Heidelberg, Victoria, Australia. 4 Address correspondence to: Marie Sinclair, Liver Transplant Unit, Austin Health, 145 Studley Road, Heidelberg, Victoria, Australia 3084. E-mail: [email protected] T.M. and M.S. contributed in data collection. M.S., M.G., P.G., and P.A. contributed in writing, reviewing, and editing of the manuscript. R.H. contributed in the statistical analysis and refining of results. Received 5 December 2013. Revision requested 30 December 2013. Accepted 19 February 2014. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0041-1337/14/9807-788 DOI: 10.1097/TP.0000000000000130
788
www.transplantjournal.com
factors that predict rejection, to individualize the use of immunosuppressive agents according to risk. Immunity is impaired in cirrhosis, and this has been attributed to defects in polymorphonucleocyte function, macrophage activity, and antibody production as well as complement deficiency (4). A functioning immune system is necessary for the development of acute cellular rejection. Therefore, sicker patients who have more severely impaired immunity may be less likely to experience rejection posttransplant (5). Low testosterone levels are found in up to 90% in men awaiting transplantation and are related to the severity of the underlying liver disease (6 Y8). The pathogenesis is multifactorial and involves suppression of the hypothalamicpituitary-testicular (HPT) axis at multiple levels (7, 9). In men awaiting liver transplantation, testosterone may be a more sensitive marker of ill health than MELD. Indeed, in a recent study, we found that low serum testosterone predicts mortality independently of MELD score and of hyponatremia (6). This suggests that serum testosterone may more closely reflect systemic illness and hence immune impairment than traditional markers of liver disease severity. In the current study, we tested the hypothesis that low pretransplant testosterone levels are associated with a Transplantation
& Volume 98, Number 7, October 15, 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Sinclair et al.
* 2014 Lippincott Williams & Wilkins
decreased risk of acute allograft rejection postliver transplant. We also hypothesized that this is independent of conventional markers of liver disease severity.
RESULTS Baseline Characteristics Baseline characteristics of the 190 consecutive men included in the study are shown in Table 1. Pretransplant assessment, which included routine testosterone studies, was performed at a median of 6 months before transplant (IQR, 3Y11). Median age at time of transplant was 53 years (IQR, 46Y58), and MELD score 19 (13Y24). Twenty-four patients (12.6%) were transplanted for autoimmune spectrum liver diseases, 83 for hepatitis C (43.7%), and 83 (43.7%) for other etiologies. Fifteen men (7.9%) had previously received a liver transplant. Seventy-five (39.5%) patients were prescribed cyclosporine as the calcineurin inhibitor component of the immunosuppressive regimen. The median total testosterone level was at 8.8 nmol/L (3.6Y14.8), and 108 patients (57%) had a low total testosterone (less than 10 nmol/L). The median free testosterone was 106 pmol/L (50Y176) and below the normal range (G230 pmol/L) in 165 patients (87%). Both total and free testosterone levels negatively correlated with MELD score (correlation, j0.28 and j0.24, respectively; PG0.0001). The median SHBG was elevated at 73 nmol/L (53.4Y97.5) (normal range, 13Y71nmol/L). There was a significant negative correlation between SHBG and MELD score (correlation, j0.18; P=0.0004). The median length of stay in hospital after liver transplantation was 16 days (12Y23). Median days spent in ICU were 3 (2Y5), with a median of 13 inpatient days in the TABLE 1.
789
general ward subsequently (9Y19). The median follow-up time for patients was 2.9 years (1.2Y4.7), with a range of 3 months to 10 years. Graft loss was seen in 11 patients (5.8%) over this time, and 22 patients died (11.6%). Acute Cellular Rejection Acute cellular rejection was diagnosed and treated in 30 patients (16%) within 3 months of liver transplantation. Men who experienced early rejection were significantly younger than those who did not (48.5 years compared with 53 years, respectively, P=0.008). Approximately 29.2% of men with liver disease due to autoimmune disease experienced rejection, with lower incidence in those with hepatitis C (8.4%) or with other etiologies (19.3%), P=0.025. The median total testosterone level in men experiencing early acute cellular rejection was 12.5 nmol/L (9.0Y19.2), which was significantly higher than that in men without rejection, 7.8 nmol/L (3.3Y13.8), P=0.003. Similarly, median free testosterone was 155 pmol/L (105Y243) in patients experiencing early rejection, significantly higher than that in men without rejection, 95pmol/L (48Y167), P=0.006. Total testosterone (correlation 0.29, P=0.04) and free testosterone levels (correlation 0.37, P=0.01) correlated significantly with severity of rejection as measured by the rejection activity index (RAI). By contrast, there was no significant association between rejection and pretransplant MELD score (P=0.33) or history of previous transplant (P=1.0). The rejection rate was lower in cyclosporine versus noncyclosporine users (6.7% vs. 21.7%, P=0.005); cyclosporine use did not retain significance after adjustment. Multivariable analysis for the outcome of transplant rejection was performed incorporating the pretransplant
Baseline demographics
Etiology n (%): Autoimmune spectrum HCV Other Median age at transplant (IQR) Median MELD (IQR) Median Na (IQR) Median total testosterone (IQR) Median SHBG (IQR) Median free testosterone (IQR) Median albumin (IQR) Median LH (IQR) Median FSH (IQR) Median creatinine (IQR) Median vitamin D (IQR) Median days in ICU posttransplant (IQR) Median days on ward posttransplant Median inpatients days posttransplant and IQR Died during follow-up: n (%) Graft loss during follow-up n (%): Years of follow-up posttransplant (IQR)
Total
No rejection
Rejection
P value
24 (12.6%) 83 (43.7%) 83 (43.7%) 53.0 (46.2Y58.0) 19.0 (13.0Y24.8) 136 (132Y139) 8.8 (3.6Y14.8) 73.0 (53Y98) 106 (50Y176) 28.0 (24.0Y34.0) 5.0 (2.8Y7.4) 6.2 (3.3Y9.7) 89.0 (71.5Y119.0) 56.0 (39.2Y82.0) 3.0 (2.0Y5.0) 13.0 (9.0Y19.0) 16.0 (12.0Y23.0) 22 (11.6%) 11 (5.8%) 2.9 (1.2Y4.7)
17 (10.6%) 76 (47.5%) 67 (41.9%) 53.0 (47.0Y58.2) 18.5 (14.0Y25.0) 135 (132Y139) 7.8 (3.3Y13.8) 70 (50Y96) 95 (48Y167) 28.0 (23.5Y33.0) 5.0 (2.8Y7.3) 5.8 (3.3Y9.7) 95.0 (72.2Y121.5) 55.0 (39.0Y79.0) 3.0 (2.0Y5.0) 12.0 (8.5Y19.0) 15.0 (11.0Y23.0) 18 (11.2%) 9 (5.6%) 2.9 (1.2Y4.8)
7 (23.3%) 7 (23.3%) 16 (53.3%) 48.5 (38.0Y54.8) 19.0 (10.5Y22.0) 137 (136Y140) 12.5 (9.0Y19.2) 86 (66Y116) 155 (105Y243) 29.0 (26.0Y38.0) 5.7 (3.8Y8.9) 7.3 (3.6Y11.4) 78.0 (69.0Y98.0) 59.0 (46.0Y104.0) 3.0 (2.0Y5.0) 15.0 (13.0Y17.0) 18.0 (15.2Y26.0) 4 (13.3%) 2 (6.7%) 2.8 (1.1Y3.9)
0.079
P values given were corrected for multiple testing by the Benjamini-Hochberg method.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
0.048 0.42 0.063 0.048 0.085 0.048 0.21 0.21 0.38 0.13 0.38 0.59 0.081 0.13 0.76 0.73 0.59
790
www.transplantjournal.com
TABLE 2.
Transplantation
& Volume 98, Number 7, October 15, 2014
Increased risk of acute cellular rejection according to testosterone levels and covariates
Parameter TT (nmol/L) cFT (pmol/L) SHBG (nmol/L) MELD Age (yr) Disease etiology
Odds ratio (95% CI) 1.07 (1.02Y1.12) 1.004 (1.0005Y1.007) 1.01 (1.002Y1.022) 0.97 (0.92Y1.02) 0.95 (0.91Y0.98) Autoimmune 1.0 HCV 0.22 (0.07Y0.73) Other 0.58 (0.21Y1.71)
P value
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
P value
0.04 0.02 0.045 0.28 0.05 V 0.01 0.30
1.06 (1.004Y1.13) b1.007 (0.99Y1.02) 0.99 (0.92Y1.05) 0.95 (0.91Y0.99) 1 1.0 2 0.25 (0.07Y0.97) 3 0.65 (0.18Y2.44)
0.03 V 0.30 0.65 0.01 V 0.04 0.50
V 1.003 (0.999Y1.007) V 0.96 (0.90Y1.02) 0.95 (0.92Y0.99) 1 1.0 2 0.20 (0.06Y0.71) 3 0.40 (0.12Y1.34)
V 0.09 V 0.18 0.02 V 0.01 0.13
For continuous variables, the odds ratios shown are based on a one-unit change in the respective parameter. Adjusted odds ratios were calculated in the presence of the other parameters listed except for SHBG, which was only used in combination with TT. TT and cFT were used alternatively in the models. TT, total testosterone; cFT, calculated free testosterone; SHBG, sex hormone-binding globulin; MELD, model for end-stage liver disease score.
variables of MELD score, total testosterone, SHBG, age at transplant, and disease etiology. A higher total testosterone level and lower age at transplant remained significantly associated with early allograft rejection in the presence of the other covariates. Odds ratios are given in Table 2. Adding cyclosporine use to the model did not diminish the predictive value of testosterone, which remained independent of this influence. A decrease of 1 nmol/L in total testosterone of 10 pmol/L in free testosterone decreased the risk of rejection by 7% and 4%, respectively (Table 2). When free testosterone was incorporated into the model instead of total testosterone and SHBG, it did not reach significance as a predictor (P=0.09, Table 2). The probability of future transplant rejection at a certain total testosterone level in the complete model, including the covariates SHBG, MELD score, age and disease etiology, is shown in Figure 1. Length of Stay in Hospital There was a trend to a longer length of stay in hospital in patients who had acute cellular rejection early posttransplant (median, 18 days versus 15 days); however, this did not reach statistical significance (P = 0.059). There was a wide variation in inpatient length of stay, IQR (12Y23).
FIGURE 1.
The total (correlation, j0.10; P=0.041) and free testosterone (j0.09, P=0.065) levels were either significantly or close to significance associated with the total inpatient length of stay. Total testosterone also correlated inversely with the length of stay in the ICU (correlation, j0.15; P=0.004). There was no significant association between testosterone levels and general ward days. Death Posttransplant Twenty-two patients died after their transplant. The presence of acute cellular rejection was similar in patients who survived as compared with those who died (15.5% vs 18.2%, P=0.98). Death was not significantly associated with either pretransplant total (P=0.66) or free (P=0.72) serum testosterone levels.
DISCUSSION The important novel finding from this study is the identification of an association between lower testosterone levels preliver transplant and a reduced risk of early allograft rejection. This association was independent of traditionally recognized markers of liver disease severity and traditional risk factors for rejection. Importantly, our study population was representative of other liver transplant cohorts, with similar rates of acute cellular rejection (2, 3) and of testosterone deficiency (7, 8). Our findings also reinforce the previously recognized increased risk of rejection in younger patients and those transplanted for autoimmune disease (10, 11). Given the observational nature of our study, we cannot infer causality or provide mechanistic insights into the association of low testosterone with a reduced risk of acute rejection. Low testosterone may be associated with reduced rejection simply because it is a marker of ill health which in turn impairs immune responses. Indeed, a previous study showed a trend toward reduced rejection risk with higher MELD score (5). Our observation that low testosterone is associated with a reduced risk of acute rejection, independent of MELD, suggests that low testosterone may be a more robust marker of poor health-associated immune suppression than MELD. There are few objective, quantifiable markers of ‘‘ill health,’’ and hence, identifying a biochemical marker that may reflect immune dysfunction should be useful. The utility of testosterone as a biomarker of ill health
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
* 2014 Lippincott Williams & Wilkins
is further supported by previous research demonstrating that in men awaiting a liver transplant, low testosterone predicts mortality independent of MELD (6). There are a number of possible mechanisms through which testosterone levels could be linked to immune function. Chronic disease induces a proinflammatory state in which HPT axis suppression may result from elevated circulating levels of cytokines such as interleukin-1, interleukin-6, and tumor necrosis factor-alpha (12). Testosterone also has been shown to have direct effects on immune function including reducing circulating antibody levels. Indeed, it has been used with some success in the treatment of autoimmune conditions (13). Few studies have examined the effect of testosterone levels on transplant rejection. Female rats have been shown to be more likely to experience rejection of allogeneic skin transplants than male rats, and this risk was subsequently reduced by testosterone treatment. Interestingly, orchidectomy with subsequent reduction in testosterone did not modify rejection rates in males, but treatment with oestradiol worsened rejection (14). A subsequent study in rats conversely demonstrated that higher testosterone levels worsened renal transplant rejection rates in male subjects, whereas estradiol was protective (15). The only report in humans linked testosterone therapy posttransplant with the development of acute rejection in three cardiac allograft recipients (16). Our results differ from a recent study in patients with chronic allograft rejection postcardiac transplantation that showed that low testosterone levels were associated with an increased rate of chronic allograft rejection (17). This study design differed from ours in that the testosterone levels were taken several years posttransplant in stable patients. Our cohort had hormone profiles measured pretransplant when patients were severely ill. The contrasting results between these two studies may reflect a different relationship between testosterone and allograft rejection at different points in time. Pretransplant, testosterone may reflect systemic illness and immunosuppression. Posttransplant, in stable patients, rising testosterone levels may act to moderate the immune system and hence reduce rejection risk, which is consistent with previously mentioned data demonstrating that testosterone can reduce antibody levels and treat autoimmune disease (13). It is well established that in the majority of cirrhotics, the HPT axis recovers at least partially after liver transplant (7, 18). This likely corresponds with patient health and consequently improving immune function. We focused on early rejection within 3 months as we hypothesized that early rejection will reflect pretransplant hormone status more closely than late rejection. Given the results observed in the cardiac transplant cohort several years posttransplant, future longitudinal studies should measure sequential testosterone levels to examine the relation between testosterone and rejection at different points in time. It is important to note that in our study, the majority of patients had low testosterone levels, with 57% having total testosterone values less than 10 nmol/L and 87% having low free testosterone less than 230 pmnol/L. Our data therefore do not provide insights into the possible impact of elevated testosterone or testosterone replacement therapy on rejection rates. Interestingly, in the multivariate analysis,
Sinclair et al.
791
only total but not free testosterone was significantly associated with rejection. SHBG binds testosterone with high affinity and has been shown to be elevated in men with cirrhosis. When hepatic decompensation occurs, SHBG levels decline (7, 8). In keeping with this observation, in our cohort, there was an inverse correlation between MELD score and SHBG, which could potentially affect free testosterone levels. In conditions which significantly affect SHBG levels, free testosterone may be a more accurate marker of androgen status than total testosterone, but this is controversial (19). Moreover, the validity of free testosterone calculation based on mass action laws (20) in the setting of marked SHBG abnormalities such as occurs in chronic liver disease is insufficiently explored. Our findings may have implications for testosterone treatment of men with chronic liver disease who have low testosterone levels. Such men may have profound hypogonadism. However, the generalized symptoms and changes to nutritional status and muscle bulk in men with severe chronic liver disease may be difficult to distinguish from those associated with hypogonadism. There are no adequately powered prospective randomized controlled trials that inform about the risks and benefits of testosterone therapy in such men. Given our finding that low testosterone is associated with reduced rejection rates, future trials of testosterone treatment in men with advanced liver disease should include an analysis of the incidence of rejection in those who undergo transplantation. One potential drawback of our study is that our findings are based on a single testosterone measurement in some cases taken several months pretransplant. However, in a previous study of men with type 2 diabetes and multiple comorbidities, 70% of men had similarly low testosterone levels when retested a median of 6 months after the initial testosterone level (21). In conclusion, lower testosterone levels preliver transplantation are independently associated with lower rates of early acute liver allograft rejection. This may reflect more severe systemic illness and/ or impaired immunity. Future prospective studies should assess whether pretransplant serum testosterone has utility in predicting clinically significant rejection and hence may ultimately prove useful to help guide greater individualization of postliver transplant immunosuppressive regimens.
MATERIALS AND METHODS We performed a database search at our institution to retrieve data on 190 consecutive men who received a liver transplant between January 2007 and June 2013. At our institution, we routinely measure testosterone levels as part of the pretransplant workup, given the high prevalence of low testosterone levels in this population (6). The immunosuppressive regimen at the Victorian liver transplant unit includes postoperative intravenous methylprednisolone for 4 days and oral prednisolone 20 mg subsequently. A calcineurin inhibitor (either tacrolimus or cyclosporine aiming for levels of 8Y12 Kg/L or 800Y1200 Kg/L, respectively) and either azathioprine or mycophenalate mofetil complete the three-drug regimen. The specific agents used in individual patients were chosen by the treating clinician with consideration of patient factors such as the presence of diabetes, renal impairment, and hepatitis C virus infection. Data collected included patient age at time of transplant and history of prior liver transplant. Disease etiology was grouped into three categories. Autoimmune disease etiologies were grouped together because of their
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
792
www.transplantjournal.com
acknowledged increased risk of rejection. These included autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis. Patients with hepatitis C were a second group because of the tendency to reduce immunosuppression early to minimize disease recurrence. All other patients were grouped together in a third category. Pretransplant biochemistry was recorded as measured using standard methodologies at the Biochemistry Department, Austin Health. This included creatinine, international normalized ratio (INR), and bilirubin to allow calculation of the MELD score, 25-hydroxyvitamin D3 (vitamin D), and albumin.
Quantification of Testosterone Levels Serum total testosterone was measured using the Access testosterone assay (Beckman Coulter, Inc., Fullerton, CA) with a minimum detection limit of 0.35 nmol/L. The intra-assay and interassay coefficients of variation (CVs) assessed for two different concentrations (4.7 and 26.0 nmol/L) were 3.9%, 4.8% and 5.7%, 5.0%, respectively. The reference range for this total testosterone assay was 10 to 27.6 nmol/L, based on gas chromatography/ mass spectrometry (GC/MS) measurements obtained from a reference panel of 124 healthy, reproductively normal young men (22). Low total testosterone levels were therefore designated in individuals with a serum level below 10 nmol/L. Sex hormone binding globulin (SHBG) levels were determined with the Immulite 2000 analyser (Diagnostics Products Corporation, Los Angeles, CA, USA). The reference range for SHBG was 13 to 71 nmol/L, and the minimum detection limit was 0.02 nmol/L. Free testosterone was calculated from total testosterone, SHBG and serum albumin based on mass action laws with Vermeulen’s formula (20). The reference range for the access-testosterone/Immulite SHBG combination for cFT was 230 to 610 pmol/L (20). Low free testosterone levels were defined as less than 230 pmol/L.
Posttransplant Outcome The complication of acute allograft rejection within 3 months of transplant was retrieved from the liver transplant database. Routine biopsies are not performed at our center but only in the context of persistent biochemical disturbance in the absence of alternative clear cause. Diagnosis required appropriate liver histology findings (rejection activity index of 5 and above) in conjunction with a clinical decision that rejection was significant and required treatment. All patients classified as having had rejection received pulse intravenous methylprednisolone. Severity was graded by RAI in all patients. Graft loss due to rejection, vascular event, or other cause was recorded, as was death due to any cause. Length of stay posttransplant was divided into intensive care days and ward days to reflect differences in disease acuity. Length of follow-up posttransplant was recorded in years from transplant.
Statistical Analysis Descriptive statistics included median and IQR (25th and 75th percentiles). Differences between two groups were analyzed using means of nonparametric Wilcoxon rank sum test or among more than two groups by Kruskal-Wallis rank sum and chi-squared test in case of frequency distributions. P values reported in Table 1 were corrected for multiple testing using the Benjamini-Hochberg method. Correlations were done using Kendall Tau rank correlation. A generalized linear model with a binomial (logit) link function and a likelihood test was used for the analysis of individual predictor(s) as well as their adjustment for multiple covariates on the binary outcome of transplant rejection. Data obtained from the model are presented as crude and adjusted odds ratios and a probability plot for the main parameter of interest (total testosterone) in the fully adjusted model. For all statistics given, PG0.05 conferred significance. The statistical software packages R for Mac (version 3.02) and Deducer 0.7Y6 with JGR 1.7Y16 were used for the analyses (23, 24).
Transplantation
& Volume 98, Number 7, October 15, 2014 REFERENCES
1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14.
15. 16. 17. 18. 19.
20. 21. 22. 23. 24.
Belle SH, Beringer KC, Detre KM. An update on liver transplantation in the United States: recipient characteristics and outcome. Clin Transplant 1995: 19. Maluf DG, Stravitz RT, Cotterell AH, et al. Adult living donor versus deceased donor liver transplantation: a 6-year single center experience. Am J Transplant 2005; 5: 149. Gruttadauria S, Vasta F, Mandala L, et al. Basiliximab in a triple-drug regimen with tacrolimus and steroids in liver transplantation. Transplant Proc 2005; 37: 2611. Christou L, Pappas G, Falagas ME. Bacterial infection-related morbidity and mortality in cirrhosis. Am J Gastroenterol 2007; 102: 1510. Onaca NN, Levy MF, Netto GJ, et al. Pretransplant MELD score as a predictor of outcome after liver transplantation for chronic hepatitis C. Am J Transplant 2003; 3: 626. Grossmann M, Hoermann R, Gani L, et al. Low testosterone levels as an independent predictor of mortality in men with chronic liver disease. Clin Endocrinol 2012; 77: 323. Handelsman DJ, Strasser S, McDonald JA, et al. Hypothalamic-pituitarytesticular function in end-stage non-alcoholic liver disease before and after liver transplantation. Clin Endocrinol 1995; 43: 331. Zietz B, Lock G, Plach B, et al. Dysfunction of the hypothalamicpituitary-glandular axes and relation to Child-Pugh classification in male patients with alcoholic and virus-related cirrhosis. Eur J Gastroenterol Hepatol 2003; 15: 495. Bannister P, Handley T, Chapman C, et al. Hypogonadism in chronic liver disease: impaired release of luteinising hormone. Br Med J (Clin Res Ed) 1986; 293: 1191. Wang YC, Wu TJ, Wu TH, et al. The risk factors to predict acute rejection in liver transplantation. Transplant Proc 2012; 44: 526. Hayashi M, Keeffe EB, Krams SM, et al. Allograft rejection after liver transplantation for autoimmune liver diseases. Liver Transplant Surg 1998; 4: 208. Jones TH, Kennedy RL. Cytokines and hypothalamic-pituitary function. Cytokine 1993; 5: 531. Pennell LM, Galligan CL, Fish EN. Sex affects immunity. J Autoimmunity 2012; 38: J282. Hirasawa K, Enosawa S. Effects of sex steroid hormones on sex-associated differences in the survival time of allogeneic skin grafts in rats. Evidence that testosterone enhances and estradiol reverses the immunosuppressive activity of cyclosporine. Transplantation 1990; 50: 637. Muller V, Szabo A, Viklicky O, et al. Sex hormones and gender-related differences: their influence on chronic renal allograft rejection. Kidney Int 1999; 55: 2011. Schofield RS, Hill JA, McGinn CJ, et al. Hormone therapy in men and risk of cardiac allograft rejection. J Heart Lung Transplant 2002; 21: 493. Caretta N, Feltrin G, Tarantini G, et al. Low serum testosterone as a new risk factor for chronic rejection in heart transplanted men. Transplantation 2013; 96: 501. Madersbacher S, Ludvik G, Stulnig T, et al. The impact of liver transplantation on endocrine status in men. Clin Endocrinol 1996; 44: 461. van den Beld AW, de Jong FH, Grobbee DE, et al. Measures of bioavailable serum testosterone and estradiol and their relationships with muscle strength, bone density, and body composition in elderly men. J Clin Endocrinol Metab 2000; 85: 3276. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84: 3666. Grossmann M, Thomas MC, Panagiotopoulos S, et al. Low testosterone levels are common and associated with insulin resistance in men with diabetes. J Clin Endocrinol Metab 2008; 93: 1834. Sikaris K, McLachlan RI, Kazlauskas R, et al. Reproductive hormone reference intervals for healthy fertile young men: evaluation of automated platform assays. J Clin Endocrinol Metab 2005; 90: 5928. Core team R. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2013. Fellows I. A data analysis GUI for R. J Stat Software 2012; 49: 1.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
AND
TRANSLATIONAL RESEARCH
Low-Serum Testosterone Levels Pre-Liver Transplantation Are Associated With Reduced Rates of Early Acute Allograft Rejection in Men Marie Sinclair,1,2,4 Peter W. Angus,1,2 Paul J. Gow,1,2 Rudolf Hoermann,2 Tamara Mogilevski,1 and Mathis Grossmann2,3 Background. Low pretransplant serum testosterone has recently been associated with increased mortality in men awaiting liver transplantation, but the potential impact on rejection has not yet been investigated. Methods. Pretransplantation serum testosterone, SHBG, and other variables were collected on 190 consecutive men who received a liver transplant between 2007 and 2013. Rates of subsequent acute cellular rejection were recorded. Multivariable analysis was performed to define variables associated with rejection and other clinically important end points. Results. Thirty (16%) of 190 men experienced acute cellular rejection. Lower pretransplant testosterone was associated with lower rejection rates, j7% (95% confidence interval [CI], j2% to j12%) per nmol/L decrease in total testosterone and j4% (95% CI, j0.5% to j7%) per 10 pmol/L decrease in free testosterone. Total testosterone (correlation 0.29, P=0.04) and free testosterone (correlation 0.37, P=0.01) correlated significantly with the histological severity of rejection. Older age at transplant (+5% [95% CI, 9%Y2%]) per year, and nonautoimmune etiology of liver disease (OR, 1.0 for autoimmune, 0.22 [95% CI, 0.07Y0.73] for hepatitis C virus, and 0.58 [95% CI, 0.21Y1.71] for other etiologies) were also associated with decreased rejection risk. In a generalized linear model including the covariates testosterone, SHBG, age, etiology, and MELD, total testosterone remained a significant predictor of rejection (adjusted OR, 1.06; P=0.03), as did age at transplant (OR, 0.95; P=0.01). Conclusion. Low preliver transplant serum testosterone independently predicts a decreased risk of acute allograft rejection. Whether testosterone is a marker of disease-associated immune dysfunction or has immune-modulatory effects requires further study. Keywords: Liver transplantation, Rejection, Testosterone. (Transplantation 2014;98: 788Y792)
urvival rates after orthotopic liver transplantation currently exceed 85% at 12 months (1). Acute cellular rejection, however, remains a frequent problem, with approximately 25% of patients experiencing an episode within 3 months of transplant (2, 3). Prevention and treatment of rejection requires high doses of immunosuppressive medications that are associated with significant side effects. Thus, there is major clinical interest in identifying pretransplant
S
The authors declare no funding or conflicts of interest. 1 Victorian Liver Transplant Unit, Austin Health, Heidelberg, Victoria, Australia. 2 Department of Medicine, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia. 3 Endocrine Unit, Austin Health, Heidelberg, Victoria, Australia. 4 Address correspondence to: Marie Sinclair, Liver Transplant Unit, Austin Health, 145 Studley Road, Heidelberg, Victoria, Australia 3084. E-mail: [email protected] T.M. and M.S. contributed in data collection. M.S., M.G., P.G., and P.A. contributed in writing, reviewing, and editing of the manuscript. R.H. contributed in the statistical analysis and refining of results. Received 5 December 2013. Revision requested 30 December 2013. Accepted 19 February 2014. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0041-1337/14/9807-788 DOI: 10.1097/TP.0000000000000130
788
www.transplantjournal.com
factors that predict rejection, to individualize the use of immunosuppressive agents according to risk. Immunity is impaired in cirrhosis, and this has been attributed to defects in polymorphonucleocyte function, macrophage activity, and antibody production as well as complement deficiency (4). A functioning immune system is necessary for the development of acute cellular rejection. Therefore, sicker patients who have more severely impaired immunity may be less likely to experience rejection posttransplant (5). Low testosterone levels are found in up to 90% in men awaiting transplantation and are related to the severity of the underlying liver disease (6 Y8). The pathogenesis is multifactorial and involves suppression of the hypothalamicpituitary-testicular (HPT) axis at multiple levels (7, 9). In men awaiting liver transplantation, testosterone may be a more sensitive marker of ill health than MELD. Indeed, in a recent study, we found that low serum testosterone predicts mortality independently of MELD score and of hyponatremia (6). This suggests that serum testosterone may more closely reflect systemic illness and hence immune impairment than traditional markers of liver disease severity. In the current study, we tested the hypothesis that low pretransplant testosterone levels are associated with a Transplantation
& Volume 98, Number 7, October 15, 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Sinclair et al.
* 2014 Lippincott Williams & Wilkins
decreased risk of acute allograft rejection postliver transplant. We also hypothesized that this is independent of conventional markers of liver disease severity.
RESULTS Baseline Characteristics Baseline characteristics of the 190 consecutive men included in the study are shown in Table 1. Pretransplant assessment, which included routine testosterone studies, was performed at a median of 6 months before transplant (IQR, 3Y11). Median age at time of transplant was 53 years (IQR, 46Y58), and MELD score 19 (13Y24). Twenty-four patients (12.6%) were transplanted for autoimmune spectrum liver diseases, 83 for hepatitis C (43.7%), and 83 (43.7%) for other etiologies. Fifteen men (7.9%) had previously received a liver transplant. Seventy-five (39.5%) patients were prescribed cyclosporine as the calcineurin inhibitor component of the immunosuppressive regimen. The median total testosterone level was at 8.8 nmol/L (3.6Y14.8), and 108 patients (57%) had a low total testosterone (less than 10 nmol/L). The median free testosterone was 106 pmol/L (50Y176) and below the normal range (G230 pmol/L) in 165 patients (87%). Both total and free testosterone levels negatively correlated with MELD score (correlation, j0.28 and j0.24, respectively; PG0.0001). The median SHBG was elevated at 73 nmol/L (53.4Y97.5) (normal range, 13Y71nmol/L). There was a significant negative correlation between SHBG and MELD score (correlation, j0.18; P=0.0004). The median length of stay in hospital after liver transplantation was 16 days (12Y23). Median days spent in ICU were 3 (2Y5), with a median of 13 inpatient days in the TABLE 1.
789
general ward subsequently (9Y19). The median follow-up time for patients was 2.9 years (1.2Y4.7), with a range of 3 months to 10 years. Graft loss was seen in 11 patients (5.8%) over this time, and 22 patients died (11.6%). Acute Cellular Rejection Acute cellular rejection was diagnosed and treated in 30 patients (16%) within 3 months of liver transplantation. Men who experienced early rejection were significantly younger than those who did not (48.5 years compared with 53 years, respectively, P=0.008). Approximately 29.2% of men with liver disease due to autoimmune disease experienced rejection, with lower incidence in those with hepatitis C (8.4%) or with other etiologies (19.3%), P=0.025. The median total testosterone level in men experiencing early acute cellular rejection was 12.5 nmol/L (9.0Y19.2), which was significantly higher than that in men without rejection, 7.8 nmol/L (3.3Y13.8), P=0.003. Similarly, median free testosterone was 155 pmol/L (105Y243) in patients experiencing early rejection, significantly higher than that in men without rejection, 95pmol/L (48Y167), P=0.006. Total testosterone (correlation 0.29, P=0.04) and free testosterone levels (correlation 0.37, P=0.01) correlated significantly with severity of rejection as measured by the rejection activity index (RAI). By contrast, there was no significant association between rejection and pretransplant MELD score (P=0.33) or history of previous transplant (P=1.0). The rejection rate was lower in cyclosporine versus noncyclosporine users (6.7% vs. 21.7%, P=0.005); cyclosporine use did not retain significance after adjustment. Multivariable analysis for the outcome of transplant rejection was performed incorporating the pretransplant
Baseline demographics
Etiology n (%): Autoimmune spectrum HCV Other Median age at transplant (IQR) Median MELD (IQR) Median Na (IQR) Median total testosterone (IQR) Median SHBG (IQR) Median free testosterone (IQR) Median albumin (IQR) Median LH (IQR) Median FSH (IQR) Median creatinine (IQR) Median vitamin D (IQR) Median days in ICU posttransplant (IQR) Median days on ward posttransplant Median inpatients days posttransplant and IQR Died during follow-up: n (%) Graft loss during follow-up n (%): Years of follow-up posttransplant (IQR)
Total
No rejection
Rejection
P value
24 (12.6%) 83 (43.7%) 83 (43.7%) 53.0 (46.2Y58.0) 19.0 (13.0Y24.8) 136 (132Y139) 8.8 (3.6Y14.8) 73.0 (53Y98) 106 (50Y176) 28.0 (24.0Y34.0) 5.0 (2.8Y7.4) 6.2 (3.3Y9.7) 89.0 (71.5Y119.0) 56.0 (39.2Y82.0) 3.0 (2.0Y5.0) 13.0 (9.0Y19.0) 16.0 (12.0Y23.0) 22 (11.6%) 11 (5.8%) 2.9 (1.2Y4.7)
17 (10.6%) 76 (47.5%) 67 (41.9%) 53.0 (47.0Y58.2) 18.5 (14.0Y25.0) 135 (132Y139) 7.8 (3.3Y13.8) 70 (50Y96) 95 (48Y167) 28.0 (23.5Y33.0) 5.0 (2.8Y7.3) 5.8 (3.3Y9.7) 95.0 (72.2Y121.5) 55.0 (39.0Y79.0) 3.0 (2.0Y5.0) 12.0 (8.5Y19.0) 15.0 (11.0Y23.0) 18 (11.2%) 9 (5.6%) 2.9 (1.2Y4.8)
7 (23.3%) 7 (23.3%) 16 (53.3%) 48.5 (38.0Y54.8) 19.0 (10.5Y22.0) 137 (136Y140) 12.5 (9.0Y19.2) 86 (66Y116) 155 (105Y243) 29.0 (26.0Y38.0) 5.7 (3.8Y8.9) 7.3 (3.6Y11.4) 78.0 (69.0Y98.0) 59.0 (46.0Y104.0) 3.0 (2.0Y5.0) 15.0 (13.0Y17.0) 18.0 (15.2Y26.0) 4 (13.3%) 2 (6.7%) 2.8 (1.1Y3.9)
0.079
P values given were corrected for multiple testing by the Benjamini-Hochberg method.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
0.048 0.42 0.063 0.048 0.085 0.048 0.21 0.21 0.38 0.13 0.38 0.59 0.081 0.13 0.76 0.73 0.59
790
www.transplantjournal.com
TABLE 2.
Transplantation
& Volume 98, Number 7, October 15, 2014
Increased risk of acute cellular rejection according to testosterone levels and covariates
Parameter TT (nmol/L) cFT (pmol/L) SHBG (nmol/L) MELD Age (yr) Disease etiology
Odds ratio (95% CI) 1.07 (1.02Y1.12) 1.004 (1.0005Y1.007) 1.01 (1.002Y1.022) 0.97 (0.92Y1.02) 0.95 (0.91Y0.98) Autoimmune 1.0 HCV 0.22 (0.07Y0.73) Other 0.58 (0.21Y1.71)
P value
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
P value
0.04 0.02 0.045 0.28 0.05 V 0.01 0.30
1.06 (1.004Y1.13) b1.007 (0.99Y1.02) 0.99 (0.92Y1.05) 0.95 (0.91Y0.99) 1 1.0 2 0.25 (0.07Y0.97) 3 0.65 (0.18Y2.44)
0.03 V 0.30 0.65 0.01 V 0.04 0.50
V 1.003 (0.999Y1.007) V 0.96 (0.90Y1.02) 0.95 (0.92Y0.99) 1 1.0 2 0.20 (0.06Y0.71) 3 0.40 (0.12Y1.34)
V 0.09 V 0.18 0.02 V 0.01 0.13
For continuous variables, the odds ratios shown are based on a one-unit change in the respective parameter. Adjusted odds ratios were calculated in the presence of the other parameters listed except for SHBG, which was only used in combination with TT. TT and cFT were used alternatively in the models. TT, total testosterone; cFT, calculated free testosterone; SHBG, sex hormone-binding globulin; MELD, model for end-stage liver disease score.
variables of MELD score, total testosterone, SHBG, age at transplant, and disease etiology. A higher total testosterone level and lower age at transplant remained significantly associated with early allograft rejection in the presence of the other covariates. Odds ratios are given in Table 2. Adding cyclosporine use to the model did not diminish the predictive value of testosterone, which remained independent of this influence. A decrease of 1 nmol/L in total testosterone of 10 pmol/L in free testosterone decreased the risk of rejection by 7% and 4%, respectively (Table 2). When free testosterone was incorporated into the model instead of total testosterone and SHBG, it did not reach significance as a predictor (P=0.09, Table 2). The probability of future transplant rejection at a certain total testosterone level in the complete model, including the covariates SHBG, MELD score, age and disease etiology, is shown in Figure 1. Length of Stay in Hospital There was a trend to a longer length of stay in hospital in patients who had acute cellular rejection early posttransplant (median, 18 days versus 15 days); however, this did not reach statistical significance (P = 0.059). There was a wide variation in inpatient length of stay, IQR (12Y23).
FIGURE 1.
The total (correlation, j0.10; P=0.041) and free testosterone (j0.09, P=0.065) levels were either significantly or close to significance associated with the total inpatient length of stay. Total testosterone also correlated inversely with the length of stay in the ICU (correlation, j0.15; P=0.004). There was no significant association between testosterone levels and general ward days. Death Posttransplant Twenty-two patients died after their transplant. The presence of acute cellular rejection was similar in patients who survived as compared with those who died (15.5% vs 18.2%, P=0.98). Death was not significantly associated with either pretransplant total (P=0.66) or free (P=0.72) serum testosterone levels.
DISCUSSION The important novel finding from this study is the identification of an association between lower testosterone levels preliver transplant and a reduced risk of early allograft rejection. This association was independent of traditionally recognized markers of liver disease severity and traditional risk factors for rejection. Importantly, our study population was representative of other liver transplant cohorts, with similar rates of acute cellular rejection (2, 3) and of testosterone deficiency (7, 8). Our findings also reinforce the previously recognized increased risk of rejection in younger patients and those transplanted for autoimmune disease (10, 11). Given the observational nature of our study, we cannot infer causality or provide mechanistic insights into the association of low testosterone with a reduced risk of acute rejection. Low testosterone may be associated with reduced rejection simply because it is a marker of ill health which in turn impairs immune responses. Indeed, a previous study showed a trend toward reduced rejection risk with higher MELD score (5). Our observation that low testosterone is associated with a reduced risk of acute rejection, independent of MELD, suggests that low testosterone may be a more robust marker of poor health-associated immune suppression than MELD. There are few objective, quantifiable markers of ‘‘ill health,’’ and hence, identifying a biochemical marker that may reflect immune dysfunction should be useful. The utility of testosterone as a biomarker of ill health
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
* 2014 Lippincott Williams & Wilkins
is further supported by previous research demonstrating that in men awaiting a liver transplant, low testosterone predicts mortality independent of MELD (6). There are a number of possible mechanisms through which testosterone levels could be linked to immune function. Chronic disease induces a proinflammatory state in which HPT axis suppression may result from elevated circulating levels of cytokines such as interleukin-1, interleukin-6, and tumor necrosis factor-alpha (12). Testosterone also has been shown to have direct effects on immune function including reducing circulating antibody levels. Indeed, it has been used with some success in the treatment of autoimmune conditions (13). Few studies have examined the effect of testosterone levels on transplant rejection. Female rats have been shown to be more likely to experience rejection of allogeneic skin transplants than male rats, and this risk was subsequently reduced by testosterone treatment. Interestingly, orchidectomy with subsequent reduction in testosterone did not modify rejection rates in males, but treatment with oestradiol worsened rejection (14). A subsequent study in rats conversely demonstrated that higher testosterone levels worsened renal transplant rejection rates in male subjects, whereas estradiol was protective (15). The only report in humans linked testosterone therapy posttransplant with the development of acute rejection in three cardiac allograft recipients (16). Our results differ from a recent study in patients with chronic allograft rejection postcardiac transplantation that showed that low testosterone levels were associated with an increased rate of chronic allograft rejection (17). This study design differed from ours in that the testosterone levels were taken several years posttransplant in stable patients. Our cohort had hormone profiles measured pretransplant when patients were severely ill. The contrasting results between these two studies may reflect a different relationship between testosterone and allograft rejection at different points in time. Pretransplant, testosterone may reflect systemic illness and immunosuppression. Posttransplant, in stable patients, rising testosterone levels may act to moderate the immune system and hence reduce rejection risk, which is consistent with previously mentioned data demonstrating that testosterone can reduce antibody levels and treat autoimmune disease (13). It is well established that in the majority of cirrhotics, the HPT axis recovers at least partially after liver transplant (7, 18). This likely corresponds with patient health and consequently improving immune function. We focused on early rejection within 3 months as we hypothesized that early rejection will reflect pretransplant hormone status more closely than late rejection. Given the results observed in the cardiac transplant cohort several years posttransplant, future longitudinal studies should measure sequential testosterone levels to examine the relation between testosterone and rejection at different points in time. It is important to note that in our study, the majority of patients had low testosterone levels, with 57% having total testosterone values less than 10 nmol/L and 87% having low free testosterone less than 230 pmnol/L. Our data therefore do not provide insights into the possible impact of elevated testosterone or testosterone replacement therapy on rejection rates. Interestingly, in the multivariate analysis,
Sinclair et al.
791
only total but not free testosterone was significantly associated with rejection. SHBG binds testosterone with high affinity and has been shown to be elevated in men with cirrhosis. When hepatic decompensation occurs, SHBG levels decline (7, 8). In keeping with this observation, in our cohort, there was an inverse correlation between MELD score and SHBG, which could potentially affect free testosterone levels. In conditions which significantly affect SHBG levels, free testosterone may be a more accurate marker of androgen status than total testosterone, but this is controversial (19). Moreover, the validity of free testosterone calculation based on mass action laws (20) in the setting of marked SHBG abnormalities such as occurs in chronic liver disease is insufficiently explored. Our findings may have implications for testosterone treatment of men with chronic liver disease who have low testosterone levels. Such men may have profound hypogonadism. However, the generalized symptoms and changes to nutritional status and muscle bulk in men with severe chronic liver disease may be difficult to distinguish from those associated with hypogonadism. There are no adequately powered prospective randomized controlled trials that inform about the risks and benefits of testosterone therapy in such men. Given our finding that low testosterone is associated with reduced rejection rates, future trials of testosterone treatment in men with advanced liver disease should include an analysis of the incidence of rejection in those who undergo transplantation. One potential drawback of our study is that our findings are based on a single testosterone measurement in some cases taken several months pretransplant. However, in a previous study of men with type 2 diabetes and multiple comorbidities, 70% of men had similarly low testosterone levels when retested a median of 6 months after the initial testosterone level (21). In conclusion, lower testosterone levels preliver transplantation are independently associated with lower rates of early acute liver allograft rejection. This may reflect more severe systemic illness and/ or impaired immunity. Future prospective studies should assess whether pretransplant serum testosterone has utility in predicting clinically significant rejection and hence may ultimately prove useful to help guide greater individualization of postliver transplant immunosuppressive regimens.
MATERIALS AND METHODS We performed a database search at our institution to retrieve data on 190 consecutive men who received a liver transplant between January 2007 and June 2013. At our institution, we routinely measure testosterone levels as part of the pretransplant workup, given the high prevalence of low testosterone levels in this population (6). The immunosuppressive regimen at the Victorian liver transplant unit includes postoperative intravenous methylprednisolone for 4 days and oral prednisolone 20 mg subsequently. A calcineurin inhibitor (either tacrolimus or cyclosporine aiming for levels of 8Y12 Kg/L or 800Y1200 Kg/L, respectively) and either azathioprine or mycophenalate mofetil complete the three-drug regimen. The specific agents used in individual patients were chosen by the treating clinician with consideration of patient factors such as the presence of diabetes, renal impairment, and hepatitis C virus infection. Data collected included patient age at time of transplant and history of prior liver transplant. Disease etiology was grouped into three categories. Autoimmune disease etiologies were grouped together because of their
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
792
www.transplantjournal.com
acknowledged increased risk of rejection. These included autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis. Patients with hepatitis C were a second group because of the tendency to reduce immunosuppression early to minimize disease recurrence. All other patients were grouped together in a third category. Pretransplant biochemistry was recorded as measured using standard methodologies at the Biochemistry Department, Austin Health. This included creatinine, international normalized ratio (INR), and bilirubin to allow calculation of the MELD score, 25-hydroxyvitamin D3 (vitamin D), and albumin.
Quantification of Testosterone Levels Serum total testosterone was measured using the Access testosterone assay (Beckman Coulter, Inc., Fullerton, CA) with a minimum detection limit of 0.35 nmol/L. The intra-assay and interassay coefficients of variation (CVs) assessed for two different concentrations (4.7 and 26.0 nmol/L) were 3.9%, 4.8% and 5.7%, 5.0%, respectively. The reference range for this total testosterone assay was 10 to 27.6 nmol/L, based on gas chromatography/ mass spectrometry (GC/MS) measurements obtained from a reference panel of 124 healthy, reproductively normal young men (22). Low total testosterone levels were therefore designated in individuals with a serum level below 10 nmol/L. Sex hormone binding globulin (SHBG) levels were determined with the Immulite 2000 analyser (Diagnostics Products Corporation, Los Angeles, CA, USA). The reference range for SHBG was 13 to 71 nmol/L, and the minimum detection limit was 0.02 nmol/L. Free testosterone was calculated from total testosterone, SHBG and serum albumin based on mass action laws with Vermeulen’s formula (20). The reference range for the access-testosterone/Immulite SHBG combination for cFT was 230 to 610 pmol/L (20). Low free testosterone levels were defined as less than 230 pmol/L.
Posttransplant Outcome The complication of acute allograft rejection within 3 months of transplant was retrieved from the liver transplant database. Routine biopsies are not performed at our center but only in the context of persistent biochemical disturbance in the absence of alternative clear cause. Diagnosis required appropriate liver histology findings (rejection activity index of 5 and above) in conjunction with a clinical decision that rejection was significant and required treatment. All patients classified as having had rejection received pulse intravenous methylprednisolone. Severity was graded by RAI in all patients. Graft loss due to rejection, vascular event, or other cause was recorded, as was death due to any cause. Length of stay posttransplant was divided into intensive care days and ward days to reflect differences in disease acuity. Length of follow-up posttransplant was recorded in years from transplant.
Statistical Analysis Descriptive statistics included median and IQR (25th and 75th percentiles). Differences between two groups were analyzed using means of nonparametric Wilcoxon rank sum test or among more than two groups by Kruskal-Wallis rank sum and chi-squared test in case of frequency distributions. P values reported in Table 1 were corrected for multiple testing using the Benjamini-Hochberg method. Correlations were done using Kendall Tau rank correlation. A generalized linear model with a binomial (logit) link function and a likelihood test was used for the analysis of individual predictor(s) as well as their adjustment for multiple covariates on the binary outcome of transplant rejection. Data obtained from the model are presented as crude and adjusted odds ratios and a probability plot for the main parameter of interest (total testosterone) in the fully adjusted model. For all statistics given, PG0.05 conferred significance. The statistical software packages R for Mac (version 3.02) and Deducer 0.7Y6 with JGR 1.7Y16 were used for the analyses (23, 24).
Transplantation
& Volume 98, Number 7, October 15, 2014 REFERENCES
1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14.
15. 16. 17. 18. 19.
20. 21. 22. 23. 24.
Belle SH, Beringer KC, Detre KM. An update on liver transplantation in the United States: recipient characteristics and outcome. Clin Transplant 1995: 19. Maluf DG, Stravitz RT, Cotterell AH, et al. Adult living donor versus deceased donor liver transplantation: a 6-year single center experience. Am J Transplant 2005; 5: 149. Gruttadauria S, Vasta F, Mandala L, et al. Basiliximab in a triple-drug regimen with tacrolimus and steroids in liver transplantation. Transplant Proc 2005; 37: 2611. Christou L, Pappas G, Falagas ME. Bacterial infection-related morbidity and mortality in cirrhosis. Am J Gastroenterol 2007; 102: 1510. Onaca NN, Levy MF, Netto GJ, et al. Pretransplant MELD score as a predictor of outcome after liver transplantation for chronic hepatitis C. Am J Transplant 2003; 3: 626. Grossmann M, Hoermann R, Gani L, et al. Low testosterone levels as an independent predictor of mortality in men with chronic liver disease. Clin Endocrinol 2012; 77: 323. Handelsman DJ, Strasser S, McDonald JA, et al. Hypothalamic-pituitarytesticular function in end-stage non-alcoholic liver disease before and after liver transplantation. Clin Endocrinol 1995; 43: 331. Zietz B, Lock G, Plach B, et al. Dysfunction of the hypothalamicpituitary-glandular axes and relation to Child-Pugh classification in male patients with alcoholic and virus-related cirrhosis. Eur J Gastroenterol Hepatol 2003; 15: 495. Bannister P, Handley T, Chapman C, et al. Hypogonadism in chronic liver disease: impaired release of luteinising hormone. Br Med J (Clin Res Ed) 1986; 293: 1191. Wang YC, Wu TJ, Wu TH, et al. The risk factors to predict acute rejection in liver transplantation. Transplant Proc 2012; 44: 526. Hayashi M, Keeffe EB, Krams SM, et al. Allograft rejection after liver transplantation for autoimmune liver diseases. Liver Transplant Surg 1998; 4: 208. Jones TH, Kennedy RL. Cytokines and hypothalamic-pituitary function. Cytokine 1993; 5: 531. Pennell LM, Galligan CL, Fish EN. Sex affects immunity. J Autoimmunity 2012; 38: J282. Hirasawa K, Enosawa S. Effects of sex steroid hormones on sex-associated differences in the survival time of allogeneic skin grafts in rats. Evidence that testosterone enhances and estradiol reverses the immunosuppressive activity of cyclosporine. Transplantation 1990; 50: 637. Muller V, Szabo A, Viklicky O, et al. Sex hormones and gender-related differences: their influence on chronic renal allograft rejection. Kidney Int 1999; 55: 2011. Schofield RS, Hill JA, McGinn CJ, et al. Hormone therapy in men and risk of cardiac allograft rejection. J Heart Lung Transplant 2002; 21: 493. Caretta N, Feltrin G, Tarantini G, et al. Low serum testosterone as a new risk factor for chronic rejection in heart transplanted men. Transplantation 2013; 96: 501. Madersbacher S, Ludvik G, Stulnig T, et al. The impact of liver transplantation on endocrine status in men. Clin Endocrinol 1996; 44: 461. van den Beld AW, de Jong FH, Grobbee DE, et al. Measures of bioavailable serum testosterone and estradiol and their relationships with muscle strength, bone density, and body composition in elderly men. J Clin Endocrinol Metab 2000; 85: 3276. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84: 3666. Grossmann M, Thomas MC, Panagiotopoulos S, et al. Low testosterone levels are common and associated with insulin resistance in men with diabetes. J Clin Endocrinol Metab 2008; 93: 1834. Sikaris K, McLachlan RI, Kazlauskas R, et al. Reproductive hormone reference intervals for healthy fertile young men: evaluation of automated platform assays. J Clin Endocrinol Metab 2005; 90: 5928. Core team R. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2013. Fellows I. A data analysis GUI for R. J Stat Software 2012; 49: 1.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
