Abstract
Introduction: Low testosterone has been demonstrated to be an independent determinant of endothelial (dys)function in men. Graft-versus-host disease (GVHD) is a major contributor to non-relapse mortality (NRM) after allogeneic stem cell transplantation (alloSCT). Vulnerability of the recipients' endothelial cell system is a novel concept to explain why a proportion of patients with acute GVHD fail to respond to escalating immunosuppressive therapy and ultimately succumb to GVHD and related complications. This retrospective study investigated the prognostic impact of pre-transplant testosterone levels on NRM after alloSCT in male patients.
Patients and methods: Between 2002 and 2014, a total of 277 male patients undergoing alloSCT at Heidelberg University (median age 55 years) provided informed consent to participate in this observational study (training cohort). 71 patients (26%) received transplants from related donors (RD). Diagnoses were acute myeloid leukemia (AML) in 108 patients (48%), myelodysplastic syndrome (MDS) in 66 patients (29%), lymphoid malignancies (lymphoma, chronic and acute lymphoid leukemia) in 75 patients (33%), and multiple myeloma in 28 patients (12%). A total of 176 patients (78%) received statin treatment post alloSCT as per institutional standard policy. For validation, an independent patient cohort of 205 men allografted for AML and MDS (median age 57 years, 18% RD, no statin treatment) at Essen University was analysed.
Pre-transplant serum samples for testosterone measurements were collected between 0 and 2 months before alloSCT and cryopreserved at -80°C. Testosterone concentrations were measured by radioimmunoassay. Pre-transplant levels of suppressor of tumorigenicity-2 (ST2) were determined by ELISA.
Overall survival (OS), incidence of relapse and NRM and were calculated from date of alloSCT to the appropriate endpoint using Cox regression analysis with cause specific hazard models for NRM and relapse. As confounding prognostic factors we included testosterone levels, age, donor type, graft type, donor sex, conditioning intensity, and disease type and stage prior to alloSCT.
Results: Median pre-transplant testosterone level in the training and validation cohort was 13.6 nmol/L (range 0.3-41.7 nmol/L) and 16.0 nmol/L (0.8-38.1 nmol/L), respectively. In the training cohort, lower pre-transplant testosterone as continuous variable was associated with shorter OS (p=0.009). Lower testosterone levels showed a trend towards higher NRM (p=0.09) and a significant association with NRM after onset of acute GVHD (p=0.02). Multivariate analysis confirmed that lower pre-transplant testosterone levels were a significant predictor of an increased NRM risk after GVHD onset (p=0.03). In the subgroup of patients not receiving statins post-transplant, lower testosterone levels were associated with increased incidence of transplant-associated microangiopathy (p=0.01). In addition, lower pre-transplant testosterone levels correlated with higher pre-transplant ST2 levels indicating endothelial vulnerability.
In the validation cohort, similar results with regard to OS (p=0.02), NRM (p=0.04), NRM after acute GVHD onset (p=0.03) in univariate analysis, and to NRM after GVHD onset (p=0.02) in multivariable analysis could be observed. The association of pre-transplant testosterone levels (in quartiles) and incidence of NRM after GVHD onset in the training and validation cohort is depicted in Figure 1A and 1B, respectively.
Conclusion: Our study suggests that low pre-transplant testosterone is associated with serological and clinical evidence for endothelial damage and is an independent risk factor for a fatal outcome of GVHD. Prospective studies in the alloSCT setting investigating testosterone and testosterone supplementation in deficient patients are highly warranted.
No relevant conflicts of interest to declare.
