Shih-Ping Liu, Department of Urology, National Taiwan University Hospital, Taipei 10002, Taiwan. E-mail: email@example.com; Han-Sun Chiang, College of Medicine, Fu Jen Catholic University, Taipei 24205, Taiwan. E-mail: firstname.lastname@example.org
Circulating endothelial progenitor cells (EPCs) are bone marrow-derived cells required for endothelial repair. A low EPC number can be considered as an independent predictor of endothelial dysfunction and future cardiovascular events. Recent evidence shows that patients with hypogonadal symptoms without other confounding risk factors have a low number of circulating progenitor cells (PCs) and EPCs, thus highlighting the role of testosterone in the proliferation and differentiation of EPCs. Here, we investigate if testosterone replacement therapy (TRT) can increase circulating EPC number in men with late onset hypogonadism. Forty-six men (age range, 40–73 years; mean age, 58.3 years) with hypogonadal symptoms were recruited, and 29 men with serum total testosterone (TT) levels less than 350 ng/dL received TRT using transdermal testosterone gel (Androgel; 1% testosterone at 5 g/day) for 12 months. Circulating EPC numbers (per 100 000 monocytes) were calculated using flow cytometry. There was no significant association between serum TT levels and the number of circulating EPCs before TRT. Compared with the number of mean circulating EPCs at baseline (9.5 ± 6.2), the number was significantly higher after 3 months (16.6 ± 11.1, p = 0.027), 6 months (20.3 ± 15.3, p = 0.006) and 12 months (27.2 ± 15.5, p = 0.017) of TRT. Thus, we conclude that serum TT levels before TRT are not significantly associated with the number of circulating EPCs in men with late onset hypogonadism. However, TRT can increase the number of circulating EPCs, which implies the benefit of TRT on endothelial function in hypogonadal men.
Late onset hypogonadism (LOH) is a clinical and biochemical syndrome associated with advancing age and characterized by a deficiency in serum testosterone levels (below the young healthy adult male reference range) (Nieschlag et al., 2005). The clinical symptoms of LOH include decreased muscle mass and strength, decreased bone mass and osteoporosis, increased central body fat, decreased libido and sexual desire, forgetfulness, loss of memory, difficulty in concentration, insomnia and a decreased sense of well-being (Wald et al., 2006). In recent years, testosterone deficiency has been linked to an increase in cardiovascular risk factors in men and has been reported to play a role in the onset and/or progression of cardiovascular disease (Traish et al., 2009). Low serum levels of testosterone in ageing men are associated with increased all-cause mortality from coronary heart disease and other vascular disorders (Jones & Saad, 2009). It has also been demonstrated that a low serum testosterone level might contribute to increased arterial stiffness, which in turn has been associated with cardiovascular risk (Hougaku et al., 2006).
An intact endothelium is critical in maintaining vascular functions of arterial relaxation and venous constriction (Foresta et al., 2005). Circulating endothelial progenitor cells (EPCs) are bone marrow-derived cells positive for CD34, CD133, vascular endothelial growth factor receptor type 2 (Foresta et al., 2006) and CD144 (E-cadherin) (La Vignera, 2011). EPCs are known to increase in response to endothelial dysfunction (La Vignera, 2011), and thus may serve as indicators of vascular health. Following endothelial insult, EPCs migrate into peripheral circulation, home to sites of neovascularization or endothelial damage and differentiate into mature endothelial cells, thereby integrating into the vasculature (Asahara et al., 1999; Dimmeler & Zeiher, 2004; Jin et al., 2006). The level of circulating EPCs has been reported to predict the occurrence of future cardiovascular events and death from cardiovascular causes (Werner et al., 2005). Similarly, a low number of EPCs is an independent predictor of atherosclerosis progression (Schmidt-Lucke et al., 2005). Long-term oral administration of testosterone has been shown to induce both endothelium-dependent and endothelium-independent vasorelaxation (Kang et al., 2002). Although reduced testosterone concentrations have been linked to the impairment of the vasculogenic reparative process mediated by EPCs (Castela et al., 2011), the direct effect of testosterone on EPCs in men with LOH is not known.
Few studies have evaluated the association between circulating EPCs and serum testosterone levels in men. Low serum testosterone levels have been reported to be associated with low number of circulating EPCs in young men with hypogonadotropic hypogonadism (HH) (Foresta et al., 2006). In their study, it was shown that testosterone replacement therapy (TRT) was able to induce an increase in EPCs through a possible direct effect on bone marrow. In another study, patients with isolated arterial erectile dysfunction (ED) and LOH not receiving androgen therapy had a higher number of endothelial microparticles (EMPs) and EPCs, and showed worse vascular parameters compared with treated hypogonadal patients, indicating the benefit of TRT in improving endothelial function (La Vignera et al., 2011a). Interestingly, patients with arterial ED and metabolic syndrome were shown to have higher EPCs and EMPs, compared with healthy men, and tadalafil was capable of increasing the number of EPCs further, suggesting the persistence of an adequate bone marrow response (La Vignera, 2011; La Vignera et al., 2011b). However, the level of circulating EPCs was not significantly different between patients with and without testosterone deficiency in a cross-sectional study evaluating patients with chronic heart failure (Florvaag et al., 2012).
LOH has been recognized recently, and TRT has shown a treatment benefit. However, to date, there is no report on the evaluation of circulating EPCs after TRT in LOH. Here, we conduct a prospective study to evaluate the efficacy of transdermal testosterone gel in men with LOH, and further determine the effect of testosterone administration on the circulating EPC number. We also investigate the association between serum testosterone levels, the number of circulating EPCs and various clinical parameters in men with LOH.
Materials and methods
This was a non-randomized, open-label, single-arm, post-marketing surveillance study to evaluate the safety and efficacy of transdermal testosterone gel (AndroGel; Solvay Pharmaceuticals, Marietta, GA, USA) in men with LOH.
The study was performed in compliance with the guidelines stated in the Declaration of Helsinki. The Institutional Review Board and Ethics Committee of the Cardinal Tien Hospital approved this study (CTH-99-2-4-023). Each subject was informed of the study rationale and procedures, and written informed consent was obtained before treatment.
Men aged between 40 and 73 years (mean age, 58.3 years) with hypogonadal symptoms were screened for eligibility for the study by their medical history and the results of physical and laboratory examinations. The inclusion criteria were hypogonadal subjects with a morning (7–11 am) serum total testosterone (TT) concentration of <350 ng/dL or free testosterone (FT) concentration of <6.5 ng/dL at the study screening visit, and with another confirmed value at the baseline visit before study treatment (conducted at least 1 week apart). Exclusion criteria included breast or prostate cancer, erythrocytosis (hematocrit, >50%), hyperviscosity, untreated obstructive sleep apnoea, severe untreated benign prostatic hypertrophy with an international prostate symptom score of >19, or uncontrolled severe heart failure. Patients who received hormone replacement therapy, anti-hormone replacement therapy, or corticosteroids within 6 months before the study screening visit were also excluded. All subjects were asked to not change their lifestyles and the medications used during the study period. Each subject was informed of the study rationale and procedures, and written informed consent was obtained before treatment.
Eligible subjects received 50-mg testosterone gel (AndroGel; equivalent to 1% testosterone in 5 g) for 12 months. The study drug was applied once daily to clean, dry and intact skin of the shoulders and upper arms and/or abdomen, and was allowed to dry for a few minutes before dressing.
Laboratory assessments including high-sensitivity C-reactive protein, serum hormones [TT and sex hormone-binding globulin (SHBG)], serum chemistry [albumin, glucose, glycosylated haemoglobin (HbA1C) and insulin], and lipid profile [total cholesterol, high density lipoprotein-C (HDL-C), low-density lipoprotein-C (LDL-C) and triglycerides] were performed at baseline and at 3, 6 and 12 months. The calculated FT and bioavailable testosterone (BT) values were obtained (Vermeulen et al., 1999). Subjects were evaluated clinically at every visit; concomitant medications and adverse events were recorded. The sexual function was evaluated through the International Index of Erectile Function (IIEF) questionnaires (Rosen et al., 1997) by using the 15-item scores and five functional domain scores. Erectile function (EF), orgasmic function (OF), sexual desire (SD), satisfaction with intercourse (SI) and overall satisfaction (OS) were assessed by summing the scores assigned to the related individual questions in each domain.
Circulating early (outgrowth) EPC measurement using flow cytometry
The analysis of EPC mobilization was modified by previous demonstration (Huang et al., 2010). Indeed, the BD fluorescence-activated cell sorting (FACS)CantoII flow cytometer (BD Bioscience, San Jose, CA, USA) was used to assess the number of circulating early EPCs. Before being stained with specific monoclonal antibodies, the mononuclear fraction was separated by centrifugation with Ficoll-Paque PREMIUM 1.077 gradient (GE Healthcare Bio-Sciences AB, Björkgatan, Uppsala, Sweden) and then the samples were washed with buffer containing phosphate-buffered saline (PBS). Then, a volume of 100 μL mononuclear fraction was incubated for 45 min in the dark with fluorescein isothiocyanate (FITC)-conjugated anti-human CD34 monoclonal antibody (mAb) (BD Bioscience), phycoerythrin (PE)-conjugated anti-human KDR mAb (BD Bioscience), and allophycocyanin (APC)-conjugated anti-human CD133 (Miltenyi Biotec, Bergisch Gladbach, Germany) and without any antibody was used as a negative control. After incubation, cells were washed with PBS before analysis. Each analysis included 100 000 events. At the intersection of the CD34 and CD133 gates, triple-positive cells were identified by the dual expression of KDR and CD133 in the CD34 gate. Quadrants were set on the basis of isotype controls. Finally flow cytometry analysis was performed using FCS Express V3 (BD Bioscience) (Fig. 1). To assess the reproducibility of the EPC measurement, circulating EPCs were measured from two separate blood samples of 10 subjects, and there was a strong correlation between the two measurements (r = 0.86, p < 0.001).
Continuous variables were expressed as mean ± standard deviation (SD), and categorical data were expressed as numbers and percentages. Statistical comparisons between the groups were tested using Chi-square test for categorical variables, and the Wilcoxon rank sum test was used for continuous variables. Wilcoxon sign rank test was used to evaluate the significant difference of variables before and after treatment. Spearman's rank correlation analysis was used to determine the relationship between variables. All statistical assessments were two sided and were considered significant at p < 0.05. Statistical analyses were performed using the spss version 15.0 statistical software (SPSS Inc., Chicago, IL, USA).
Baseline characteristics of hypogonadal and eugonadal men
Forty-six men with hypogonadal symptoms were screened. Table 1 shows the baseline characteristics of 29 hypogonadal men with serum TT levels of <350 ng/dL (average age, 58.0 ± 10.1 years) in comparison with 17 eugonadal men with serum TT levels of >350 ng/dL (average age, 57.5 ± 9.1 years). There was no significant difference in age, body mass index (BMI) and waist circumference (WC) between the two groups. The baseline testosterone levels in the hypogonadal group were significantly lower than those in the eugonadal group (TT: 272.2 ± 54.0 ng/dL vs. 407.1 ± 26.1; FT: 5.63 ± 1.85 ng/dL vs. 8.20 ± 1.01 ng/dL; BT: 144.0 ± 48.0 ng/dL vs. 206.2 ± 24.4 ng/dL respectively) (p < 0.001 for all). In addition, the baseline SHBG levels in the hypogonadal group were significantly lower than those in the eugonadal group (27.1 ± 9.4 vs. 31.7 ± 5.5 nmol/L respectively; p = 0.04). The number of EPCs between the two groups did not differ significantly at baseline. No significant differences in other clinical parameters were observed between the two groups. In hypogonadal group, antihypertensives, lipid lowering and antidiabetic medication were used in 7, 3 and 4 men. The doses had remained stable for more than 3 months prior to enrolment, and were held constant during the trial. Co-morbidities such as cardiovascular disease and diabetes were found in 10 (34.4%) of hypogonadal men.
Table 1. Comparison of baseline characteristics between men with serum total testosterone levels less than 350 ng/dL (hypogonadal) or greater than 350 ng/dL (eugonadal)
Total (N = 46)
Men with serum TT < 350 ng/dL (N = 29)
Men with serum TT > 350 ng/dL (N = 17)
BMI: body mass index; BT: bioavailable testosterone; CRP: C-reactive protein; EPCs: endothelial progenitor cells; FT: free testosterone; HbA1C: haemoglobin A1C; HDL: high-density lipoprotein; LDL: low-density lipoprotein; LH: luteinizing hormone; SHBG: serum sex hormone-binding globulin; TG: triglycerides; TT: total testosterone; WC: waist circumference. *Indicates statistical significance, p < 0.05.
The correlation of number of EPCs and IIEF-15 score with baseline clinical parameters is shown in Table 2. A significant positive correlation was established between EPCs and insulin and between EPCs and IIEF-15 score. There was no significant association between serum TT level and circulating EPC number before TRT. EPC number correlated positively with baseline insulin level (r = 0.14, p < 0.05) and IIEF-15 score (r = 0.28, p < 0.05). The IIEF-15 score also correlated positively with TT, FT, and BT levels (r = 0.43, 0.11 and 0.05 respectively; p < 0.05 for all).
Table 2. Correlation analysis of endothelial progenitor cells and International Index of Erectile Function-15 score with various clinical parameters
BMI: body mass index; BT: bioavailable testosterone; CRP: C-reactive protein; EPCs: endothelial progenitor cells; FT: free testosterone; HbA1C: haemoglobin A1C; HDL: high-density lipoprotein; IIEF-15: International Index of Erectile Function-15; LDL: low-density lipoprotein; SHBG: serum sex hormone-binding globulin; TG: triglycerides; TT: total testosterone; WC: waist circumference. *Indicates statistical significance, p < 0.05.
Change in clinical parameters, sexual characteristics and EPC number following TRT
Hypogonadal men in the study received transdermal testosterone gel (Androgel) for a duration of 3, 6 and 12 months, and the clinical and sexual characteristics and the number of circulating EPCs (calculated per 100 000 monocytes) were evaluated at 3, 6 and 12 months. The comparison of the parameters before (baseline) and after TRT is shown in Table 3. The TT, FT and BT levels at 3 months (418.2 ± 256.1, 9.96 ± 6.76 and 233.7 ± 169.9 ng/dL respectively) and 6 months (499.2 ± 391.4, 13.55 ± 12.68 and 326.2 ± 298.3 ng/dL respectively) of TRT were significantly higher than those at baseline (272.2 ± 54.0, 5.63 ± 1.85 and 144.0 ± 48.0 ng/dL respectively) (p < 0.05 for all). After 12 months of TRT, the TT and FT levels were significantly higher (419.2 ± 314.2 and 11.72 ± 9.35 ng/dL) than those at baseline (272.2 ± 54.0 and 5.63 ± 1.85 ng/dL) (p < 0.05 for both). Although TRT increased the BT levels at 12 months as compared with those at baseline, the difference was not statistically significant. The IIEF-15 score significantly improved at 3 months (35.8 ± 19.0) and at 6 months (38.2 ± 18.9) of TRT when compared with that at baseline (29.9 ± 17.5; p < 0.05). There was a significant increase in the erectile function at 3 months (15.9 ± 8.7) and 6 months (16.8 ± 9.4) than at baseline (13.4 ± 8.3; p < 0.05). TRT did not induce any significant changes in surrogate markers related to atherosclerosis and diabetes (Table 3).
Table 3. Comparison of clinical parameters and sexual characteristics before and after testosterone replacement therapy (TRT)
Baseline (N = 29)
3 months (N = 29)
6 months (N = 24)
12 months (N = 20)
BMI: body mass index; BT: bioavailable testosterone; CRP: C-reactive protein; EPCs: endothelial progenitor cells; FT: free testosterone; HbA1C: haemoglobin A1C; HDL: high-density lipoprotein; IIEF-15: International Index of Erectile Function-15; LDL: low-density lipoprotein; SHBG: serum sex hormone-binding globulin; TG: triglycerides; TT: total testosterone; WC, waist circumference.*Indicates statistical significance, p < 0.05.
As seen in Table 3 and Fig. 2, compared with the mean circulating EPC number at baseline (10.13 ± 6.67), the number was significantly higher after 3 months (14.2 ± 9.7; p = 0.027), 6 months (16.6 ± 17.2; p = 0.006) and 12 months (20.3 ± 15.1; p = 0.017) of TRT.
Our results show that pre-TRT serum testosterone levels had no significant association with the number of circulating EPCs, irrespective of whether serum TT, FT, or BT level was analysed. However, following 3, 6 and 12 months of TRT, testosterone levels and EPC numbers increased significantly compared with that at baseline (with the exception of BT levels at 12 months of TRT). In addition, TRT significantly improved sexual characteristics such as erectile function and the IIEF-15 score. To our knowledge, this is the first long-term study to report the benefit of TRT in increasing the number of EPCs in men with LOH.
The integrity of the endothelial cell layer is of critical importance for the proper functioning of the vessel and for the prevention of vascular disorders (Op den Buijs et al., 2004). The damaged endothelium can be replaced in two ways: through the proliferation of surrounding ECs or through the proliferation and differentiation of bone-marrow-derived circulating EPCs (Op den Buijs et al., 2004). Because the overall rate of endothelial cell turnover is low (Schwartz & Benditt, 1976), EPCs may serve as an important source to replace the damaged endothelium. It has been suggested that the vascular repair of the injured endothelial monolayer occurs through the regenerative activity of EPCs (Dimmeler & Zeiher, 2004). Interestingly, in recent years, a reduction in the number of EPCs has been demonstrated in a number of pathological conditions such as cardiovascular disease, ED, diabetes and stroke (Loomans et al., 2004; Fadini et al., 2005; Foresta et al., 2005; Schmidt-Lucke et al., 2005; Werner et al., 2005; Bumhakel et al., 2006; Chu et al., 2008), which are conditions characterized by endothelial dysfunction. In particular, reduced EPCs were identified as an independent predictor of death from cardiovascular causes (Werner et al., 2005). These findings suggest that EPCs may represent a link between cardiovascular risk factors, endothelial dysfunction and ED (Bumhakel et al., 2006). Thus, an increase in EPCs may suggest an improvement in the endothelial function of hypogonadal men.
A previous study by Foresta et al. (2006) has demonstrated that TRT treatment administered for a duration of 6 months is capable of inducing a significant increase in circulating EPCs (with respect to baseline) in men with HH, possibly through a direct effect on the bone marrow. In this study, 10 young patients with HH (mean age, 28.6 ± 3.1 years) were treated with T gel therapy (50 mg/day) for 6 months, and the levels of circulating PCs and EPCs were evaluated. Our results are consistent with those obtained in the study by Foresta et al.; nevertheless, there are certain differences, which may be attributed to the different study populations. Our study included elderly patients with LOH who had more cardiovascular risk factors than the younger HH patients recruited in the study by Foresta et al. These risk factors may also influence the number of circulating EPCs. Therefore, the number of circulating EPCs at baseline may not be a good surrogate marker for LOH. These results were similar to previous findings that testosterone deficiency does not appear to impact the regenerative EPCs in patients with chronic heart failure (Florvaag et al., 2012). In addition, the patients in our study received Androgel for a longer duration (12 months). Furthermore, our study is the first to assess the correlation of EPC numbers with clinical and sexual parameters and to evaluate the effect of TRT on sexual characteristics.
We showed that TRT can increase the number of circulating EPCs in men with LOH. Although the exact mechanism of the effect of testosterone on EPCs is not clear, it has been reported that androgen receptor has been demonstrated to be widely expressed in the bone marrow and in particular on CD34-positive cells, which are the common precursors of many progenitor cells such as EPCs (Foresta et al., 2006). Androgens [synthetic androgen methyltrienolone (R1881)] also induce an increase in the proliferation, migration and colony formation activity of EPCs via an androgen receptor-mediated pathway (Foresta et al., 2008). Testosterone might play a role in the mechanism of EPCs release from bone marrow. Our study also showed a significant positive correlation between EPCs and insulin level, and EPCs and IIEF-15 score at baseline. These results support previous findings that EPC levels are significantly reduced in diabetes and ED (Loomans et al., 2004; Fadini et al., 2005; Foresta et al., 2005). Importantly, TT, FT and BT levels correlated positively with IIEF-15 score, which is a widely used, multi-dimensional, self-reporting tool for the assessment of ED and for the diagnostic evaluation of ED severity (Rosen et al., 2002). The exact mechanism underlying the effect of testosterone on ED is not known (Castela et al., 2011); however, a threshold level of testosterone is considered necessary for normal erectile function (Shabsigh et al., 2006). Hypogonadism may also play an important role in the pathophysiology of ED, and TRT in combination with phosphodiesterase type 5 (PDE5) inhibitors has proven to be effective in hypogonadal men with ED (Shabsigh et al., 2006; Foresta et al., 2009). It has been reported that TRT administered with sildenafil significantly improved erectile function and other sexual parameters in sildenafil-unresponsive hypogonadal men (Shabsigh et al., 2008). In agreement with these findings, we observed a significant improvement in IIEF-15 scores and erectile function in hypogonadal men after TRT. Recently, it has been shown that testosterone in combination with vardenafil, but not vardenafil alone, significantly increases the numbers of EPCs and PCs in hypogonadal men, suggesting that testosterone positively modulates PDE5 in the bone marrow (Foresta et al., 2009).
We have previously reported that transdermal testosterone gel treatment for hypogonadal patients can improve their sexual dysfunction mainly through restoring erectile function (Chiang et al., 2009). This study also provided similar results in terms of improvement in erectile function, as assessed by IIEF-15 scores. We also demonstrated the increase in circulating EPCs after TRT. These results reflect the important role of endothelial function in erectile function and may also imply that the improvement of erectile function after TRT may occur through the improvement of endothelial function.
It is known that low serum testosterone levels, as a result of hypogonadism or androgen-deprivation therapy in men with prostate carcinoma, produce adverse effects on cardiovascular health (Traish et al., 2009). Androgen deficiency is associated with increased levels of total cholesterol and low-density lipoprotein, increased production of proinflammatory factors, increased thickness of the arterial wall and is known to contribute to endothelial dysfunction. TRT is capable of restoring arterial vasoreactivity; reducing the levels of proinflammatory cytokines, total cholesterol, triglyceride and high-density lipoprotein; and improving endothelial function (Traish et al., 2009). In three randomized, placebo-controlled studies of hypogonadal men with metabolic syndrome, TRT reduced fasting glucose, improved insulin resistance, lipid profiles and WC; and improved surrogate markers of atherosclerosis and inflammation (Aversa et al., 2010; Kalinchenko et al., 2010; Jones et al., 2011). In this study, we did not find significant changes in any of these parameters at baseline or after TRT in hypogonadal men (serum testosterone, <350 ng/dL), which could be attributed to the small study population. Conducting a large-scale study with a longer duration of TRT (>12 months) may better discern the effects of TRT on markers of atherosclerosis, such as cholesterol, triglycerides and HDL cholesterol.
Hypogonadism and ED are the most common disorders observed in the ageing male (Castela et al., 2011), which was our study population. A testosterone decline has also been associated with chronic medical diseases such as metabolic syndrome, diabetes mellitus and cardiovascular disease, which are conditions with an increased morbidity and mortality (Jones, 2007). Moreover, EPCs have been proposed as markers of endothelial dysfunction (La Vignera et al., 2011a). In view of this emerging evidence and our results, androgen replacement therapy could potentially reduce the risk of the aforementioned disorders in hypogonadal men. When performed with caution and with careful monitoring for prostate diseases, TRT may be successfully used for improving endothelial dysfunction in men with LOH. The major limitation of our study was lack of a placebo group, and the fact that we cannot eliminate the possible placebo effect after treatment.
Before initiating TRT, serum TT levels had no significant association with circulating EPCs in men with LOH. TRT increased the number of circulating EPCs, which implies the benefit of TRT on endothelial function in ageing men with LOH. TRT also improved IIEF-15 scores and erectile function in hypogonadal men.
This work was supported by the National Science Council (NSC-100-2314-B-567-001-MY3), Taiwan, and by the Cardinal Tien Hospital (CTH-100-1-2A16 and CTH-101-1-2B09). We would like to acknowledge Cactus Communications for their writing and editing services.