Flora F Ip[1,2], Irene di Pierro[1], Ross Brown[3], Ilona Cunningham[4], David J Handelsman[1] and Peter Y Liu[1]

[1]Department of Andrology, Concord Hospital and ANZAC Research Institute, University of Sydney, Sydney, New South Wales 2139, Australia, [2]Department of Endocrinology, Concord Hospital, Sydney, New South Wales 2139, Australia, [3]Department of Hematology, Royal Prince Alfred Hospital, Sydney, New South Wales 2050, Australia and [4].Department of Hematology, Concord Hospital, Sydney, New South Wales 2139, Australia.

Wales 2050, Australia and Department of Hematology, Concord Hospital, Sydney, New South Wales 2139, Australia (Correspondence should be addressed to P Y Liu who is now at Woolcock Institute of Medical Research, University of Sydney, Glebe, New South Wales 2037, Australia; E-mail: pliu@mail.usyd.edu.au.

Abstract

Objectives: Testosterone formulations that have more steady-state pharmacokinetics, such as subcutaneously implanted testosterone pellets, may cause less erythrocytosis than i.m. injections of shorter acting androgen esters. We, therefore, sought to define the prevalence, predictors, and proximate basis (role of erythropoietin) for polycythemia (hematocrit >0.50) in hypogonadal men receiving testosterone implants long term.

Design: A cross-sectional study was conducted in an academic andrology center with a longitudinal subgroup analysis.

Patients: A total of 158 hypogonadal men aged 14-84 years (mean age 46.7 years) treated on average for 8 years (range 0-21 years).

Measurements: Trough blood testosterone and hematocrit. Serial serum erythropoietin concentrations were measured in 16 volunteers.

Results: Positive univariate associations between polycythemia (hematocrit >0.50) and log(testosterone) (odds ratio (OR) 24.7, 95% confidence interval (CI): 4.3, 141.2, P<0.01) and age (OR 1.1, 95% CI: 1.0, 1.1, P = 0.03) and a borderline relationship with current smoking (OR 4.2, 95% CI: 0.9, 20.0, P=0.08) were unveiled. A sensitivity analysis using alternate definitions of polycythemia was performed to capture all potential covariants. Multivariate regression analysis incorporating all potential covariants disclosed the independent OR of developing polycythemia (after adjusting for smoking and age) for log(testosterone) to be 15.0 (95% CI: 2.5, 90.8). Duration of testosterone therapy did not alter the risk of polycythemia. A direct relationship between testosterone and erythropoietin was observed (P=0.05).

Conclusions: Higher trough serum testosterone concentrations but not duration of treatment predict the development of polycythemia in men receiving long-acting depot testosterone treatment.

Introduction

Secondary polycythemia can occur after testosterone therapy, although the risk varies depending on the actual formulation. Randomized studies in hypogonadal men show that more erythrocytosis is induced with i.m. injections of 200 mg testosterone enanthate every 2 weeks compared with 5 mg testosterone patches applied transdermally daily (1) or with 100 mg compared with 50 mg of transdermal testosterone (2). On the other hand, i.m. injections of 250 mg testosterone enanthate every 3 weeks, oral testosterone undecanoate 160 mg/day, or subcutaneous implantation of 1200 mg crystallized testosterone are equivalent (3). As a unifying hypothesis, these data suggest that both the dose and pharmacokinetics of specific testosterone formulations (rather than the actual route of administration) are important because both influence the relative amount of time blood testosterone concentrations remain in the supraphysio- logical range. Indeed, others have postulated that the increased rates of polycythemia with short-acting testosterone esters are due to supraphysiological levels of testosterone (1, 4, 5). If true, longer acting testosterone preparations such as testosterone implants and i.m. injections of testosterone undecanoate, which do not cause prolonged supraphysiological testosterone levels when administered at appropriate doses (6), may therefore result in a lesser degree of erythrocytosis. In support of this hypothesis, direct relationships between testosterone and hematocrit have already been demonstrated (2, 3, 7). 

A number of other biological factors modulate the relationship between testosterone and hematocrit. Of these, age (7), smoking (8, 9), and obesity (10) are the most important factors. Increasing age is associated with a well-defined decline in systemic testosterone exposure (11, 12) and a less well-defined decline in hematocrit in very old age (13, 14). Current cigarette smoking is associated with both increased testosterone concentrations (9, 11, 15) and polycythemia (8). Finally, body mass index (BMI) is inversely related to blood testosterone concentrations (11, 12) and confounds measurement of red cell mass (16).

Although these biological factors modulate the relationship between testosterone and erythrocytosis, the exact mechanism or mechanisms by which this occurs requires further evaluation. Erythropoietin is the key regulator of erythrocytosis and therefore a likely candidate. However, the currently available data examining its role are contradictory (7, 17-20). Whether these contradictory findings arise from differing study populations, study design, duration of therapy, or pharmacokinetics of testosterone administration are unknown. Indeed, the three available randomized studies ceased existing testosterone therapy for only 4-6 (1, 2) or 12 weeks prior (3), which may not have been sufficient, given that a single hematopoietic cycle requires 90 days to complete. Furthermore, only one large long-term study of androgen therapy using longer acting preparations is currently available (10), and there are no larger studies of longer duration examining subcutaneously implanted testosterone pellets.

For these reasons, our aim was to define the prevalence, predictors, and potential mediators of polycythemia in androgen-deficient men being treated long term with subcutaneous testosterone pellets. Such information would be useful in formulating surveillance strategies for long-term testosterone replacement with longer acting preparations as well as to inform understanding of the physiological actions of testosterone.

Materials and methods

One hundred and fifty-eight men with established hypogonadism due to hypothalamo-pituitary or testicular disease (excluding andropause) were treated with subcutaneously implanted testosterone pellets at the Department of Andrology (Concord Hospital, Sydney, Australia) from March to November 2006. Men received testosterone pellet therapy primarily based on patient preference and had easy access to alternative forms of testosterone therapy if desired. A cross-sectional analysis was undertaken to define the characteristics of these men in order to investigate the effects of testosterone pellet therapy on hematocrit levels. Ethical approval was provided by the Sydney

South West Area Health Service Human Research Ethics Committee (Concord Hospital).

Blood was collected immediately prior to implantation of testosterone for later measurement of reproductive hormones (total testosterone and sex hormone-binding globulin, SHBG) and hematocrit. Anthropometric data (including height and weight) as well as demographic data (including age and smoking status) were collected during each visit.

Mean corpuscular volume (MCV) and red cell count (RCC) were measured by Cell-Dyn 4000 (Abbot Diagnostics) from which the hematocrit was calculated using the formula: MCVXRCC/100. The coefficient of variation (CV) of this assay was 1.2%. Hematocrit exceeding 0.50 was defined as polycythemia in accordance with our laboratory reference range, as there is no universally accepted definition of polycythemia in investigating androgen-induced erythrocytosis. However, other thresholds were also evaluated in a sensitivity analysis. Trough testosterone assays were performed using the ImmuLite 2000 (DPC, Alameda, CA, USA). The inter-assay CV was 10.5% and the intra-assay CV was 8.7% for testosterone. Serum erythropoietin assays were performed on a Beckman Coulter UniCel 800 (Beckman Coulter, Fullerton, CA, USA). The inter-assay CV of this assay was 4.2% and the intra-assay CV was 3.3%.

In the longitudinal study, 16 men volunteered to allow blood to be collected serially (generally monthly) to explore the relationship between testosterone and erythropoietin. In these samples, erythropoietin and reproductive hormones were also measured.

Statistical analysis

Age, trough testosterone concentrations, SHBG, BMI, height and weight, smoking status, diagnosis, and duration of treatment were considered to be possible predictors of polycythemia. The duration of treatment was calculated based on an average 6-month duration between implantation of testosterone pellets (6). Smoking and primary diagnosis were modeled as categorical variables. Univariate analyses were conducted on the covariates listed above.

Logistic regression was used to assess factors that alter the odds of polycythemia (thereby modeling hematocrit categorically). Multivariate models were built by forward stepwise logistic regression and confirmed by best subset analysis. A sensitivity analysis was also performed using different cutpoints for the definition of polycythemia to ensure that all potential covariants that might modify the relationship between testosterone and hematocrit were captured.

In order to confirm the results of logistic regression, univariate and multivariate linear regressions were then performed analogously to define continuous relationships of study variables with hematocrit.

In the longitudinal substudy, data were analyzed using the mean or mean change of repeated measurements. Nonparametric Spearman correlation statistics, which are robust to outliers, were calculated to define the association between blood testosterone with hematocrit and erythropoietin.

Data are expressed as mean + S.E.M. For categorical variables, the number in each group and the percentage of the total are given. Two-tailed P values of <0.05 were considered statistically significant. Analyses were executed with proc logistic, proc reg, and proc corr using SAS version 9.1 (SAS Institute, Cary, NC, USA).

Results

Baseline characteristics of the 158 men in the cross-sectional analysis are shown in Table 1. Eight of 158 men (5.1%) developed polycythemia (hematocrit exceeding 0.50 in accordance with our laboratory reference range). No cardiac or thromboembolic events occurred in those with secondary polycythemia.

Table 1. Baseline characteristics of 158 men in cross-sectional analysis.

Logistic regression

Univariate regression consistently revealed an increased odds of polycythemia with higher trough blood testos-terone concentrations, using the three cutpoints (hematocrit in excess of 0.49, 0.50, and 0.51) chosen for the sensitivity analysis (Table 2). Trough blood testosterone concentrations exhibited both the highest odds for polycythemia and the greatest significance amongst all covariables. Age was also a consistent univariate predictor of polycythemia (P < 0.03), whereas smoking history (current versus not) was a significant risk factor only for marked polycythemia (P = 0.04). Duration of treatment was not a risk factor for the development of polycythemia in any model when analyzed as a continuous variable, or dichotomized according to whether treatment duration was more than 1 year, or not (data not shown). Similarly, BMI did not increase the odds of polycythemia when analyzed as a continuous variable, or dichotomized according to whether obesity (BMI> 30 kg/m²) was present or not (data not shown).

Table 2. Univariate odds for polycythemia.

Multivariate analyses were performed to assess the independent odds of polycythemia after adjusting for significant factors identified from the univariate analyses (Table 3). Trough testosterone concentrations remained a highly statistically significant predictor of polycythemia in all models (P < 0.001, for each) after adjustment for presumptive factors such as age or smoking. Age (P = 0.02) and smoking (P=0.03) were significant predictors, but only with cutpoints of 0.49 and 0.51 respectively. Other presumptive factors (including obesity) were not independently significant, when added to any combination of testosterone, age, and smoking (not shown).

Table 3. Multivariate odds for polycythemia.

Linear regression

Univariate linear regression confirmed that increasing trough blood testosterone concentration and age were the only factors significantly related to increasing hematocrit (Fig. 1). Multivariate linear regression disclosed an independent relationship only with trough blood testosterone (Table 4).

Table 4. Multivariate linear regressions with hematocrit.

Figure 1. Univariate linear regressions with hematocrit. The linear relationship between testosterone (left upper), age (right upper), treatment duration (left lower), and smoking status (right lower) and hematocrit is plotted for 158 men. Pearson’s correlation coefficient and P values are shown. Regression lines are plotted for significant relationships.

Longitudinal analysis of effects of erythropoietin

Sixteen men underwent serial blood sampling for measurement of erythropoietin, testosterone, and hematocrit. These men were on average 49.8 years of age (range: 26.8-83.9 years) with a BMI of 28.9 kg/m² (22.6-39.3 kg/m²) and treated on average for 7 years (0-20 years). Four of the 16 men were current smokers.

Mean testosterone and mean hematocrit (P < 0.01), as well as the mean change in testosterone and mean change in hematocrit levels (P = 0.03), were significantly related (Fig. 2). Mean serum erythropoietin levels were weakly associated with mean serum testosterone (P=0.05), but no association between the mean changes in serum testosterone and mean changes in erythropoietin levels were observed (P= 0.45; Fig. 2).

Figure 2. Univariate relationships between testosterone, hematocrit, and erythropoietin. The univariate relationships between testosterone and hematocrit (bottom left), testosterone and erythropoietin (bottom right), and change in testosterone and change in hematocrit (upper left), and change in testosterone and change in erythropoietin (upper right) are shown for 16 men. P values from nonparametric Spearman correlation are shown. See text for further details.

Discussion

The prevalence of secondary polycythemia was 5.1% in hypogonadal men treated with testosterone pellets for up to 21 years. Other studies have reported the rates of secondary polycythemia that range from 2.5 to over 40% depending on the dose and testosterone formulation (1-3, 10, 21, 22). These individuals are still likely to be at increased risk for cardiovascular, cerebrovascular, and peripheral vascular mortality (23) or events that may occur due to augmented blood viscosity associated with polycythemia, despite testosterone-associated thromboembolic events not yet being reported in those receiving testosterone replacement therapy (4).

Higher trough blood testosterone concentration is the most important factor associated with higher hematocrit, consistent with other reports (2, 7, 10) and studies demonstrating a linear, dose-response relationship between testosterone and hematocrit (2, 3, 7). We show additionally through multivariate analysis that higher trough blood testosterone concentrations remain an important and significant risk for polycythemia, independently of a range of confounding clinical and biochemical factors. Together, these data suggest that the lowest efficacious dose required to achieve control of clinical features of androgen deficiency, e.g. hypogonadal symptoms, bone loss, fractures, should be prescribed in order to reduce the risk of polycythemia.

The duration of treatment was not associated with the development of secondary polycythemia. This novel finding has not been previously reported and may relate to the long average treatment duration of 8 years (range 0-21 years) experienced by our volunteers. In contrast, the next largest study treated men for <2 years on average and up to 9.5 years (10), and the available prospective randomized trials treated men for < 1 year (1-3). These data provide some reassurance that prolonged treatment is not associated with adverse hematological outcomes, and is important given that hypogonadism requires lifelong therapy.

Increasing age was associated with increased odds for the development of polycythemia in androgen- deficient men treated with testosterone. Physiological experiments where normal men are rendered acutely androgen deficient, and then treated with graded and matched doses of testosterone confirm this age association (7).

Although testosterone therapy in androgen-deficient men clearly stimulates erythrocytosis and increases the odds of polycythemia fourfold (24), the mechanism by which this occurs remains elusive. We show a direct relationship between testosterone and immuno- reactive erythropoietin, which was of borderline significance (P = 0.05), suggesting but not proving that erythropoietin concentrations may be a mechanism through which erythrocytosis is stimulated. Androgens also increase tissue sensitivity to erythropoietin in adults with renal failure (25), and this mechanism may also occur in androgen-deficient men treated with testosterone even in the absence of any change in bioactivity or immunoreactive erythropoietin. However, we did not assess erythropoietin sensitivity. On the other hand, physiological experiments in normal men do not show a relationship between testosterone dose and immunoreactive erythropoietin after 5 months of treatment (7). Whether these discrepant findings are due to study populations (acute versus chronic androgen deficiency), duration of therapy or other factors is unknown.

Strengths of this study include the large number of hypogonadal men studied and the long treatment duration, which are double the sample size and four times the treatment duration of the next largest study of a long-acting testosterone formulation (10). Caveats include the lack of determination of a number of factors which may influence erythropoiesis, including iron stores, folate, and renal function. However, these factors were not measured because they were considered likely to be normal in our population of otherwise well community-dwelling men.

We conclude that higher trough serum testosterone levels but not duration of treatment predict the development of polycythemia in men receiving long-acting depot testosterone treatment. These data inform clinical practice. Further investigation of the mechanisms by which polycythemia develops is required.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector. PYL was supported by Cancer Development Award 511929 from the National Health and Medical Research Council of Australia.

Acknowledgements

The authors thank Ann Conway, Susan Wishart, and all the staff of the Department of Andrology, Concord Hospital for the clinical care of the patients who volunteered for this study.

References

Dobs AS, Meikle AW, Arver S, Sanders SW, Caramelli KE & Mazer NA. Pharmacokinetics, efficacy, and safety of a permeation- enhanced testosterone transdermal system in comparison with bi-weekly injections of testosterone enanthate for the treatment of hypogonadal men. Journal of Clinical Endocrinology and Metabolism 1999 84 3469-3478. 

Wang C, Swerdloff RS, Iranmanesh A, Dobs A, Snyder PJ, Cunningham G, Matsumoto AM, Weber T & Berman N. Transdermal testosterone gel improves sexual function, mood, muscle strength, and body composition parameters in hypo- gonadal men. Journal ofClinical Endocrinology and Metabolism 2000 85 2839-2853. 

Jockenhovel F, Vogel E, Reinhardt W & Reinwein D. Effects of various modes of androgen substitution therapy on erythropoiesis. European Journal of Medical Research 1997 2 293-298. 

Rhoden EL & Morgentaler A. Risks of testosterone-replacement therapy and recommendations for monitoring. New England Journal of Medicine 2004 350 482-492. 

Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS & Montori VM. Testosterone therapy in adult men with androgen deficiency syndromes: an Endocrine Society clinical practice guideline. Journal of Clinical Endocrinology and Metabolism 2006 91 1995-2010. 

Kelleher S, Howe C, Conway AJ & Handelsman DJ. Testosterone release rate and duration of action of testosterone pellet implants. Clinical Endocrinology 2004 60 420-428. 

Coviello AD, Kaplan B, Lakshman KM, Chen T, Singh AB & Bhasin S. Effects of graded doses of testosterone on erythropoiesis in healthy young and older men. Journal of Clinical Endocrinology and Metabolism 2008 93 914-919. 

Yanbaeva DG, Dentener MA, Creutzberg EC, Wesseling G & Wouters EF. Systemic effects of smoking. Chest 2007 131 1557-1566. 

Shiels MS, Rohrmann S, Menke A, Selvin E, Crespo CJ, Rifai N, Dobs A, Feinleib M, Guallar E & Platz EA. Association of cigarette smoking, alcohol consumption, and physical activity with sex steroid hormone levels in US men. Cancer Causes & Control 2009 20 877-886. 

Zitzmann M & Nieschlag E. Androgen receptor gene CAG repeat length and body mass index modulate the safety of long-term intramuscular testosterone undecanoate therapy in hypogonadal men. Journal of Clinical Endocrinology and Metabolism 2007 92 3844-3853. 

Wu FC, Tajar A, Pye SR, Silman AJ, Finn JD, O’Neill TW, Bartfai G, Casanueva F, Forti G, Giwercman A, Huhtaniemi IT, Kula K, Punab M, Boonen S & Vanderschueren D. Hypothalamic- pituitary-testicular axis disruptions in older men are differentially linked to age and modifiable risk factors. Journal of Clinical Endocrinology and Metabolism 2008 93 2737-2745. 

Liu PY, Beilin J, Meier C, Nguyen TV, Center JR, Leedman PJ, Seibel MJ, Eisman JA & Handelsman DJ. Age-related changes in serum testosterone and sex hormone binding globulin in Australian men: longitudinal analyses of two geographically separate regional cohorts. Journal of Clinical Endocrinology and Metabolism 2007 92 3599-3603. 

Caprari P, Scuteri A, Salvati AM, Bauco C, Cantafora A, Masella R, Modesti D, Tarzia A & Marigliano V. Aging and red blood cell membrane: a study of centenarians. Experimental Gerontology 1999 34 47-57. 

Sansoni P, Cossarizza A, Brianti V, Fagnoni F, Snelli G, Monti D, Marcato A, Passeri G, Ortolani C & Forti E. Lymphocyte subsets and natural killer cell activity in healthy old people and centenarians. Blood 1993 82 2767-2773. 

Svartberg J, Midtby M, Bonaa KH, Sundsfjord J, Joakimsen RM & Jorde R. The associations of age, lifestyle factors and chronic disease with testosterone in men: the Tromso Study. European Journal of Endocrinology 2003 149 145-152. 

Leslie WD, Dupont JO & Peterdy AE. Effect of obesity on red cell mass results. Journal of Nuclear Medicine 1999 40 422-428. 

Dickerman RD, Pertusi R, Miller J & Zachariah NY. Androgen- induced erythrocytosis: is it erythropoietin? American Journal of Hematology 1999 61 154-155. 

Alexanian R. Erythropoietin and erythropoiesis in anemic man following androgens. Blood 1969 33 564-572. 

Schooley JC. Inhibition of erythropoietic stimulation by testosterone in polycythemic mice receiving anti-erythropoietin. 

Proceedings of the Society for Experimental Biology and Medicine 1966 122 402-403. 

Cui YG, Tong JS, Pan QQ, Di FS, Jia Y, Feng T, Liu Y, Wang XH & Zhang GY. Effect of androgen on erythropoietin in patients with hypogonadism. Zhonghua Nan Ke Xue 2003 9 248-251. 

Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Lenrow DA, Holmes JH, Dlewati A, Santanna J, Rosen CJ & Strom BL. Effect of testosterone treatment on body composition and muscle strength in men over 65 years of age. Journal of Clinical Endocrinology and Metabolism 1999 84 2647-2653. 

Steidle C, Schwartz S, Jacoby K, Sebree T, Smith T & Bachand R. AA2500 testosterone gel normalizes androgen levels in aging males with improvements in body composition and sexual function. Journal of Clinical Endocrinology and Metabolism 2003 88 2673-2681. 

Brown DW, Giles WH & Croft JB. Hematocrit and the risk of coronary heart disease mortality. American Heart Journal 2001 142 657-663. 

Calof OM, Singh AB, Lee ML, Kenny AM, Urban RJ, Tenover JL & Bhasin S. Adverse events associated with testosterone replacement in middle-aged and older men: a meta-analysis of randomized, placebo-controlled trials. Journals of Gerontology Series A: Biological Sciences and Medical Sciences 2005 60 1451-145 7. 

Ballal SH, Domoto DT, Polack DC, Marciulonis P & Martin KJ. Androgens potentiate the effects oferythropoietin in the treatment of anemia of end-stage renal disease. American Journal of Kidney Diseases 1991 17 29-33.