Abstract
Objective
Indirect evidence suggests that the effects of testosterone on fat mass in men are dependent on aromatization to estradiol (E2). However, no controlled study has assessed the effects of E2 in the absence of testosterone.
Design
Six-month randomized, placebo-controlled trial with the hypothesis that men randomized to E2 would reduce their fat mass.
Methods
Seventy-eight participants receiving androgen deprivation therapy for prostate cancer were randomized to 0.9 mg of 0.1% E2 gel per day, or matched placebo. Dual x-ray absorptiometry body composition was measured at baseline, month 3, and month 6. The primary outcome was total fat mass.
Results
Serum E2 increased in the estradiol group over 6 months compared to placebo, and mean-adjusted difference (MAD) was 207 pmol/L (95% CI: 123–292), P < 0.001. E2 treatment changed total fat mass, MAD 1007 g (95% CI: 124–1891), but not significantly, so P = 0.09. There were other consistent non-significant trends toward increased proportional fat mass, MAD 0.8% (95% CI: 0.0–1.6), P= 0.15; gynoid fat, MAD 147 g (95% CI: 2–293), P = 0.08; visceral fat, 53 g (95% CI: 1–105) P = 0.13; and subcutaneous fat, MAD 65 g (95% CI: 5–125), P = 0.11. Android fat increased, MAD 164 g (95% CI: 41–286), P = 0.04.
Conclusion
Contrary to our hypothesis, we provide suggestive evidence that E2 acting in the absence of testosterone, may increase total and regional fat mass in men. Given the premature closure of clinical trials due to the COVID pandemic, this potentially important observation should encourage additional studies to confirm or refute whether E2 promotes fat expansion in the absence of testosterone.
Introduction
Exogenous testosterone administration reduces fat mass in hypogonadal men (1, 2, 3) and in older men with age-associated low serum testosterone (4, 5). By comparison, aromatase inhibitors, which increase serum testosterone and reduce estradiol (E2), do not decrease fat mass in older men with low testosterone (6) raising the question of whether the effect of testosterone on fat mass is aromatization-dependent (7). Supporting this concept are observations that men with germline loss of function of the estrogen receptor-alpha (ER-alpha) (8) or aromatase (9) genes display excess adiposity and insulin resistance, and experimental data suggesting that intact aromatization is required to prevent fat gain induced by short-term hypogonadism (10, 11). In women, E2 replacement reduces the postmenopausal accumulation of central fat (12). However, the hypothesis that E2 prevents fat accumulation in men, independently of testosterone, has not been tested experimentally by investigating E2 actions in the absence of testosterone.
Mouse studies are generally supportive of the notion that regulatory effects of testosterone on fat mass in adult males is, at least in part, dependent on the actions of E2. The male global ER-alpha knockout mouse displays age-dependent accumulation of white adipose tissue due to an increase in adipocyte size and number that is not due to increased food consumption, but rather lower overall energy expenditure (13, 14). Male aromatase knockout mice (15) display a similar phenotype. Tissue-specific knockout models have highlighted direct and indirect effects for E2 acting on ER-alpha in adipose tissue, brain, and possibly skeletal muscle in preventing fat accumulation and inflammation but also that E2 acting via ER-alpha is necessary for proliferation and differentiation of pre-adipocytes during development (16). Experiments in orchiectomized adult male mice have shown that E2 is important in prevention of fat accumulation, possibly with a visceral depot-specific effect (17, 18, 19).
Androgen deprivation therapy (ADT) for prostate cancer is the most frequent contemporary cause of severe sex steroid deficiency in older men, with serum testosterone and its substrate E2 circulating at near castrate concentrations (20). Medical castration is used for up to 3 years in the adjuvant setting in combination with curative-intent radiotherapy, and as palliative treatment for metastatic disease (21, 22). ADT is usually achieved by administration of gonadotropin-releasing hormone (GnRH) analogs, representing the only clinical situation where standard clinical care allows for prolonged deferral of testosterone replacement. This creates the opportunity for the isolated effect of E2 to be observed in the absence of testosterone.
Men commencing ADT typically gain fat mass and lose lean mass. In a meta-analysis of seven uncontrolled observational studies (325 participants), ranging in duration from 3 to12 months, the pooled estimate of mean change in fat mass by dual x-ray absorptiometry (DXA) was 7.7% (95% CI: 4.3–11.2%, P < 0.0001) (23). In six studies (260 participants), mean change in lean mass was −2.8% (95% CI: −3.6 to −2.0%), P < 0.00001 (15). In a controlled experiment enrolling healthy older male volunteers, 16-weeks of ADT increased fat mass by 12% and reduced lean mass by 2% (3).
In this trial, we aimed to assess the effect of transdermal E2 on fat mass over 6 months in men undergoing medical castration with ADT. The E2 dose was selected to target a minimum circulating E2 concentration in the range present in eugonadal men (E2 ‘add-back’). We hypothesized that men randomized to E2 add-back, compared to those randomized to placebo, would reduce their fat mass.
Subjects and methods
We conducted a 6-month, randomized, double-blinded, placebo-controlled, parallel-group trial at Austin Health, a tertiary referral hospital affiliated with The University of Melbourne. Participants were recruited from outpatient clinics from November 2017 to February 2020 until recruitment was terminated prematurely due to the COVID-19 pandemic. Men were eligible for the study if they had been receiving GnRH agonists or antagonists for prostate cancer for a minimum of 4 weeks, with that therapy intended to continue for at least a further of 6 months. Exclusion criteria were impaired performance status (Eastern Cooperative Oncology Group Performance Status (ECOG) >2); bone metastases at scanning sites; history of venous thromboembolism (VTE); breast cancer; prior antiresorptive or strontium ranelate use or current indication for such therapy; systolic blood pressure >160 or diastolic blood pressure >100; New York Heart Association class 3 or 4 angina or heart failure; stroke, transient ischaemic attack, myocardial infarction, or angina within 12 months; current oral glucocorticoid treatment; prior chemotherapy; and alcohol or illicit drug abuse.
The trial protocol was approved by the Austin Health Human Research Ethics Committee (HREC/16/Austin/98) and each participant provided written informed consent. The trial was preregistered with the Australian New Zealand Clinical Trials Registry (identifier 12614000689673). We followed the Consolidated Standards of Reporting Trials checklist in reporting this randomized trial (24).
Participants were randomly allocated, in a concealed fashion by clinical pharmacists independent of trial investigators, to two intervention groups: E2 gel 1 mL (0.9 mg) per day or matching placebo gel 1 mL per day. We previously observed that E2 gel 0.9 mg daily increased the minimum serum E2 of men undergoing ADT into the reference range for healthy older men (25). Randomization occurred as follows: First, participants were stratified by ADT duration (≤3 or >3 months) and then by eligibility to undergo brain MRI (data to be reported separately). Secondly, participants were allocated by restricted randomization, using a computer-generated randomization scheme in blocks of size 4, to E2 or placebo in a ratio of 1:1.
E2 gel was Sandrena™ 1 mg/g E2 (Aspen Pharmacare, St Leonards, Australia). Placebo gel was a-gel™ (Fresenius Kabi, North Ryde, Australia) and matched the E2 gel for color , smell, and consistency. E2 and placebo were re-packaged into identical 10 mL syringes by pharmacy, with instructions to apply 1 mL each morning to the skin of the upper arms or abdomen. The concealed treatment allocation maintained the blinding of participants, investigators, and clinicians during treatment, with blinding maintained until data analysis, after the database had been cleaned and locked. Used syringes were retrieved to calculate gel usage.
Study visits were conducted at baseline, month 1, month 3, and month 6 of the intervention. Gel syringes were collected during each visit and residual volume was recorded to assess adherence. Adverse events were graded according to Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 (26).
Self-reported time spent performing weight-bearing exercise was recorded at baseline and average daily step count and proportion of awake time spent doing non-sedentary activity were measured for 1 week at baseline and 6 months by accelerometer (GT3x, Actigraph, Pensacola, USA). Participants were given standard advice at baseline and throughout the study encouraging regular cardiovascular and resistance exercise.
Dual X-ray absorptiometry (DXA)
Body composition was measured using the Hologic Horizon A DXA system, software version APEX 13.5.6 (Hologic Inc, Bedford, USA). The software reports total body and regional fat mass and lean mass. The android region is that component of the trunk from the top of the iliac crests to a height 20% of the distance from the iliac crest to the chin (27). The gynoid region is bordered superiorly by a horizontal line drawn below the iliac crest at a distance of 1.5 times the height of the android region. The gynoid region extends inferiorly for a distance of two times the height of the android region, such that it incorporates the hips and upper thighs (28). Visceral fat mass is estimated from a subregion of the android region extending 5.2 cm superiorly from the top of the iliac crests. The software estimates the total circumferential subcutaneous fat in this region by extrapolating from measured subcutaneous fat between the skin and abdominal wall musculature on both sides. This estimate of subcutaneous fat is subtracted from the measured total fat in this region to yield an estimate of visceral fat (27, 29). Coefficients of variation (CV) for all body composition parameters using this DXA system in our hands were <3% (30).
Anthropometry
At each study visit, height, and weight were measured. Waist circumference was measured according to World Health Organisation guidelines (31). Grip strength was measured in triplicate bilaterally to the nearest kilogram using a hand grip dynamometer (Jamar 5030J1, Sammons Preston, Bolingbrook, USA). The best of the six measurements was used for statistical analysis (32).
Blood samples
Fasting morning pre-dose blood samples were drawn at each visit. Serum was stored at −80°C for batched analysis of sex steroid profile by liquid chromatography mass spectrometry (LCMS/MS) (33). Limits of detection (LOD) and quantification (LOQ) were 11 pmol/L and 18 pmol/L for E2; 4 pmol/L and 11 pmol/L for estrone (E1); and 0.03 nmol/L and 0.09 nmol/L for testosterone, respectively. CV were 5–10, 4–9, and 2–8%, respectively, for within-run reproducibility at three levels of quality control samples.
Other biochemistry was performed using routine procedures at the biochemistry department, Austin Health. Sex hormone-binding globulin (SHBG) was measured by immunoassay (Roche Diagnostics) with interassay CV of 6% at 21 nmol/L and 6% at 40 nmol/L. Fasting glucose was measured using a hexokinase photometric assay (Roche Diagnostics) with CV of 1.5% at 4.8 mmol/L and 15.5 mmol/L. Insulin and C-peptide were measured by electrochemiluminescence immunoassay (Roche Diagnostics). CV were 4% at 16.3 mIU/L and 5% at 154 mIU/L for insulin and 4.5% at 2.5 nmol/L and 6.8% at 0.55 nmol/L for C-peptide, respectively. Fasting glucose and C-peptide were used to calculate insulin resistance using homeostatic model assessment of insulin resistance (HOMA2-IR) as described (34). Glycated hemoglobin (HbA1c) was measured using turbidimetric inhibition immunoassay (Roche Diagnostics) with CV of 2.7% at 38 mmol/mol and 1.9% at 80 mmol/mol, respectively. Insulin-like growth factor-1 (IGF-1) was measured using chemiluminescence immunoassay (DiaSorin) with CV of 10% at 11.4 nmol/L and 8.5% at 42.2 nmol/L, respectively. Prostate-specific antigen (PSA) was measured using electrochemiluminescent immunoassay (Roche Diagnostics) with a minimum detection limit of 0.03 mcg/L and CV of 2.3% at 0.9 mcg/L and 2.1% at 19.5 mcg/L, respectively. Total cholesterol, HDL-cholesterol, and triglycerides were measured by enzymatic assay (Roche Diagnostics). CV were 2.1% at 2.8 mmol/mol and 1.6% at 6.9 mmol/mol, 2.9% at 0.7 mmol/mol and 2.0% at 1.4 mmol/mol, and 2.9% at 1.0 mmol/mol and 2.0% at 2.0 mmol/mol, respectively. LDL cholesterol was calculated using the Friedewald equation.
Study design and pre-specified outcomes
The pre-specified primary outcome was total fat mass. This study assessed a second, independent primary endpoint of total volumetric bone mineral density at the distal tibia (to be reported separately). Pre-specified secondary endpoints of this study were visceral and subcutaneous abdominal fat, lean body mass, and insulin resistance measured by HOMA2-IR.
Power analyses
Expected changes in total fat mass after 6 months of ADT were derived from a meta-analysis of observational studies ranging from 3 to12 months duration (23). This study reported an average difference of 7.7% (95% CI 4.3–11.2) in total fat mass, and the s.d. at baseline ranged from 4.8 to 7.1%. Using a conservative difference of 4.3% and s.d. of 7.1% yielded 43 participants per group to provide 80% power at a level of significance of 0.05 to detect an increase in total fat mass over 6 months without treatment, assuming the mean response in the E2 group was ‘no change’. We aimed to recruit 54 participants per group to allow for 20% attrition.
Statistical analysis
Data were partly non-normally distributed and are reported as median and interquartile range (IQR). The Wilcoxon test was used to compare baseline characteristics between groups, or, in case of frequencies, chi-square test and Fisher’s exact test (with low numbers) were used. Correlations are based on Kendall’s rank correlation. The treatment effect of E2 administration vs placebo over 6 months was evaluated for body composition and cardiometabolic risk markers. The primary analysis was intention to treat including all subjects that were randomized as per their assigned group, all of whom commenced the intervention. Per-protocol analysis was additionally done as a sensitivity analysis including only participants who completed the trial and adhered to the protocol (Fig. 1). Trial outcomes are reported as the mean-adjusted difference (MAD) between the E2 group and placebo group, surrounded by a profiled 95% CI, over the course of the trial. The significance level was tested as a single P value over all time points. MADs and P values were determined with the use of repeated measures mixed-effects models including interaction of time point with treatment group (the treatment effect), treatment group, time points, strata of ADT duration (≤3 or >3 months), and baseline values of the respective variable as fixed effects, and subject as a random effect (35). When repeated measures were not available, the mixed model was substituted by a generalized linear model adjusted for strata of ADT duration (≤3 or >3 months), age and BMI. In a sensitivity analysis, using a Bayesian model with the same formula argument as the mixed model, we tested the posterior probability for a one-sided hypothesis that the estimated MAD between the groups at 6 months would be less than 0, our original hypothesis (36). For sex steroid analysis, values below the limit of detection (LOD) were assigned the value of the LOD. Because of the interrelated outcomes, no adjustments for multiple testing were made. A two-sided P -value <0.05 was considered indicative of statistical significance. Analyses were performed with the R statistical base package (version 4.0.4 for Mac) and the added packages lme4 1.1–26 and effects 4.1–4 (35, 37, 38).
Study flow diagram.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Study flow diagram.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Study flow diagram.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Results
In the study, 255 men had preliminary assessments for eligibility and 39 were randomized to each intervention (Fig. 1). The last screened participant was not randomized because the human research ethics committee ordered cessation of new clinical trial visits due to the COVID-19 pandemic. In the placebo group, one participant withdrew from the trial due to anxiety about attending study visits during the pandemic, and 38 completed follow-up at 6 months. In the E2 group, one man withdrew (failed to attend study visits) and another became unwell due to new-onset brain metastases, such that 37 men completed the follow-up (Fig. 1).
Baseline characteristics of the groups are shown in Table 1. Participants had been receiving ADT for a median of 4.4 months (IQR: 2–10, range: 1–50) prior to randomization and 10% had received a prior course. Participants had castrate testosterone concentrations at baseline and throughout the trial. Serum E2 increased in the E2 group over 6 months compared to the placebo group, MAD 207 pmol/L (95% CI: 123–292), P < 0.001 (Fig. 2 and Table 2).
Adjusted mean and 95% CI of serum estradiol (E2) concentration (pmol/L) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 207 pmol/L (95% CI: 123–292), P < 0.001. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Adjusted mean and 95% CI of serum estradiol (E2) concentration (pmol/L) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 207 pmol/L (95% CI: 123–292), P < 0.001. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Adjusted mean and 95% CI of serum estradiol (E2) concentration (pmol/L) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 207 pmol/L (95% CI: 123–292), P < 0.001. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Baseline characteristics. Data are presented as n (%), mean ± s.d., or as median (IQR).
| Placebo group (n = 39) | E2 group (n = 39) | P value | |
|---|---|---|---|
| Age (years) | 72.1 ± 6.7 | 71.7 ± 7.8 | 0.82 |
| ECOG | 0.73 | ||
| 0 | 35 (90%) | 35 (90%) | |
| 1 | 3 (8%) | 3 (8%) | |
| 2 | 1 (3%) | 1 (3%) | |
| Conceived | 36 (92%) | 36 (92%) | 1.00 |
| Educational (years) | 14.4 ± 4.8 | 14.4 ± 4.7 | 0.91 |
| Body mass index (kg/m2) | 29.2 (27.4; 32.0) | 28.2 (26.3; 32.7) | 0.79 |
| Prostate cancer stage | 0.82 | ||
| Localized | 13 (33%) | 10 (26%) | |
| Locally advanced | 6 (15%) | 10 (26%) | |
| Local recurrence | 1 (3%) | 1 (3%) | |
| M0 | 9 (23%) | 7 (18%) | |
| Metastatic CSPC | 10 (26%) | 10 (26%) | |
| Metastatic CRPC | 0 (0%) | 1 (3%) | |
| PSA | 0.17 (0.04; 0.89) | 0.24 (0.04; 2.03) | 0.49 |
| Prostatectomy | 18 (46%) | 13 (13%) | 0.36 |
| Radiotherapy (previous or concurrent | 34 (87%) | 31 (80%) | 0.54 |
| Prior course of ADT | 5 (13%) | 3 (8%) | 0.71 |
| ADT duration (months) | 4.7 (2.1; 9.9) | 4.0 (2.0; 8.8) | 0.59 |
| Planned ADT duration | 0.84 | ||
| 6 months | 5 (13%) | 3 (8%) | |
| 12 months | 0 | 1 (3%) | |
| 18 months | 8 (21%) | 6 (15%) | |
| 24 months | 8 (21%) | 14 (36%) | |
| 36 months | 2 (5%) | 1 (3%) | |
| Indefinite | 15 (38%) | 13 (33%) | |
| Undefined | 1 (3%) | 1 (3%) | |
| Smoking status | |||
| Current | 2 (5%) | 0 | 0.49 |
| Ex-smoker | 19 (49%) | 20 (51%) | 1.0 |
| Never smoked | 18 (46%) | 19 (49%) | 0.82 |
| Pack years | 6.0 (0.0; 17.5) | 5.0 (0.0; 16.5) | 0.85 |
| Alcohol (standard drinks/week) | 3 (1; 7) | 5 (1; 8) | 0.60 |
| Non-white race | 0 | 2 (5%) | 0.49 |
| Medical comorbidities | |||
| Diabetes | 7 (18%) | 5 (13%) | 0.75 |
| Hypertension | 31 (79%) | 25 (64%) | 0. 44 |
| Prior chronic glucocorticoid | 3 (8%) | 0 | 0.24 |
| Bariatric surgery | 0 | 0 | 1 |
| Chronic liver disease | 1 (3%) | 1 (3%) | 1 |
| Stroke | 0 | 1 (3%) | 1 |
| Ischaemic heart disease | 4 (10%) | 9 (23%) | 0.22 |
| Heart failure | 1 (3%) | 0 | 1 |
| Chronic kidney disease stage 3 | 5 (13%) | 3 (8%) | 0.46 |
| Chronic kidney disease stage 4 or 5 | 0 | 0 | 1 |
| Gynaecomastia | |||
| Total | 12 (31%) | 11 (28%) | 0.87 |
| CTCAE Grade 1 | 11 | 9 | |
| CTCAE Grade 2 | 1 | 2 | |
| Nipple tenderness only symptom | 0 | 2 | |
| CTCAE Grade 3 | 0 | 0 | |
| Medications, n (%) | |||
| GnRH agonist | 35 (90%) | 36 (92%) | 1 |
| GnRH antagonist | 5 (13%) | 3 (8%) | 1 |
| Anti-androgen | 2 (5%) | 2 (5%) | 1 |
| Testosterone (nmol/L) | 0.3 (0.2; 0.4) | 0.3 (0.2; 0.5) | 0.76 |
| Estradiol (pmol/L) | 83 (61; 100) | 82 (67; 102) | 0.56 |
| Activity | |||
| Weight-bearing exercise (h/week) | 0.00 (0.00; 1.00) | 0.00 (0.00; 1.25) | 0.67 |
| Daily step count | 6007 (4748; 8576) | 6226 (4255; 7352) | 0.65 |
| Non-sedentary time (% of awake) | 13.3 (9.6; 16.6) | 13.6 (9.9; 16.1) | 0.99 |
ADT, androgen deprivation therapy; CRPC, castration-resistant prostate cancer; CSPC, castration-sensitive prostate cancer; CTCAE, Common Terminology Criteria for Adverse Events version 4.03; E2, estradiol; ECOG, Eastern Cooperative Oncology Group Performance Status; GnRH, gonadotropin-releasing hormone; M0, biochemically recurrent prostate cancer without detectable metastases on conventional.
Sex steroid concentrations.
| Placebo group* (n = 39) | E2 group* (n = 39) | MAD (E2 vs placebo) (95% CI) | P value1 | |
|---|---|---|---|---|
| Estradiol (pmol/L) | <0.001 | |||
| 0 months | 83 (61; 100) | 82 (67; 102) | ||
| 1 month | 83 (72; 124) | 226 (147; 397) | 253 (169 to 336) | |
| 3 months | 86 (68; 116) | 238 (155; 342) | 217 (133 to 301) | |
| 6 months | 94 (74; 136) | 265 (188; 385) | 207 (123 to 292) | |
| Estrone (pmol/L) | <0.001 | |||
| 0 months | 72 (53; 96) | 63 (46; 87) | ||
| 1 month | 60 (51; 85) | 141 (103; 196) | 99 (70 to 128) | |
| 3 months | 67 (48; 75) | 138 (104; 190) | 113 (84 to 143) | |
| 6 months | 72 (58; 97) | 163 (104; 222) | 116 (87 to 146) | |
| Testosterone (nmol/L) | 0.25 | |||
| 0 months | 0.3 (0.2; 0.4) | 0.3 (0.2; 0.5) | ||
| 1 month | 0.3 (0.2; 0.4) | 0.3 (0.2; 0.5) | 0.0 (−0.9 to 1.0) | |
| 3 months | 0.3 (0.2; 0.4) | 0.2 (0.2; 0.3) | −0.5 (−1.5 to 0.5) | |
| 6 months | 0.3 (0.2; 0.5) | 0.3 (0.2; 0.4) | −0.8 (−1.8 to 0.1) |
*Values are presented as median (IQR); 1The significance level was tested as a single P value between groups over all time points.
Bold values indicate statistical significance (P < 0.05)
E2, estradiol; MAD, mean-adjusted difference.
Body composition by DXA
The primary endpoint of this trial was total fat mass over 6 months. E2 add-back, compared to placebo, did not significantly change total fat mass, although the CI suggests a likely increase in the E2 group, MAD 1007 g (95% CI: 124–1891), P = 0.09 (Fig. 3). E2 did not significantly change the proportional fat mass (%), MAD 0.8% (95% CI: 0.0–1.6), P = 0.15. E2 increased the android fat, MAD 164 g (95% CI: 41–286), P = 0.04. There was no significant difference between groups in gynoid fat, with the CI suggesting that gynoid fat increased in the E2 group, MAD 147 g (95% CI: 2–293), P = 0.08. Android gynoid fat ratio did not change. Changes in visceral adipose tissue (VAT) mass and subcutaneous adipose tissue (SAT) mass were also non-significant but suggestive of an increase due to E2 treatment, MAD 53 g (95% CI: 1–105) P = 0.13 and MAD 65 g (95% CI: 5–125), P = 0.11, respectively. E2 had no effect on lean mass, MAD −259 g (95% CI: −1218 to 700), P = 0.72 (Table 3).
Adjusted mean and 95% CI of fat mass (g) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 1007 g (95% CI: 124–1891), P = 0.09. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Adjusted mean and 95% CI of fat mass (g) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 1007 g (95% CI: 124–1891), P = 0.09. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Adjusted mean and 95% CI of fat mass (g) by group and study visit. Mean-adjusted difference, E2 group vs placebo group, 1007 g (95% CI: 124–1891), P = 0.09. The significance level was tested as a single P value between groups over all time points.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0663
Body composition and anthropometry.
| Placebo group* (n = 39) | E2 group* (n = 39) | MAD (E2 vs placebo) (95% CI) | P value1 | |
|---|---|---|---|---|
| Total fat mass (g) | 0.09 | |||
| 0 months | 31 612 (25 105; 34 791) | 27 606 (24 577; 36 518) | ||
| 3 months | 31 620 (27 335; 33 799) | 28 843 (25 848; 38 071) | 568 (−312 to 1446) | |
| 6 months | 31 375 (26 322; 35 597) | 30 200 (26 128; 37 812) | 1007 (124 to 1891) | |
| Proportion fat mass (%) | 0.15 | |||
| 0 months | 33.8 (31.0; 38.1) | 34.1 (30.3; 36.5) | ||
| 3 months | 35.4 (31.8; 37.7) | 34.1 (31.7; 37.0) | 0.3 (−0.5 to 1.1) | |
| 6 months | 34.7 (32.3; 38.2) | 35.5 (32.3; 37.9) | 0.8 (0.0 to 1.6) | |
| Total lean mass (g) | 0.72 | |||
| 0 months | 58 001 (53 060; 62 540) | 58 233 (53 833; 62 286) | ||
| 3 months | 57 613 (54 019; 60 918) | 57 767 (54 184; 63 721) | 137 (−818 to 1092) | |
| 6 months | 57 145 (52 907; 61 501) | 56 095 (52 779; 62 719) | −259 (−1218 to 700) | |
| Android fat mass (g) | 0.04 | |||
| 0 months | 2820 (2188; 3566) | 2543 (2248; 3614) | ||
| 3 months | 2877 (2415; 3416) | 2602 (2209; 3899) | 77 (−46 to 199) | |
| 6 months | 2892 (2412; 3557) | 2742 (2342; 3715) | 164 (41 to 286) | |
| Gynoid fat mass (g) | 0.08 | |||
| 0 months | 4597 (3812; 5210) | 4398 (3891; 5232) | ||
| 3 months | 4675 (3930; 5090) | 4686 (4044; 5673 | 142 (−3 to 187) | |
| 6 months | 4815 (4082; 5349) | 4670 (4185; 5530) | 147 (2 to 293) | |
| VAT mass (g) | 0.13 | |||
| 0 months | 874 (695; 1084) | 819 (641; 1105) | ||
| 3 months | 888 (750; 990) | 843 (670; 1083) | 17 (−31 to 65) | |
| 6 months | 908 (725; 1078) | 875 (674; 1157) | 49 (1 to 97) | |
| VAT area (cm2) | 0.13 | |||
| 0 months | 181 (144; 225) | 170 (133; 229) | ||
| 3 months | 184 (156; 205) | 175 (139; 225) | 4 (−6 to 13) | |
| 6 months | 188 (150; 224) | 181 (140; 240) | 10 (0 to 20) | |
| SAT mass (g) | 0.11 | |||
| 0 months | 1443 (1145; 1726) | 1374 (1154; 1714) | ||
| 3 months | 1500 (1232; 1676) | 1413 (1195; 1813) | 35 (−25 to 95) | |
| 6 months | 1502 (1187; 1779) | 1484 (1235; 1800) | 65 (5 to 125) | |
| SAT area (cm2) | 0.11 | |||
| 0 months | 299 (238; 358) | 285 (239; 356) | ||
| 3 months | 311 (255; 348) | 293 (248; 376) | 7 (−5 to 19) | |
| 6 months | 312 (246; 369) | 308 (256; 373) | 14 (1 to 26) | |
| Weight (kg) | 0.51 | |||
| 0 months | 89.5 (79.6; 93.9) | 85.4 (78.4; 96.2) | ||
| 1 month | 88.1 (78.5; 92.3) | 85.6 (78.5; 98.0) | 0.5 (−0.6 to 1.5) | |
| 3 months | 87.8 (79.2; 92.6) | 84.6 (78.9; 98.3) | 0.6 (−0.4 to 1.7) | |
| 6 months | 88.7 (79.1; 93.5) | 85.9 (79.5; 100.0) | 0.8 (−0.3 to 1.8) | |
| Waist circ (cm) | 0.74 | |||
| 0 months | 104 (98; 113) | 102 (95; 113) | ||
| 1 month | 104 (97; 114) | 102 (96; 112) | 0 (−1 to 1) | |
| 3 months | 104 (97; 111) | 104 (95; 113) | 1 (−1 to 2) | |
| 6 months | 102 (96; 114) | 104 (94; 112) | 0 (−1 to 1) | |
| Grip strength (kg) | 0.10 | |||
| 0 months | 36.0 (34.0; 41.5) | 38.0 (34.5; 42.0) | ||
| 1 month | 36.0 (32.5; 41.0) | 37.0 (34.0; 42.0) | ||
| 3 months | 35.5 (33.2; 39.0) | 38.0 (34.0; 43.0) | 0.3 (−1.2 to 1.8) | |
| 6 months | 34.5 (31.2; 37.5) | 36.0 (34.0; 43.0) | 1.6 (0.1 to 3.1) | |
| Daily step count | 0.66 | |||
| 0 months | 6007 (4748; 8576) | 6226 (4255; 7352) | ||
| 6 months | 5867 (4078; 6982) | 5905 (4443; 6796) | −220 (−1206 to 766) | |
| Non-sedentary time% | 0.82 | |||
| 0 months | 13.3 (9.60; 16.6) | 13.6 (9.88; 16.1) | ||
| 6 months | 12.9 (9.52; 15.2) | 12.7 (10.4; 14.8) | −0.2 (−1.8 to 1.5) |
*Values are presented as median (IQR); 1The significance level was tested as a single P value between groups over all time points.
Bold values indicates a value P< 0.05
Circ, circumference; MAD, mean-adjusted difference; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.
In the cohort, overall there was no significant correlation between ADT duration prior to randomization and baseline total fat mass, tau = 0.13, P = 0.09. Randomization was stratified by ADT duration (≤3 vs >3 months) and our statistical model included these strata of ADT duration. In our study, ADT duration prior to study did not significantly modify the effects of E2 on fat mass.
Anthropometry
E2 had no effect on body weight, waist circumference, or grip strength (Table 3).
Metabolic markers
There were no significant differences between groups in SHBG, with the CI suggesting that SHBG increased in the E2 group, MAD 5.5 nmol/L (95% CI: 0.3–10.6), P = 0.10. IGF-1, MAD −2.2 nmol/L (95% CI: −4.6 to −0.2), P = 0.17, lipid parameters, fasting glucose, HbA1c, and HOMA2-IR were not different between groups (Table 4).
Metabolic markers.
| Placebo group* (n = 39) | E2 group* (n = 39) | MAD (E2 vs placebo) (95% CI) | P value1 | |
|---|---|---|---|---|
| FBGL (mmol/L) | 0.56 | |||
| 0 months | 5.6 (5.2; 6.2) | 5.4 (5.2; 6.3) | ||
| 3 months | 5.6 (5.1; 6.2) | 5.4 (4.7; 6.2) | −0.1 (−0.6 to 0.3) | |
| 6 months | 5.6 (5.1; 6.3) | 5.2 (4.9; 5.9) | 0.1 (−0.3 to 0.6) | |
| HbA1c (mmol/mol) | 0.76 | |||
| 0 months | 36.0 (35.0; 41.5) | 38.0 (34.0; 42.6) | ||
| 3 months | 37.0 (35.0; 39.0) | 37.0 (34.0; 40.3) | −0.7 (−2.7 to 1.4) | |
| 6 months | 37.5 (35.0; 40.8) | 38.0 (33.0; 40.0) | 0.0 (−2.0 to 2.1) | |
| HOMA2-IR | 0.83 | |||
| 0 months | 1.6 (1.1; 2.5) | 1.6 (1.2; 2.4) | ||
| 3 months | 1.7 (1.1; 2.4) | 1.4 (0.9; 2.3) | −0.2 (−0.9 to 0.5) | |
| 6 months | 1.6 (1.0; 2.4) | 1.4 (1.0; 2.3) | 0.1 (−0.7 to 0.8) | |
| TC (mmol/L) | 0.33 | |||
| 0 months | 5.0 (4.3; 5.7) | 4.9 (4.1; 5.8) | ||
| 3 months | 5.2 (4.1; 5.8) | 4.6 (4.0; 5.3) | −0.2 (−0.5 to 0.1) | |
| 6 months | 5.0 (4.3; 5.7) | 4.4 (3.8; 5.3) | 0.2 (−0.5 to 0.1) | |
| LDL-C (mmol/L) | 0.55 | |||
| 0 months | 2.8 (1.9; 3.3) | 2.5 (1.9; 3.2) | ||
| 3 months | 2.9 (2.0; 3.5) | 2.6 (1.9; 3.1) | −0.13 (−0.40 to 0.14) | |
| 6 months | 2.6 (2.3; 3.3) | 2.4 (1.7; 3.0) | −0.13 (−0.39 to 0.14) | |
| HDL-C (mmol/L) | 0.52 | |||
| 0 months | 1.5 (1.2; 1.8) | 1.5 (1.2; 1.8 | ||
| 3 months | 1.5 (1.2; 1.7) | 1.5 (1.3; 1.7 | −0.00 (−0.09 to 0.09) | |
| 6 months | 1.4 (1.2; 1.6) | 1.4 (1.2; 1.8 | 0.04 (−0.04 to 0.13) | |
| Triglycerides (mmol/L) | 0.26 | |||
| 0 months | 1.4 (1.0; 2.1) | 1.4 (1.2; 1.8) | ||
| 3 months | 1.4 (0.9; 2.0) | 1.5 (1.1; 1.8) | −0.08 (−0.32 to 0.16) | |
| 6 months | 1.6 (1.1; 2.1) | 1.3 (1.0; 1.6) | −0.20 (−0.44 to 0.04) | |
| IGF-1 (nmol/L) | 0.17 | |||
| 0 months | 23.1 (17.5; 25.8) | 19.9 (17.0; 26.3) | ||
| 3 months | 23.0 (18.2; 28.0) | 19.2 (15.4; 24.5) | −1.6 (−4.0 to 0.7) | |
| 6 months | 22.8 (18.0; 26.3) | 17.9 (14.4; 21.7) | −2.2 (−4.6 to 0.2) | |
| SHBG (nmol/L) | 0.10 | |||
| 0 months | 53.0 (35.5; 71.0) | 56.0 (44.5; 75.0) | ||
| 3 months | 50.5 (34.8; 68.0) | 60.0 (39.5; 70.0) | 1.2 (−4.0 to 6.4) | |
| 6 months | 45.5 (34.0; 61.8) | 54.0 (43.0; 69.0) | 5.5 (0.3 to 10.6) |
*Values presented as median (IQR); 1The significance level was tested as a single P value between groups over all time points.
E2, estradiol; FBGL, fasting blood glucose level; HOMA2-IR, homeostatic model assessment of insulin resistance 2; IGF-1, insulin-like growth factor 1; MAD, mean adjusted difference; SHBG, sex hormone-binding globulin; TC, total cholesterol.
Sensitivity analyses
In a sensitivity per-protocol analysis excluding men with protocol violations (Fig. 1), E2 add-back, compared to placebo, did not significantly increase total fat mass, MAD 1054 g (95% CI: 28–2080), P = 0.14. The SHBG increase remained non-significant, MAD 6.9 nmol/L (95% CI: 1.0–12.9), P = 0.07. Other body composition, anthropometric, and metabolic marker outcomes were similar (data not shown). In another sensitivity analysis, the Bayesian posterior probability for our original hypothesis, namely a negative difference between the treatment group and placebo group at 6 months, was very low (<1%). Investigating the mechanisms of E2 action, in E2-treated patients, IGF-1 significantly declined over time, by −2.4 nmol/L (95% CI: −4.0 to −0.8), P = 0.02.
Adverse events
Adverse events are shown in Table 5. There were no serious adverse events detected other than prostate cancer progression which occurred in two participants in the placebo arm and one participant in the E2 arm. There were no arterial or venous thrombosis events. Two participants in the E2 arm ceased the intervention because of adverse effects on the breast. Incident CTCAE grade 1 or grade two gynaecomastia was twice as common in the E2 arm (44 vs 21%) and nipple tenderness was reported in 11 (28%) of participants in the E2 arm and 1 (3%) participant in the placebo group. Exacerbation of hypertension was more common in the placebo arm, occurring in seven (18%) participants and no participants in the E2 arm.
Incident adverse events. Data are presented as n (%).
| Adverse event | Placebo group (n = 39) | E2 group (n = 39) | P value1 |
|---|---|---|---|
| Overall | |||
| Serious adverse event2 | 2 (5%) | 1 (3%) | 1 |
| Drug cessation due to adverse event | 0 | 2 (5%) | 0.49 |
| Prostate cancer progression3 | 2 (5%) | 1 (3%) | 1 |
| Fracture | 0 | 0 | 1 |
| VTE | 0 | 0 | 1 |
| Acute coronary syndrome | 0 | 0 | 1 |
| Stroke/TIA | 0 | 0 | 1 |
| Gynaecomastia | |||
| Grade 1 | 7 (18%) | 3 (8%) | 0.31 |
| Grade 2 | 1 (3%) | 14 (36%) | 0.31 |
| Grade 3 | 0 | 0 | 1 |
| Total | 8 (21%) | 17 (44%) | 0.03 |
| Nipple tenderness | 1 (3%) | 11 (28%) | 0.003 |
| Rash | |||
| Grade 1 | 8 (21%) | 4 (10%) | 0.35 |
| Grade 2+ | 0 | 0 | 1 |
| Rash at gel site | 0 | 0 | 1 |
| New non-prostate malignancy | 0 | 0 | 1 |
| Hypertension exacerbation4 | 7 (18%) | 0 | 0.01 |
| Headache | |||
| Grade 1 | 4 (10%) | 3 (8%) | 1 |
| Grade 2+ | 0 | 0 | 1 |
1The significance level was tested as a single P value between groups over all time points. 2Serious adverse events are defined as events of Common Terminology Criteria for Adverse Events v4.03 Grade 3 or above. Incident serious adverse events were only those of prostate cancer progression. 3Prostate cancer progression is defined as incident biochemical failure, resistance to castration, or new metastases. Both events in the placebo group occurred in a single participant who had biochemical failure at month 1 and development of new metastases at month 6. 4Hypertension exacerbation is defined as: development of prehypertension when previously normal; symptomatic increase by >20 mmHg diastolic OR increase to >140/90 if previously normal; hypertension requiring more intensive therapy than previously used; life-threatening hypertension; or death due to hypertension.
E2, estradiol; TIA, transient ischaemic attack; VTE, venous thromboembolism.
Discussion
In this randomized controlled trial (RCT), we aimed to assess the effect of E2 add-back on fat mass in men rendered castrate by ADT for prostate cancer. Contrary to our hypothesis, E2 did not reduce fat mass over 6 months. On the contrary, while not statistically significant, we observed a consistent trend in the opposite direction on several measures of body fat. Men assigned to E2, compared to placebo, demonstrated trends toward increased total fat mass with a pattern of concordant trends in regional android fat, gynoid fat, VAT, and SAT mass which all had positive 95% CIs at 6 months and overall P values around 0.1.
We are unaware of any previous studies in men designed to measure the effects of E2 on fat mass in the absence of testosterone. The only previous study of body composition change during E2 add-back in men receiving ADT was of 8 weeks duration, included 18 participants and no control group, and reported no difference in body composition from baseline to study end (39).
We hypothesized that E2 add-back would prevent fat gain in men undergoing ADT because multiple lines of evidence implicate a role for E2 in male fat regulation. For example, the male global ER-alpha knockout mouse which has high serum testosterone and E2 (14) displays age-dependent accumulation of white adipose tissue, due to an increase in adipocyte size and number (13). Men with rare inactivating germline mutations of the ER-alpha (8) or aromatase (9) genes are similarly characterized by excess adiposity and insulin resistance. Although, these congenital lesions cannot differentiate developmental from homeostatic effects of E2, they are supported by medical castration experiments in healthy adults. In such a study by Finkelstein and colleagues among healthy young men, the effects of E2 on fat were inferred by treatment with ADT for 16 weeks, and randomization to graded testosterone add-back, or placebo, with or without aromatase inhibitor to distinguish effects of testosterone that are and are not aromatization dependent, respectively, although aromatase inhibitors only reduce E2 by about 90% (10) and allocation to the aromatase inhibitor cohort was not randomized. In that study, E2 deficiency was inferred to be the primary mediator of increases in subcutaneous, intraabdominal, and total body fat (%). Other experiments using temporary ADT with testosterone add-back and aromatase inhibition also determined an E2-dependent effect to decrease fat mass (11, 40).
In this trial, we refute the hypothesis of a fat-decreasing effect of E2 add-back in men on ADT. E2 also did not lower HOMA2-IR as hypothesized, and it did not have the expected effect to lower HDL-C. Unexpectedly, E2 add-back fell short of significantly lowering IGF-1 and raising SHBG, over placebo, although the 95% CIs at 6 months were suggestive of such effects. One reason for failing to confirm our primary hypothesis could have been insufficient target organ estrogenization. However, E2 did produce expected (adverse) effects at the breast, and significantly reduced bone resorption as determined by measuring serum beta carboxyl-terminal type 1 collagen telopeptide (bone data to be reported separately). Premature closure of the study to the COVID-19 pandemic is another possible explanation for these findings.
We selected an E2 dose that was previously shown in men receiving ADT to produce minimum serum E2 concentrations within the reference range for healthy older men (25). In that study, E2 (combined 0.9 and 1.8 mg doses) did increase SHBG. In this current trial, trough E2 concentrations in the E2 group at 6 months (median 265 pmol/L, IQR 188–385) were similar to those in our previous study (25). Smoothed age-specific reference ranges for Australian men, using the same E2 assay, have been reported as 37–195 pmol/L (men <65 years), 22–165 pmol/L (75–85 years), and 22–173 pmol/L (>85 years) (41). Using another LCMS/MS assay, the reference range for Danish men < 60 years was approximately 51–151 pmol/L (42). By comparison, the E2 concentrations recommended for transgender females, reflecting the premenopausal female range, are 367–734 pmol/L (43). We have therefore achieved serum E2 concentrations at or above those seen in healthy eugonadal older men. However, given that physiologically, 50–75% of E2 in men is derived from extragonadal aromatization of testosterone, serum E2 concentrations in eugonadal men reflect diffusion down a concentration gradient of E2 that has been produced and escaped local metabolism in peripheral tissues (44). The intervention in this study may therefore represent a sub-physiological E2 replacement, particularly in tissues with high aromatase activity such as fat.
One important difference of our current study compared to all previous studies examining the effects of E2 on fat mass in men is that by virtue of our experimental design, we examined the effects of E2 in the absence of testosterone. Given the consistently positive CIs surrounding our measures of fat mass at 6 months, the potentially important observation of fat gain in the E2 arm over placebo should encourage larger or longer studies to confirm or refute a fat-promoting effect of E2 in the absence of testosterone. Gender-affirming hormone therapy provides another model in which the effects of estrogens, usually E2, can be observed in the context of very low testosterone concentrations, in people with male natal sex. Over 12–24 months, transfeminine hormone therapy, typically with E2 in combination with cyproterone acetate or spironolactone, targeting E2 and testosterone concentrations similar to cisgender premenopausal females (43), is associated with an average increase in total body fat mass and proportion, a reduction in lean body mass, and an increase in gynoid and android fat proportion (45, 46, 47). However, studies in transwomen to date are limited by their observational, cross-sectional, and largely uncontrolled nature (48).
The controlled data on the effects of E2 in men generated to date are derived from paradigms in which testosterone is present. Testosterone has lipolytic and anti-adipogenic effects including via increasing lipolytic myokine production (49) and, in rodents, by diversion of pluripotent stem cells away from adipogenesis and toward myogenesis (50, 51), effects which might offset or prevent gains in fat mass due to E2. In Finkelstein’s study, discussed above, E2 deficiency was inferred to be the primary mediator of increases in body fat proportion. When E2 is present, testosterone add-back to men rendered hypogonadal, reduces fat gain in a dose-dependent fashion (3, 10). It is possible that E2 has a permissive role, perhaps with a low-dose threshold, to facilitate primary actions of testosterone to reduce fat mass. This notion is consistent with the findings of a pilot study by Colleluori et al., reporting that among obese men subjected to weight loss and randomized to the aromatase inhibitor anastrozole, or placebo, for 6 months, the former group obtained a higher fat mass loss (52).
Another potential mechanism by which the direction of the effect of E2 on fat mass in men could depend on the presence of testosterone, is via interactions of these hormones with the growth hormone (GH) – IGF-1 axis (53). E2 stimulates GH in men (54) while non-aromatisable androgens do not (55). It is unclear whether intrapituitary aromatization is required or important (54, 56). E2 reduces the hepatic IGF-1 response to GH such that the net effect of E2 on IGF-1 depends on relative exposure of the hypothalamus/pituitary and liver (57). In this study, we did observe a significant decline over time in IGF-1 in E2-treated participants. Androgens potentiate many of the actions of GH including IGF-1 production and fat oxidation (53). It has been suggested that the fat gain observed in men on testosterone replacement in the aromatase inhibitor cohort of Finkelstein’s and other similar studies could have been mediated by suppression of GH (58). It could be that in the context of profound testosterone deficiency studied here, transdermal E2 add-back is insufficient to rescue deficient GH production yet does exert a peripheral effect in suppressing GH-induced lipolysis.
Strengths and limitations of the study
This was a rigorously conducted trial of E2 add-back for men undergoing ADT for prostate cancer, an intervention never before subject to a randomized controlled trial for the purpose of evaluating body composition. The trial was stopped early due to the COVID-19 pandemic with 75 men completing the study. This fell short of our recruitment target which estimated that 86 completers were needed to achieve adequate power. Despite this, and the lack of significant estrogen-dependent effects on biochemical markers, serum E2 concentrations were at or above those in eugonadal men, and we observed consistent trends suggesting that such E2 dosing in the absence of testosterone may increase rather than decrease total and regional fat mass. Of note in a sensitivity analysis, the Bayesian posterior probability for our original hypothesis, namely a negative difference between the treatment group and placebo group at 6 months, was very low (<1%). While we speculate on possible mechanisms by which effects of E2 on fat mass may depend on the presence of testosterone, our study was not designed to confirm or refute these mechanisms. For example, IGF-1 was not a pre-specified endpoint and although we offer some speculation on interactions of E2 and testosterone via the somatotrophic axis, our trial was not designed to investigate interactions of E2 and testosterone via the somatotrophic axis. Moreover, while it is possible that E2, at a lower concentration may have a permissive role in the actions of testosterone to reduce fat mass, our study was not designed to explore potential threshold effects. Moreover, given the intracrine metabolism of sex steroids, provision of exogenous E2 treatment may not necessarily restore physiologic concentrations of E2 in target tissues. Finally, our study enrolled older men with prostate cancer, and it is possible that the effects of E2, in the absence of testosterone, on fat mass may be different in other populations, such as, for example, young healthy men. Of note, men receiving GnRH analog treatment for prostate cancer represent a group of men for whom the ethical requirement to correct androgen deficiency is absent and for whom, based on a hypothesis of net benefit, we provided E2 add-back for 6 months. While a 6-month study of experimental castration in healthy men with graded E2 add-back in the absence of testosterone would be ideal, ethical constraints may make such a study difficult to conduct. Future studies should also be designed to explore the question whether E2 has depot-specific effects of fat mass in men.
Participants in this trial had been receiving ADT for a median of 4.4 months (range 1–50 months) prior to randomization and 10% had received a prior course. It is possible that ADT-induced increases in body fat occuring prior to randomization obscured an effect of E2 to prevent fat gain and that such an effect might have been detected if all men were randomized very soon after ADT commencement. However, a significant correlation between ADT duration prior to randomization and baseline total fat mass could not be confirmed in the actual data. In the RCT, we adjusted for ADT duration, including it as a covariable.
Participants in this trial were a relatively homogenous cohort of predominantly Caucasian men (consistent with the local population) with good functional status, which might limit generalizability to other ethnic groups, and to sicker men. While we measured activity levels for 1 week at baseline and study end, these may not have reflected average activity during the 6-month trial.
Significance and conclusions
The unique experimental design of this study allowed us to examine the effects of E2 on male fat mass in the absence of testosterone. We refute our hypothesis of a direct E2 effect to reduce fat mass, but provide suggestive evidence that, contrary to our hypothesis, E2, without testosterone, may increase, rather than decrease total and regional fat mass over 6 months. While preliminary, these findings provide insights into biological actions of E2 in men, which may be more complex than traditionally assumed. While work by Finkelstein et al., and others, discussed above, has highlighted the importance of E2 in determining fat distribution and accumulation in combination with testosterone replacement, this RCT suggests E2 might be permissive for actions of androgens in preventing fat accumulation but that E2 itself is not solely responsible for these effects. This insight, if confirmed would be relevant for designing optimal hormone replacement regimens for hypogonadal men. Moreover, E2, albeit in higher doses, is currently under investigation as a sole mode of ADT (59). Given that E2 as a sole mode of ADT reduces testosterone to castrate concentrations, a better understanding of the effects of E2 on fat mass and cardiometabolic risk factors has clinical relevance, particularly since cardiovascular events are a common cause of death in men with early prostate cancer.
Declaration of interest
M G has received research funding from Bayer Pharma, Novartis, Weight Watchers, and Lilly and speaker’s honoraria from Besins Healthcare. A S C has received funding from Besins Healthcare for investigator-initiated studies utilizing estradiol and progesterone. D J H has received institutional funding for investigator-initiated testosterone pharmacology research from Besins Healthcare and Lawley and has provided expert testimony in anti-doping and professional standards tribunals and testosterone litigation. N R, R H, and J D Z declare that they have no conflict of interest. Mathis Grossmann is on the editorial board of EJE. Mathis Grossmann was not involved in the review or editorial process for this paper, on which they are listed as an author.
Funding
This study was funded by a project grant from the National Health and Medical Research Council (NHMRC) (#APP1099173). A S C is supported by a NHMRC Early Career Fellowship (#1143333). N R is supported by an Australian Postgraduate Award.
Clinical trial registration
The study was preregistered with the Australian New Zealand Clinical Trials Registry (identifier 12614000689673).
References
- 1↑
Wang CC, Swerdloff RSR, Iranmanesh AA, Dobs AA, Snyder PJP, Cunningham GG, Matsumoto AMA, Weber TT, Berman NN & Testosterone Gel Study Group. Transdermal testosterone gel improves sexual function, mood, muscle strength, and body composition parameters in hypogonadal men. Journal of Clinical Endocrinology and Metabolism 2000 85 2839–2853. (https://doi.org/10.1210/jcem.85.8.6747)
- 2↑
Wang C, Cunningham G, Dobs A, Iranmanesh A, Matsumoto AM, Snyder PJ, Weber T, Berman N, Hull L, Swerdloff RS. Long-term testosterone gel (AndroGel) treatment maintains beneficial effects on sexual function and mood, lean and fat mass, and bone mineral density in hypogonadal men. Journal of Clinical Endocrinology and Metabolism 2004 89 2085–2098. (https://doi.org/10.1210/jc.2003-032006)
- 3↑
Finkelstein JS, Lee H, Burnett-Bowie SM, Darakananda K, Gentile EC, Goldstein DW, Prizand SH, Krivicich LM, Taylor AP & Wulczyn KE et al. Dose-response relationships between gonadal steroids and bone, body composition, and sexual function in aging men. Journal of Clinical Endocrinology and Metabolism 2020 105 2779–2788. (https://doi.org/10.1210/clinem/dgaa318)
- 4↑
Page ST, Amory JK, Bowman FD, Anawalt BD, Matsumoto AM, Bremner WJ, Tenover JL. Exogenous testosterone (T) alone or with finasteride increases physical performance, grip strength, and lean body mass in older men with low serum T. Journal of Clinical Endocrinology and Metabolism 2005 90 1502–1510. (https://doi.org/10.1210/jc.2004-1933)
- 5↑
Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Lenrow DA, Holmes JH, Dlewati A, Santanna J & Rosen CJ et al.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. (https://doi.org/10.1210/jcem.84.8.5885)
- 6↑
Burnett‐Bowie SA, Roupenian KC, Dere ME, Lee H, Leder BZ. Effects of aromatase inhibition in hypogonadal older men: a randomized, double‐blind, placebo‐controlled trial. Clinical Endocrinology 2009 70 116–123. (https://doi.org/10.1111/j.1365-2265.2008.03327.x)
- 7↑
Russell N, Grossmann M. MECHANISMS IN ENDOCRINOLOGY: Estradiol as a male hormone. European Journal of Endocrinology 2019 181 R23–R43. (https://doi.org/10.1530/EJE-18-1000)
- 8↑
Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B, Williams TC, Lubahn DB, Korach KS. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. New England Journal of Medicine 1994 331 1056–1061. (https://doi.org/10.1056/NEJM199410203311604)
- 9↑
Jones MEE, Boon WC, McInnes K, Maffei L, Carani C, Simpson ER. Recognizing rare disorders: aromatase deficiency. Nature Clinical Practice: Endocrinology and Metabolism 2007 3 414–421. (https://doi.org/10.1038/ncpendmet0477)
- 10↑
Finkelstein JS, Lee H, Burnett-Bowie SAM, Pallais JC, Yu EW, Borges LF, Jones BF, Barry CV, Wulczyn KE & Thomas BJ et al. Gonadal steroids and body composition, strength, and sexual function in men. New England Journal of Medicine 2013 369 1011–1022. (https://doi.org/10.1056/NEJMoa1206168)
- 11↑
Chao J, Rubinow KB, Kratz M, Amory JK, Matsumoto AM, Page ST. Short-term estrogen withdrawal increases adiposity in healthy men. Journal of Clinical Endocrinology and Metabolism 2016 101 3724–3731. (https://doi.org/10.1210/jc.2016-1482)
- 12↑
Tchernof A, Calles-Escandon J, Sites CK, Poehlman ET. Menopause, central body fatness, and insulin resistance: effects of hormone-replacement therapy. Coronary Artery Disease 1998 9 503–511. (https://doi.org/10.1097/00019501-199809080-00006)
- 13↑
Heine PA, Taylor JA, Iwamoto GA, Lubahn DB, Cooke PS. Increased adipose tissue in male and female estrogen receptor-α knockout mice. PNAS 2000 97 12729–12734. (https://doi.org/10.1073/pnas.97.23.12729)
- 14↑
Callewaert F, Venken K, Ophoff J, Gendt KD, Torcasio A, Lenthe GH, Oosterwyck HV, Boonen S, Bouillon R & Verhoeven G et al. Differential regulation of bone and body composition in male mice with combined inactivation of androgen and estrogen receptor‐α. FASEB Journal 2009 23 232–240.
- 15↑
Jones ME, Thorburn AW, Britt KL, Hewitt KN, Wreford NG, Proietto J, Oz OK, Leury BJ, Robertson KM & Yao S et al. Aromatase-deficient (ArKO) mice have a phenotype of increased adiposity. PNAS 2000 97 12735–12740. (https://doi.org/10.1073/pnas.97.23.12735)
- 16↑
Cooke PS, Nanjappa MK, Ko C, Prins GS, Hess RA. Estrogens in male physiology. Physiological Reviews 2017 97 995–1043. (https://doi.org/10.1152/physrev.00018.2016)
- 17↑
Movérare‐Skrtic S, Venken K, Andersson N, Lindberg MK, Svensson J, Swanson C, Vanderschueren D, Oscarsson J, Gustafsson JA, Ohlsson C. Dihydrotestosterone treatment results in obesity and altered lipid metabolism in orchidectomized mice. Obesity 2006 14 662–672. (https://doi.org/10.1038/oby.2006.75)
- 18↑
Kim NR, David K, Corbeels K, Khalil R, Antonio L, Schollaert D, Deboel L, Ohlsson C, Gustafsson JÅ & Vangoitsenhoven R et al. Testosterone reduces body fat in male mice by stimulation of physical activity via extrahypothalamic ERα signaling. Endocrinology 2021 162 bqab045. (https://doi.org/10.1210/endocr/bqab045).
- 19↑
Sebo ZL, Rodeheffer MS. Testosterone metabolites differentially regulate obesogenesis and fat distribution. Molecular Metabolism 2020 44 1–9.
- 20↑
Grossmann M, Hamilton EJ, Gilfillan C, Bolton D, Joon DL, Zajac JD. Bone and metabolic health in patients with non-metastatic prostate cancer who are receiving androgen deprivation therapy. Medical Journal of Australia 2011 194 301–306. (https://doi.org/10.5694/j.1326-5377.2011.tb02979.x)
- 21↑
Mottet N, Bellmunt J, Bolla M, Briers E, Cumberbatch MG, Santis MD, Fossati N, Gross T, Henry AM & Joniau S et al.EAU-ESTRO-SIOG guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent. European Urology 2017 71 618–629. (https://doi.org/10.1016/j.eururo.2016.08.003)
- 22↑
Cornford P, Bellmunt J, Bolla M, Briers E, Santis MD, Gross T, Henry AM, Joniau S, Lam TB & Mason MD et al.EAU-ESTRO-SIOG guidelines on prostate cancer. Part II: treatment of relapsing, metastatic, and castration-resistant prostate cancer. European Urology 2017 71 630–642. (https://doi.org/10.1016/j.eururo.2016.08.002)
- 23↑
Haseen F, Murray LJ, Cardwell CR, O’Sullivan JM, Cantwell MM. The effect of androgen deprivation therapy on body composition in men with prostate cancer: systematic review and meta-analysis. Journal of Cancer Survivorship: Research and Practice 2010 4 128–139. (https://doi.org/10.1007/s11764-009-0114-1)
- 24↑
Russell N, Hoermann R, Cheung AS, Zajac JD, Handelsman DJ, Grossmann M. Supplementary Table_CONSORT checklist_Estradiol and fat mass.docx. Figshare, 2021. (https://doi.org/10.6084/m9.figshare.14700837.v1).
- 25↑
Russell N, Hoermann R, Cheung AS, Ching M, Zajac JD, Handelsman DJ, Grossmann M. Short-term effects of transdermal estradiol in men undergoing androgen deprivation therapy for prostate cancer: a randomized placebo-controlled trial. European Journal of Endocrinology 2018 178 565–576. (https://doi.org/10.1530/EJE-17-1072)
- 26↑
NIH National Cancer Institute. Common terminology criteria for adverse events version 4.03, 2010. (available at: https://nciterms.nci.nih.gov/ncitbrowser/pages/advanced_search.jsf?dictionary=CTCAE&version=4.03). Accessed on 4 July 2017.
- 27↑
Micklesfield LK, Goedecke JH, Punyanitya M, Wilson KE, Kelly TL. Dual‐energy X‐ray performs as well as clinical computed tomography for the measurement of visceral fat. Obesity 2012 20 1109–1114. (https://doi.org/10.1038/oby.2011.367)
- 28↑
Stults-Kolehmainen MA, Stanforth PR, Bartholomew JB, Lu T, Abolt CJ, Sinha R. DXA estimates of fat in abdominal, trunk and hip regions varies by ethnicity in men. Nutrition and Diabetes 2013 3 e64–e64. (https://doi.org/10.1038/nutd.2013.5)
- 29↑
Dias KA, Ramos JS, Wallen MP, Davies PSW, Cain PA, Leong GM, Ingul CB, Coombes JS, Keating SE. Accuracy of longitudinal assessment of visceral adipose tissue by dual-energy X-ray absorptiometry in children with obesity. Journal of Obesity 2019 2019 2193723. (https://doi.org/10.1155/2019/2193723)
- 30↑
Cheung YM, Roff G, Grossmann M. Precision of the hologic horizon A dual energy X-ray absorptiometry in the assessment of body composition. Obesity Research and Clinical Practice 2020 14 514–518. (https://doi.org/10.1016/j.orcp.2020.10.005)
- 31↑
World Health Organization. Waist Circumference and Waist-Hip Ratio: Report of a WHO Expert Consultation, Geneva, 8–11 December 2008. Geneva: World Health Organization, 2011.
- 32↑
Roberts HC, Denison HJ, Martin HJ, Patel HP, Syddall H, Cooper C, Sayer AA. A review of the measurement of grip strength in clinical and epidemiological studies: towards a standardised approach. Age and Ageing 2011 40 423–429. (https://doi.org/10.1093/ageing/afr051)
- 33↑
Harwood DT, Handelsman DJ. Development and validation of a sensitive liquid chromatography–tandem mass spectrometry assay to simultaneously measure androgens and estrogens in serum without derivatization. Clinica Chimica Acta: International Journal of Clinical Chemistry 2009 409 78–84. (https://doi.org/10.1016/j.cca.2009.09.003)
- 34↑
Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care 2004 27 1487–1495. (https://doi.org/10.2337/diacare.27.6.1487)
- 35↑
Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 2015 67 1–48.
- 36↑
Bürkner PC Advanced Bayesian multilevel modeling with the R package brms. R Journal 2018 10 395–411. (https://doi.org/10.32614/RJ-2018-017)
- 37↑
R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2021. (available at: https://www.R-project.org/)
- 38↑
Fox J, Weisberg S. Visualizing fit and lack of fit in complex regression models with predictor effect plots and partial residuals. Journal of Statistical Software 2018 87 1–27. (https://doi.org/10.18637/jss.v087.i09)
- 39↑
Purnell JQ, Bland LB, Garzotto M, Lemmon D, Wersinger EM, Ryan CW, Brunzell JD, Beer TM. Effects of transdermal estrogen on levels of lipids, lipase activity, and inflammatory markers in men with prostate cancer. Journal of Lipid Research 2006 47 349–355. (https://doi.org/10.1194/jlr.M500276-JLR200)
- 40↑
Juang PS, Peng S, Allehmazedeh K, Shah A, Coviello AD, Herbst KL. Testosterone with dutasteride, but not anastrazole, improves insulin sensitivity in young obese men: a randomized controlled trial. Journal of Sexual Medicine 2014 11 563–573. (https://doi.org/10.1111/jsm.12368)
- 41↑
Handelsman DJ, Yeap B, Flicker L, Martin S, Wittert GA, Ly LP. Age-specific population centiles for androgen status in men. European Journal of Endocrinology 2015 173 809–817. (https://doi.org/10.1530/EJE-15-0380)
- 42↑
Frederiksen H, Johannsen TH, Andersen SE, Albrethsen J, Landersoe SK, Petersen JH, Andersen AN, Vestergaard ET, Schorring ME & Linneberg A et al. Sex-specific estrogen levels and reference intervals from infancy to late adulthood determined by LC-MS/MS. Journal of Clinical Endocrinology and Metabolism 2020 105 754–768. (https://doi.org/10.1210/clinem/dgz196)
- 43↑
Hembree WC, Cohen-Kettenis PT, Gooren L, Hannema SE, Meyer WJ, Murad MH, Rosenthal SM, Safer JD, Tangpricha V, T’Sjoen GG. Endocrine treatment of gender-dysphoric/gender-incongruent persons: an Endocrine Society* clinical practice guideline. Journal of Clinical Endocrinology and Metabolism 2017 102 3869–3903. (https://doi.org/10.1210/jc.2017-01658)
- 44↑
Russell N, Cheung A, Grossmann M. Estradiol for the mitigation of adverse effects of androgen deprivation therapy. Endocrine-Related Cancer 2017 24 R297–R313. (https://doi.org/10.1530/ERC-17-0153)
- 45↑
Klaver M, Dekker MJHJ, Mutsert de R, Twisk JWR, Heijer den M. Cross‐sex hormone therapy in transgender persons affects total body weight, body fat and lean body mass: a meta‐analysis. Andrologia 2017 49 e12660. (https://doi.org/10.1111/and.12660)
- 46↑
Klaver M, Blok de CJM, Wiepjes CM, Nota NM, Dekker MJHJ, Mutsert de R, Schreiner T, Fisher AD, T’Sjoen G, Heijer den M. Changes in regional body fat, lean body mass and body shape in trans persons using cross-sex hormonal therapy: results from a multicenter prospective study. European Journal of Endocrinology 2018 178 163–171. (https://doi.org/10.1530/EJE-17-0496)
- 47↑
Velzen van DM, Nota NM, Simsek S, Conemans EB, T’Sjoen G, Heijer den M. Variation in sensitivity and rate of change in body composition: steps toward individualizing transgender care. European Journal of Endocrinology 2020 183 529–536. (https://doi.org/10.1530/EJE-20-0609)
- 48↑
Spanos C, Bretherton I, Zajac JD, Cheung AS. Effects of gender-affirming hormone therapy on insulin resistance and body composition in transgender individuals: a systematic review. World Journal of Diabetes 2020 11 66–77. (https://doi.org/10.4239/wjd.v11.i3.66)
- 49↑
Basaria S, Bhasin S. Targeting the skeletal muscle-metabolism axis in prostate-cancer therapy. New England Journal of Medicine 2012 367 965–967. (https://doi.org/10.1056/NEJMcibr1203160)
- 50↑
Singh R, Artaza JN, Taylor WE, Gonzalez-Cadavid NF, Bhasin S. Androgens stimulate myogenic differentiation and inhibit adipogenesis in C3H 10T1/2 pluripotent cells through an androgen receptor-mediated pathway. Endocrinology 2003 144 5081–5088. (https://doi.org/10.1210/en.2003-0741)
- 51↑
Singh R, Artaza JN, Taylor WE, Braga M, Yuan X, Gonzalez-Cadavid NF, Bhasin S. Testosterone inhibits adipogenic differentiation in 3T3-L1 cells: nuclear translocation of androgen receptor complex with β-catenin and T-cell factor 4 may bypass canonical Wnt signaling to down-regulate adipogenic transcription factors. Endocrinology 2006 147 141–154. (https://doi.org/10.1210/en.2004-1649)
- 52↑
Colleluori G, Chen R, Turin CG, Vigevano F, Qualls C, Johnson B, Mediwala S, Villareal DT, Armamento-Villareal R. Aromatase inhibitors plus weight loss improves the hormonal profile of obese hypogonadal men Without causing major side effects. Frontiers in Endocrinology 2020 11 277. (https://doi.org/10.3389/fendo.2020.00277)
- 53↑
Birzniece V, Ho KKY. Sex steroids and the GH axis: implications for the management of hypopituitarism. Best Practice and Research: Clinical Endocrinology and Metabolism 2017 31 59–69. (https://doi.org/10.1016/j.beem.2017.03.003)
- 54↑
Roelfsema F, Yang RJ, Takahashi PY, Erickson D, Bowers CY, Veldhuis JD. Aromatized estrogens amplify nocturnal growth hormone secretion in testosterone-replaced older hypogonadal men. Journal of Clinical Endocrinology and Metabolism 2018 103 4419–4427. (https://doi.org/10.1210/jc.2018-00755)
- 55↑
Keenan BS, Richards GE, Ponder SW, Dallas JS, Nagamani M, Smith ER. Androgen-stimulated pubertal growth: the effects of testosterone and dihydrotestosterone on growth hormone and insulin-like growth factor-I in the treatment of short stature and delayed puberty. Journal of Clinical Endocrinology and Metabolism 1993 76 996–1001. (https://doi.org/10.1210/jcem.76.4.8473416)
- 56↑
Rochira V, Zirilli L, Maffei L, Premrou V, Aranda C, Baldi M, Ghigo E, Aimaretti G, Carani C, Lanfranco F. Tall stature without growth hormone: four male patients with aromatase deficiency. Journal of Clinical Endocrinology and Metabolism 2010 95 1626–1633. (https://doi.org/10.1210/jc.2009-1743)
- 57↑
Leung KC, Doyle N, Ballesteros M, Sjogren K, Watts CKW, Low TH, Leong GM, Ross RJM, Ho KKY. Estrogen inhibits GH signaling by suppressing GH-induced JAK2 phosphorylation, an effect mediated by SOCS-2. PNAS 2003 100 1016–1021. (https://doi.org/10.1073/pnas.0337600100)
- 58↑
Birzniece V Hepatic actions of androgens in the regulation of metabolism. Current Opinion in Endocrinology, Diabetes, and Obesity 2018 25 201–208. (https://doi.org/10.1097/MED.0000000000000405)
- 59↑
Langley RE, Cafferty FH, Alhasso AA, Rosen SD, Sundaram SK, Freeman SC, Pollock P, Jinks RC, Godsland IF & Kockelbergh R et al. Cardiovascular outcomes in patients with locally advanced and metastatic prostate cancer treated with luteinising-hormone-releasing-hormone agonists or transdermal oestrogen: the randomised, phase 2 MRC PATCH trial. Lancet: Oncology 2013 14 306–316. (https://doi.org/10.1016/S1470-2045(1370025-1)
