AndroGel Evidence Base Graded by GRADE: What the Clinical Trials Actually Show

What Is AndroGel and Why Does GRADE Matter Here

AndroGel is a hydroalcoholic testosterone gel applied daily to the upper arms, shoulders, or abdomen. It has been FDA-approved since 2000 for classic male hypogonadism, defined as consistently low serum testosterone combined with signs or symptoms of deficiency. Applying GRADE (Grading of Recommendations, Assessment, Development and Evaluation) to its evidence base matters because the drug is among the most prescribed testosterone formulations in the United States, yet its long-term outcome data remain thinner than its widespread use might suggest.

GRADE grades evidence across four levels: High, Moderate, Low, and Very Low. These levels reflect risk of bias, inconsistency, indirectness, imprecision, and publication bias across the body of trials. A drug can have a strong FDA indication and still carry only Moderate or Low GRADE certainty for specific outcomes when trials are short, populations are narrow, or surrogate endpoints substitute for patient-centered outcomes.

How GRADE Applies to Testosterone Gel Specifically

Most AndroGel trials used serum testosterone normalization as the primary endpoint. Serum T is a surrogate, not a patient-centered outcome like fracture rate or cardiovascular event rate. GRADE penalizes surrogate-dominated evidence by at least one level, which is why clinical outcome domains (sexual function, bone, cognition, cardiovascular) each earn their own GRADE rating below.

The Regulatory Foundation

The FDA approved AndroGel 1% in 2000 and AndroGel 1.62% in 2011. Both approvals rested on pharmacokinetic studies demonstrating that daily topical application achieved and maintained serum testosterone within the 300 to 1,000 ng/dL eugonadal reference range in most men. Those PK studies were not powered for clinical outcomes, a limitation that the T-Trials were later designed to address [1].

The T-Trials: The Highest-Quality Evidence Available

The Testosterone Trials (T-Trials) represent the best-controlled evidence for AndroGel to date. Published in the New England Journal of Medicine in 2016, the coordinated T-Trials enrolled 788 men aged 65 or older with serum testosterone below 275 ng/dL and at least one of three symptom domains: sexual dysfunction, low vitality, or impaired physical function [1]. Participants received testosterone gel (titrated to maintain serum T at 500 to 1,000 ng/dL) or placebo for 12 months.

The T-Trials were not a single trial. They were seven simultaneous, nested randomized controlled trials, each addressing a different outcome domain. This design gave GRADE reviewers domain-specific evidence rather than a single composite endpoint.

Sexual Function Domain (GRADE: Moderate)

The Sexual Function Trial within T-Trials found a statistically significant improvement in sexual activity, sexual desire, and erectile function with testosterone gel versus placebo [1]. The mean increase in the Psychosexual Daily Questionnaire sexual desire score was 0.58 points (95% CI 0.39 to 0.77, P<0.001). This result was consistent across pre-specified subgroups.

GRADE rates this evidence as Moderate rather than High because: (1) the trial lasted only 12 months, which limits conclusions about durability; (2) the population was restricted to men 65 or older with very low baseline T, reducing generalizability to younger men or those with borderline low T; and (3) effect sizes, while statistically significant, were modest in absolute terms.

The Endocrine Society 2018 clinical practice guideline states: "We suggest offering testosterone therapy to men with hypogonadism who have sexual dysfunction" (Recommendation 3.1, weak recommendation, low-quality evidence for most sexual outcomes) [2]. The discrepancy between the T-Trials moderate-quality finding and the Society's "low-quality" label reflects differences in the full evidence base the Society considered beyond just the T-Trials.

Physical Function Domain (GRADE: Low)

The Physical Function Trial failed to meet its primary endpoint. Men receiving testosterone did not achieve a statistically significant improvement in the 6-minute walk distance compared with placebo (between-group difference 10.7 m, 95% CI -3.7 to 25.1 m, P=0.14) [1]. The vitality trial also showed only modest benefit that did not reach statistical significance on the primary composite.

GRADE rates physical function evidence as Low for AndroGel based on T-Trials: the trial was negative on its primary outcome, and the confidence interval does not rule out a clinically meaningful benefit or harm.

Cognitive Function Domain (GRADE: Low)

The Cognitive Function Trial found no significant benefit of testosterone on any cognitive outcome measured over 12 months, including memory, executive function, and spatial ability [3]. The Endocrine Society guideline does not recommend testosterone therapy specifically to improve cognition [2].

Bone Density Evidence (GRADE: Moderate)

The Bone Trial within T-Trials showed that testosterone gel increased volumetric bone mineral density at the spine by 7.5% and trabecular bone score improved significantly compared with placebo over 12 months [4]. These are surrogate outcomes; the trial was not powered to detect fracture reduction.

A 2017 analysis by Snyder et al. Reported: "Testosterone treatment significantly increased volumetric bone mineral density and estimated bone strength" (P<0.001 for both spine and hip sites) [4].

GRADE rates this Moderate because: the imaging endpoints are validated surrogates, the trial was adequately randomized and blinded, and the effect sizes were clinically plausible. The downgrade from High reflects the absence of fracture endpoint data and the 12-month observation window.

Clinical Relevance of the Bone Finding

Hypogonadal men have roughly 2.5-fold higher fracture risk compared with eugonadal peers in epidemiological cohorts [5]. Whether testosterone gel translates the observed BMD gains into actual fracture reduction remains unproven. Clinicians treating osteoporotic hypogonadal men should consider co-prescribing bisphosphonate therapy rather than relying on testosterone alone, per current osteoporosis guidelines from the Endocrine Society [2].

Cardiovascular Evidence (GRADE: Low)

Cardiovascular safety is the most contested domain in AndroGel's evidence base. The FDA issued a drug safety communication in 2015 requiring all testosterone product labeling to carry a warning about possible increased risk of heart attack and stroke [6]. That communication was triggered in part by a 2010 placebo-controlled trial (the TOM Trial, N=209) that was stopped early due to excess cardiovascular events in the testosterone arm [7].

The TOM Trial Signal

The TOM (Testosterone in Older Men with Mobility Limitations) Trial enrolled older men with mobility limitations and found that the testosterone gel group experienced significantly more cardiovascular events (23 vs. 5 in placebo, P=0.001) over a mean follow-up of only 5.9 months [7]. The TOM population was at high baseline cardiovascular risk, which limits generalizability but raises a legitimate safety signal.

GRADE rates this cardiovascular signal as contributing Low-quality evidence of harm: the trial was stopped early (which inflates effect sizes), the population was selected for frailty, and the absolute event numbers were small.

TRAVERSE Trial Context

The TRAVERSE trial (N=5,246 men aged 45 to 80, mean follow-up 33 months) was designed specifically to assess cardiovascular safety of testosterone replacement. Preliminary results presented in 2023 and published in the New England Journal of Medicine showed that testosterone therapy was non-inferior to placebo for major adverse cardiovascular events (MACE: 182 events [7.0%] in testosterone vs. 190 [7.3%] in placebo; HR 0.96, 95% CI 0.78 to 1.17) [8]. This represents the strongest cardiovascular safety evidence to date.

However, TRAVERSE also found a statistically significant increase in pulmonary embolism (1.0% vs. 0.5%, P=0.03) and atrial fibrillation (3.5% vs. 2.4%, P=0.001) in the testosterone arm [8]. These findings raise caution for men with pre-existing thromboembolic risk or atrial fibrillation history.

GRADE for cardiovascular outcomes now reads: Low to Moderate depending on the specific endpoint. MACE non-inferiority is supported by one large RCT (Moderate for that specific endpoint). Atrial fibrillation and VTE signals remain Low-quality because they were secondary endpoints in TRAVERSE and require replication.

Pharmacokinetics and Dose-Response: What Clinicians Need to Know

Serum Testosterone Normalization

AndroGel 1.62% at the starting dose of 40.5 mg/day achieves mean steady-state serum testosterone of approximately 520 ng/dL in most hypogonadal men [9]. The dose may be titrated to 20.25 mg/day (minimum) or 81 mg/day (maximum) based on morning serum T drawn 2 hours after application. The FDA-approved target range is 300 to 1,000 ng/dL.

Bioavailability through intact skin is approximately 10% of applied dose, which is lower than intramuscular testosterone cypionate but provides a steadier daily serum T profile without the supraphysiologic peaks seen with weekly injections [10].

Dihydrotestosterone Elevation

Transdermal testosterone raises dihydrotestosterone (DHT) disproportionately compared with intramuscular routes because skin contains high 5-alpha-reductase activity. In the T-Trials, DHT rose by a mean of 60% above baseline [1]. Elevated DHT may increase benign prostatic hyperplasia symptom scores and is a theoretical concern for prostate cancer progression, though no trial has demonstrated a causal link to prostate cancer incidence [2].

Secondary Exposure Risk

The FDA label mandates a boxed warning about secondary testosterone exposure to women and children through skin contact [6]. Studies show that gel-to-skin transfer is detectable within 15 minutes of unprotected contact and that covering the application site with clothing reduces transfer by approximately 90 to 95%. Washing the site with soap and water before contact reduces transfer further [6].

Current Guideline Positions on AndroGel

AUA 2018 Guidelines

The American Urological Association 2018 guideline on testosterone deficiency recommends that clinicians offer testosterone therapy to symptomatic men with consistently low serum testosterone (two morning measurements below 300 ng/dL). The guideline states no preference for delivery route (injectable vs. Topical vs. Other) provided eugonadal serum levels are achieved [11]. Topical gels including AndroGel are listed as first-line options alongside injectable testosterone cypionate and enanthate.

Endocrine Society 2018 Clinical Practice Guideline

The Endocrine Society guideline recommends testosterone therapy for men with classic hypogonadism who have signs or symptoms of androgen deficiency. It specifies: "We suggest against starting testosterone therapy in patients who are actively trying to have children, have breast or prostate cancer, have a palpable prostate nodule or induration, have an elevated PSA (>4 ng/mL), or have untreated severe obstructive sleep apnea" [2].

The guideline further states: "We recommend that clinicians check serum testosterone levels 3 to 6 months after initiating treatment" and defines treatment success as morning serum T in the mid-normal range (400 to 700 ng/dL) [2].

EAU Guidelines on Male Hypogonadism

The European Association of Urology 2023 guidelines rate testosterone gel as a Grade B recommendation for male hypogonadism treatment, equivalent in strength to injectable formulations, noting that patient preference and adherence should guide route selection [12].

Comparing AndroGel to Other Testosterone Formulations

AndroGel is not the only option. The evidence base for different formulations varies, and clinicians should match the route to the patient's profile.

| Formulation | GRADE for T Normalization | Key Practical Difference | |---|---|---| | AndroGel 1% / 1.62% (topical) | Moderate | Daily application; secondary transfer risk | | Testosterone cypionate IM | Moderate | Weekly or biweekly injection; peak-trough variation | | Testosterone undecanoate IM (Aveed) | Moderate | Every-10-week injection after loading; REMS program required | | Testosterone pellets (subcutaneous) | Low | Every 3 to 6 months; minor surgical procedure | | Testosterone nasal gel (Natesto) | Low | Three-times-daily dosing; no transfer risk |

Injectable testosterone cypionate generally produces higher average serum T concentrations with greater variability. For men who prefer daily mood and energy stability, topical gels tend to be preferred, though adherence data across formulations are limited.

Absolute Contraindications and Monitoring Requirements

Contraindications

Men with known or suspected breast cancer, active prostate cancer, a hematocrit above 50%, untreated severe sleep apnea, or uncontrolled heart failure should not receive AndroGel [2, 6]. Men desiring fertility should use gonadotropin-based therapy instead, as exogenous testosterone suppresses LH and FSH, reducing intratesticular testosterone and spermatogenesis [2].

Recommended Monitoring Schedule

Per the Endocrine Society 2018 guideline, clinicians should check:

• Serum testosterone at 3 to 6 months after initiation, then annually
• Hematocrit at 3 to 6 months, then annually (hold or reduce dose if hematocrit exceeds 54%)
• PSA at 3 to 6 months, then annually in men over 40
• Bone mineral density at 1 to 2 years in men with osteoporosis or low-trauma fracture history
• Lipid panel at baseline and 12 months

A hematocrit rise above 54% is the most common laboratory adverse effect of testosterone therapy, occurring in approximately 5.8% of T-Trials participants on active treatment versus 1.0% on placebo [1].

Specific Populations: Evidence Gaps

Younger Hypogonadal Men (Age <50)

The T-Trials enrolled only men aged 65 or older, leaving a direct GRADE gap for younger hypogonadal men. Most data in this subgroup come from open-label PK studies and small randomized trials. GRADE for benefit in younger men would be rated Low for most clinical outcomes due to indirectness from the older-men evidence base.

Type 2 Diabetes and Metabolic Syndrome

A 2016 meta-analysis (N=1,562 across 19 RCTs) found that testosterone therapy reduced fasting glucose by a mean of 1.0 mmol/L and HbA1c by 0.87% in hypogonadal men with type 2 diabetes or metabolic syndrome (P<0.001 for both) [13]. GRADE for glycemic benefit is Moderate based on this pooled evidence, though no large trial has used a diabetes-specific primary endpoint.

Obesity-Related Hypogonadism

Men with functional hypogonadism driven by obesity (BMI >35) may experience testosterone normalization through weight loss alone. A 2019 RCT found that a 10% weight reduction achieved through diet restored eugonadal testosterone in 53% of obese hypogonadal men without testosterone therapy [14]. Clinicians should address obesity before or alongside initiating AndroGel in this population.

GRADE Summary Table for AndroGel Outcomes

| Outcome Domain | GRADE Certainty | Direction | Notes | |---|---|---|---| | Serum T normalization | High | Benefit | Consistent across PK and RCT data | | Sexual function | Moderate | Benefit | T-Trials; 12-month data only | | Bone mineral density | Moderate | Benefit | Surrogate; no fracture RCT | | Physical function | Low | No clear benefit | T-Trials primary endpoint negative | | Cognition | Low | No benefit | T-Trials cognitive domain negative | | Glycemic control (T2DM) | Moderate | Benefit | Meta-analysis; no dedicated primary endpoint trial | | MACE (cardiovascular events) | Moderate | Non-inferior to placebo | TRAVERSE; single trial | | Atrial fibrillation | Low | Possible harm | Secondary endpoint in TRAVERSE | | Venous thromboembolism | Low | Possible harm | Secondary endpoint in TRAVERSE | | Erythrocytosis | Moderate | Harm | Consistent across trials; hematocrit-dependent |