AndroGel Side Effects: Severity Distribution by Patient Phenotype

What the FDA Label Says About AndroGel Adverse Events

The FDA-approved prescribing information for AndroGel 1.62% identifies application-site reactions, headache, emotional lability, and hematocrit elevation as the most common adverse events in controlled trials [1]. The label carries a Black Box Warning about secondary exposure in women and children, a category of harm that is mechanistically distinct from pharmacologic toxicity in the treated patient [1].

In the 182-patient, 182-day key trial supporting AndroGel 1.62% approval, the most frequent treatment-emergent adverse events (TEAEs) were application-site reactions (5.5%), headache (4.4%), and increased hematocrit (3.8%) [1]. These rates appear modest, but the trial excluded patients with hematocrit above 50%, baseline PSA above 4 ng/mL, and a history of cardiovascular events within the prior 6 months, which means the enrolled population was lower-risk than many real-world TRT candidates.

How the Label Severity Tiers Map to Clinical Practice

The label groups events into "common" (occurring in 2% or more of subjects and exceeding placebo) and "serious" (requiring intervention or resulting in hospitalization). Grade-level severity mapping used in oncology (CTCAE) is not applied in the AndroGel label, but the Testosterone Trials (TTrials) used a modified serious-adverse-event definition that maps reasonably onto CTCAE grade 3 or higher [2].

Clinicians should treat any hematocrit above 54%, any PSA velocity above 0.75 ng/mL per year, and any symptomatic venous thromboembolism as grade 3 equivalents requiring immediate dose reduction or discontinuation.

Mild-to-Moderate Adverse Events: Who Gets Them

Mild adverse events from AndroGel are common, self-limiting, and phenotype-agnostic. Skin reactions dominate this tier.

Application-Site Reactions

Application-site erythema, pruritus, and dry skin occur in approximately 3 to 7% of users across trials [1]. These reactions are more frequent in men with atopic dermatitis or sensitive skin, and in those who apply the gel immediately after showering to still-damp skin. Switching from the 1% to the 1.62% formulation reduces total alcohol vehicle volume by roughly 40%, which may lower irritation frequency in reactive individuals [3].

Acne and Sebaceous Changes

Acne emerges in 3 to 8% of patients, most commonly in men under 40 years old who are starting from a relatively low baseline testosterone level and experience a steep upward shift in serum DHT [4]. The sebaceous gland response is androgen-receptor mediated and dose-dependent. Men with a personal or family history of cystic acne should be counseled before initiation that topical retinoids may be needed adjunctively.

Mood and Sleep Changes

Emotional lability, irritability, and mild sleep disruption occur in a subset of users, particularly during the first 4 to 8 weeks as the hypothalamic-pituitary-gonadal axis adjusts [2]. The TTrials (N=788 men aged 65 or older) reported no statistically significant difference in depressive symptom scores between testosterone and placebo at 12 months using the PHQ-9, suggesting that mood effects at standard doses are modest in older men [2].

Moderate-to-Severe Adverse Events: Phenotype Matters Significantly

Hematologic and cardiovascular adverse events carry real morbidity. Their probability is not evenly distributed.

Polycythemia: The Hematocrit Phenotype

Erythrocytosis (hematocrit above 50%, or above 54% by Endocrine Society threshold) is the most common laboratory adverse event requiring dose modification in clinical practice [5]. The Endocrine Society 2018 guideline states: "We suggest checking hemoglobin and hematocrit at baseline, at 3 to 6 months after initiating treatment, and then annually. If the hematocrit exceeds 54%, stop therapy until the hematocrit decreases to a safe level" [5].

Men at highest risk for polycythemia share a recognizable phenotype: baseline hematocrit of 46% or higher, current or former smoking history, obstructive sleep apnea (OSA), residence at high altitude, and chronic obstructive pulmonary disease. In a 2020 retrospective cohort of 1,114 men on injectable or topical testosterone, OSA was the single strongest independent predictor of hematocrit elevation above 52%, with an odds ratio of 3.1 (95% CI 2.0 to 4.8) [6]. Topical formulations like AndroGel produce lower and more stable serum testosterone peaks than weekly or biweekly injections, which translates to a modestly lower polycythemia rate compared with testosterone cypionate or enanthate [5].

Cardiovascular Events: The Cardiometabolic Phenotype

The cardiovascular safety of testosterone therapy remains actively debated. The TRAVERSE trial (N=5,204, median follow-up 33 months), published in the New England Journal of Medicine in 2023, found that testosterone therapy did not increase the rate of major adverse cardiovascular events (MACE) compared with placebo in men with hypogonadism and pre-existing or high-risk cardiovascular disease (HR 0.96, 95% CI 0.78 to 1.17) [7]. This was a non-inferiority finding, not a demonstration of cardiovascular benefit.

TRAVERSE also found a statistically significant increase in atrial fibrillation (3.5% vs. 2.4%), pulmonary embolism (0.9% vs. 0.5%), and acute kidney injury in the testosterone arm [7]. These signals matter for specific phenotypes.

Men who carry the highest absolute cardiovascular risk from AndroGel share these features: age above 65, established coronary artery disease or peripheral arterial disease, baseline hematocrit above 46%, active atrial fibrillation or prior VTE, and uncontrolled hypertension (systolic blood pressure above 160 mmHg). For this group, initiating AndroGel requires a documented informed-consent conversation about the TRAVERSE AF signal and a plan for hematocrit monitoring at 3 months.

Venous Thromboembolism

The FDA added a VTE warning to all testosterone products in 2014 following a review of spontaneous post-market reports [8]. The biological mechanism is plausible: testosterone increases erythropoiesis, which raises blood viscosity, and supraphysiologic levels stimulate platelet aggregation pathways. Clinically, VTE risk appears concentrated in men with inherited or acquired thrombophilias (Factor V Leiden, antiphospholipid antibody syndrome, prior DVT) and in those who develop polycythemia on therapy [8].

A 2016 FDA Drug Safety Communication confirmed the warning and recommended that prescribers evaluate patients for VTE if they report leg swelling, pain, or dyspnea, even months after stable TRT [8].

Prostate-Related Adverse Events by Phenotype

PSA Rise and Prostate Volume

Testosterone therapy reliably increases serum PSA by a mean of 0.3 to 0.5 ng/mL within the first 3 to 6 months of treatment in eugonadal-range restoration [5]. The Endocrine Society guideline specifies referral to urology for PSA rise above 1.4 ng/mL above baseline within any consecutive 12-month period, or any confirmed PSA above 4.0 ng/mL [5]. Men over 40 with a baseline PSA between 2.5 and 3.9 ng/mL, or those with first-degree relatives with prostate cancer, represent a higher-surveillance phenotype and warrant a urology consultation before initiation rather than after a PSA spike.

Lower Urinary Tract Symptoms

Benign prostatic hyperplasia (BPH) related symptoms can worsen with testosterone therapy in men who have baseline IPSS scores above 19, significant post-void residual urine, or concurrent alpha-1-blocker therapy [5]. This phenotype requires close IPSS monitoring at 3 and 6 months.

Secondary Exposure: A Distinct Safety Category

Who Is at Risk

The FDA Black Box Warning on AndroGel specifically addresses virilization in female partners and pediatric contacts through skin-to-skin transfer [1]. Virilization in women exposed to AndroGel-treated men has been documented in case reports, with signs including clitoral enlargement, voice deepening, and acne appearing after months of consistent exposure through shared bedding or direct skin contact [9].

Children are more sensitive because they have lower body mass and their androgen receptors are highly responsive. Reported pediatric cases involved accelerated bone age, pubic hair development in children as young as 2 years, and in one case, aggressive behavior in a 4-year-old boy whose father applied gel to his chest and held the child before the gel dried [9].

Mitigation is straightforward: apply gel to shoulders, upper arms, or abdomen; allow 5 to 6 minutes drying time; cover the site with clothing; wash hands thoroughly; and shower before prolonged skin contact with a partner or child.

Partner Serum Level Data

In a pharmacokinetic sub-study of AndroGel 1%, female partners of treated men showed measurable serum testosterone levels above the normal female range (above 70 ng/dL) when skin-to-skin contact occurred within 2 hours of application without a barrier garment [10]. Levels returned to baseline when the barrier-garment protocol was followed consistently [10].

Severity Distribution Summary by Patient Phenotype

The table below organizes the clinically meaningful adverse events into a phenotype-risk matrix based on the trial data and FAERS signals reviewed above.

| Adverse Event | Severity Tier | Highest-Risk Phenotype | Monitoring Interval | |---|---|---|---| | Application-site reaction | Mild | Atopic skin, high-alcohol vehicle | As needed | | Acne | Mild to moderate | Age <40, high DHT sensitivity | 3-month visit | | Mood / sleep change | Mild | Rapid testosterone escalation | 4 to 8 weeks | | Erythrocytosis | Moderate to severe | OSA, smoker, baseline HCT >46% | 3 months, then annually | | Atrial fibrillation | Moderate to severe | Age >65, pre-existing AF risk | ECG if palpitations | | VTE | Severe | Thrombophilia, polycythemia | Symptom-based | | PSA rise | Moderate (monitoring trigger) | Age >50, family Hx prostate Ca | 3 to 6 months | | Secondary virilization | Severe (third party) | Female partner, child contact | Education at initiation |

FAERS Post-Market Signal Overview

The FDA Adverse Event Reporting System (FAERS) database through Q4 2023 shows that testosterone-containing products collectively rank among the top 10% of drugs by cardiovascular serious-outcome reports in adult males [8]. For gel formulations specifically, the most-reported serious outcomes are polycythemia, PE, and MI, in that order by report count. FAERS data carry well-known limitations: they capture voluntary reports, lack denominators, and are subject to notoriety bias following regulatory communications.

The 2014 FDA safety review that prompted the testosterone label update analyzed spontaneous reports and two observational studies, one of which (Vigen et al., JAMA 2013) has since been criticized for methodologic errors, including inclusion of women in a men-only dataset [11]. The updated 2015 FDA testosterone advisory committee did not mandate a complete cardiovascular contraindication but required a label update stating that cardiovascular risk has not been established [8].

What FAERS Cannot Tell Us

FAERS signal counts do not translate directly to incidence rates. A clinical trial like TRAVERSE, with its prospective design and 5,204 randomized men, provides stronger evidence for incidence rate estimation than any FAERS analysis. Clinicians should weight TRAVERSE data above spontaneous report counts when counseling patients, while remaining alert to new FAERS signals that might represent events outside TRAVERSE's eligibility criteria.

Initiating and Monitoring AndroGel Safely Across Phenotypes

Baseline Workup Checklist

Before starting AndroGel, the Endocrine Society recommends obtaining: two morning serum testosterone levels below 300 ng/dL on separate days, serum LH and FSH (to confirm primary vs. Secondary hypogonadism), PSA, hematocrit, and a documented discussion of fertility preservation if the patient is of reproductive age [5]. Men with active or recent prostate cancer, hematocrit above 50% at baseline, severe untreated OSA, or a history of VTE are candidates for an alternative approach or specialist input before TRT initiation.

Dose Adjustment and Switching Logic

AndroGel 1.62% is typically started at 40.5 mg (2 pump actuations) applied once daily to the shoulders or upper arms. After 14 days, serum testosterone is checked 2 to 8 hours after morning application. If the level remains below 300 ng/dL, the dose can be increased to 60.75 mg (3 actuations); if above 1,050 ng/dL, it should be reduced to 20.25 mg (1 actuation) [1]. Maintaining trough testosterone in the 400 to 700 ng/dL range minimizes both under-treatment symptoms and supraphysiologic hematologic adverse events.

Men who develop erythrocytosis on gel may benefit from switching to a lower-frequency regimen rather than discontinuation, since injectable formulations with longer intervals allow intentional dose-interval titration. Alternatively, therapeutic phlebotomy can be considered in men whose hematocrit reaches 54% to 56% if dose reduction alone is insufficient, though phlebotomy is an off-label intervention in this context [5].

In the Testosterone Trials, testosterone gel (applied daily) produced a mean serum testosterone of 489 ng/dL at 12 months, compared with 234 ng/dL in the placebo group, and the serious adverse event rate was 23.6% in the testosterone group versus 20.9% in the placebo group (not statistically significant at that sample size) [2].