AndroGel FAERS Safety Signals: FDA Post-Market Surveillance Data and What It Means

What FAERS Data Tells Us About AndroGel

The FDA Adverse Event Reporting System is a passive surveillance database that collects voluntary reports of adverse events associated with marketed drugs. For AndroGel (testosterone gel), FAERS became a primary early-warning tool that detected safety signals years before confirmatory trials delivered results. FAERS data is not proof of causation. It is a signal-detection system.

Between 2000 and 2014, FAERS accumulated a disproportionate number of cardiovascular adverse event reports linked to testosterone products, including AndroGel. A 2014 pharmacovigilance analysis identified that myocardial infarction, stroke, and death were among the most frequently reported serious outcomes for testosterone therapies [1]. The reporting odds ratio for cardiovascular events associated with testosterone exceeded the background rate for comparable drug classes in the database. The FDA responded with a January 2014 safety communication announcing an investigation into cardiovascular risk with all approved testosterone products [2].

Two observational studies accelerated the FDA's response. Vigen et al. published a retrospective cohort analysis in JAMA (2013) showing a 29% increase in the composite of death, MI, and stroke among men receiving testosterone prescriptions [3]. Finkle et al. reported in PLoS ONE (2014) that the risk of nonfatal MI doubled in the 90 days following a testosterone prescription in men aged 65 and older [4]. Both studies relied on claims data and had methodological limitations, but the FAERS signal and the observational findings together moved the FDA to act.

The 2015 Label Revision: What Changed

FDA required all testosterone product manufacturers to revise labeling in March 2015. The changes were substantial. The agency narrowed the approved indication from broad "hypogonadism" to men with low testosterone caused by specific, documented medical conditions: disorders of the testes, pituitary, or hypothalamus [2]. Age-related testosterone decline alone no longer met the labeled indication.

The revised label also added a warning about possible increased risk of heart attacks and strokes. This was not a boxed warning but a precaution-level addition to the Warnings and Precautions section. FDA simultaneously mandated that manufacturers conduct post-market studies to clarify the cardiovascular risk. The Endocrine Society's 2018 clinical practice guideline echoed this caution, recommending testosterone therapy only for men with "unequivocally low serum testosterone concentrations" combined with consistent symptoms [5].

Dr. Robert Califf, then FDA Deputy Commissioner, stated in the 2015 announcement: "Patients should be made aware of the possible increased cardiovascular risk when deciding whether to use or continue testosterone therapy" [2]. The label change reflected a broader FDA stance that testosterone prescribing had expanded beyond the evidence base, with an estimated 25% of men initiating testosterone therapy never having a baseline testosterone level checked [6].

Cardiovascular Signal: From FAERS to TRAVERSE

The cardiovascular safety question dominated the AndroGel regulatory narrative for nearly a decade. FAERS data showed cardiovascular events as the most frequently reported serious adverse outcome category. But passive surveillance cannot establish causality.

The TRAVERSE trial, published in the New England Journal of Medicine in 2023, was the definitive answer. This randomized, double-blind, placebo-controlled trial enrolled 5,246 men aged 45 to 80 with hypogonadism and pre-existing or high risk of cardiovascular disease [7]. The primary endpoint was time to first major adverse cardiovascular event (MACE): a composite of cardiovascular death, nonfatal MI, and nonfatal stroke.

Results were reassuring for the primary endpoint. The hazard ratio for MACE was 0.96 (95% CI: 0.78 to 1.17), meeting the pre-specified non-inferiority margin [7]. Testosterone did not increase the rate of heart attacks, strokes, or cardiovascular death compared to placebo in this high-risk population.

The trial used transdermal testosterone gel (1.62%), making its findings directly applicable to AndroGel users. Mean follow-up was 33 months. The testosterone group maintained mean serum testosterone levels in the mid-normal range (350 to 500 ng/dL) throughout the study.

Pulmonary Embolism and Venous Thromboembolism

While TRAVERSE resolved the MACE question, it raised a new concern. Pulmonary embolism occurred in 0.9% of testosterone-treated men vs. 0.5% of placebo recipients [7]. This finding aligned with earlier FAERS signals and a 2014 FDA safety communication that had already added venous thromboembolism (VTE) to the testosterone product labeling [2].

The biological mechanism is plausible. Testosterone stimulates erythropoiesis, increasing red blood cell mass and blood viscosity. Polycythemia (hematocrit >54%) is the most common laboratory abnormality reported in FAERS for topical testosterone products. In the T-Trials, a coordinated set of seven NIH-funded studies enrolling 790 men aged 65 and older, hematocrit elevations above 54% occurred in 3.2% of testosterone-treated men vs. 0.5% of placebo recipients within the first 12 months [8].

Higher doses produce more erythrocytosis. A dose-ranging pharmacokinetic study of testosterone gel reported polycythemia rates of up to 18.2% at supratherapeutic levels [9]. Current Endocrine Society guidelines recommend checking hematocrit at baseline, at 3 to 6 months, then annually, with dose reduction or phlebotomy if hematocrit exceeds 54% [5].

Dr. Shalender Bhasin, principal investigator of the TRAVERSE trial and professor at Harvard Medical School, noted: "The increase in pulmonary embolism warrants careful monitoring, and clinicians should assess thromboembolic risk factors before initiating testosterone therapy" [7].

Secondary Exposure and Transference Risk

AndroGel carries a boxed warning for secondary exposure, the most prominent safety feature on its FDA-approved label [10]. This warning is unique to topical testosterone formulations and does not apply to injectable testosterone cypionate or enanthate.

FAERS reports of virilization in children and women exposed to testosterone gel through skin-to-skin contact prompted the initial boxed warning in 2009. Reported effects in children included premature pubic hair development, advanced bone age, increased aggression, and penile or clitoral enlargement. Some cases involved children as young as 9 months old.

The label mandates specific risk-mitigation steps: apply the gel only to clean, dry, intact skin of the shoulders and upper arms; cover the application site with clothing after the gel dries; wash hands immediately after application; and wash the application site before any skin-to-skin contact with another person [10]. Despite these instructions, FAERS continues to receive transference reports, suggesting real-world adherence to the instructions is inconsistent.

The transference signal is a reminder that FAERS captures adverse events that clinical trials may undercount. Trials control application technique and monitor household contacts. Real-world prescribing does not.

Liver, Prostate, and Psychiatric Signals

FAERS data also contains reports of hepatotoxicity, prostate events, and psychiatric effects associated with testosterone products, though these signals are smaller and less clearly attributed to AndroGel specifically.

Hepatotoxicity. Severe liver injury is primarily associated with oral 17-alpha-alkylated androgens, not transdermal testosterone. FAERS reports of liver injury with AndroGel are rare and generally confounded by concurrent medications or pre-existing liver disease. The current AndroGel label does not carry a liver warning, distinguishing it from oral methyltestosterone [10].

Prostate. The FDA label requires monitoring for prostate cancer. The TRAVERSE trial reported higher rates of prostate biopsy in the testosterone group (3.5% vs. 2.8%) but no statistically significant difference in diagnosed prostate cancer (0.5% vs. 0.4%) [7]. The Endocrine Society guideline recommends PSA measurement at 3 to 6 months and then per age-appropriate screening guidelines [5]. Testosterone therapy remains contraindicated in men with known prostate cancer or breast cancer.

Psychiatric effects. Mood changes, irritability, and aggression are listed in the AndroGel label. FAERS reports of psychiatric adverse events exist but represent a small fraction of total reports. The T-Trials found no significant difference in depressive symptoms between testosterone and placebo groups, and modest improvements in sexual function and vitality domains [8].

How to Read FAERS Data Correctly

FAERS data requires careful interpretation. Several limitations apply to every signal discussed in this article.

Reporting is voluntary. Health care professionals and consumers can submit reports, but there is no requirement to do so. This creates reporting bias: serious events are more likely to be reported than mild ones, and media attention (such as the testosterone-cardiovascular controversy) can spike reporting rates independent of true incidence changes.

FAERS cannot calculate incidence rates. The denominator (total number of patients taking a drug) is unknown. A drug with more reports may simply be more widely prescribed, not more dangerous.

Duplicate reports exist. The same event may be reported by a physician, a pharmacist, and the patient. FDA staff deduplicate where possible, but the process is imperfect.

Causation is not established. A report of an MI in a patient taking AndroGel does not mean AndroGel caused the MI. The patient population prescribed testosterone (older men with comorbidities) has a high baseline rate of cardiovascular events regardless of treatment.

These limitations do not make FAERS data useless. FAERS detected the cardiovascular and VTE signals that eventually led to the TRAVERSE trial and multiple label changes. The system works as intended: it generates hypotheses that randomized trials then test. For AndroGel, TRAVERSE largely resolved the MACE hypothesis (no increased risk) while confirming the VTE signal (increased pulmonary embolism).

Current Label Status and Monitoring Requirements

The 2023 AndroGel label reflects the cumulative effect of two decades of FAERS signals, observational studies, and the TRAVERSE trial. The current label includes a boxed warning for secondary exposure in children, warnings for cardiovascular risk and VTE, and precautions for polycythemia, sleep apnea, and lipid changes [10].

Prescribers using AndroGel should follow this monitoring protocol, consistent with the Endocrine Society 2018 guideline [5]:

• Confirm the diagnosis with two morning total testosterone measurements <300 ng/dL before starting therapy
• Measure hematocrit at baseline, 3 to 6 months, then annually
• Check PSA at baseline and 3 to 6 months, then per screening guidelines
• Monitor testosterone levels at 3 to 6 months, targeting 400 to 700 ng/dL on trough measurement
• Assess bone mineral density at baseline if osteoporosis risk factors are present
• Discontinue therapy and refer for hematology if hematocrit exceeds 54%

Patients should receive written instructions on gel application technique and household member protection. The FDA's REMS (Risk Evaluation and Mitigation Strategy) for testosterone gel includes a medication guide that must be dispensed with each prescription [2].

Men currently using AndroGel who have questions about the FAERS data or label warnings should discuss their individual risk profile with a board-certified endocrinologist or urologist, because the absolute risk increase for VTE in TRAVERSE was 0.4 percentage points over a mean follow-up of 33 months [7].