HIV antiretrovirals in FAERS: 205,550 adverse-event reports by the numbers

The FDA Adverse Event Reporting System (FAERS) holds 205,550 individual safety reports that name at least one HIV antiretroviral — a tenofovir prodrug (TDF or TAF), emtricitabine, lamivudine, dolutegravir, bictegravir, an older agent such as efavirenz or zidovudine, or another antiretroviral — in any role, from the first structured records through the openFDA extract dated June 8, 2026. Emtricitabine anchors the class at 92,656 reports, followed by tenofovir disoproxil fumarate (TDF) at 78,514 and lamivudine at 53,632; tenofovir alafenamide (TAF), the newer prodrug, sits at 44,606. Of the class total, 158,461 reports — 77.1 percent — carry an FDA seriousness flag, and 16,311 (7.9 percent) record a death flag or a fatal outcome. The reaction list is unusual for a chronic-therapy class: it is led by bone density decreased (21,577), chronic kidney disease (14,007), renal failure (12,459), and osteonecrosis (11,637) — a bone-and-kidney cluster that maps onto the tenofovir TDF safety history and, as the reporter mix shows, has been heavily amplified by product-liability litigation.

This article is a class-level descriptive read of HIV antiretroviral reporting in FAERS, written for pharmacovigilance, medical-affairs, and HIV market-access teams who meet these numbers in formulary reviews, TDF-to-TAF positioning, PrEP policy, and real-world-evidence dossiers. Every figure here is computed from the public openFDA FAERS extract (20,328,575 total reports, export dated 2026-06-08). It is not a disproportionality signal analysis, it is not a safety claim, and it is not incidence data. FAERS is a spontaneous-reporting system: it captures what was reported, never what was caused, and it has no exposure denominator. For the class-wide FAERS methodology and the database aggregate picture, see the companion analysis of 20 million FAERS reports and the FAERS report-reading workflow; for adjacent class cuts, see checkpoint inhibitors and JAK inhibitors in FAERS.

Methodology, and one deliberate exclusion

A report is counted for the class when an antiretroviral name appears anywhere in that report's drug list — as a suspect, concomitant, or interacting drug — using a substring match across the drug_substances, drug_generics, and drug_brands fields. Because a single report can name several drugs, and single-tablet regimens co-formulate three or four antiretrovirals, the per-molecule totals sum to far more than the 205,550 class-level (union, deduplicated) reports. A "serious" report is any report the FDA coded as serious; a report counts toward death when the seriousness field carries a death flag or the outcome field carries a Fatal outcome. Reactions are MedDRA Preferred Terms, case-normalized. One deliberate exclusion shapes this cohort: ritonavir is left out, even though it is a legitimate protease-inhibitor booster in HIV regimens, because its FAERS footprint is now dominated by the COVID-19 antiviral Paxlovid (nirmatrelvir/ritonavir) — including ritonavir inflates the class by roughly 57,000 reports and makes "covid-19" the single most-reported reaction, which would misrepresent an HIV analysis. FAERS does not link a reaction to a specific drug within a multi-drug report, so the reaction counts are report-level for reports that name an antiretroviral; they are not proof of causation.

Reporting volume, by molecule

The tenofovir split is the headline. Combined, TDF (78,514) and TAF (44,606) account for 123,120 mentions — the backbone of the class — and TDF still leads TAF despite having been largely replaced by TAF in guidelines, because TDF carries two decades of exposure and the entire bone/kidney litigation load. Emtricitabine leads on raw count only because it is co-formulated into nearly every tenofovir-based single-tablet regimen. Integrase inhibitors dolutegravir (26,502) and bictegravir (19,986) — the modern anchor drugs — report in proportion to their large but more recent exposure, not to disproportionate toxicity.

Seriousness, mortality, and demographics

The male skew (54.9 percent, versus 26.6 percent female with the remainder unspecified) tracks HIV epidemiology in the reporting population, and the median age of 39 reflects a chronic-therapy cohort rather than an elderly one. The 77.1 percent serious share is high for a well-tolerated modern therapy class, and the reason is visible in the reaction list and the reporter mix below: a large share of these reports are litigation filings that code serious bone and renal injuries.

The reaction list is a tenofovir-and-litigation signature

Three coherent clusters sit inside this list. The bone-and-kidney cluster — bone density decreased, osteoporosis, osteonecrosis, multiple fractures, chronic kidney disease, renal failure, Fanconi syndrome (1,514) — is the classic tenofovir disoproxil (TDF) signature: high systemic tenofovir exposure with TDF drives proximal renal tubulopathy and bone-mineral loss, effects that TAF's ~90-percent-lower plasma exposure largely avoids. The neuropsychiatric cluster — anxiety, emotional distress, anhedonia, depression (4,141) — is the efavirenz signature, the CNS-penetrant NNRTI long associated with vivid dreams, mood disturbance, and suicidal ideation. The pregnancy cluster — foetal exposure during pregnancy (8,333), maternal exposure during pregnancy (4,601) — reflects the intense post-2018 dolutegravir surveillance described below. Treatment-failure terms (virologic failure 4,721, drug resistance 4,057) round out the picture.

But the raw magnitude of the bone/renal terms is not a clean clinical incidence — it is litigation-shaped, which the reporter mix makes explicit.

Why lawyers file 13.3 percent of these reports

In the 20-million-report FAERS database-wide overview, lawyer-initiated reports are a low-single-digit share. In the HIV antiretroviral class they are 13.3 percent — 27,426 reports — the fingerprint of the tenofovir (TDF) product-liability litigation, in which mass-tort firms recruited patients who took TDF-based drugs and later developed kidney or bone injury and filed adverse-event reports to support their claims. That is why bone density decreased, osteonecrosis, tooth loss, and multiple fractures sit at the very top of the reaction list, and why any read of this class must separate the clinical signal from the litigation signal. (The reproducible lawyer share here is 13.3 percent; earlier internal drafts overstated it as roughly a quarter.) For pharmacovigilance and access analysts, the practical rules are: discount the bone/renal magnitudes for litigation inflation, expect duplicate filings across litigation stages, and weight physician- and HCP-sourced reports when estimating real clinical event patterns.

TDF versus TAF, in one table

TDF leads TAF on report volume not because it is used more today — it is not — but because it holds the historical exposure and the entire litigation load. TAF's counts come almost entirely from current high-market-share regimens (Biktarvy, Descovy) and reflect exposure, not a comparable toxicity signal. For P&T committees, this is the core evidence point behind preferring TAF-based single-tablet regimens despite higher acquisition cost: the downstream cost of managing TDF-associated CKD, fractures, and osteoporosis is real, and the FAERS record — even litigation-inflated — is consistent with it.

The efavirenz neuropsychiatric signal

Anxiety (9,335), emotional distress (8,798), and anhedonia (7,698) keep the neuropsychiatric signal near the top of the class list even though efavirenz has been displaced from first-line guidelines. Efavirenz, a CNS-penetrant NNRTI, is historically responsible for these reports; the signal persists because efavirenz remains in wide generic use in low- and middle-income settings, because some neuropsychiatric sequelae are chronic, and because integrase inhibitors — dolutegravir in particular — carry their own smaller insomnia/anxiety/depression tolerability signal that is actively monitored. For payer teams, this cluster is part of the value case for modern INSTI regimens: fewer treatment-limiting CNS effects and fewer discontinuations, against their higher list price.

Dolutegravir and the neural-tube-defect history, stated correctly

The 8,333 "foetal exposure during pregnancy" reports are the footprint of one of the clearest pharmacovigilance case studies of the past decade:

• May 2018: The Tsepamo surveillance study in Botswana reported a neural-tube-defect (NTD) signal — 4 cases in 426 births (0.94 percent) — in infants of women taking dolutegravir at conception, prompting FDA and WHO safety alerts and a label warning.
• 2020 update: With a far larger denominator, the NTD prevalence with dolutegravir at conception fell to 0.19 percent (7 of 3,591 births). The correct comparators from the same analysis were roughly 0.11 percent for non-dolutegravir antiretrovirals at conception and about 0.07 percent for HIV-negative pregnancies — that is, dolutegravir at conception was no longer distinguishable from other antiretrovirals, and 0.07 percent is the HIV-negative background rate, not the "non-dolutegravir" rate.
• Guideline reversal: On the stabilized data, the FDA, WHO, and HHS moved dolutegravir to a preferred first-line option in pregnancy, citing its high resistance barrier and rapid viral suppression.

The 8,333 FAERS reports therefore reflect mandated surveillance and registry tracking of a resolved-to-reassuring signal, not an active teratogenicity signal — a distinction access and clinical teams must preserve when they encounter the count.

PrEP and the long-acting injectable shift

Pre-exposure prophylaxis is a distinct slice of the class: "prophylaxis against HIV infection" is the recorded indication in 10,374 reports, on top of the treatment population. Two access-relevant patterns sit here. First, oral TDF/emtricitabine (Truvada) and TAF/emtricitabine (Descovy) PrEP carry the same tenofovir bone/renal considerations as treatment, which is why Descovy PrEP is positioned for patients where TDF renal or bone risk is a concern. Second, the class is shifting to long-acting injectables: cabotegravir (14,754 mentions across its treatment and Apretude PrEP uses) and, more recently, lenacapavir (981 mentions, rising) move HIV prevention and treatment from daily oral adherence to every-other-month or twice-yearly dosing. For long-acting agents the reproducible reporting concerns move from organ toxicity toward injection-site reactions and the resistance risk of sub-therapeutic drug levels during an interrupted dosing schedule — a different pharmacovigilance and adherence-support profile than the oral backbone, and one access teams should expect to grow in future extracts.

The hepatitis B overlap and the discontinuation-flare warning

Several antiretroviral backbones are also hepatitis B drugs: tenofovir (both TDF and TAF), lamivudine, and emtricitabine are active against HBV, and the FAERS record reflects it — hepatitis B is the recorded indication in 5,644 reports and chronic hepatitis B in 2,899. This overlap carries a specific, boxed safety point that access and clinical teams must not lose: abrupt discontinuation of these agents in a patient co-infected with HBV can precipitate a severe, sometimes fatal, acute hepatitis B flare, and the labels carry a boxed warning to that effect. The practical consequences are concrete — HBV status should be established before starting or stopping a tenofovir- or lamivudine/emtricitabine-containing regimen, and any formulary switch, prior-authorization lapse, or specialty-pharmacy gap that interrupts therapy in a co-infected patient is a clinical hazard, not merely an administrative one. It is the mirror image of the tenofovir toxicity story: the same molecules that generate the bone/renal litigation signal are also protective HBV therapy whose interruption is dangerous.

How the HIV class compares to other specialty classes

These are matched-scope, full-history, any-role figures for all three classes (see the checkpoint and JAK cuts). The HIV class's distinguishing feature is not mortality — checkpoint inhibitors, an advanced-cancer class, are far higher — but its uniquely high lawyer-reporter share and its chronic organ-toxicity (bone/renal) signature, both of which trace to the tenofovir TDF litigation.

The reporting trajectory, by year

Reporting climbed steadily through the 2010s with the expansion of single-tablet regimens, then spiked sharply in 2020 to 31,069 — roughly double the 2019 volume — before settling back into a band of about 13,000–16,000 reports per year. That 2020 bulge is not a clinical safety event; it coincides with the peak of tenofovir-disoproxil (TDF) bone- and kidney-injury litigation filings, the same litigation that puts lawyers at 13.3 percent of the class and drives the bone/renal terms to the top of the reaction list. Reading the trajectory without that context would misattribute a litigation reporting wave to a change in drug safety. (Years omitted from the table fall on the same trend line; 2026 is a partial year through the extract cut-off.)

Adherence, resistance, and backbone durability

The treatment-failure terms — virologic failure (4,721), drug resistance (4,057), and product dose omission issue (5,446) — carry a different lesson than the toxicity clusters. In HIV, unlike most chronic therapies, a lapse in adherence is not merely a temporary loss of symptom control: sub-therapeutic drug levels select resistance mutations that can permanently retire a regimen and narrow future options, which is why these terms belong to the safety conversation at all. This is the reproducible evidence behind adherence-support and benefit-design arguments — that copay accumulators, maximizers, prior-authorization delays, and specialty-pharmacy gaps which interrupt an antiretroviral backbone are a resistance risk, not just an out-of-pocket inconvenience — and it is one reason modern high-barrier integrase regimens, which are more forgiving of missed doses, carry value beyond their WAC.

What this means for pharmacovigilance, medical-affairs, and access teams

• Separate the clinical signal from the litigation signal. The bone/renal terms lead the list partly because 13.3 percent of reports are lawyer-filed; do not read the magnitudes as clinical incidence.
• The TDF-to-TAF value case is defensible. The tenofovir bone/renal signature supports preferred positioning of TAF-based single-tablet regimens despite higher WAC, given the downstream cost of TDF-associated CKD and fractures.
• Preserve fast neuropsychiatric-switch pathways. The efavirenz cluster — and the smaller dolutegravir CNS signal — argue for rapid formulary exceptions to a neutral backbone rather than access delays that risk virologic failure.
• Read the pregnancy count as surveillance, not signal. The 8,333 foetal-exposure reports reflect a resolved dolutegravir NTD story now guideline-preferred in pregnancy.
• Exclude the ritonavir/Paxlovid artifact from HIV analyses. Any class read that includes ritonavir will be swamped by COVID-19 reports; state the exclusion.

FAQ

Does FAERS prove tenofovir, efavirenz, or dolutegravir caused these events?

No. FAERS is spontaneous and denominator-free; a report reflects suspicion, not causation, and is subject to reporting bias, missing data, and duplicate (often litigation-driven) filings. The bone/renal cluster is consistent with the known tenofovir TDF mechanism, but its magnitude here is inflated by product-liability reporting.

Why did you exclude ritonavir from the HIV class?

Ritonavir is a legitimate HIV booster, but its current FAERS footprint is dominated by Paxlovid (nirmatrelvir/ritonavir) for COVID-19. Including it adds roughly 57,000 reports and makes COVID-19 the top reaction, which would misrepresent an HIV antiretroviral analysis. Cobicistat, an HIV/PrEP-only booster, is retained.

Does TAF being lower-count than TDF mean TAF is riskier per patient?

No — the opposite. TDF's higher count reflects two decades of exposure plus the entire litigation load; TAF's counts come from current high-share regimens and reflect exposure, not a comparable toxicity signal. TAF's systemic tenofovir exposure is roughly 90 percent lower than TDF's.

Is the dolutegravir pregnancy signal a reason to avoid it in pregnancy?

No. The high foetal-exposure count reflects mandated surveillance after the 2018 alert; long-term Tsepamo data brought the NTD prevalence down to ~0.19 percent and then in line with other antiretrovirals, and dolutegravir is now a preferred first-line option in pregnancy per FDA, WHO, and HHS.

Should hepatitis B status change how a regimen is switched?

Yes. Tenofovir (TDF and TAF), lamivudine, and emtricitabine are also hepatitis B drugs, and abrupt discontinuation in a co-infected patient can trigger a severe, sometimes fatal HBV flare — the labels carry a boxed warning. Establish HBV status before starting or stopping these agents, and treat any prior-authorization lapse or specialty-pharmacy gap that interrupts therapy in a co-infected patient as a clinical hazard, not an administrative one.

Sources

• US FDA, openFDA FAERS (Adverse Event Reporting System) extract, export dated 2026-06-08 (20,328,575 total reports). Class-union, per-molecule, outcome, reaction, reporter, and indication aggregates computed by PharmaDossier from the public FAERS extract using an any-role substring match across drug substance, generic, and brand fields, with ritonavir excluded to remove the Paxlovid/COVID-19 confounder. https://open.fda.gov/data/faers/
• Zash R, Holmes L, Diseko M, et al. "Neural-Tube Defects and Antiretroviral Treatment Regimens in Botswana." N Engl J Med 2019;381(9):827-840 — the Tsepamo surveillance study; subsequent updates reported dolutegravir-at-conception NTD prevalence of 0.19% (7/3,591) by April 2020 and 0.11% (10/9,460) by March 2022, converging with non-dolutegravir ART (~0.11%) and near the HIV-negative baseline (~0.07%). https://pubmed.ncbi.nlm.nih.gov/31329379/
• US FDA, Drug Safety Communication (May 2018): FDA warns of serious birth defects with dolutegravir (HIV) at conception (historical alert, later relaxed). https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-serious-birth-defects-can-occur-if-mother-takes
• HHS Clinical Info, Guidelines for the Use of Antiretroviral Agents in Adults and Adolescents with HIV — Table of common and severe adverse effects (tenofovir renal/bone, efavirenz neuropsychiatric). https://clinicalinfo.hiv.gov/en/guidelines/hiv-clinical-guidelines-adult-and-adolescent-arv/adverse-effects-antiretroviral-medications-full
• US FDA, prescribing information for Viread (TDF), Vemlidy/Descovy (TAF), Sustiva (efavirenz), and Tivicay (dolutegravir) — renal, bone, neuropsychiatric, and pregnancy sections. https://www.accessdata.fda.gov/scripts/cder/daf/