Category: Drug Safety Communications

This section covers FDA Drug Safety Communications and related regulatory updates. Posts in this category summarize new safety alerts, label changes, and emerging evidence that may affect clinical decision making.

Tavneos (Avacopan) and Serious Liver Injury: What Patients and Clinicians Need to Know, Including What the FDA Isn’t Saying Publicly

⚠️ Key Safety Summary: Read This First On March 31, 2026, the FDA issued a Drug Safety Communication identifying 76 cases of drug-induced liver injury (DILI) linked to Tavneos (avacopan), including 8 deaths and 54 hospitalizations. Seven biopsy-confirmed cases involved vanishing bile duct syndrome (VBDS), a potentially irreversible liver condition, and 3 of those were fatal. Median time from starting Tavneos to liver injury onset: 46 days. Most cases occurred within 60 days. Updated monitoring requirements: Patients on Tavneos should have liver function tests every 2 weeks for the first month, then monthly for 5 months. Discontinue immediately if ALT/AST exceed 3 times the upper limit of normal (ULN) or ALP exceeds 2 times ULN. Critical context: The FDA had already requested in January 2026 that Amgen voluntarily withdraw Tavneos from the U.S. market. Amgen refused. The drug remains available.

This story is more complicated than a standard FDA drug safety alert. Tavneos (avacopan) was already under significant regulatory pressure when the March 31, 2026 safety communication was issued, and understanding why the drug is still on the market requires knowing the full context, not just the liver injury numbers.

The liver injury signal is real and serious. Eight people have died. But the situation patients and clinicians are navigating is also one in which a manufacturer has declined a federal request to remove a drug from shelves, both U.S. and European regulators are reviewing the integrity of the clinical trial data that supported the drug’s original approval, and a major watchdog organization is publicly challenging the FDA’s failure to escalate. All of this is happening while people with a serious, life-threatening autoimmune disease continue to be treated with, and in some cases depend on, the drug in question.

Here is the whole picture.

The Disease Tavneos Treats: ANCA-Associated Vasculitis

ANCA-associated vasculitis (AAV) is a group of rare autoimmune diseases in which the body’s immune system attacks and destroys small-to-medium blood vessels throughout the body. The two forms Tavneos is approved to treat are granulomatosis with polyangiitis (GPA, formerly called Wegener’s granulomatosis) and microscopic polyangiitis (MPA).

These are serious diseases. Untreated or inadequately managed, they can destroy kidney function, damage the lungs, and be fatal. The standard treatment for decades has involved high-dose glucocorticoids (steroids) combined with immunosuppressants such as cyclophosphamide or rituximab. These regimens work, but carry significant toxicity of their own. Chronic steroid use is associated with infection, bone loss, diabetes, and cardiovascular disease. Any therapy that could reduce steroid burden while maintaining disease control represents a genuine clinical advance.

Avacopan works by blocking the complement C5a receptor, which plays a role in driving neutrophil-mediated inflammation in AAV. The ADVOCATE Phase 3 trial showed it could achieve non-inferior remission rates compared with prednisone tapering at week 26, and superior sustained remission at week 52, with significantly less glucocorticoid exposure. Published in the New England Journal of Medicine in 2021, those results drove FDA approval. That is a clinically meaningful result for a disease where the side effects of standard treatment are themselves a major burden.

The Liver Injury Signal: What the Data Shows

Hepatotoxicity, meaning drug-induced liver injury, was not a surprise finding with avacopan. It was identified in premarket clinical trials and included in the drug’s prescribing information as a warning from the time of approval in 2021. In the ADVOCATE trial itself, 5.4% of patients in the avacopan arm experienced serious adverse events related to liver function, compared with 3.6% in the prednisone arm.

What changed, and what the March 31, 2026 safety communication addresses, are two new and more severe categories of concern that emerged in the postmarketing period:

Category	Number of Cases	Outcomes
All DILI cases (reasonable causal evidence)	76 total	74 serious outcomes
Hospitalizations	54	—
Deaths	8	All fatal by definition
Cholestatic or mixed injury pattern	38 of 60 with lab data	Elevated ALP + bilirubin
Biopsy-confirmed VBDS	7	All hospitalized; 3 fatal
Median time to onset	46 days	Range: 22 to 140 days
Cases from Japan	66 of 76	Largest concentration globally

Source: FDA Drug Safety Communication. March 31, 2026.

What is vanishing bile duct syndrome (VBDS)? VBDS is a rare and serious condition in which the small bile ducts inside the liver are progressively destroyed. Bile, the digestive fluid produced by the liver, can no longer drain properly, leading to a backup of bile acids in the liver and bloodstream. It is called “vanishing” because on liver biopsy, the small intrahepatic bile ducts that are normally present in portal tracts have disappeared. The resulting damage can be permanent and may eventually progress to cirrhosis or liver failure if not caught early. VBDS is most commonly caused by drug-induced liver injury, immune-mediated disorders, infections, and malignancy. It is distinctly different from the transient transaminase elevations seen in many drug reactions. It is a structural injury to the bile duct architecture itself. Clinically, patients typically present with jaundice (yellowing of skin or eyes), pruritus (intense itching that is often worse at night), and fatigue. In the avacopan VBDS cases, the majority occurred within 60 days of starting treatment.

Why Are 87% of Cases From Japan?

The geographic concentration of DILI and VBDS cases is one of the most striking features of this safety signal. Of 76 total DILI cases, 66 were reported from Japan, approximately 87%. Of the 7 biopsy-confirmed VBDS cases, 6 were from Japan.

Amgen has noted that VBDS cases from Japan primarily involved patients aged 65 and older. Several factors may contribute to the geographic pattern, none of which are definitively established:

Pharmacogenomic differences: Japanese patients may have different expression profiles or activity levels for the drug-metabolizing enzymes responsible for avacopan clearance, potentially altering hepatic drug exposure.

AAV epidemiology: MPA is substantially more prevalent in Japan than GPA compared with Western countries, and the two conditions may involve different baseline inflammatory profiles affecting hepatic susceptibility.

Concomitant medications: Patients in Japan may more frequently receive certain co-medications. Antibiotics such as trimethoprim/sulfamethoxazole, commonly given as infection prophylaxis in immunocompromised patients, have themselves been associated with DILI and may interact synergistically.

Post-marketing surveillance intensity: Japan has a notably rigorous pharmacovigilance system, and some of the apparent geographic concentration may reflect more systematic case capture rather than true biological difference.

Age and comorbidity profile: The older age of most Japanese VBDS cases may reflect a population with greater baseline hepatic vulnerability.

A case report published in Annals of Internal Medicine: Clinical Cases documented VBDS in a 74-year-old patient with MPA treated with avacopan, with a Naranjo Adverse Drug Reaction score of 6 (probable causality). That report noted the importance of monitoring compliance: in that case, liver enzyme testing had been inadvertently delayed, which may have contributed to the severity of the injury. It also noted that prior DILI episodes may increase vulnerability to subsequent drug-related liver injury.

The Bigger Story: FDA Requested Withdrawal. Amgen Said No.

The March 31 safety communication cannot be read in isolation. Six weeks earlier, on January 16, 2026, the FDA had privately requested that Amgen voluntarily withdraw Tavneos from the U.S. market. Amgen disclosed this publicly in February, and on January 28 formally informed the FDA it would not comply.

The FDA’s withdrawal request cited two concerns. The first was hepatotoxicity, specifically the emerging DILI and VBDS signal that became the subject of the March safety communication. The second was a data integrity issue: the FDA raised questions about a process followed by ChemoCentryx (the original developer) to re-adjudicate primary endpoint results for 9 of the 331 patients in the ADVOCATE trial, the sole pivotal study supporting avacopan’s approval.

What was the ADVOCATE endpoint re-adjudication controversy? The ADVOCATE trial used the Birmingham Vasculitis Activity Score (BVAS) to assess disease activity. The primary endpoint was remission (BVAS = 0) at week 26. After investigators originally scored certain patients, a post-hoc adjudication committee reviewed and changed the scores for 9 patients. The FDA first raised concerns about this process during the original 2021 review, and the FDA advisory committee vote on avacopan’s approval was close. These disputes were later publicly aired in a civil investor lawsuit (Amgen won in August 2025) and triggered the EMA’s January 2026 safety review of avacopan in Europe. The ADVOCATE re-adjudication issue is about whether the efficacy data supporting avacopan’s approval was handled appropriately, specifically whether changing endpoint scores for 9 patients affected the trial’s outcome assessment in ways that could have influenced regulatory decisions.

Amgen’s response has been consistent: the company maintains that Tavneos has a favorable benefit-risk profile, that liver toxicity is a known and labeled risk, and that more than 7,000 patients have been treated with it since approval. The company has indicated it submitted a proposed label update to the FDA in 2024 to add VBDS to the prescribing information, a request that was still pending at the time of the March 31, 2026 safety communication.

It is also contextually relevant that Tavneos generated $459 million in sales in 2025, growing 62% year-over-year, making it one of Amgen’s fastest-growing products. That commercial context is not dispositive in evaluating Amgen’s position, but it is part of the full picture of why a voluntary withdrawal request would face resistance.

Dr. Robert Steinbrook, Health Research Group Director at Public Citizen, stated in a formal release on March 31, 2026 that if the FDA in January 2026 requested that avacopan be voluntarily withdrawn from the U.S. market, the agency needed to explain publicly why it had not made that request publicly, and why the prescribing information did not yet include a boxed warning for the risk of fatal liver disease. He characterized these as urgent questions for the FDA to answer.

That criticism represents a legitimate and so far unanswered question about regulatory transparency. The FDA’s standard tools when a company declines voluntary withdrawal are limited: the agency can initiate mandatory withdrawal proceedings, but these require formal regulatory steps and take time, and no such proceedings have been announced.

For Patients Currently Taking Tavneos

If you are taking Tavneos for ANCA-associated vasculitis, the most important message is this: do not stop the medication on your own without speaking to your rheumatologist first. Stopping avacopan abruptly without a transition plan in a patient with active GPA or MPA could allow disease flare, which carries its own serious risks including kidney damage.

Symptoms of liver injury: seek care immediately if you develop any of these Unusual fatigue or weakness that is more than your baseline; nausea or vomiting without another clear cause; itching (pruritus), especially if persistent or worse at night; yellowing of your skin or the whites of your eyes (jaundice); light-colored or pale stools; dark, tea-colored urine; pain or swelling in the upper right abdomen. These symptoms can appear within the first six weeks of treatment. Do not wait for your next scheduled appointment. Contact your provider the same day.

If you have concerns about whether to continue Tavneos given the safety communication, that conversation belongs with your rheumatologist specifically. The benefit-risk calculation is individual: it depends on your disease severity, how well your vasculitis is controlled, your liver function baseline, and whether alternative regimens are viable for your situation. The Vasculitis Foundation has issued a patient-facing update on this situation and is a good resource for community support and current information.

Adverse events should be reported to FDA MedWatch at 1-800-332-1088 or online at fda.gov/safety/medwatch.

For Clinicians: Monitoring Protocol and Decision Framework

The FDA’s updated monitoring recommendations are specific and represent an intensification of the original labeling:

Time Period	Monitoring Frequency	Action Threshold
First month of treatment	Every 2 weeks	Discontinue if ALT/AST greater than 3 times ULN, or ALP greater than 2 times ULN
Months 2 to 6	Monthly	Same thresholds; monitor for new symptoms
After 6 months	As clinically indicated	Remain vigilant for late-onset cases
Any time	Immediately on symptoms	Jaundice or pruritus: discontinue and refer to hepatology

Key clinical judgment points for prescribers:

The DILI pattern in most cases is cholestatic or mixed, characterized by elevated alkaline phosphatase (ALP) and total bilirubin rather than isolated transaminase elevation. This pattern can progress more insidiously than hepatocellular injury. Do not wait for transaminase elevation alone to act.

VBDS is a structural, potentially irreversible injury. If a patient develops jaundice or persistent pruritus, discontinue avacopan promptly and refer to hepatology without waiting to see if liver enzymes normalize on repeat testing.

Prior DILI episodes from any cause may increase susceptibility to subsequent drug-induced liver injury. Take a thorough medication and hepatic history before initiating avacopan.

The median onset of 46 days means the highest-risk window falls exactly in the first two months. The biweekly monitoring in month one is not optional in the current regulatory and clinical context.

Document your monitoring compliance carefully. The published VBDS case report noted that a two-week delay in scheduled liver monitoring may have contributed to injury severity in that patient.

For patients with uncontrolled disease where avacopan provides meaningful clinical benefit, the benefit-risk calculation may still favor continuation with rigorous monitoring. For patients in stable remission or with other viable options, the calculus is different. Consult ACR vasculitis guidelines and consider hepatology co-management when initiating or continuing therapy given the current safety signal.

What Happens Next

EMA review

The European Medicines Agency launched a review of avacopan in January 2026, citing concerns about data integrity in the ADVOCATE trial. The EMA will evaluate all available evidence to determine whether the handling of trial data affects the overall benefit-risk profile of avacopan for European patients. That review is ongoing.

FDA mandatory withdrawal authority

When a pharmaceutical company declines a voluntary withdrawal request, the FDA can initiate mandatory withdrawal proceedings under 21 U.S.C. §355(e), but this requires a formal process that includes opportunity for hearing. The FDA has not announced it is pursuing this path for avacopan. Whether and when the agency escalates remains an open question.

Label update (pending)

Amgen submitted a proposed label update in 2024 to add VBDS to the prescribing information. As of the March 31, 2026 safety communication, that label change had not yet been approved. The current label includes a hepatotoxicity warning but does not specifically mention VBDS by name.

Boxed warning

Public Citizen has called for a boxed warning, the FDA’s strongest label alert, reserved for serious or life-threatening risks, for fatal liver disease associated with avacopan. The current label does not carry a boxed warning for hepatotoxicity. Given eight confirmed deaths, the question of whether the FDA will move in this direction in any eventual label update bears watching.

For related coverage of how FDA regulatory tools, including drug safety communications, mandatory withdrawal authority, and label change procedures, work in practice, see our earlier post on what the FDA’s contrasting decisions on camizestrant and vepdegestrant reveal about regulatory evidence standards.

If you are a patient or caregiver navigating this situation:

The Vasculitis Foundation is actively tracking developments and providing patient-centered guidance on the Tavneos situation. The American College of Rheumatology maintains current treatment guidelines for AAV. Adverse events can be reported directly to the FDA MedWatch system at 1-800-332-1088 or online.

Sources

FDA Drug Safety Communication: FDA Identifies Cases of Serious Liver Injury in Patients Taking Tavneos (avacopan). March 31, 2026. fda.gov.

Tavneos original FDA approval: FDA approves avacopan for ANCA-associated vasculitis. October 7, 2021. fda.gov.

Amgen prescribing update: Important Update Regarding TAVNEOS (avacopan). Amgen. April 2026.

ADVOCATE Phase 3 trial: Jayne DRW, Merkel PA, Schall TJ, Bekker P. Avacopan for the Treatment of ANCA-Associated Vasculitis. N Engl J Med. 2021;384:599-609.

ADVOCATE trial registration: NCT02994927. ClinicalTrials.gov.

VBDS case report: Annals of Internal Medicine: Clinical Cases. Avacopan Causing Vanishing Bile Duct Syndrome in an Adult Patient With Microscopic Polyangiitis. 2024. doi:10.7326/aimcc.2024.0602

Withdrawal controversy: Medscape. What’s at Issue in the FDA’s Request to Withdraw Avacopan? February 23, 2026.

Medscape alert coverage: Medscape. FDA Issues Alert on Liver Injuries Linked to Vasculitis Drug, Following Withdrawal Request. March 31, 2026.

Amgen refusal: Pharmaphorum. Amgen baulks at FDA request to withdraw Tavneos. February 2026.

Public Citizen statement: Public Citizen. FDA’s Avacopan Alert Raises More Questions Than It Answers. March 31, 2026.

EMA review: European Medicines Agency. EMA starts review of Tavneos, a medicine for rare autoimmune diseases GPA and MPA. January 30, 2026.

Vasculitis Foundation patient update: Important Information for the Vasculitis Community Regarding TAVNEOS (avacopan). vasculitisfoundation.org. April 2026.

The Rheumatologist: Kaufman MB. Avacopan Under Scrutiny by FDA, EMA Due to Data Concerns. February 21, 2026. the-rheumatologist.org.

ACR vasculitis guidelines: American College of Rheumatology. Guidelines for the Management of Antineutrophil Cytoplasmic Antibody-Associated Vasculitis. rheumatology.org.

MedWatch adverse event reporting: FDA MedWatch Safety Reporting Portal. fda.gov/safety/medwatch.

Patient resources: Vasculitis Foundation | American College of Rheumatology | FDA MedWatch | NIAMS Vasculitis Information

Disclaimer: Health Evidence Digest provides general information about drug safety communications and health research for educational purposes. This content is not a substitute for professional medical advice, diagnosis, or treatment. Patients should not make changes to their medication without consulting their healthcare provider. All clinical decisions should account for individual patient circumstances, disease severity, and current treatment guidelines.

April 22, 2026

The FDA Just Removed Boxed Warnings From Six Hormone Therapy Products. Here Is What Changed, What Stayed, and Why It Matters.

📌 The essentials On February 12, 2026, the FDA approved drug labeling changes to six menopausal hormone therapy products, removing risk statements related to cardiovascular disease, breast cancer, and probable dementia from their boxed warnings. This is the first batch of approvals. At the FDA’s request, 29 drug companies have submitted proposed labeling changes, with additional approvals expected as submissions are reviewed. The six products affected: Bijuva (estradiol and progesterone), Divigel (estradiol gel), Cenestin (synthetic conjugated estrogens A), Enjuvia (synthetic conjugated estrogens B), Prometrium (progesterone), and Estring (estradiol vaginal system). What was NOT removed: the boxed warning for endometrial cancer remains on systemic estrogen-alone products. This is a clinically important retained warning that all prescribers and patients should be aware of. Context: the FDA initiated this label review in November 2025 following a comprehensive scientific literature review, an expert advisory panel in July 2025, and a public comment period. The 29-company label update process is ongoing.

Menopausal hormone therapy has had a complicated two decades in American medicine. In the early 2000s, the Women’s Health Initiative (WHI), a landmark landmark trial, reported findings that were widely interpreted as showing hormone therapy increased risks of breast cancer, heart disease, stroke, and blood clots. The FDA added boxed warnings to hormone therapy products reflecting those risks. Prescribing dropped dramatically, and it never fully recovered.

The scientific community has spent years re-examining those findings. The primary methodological critique has been consistent: the average age of WHI participants was 63, more than a decade past the average age of menopause onset. The formulation used (conjugated equine estrogen plus medroxyprogesterone acetate, a synthetic progestin) is not representative of current prescribing, which increasingly favors body-identical hormones. And subsequent analyses specifically in younger women who started HRT closer to menopause onset showed a substantially different risk profile.

On February 12, 2026, the FDA acted on that updated evidence. It approved the removal of cardiovascular disease, breast cancer, and probable dementia risk statements from the boxed warnings of six hormone therapy products. The action is the first step of a broader process affecting 29 drug products.

What a Boxed Warning Is and Why Its Removal Matters

A boxed warning is the FDA’s most prominent safety communication on a drug label. It appears inside a bold black border at the top of the prescribing information and is reserved for serious or life-threatening risks. Its presence signals to prescribers that the drug carries risks serious enough to warrant special consideration and monitoring.

For many prescribers and patients, a boxed warning is functionally a deterrent. The presence of a cancer warning or heart disease warning on hormone therapy labels has contributed to underutilization for more than two decades. The utilization gap is stark: in 2020, approximately 41 million U.S. women were between ages 45 and 64, yet only about 2 million women ages 46 to 65 received a hormone therapy prescription. That is roughly 1 in 20.

Removing those specific risk statements from the boxed warning reflects the FDA’s conclusion, after comprehensive scientific review, that the evidence no longer supports placing cardiovascular disease, breast cancer, and probable dementia in the most prominent warning position on the label.

What Was Removed and What Was Retained

This distinction is clinically critical and was largely absent from most coverage of this announcement.

Removed from boxed warnings (for the six approved products):

Risk statements related to cardiovascular disease
Risk statements related to breast cancer
Risk statements related to probable dementia

Retained and not changed:

The boxed warning for endometrial cancer remains on systemic estrogen-alone products. This warning is not being removed because it reflects a well-established, causal biological relationship: unopposed estrogen in women with an intact uterus increases the risk of endometrial hyperplasia and endometrial cancer. This is why women with an intact uterus who use systemic estrogen-alone therapy require concurrent progestogen. This warning is not a legacy artifact of the WHI. It reflects established reproductive endocrinology and has not been reconsidered in this process.
Contraindications remain in the prescribing information for all updated products, including active arterial thromboembolic disease (stroke, myocardial infarction), known or suspected breast cancer, known or suspected estrogen-dependent neoplasia, undiagnosed abnormal uterine bleeding, active liver dysfunction, and known hypersensitivity.

Why the endometrial cancer warning staying matters Estrogen stimulates growth of the uterine lining. Without progesterone or progestogen to counteract that stimulation, prolonged unopposed estrogen exposure in women with a uterus can cause endometrial hyperplasia and increase the risk of endometrial cancer. This is not a disputed risk. It is a well-characterized biological mechanism that has been known since the 1970s. The current label update specifically preserves this warning for systemic estrogen-alone products (such as estradiol patches, oral estrogens, and estradiol gels like Divigel). Women with an intact uterus who use systemic estrogen therapy must be prescribed a progestogen alongside it. This requirement does not change with the February 2026 label updates.

The Scientific Evidence Behind the Label Changes

The FDA’s decision followed a systematic process that included an expert advisory panel in July 2025, a comprehensive literature review, and a public comment period. The relevant evidence base is substantial and has been building for years.

The WHI reanalysis and the timing hypothesis

The Women’s Health Initiative enrolled women with an average age of 63. When researchers reanalyzed the WHI data stratifying by age at enrollment and time since menopause, a different picture emerged. Women who started HRT within 10 years of menopause onset or before age 60 showed different outcomes than those who started more than 10 years post-menopause. This is sometimes called the “timing hypothesis” or “window of opportunity.”

The Kronos Early Estrogen Prevention Study (KEEPS) enrolled women aged 42 to 58 within 36 months of their final menstrual period and found that low-dose oral or transdermal estradiol started early in the menopausal transition did not increase cardiovascular risk over four years.

The Early versus Late Intervention Trial with Estradiol (ELITE) randomized women to estradiol or placebo based on time since menopause (fewer than 6 years, or more than 10 years). It found that estradiol slowed progression of subclinical atherosclerosis in women who started within 6 years of menopause but not in those who started later, directly supporting the timing hypothesis.

The Menopause Society’s 2022 position statement concluded that the benefits of hormone therapy outweigh the risks for most healthy symptomatic women who begin treatment under age 60 or within 10 years of menopause, and that the WHI data should not be applied indiscriminately to all hormone therapy users.

The FDA’s February 2026 label update represents the regulatory translation of this accumulated body of evidence.

What the updated labels say about timing

The FDA’s updated labeling recommendation is to start systemic HRT within 10 years of menopause onset or before age 60. This is not an absolute contraindication for later initiation, but it is a clinical guidance point that the label will now include explicitly.

The benefit claims: what the evidence actually supports

The HHS fact sheet accompanying this announcement cited risk reductions including cardiovascular diseases by up to 50%, Alzheimer’s disease by 35%, and bone fractures by 50 to 60%. These figures come from observational analyses and secondary endpoint analyses from the relevant literature. They represent the upper estimates of potential benefit from studies of women who initiated HRT at younger ages and are not uniformly established as causal effect sizes across all formulations and populations. The fracture reduction data is the most consistently supported. The cardiovascular benefit in the timing window is biologically plausible and supported by the evidence above. The Alzheimer’s protection data is promising but still the subject of ongoing research. The appropriate interpretation of these figures is that they reflect the potential magnitude of benefit in women who use HRT appropriately, not guaranteed outcomes for every individual patient.

Which Products Are Affected and Their Updated Labels

The six products in the first batch of approved labeling changes represent all four major categories of menopausal HRT:

Product	Category	Active ingredient	Updated label
Bijuva	Systemic combination (estrogen plus progestogen)	Estradiol and progesterone	Link
Divigel	Systemic estrogen-alone (topical gel)	Estradiol	Link
Cenestin	Systemic estrogen-alone (oral)	Synthetic conjugated estrogens A	Link
Enjuvia	Systemic estrogen-alone (oral)	Synthetic conjugated estrogens B	Link
Prometrium	Systemic progestogen-alone	Progesterone (micronized)	Link
Estring	Topical vaginal estrogen (ring)	Estradiol vaginal system	Link

The remaining 23 drug companies that submitted proposed labeling changes are still in the review process. Their products will be updated as submissions are approved. The FDA’s menopausal hormone therapies updated prescribing information page will track additional products as they are approved.

What This Means for Women and Clinicians

For women currently using or considering HRT

The removal of boxed warnings for cardiovascular disease, breast cancer, and probable dementia does not mean these risks are zero. It means the FDA has concluded that the evidence does not support placing them in the most prominent risk-warning position for the indicated population (women using HRT appropriately, typically within 10 years of menopause onset and before age 60).

HRT remains an individualized decision that should account for personal risk factors including:

Personal or family history of breast cancer
History of blood clots, stroke, or cardiovascular disease
Liver disease
Undiagnosed vaginal bleeding
Time since menopause and age at initiation

The Menopause Society’s practitioner finder can help identify clinicians with menopause medicine certification who are equipped to discuss the updated evidence and individual risk profiles.

For clinicians

The label changes affect prescribing information and patient counseling conversations. Women who have avoided HRT because of the boxed warnings may now re-engage with the question, and clinicians should be prepared to have evidence-based conversations about individualized benefit-risk assessment. The retained endometrial cancer warning for estrogen-alone products, the continued contraindications, and the timing guidance are all still present in the label and remain clinically operative.

What has not changed

HRT still requires a prescription. Contraindications still apply. The timing guidance (within 10 years of menopause onset or before age 60 for systemic HRT) is a recommendation, not an absolute cutoff. Women with an intact uterus still need progestogen with systemic estrogen. The conversations between clinician and patient remain the appropriate framework for HRT decisions.

For related coverage on how this label change fits into the broader reassessment of hormone therapy in women’s health, see our posts on vaginal estrogen therapy safety in endometrial cancer survivors, hormone therapy underuse in women with premature ovarian insufficiency, and new 2026 cervical cancer screening guidelines allowing self-collection.

Sources

FDA press announcement: FDA Approves Labeling Changes to Menopausal Hormone Therapy Products. FDA.gov. February 12, 2026.

HHS announcement (November 2025): HHS Advances Women’s Health, Removes Misleading FDA Warnings on Hormone Replacement Therapy. FDA.gov. November 10, 2025.

HHS fact sheet: FACT SHEET: FDA Initiates Removal of “Black Box” Warnings from Menopausal Hormone Replacement Therapy Products. HHS.gov. November 10, 2025.

FDA updated prescribing information tracker: Menopausal Hormone Therapies with Updated Prescribing Information. FDA.gov.

Updated label: Bijuva: accessdata.fda.gov.

Updated label: Divigel: accessdata.fda.gov.

Updated label: Cenestin: accessdata.fda.gov.

Updated label: Enjuvia: accessdata.fda.gov.

Updated label: Prometrium: accessdata.fda.gov.

Updated label: Estring: accessdata.fda.gov.

Contemporary OB/GYN coverage: FDA updates labels on multiple menopausal hormone therapies. contemporaryobgyn.net. February 12, 2026.

Pharmacy Times coverage: FDA Approves Drug Labeling Changes to 6 Menopausal Hormone Therapy Products. pharmacytimes.com. 2026.

Prism News coverage: FDA narrows boxed warnings for six menopausal hormone therapies. prismnews.com. February 13, 2026.

KEEPS trial: Harman SM et al. KEEPS: The Kronos Early Estrogen Prevention Study. Climacteric. 2005. PMC3678904.

ELITE trial: Hodis HN et al. Vascular Effects of Early versus Late Postmenopausal Treatment with Estradiol. N Engl J Med. 2016.

Menopause Society 2022 position statement: The 2022 Menopause Society Position Statement on Hormone Therapy. menopause.org.

WHI background: Women’s Health Initiative. National Heart, Lung, and Blood Institute.

Menopause Society practitioner finder: Find a Menopause Healthcare Practitioner.

Disclaimer: Health Evidence Digest provides general information about FDA regulatory actions and health research for educational purposes. This content is not a substitute for professional medical advice, diagnosis, or treatment. Decisions about hormone therapy should be made in consultation with a qualified healthcare provider who can evaluate your individual health history, risk factors, and symptom burden.

April 14, 2026

The FDA Just Added a New Warning to All Carbidopa/Levodopa Parkinson’s Medications. Here Is What It Says and Why It Matters.

⚠️ Key Safety Summary On March 20, 2026, the FDA issued a Drug Safety Communication requiring new warnings on all carbidopa/levodopa-containing medications for Parkinson’s disease, addressing the risk of vitamin B6 (pyridoxine) deficiency and associated seizures. What the FDA found: 14 confirmed cases of seizures linked to vitamin B6 deficiency in patients on carbidopa/levodopa products, including 2 fatalities. All cases involved levodopa doses exceeding 1,000 mg daily. Latency periods ranged from 23 to 132 months, meaning the problem can develop years into treatment. The seizures are clinically distinctive: they do not respond to standard anti-seizure medications but resolve after vitamin B6 supplementation. In 9 of 9 patients who received B6 supplementation, seizures fully resolved. In some cases, progression to status epilepticus was observed before the diagnosis was made. What is now required: healthcare professionals must evaluate baseline vitamin B6 levels before starting treatment, monitor periodically during treatment, and supplement as needed. If you or someone you care for is on a carbidopa/levodopa medication: do not stop the medication without medical guidance. Report new symptoms (seizures, confusion, numbness, tingling) to your prescriber promptly.

Parkinson’s disease affects approximately 1 million Americans and is the second most common neurodegenerative condition in the United States. The cornerstone of pharmacological management is levodopa, which the brain converts to dopamine to compensate for the progressive loss of dopamine-producing neurons. Levodopa is almost always combined with carbidopa, which prevents levodopa from being metabolized in the bloodstream before it reaches the brain, allowing lower and more effective doses.

These medications work well. Many patients take them for years or decades. But a drug safety communication issued by the FDA on March 20, 2026 has established, for the first time with a formal labeling requirement, that the same mechanism that makes carbidopa/levodopa effective is also depleting vitamin B6, and that this depletion can, in some patients, cause seizures that clinicians may not immediately recognize as B6-related.

What the FDA Found: 14 Cases, 2 Deaths, and a Seizure Type That Resists Standard Treatment

The FDA’s safety review identified 14 cases of seizures linked to vitamin B6 deficiency in patients using carbidopa/levodopa products. Thirteen were postmarketing reports submitted to the FDA Adverse Event Reporting System (FAERS) database; one was identified in the medical literature.

Because the FDA notes that its review relied on spontaneous case reports and that similar cases may exist unreported, the 14 confirmed cases likely underestimate the true incidence.

Key characteristics of the 14 cases:

Feature	Detail
Total confirmed cases	14
All cases: levodopa dose	Above 1,000 mg daily in every case
Higher doses (above 1,500 mg)	Associated with shorter time to deficiency development
Latency period (treatment start to seizure onset)	23 to 132 months
Seizure type	Focal onset seizures with secondary generalization; some progressed to status epilepticus
Response to standard anti-seizure medications	Did not respond in the majority of cases
Response to vitamin B6 supplementation	9 of 9 treated patients had complete resolution
Fatalities	2, both with documented low vitamin B6 levels and poorly controlled seizures
Formulations involved	Oral tablets and enteral suspension
Cases involving Stalevo or Vyalev	None confirmed, but biological plausibility supports similar risk

Source: FDA Drug Safety Communication, March 20, 2026.

The status epilepticus finding is the most urgent clinical detail

The FDA communication notes that progression to status epilepticus was observed in some of the reported cases. Status epilepticus is a neurological emergency, defined as a seizure lasting more than five minutes or two or more seizures without full recovery of consciousness between them. It carries significant risk of brain injury and death if not treated rapidly. In the context of this safety communication, the critical clinical implication is that if a patient on carbidopa/levodopa presents with new-onset seizures that do not respond to conventional anti-seizure medications, vitamin B6 deficiency should be in the differential diagnosis immediately, not as a last resort.

The seizure pattern described is consistent with vitamin B6-dependent epilepsy: focal onset seizures with secondary generalization that resist standard anticonvulsants and resolve with pyridoxine. This pattern has a known biochemical mechanism.

The supporting evidence for B6 deficiency in these cases

Beyond the seizures themselves, the FDA review found additional laboratory and clinical evidence of vitamin B6 deficiency in the reviewed cases:

Elevated homocysteine levels in four cases (B6 is required for homocysteine metabolism; deficiency causes it to accumulate)
Microcytic or normocytic anemia in three cases (B6 is required for heme synthesis)
Neuropsychiatric symptoms in four cases

This multi-system evidence of B6 deficiency strengthens the FDA’s conclusion that “there is reasonable evidence of a causal association between drug products containing carbidopa/levodopa and vitamin B6 deficiency-associated seizures.”

Why Carbidopa/Levodopa Depletes Vitamin B6: The Mechanism

Understanding the biochemical basis of this safety signal helps explain both why it is real and why it has been underrecognized.

Vitamin B6 (pyridoxine) is a water-soluble vitamin with a critical role in over 100 enzymatic reactions in the body, particularly in amino acid metabolism and neurotransmitter synthesis. Its active form, pyridoxal-5′-phosphate (PLP), is the biologically active cofactor.

Carbidopa/levodopa depletes B6 through two mechanisms operating simultaneously:

First: The conversion of levodopa to dopamine (and its subsequent metabolism) consumes PLP as a cofactor. At higher levodopa doses, this ongoing metabolic consumption exceeds normal dietary intake and tissue stores.

Second: Carbidopa itself binds to and inactivates pyridoxal phosphate. This is not a side effect unique to therapeutic use. The structural interaction between carbidopa and PLP is well-characterized biochemically, and it was the reason that earlier levodopa formulations without carbidopa actually caused peripheral B6 depletion through a somewhat different pathway.

The combined effect is a functional reduction in available B6 that compounds over time at higher doses. The long latency period (23 to 132 months) reflects this gradual depletion dynamic: it takes time for dietary intake and tissue stores to be overwhelmed.

What Products Are Affected

The new warning requirement applies to all drug products containing carbidopa and levodopa. These include:

Brand name	Components	Formulation
Sinemet, Sinemet CR	Carbidopa/levodopa	Oral tablets (immediate and controlled release)
Rytary	Carbidopa/levodopa	Oral extended-release capsules
Dhivy	Carbidopa/levodopa	Oral tablets
Crexont	Carbidopa/levodopa	Oral extended-release capsules
Duopa	Carbidopa/levodopa	Enteral suspension (continuous infusion)
Stalevo	Carbidopa/levodopa/entacapone	Oral tablets
Vyalev	Foscarbidopa/foslevodopa	Subcutaneous infusion

No confirmed cases of B6-deficiency seizures were found with Stalevo or Vyalev, which the FDA attributes to lower usage patterns, more recent approval dates, and different dosing requirements. However, the FDA explicitly states that biological plausibility supports similar risk across all formulations, noting that vitamin B6 deficiency was observed in the clinical trials for the injectable product. The warning applies to all formulations.

For Patients and Caregivers

Do not stop your medication without speaking to your neurologist or prescriber first

Carbidopa/levodopa is essential for managing Parkinson’s symptoms. Stopping abruptly can cause serious problems including a Parkinson’s crisis (sudden, severe worsening of symptoms), and in rare cases, a potentially life-threatening condition called neuroleptic malignant-like syndrome from abrupt withdrawal. The solution to B6 depletion is monitoring and supplementation, not stopping the medication.

What to watch for and report to your doctor immediately

Contact your prescriber promptly if you experience any of the following:

New or unexplained seizures of any kind
Confusion or worsening cognitive symptoms
Depression that seems different from your baseline
Numbness, tingling, sharp pains, or weakness in the hands, feet, or limbs
Mouth sores, inflammation of the tongue or lips, or skin changes
Persistent fatigue

If a seizure occurs and does not respond to emergency anti-seizure treatment, inform emergency responders and hospital staff that you are taking carbidopa/levodopa and that vitamin B6 deficiency-associated seizures should be considered. The seizures described in the FDA cases did not respond to conventional anticonvulsants but resolved with B6 supplementation.

Ask your provider about monitoring

Based on the new FDA warning, your prescriber should now:

Check your vitamin B6 level before starting or continuing treatment (particularly if you have been on levodopa for a long time or are on high doses)
Recheck B6 levels periodically during treatment
Consider whether vitamin B6 supplementation is appropriate for your situation

Do not self-supplement with high-dose B6 without medical guidance. While standard dietary doses of B6 are safe, high-dose pyridoxine supplementation (above 50 to 100 mg per day for extended periods) carries its own risk of peripheral neuropathy and should be managed by your provider.

Resources for people living with Parkinson’s disease

The Parkinson’s Foundation and the Michael J. Fox Foundation for Parkinson’s Research both maintain current patient-facing information on Parkinson’s medications, including updates on drug safety communications. The FDA’s safety communication page is updated as new information becomes available.

For Healthcare Professionals

The FDA’s new labeling requirement translates to the following clinical actions:

When	Action
Before starting carbidopa/levodopa	Evaluate baseline vitamin B6 (pyridoxine) and PLP levels
Periodically during treatment	Recheck B6 levels, frequency guided by dose and clinical status
If new symptoms appear	Evaluate B6 levels promptly regardless of time on therapy
If seizures occur	Recognize that B6 deficiency-associated seizures do not respond to standard anticonvulsants; vitamin B6 administration is the treatment
When titrating to doses above 1,000 mg levodopa	Higher vigilance warranted; doses above 1,500 mg levodopa associated with shorter time to deficiency
If adding anti-seizure medications	Be aware that select anticonvulsants can worsen B6 deficiency

The seizure type is distinctive and should be recognized: focal onset with secondary generalization, consistent with pyridoxine-dependent epilepsy, unresponsive to conventional anticonvulsants, and resolving with B6 administration. Elevated homocysteine, microcytic or normocytic anemia, and neuropsychiatric symptoms are supporting laboratory findings that can help confirm the diagnosis when seizures occur in the context of carbidopa/levodopa use.

Reporting Adverse Events

Patients and healthcare professionals should report adverse events related to carbidopa/levodopa products to the FDA MedWatch program online at fda.gov/safety/medwatch or by calling 1-800-332-1088. Reporting contributes to the pharmacovigilance database that identifies safety signals like this one. The FDA’s 14-case dataset almost certainly undercounts the true incidence of this adverse effect; reporting by clinicians who encounter it helps the agency characterize the full scope.

For related coverage of drug safety communications and post-market safety monitoring on Health Evidence Digest, see our post on the Tavneos (avacopan) serious liver injury warning and what it means when the FDA and a manufacturer disagree about risk and our analysis of why the FDA required post-marketing studies for Foundayo after its accelerated approval.

Sources

FDA Drug Safety Communication (primary source): FDA Is Requiring Warning about Vitamin B6 Deficiency and Associated Seizures for Drug Products Containing Carbidopa/Levodopa. March 20, 2026. fda.gov.

FDA DSC PDF: Drug Safety Communication PDF. fda.gov/media/191605.

FAERS reporting system: Questions and Answers: FDA’s Adverse Event Reporting System. fda.gov.

Vitamin B6 (pyridoxine) overview: Vitamin B6: Fact Sheet for Health Professionals. NIH Office of Dietary Supplements.

Homocysteine and B6: Homocysteine. StatPearls. NCBI.

Status epilepticus: Status Epilepticus. StatPearls. NCBI.

Parkinson’s disease overview: Parkinson’s Disease. NINDS.

Levodopa withdrawal syndrome: Neuroleptic Malignant Syndrome and Parkinsonism-Hyperpyrexia. StatPearls. NCBI.

MedWatch reporting: FDA MedWatch. fda.gov/safety/medwatch.

Patient resources: Parkinson’s Foundation | Michael J. Fox Foundation

Disclaimer: Health Evidence Digest provides general information about FDA drug safety communications and health research for educational purposes. This content is not a substitute for professional medical advice. Patients taking carbidopa/levodopa-containing medications should not discontinue treatment without consulting their neurologist or prescribing clinician. Report new symptoms to your healthcare provider.

April 11, 2026

GLP-1 Drugs and Suicidal Thoughts: The FDA Reviewed 107,910 Patients and 2.2 Million Real-World Users. What It’s Now Asking Drug Makers to Remove.

📌 Where This Review Stands as of January 2026: Read This First January 13, 2026: The FDA issued an updated Drug Safety Communication concluding that its comprehensive review, spanning 91 clinical trials, 107,910 patients, and a real-world cohort of 2.24 million users, found no evidence of increased suicidal behavior or ideation with GLP-1 receptor agonists. The FDA has formally requested removal of suicidality warnings from the prescribing information for semaglutide, tirzepatide, and liraglutide. This is the FDA’s clearest signal that it considers the concern resolved. The European Medicines Agency reached a similar conclusion in April 2024 after its own independent review. Important nuance: the FDA’s update also examined anxiety, depression, irritability, and psychosis, and found no increased risk for these psychiatric adverse events either. Patients should not stop GLP-1 medications without consulting their healthcare provider. If you are experiencing mood changes or distressing thoughts, contact your clinician and call or text 988.

When a drug is used by tens of millions of people, any signal in the adverse event database draws attention, even when that signal does not establish causation. That is the story of how GLP-1 receptor agonists and suicidality became a major pharmacovigilance question, why it took more than two years and two regulatory agencies to work through the evidence, and why the answer that emerged from one of the largest drug safety analyses ever conducted matters for patients and clinicians navigating this question.

The conclusion: the FDA reviewed the evidence systematically and comprehensively. It found no causal link. It is now asking manufacturers to remove the suicidality warning language. Getting to that conclusion required understanding why the signal appeared, why it was worth investigating seriously, and why the initial reports did not mean what some feared.

How the Signal Emerged and Why It Was Worth Investigating

GLP-1 receptor agonists, the class that includes semaglutide (Ozempic, Wegovy), tirzepatide (Mounjaro, Zepbound), liraglutide (Victoza, Saxenda), dulaglutide (Trulicity), and others, have become among the most widely prescribed drugs in modern medicine. More than 40 million Americans have taken semaglutide alone. At that scale, rare adverse events generate measurable counts in pharmacovigilance databases even when those events occur at the same rate as in untreated patients.

The concern originated in Europe. In 2023, the European Medicines Agency received reports of suicidal thoughts and self-injury in patients taking semaglutide and liraglutide. The EMA initiated a formal safety review. The FDA, which had already added precautionary language to certain GLP-1 labels, also began a systematic evaluation. In January 2024, the FDA announced preliminary findings with no evidence of a causal link, but stated the review was not yet complete. The January 13, 2026 Drug Safety Communication represents the completed review.

Why the initial signal was plausible enough to investigate seriously Several biological mechanisms made a GLP-1/suicidality connection worth examining. GLP-1 receptors are expressed not just in the pancreas and gut but throughout the brain, including in regions involved in mood regulation, reward processing, and appetite. Semaglutide is known to reduce what is sometimes described as “food noise,” the intrusive, preoccupying thoughts about food. Whether this appetite-suppressing CNS effect could, in some patients, contribute to a broader hedonic blunting or mood change was a reasonable hypothesis to test. Additionally, rapid weight loss from any cause is associated with increased risk of mood disruption, nutritional deficiency, and in some cases a paradoxical worsening of psychological wellbeing when results do not match expectations. Patients who begin GLP-1 therapy with unrealistic expectations and experience slower-than-expected weight loss may be particularly vulnerable. An umbrella review published in Healthcare (MDPI) in 2025 specifically identified unrealistic treatment expectations as a factor associated with suicidal ideation in GLP-1 users, independently of any pharmacological mechanism. The precedent from lorcaserin (Belviq), a weight-loss drug withdrawn in 2020 due to cancer risk that had also carried a suicidality warning, created additional regulatory pressure to examine this class carefully. And context matters: obesity and type 2 diabetes both independently elevate the baseline risk of depression and suicidality, meaning any population study in GLP-1 users is working against a backdrop of elevated baseline risk.

What the FDA’s Review Actually Examined

The January 2026 Drug Safety Communication represents one of the most comprehensive pharmacovigilance analyses ever conducted for a drug class. The FDA’s review incorporated three distinct streams of evidence.

1. FAERS adverse event report review

The FDA conducted detailed reviews of suicidal thoughts and action reports in its Adverse Event Reporting System (FAERS) database. It found that the reports were limited in detail and that suicidal events could be explained by other factors, including the underlying conditions for which GLP-1s are prescribed. A peer-reviewed FAERS analysis by McIntyre et al. (2024), applying Bradford Hill criteria for causality assessment, found disproportionate reporting of suicidal ideation with semaglutide and liraglutide, but no disproportionate reporting of actual suicidal behavior, suicide attempts, or completed suicide for any GLP-1. Their conclusion: no causal link exists when confounders are considered.

2. Meta-analysis of 91 clinical trials (107,910 patients)

The FDA conducted a meta-analysis pooling data from 91 placebo-controlled clinical trials of GLP-1 medications: 60,338 patients treated with GLP-1s and 47,572 treated with placebo. This is among the largest drug safety meta-analyses ever assembled for a single question. The results did not demonstrate an increased risk of suicidal behavior or ideation with GLP-1s. The analysis also examined related psychiatric outcomes, including anxiety, depression, irritability, and psychosis, and found no increased risk for any of these either.

3. Sentinel System real-world cohort study (2.24 million users)

The FDA’s Sentinel System, a large database of health insurance claims and patient health records, was used to conduct a retrospective cohort study comparing the risk of intentional self-harm between new users of GLP-1 receptor agonists and new users of SGLT2 inhibitors. The study included 1,161,983 GLP-1 users and 1,081,155 SGLT2 users. This real-world analysis found no increased risk of intentional self-harm with GLP-1 use.

Evidence stream	Scale	Finding
FAERS adverse event review	Detailed case-by-case review	No clear causal relationship; limited report detail; events explainable by underlying conditions
Clinical trial meta-analysis	91 trials, 107,910 patients (60,338 GLP-1 vs 47,572 placebo)	No increased risk of suicidal ideation, behavior, anxiety, depression, irritability, or psychosis
Sentinel real-world cohort	2,243,138 users (1.16M GLP-1 vs 1.08M SGLT2 inhibitor)	No increased risk of intentional self-harm with GLP-1 use
EMA independent review (April 2024)	EU-based pharmacovigilance and clinical trial data	Evidence insufficient to support causal association; no label change warranted

Beyond the FDA: What Independent Research Shows

The regulatory conclusions are consistent with the direction of the independent peer-reviewed literature, which has grown substantially since 2023.

The Nature Medicine study (semaglutide vs. comparators, n=240,618)

A study published in Nature Medicine, led by researchers at Case Western Reserve University and the National Institute on Drug Abuse, used electronic health records from 240,618 patients with overweight or obesity prescribed semaglutide or a non-GLP-1 medication for weight management. Semaglutide was associated with a lower risk of incident suicidal ideation (HR 0.27) and recurrent suicidal ideation (HR 0.44) compared with other agents. Dr. Nora Volkow, MD, Director of the National Institute on Drug Abuse (NIDA), was explicit about the appropriate interpretation of this finding in the publication’s commentary: the study does not indicate that taking semaglutide could reduce the risk of suicidal ideation or that it actively protects against suicidal ideation, as such a conclusion would require clinical trials.

Mendelian randomization: no genetic link

A Mendelian randomization study by Nguyen (2024) used genetic variants associated with GLP-1 pathway activity as proxies for long-term drug-like exposure. This approach can detect causal biological effects independent of confounders. The study found no causal evidence linking the glycemic and BMI-lowering effects of GLP-1 receptor agonists to an increased risk of suicide attempts, either in the general population or in individuals with pre-existing mental health conditions.

Systematic reviews and meta-analyses

A 2025 systematic review in Diabetes/Metabolism Research and Reviews (Bushi et al.) pooled observational cohort and case-control studies. It found no consistent association between GLP-1 use and increased suicidal ideation or behavior in the general diabetes or obesity population. An umbrella review published in Healthcare (MDPI) in 2025 synthesized 12 reviews and meta-analyses and found results consistent across studies, with no significant association between GLP-1 use and suicidality. A meta-analysis of 31 RCTs involving approximately 85,000 patients found no significant difference in psychiatric disorders (OR 0.97; 95% CI 0.83 to 1.15) or suicidal behavior (OR 0.86) between GLP-1 and placebo arms.

The Critical Confounding Issue: Obesity and Diabetes Independently Elevate Suicide Risk

This point deserves its own section because it is the single most important piece of context for interpreting any report connecting GLP-1 drugs to suicidality.

Obesity is independently associated with elevated rates of depression, anxiety, and suicidal ideation. Type 2 diabetes is independently associated with the same. People prescribed GLP-1 receptor agonists are, by definition, drawn from a population with higher baseline risk of psychiatric adverse events than the general population. When a drug is given to millions of people from a higher-risk population, adverse events including suicidal ideation will be reported, even if the drug itself plays no causal role.

This is called confounding by indication: the condition being treated (obesity, diabetes) is itself a risk factor for the outcome being studied (suicidal ideation). Properly controlled studies, using comparator groups of patients with the same underlying conditions on different medications, or using the Mendelian randomization approach to remove confounders entirely, consistently fail to find excess suicidal risk attributable to GLP-1 drugs.

The co-prescription signal: a nuance worth knowing One finding from the Bushi et al. systematic review warrants specific clinical attention: patients co-prescribed antidepressants showed a substantially higher reporting odds ratio for suicidal ideation (ROR 4.45; 95% CI 2.52 to 7.86), and those co-prescribed benzodiazepines showed a similarly elevated signal (ROR 4.07; 95% CI 1.69 to 9.82). This suggests that individuals with pre-existing mental health conditions who are also taking GLP-1 drugs may warrant closer psychiatric monitoring. This does not mean GLP-1 drugs caused the suicidality in these patients. It almost certainly reflects the fact that patients already on antidepressants or benzodiazepines have higher baseline psychiatric vulnerability. But it is a reasonable clinical signal: if a patient is already being treated for depression or anxiety and starts a GLP-1, maintaining existing mental health follow-up is sensible practice regardless of the FDA’s label change.

What the FDA Is Asking Drug Makers to Remove and What Stays

The January 2026 Drug Safety Communication is explicit: the FDA is requesting removal of suicidal behavior and ideation warnings from the prescribing information for semaglutide, tirzepatide, and liraglutide. This is a formal regulatory request to manufacturers, not a unilateral FDA label change. Manufacturers submit the proposed language and the FDA approves it. The practical effect is the same.

What stays in the label: the established, class-wide safety profile for GLP-1s including gastrointestinal effects, thyroid C-cell tumor risk for injectable GLP-1s, pancreatitis risk, and the contraindications relevant to each specific drug. None of these are affected by the suicidality review.

The FDA also emphasized that despite requesting label removal, clinicians should continue to be prepared to discuss the findings with patients and to refer individuals who report suicidal ideation or behavior to mental health professionals, not because the drug is causing it, but because the population taking these drugs has higher baseline psychiatric risk that warrants clinical awareness.

For Patients: What This Means in Practice

Do not stop your GLP-1 medication based on the suicidality concern. The evidence does not support a causal link. Stopping abruptly without medical guidance can worsen blood sugar control or cause rebound weight gain. Read our post on what happens when semaglutide is discontinued for more context on managing GLP-1 therapy transitions.

Do tell your prescribing clinician about any changes in mood, new or worsening depression, or distressing thoughts. This applies to any medication in any patient and is not specific to GLP-1s, but it is always important to keep your care team informed.

If you already have a history of depression or anxiety, ensure your mental health provider is aware you are starting a GLP-1. Not because of a drug risk, but because any significant metabolic change, body image shift, or treatment expectation mismatch can affect mental health in people already managing psychiatric conditions.

If you are having thoughts of self-harm or suicide, call or text 988 (U.S. Suicide and Crisis Lifeline) any time, or go to your nearest emergency department. Do not wait for a scheduled appointment.

The label change does not mean GLP-1 drugs are risk-free. They have well-documented side effects including gastrointestinal effects, injection site reactions, and for some patients, hair loss. The specific concern about suicidality, however, has been investigated extensively and not confirmed.

For Clinicians: Practical Takeaways

The FDA’s label change request does not eliminate the clinical rationale for psychiatric monitoring in GLP-1 users. The population prescribed these drugs has elevated baseline psychiatric risk, and the co-prescription signal (antidepressants, benzodiazepines) warrants awareness even if it likely reflects underlying vulnerability rather than a drug effect.

Discuss realistic treatment expectations before initiating GLP-1 therapy. The umbrella review found that unrealistic expectations and disappointment with weight loss results were associated with elevated suicidal risk signals. Setting evidence-based expectations, specifically that 10 to 15% weight loss is typical and results vary, is now both good clinical practice and supported by the safety literature.

The review’s psychiatric adverse event finding is broader than suicidality. No increased risk of anxiety, depression, irritability, or psychosis was found across 91 trials. This is useful context for reassuring patients who ask about neuropsychiatric effects broadly.

For patients with a history of suicidality or active psychiatric illness who are candidates for GLP-1 therapy, consider co-management with their mental health provider, not because of drug risk but because of the baseline complexity.

For related HED coverage of GLP-1 medications and their evolving indications, see our posts on GLP-1 medications and PCOS fertility research in 2026, the approval of Wegovy HD (semaglutide 7.2 mg) and the STEP UP trial data, and the FDA’s post-marketing study requirements for Foundayo (orforglipron).

If you or someone you know is struggling with mental health right now, please reach out to the 988 Suicide and Crisis Lifeline by calling or texting 988, available 24 hours a day, 7 days a week in the U.S.

Sources

FDA Drug Safety Communication (Jan 2026): Update on FDA’s Ongoing Evaluation of Reports of Suicidal Thoughts or Actions in Patients Taking Certain Type 2 Diabetes and Obesity Medicines. January 13, 2026. fda.gov.

Pharmacy Times coverage: FDA Finds No Increased Suicide Risk With GLP-1 Medications, Requests Removal of Warning Labels. pharmacytimes.com. 2026.

EMA review conclusion: European Medicines Agency. Assessment report on GLP-1 receptor agonists and risk of suicidal behavior and ideation. April 2024.

McIntyre et al. FAERS analysis: McIntyre RS, Mansur RB, Rosenblat JD, Kwan ATH. Association between GLP-1 RAs and suicidality using Bradford Hill criteria. Expert Opin Drug Saf. 2024;23(1):47-55. doi:10.1080/14740338.2023.2295397

Nature Medicine cohort: Wang W, Volkow ND, Berger NA, et al. Association of semaglutide with risk of suicidal ideation in a real-world cohort. Nat Med. 2024. doi:10.1038/s41591-023-02750-1

Bushi et al. systematic review: Bushi G et al. Association of GLP-1 Receptor Agonists With Risk of Suicidal Ideation and Behaviour: A Systematic Review and Meta-Analysis. Diabetes Metab Res Rev. 2025. doi:10.1002/dmrr.70037

Umbrella review: Evaluating Suicidal Risk in GLP-1RA Therapy: An Umbrella Review of Meta-Analytic Evidence. Healthcare (MDPI). 2025;13(22):2958. doi:10.3390/healthcare13222958

ScienceDirect systematic review: The effect of GLP-1 receptor agonists on measures of suicidality: A systematic review. ScienceDirect. 2025. doi:10.1016/j.jad.2025.01.007

Lancet eClinical case-time-control: Suicide and suicide attempt in users of GLP-1 receptor agonists: a nationwide case-time-control study. EClinicalMedicine. 2024. doi:10.1016/j.eclinm.2024.102983

Mendelian randomization: Nguyen et al. Causal assessment of GLP-1 RA effects and suicide attempts via Mendelian randomization. 2024.

FDA Sentinel System: FDA’s Sentinel Initiative. fda.gov/safety/fdas-sentinel-initiative.