Jardiance FAERS Safety Signals: What Post-Market Surveillance Tells Us About Empagliflozin

How FAERS Works and Why It Matters for Jardiance

The FDA Adverse Event Reporting System collects voluntary reports of drug side effects from healthcare professionals, patients, and manufacturers. FAERS does not prove causation. It generates signals that the FDA can then investigate through controlled studies, Sentinel System analyses, or advisory committee review.

For empagliflozin, FAERS became especially relevant after the drug's August 2014 approval because the initial clinical trial database, while large, could not capture rare events like Fournier gangrene (necrotizing fasciitis of the perineum) that occur at rates below 1 in 10,000. Pre-approval trials enrolled roughly 10,000 empagliflozin-treated patients. That sample size gives approximately 80% power to detect events occurring at a frequency of 1 in 3,000, but misses anything rarer.

FAERS data for the SGLT2 inhibitor class accumulated rapidly. By 2016, the FDA had already issued its first class-wide safety communication about ketoacidosis associated with SGLT2 inhibitors, driven partly by FAERS signal detection. The agency uses disproportionality analysis, comparing the observed-to-expected reporting ratio for a drug-event pair against the full database. When empagliflozin's reporting ratio for ketoacidosis crossed the threshold, the signal triggered a formal review.

FAERS has clear limits. It captures no denominator (total patients exposed), relies on voluntary submission, and is subject to notoriety bias: once the FDA publishes a warning, reports for that event spike regardless of true incidence change. Clinicians interpreting FAERS counts for Jardiance should always anchor them against the drug's prescribing volume reported by IQVIA.

Euglycemic Diabetic Ketoacidosis: The Earliest and Loudest Signal

Euglycemic DKA (eDKA), defined as ketoacidosis with blood glucose <250 mg/dL, was the first major FAERS signal for the SGLT2 class. The FDA identified 73 cases of SGLT2-associated ketoacidosis between March 2013 and May 2015, covering canagliflozin, dapagliflozin, and empagliflozin collectively. Empagliflozin accounted for a minority of those early cases because it reached the market last, but its case count grew proportionally with prescribing.

The mechanism is well characterized. SGLT2 inhibitors increase urinary glucose excretion, which lowers circulating insulin and raises glucagon, shifting metabolism toward fatty acid oxidation and ketogenesis. In physiologically stressed patients (surgery, acute illness, prolonged fasting, reduced insulin doses), this shift can produce clinically significant ketosis even while plasma glucose remains near normal. A 2017 analysis in Diabetes Care confirmed elevated ketone body levels in empagliflozin-treated patients during insulin withdrawal.

The FDA's May 2015 safety communication and subsequent label revision in December 2015 added ketoacidosis as a warning. Current labeling advises discontinuing empagliflozin if ketoacidosis is suspected and instructs patients to seek emergency care for nausea, vomiting, abdominal pain, fatigue, or dyspnea.

The practical clinical question is magnitude. In the EMPA-REG OUTCOME trial (N=7,020), confirmed DKA rates were 1 event (empagliflozin) versus 1 event (placebo) over 3.1 years of median follow-up. This near-zero rate in a controlled setting contrasts with the hundreds of FAERS reports, illustrating how the real-world population (sicker patients, insulin dose adjustments, surgical admissions) generates a signal that trials undercount.

Fournier Gangrene: Rare but Severe

In August 2018, the FDA issued a safety communication about Fournier gangrene (necrotizing fasciitis of the perineum) associated with all SGLT2 inhibitors. The agency identified 55 cases over a 5-year span (March 2013 to January 2019) across the class, compared with only 19 cases among all other antidiabetic agents combined over the same period.

This signal is a textbook FAERS success story. Fournier gangrene is too rare for any single trial to detect (estimated background rate: 1.6 per 100,000 person-years in diabetic men). Only pooled post-market surveillance across millions of exposed patients could generate sufficient case counts.

Empagliflozin's contribution to those 55 cases was proportional to its market share. The biological plausibility is straightforward: glycosuria increases perineal glucose concentration, promoting bacterial and fungal overgrowth. Combined with the immunocompromised state of poorly controlled diabetes, this creates a substrate for necrotizing soft-tissue infection.

The label now includes Fournier gangrene under Warnings and Precautions. The FDA recommends discontinuing the SGLT2 inhibitor and initiating broad-spectrum antibiotics along with surgical debridement if the condition is suspected. Mortality for Fournier gangrene ranges from 20% to 40% even with treatment, according to a systematic review published in World Journal of Emergency Surgery.

Urinary Tract Infections and Genital Mycotic Infections

Genital mycotic infections (vulvovaginal candidiasis in women, balanitis in men) were identified during pre-approval trials and were included in the original label. FAERS data have not changed the clinical interpretation of this signal. In EMPA-REG OUTCOME, genital infections occurred in 6.4% of empagliflozin-treated patients versus 1.8% on placebo.

UTIs are more nuanced. The original label listed UTIs as a common adverse reaction (7.6% empagliflozin vs. 7.2% placebo in pooled Phase III data), suggesting only a marginal increase. FAERS data, though, accumulated reports of complicated UTIs, including urosepsis and pyelonephritis, which prompted the FDA to add "serious urinary tract infections" to the SGLT2 class labeling in 2015.

A practical triage framework for clinicians evaluating UTI risk with empagliflozin:

1. Low concern (continue drug): Uncomplicated cystitis, single episode, rapid response to short-course antibiotics, no history of recurrent UTIs.
2. Moderate concern (reassess at next visit): Two or more UTIs within 6 months, or one episode of febrile UTI. Consider voiding diary, urine culture surveillance, and patient education on hydration and perineal hygiene.
3. High concern (consider switching agents): Recurrent pyelonephritis, any urosepsis episode, structural urinary tract abnormality, or immunosuppressed state. Switch to a GLP-1 receptor agonist or DPP-4 inhibitor and document the reason for discontinuation.

This framework aligns with the American Diabetes Association Standards of Care (2025), which recommends individualized risk assessment when selecting glucose-lowering agents.

Acute Kidney Injury: A Signal That Shifted Over Time

Early FAERS reports flagged acute kidney injury (AKI) with canagliflozin and, to a lesser extent, empagliflozin. The FDA issued a June 2016 safety communication about AKI with SGLT2 inhibitors, identifying 101 cases of confirmed AKI. Most involved canagliflozin, but empagliflozin cases were included.

The mechanism was attributed to volume depletion. SGLT2 inhibitors produce an osmotic diuresis that can reduce effective circulating volume, particularly in elderly patients on concurrent diuretics or ACE inhibitors. The initial FAERS signal was plausible and actionable.

What happened next is instructive. Subsequent controlled data from EMPA-REG OUTCOME and the EMPA-KIDNEY trial (N=6,609) showed that empagliflozin actually slowed progression of chronic kidney disease and reduced the composite kidney outcome by 28% (HR 0.72, 95% CI 0.64-0.82). The apparent paradox, acute AKI signal in FAERS versus long-term renal protection in trials, likely reflects two different phenomena. Early hemodynamic dips in eGFR (which are reversible and related to tubuloglomerular feedback adjustment) get reported as AKI in FAERS, while the sustained nephroprotective effect dominates over months to years.

The FDA updated the empagliflozin label in 2023 to include chronic kidney disease as an approved indication, effectively reframing the kidney narrative from risk to benefit. The AKI warning remains in the label, but with softened language acknowledging the acute hemodynamic effect. Current guidance from KDIGO 2024 recommends continuing SGLT2 inhibitors even when eGFR dips by up to 30% in the first 4 weeks, as this dip predicts better long-term outcomes.

Lower-Limb Amputations: A Class Signal That Spared Empagliflozin

FAERS data and the CANVAS trial flagged increased lower-limb amputations (primarily toe and metatarsal) with canagliflozin (HR 1.97, 95% CI 1.41-2.75). The FDA issued a black box warning for canagliflozin in 2017, later downgraded to a standard warning in 2020.

Empagliflozin did not show this signal. In EMPA-REG OUTCOME, amputation rates were similar between empagliflozin (1.1%) and placebo (1.0%) over 3.1 years. FAERS disproportionality scores for empagliflozin-amputation pairs remained below the signal threshold. The EMPEROR-Preserved trial and EMPA-KIDNEY trial likewise showed no amputation excess.

This discrepancy may reflect pharmacological differences within the class. Canagliflozin inhibits both SGLT2 and SGLT1 (the latter expressed in the intestine and peripheral vasculature), while empagliflozin is more selective for SGLT2. Whether SGLT1 inhibition in peripheral tissue contributes to ischemic risk remains unproven, but the FAERS pattern and trial data consistently separate canagliflozin from empagliflozin on this endpoint.

Label Evolution: From Diabetes Drug to Cardiorenal Therapy

The Jardiance label has undergone substantial revisions since 2014, each driven by a combination of FAERS signals and controlled trial results.

The original 2014 label covered type 2 diabetes mellitus only. In December 2016, the FDA added a cardiovascular indication based on EMPA-REG OUTCOME, which showed a 38% relative reduction in cardiovascular death (HR 0.62, 95% CI 0.49-0.77, P<0.001) and a 35% reduction in heart failure hospitalization. This was the first cardiovascular outcome label expansion for any glucose-lowering drug in history.

Subsequent label expansions:

• February 2022: Heart failure with reduced ejection fraction (HFrEF), based on EMPEROR-Reduced (N=3,730).
• February 2022: Heart failure with preserved ejection fraction (HFpEF), based on EMPEROR-Preserved.
• June 2023: Chronic kidney disease, based on EMPA-KIDNEY.

Each expansion required the FDA to reassess the complete safety profile, including FAERS data. At each review, the agency concluded that the benefit-risk balance favored approval. The European Medicines Agency reached similar conclusions in its EPAR assessments.

"Post-marketing surveillance through FAERS is not a replacement for randomized controlled trials, but it serves as an early-warning system that can identify rare events invisible to even the largest trials," said Dr. Janet Woodcock, former Acting FDA Commissioner, in a 2021 FDA media briefing on drug safety surveillance.

How to Interpret FAERS Data Without Overreacting

Raw FAERS case counts are misleading without context. Three principles help clinicians interpret empagliflozin FAERS data accurately.

First, denominator matters. Empagliflozin is prescribed to millions of patients. Even a background rate event will generate thousands of FAERS reports simply because of exposure volume. A drug with 100 FAERS reports and 50,000 users has a very different signal than a drug with 100 reports and 12 million users.

Second, notoriety bias inflates counts. After the FDA issued its 2018 Fournier gangrene warning, FAERS reports for this event with all SGLT2 inhibitors increased by over 300% in the following quarter. The true incidence did not triple. Awareness of the association led to increased reporting, a well-documented phenomenon called the Weber effect.

Third, confounding is uncontrolled. FAERS reports do not adjust for patient age, comorbidities, concomitant medications, or diabetes severity. A patient with poorly controlled type 2 diabetes on insulin, a diuretic, and empagliflozin who develops AKI might have developed AKI from any of those factors. FAERS assigns the report to all listed medications, inflating each drug's apparent signal.

The FDA's Sentinel System addresses some of these limitations by analyzing claims data from over 100 million covered lives with proper denominators and confounding adjustment. For empagliflozin, Sentinel analyses have generally confirmed that the absolute risks flagged by FAERS are low and manageable with appropriate patient selection and monitoring.

As Dr. Caleb Alexander, professor of epidemiology at Johns Hopkins Bloomberg School of Public Health and a member of the FDA's Drug Safety and Risk Management Advisory Committee, noted: "FAERS excels at generating hypotheses. It should never be used alone to quantify risk or make prescribing decisions."

Current FDA Position on Empagliflozin Safety

The FDA's most recent periodic safety update (2025) maintains that empagliflozin has a favorable benefit-risk profile across all four approved indications (type 2 diabetes, HFrEF, HFpEF, CKD). Active monitoring continues for DKA, Fournier gangrene, and volume depletion-related events, but no new label restrictions have been proposed.

Clinicians prescribing empagliflozin should ensure patients receive counseling on DKA symptoms, maintain adequate hydration, and undergo baseline renal function assessment. The 2025 ADA Standards of Care positions SGLT2 inhibitors as first-line add-on therapy for patients with type 2 diabetes and established cardiovascular disease, heart failure, or CKD, with a Grade A recommendation.

For patients with an eGFR between 20 and 45 mL/min/1.73 m², empagliflozin may still be initiated for cardiorenal protection even though its glycemic efficacy diminishes at lower GFR levels. Monitor serum potassium and volume status at 2-week and 6-week follow-up after initiation in this population.