Deoxycholic Acid Injection (Kybella): Development, Safety Evolution and Pharmacovigilance

Deoxycholic acid injection illustrates how a familiar endogenous molecule can acquire a new, highly local pharmacological purpose. Deoxycholic acid is a secondary bile acid; the product-development story is not the discovery of a new molecule, but the creation and clinical testing of a standardised injectable formulation intended to destroy adipocytes in a defined area. The investigational product was called ATX-101 and became Kybella in the United States.

The narrowness of the authorised use matters. The US product is indicated for adults with moderate-to-severe convexity or fullness associated with submental fat. It is not approved for subcutaneous fat elsewhere, and the current US label says such use is not recommended. Injection outside the studied area exposes the product to different anatomy and is not supported by the efficacy and safety evidence used for approval.[1,2]

The safety story has also evolved. The 2015 approval review focused on local reactions and procedural risks, especially marginal mandibular nerve injury, dysphagia and skin ulceration. In September 2026, FDA approved new US label language after reviewing spontaneous reports and medical literature: serious adverse reactions have been reported after use outside the submental region, and injection-site nodules or masses may persist and require medical intervention. The nodule/mass reports include use both within and outside the approved area.[1–3] These findings are important for surveillance, but report counts do not establish incidence or prove that every reported event was caused by the drug.

Scope, Product Identity and Regulatory Context

This article follows the deoxycholic acid injection programme from early development to the current US safety position. It focuses on product history, clinical evidence, regulatory and commercial transitions, and practical pharmacovigilance. It is not a guide to performing cosmetic injections or a substitute for current local prescribing information.

Kybella is a 10 mg/mL solution for subcutaneous use. FDA describes deoxycholic acid as a cytolytic agent: when injected into submental fat under the labelled conditions, it physically disrupts cell membranes and causes cell lysis. The label states that the safety and effectiveness of use outside the submental region have not been established and that such use is not recommended.[2] FDA also reports that a generic deoxycholic acid injection is now approved for the same submental-fat indication; the brand-specific label and case identity should still be distinguished from reports involving a generic or an unapproved product.[1]

The regulatory statements below principally describe the United States. They should not be automatically applied to another jurisdiction's authorisation, product information, reporting rules or medical-practice framework. For global safety systems, the relevant local reference safety information (RSI), marketing authorisation and current regulatory communications must be checked separately.

From Endogenous Bile Acid to a Local Injectable

Deoxycholic acid is part of human bile-acid physiology. The developmental idea behind ATX-101 was to exploit its cell-membrane-disrupting property at a controlled injection site. The product was developed as a proprietary formulation of sodium deoxycholate, rather than as a new molecular entity discovered for this indication. FDA's 2015 quality review documents an important manufacturing transition during development: the drug substance was initially purified from bovine and ovine bile, after which the applicant developed a chemical synthesis process for free deoxycholic acid; FDA found the materials structurally identical in solution and considered the submitted quality specification adequate. This history is relevant to product provenance, but should not be used to infer the current API supplier or manufacturing site.[4,5]

This distinction is important. The therapeutic concept was a localized cytolytic treatment for a small, accessible fat deposit, with treatment effect depending on dose, injection depth, spacing and anatomical placement. It was not a systemic weight-management treatment. The biological activity that makes local adipocyte destruction possible also explains why misplaced or overly superficial injections can damage non-adipose tissue and why proximity to nerves, salivary glands, lymph nodes, muscles and vessels is clinically important.[2,4]

The US investigational new drug application for ATX-101 was opened in December 2007. Early clinical work evaluated dose, concentration, injection patterns, pharmacokinetics and local tissue response. In a small histology study, injection into abdominal fat produced adipocyte destruction followed by a local immune response and thickening of fibrous septae, findings consistent with the proposed local mechanism. These early observations supported further development; they were not proof that every anatomical site would have the same safety or cosmetic outcome.[4]

Development and Commercial Rights

Kythera was founded in 2004 as a biotechnology company focused on prescription products for aesthetic medicine. Its 2007–2008 phase IIa studies explored submental injections and helped define patient and clinician assessment scales. Bayer later collaborated on development and commercialisation outside the United States and Canada. Kythera's 2012 public filing described Bayer's European phase III programme alongside Kythera's North American studies, showing that development was initially distributed across the companies and regions.[4]

In March 2014, Kythera reacquired the rights to develop and commercialise ATX-101 outside the United States and Canada from Bayer for reported consideration of $84 million, comprising stock and a promissory note, with additional potential sales milestones. The transaction consolidated global rights with Kythera ahead of launch and regulatory submissions.[5] This history helps explain why early European trials and later US approval materials may refer to different sponsors, trial cohorts and regulatory pathways even though they concern the same investigational product.

The business history then changed quickly. FDA approved Kybella on 29 April 2015, and US commercial sales began in June. In June 2015 Allergan agreed to acquire Kythera in a transaction announced at approximately $2.1 billion; the acquisition closed on 1 October 2015. AbbVie completed its acquisition of Allergan in May 2020, bringing Allergan's aesthetics business into AbbVie. Kybella is currently presented within the Allergan Aesthetics portfolio, an AbbVie company.[6–8,11,12,14] These transactions explain the change in corporate stewardship; they do not alter the product's approved indication or establish a new safety conclusion.

Clinical Development: From Dose Finding to Pivotal Trials

The first development question was whether a locally injected bile acid could reduce submental fat in a measurable and acceptable way. Phase I work examined single-dose safety, concentration, dose escalation, systemic exposure and injection location. In the FDA review, mean plasma deoxycholic acid rose shortly after treatment and returned to endogenous baseline on average by 24 hours. This supported a predominantly local treatment concept, but brief systemic exposure does not mean that local tissue injury is impossible or that all uses outside the studied region are safe.[4]

Three randomised, double-blind, placebo-controlled phase II studies involving 284 participants explored dose and treatment regimens. They also helped validate clinician-rated and patient-rated submental-fat scales. The 2 mg/cm² dose generally performed better than the lower dose in the phase IIb study and informed the pivotal programme. The company filings report that the phase II studies were generally characterised by transient local injection-site reactions, particularly pain, swelling, bruising, numbness, redness and induration.[4]

The pivotal programme included European phase III trials and two North American trials known as REFINE-1 and REFINE-2. Across the two North American trials, 1,019 participants were included in the main safety analysis: 513 received deoxycholic acid and 506 received placebo. Participants could receive up to six treatment sessions at least four weeks apart. The primary outcomes combined clinician and patient ratings, with imaging-based measures included in subsets. These were aesthetic outcomes in selected adults with moderate-to-severe submental fullness, not outcomes for treatment of obesity, generalised adiposity or fat deposits elsewhere.[2,4,9,10]

In the two pivotal US/Canadian studies, a one-grade improvement on both clinician and patient scales was achieved by approximately 70% and 66.5% of treated participants in the respective trials, compared with 18.6% and 22.2% receiving placebo. A two-grade improvement occurred in 13.4% and 18.6% versus less than 0.1% and 3.0%, respectively. The trials therefore demonstrated a treatment effect for the studied submental indication, while also showing that many participants did not achieve the larger two-grade response.[2,4] The results should be interpreted in the context of the selected trial population, repeated treatments and local procedure-related effects.

Approval and the Initial Benefit–Risk Position

FDA approved the new drug application for Kybella on 29 April 2015 for improvement in the appearance of moderate-to-severe convexity or fullness associated with submental fat in adults. FDA's multidisciplinary summary review recommended approval for that specific indication. A Dermatologic and Ophthalmic Drugs Advisory Committee discussed the evidence in March 2015 and unanimously considered the benefit–risk evidence supportive of approval. FDA concluded that a Risk Evaluation and Mitigation Strategy was not necessary at that time and relied on prescribing information, patient information and pharmacovigilance, with particular attention to marginal mandibular nerve injury.[4,6]

The original benefit–risk assessment did not mean that treatment was free of meaningful burden. Local injection reactions were common. In pooled pivotal trials, injection-site swelling occurred in 87% of treated subjects, bruising in 72%, pain in 70%, numbness in 66%, and induration in 23%. These events were more frequent than with placebo for several reactions and could persist beyond 30 days. About 7% of treated participants discontinued because of adverse events, compared with 1% in the placebo group. The label also describes marginal mandibular nerve paresis, dysphagia and ulceration/necrosis as important risks.[2,4]

The pivotal programme was too small and too short to characterise every rare, delayed or use-dependent complication that might arise after broad commercial use. FDA's 2015 review recorded five deaths during the clinical trials, none considered related to ATX-101. It also noted a recovered serious mandibular nerve injury and discussed the value of injection instructions and post-marketing monitoring. A later serious event report should not be dismissed because it was not observed in the trials; nor should a post-marketing report be treated as proof of causation without case-level assessment.[4]

Safety Profile and How It Has Evolved

The current US label includes several risks with distinct mechanisms and management implications. Marginal mandibular nerve injury can present with an asymmetric smile or facial muscle weakness. Dysphagia may occur, and the label advises avoiding use in patients with pre-existing dysphagia because it may worsen. Bruising/haematoma deserves attention in patients with bleeding abnormalities or antiplatelet/anticoagulant treatment. Superficial dermal injection can cause ulceration and necrosis. Tissue damage and vascular injury can occur when injections are placed near vulnerable structures or inadvertently into a vessel.[2]

Infection, alopecia and scarring are also included among reported injection-site conditions. Nodules are defined in the revised label as 3 cm or smaller and masses as larger than 3 cm. Some may not resolve and may require medical intervention. The label directs discontinuation if an injection-site ulceration, necrosis, nodule, mass or infection develops after administration. For pharmacovigilance, this is a product-information instruction for healthcare providers; case processing should preserve the reporter's original term and capture clinical description, dimensions, onset, duration, treatment and outcome rather than replacing the report with a database term alone.[2]

The revised label therefore adds specificity to the risk profile rather than changing the authorised site. The indication remains submental fat in adults, and use elsewhere remains not approved and not recommended. The 2026 communication also warns about unapproved administration outside the submental region, where serious reports have included blurred or reduced vision after periorbital use, involuntary eye contractions, muscle or nerve injury, facial paresis, paresthesia, weakness, numbness and difficulty walking.[1,2]

The September 2026 FDA Safety Review

On 15 September 2026, FDA announced approval of additional Kybella label information after review of FAERS reports, now called the FDA Adverse Event Monitoring System (AEMS), and medical literature. FDA identified 129 cases of adverse events associated with deoxycholic acid injections administered outside the approved submental region since the 2015 Kybella approval. Among the serious reports were blurred or reduced vision after periorbital use and muscle or nerve injury, including facial paresis and paresthesia.[1]

FDA separately identified 117 cases of unresolved injection-site nodules or masses associated with deoxycholic acid injections. At the time of reporting, the unresolved lesions had been present for a mean of 143 days, approximately five months. FDA says nodules and masses have been reported following use in the approved submental region and outside it, and they may not resolve over time or may require medical intervention.[1] The communication describes cases identified from reports and literature; it does not provide an exposed-patient denominator from which incidence can be calculated.

The regulatory change involved two supplemental applications. FDA's approval letter states that supplement S-006, submitted in October 2024, added injection-site mass to the postmarketing adverse reactions section. Supplement S-007, received in July 2026, added a new warning and patient-information language about serious adverse reactions associated with unapproved use outside the submental region. FDA identified the relevant serious reactions during its review of postmarketing cases from FAERS and literature.[3] The current label lists September 2026 as the latest major revision for relevant Warnings and Precautions sections.[2]

Three evidence statements should remain distinct:

Spontaneous reports and case literature are valuable for detecting uncommon, serious or delayed outcomes, especially when trial follow-up was limited. They are subject to underreporting, stimulated reporting, missing clinical detail, duplicate submissions and uncertain product attribution. The FDA label itself cautions that voluntarily reported post-approval events come from a population of uncertain size, so their frequency cannot be reliably estimated and causality cannot always be established.[1,2] The appropriate interpretation is that FDA identified a safety signal and updated product information, not that the available case counts quantify risk.

The 2015 review had already noted concern about possible use outside the studied indication, but at that time FDA reported no data showing an increased risk relative to the proposed indication and treated that issue as a matter of medical practice. The 2026 action reflects later post-marketing evidence and does not mean the original trial findings were false; it demonstrates how exposure patterns after launch can reveal risks that were not adequately characterised at approval.[4]

Practical Case Assessment

A useful case assessment starts by establishing what product was administered and where. Reports may use “Kybella,” deoxycholic acid, a generic product, an imported product, or a non-FDA-approved “fat-dissolving” injection. Product name, manufacturer, package or lot information, source of supply and the reporter's certainty should be captured. Do not assume that all products marketed or described as deoxycholic acid injections are the approved product or have the same formulation, manufacturing controls or regulatory status. FDA has separately warned about unapproved injection-lipolysis products.[1,13]

For an event following Kybella use, document the treatment indication and anatomical site in the reporter's words, then clarify whether the injected area was submental or elsewhere. Record dates, number of sessions, interval, estimated dose and volume when available, onset and evolution of symptoms, injection-site description, concomitant procedures or products, relevant medical history, treatment, specialist evaluation, outcome and any recurrence after repeat administration. For a persistent nodule or mass, size and duration are especially useful because the current label distinguishes nodules from masses by diameter and notes that some lesions may persist or require intervention.[1,2]

Assessment should distinguish the clinical event from the use context. An administration outside the approved submental site is a relevant exposure circumstance; it is not itself an adverse event. The actual event might be nerve injury, visual symptoms, dysphagia, tissue damage or a persistent lesion. Conversely, a nodule after labelled submental use is still relevant even when administration appears consistent with the US indication. Do not classify all such cases as misuse or assume technique error unless the available evidence supports that assessment.

Product causality and procedure contribution may coexist. The molecule's cytolytic effect, injection depth, volume, grid placement, proximity to vulnerable anatomy, infection, vascular injury, local tissue response and the patient's anatomy can all be relevant. A report can support several plausible contributors without allowing a definitive conclusion. Record each as an assessed factor, not as established cause, and seek follow-up that can clarify chronology, examination findings, imaging or specialist intervention where feasible.

Signal Evaluation and Governance

The main surveillance question is not simply whether the case count rose after a label update. It is whether the observed pattern is clinically meaningful and sufficiently reliable to inform action. Stratify cases by product and manufacturer, approved versus outside-submental location, reported event, seriousness, time to onset, lesion persistence, intervention, outcome, treatment session and available denominator. Separate labelled-site from outside-site use, and distinguish deoxycholic acid reports from reports involving other active ingredients or unapproved mixed products.

For a rising report pattern, review duplicates and follow-up versions, case narratives, product traceability, reporting source, exposure opportunity and possible publicity effects. FDA's 129 and 117 are case counts from identified evidence, not rates. Do not divide them by the number of Kybella vials sold unless the exposure measure is appropriate to the case definition, time window, geography and product mix; even a crude reporting rate would not be an incidence estimate. If the organisation has reliable procedure-volume data, it may support contextual analysis, but its coverage and limitations should be explicit.

An actionable governance record should show the question assessed, data cut, search strategy, case definition, deduplication approach, relevant literature, local label comparison, limitations, clinical review, decision rationale, escalation and follow-up owner. Consider whether reporting obligations, local product information updates, healthcare professional communication, training, product-quality assessment or risk-minimisation review apply in the relevant jurisdiction. The FDA's 2026 action should be recorded as a US regulatory change; the impact on other territories requires separate RSI and regulatory assessment. For the general analytical sequence, connect this review to the QPPV.com guides on signal detection and signal validation.

Illustrative Learning Case

A clinic reports that a patient developed a firm, painful 4 cm mass after a second treatment with deoxycholic acid injections in the lower face. The product was described as “Kybella,” but the package is not available. The patient also had swelling and altered smile movement. The report does not identify the exact injection site, interval, lot, treating professional or whether the mass was cultured or imaged.

The first task is not to conclude that the event proves a class-wide effect. Clarify product identity and source, exact location, administration dates, event chronology, clinical findings, treatment, outcome, and whether a second report may exist through the clinic, patient or manufacturer. Capture the mass and the altered smile as separate clinical concepts, while assessing whether they could share an anatomical or procedural contributor. Determine whether the use was within the labelled submental region and compare the case with the applicable local label. If the product cannot be verified, preserve that uncertainty and consider product-quality or unapproved-product pathways. Escalate according to seriousness and local procedures, then document why the case does or does not change the aggregate signal assessment.

Common Analytical and Process Failures

• Treating 129 outside-region cases or 117 unresolved lesions as incidence estimates.
• Combining Kybella, an approved generic, foreign products and unapproved injection mixtures without preserving product identity.
• Treating “outside the submental region” as the adverse event instead of an exposure circumstance.
• Assuming a case was caused by poor injection technique without supporting clinical evidence.
• Missing the approved-site cases of persistent nodules or masses because attention is concentrated on unapproved use.
• Coding a patient-reported “lump” as a confirmed mass without obtaining dimensions or a clinical description.
• Using the September 2026 US label as if it automatically changed the RSI in every country.
• Treating a public label change as proof that all reports were causally established.
• Omitting data cut, case definition, follow-up and duplicate handling from the signal review record.

QPPV and Safety-System Checklist

The following are recommended system controls for a marketing authorisation holder or safety organisation. They are not presented as new FDA legal requirements.

• Maintain a current product dictionary that distinguishes Kybella, approved generic deoxycholic acid products, other products and uncertain product identity.
• Ensure intake forms and follow-up prompts can capture injected anatomical site, treatment indication, number of sessions, dates, dose or volume, lot and source where available.
• Preserve the verbatim report, then code the clinical event and relevant exposure circumstances separately.
• Route serious visual, neurological, swallowing, vascular, necrotic, infectious and persistent-lesion reports for appropriate medical review.
• Assess reports against the applicable local product information and document jurisdiction-specific label differences.
• Reconcile reports received from patients, clinics, medical information, product-quality channels, literature and vendors when those channels operate in the organisation.
• Maintain traceable evidence of signal detection, validation, decision-making, escalation, regulatory reporting and any follow-up action.
• Reassess training and risk controls when the procedure, product source, label, or pattern of use changes.

Key Takeaways

• Deoxycholic acid was not invented for Kybella; ATX-101 was a product-development programme that adapted its local cytolytic effect into a standardised injectable for submental fat.
• Phase I–III development established efficacy and a common local-reaction profile for the studied population and anatomical site. It did not establish safety or effectiveness for other fat deposits.
• The commercial history moved from Kythera's development, through Bayer's collaboration and later transfer of ex-US/Canada rights, to Allergan and then AbbVie ownership.
• FDA's September 2026 action added information about serious reports following use outside the submental region and about potentially persistent injection-site nodules or masses reported both inside and outside that region.
• The FDA case counts identify a safety concern; without a suitable exposure denominator and complete clinical data, they do not quantify incidence or prove causation.
• Good pharmacovigilance preserves product identity, site and procedure context, clinical details, local RSI, uncertainty and the reasoning behind signal decisions.

References

1. U.S. Food and Drug Administration. FDA Approves Additional Information in Labeling for Kybella (Deoxycholic Acid) Injection Warning of Adverse Reactions Associated with Unapproved Use. Drug Safety Communication, 15 September 2026. FDA page
2. U.S. Food and Drug Administration. Kybella (deoxycholic acid) injection: Prescribing Information, revised September 2026. FDA label PDF
3. U.S. Food and Drug Administration. Approval letter for NDA 206333, supplements S-006 and S-007. FDA approval letter
4. U.S. Food and Drug Administration. NDA 206333: Summary Review. 2015. FDA NDA review file
5. Kythera Biopharmaceuticals, Inc. Form 10-K for 2013: ATX-101 development and acquisition of ex-US/Canada rights from Bayer. U.S. Securities and Exchange Commission. SEC filing
6. U.S. Food and Drug Administration. CY 2015 CDER Drug and Biologic Calendar Year Approvals. FDA approvals list
7. Kythera Biopharmaceuticals, Inc. Allergan to Acquire KYTHERA Biopharmaceuticals. 17 June 2015. Filed with the SEC. SEC exhibit
8. AbbVie. AbbVie Completes Transformative Acquisition of Allergan. 8 May 2020. AbbVie announcement
9. Jones DH, Carruthers J, Joseph JH, et al. REFINE-1, a multicenter, randomized, double-blind, placebo-controlled, phase 3 trial with ATX-101, an injectable drug for submental fat reduction. Dermatologic Surgery. 2016;42(1):38–49. PubMed
10. Humphrey S, Sykes J, Kantor J, et al. ATX-101 for reduction of submental fat: a phase III randomized controlled trial (REFINE-2). Journal of the American Academy of Dermatology. 2016. PubMed
11. Allergan plc. Annual Report on Form 10-K for 2017. Reports completion of the Kythera acquisition on 1 October 2015. Annual report PDF
12. Allergan Aesthetics, an AbbVie company. KYBELLA product information and portfolio page. Allergan Aesthetics
13. U.S. Food and Drug Administration. Using Fat-Dissolving Injections That Are Not FDA Approved Can Be Harmful. FDA consumer information
14. Kythera Biopharmaceuticals, Inc. Quarterly Report on Form 10-Q for the period ended 30 June 2015. Reports US product sales commenced in late June 2015. SEC filing

Regulatory Note

This article is educational and describes US regulatory and safety information available through 3 October 2026. The current local authorisation, product information and regulatory communications should be checked before applying any conclusion in another jurisdiction. Clinical decisions must be made by qualified healthcare professionals using the patient's circumstances and current local prescribing information.