Does Kybella Work? What the Trials Prove, and What They Never Tested

Kybella works on one tissue, and the paperwork says so in a single line. In the two phase 3 trials the FDA reviewed, 68.2 percent of treated patients improved by at least one grade on both the clinician-rated and patient-rated submental fat scales, against 20.5 percent on placebo; 16.0 percent improved by two grades or more, against 1.5 percent. An MRI substudy found 43 percent of treated patients lost at least 10 percent of their submental fat volume, against 5 percent on placebo. Deoxycholic acid dissolves the membranes of the fat cells it contacts and does nothing to any other structure. The FDA-approved label states that safe and effective use outside the submental region has not been established, and no trial has tested the drug against loose skin, against platysmal bands, or against the fat that sits beneath the platysma muscle.

Which turns the question around. The evidence for the drug is settled and public; the open variable is the chin in front of the mirror. The REFINE trials enrolled adults whose submental convexity was graded 2 or 3 on a five-point clinician scale and screened out anyone with excessive skin laxity. The label instructs the injector to palpate the area before every session to confirm there is preplatysmal fat to treat at all. That instruction exists because the same silhouette can be produced by four different things, and the injection only addresses one of them.

What "working" meant to the people who ran the trials

The FDA label describes two randomized, double-blind, placebo-controlled trials with 513 subjects treated with Kybella and 506 given placebo, 1,022 adults in total. Participants were 19 to 65 years old, 85 percent women, with a mean body mass index of 29.

Success was defined in advance as a composite: an improvement of at least one grade on both the Clinician-Reported and the Patient-Reported Submental Fat Rating Scale, each running from 0 to 4. In REFINE-1, published by Jones and colleagues in Dermatologic Surgery in 2016, 70.0 percent of treated patients cleared that bar against 18.6 percent on placebo. In REFINE-2, published by Humphrey and colleagues in the Journal of the American Academy of Dermatology the same year, the figures were 66.5 percent and 22.2 percent.

The two-grade bar tells a different story. REFINE-1 recorded 13.4 percent of treated patients reaching it; REFINE-2 recorded 18.6 percent. Sources diverge slightly on the comparator: the FDA label lists the REFINE-1 placebo rate as under 0.1 percent, while the published trial reports zero. Either way, a substantial visible change happened in fewer than one treated patient in five.

How much fat actually leaves, measured by machine rather than eye

Rating scales record impressions. MRI records volume, and the trials scanned a subset: 449 of the 1,022 participants, at preselected centers. The threshold for an "MRI responder" was set before the trials began at a 10 percent reduction in submental fat volume, a cutoff carried over from an earlier ATX-101 study for statistical powering.

Pooled, 43 percent of treated patients crossed it, against 5 percent on placebo. The individual trials sit either side of that average, and the gap between them is worth naming rather than smoothing: REFINE-1 reported 46.3 percent versus 5.3 percent, REFINE-2 reported 40.0 percent versus 5.1 percent. The FDA label publishes only the rounded pooled figures.

Read the other way, the majority of scanned patients who received the drug did not reach a 10 percent volume reduction. Both statements come from the same dataset.

What a full course actually involves

The dosing is unusually specific for an aesthetic treatment, because it is a drug rather than a device. The label prescribes an area-adjusted dose of 2 mg per square centimetre, delivered as 0.2 mL per injection, at sites spaced 1 cm apart. A single session may use up to 50 injections, or 10 mL, whichever comes first. Up to 6 sessions may be given, at intervals no less than 1 month apart.

The label's own sequence for a session runs like this:

1. Palpate the submental area to confirm sufficient preplatysmal fat and rule out other causes of fullness.
2. Outline the planned treatment area with a surgical pen.
3. Apply a 1 cm injection grid to mark every injection site.
4. With the patient tensing the platysma, pinch the submental fat and inject 0.2 mL into the preplatysmal fat beside each mark, using a 30-gauge or smaller half-inch needle.
5. Repeat at intervals of at least one month, up to a maximum of six sessions.

Volumes drop as the fat does. The label reports an average of 6.4 mL at the first session, falling to 4.4 mL at the sixth among those who completed the full course. Fifty-nine percent of trial subjects received all six treatments.

Practice looks different from protocol. Sagar Patel and Russell Kridel, reporting in JAMA Facial Plastic Surgery in 2018, surveyed 102 members of the American Academy of Facial Plastic and Reconstructive Surgery and found practitioners administering a mean of 2.6 sessions, against 4.1 in the trials. Just over a quarter of respondents, 27.4 percent, named patient cost as the primary reason treatment stopped.

The side effects that belong on the consent form

Marginal mandibular nerve injury, which presents as an asymmetric smile or weakness of the facial muscles, occurred in 4 percent of Kybella-treated subjects, 20 of 513, against under 1 percent on placebo. The label records that every case resolved without intervention, over a range of 1 to 298 days, with a median of 44 days. Dysphagia occurred in 2 percent, 10 of 513, resolving over 1 to 81 days with a median of 3.

Injections placed too superficially, into the dermis rather than the fat, can cause skin ulceration and necrosis. Cases of injection-site alopecia have been reported, and the label advises considering withholding further treatments until it resolves.

When the FDA announced the approval on 29 April 2015, Amy G. Egan, M.D., M.P.H., then deputy director of the Office of Drug Evaluation III in the agency's Center for Drug Evaluation and Research, put the boundary plainly: "It is important to remember that Kybella is only approved for the treatment of fat occurring below the chin, and it is not known if Kybella is safe or effective for treatment outside of this area."

Kybella, liposuction, and the procedures aimed at skin

No randomized trial has compared deoxycholic acid injection with submental liposuction directly. Patel and Kridel built the closest available comparison by pooling the phase 3 data against surveyed surgical fees.

| | Deoxycholic acid injection | Submental liposuction | Skin and platysma procedures | |---|---|---|---| | Targets | Preplatysmal fat only | Preplatysmal fat, and deeper compartments with an open approach | Skin laxity; platysmal bands | | Strongest evidence | Two placebo-controlled phase 3 trials, 1,022 adults | Long surgical record; no randomized trial against deoxycholic acid | Microfocused ultrasound cleared by the FDA in October 2012 for lifting submental and neck tissue; surgical series for platysmaplasty | | Sessions | Up to 6, at least a month apart | One | One for surgery; device protocols vary | | Reported cost | ASPS average of $1,157 per non-surgical fat-reduction treatment; Patel and Kridel computed $6,426.35 for the 186 mg trial average, at a surveyed $691.04 per 20 mg vial | Patel and Kridel: $2,976.56 mean total; ASPS liposuction surgeon fee average $4,711 | ASPS neck lift surgeon fee average $7,885 | | Where it fails | Loose skin, platysmal bands, subplatysmal fat | Patients unwilling to have surgery | Removes no fat |

The two cost figures for the same drug are not a contradiction so much as two different questions. The American Society of Plastic Surgeons publishes an average practitioner fee of $1,157 for non-surgical fat reduction, a category that also covers cryolipolysis and radiofrequency, and that figure covers one treatment. Patel and Kridel priced a trial-length course. Their recommendation was specific: clinicians should consider submental liposuction "a superior alternative in regard to cost and recovery" for patients who will need more than two vials per treatment and more than three sessions.

For fullness that is skin or muscle rather than fat, neither option applies. Expert consensus in the dermatologic literature holds that patients with long or wide platysmal bands achieve suboptimal outcomes with deoxycholic acid and may need neuromodulator injections or surgery. The FDA cleared microfocused ultrasound for non-invasive lifting of submental and neck tissue in October 2012, a separate indication addressing a separate tissue.

How long the reduction lasts once the sessions stop

Shannon Humphrey, MD, a dermatologist at the University of British Columbia, led a follow-up of REFINE participants published in Aesthetic Surgery Journal in 2021. It tracked 224 patients — 113 treated, 111 on placebo — for three years after their final session. Among treated patients who had responded, the improvement held in 86.4 percent at year one, 90.6 percent at year two and 82.4 percent at year three. The authors' conclusion: "Improvements in submental contour achieved with ATX-101 are maintained for up to 3 years in the majority of patients."

The placebo arm is the useful control on that number. Its own responders held their improvement at 56.8, 73.8 and 65.0 percent across the same three years, which suggests the retention figures partly measure how stable a graded rating is over time, not only how permanently fat cells are gone.

Skin laxity, tracked with a validated four-point scale, was unchanged or improved from baseline in most patients in both arms: 81 percent versus 75 percent at year one, 78 versus 76 at year two, 75 versus 74 at year three. Nobody with grade 4 laxity was in the trials to begin with.

Frequently asked questions

How long do Kybella results last?

In a three-year follow-up of REFINE trial patients published in Aesthetic Surgery Journal, responders maintained their improvement in 86.4 percent of cases at one year, 90.6 percent at two years and 82.4 percent at three. Destroyed fat cells do not regenerate, though remaining cells in the area can still enlarge with weight gain.

Is Kybella worth the money?

That depends on how many vials a chin needs. Patel and Kridel, in JAMA Facial Plastic Surgery, concluded that submental liposuction is the better value for anyone requiring more than two vials per session and more than three sessions. Below that threshold, the injection avoids surgery at comparable cost.

Is there anything better than Kybella?

No randomized trial has compared Kybella against submental liposuction or cryolipolysis directly, so "better" has no head-to-head evidence behind it. Liposuction treats larger fat volumes in one session. For loose skin or platysmal bands, neither approach helps, and surgical or energy-based treatment is the relevant option.

What are the negatives of Kybella?

The FDA label reports marginal mandibular nerve injury, causing an uneven smile, in 4 percent of trial subjects, resolving over a median of 44 days. Dysphagia occurred in 2 percent. Injections placed too superficially can cause ulceration and necrosis. Swelling, bruising, numbness and pain follow most sessions, and a course can require six visits.

How much does Kybella cost?

The American Society of Plastic Surgeons lists an average practitioner fee of $1,157 for non-surgical fat reduction, covering one treatment and excluding related expenses. A full course multiplies that. Patel and Kridel calculated $6,426.35 for the average drug quantity used across the phase 3 trials, at $691.04 per 20 mg vial.

Does Kybella work for the neck?

The approval covers submental fat, the pocket beneath the chin, and nothing below it. The FDA label states that safe and effective use for subcutaneous fat outside the submental region has not been established. Fullness lower on the neck, or caused by skin laxity and platysmal banding, falls outside what any trial tested.

Does my fullness come from fat, skin, or neck anatomy?

Only palpation answers this, which is why the label instructs injectors to pinch the area before every session and confirm preplatysmal fat is present. Fat that pinches between two fingers may respond. Skin that folds without a fat layer beneath it, visible vertical bands, or a low hyoid position will not.