Botulinum toxin: the effect over placebo was clear in two phase 3 trials, both run by the manufacturer

Two phase 3 trials (ULTRA I and ULTRA II) in adults aged 18 and over. Participants with moderate-to-severe glabellar, forehead and lateral canthal lines at maximum contraction were randomised 2:1:1 to up to 64 units of incobotulinumtoxinA or placebo.

Three hundred sixty-two and 368 people respectively — 730 in total — received treatment. In both studies the treatment was significantly more effective than placebo on the primary endpoints and the key secondary endpoints (P < .0001 for each). Results in the open-label extension were consistent.

Assessment used investigator ratings on five-point scales. That the endpoint is a trained assessor's score rather than self-reported improvement makes this result sturdier.

What remains is funding. These are a manufacturer's registration trials, and that is why this entry sits at moderate.

A different indication from wrinkles. And the evidence here is thicker than on the wrinkle side.

A 2007 North American multicentre study followed 322 subjects for 52 weeks, double-blind. Participants were randomised to botulinum toxin type A at 75 U or 50 U per axilla, or to placebo.

The results.

• Reaching a 2-point improvement on the 4-point Hyperhidrosis Disease Severity Scale (HDSS): 75% in the 75-U group, 75% in the 50-U group, 25% in the placebo group (P < .001)
• The finding was corroborated by gravimetric measurement — actually weighing the sweat
• Median duration of effect: 197 days at 75 U, 205 days at 50 U, 96 days on placebo

Two things have to be read together.

One, 75 U and 50 U came out the same. The response rate was 75% in both, and the duration was in fact longer at the lower dose (205 days against 197). Raising the dose by half did not separate the results.

Two, 25% of the placebo group reached a 2-point improvement, and it lasted 96 days. One in four people injected with something inert dropped two steps on a severity scale, and stayed there for over three months.

What that figure means on this site: on other procedure pages we write, repeatedly, that without a control group the treatment's own contribution cannot be known. Here it is there in numbers — the gap between 75% and 25% is the drug's share, and the 25% is the share of the needle, the expectation and the passage of time.

A 2026 phase 3, multicentre, randomised, double-blind, placebo-controlled study in Chinese patients looked at the same indication. We could not verify the participant count from the abstract, so that field is left blank.

The limits, too. The authors of the 2007 trial state that the effect of total surface area involvement was not evaluated, and both trials are manufacturer-funded. So by this site's rule it does not rise to 'strong evidence' — the deoxycholic acid page applies the same rule.

The Korean drug register lists 59 products with botulinum toxin as the active ingredient, all prescription-only (checked 16 September 2026). Products such as Nabota (Clostridium botulinum toxin type A) are among them.

It requires a doctor's prescription and a procedure. This page does not cover units, injection sites or intervals — those are settled in consultation.

A 2019 double-blind controlled pilot study in Dermatol Ther, from Iran. It is the same trial as on the intralesional corticosteroid page — do not count it as two badges.

20 patients with a cumulative 40 keloids completed it.

• one arm — intralesional triamcinolone alone
• the other — intralesional triamcinolone plus botulinum toxin A

The result split in two.

• shape — no significant difference between the arms in height, vascularisation, pliability and pigmentation scores
• symptoms — the arm with botulinum toxin added was significantly better than triamcinolone alone for pain and pruritus (p < 0.001)

Now how to read it.

First, this entry's endpoint is what the patients felt. Pain and itch are what a person tells you. This site marks that as 'self-reported', and an endpoint like that on its own does not raise a grade. So even with a p < 0.001 attached, this entry stops at 'early evidence'.

Second, but there is no reason to think the result small either. With keloids, pain and itch are often the very reason someone comes to a clinic. If it hurts and itches less while the shape stays the same, that is a meaningful change. When this site writes that self-assessment is a weak endpoint, it does not mean the figure does not matter; it means it measures something different.

Third, the paper's title speaks for the other side. It reads 'Intralesional triamcinolone alone or in combination with botulinum toxin A is ineffective for the treatment of formed keloid scar.' That is, a title set by the shape results.

And that title needs care: this trial has no untreated control. Both arms received steroid. That is set out on the intralesional corticosteroid page.

The sentence this page can write is this: when botulinum toxin was added to a keloid that was already receiving steroid, the shape did not change and the symptoms improved.

Fourth, 20 patients, and the authors called it a pilot study. With 40 keloids there are more lesions than people.

Fifth, the abstract does not state the duration.

Sixth, the authors' conclusion is 'it should be used adjuvantly'. They write that future studies should be performed in the adjuvant setting after excision.

So it sits at 'early evidence'. Do not read it as the same badge as this page's other entries — expression lines and sweating. What holds them up is not of the same thickness.