What the MHRA actually said

On 15 July 2026 the MHRA issued a Drug Safety Update covering all botulinum toxin type A products, following rare reports of iatrogenic botulism after therapeutic and cosmetic use. The advice is straightforward: symptoms typically appear within days but may take up to four weeks; they include difficulty swallowing, slurred speech, breathing difficulty and generalised muscle weakness; and patients should seek immediate medical attention if they occur. Product information and patient leaflets have been updated accordingly.

The agency also restates known risk amplifiers: underlying neurological or neuromuscular disorders, a history of dysphagia or aspiration, high doses, and use outside licensed indications or into unapproved sites. Serious reactions attributed to distant spread have been monitored and reflected in product information since 2007.

None of this is new pharmacology. What is new is emphasis, and emphasis has consequences — for consent forms, for insurers, and for how a specialty explains an adverse event to a patient who has read a newspaper.

Where the criticism sits

The warning is framed around the molecule. Toxin may spread; therefore watch for symptoms. That framing is appropriate for a regulator, whose remit is the product licence, not the practitioner's hand. But read as clinical guidance it is close to fatalistic: it implies that iatrogenic botulism is an intrinsic and largely unmodifiable property of treatment.

The published clinical pattern does not support that reading. Swallowing dysfunction is not distributed evenly across facial treatment sites. It is concentrated where the needle is placed near, or within, the muscles of deglutition. In therapeutic practice that is the cervical dystonia population; in cosmetic practice it is the neck.

The variable the warning does not discuss — because it cannot — is competence. Dose selection, dilution, injection depth, needle angle, vector and lateral versus medial placement are decided by the injector, not by the manufacturer. Any safety communication that omits them describes the hazard while leaving the largest controllable determinant of exposure unaddressed.

The evidence that dysphagia is a neck problem

In cervical dystonia, dysphagia is among the most common adverse effects, reported across series at roughly 10–40% depending on dose, target muscle and guidance technique. A drug-class analysis of prescribing information for botulinum toxin type A products in cervical dystonia confirms dysphagia and neck muscle weakness as recognised class effects rather than product-specific anomalies, and instrumental studies using videofluoroscopy have documented measurable change in swallowing parameters after injection even where patients do not spontaneously complain.

In cosmetic practice the reports are sparser but strikingly consistent in location. Dysphagia has been described after botulinum toxin for neck rejuvenation, and mild-to-moderate dysphagia has been reported after very low-dose abobotulinumtoxinA for platysmal bands — at an onabotulinumtoxinA-equivalent of approximately 35 units, well below the doses usually invoked when clinicians reassure themselves that a total unit count is safe.

That last case matters most. If dysphagia can follow 35 unit-equivalents in the neck while 60 units in the upper face routinely does not produce it, then total dose is a poor single explanation. What differs is not the quantity of toxin but its anatomical relationship to the pharyngeal and strap musculature.

Depth, angle and vector: why the platysma is unforgiving

The platysma is a broad, thin subcutaneous sheet, frequently one to two millimetres thick, lying immediately superficial to the deep cervical fascia and the infrahyoid strap muscles. Anatomical injection guidance for neck aesthetics repeatedly emphasises superficial, intradermal-to-immediately-subplatysmal placement using small volumes and a low-angle approach precisely because the safety margin is measured in millimetres, not centimetres.

Depth cannot be discussed without entry angle and vector. A needle entering the same skin point at 90 degrees versus 30 degrees, advanced the same five millimetres, terminates in two different anatomical compartments. Perpendicular entry into a pinched band is a fundamentally different act from a shallow tangential pass along a tented band, even though both are documented in a chart as “platysma”.

Volume compounds the error. High-dilution technique delivers a larger fluid bolus per unit, and fluid follows the path of least resistance — which, in the anterior neck, is medially and deep along fascial planes toward exactly the muscles nobody intended to weaken. Low-volume, laterally biased, superficial placement is not a stylistic preference; it is the mechanism by which the complication is avoided.

Guidance changes the number

The most instructive dataset in this discussion is not a cosmetic one. In a cervical dystonia cohort, 34 episodes of dysphagia occurred across 98 injection sessions performed with electromyographic guidance — a cumulative rate of 34.7%. When ultrasound imaging was added to the same operators treating the same indication, dysphagia occurred in 0 of 27 sessions.

The molecule did not change. The licence did not change. The dose range did not change. What changed was the accuracy of placement relative to muscle thickness and to the structures beneath it. That is the clearest available demonstration that a substantial share of what gets recorded as “spread” is, in mechanistic terms, misplacement.

Extrapolating a small therapeutic series to cosmetic practice requires caution, and it should not be read as a mandate for routine ultrasound in aesthetic neck treatment. It should be read as evidence that placement precision is a modifiable determinant of a complication currently being discussed as if it were purely stochastic.

What should change in practice

Consent should name dysphagia specifically for any neck treatment, alongside the MHRA's four-week symptom window and the instruction to seek urgent medical assessment for swallowing, speech or breathing difficulty. Generic wording about “spread to nearby muscles” no longer meets the standard the regulator has set.

Case selection should exclude, or refer, patients with pre-existing dysphagia, aspiration history, neuromuscular disease, prior neck surgery or radiotherapy, and those already receiving therapeutic botulinum toxin elsewhere within the dosing interval. Cumulative exposure across indications is easy to overlook when treatments are delivered by different clinicians.

Technique should default to superficial placement into or immediately beneath the platysma, small aliquots, conservative total dosing, lateral rather than midline emphasis, and avoidance of deep perpendicular entry over the anterior neck. Where a band cannot be lifted away from deeper tissue, the safer decision is to reduce dose or decline rather than to inject blind at depth.

Training should treat the neck as a distinct high-consequence zone rather than as a continuation of lower-face work. Practitioners who have only ever been taught upper-face injecting should not be adding platysma treatment on the basis of a product leaflet and a weekend.

Position

Iatrogenic botulism after cosmetic botulinum toxin is rare, and the MHRA is right to strengthen patient-facing warnings. Rarity is not an argument against vigilance, and clinicians who dismiss the update as media noise are making the specialty's case for statutory licensing on its behalf.

But the corollary deserves equal airtime. The overwhelming majority of cosmetic dysphagia reports arise from one anatomical region, treated with too much product, too deeply, too close to the midline, by practitioners whose training did not cover the fascial relationships of the anterior neck. Framing this as unpredictable pharmacological spread is comfortable, because nobody is responsible for the unpredictable.

Toxin spreads. It has always spread. The variable that is actually under discussion is where it was put.