EDITOR,Botulinum neurotoxin A, when injected into striated muscle, prevents acetylcholine release causing flaccid paresis of the muscle.¹ Scott termed this procedure "chemodenervation" and introduced the technique for treatment of strabismus in 1979.² The procedure has been reported to be safe with no systemic and minimal local complications. Complications commonly reported following botulinum injection to horizontal muscle are ptosis and vertical muscle involvement.³ We report a case of globe perforation following botulinum injection into the medial rectus muscle in a woman with high myopia.

CASE REPORT
A 46 year old woman presented with a 6 year history of diplopia, worse on dextroversion. Prism treatment by her optometrist had partly alleviated her symptoms. She was found to have bilateral abduction weakness, worse on the right side. She was a high myope wearing a 19.5 D lens in her right eye and 17 D in her left eye. She was normotensive, normoglycaemic, and had no neurological abnormalities on clinical examination. A decision was made to perform botulinum injection to her right medial rectus muscle to achieve temporary, or possibly longer lasting, symptomatic improvement.

After informed consent the eye was anaesthetised with 1% amethocaine hydrochloride eye drops. Under electromyographic (EMG) control a 27 gauge monopolar needle was advanced into the medial rectus muscle with no undue resistance. A good EMG response was obtained from the muscle. The signal increased on adduction and decreased on abduction of the eye. Botulinum toxin 2.5 units was injected. The injection was performed by a surgeon experienced in using the technique (BWF).

Immediately after treatment the patient complained of red floaters in the right eye "like drops of blood". Fundus examination revealed a small vitreous haemorrhage with a small retinal haemorrhage in the nasal retina at the equator (Fig 1). There was no obvious retinal tear. A diagnosis of globe perforation was made. There was poor uptake of laser energy by the atrophic retinal pigment epithelium and cryotherapy was therefore performed to the probable entry site. On follow up the retina remained flat. The motility response to the botulinum injection was poor. Paralysis of accommodation and dilatation of the pupil developed which resolved 10 months after the procedure.

View larger version (77K): [in this window] [in a new window]   Figure 1   Disc and nasal retina of the right eye showing retinal haemorrhage following botulinum injection.

COMMENT
Botulinum injection is regarded as a procedure with a low incidence of morbidity.⁴ Post injection ptosis and diplopia are transient complications. No eye has been reported to have lost vision as a result of botulinum injection. Accidental perforation of the globe is an acknowledged complication of peribulbar and retrobulbar anaesthesia⁵) and strabismus surgery.⁶ High myopia is a strong risk factor for globe perforation in peribulbar anaesthesia.⁷ Our case shows that globe perforation can occur with botulinum injection as with any other periocular surgical procedure.

In our patient the injection was performed by an experienced surgeon and the procedure was performed in a standard fashion using EMG control. Before injection there was good muscle signal. It has been shown that EMG signals may be recorded in most cases once the needle contacted the conjunctiva.⁸ Using a monopolar electrode needle, the reference electrode is located centimetres away from the active electrode and the needle can record signals that are closer to the needle electrode than the reference electrode even if they are several millimetres away from the needle electrode. In our patient it is likely that the tip of the needle was in the vitreous cavity at some stage during the procedure and an EMG response was present at that time. The most common potentially vision impairing complication of globe perforation is retinal detachment. Retinal detachment has been reported in globe perforations associated with peribulbar anaesthesia and after strabismus surgery.⁹ The theoretical risk of causing a globe perforation is greater with botulinum injection into an extraocular muscle than it is with peribulbar injection. The needle enters the muscle just behind the insertion and the sclera is at its thinnest at this part (0.3 mm). Any movement of the eye by the patient, with the needle in this position may result in an inadvertent perforation of the globe, especially in patients who are likely to have thin sclera, such as high myopes. Demonstration of increased signal by movement of the eye into the field of action of the muscle to be injected should probably be avoided for this reason.

Management of patients with scleral perforations is controversial. Some authors recommend that they should be treated with indirect diode laser or transscleral cryotherapy regardless of the depth of perforation, to reduce the incidence of retinal detachment.⁹ However, animal experiments have found a higher incidence of retinal detachment following heavy cryotherapy,¹⁰ and suggest cryotherapy should be used only if there is vitreous haemorrhage or the patient has a predisposing risk factor for retinal detachment. Our patient was a high myope and she had a small retinal and vitreous haemorrhage which increased her risk of developing a detachment. Globe perforation, although rare, is a complication that can occur with botulinum injection into an extraocular muscle and surgeons doing the procedure and their patients should be aware of this. The risk is higher in myopic eyes, as the equator of the globe is more posterior than usual, and the sclera thinner.

Botulinum toxin injection to an extraocular muscle should be approached with extreme caution in highly myopic eyes, and all movement of the eye should be avoided during the procedure.