Vitreous hemorrhage following inadvertent intra-ocular injection of botulinum toxin

Sir,

Botulinum Toxin Type A is routinely used in Strabismus practice for various indications and by various techniques ranging from electromyography (EMG) assisted to transconjunctival injection under topical anesthesia.[1,2] Potential complications of this seemingly simple procedure should be kept in mind while performing the same.

A 30-year-old lady was taken up for 4 Units (0.1 ml) injection of Botulinum Toxin into the Right Medial Rectus (RMR) muscle for 2-month-old Right Lateral Rectus (RLR) Palsy associated with tubercular meningitis. The procedure was attempted transconjunctivally under topical anesthesia, parallel to muscle insertion with a 30G needle. Upon injection the patient complained of severe pain and the intra-ocular pressure was noted to be markedly raised. Immediate indirect ophthalmoscopy confirmed intra-ocular penetration with central retinal artery pulsations and a pre-retinal hemorrhage medially. Immediate paracentesis was performed, followed by cryo-application over the injection site. Post-operatively the patient had a dispersed vitreous haemorrhage and underwent repeat ultrasonography to rule out retinal detachment. The hemorrhage cleared by conservative management and the visual acuity returned to pre-injection level over 2 weeks. A retinal break with surrounding cryo-reaction was visualised at the site of entry.

There are very few reports of intra-ocular injection of botulinum.[3,4,5] All confirm the non-toxic nature of botulinum to intra-ocular tissues. This is the first report (pubmed search) where vitreous hemorrhage was noted which cleared spontaneously.

On analysis we understand that injection of botulinum requires more caution than any other peri-ocular procedure for the following reasons:

• The associated deviation makes reaching the desired site difficult as in this case the RLR palsy with large esotropia caused us to introduce our forcep deep near the medial canthus almost blindly

• The pain associated with holding the extra-ocular muscle is significant causing the patient to squeeze the lids making the procedure difficult

• The need to inject near the ocular surface of the muscle (to block the nerve which enters from the bulbar surface) takes one closer to the sclera which is already thin below the muscle belly.[6]

It would thus be logical to inject the muscle after a retro or a peribulbar block in the absence of EMG which would take care of the first two concerns (above) associated with this procedure. Using a 26-G needle rather than a 30-G and injecting perpendicular to the muscle rather than parallel to it would also decrease the possibility of intra-ocular penetration. We were fortunate to escape serious complications which could have occurred in our patient and probably immediate paracentesis and cryopexy were decisive.