Sir,

We read with interest the article by Drew et al,1 and would like to comment on the authors’ clinical management and conclusions.

In this patient, the authors elected not to investigate further for immunodeficiency. This is a fair approach in developed countries for healthy children with a sentinel herpes zoster ophthalmicus (HZO) event, as most do not have immunodeficiency or malignancy. However, in India and Africa, where human immunodeficiency virus (HIV) infection is epidemic, the first presentation of herpes zoster may be an indicator of HIV infection.2, 3 We would like to clarify this patient's background and whether HIV status was assessed, as this would be of clinical importance if she had originated from the Indian subcontinent.2 Furthermore, adult patients with HZO-associated ophthalmoplegia were significant for HIV infection.4 HZO was the initial clinical manifestation in some cases, suggesting that reliance on clinical ‘stigmata’ of immunodeficiency as a guide to testing may be unreliable. None of the children reported with HZO-associated ophthalmoplegia had been tested for HIV, and therefore, the likelihood of underlying immunodeficiency in this condition is still unknown.

The authors mentioned that this patient received acyclovir and prednisolone to ‘hasten resolution of her cranial nerve palsies’. Although early acyclovir treatment is effective for reduction of the duration of rash, incidence of pseudodendritic and immune stromal keratopathy and incidence and severity of postherpetic neuralgia, there is little evidence for its role in improving outcomes in HZO-associated ophthalmoplegia.5 Likewise, there is little consensus on the efficacy of steroids. Good recovery of ophthalmolplegia and ptosis in both adults and children possibly reflects an intrinsic disease course towards resolution even without treatment. The rarity of paediatric zoster has prevented the conduct of prospective studies evaluating treatment options; however, we recognise that acyclovir and steroid administration is common in clinical practice.

Despite improvement in ophthalmoplegia and ptosis, mydriasis persisted. This has been consistently observed in patients with HZO and pupil dilation.3, 4 We hypothesise that orbital inflammation may have a relatively greater impact on the smaller parasympathetic C fibres to the sphincter pupillae than on the larger Aδ motor fibres innervating the extraocular muscles. Mydriasis should be recognised as an important neuro-ophthalmic sequelae of HZO.