Sir,

We present the first case of multiple uniocular breaks in Descemet’s membrane secondary to presumed non-accidental injury (NAI).

Case report

The first of twins, born at 35 weeks by spontaneous unassisted vaginal delivery, presented with a 2-week history of unexplained corneal haze in her left eye at the age of 4 months. Topical antibiotics and steroids were prescribed and she subsequently failed to attend. At 8 months, she was admitted to a second hospital with subdural haemorrhage, rib fractures, retinal haemorrhages and facial bruising. A diagnosis of NAI was made.

At 16 months she attended this unit with left amblyopia, eccentric fixation and esotropia. Her visual acuity was 0.2 logmar in the right eye and 1.0 in the left (Kay pictures, UK). There was no significant refractive error. She had a faint stromal scar medially in the left eye with multiple Descemet’s breaks seen on slit-lamp, the largest being vertically in the visual axis (Figure 1). Her cornea was of normal diameter, spherical, with no evidence of keratoconus. Her intraocular pressure was normal. No abnormality was detected in the right eye. Connective tissue disorders were excluded by a paediatrician. Her ocular appearance is unchanged on follow-up.

Figure 1
figure 1

Left cornea on retroillumination demonstrating breaks in Descemet’s membrane highlighted with red arrows.

Full size image

Discussion

Ocular symptoms and signs are the presenting feature in 1–5% of child abuse cases.1 It is recognised that 31% of cases are misdiagnosed on their first presentation and many of them go onto suffer repeated and escalating episodes of violence,2 as in this case.

It is likely that the Descemet’s breaks in our case were not seen at presentation due to the overlying corneal haze secondary to acute hydrops, but that they were there from the unexplained injury at age 4 months. Indeed the history of sudden onset of corneal haze would be explained by the Descemet’s breaks. Other causes of Descemet’s breaks, such as forceps injury at birth,3 prolonged labour, keratoconus and congenital glaucoma4 were excluded by history and examination. Posterior polymorphous dystrophy was also considered and excluded.5

We recommend increased awareness and the inclusion of non-accidental injury as a differential diagnosis for unexplained Descemet’s breaks with corneal haze in a child.