Ophthalmological consequences of cyanotic congenital heart disease: vascular parameters and nerve fibre layer.

BACKGROUND

This study investigated the long-term ophthalmological consequences of cyanotic congenital heart disease (CHD).

DESIGN

Cross-sectional study, tertiary referral setting.

PARTICIPANTS

Thirteen adults with cyanotic CHD (40 ± 4 years). Age- and sex-matched healthy controls underwent aspects of the protocol.

METHODS

Cyanosed subjects had a full ophthalmic examination, visual fields, scanning laser ophthalmoscopy and optical coherence tomography to assess retinal nerve fibre layer (RNFL), retinal photography and cerebral magnetic resonance imaging (MRI).

MAIN OUTCOME MEASURES: RNFL thickness and quantitative analysis of retinal vessels with fractal dimension, branching and central retinal arterial equivalent (CRAE) and central retinal venous equivalent (CRVE).

RESULTS

No abnormalities of anatomy, motility, intraocular pressure or anterior segments were detected apart from one subject who had bilateral cataracts. Corrected visual acuity was normal in all but one cyanosed subject. Clinical examination revealed dilated retinal vasculature in 12/13 cyanosed subjects and increased tortuosity in 8/13. In the setting of cyanosis, skeletonized retinal arterial and venous beds had higher fractal dimension and increased branching (P ≤ 0.01, n = 11 for all); retinal vessels were dilated (CRAE: 227 vs. 183, n = 11, P < 0.0001; CRVE: 254 vs. 221, n = 11, P = 0.01). Visual fields showed scotomas in two subjects associated with RNFL thinning. No disc oedema was detected. 6/13 subjects' RNFL thickness fell below the normal 95% confidence interval in at least one sector without explanatory cerebral pathology (P < 0.0001, n = 13). Mean RNFL thickness correlated with MRI cerebral white matter volume (R = 0.67, P = 0.035).

CONCLUSIONS

Cyanosed subjects had vessel dilatation, increased branching and tortuosity. RNFL and visual field thresholds were reduced suggesting impaired neuro-ophthalmological functioning.