Retinal oximetry in central retinal artery occlusion

dizziness when she woke up and realized that she had almost completely lost her vision on the left eye. She has suffered from polymyalgia rheumatica and had received oral prednisolone until 2 months earlier. The following day she had bare light perception and a pupillary afferent defect on the left eye. The left fundus showed central retinal artery occlusion with segmented blood columns (‘boxcarring’, Fig. 1) in the arterioles and venules, marked ischaemic oedema and a cherry red spot in the fovea. The right eye had normal visual acuity of 1.0. The intraocular pressure was normal in both eyes. Sedimentation rate was only 26 and 29 mm ⁄h on repeated measurements (Elfarsson et al. 2010). Given the history and arterial occlusion and a normal ultrasound examination of heart and carotid arteries, a tentative diagnosis of temporal arteritis was made. Treatment was started with prednisolone tablets 1 mg ⁄kg and a temporal artery biopsy obtained. Histological examination confirmed temporal giant cell arteritis (Fig. 1) and prednisolone treatment was continued. Ultrasound examination of the heart and carotid arteries was performed, keeping in mind that Miyazawa et al. (2011) reported that carotid stenosis is frequently associated with central retinal artery occlusion. Also, Mansour & Younis (2011) and Pournaras et al. (2010) suggested that ocular paracentesis and hypotensive treatment may be considered, even this late (>24 h) following the onset of central retinal artery occlusion. Spectrophotometric retinal oximetry (Hardarson et al. 2006) was performed on both eyes approximately 30 h after she noticed the vision loss. Retinal arterioles showed oxygen saturation of 71 ± 9% (mean ± SD for major arterioles) and the venular oxygen saturation was 63 ± 9% (Fig. 1). The corresponding values for the normal fellow eye were 95 ± 5% (arterioles) and 66 ± 8% (venules). The low arteriovenous difference in the affected eye is consistent with stagnant blood in the vasculature, cell death and decreased oxygen consumption. One month later the eye remained blind, although the perfusion of the retinal circulation had been restored. The oxygen saturation in the arterioles had returned to 100 ± 4%, while the venous saturation was 54 ± 5% (Fig. 1). At this time, the arteriolar saturation in the fellow eye was 100 ± 4% and the venular saturation 60 ± 6%. The pathogenesis of central retinal artery occlusion has been extensively studied (Hayreh 2011) although few studies are available on the influence of occlusion on human retinal oxygenation. Hammer et al. (2009) investigated branch and central retinal artery occlusions with a two wavelength oximeter. At the time of diagnosis, which was within 48 h of the occlusion, they found low arterial saturation (78%) in the occluded branch artery, similar to our findings, and a considerable increase (91%) 5 days later, after treatment with pentoxifylline for improvement of blood flow. Saturation after central retinal artery occlusion was, however, initially close to normal (93%) although it did increase after 5 days and treatment (103%). Venous saturation in central retinal artery occlusion was initially slightly below normal (55%) and had increased to 70% 5 days later. A later paper by the same group confirmed the results for occluded branch retinal arterioles (Gehlert et al. 2010). Spectrophotometric retinal oximetry is a relatively new form of diagnostic imaging. It can clearly show decreased arterial oxygen saturation following central retinal artery occlusion and may thus become helpful in diagnosing this disorder. It also demonstrates the return of blood flow and may Diagnosis and Therapy in Ophthalmology