Effect of ophthalmic artery collateral flow on cerebral hemodynamics in carotid artery occlusions.

Aim: In internal carotid artery (ICA) occlusions a reversal flow in the ophthalmic artery (OA) has been considered both an important collateral as well as a sign of inadequate intracranial cross-flow. The aim of the study was to evaluate the effect of OA collateral on the cerebral haemodynamics using ultrasound and CCT. Methods: In 540 patients with a total of 577 ICA occlusions the presence of an OA collateral was investigated with periorbital continuous-wave Doppler sonography and the findings were correlated with the ipsilateral cerebrovascular reserve capacity (CVR), measured with the Transcranial Doppler (TCD) CO2 test. Moreover, in 128 of 577 ICA occlusions the presence of an OA collateral was compared with the type of infarction in CCT and the neurological deficit. Results: In 577 ICA occlusions an OA collateral was found in 59.4% (n = 343), whereas there was no or anterograde OA flow in 40.6% (n = 234). A significant correlation could be demonstrated between the CO, reactivity and an OA collateral, which was present in 51.3% ipsilateral to a normal CVR (n = 175), in 66.9% ipsilateral to a diminished CVR (n = 115) and in 82.8% ipsilateral to an exhausted CVR (n = 53). Patients with OA collateral had a 15.5% risk of developing an exhausted CO2 reactivity, while there was only a 4.7% risk in the remaining patients (p< 0.001). In the 23 patients with haemodynamic infarctions in CCT the highest percentage of an OA collateral (87%) was found; no correlation could be demonstrated in these cases between the degree of neurological impairment and OA collateral flow. Conclusion: Patients with ICA occlusions and an OA collateral are subject to a more than a threefold risk of suffering from an exhausted CO2 reactivity in comparison with those revealing no reversal OA flow. Nevertheless, in patients with ICA occlusions and an OA collateral, CO2 reactivity was found to be normal in 51%, implying that in individual cases ultrasonic investigation of the periorbital flow is in itself insufficient to estimate the haemodynamic stroke risk and therefore cannot replace the TCD CO2 testing.