Background: The competing interests of preventing recurrent ischemic optic neuropathy (ION) and minimizing medication side effects make corticosteroid dose reduction in giant cell arteritis (GCA) a difficult problem. The authors sought to determine whether any factors were predictive of recurrent ION.
Methods: Retrospective review of the records of 100 consecutive patients with biopsy-proven giant cell arteritis diagnosed in two Australian hospitals between 1988 and 1998. Among 67 patients who met inclusion criteria for ION in GCA, seven patients had recurrent ION. We compared the seven patients with recurrent ION to the 60 patients with nonrecurrent ION in terms of age, gender, mode of corticosteroid delivery, initial visual acuity in the affected eye, prevalence of bilateral ION, initial erythrocyte sedimentation rate (ESR) level, and rate of corticosteroid dose reduction. In the recurrent ION group, we documented the timing of the recurrence in relation to corticosteroid dose, elevation in acute phase reactants, and relapse of systemic symptoms.
Results: We found recurrent ION in GCA in 10% of our cohort, higher than has been previously reported. Recurrences, all of which were ipsilateral, occurred from 3 to 36 months (median 8 months) after the initial ION. None of the clinical indicators the authors examined differed between the two groups. Six of seven patients with recurrent ION had elevations in ESR or C-reactive protein or a new headache at the time of ION recurrence, but in only one of these patients were these features recognized as preceding the recurrent ION. One patient had neither an elevation in acute phase reactants nor a relapse in systemic symptoms of GCA at the time of ION recurrence.
Conclusions: Recurrent ION in GCA is difficult to predict. Although elevated acute phase reactants or new systemic symptoms consistent with GCA were present in six (83%) of our patients with ION recurrence, in only one patient (17%) did these events occur with enough lead time to allow caregivers to act preemptively. Thus, even very close monitoring of GCA patients with ION may not predict ION recurrence.