Abstract
Purpose :
While fluorescein angiography (FA) and optical coherence tomography (OCT) are both considered important in the primary diagnostic workup of patients with neovascular age related macular degeneration (nAMD), it is unclear whether routinely ommiting the relatively invasive FA in the follow up of these patients carries the risk of overlooking the need for retreatment. To quantify the risk of missing an active lesion in a routine clinical health care setting, a trial of diagnostic agreement and of the resulting therapeutic decisions based on different diagnostic modalities was performed. The indication for retretament was thus first solely based on SD-OCT (spectral domain OCT), then on OCT plus FA and finally on OCT, FA and ophthalmoscopy.
Methods :
A total of 139 consecutive patients of a follow-up clinic for patients with nAMD seen for the decision of whether to reinject ranibizumab were included. (63% female, median age 80 years (51 to 96). For each patient SD-OCT volume scans and a FA were performed. The indication for a possible re-treatment was then first based exclusively on the OCT examination, then using the additional information from the FA and finally by also considering the findings of the fundus examination.
A sample size of at least 130 eyes was calculated. The analysis was based on Cohen’s kappa for pair-wise agreement with a 95% confidence interval and on the McNemar test for symmetry at a 5% significance.
Results :
The concordance of OCT- vs OCT/FFA based decisions was estimated 0.90 (Cohen’s kappa 95%, CI 0.81; 0.97). In 5% of the patients an indication for treatment was seen after the additional information of the FA, implying a locally significant deviation in decision patterns (McNemar p=0.016). Concordance of OCT vs OCT/FFA/ophthalmoscopy based decision was furthermore estimated 0.85 (95%, CI 0.76; 0.94). In contrast to the solely OCT-based decisions, treatment was seen indicated in an additional 6% of the patients, whereas in 2 patients (1%) a previous decision for treatment was taken back (McNemar p=0,109).
Conclusions :
Despite encouraging agreement between OCT, OCT/FFA and OCT/FFA/ophthalmoscopy based treatment decision patterns for anti-VEGF therapy, the exclusively OCT based diagnostic procedure carries a notable risk for missing active lesions in nAMD and thereby the risk of missing necessary treatment.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.