Comparison of Cirrus and Spectralis Optical Coherence Tomography: Factors Associated With Differences in Retinal Nerve Fiber Thickness Measurement (P5.312)

Abstract

OBJECTIVE: To evaluate factors associated with differences in retinal nerve fiber layer (RNFL) thickness measurement between two optical coherence tomography (OCT) devices. BACKGROUND: RNFL thickness, an important outcome measure in Multiple Sclerosis (MS), can be captured by different OCT devices that vary in their agreement within patients and populations. METHODS: OCT scans (optic disc protocol) were performed on both Cirrus and Spectralis devices on the same day for 324 patients and 28 controls. Pearson correlation, paired t-test and Bland-Altman analyses were used to assess correlations, differences and agreement between devices. Measurements for right eyes only were analyzed. RESULTS: 94[percnt] of patients had relapsing remitting MS (RRMS), 5[percnt] had secondary progressive MS (SPMS) and 38[percnt] of eyes had a history of optic neuritis (ON). RNFL thickness by Spectralis was thicker than Cirrus in 66[percnt] of the eyes, exactly the same in 8[percnt] and thinner in 26[percnt]. RNFL thickness measurements highly correlated between the devices in control eyes, all disease subtypes, and eyes with and without ON (r=0.94-0.97, p<0.001). The mean difference in RNFL thickness between devices was significant in controls (2.6 microns, p<0.001), RRMS patients (1.9 microns, p<0.001) and in eyes without ON (2.9 microns, p<0.001), but not for SPMS patients (-0.9 microns, p=0.5) nor for eyes with a history of ON (-0.48 p=0.4). The difference was more pronounced at lower Cirrus signal strength: 1.8 microns for signal 蠅8 (p<0.001) vs. 3.8 for <8 (p<0.001). The limit of agreement (LOA) between devices was wide in all patient subgroups with the smallest LOA for controls (-9.26 to 4.02 microns). CONCLUSION: Cirrus and Spectralis have small population mean differences, notably in eyes with more damage, but have poor individual agreement. Improved signal quality may allow interchangeability between devices only in select populations but not on an individual level.

Disclosure: Dr. Wilson has nothing to disclose. Dr. Chahin has nothing to disclose. Dr. Jacobs has received personal compensation for activities with Biogen Idec, Questcor Pharmaceuticals and Teva as a consultant and/or speaker. Dr. Markowitz has received personal compensation for activities with Bayer Pharmaceuticals Corporation, Biogen Idec, Novartis, Roche Diaganostics Corporation, Genzyme, Genentech, EMD Serono, and Teva Neuroscience as a consultant. Dr. Calabresi has received personal compensation for activities with Abbott, Vaccinex, and Vertex. Dr. Calabresi has received research support from Novartis and Biogen Idec. Dr. Frohman has participated in the speakers' bureau for Teva Neurosciences, Acorda, and Novartis and has received consulting fees from TEVA Neurosciences, and Acorda., Dr. Galetta has received personal compensation for activities with Biogen Idec, Genzyme, and Vaccinex. Dr Balcer received personal compensation from Biogen Idec and consulting for Biogen Idec, Vaccinex and Genzyme. She is on a clinical trial advisory board for Biogen-Idec.,