Comparison of Spectralis and Cirrus spectral domain optical coherence tomography for the objective morphometric assessment of the neuroretinal rim width.

Purpose: The assessment of cup-disc ratio as a surrogate parameter for the neuroretinal rim width (NRW) of the optic nerve is well established, but prone to human error and imprecision. Objective assessment of the NRW is provided by spectral domain optical coherence tomography (SD-OCT). This study is the first to systematically compare NRW measurements acquired with the Carl Zeiss Meditech Cirrus HD-OCT 5000 and the Heidelberg Engineering Spectralis SD-OCT.

Methods: In this cross-sectional study, 20 eyes of each 20 glaucoma patients and 20 age-matched healthy controls underwent ophthalmic examination, SD-OCT imaging, and computer perimetry. Regression analyses were performed for the NRW comparability and the effect of the rotational alignment disconcordance (RAD), receiver-operating characteristics (ROC) for NRW-based healthy glaucoma discrimination capability, and Pearson's correlation for covariate association.

Results: Mean NRW differences were 8 ± 48 μm (p = 0.4528), 91 ± 80 μm (p < 0.01), and 49 ± 77 μm (p < 0.001) in the glaucoma, healthy, and whole group. On average, the Cirrus showed higher NRW values (+ 50 μm) than the Spectralis, this difference increased with values starting with 159 μm. Discrimination ROC were 1.0 (Spectralis) and 0.9675 (Cirrus). RAD showed very little effect on NRW (R = 0.9661, p < 0.001). NRW-covariate correlation was highly significant (p < 0.001) with both devices for clinical cup/disc ratio, calculated rim width, visual field mean, and pattern deviations.

Conclusions: Our results suggest to only cautiously compare Spectralis and Cirrus NRW measurements only in patients with morphologically manifest glaucoma. For morphological progression analysis, we recommend the continuous usage of the same device.