Retinal nerve fiber layer imaging with spectral-domain optical coherence tomography: analysis of the retinal nerve fiber layer map for glaucoma detection

Objective: To evaluate the diagnostic performance of the retinal nerve fiber layer (RNFL) thickness deviation map imaged by a spectral-domain optical coherence tomography (OCT; Cirrus HD-OCT, Carl Zeiss Meditec Inc, Dublin, CA) and compare its sensitivity and specificity for glaucoma detection with circumpapillary RNFL measurement derived from the standard 3.46 mm diameter circle scan.

Design: Prospective, cross-sectional study.

Participants: We included 102 normal subjects and 121 glaucoma patients.

Methods: One eye from each individual was imaged with Cirrus HD-OCT and Stratus OCT (Carl Zeiss Meditec Inc.). Glaucoma was defined based on the presence of visual field defects with the Humphrey visual field analyzer (Carl Zeiss Meditec Inc.). A scoring system (0-5) was developed to analyze the RNFL thickness deviation map taking the defect size, shape, depth, location, and distance from the disc margin into consideration. Each of these features was scored independently by a masked observer with a highest total score of 5 (glaucomatous RNFL defect) and a lowest score of 0 (no RNFL defect). Sensitivity and specificity were computed with a score of > or =3, > or =4, or =5. The diagnostic performance of circumpapillary RNFL measurement was analyzed with clock-hour and average RNFL thickness categorical classification.

Main outcome measures: Diagnostic sensitivity and specificity.

Results: The sensitivities of the RNFL thickness deviation map ranged between 95.0% and 97.5%. There were significant differences in specificity between a map score of 5, a map score of > or =4 (87.3%), and a map score > or =3 (72.5%; P< or =0.014). A map score of 5 attained a significantly higher sensitivity (95.0%) compared with clock-hour or average RNFL thickness categorical classification by Stratus OCT or Cirrus HD-OCT (46.3%-88.4%; P< or =0.033) at a comparable level of specificity (95.1%), except when glaucoma was detected as having > or =1 clock-hour at the < or =5% level by Cirrus HD-OCT in which an equally high sensitivity (93.4%) was found but at the expense of a significantly lower specificity (83.3%; P<0.001).

Conclusions: Analysis of the RNFL thickness deviation map provides additional spatial and morphologic information of RNFL damage and significantly improves the diagnostic sensitivity for glaucoma detection compared with conventional circumpapillary RNFL measurement.