Neuroretinal rim in non-glaucomatous large optic nerve heads: a comparison of confocal scanning laser tomography and spectral domain optical coherence tomography

Background/aims To compare margin-based rim area measurements from confocal scanning laser tomography (CSLT) with Bruch's membrane opening (BMO)-based measurements from spectral domain optical coherence tomography (SD-OCT) by analysis of a group of nonglaucomatous eyes with large optic discs, so-called macrodiscs (disc size >2.45 mm(2) in CSLT). Objective is to create a reference base for large optic nerve heads in SD-OCT diagnostics. Methods 102 eyes received CSLT and SD-OCT measurements and clinical examination on the same day. Visual field testing confirmed absence of glaucomatous defects. Statistical and correlation analysis was performed for rim area by CSLT as well as retinal nerve fibre layer thickness (RNFLT) and BMO minimal rim width (BMO-MRW) by SD-OCT. Results Mean disc size in CSLT was 2.98 +/- 0.4 mm(2) (range 2.45-4.23), mean rim area of 1.55 +/- 0.4 mm(2). BMO area was 2.51 +/- 0.33 mm2 (range 1.61-3.51), mean global RNFLT was 79.55 +/- 17.2 mu m, mean global BMO-MRW was 234.84 +/- 48.3 mu m. Correlation of BMO-MRW to global RNFLT was stronger (r=0.60, p<10(-5)) than correlation of CSLT rim area to global RNFLT (r=0.26, p=0.24). BMO-MRW and CSLT rim area correlated with r=0.59 (p<10(-5)). BMO-MRW and RNFLT significantly decreased with increasing age (p<0.001). Annual loss of BMO-MRW was 0.8 mu m/year (R-2=0.14, p<0.001), loss of RNFLT was 0.27 mu m/year (R-2=0.17, p=0.001). Conclusions In large optic discs, BMO-MRW is thinner compared with normal-sized discs and correlates better than CSLT parameters with the RNFLT. An age-depended loss of BMO-MRW needs to be taken into account in evaluation of the neuroretinal rim.