Factors Affecting Cone Photoreceptor Identification in RPGR-Associated Retinopathy

Abstract

Purpose: Cone photoreceptor quantification has been carried out in normal subjects using either manual or semi automated methods. Natural history studies and clinical trials of subjects with degenerative retinal diseases need to have a reliable and objective method of assessing the cone mosaic if this technique is to be incorporated into trial protocols. We present data on cone counting reliability in subjects with RPGR-associated retinopathy.

Methods: Fourteen subjects (8-51 years) with molecularly confirmed RPGR-associated retinopathy were imaged using a custom built adaptive optics scanning light ophthalmoscope (AOSLO) designed to simultaneously acquire images in confocal and split detection imaging modes. Cone photoreceptors were manually identified by two graders (G1 and G2) in the confocal and split detection images acquired at 42 parafoveal regions of interest. All 84 images were graded twice by each observer. Reliability of cone cell identification was compared between trials, between observers and between imaging modes. Statistical significance for all tests was assessed at the 5% level.

Results: Mean variance between trials was 9% for G1 and 16% for G2 across all images, however the mean number of photoreceptors identified was not statistically significantly different for either grader (p=0.1 for G1, p = 0.06 for G2, paired t-test). There was not a significant difference in mean variance between imaging modes (p=0.28, paired t-test). There was no statistically significant difference in mean variance between trials for confocal (5 ± 34%) or split detection (2 ± 6%).

Conclusions: While there was no statistical difference in mean variance between split detection and confocal images, the standard deviation of the variance was significantly different. Therefore split detection greatly facilitated the reliability of photoreceptor identification. The significant difference between graders will require further investigation to identify the underlying contributing factors, with refinement of training and certification methods likely to also help address this difference and thereby allow this image analysis modality to enter the reading center-based format for future large multi-center trials.