Response to: Gharbiya et al. (2014) J Alzheimers Dis 40, 907-917

13 November 2014

We are writing this letter in reference to the recent paper by Gharbiya and coworkers titled “Choroidal thinning as a new finding in Alzheimer's disease: evidence from enhanced depth imaging spectral domain optical coherence tomography” [1].

Optical coherence tomography (OCT) is widely accepted to measure structural changes in retinal layers for the assessment of retinal diseases such as glaucoma and diabetic macular edema in clinical practice. Beyond clinical utility and research studies in ocular diseases, OCT is also being increasingly used in clinical neurological studies to investigate changes in the retinal neuronal and axonal layers, such as the retinal nerve fiber layer (RNFL) [2-12]. With recent advances in OCT technology allowing better image acquisition of the posterior/deeper layers, as Gharbiya et al. have reported, thickness of the choroidal layer can be quantified as well, using enhanced depth imaging in spectral-domain OCT. In their study, the authors first reported a novel finding of significant reduction in choroid thickness in patients with Alzheimer’s disease (AD) compared with healthy controls. They also compared central subfield retinal thickness and average peripapillary RNFL thickness, reporting no significant changes in AD, inconsistent with previous reports.

In this letter, we would like to highlight that signal strength (or signal-to-noise ratio or image quality index), a quantifiable measure of image quality applied to the OCT scan, should be considered in the interpretation of RNFL measurements. The authors mention that the discrepancy in their results with previous literature may be due to patient dependent factors such as media opacity and optic disc anatomy (e.g., optic disc size), operator and instrument dependent factors, or magnification errors due to axial length, refractive error, and corneal curvature. However, a more fundamental, controllable, and yet grossly neglected factor that can directly impact measurements from OCT is signal strength.

RNFL measurements are significantly correlated to signal strength in the two main OCT systems in use, the time domain-OCT and spectral domain-OCT [13-15]. Studies using time domain-OCT have shown robust positive correlations between signal strength and RNFL thickness measurements [13, 16-18], which remain even after imposing minimum signal strength for scan selection [16, 17]. Although varying arbitrary cut-off values have been implemented as exclusion criteria in previous OCT studies in AD [7, 9, 12], this suggests that even in high quality scans, signal strength should still be taken into consideration when comparing RNFL thickness in any cohort. As RNFL thickness is likely to be underestimated at low signal strength, RNFL defects may be falsely identified, and this has been shown to influence glaucoma prediction [19]. Signal strength also affects the measurement of other optic nerve head (ONH) parameters, including cup volume and disc area, with signal strength positively correlated with measurements [20]. Even in the newer spectral domain-OCT system, the effect of signal strength on RNFL thickness measurements and interpretation of the RNFL thickness deviation map remain [15], with signal strength an important predictor of RNFL measurement variability [21]. In contrast, Rao et al. has shown that signal strength also has a significant effect on other ONH measurements, but not RNFL and macular parameters [22]. Nevertheless, these studies highlight that regardless of the OCT system used, signal strength remains an important factor influencing measurements.

Gharbiya et al. have compared their results against previous studies on RNFL thickness in AD (as seen in Table 4) [1]. Despite the established effect signal strength has on OCT thickness measurements, studies in both time domain-OCT [2-7, 11] and spectral domain-OCT [10, 23, 24] systems have neither incorporated signal strength into their exclusion criteria, nor adjusted for signal strength in statistical models. This suggests that differences in RNFL thicknesses between AD patients and controls are not directly comparable between studies, as signal strength has not been accounted for.

In summary, level of signal strength is an important factor that needs to be taken into consideration for interpretation of RNFL and ONH measurements, and possibly even in choroidal thickness measurements. This is especially important in demonstrating whether OCT is capable of detecting disease progression over time in future studies, and also in patients with AD who are more prone to poorer compliance during scan acquisition, easily compromising scan quality. Although many studies have reported RNFL thinning in AD patients, underestimation of RNFL thickness due to reduced signal strength can amplify the margin of error in RNFL measurements of AD patients. The failure to acknowledge the significant impact of signal strength on OCT measurements will compromise the quality of emerging research on neurodegenerative diseases.

Yi Lin Ong1, Yi Ting Ong1, Carol Y Cheung1, 2
1Singapore Eye Research Institute, Singapore National Eye Centre, Singapore
2Ophthalmology and Visual Sciences Academic Clinical Programme, Duke-NUS Graduate Medical School, National University of Singapore, Singapore

References
[1] Gharbiya M, Trebbastoni A, Parisi F, Manganiello S, Cruciani F, D'Antonio F, De Vico U, Imbriano L, Campanelli A, De Lena C (2014) Choroidal thinning as a new finding in Alzheimer's disease: evidence from enhanced depth imaging spectral domain optical coherence tomography. J Alzheimers Dis 40, 907-917.
[2] Berisha F, Feke GT, Trempe CL, McMeel JW, Schepens CL (2007) Retinal abnormalities in early Alzheimer's disease. Invest Ophthalmol Vis Sci 48, 2285-2289.
[3] Parisi V, Restuccia R, Fattapposta F, Mina C, Bucci MG, Pierelli F (2001) Morphological and functional retinal impairment in Alzheimer's disease patients. Clin Neurophysiol 112, 1860-1867.
[4] Iseri PK, Altinaş O, Tokay T, Yüksel N (2006) Relationship between cognitive impairment and retinal morphological and visual functional abnormalities in Alzheimer disease. J Neuroophthalmol 26, 18-24.
[5] Paquet C, Boissonnot M, Roger F, Dighiero P, Gil R, Hugon J (2007) Abnormal retinal thickness in patients with mild cognitive impairment and Alzheimer's disease. Neurosci Lett 420, 97-99.
[6] Lu Y, Li Z, Zhang X, Ming B, Jia J, Wang R, Ma D (2010) Retinal nerve fiber layer structure abnormalities in early Alzheimer's disease: evidence in optical coherence tomography. Neurosci Lett 480, 69-72.
[7] Kesler A, Vakhapova V, Korczyn AD, Naftaliev E, Neudorfer M (2011) Retinal thickness in patients with mild cognitive impairment and Alzheimer's disease. Clin Neurol Neurosurg 113, 523-526.
[8] Danesh-Meyer HV, Birch H, Ku JY, Carroll S, Gamble G (2006) Reduction of optic nerve fibers in patients with Alzheimer disease identified by laser imaging. Neurology 67, 1852-1854.
[9] Kirbas S, Turkyilmaz K, Anlar O, Tufekci A, Durmus M (2013) Retinal nerve fiber layer thickness in patients with Alzheimer disease. J Neuroophthalmol 33, 58-61.
[10] Marziani E, Pomati S, Ramolfo P, Cigada M, Giani A, Mariani C, Staurenghi G (2013) Evaluation of retinal nerve fiber layer and ganglion cell layer thickness in Alzheimer's disease using spectral-domain optical coherence tomography. Invest Ophthalmol Vis Sci 54, 5953-5958.
[11] Moschos MM, Markopoulos I, Chatziralli I, Rouvas A, Papageorgiou SG, Ladas I, Vassilopoulos D (2012) Structural and functional impairment of the retina and optic nerve in Alzheimer's disease. Curr Alzheimer Res 9, 782-788.
[12] Larrosa JM, Garcia-Martin E, Bambo MP, Pinilla J, Polo V, Otin S, Satue M, Herrero R, Pablo LE (2014) Potential new diagnostic tool for Alzheimer's disease using a linear discriminant function for Fourier domain optical coherence tomography. Invest Ophthalmol Vis Sci 55, 3043-3051.
[13] Cheung CY, Leung CK, Lin D, Pang CP, Lam DS (2008) Relationship between retinal nerve fiber layer measurement and signal strength in optical coherence tomography. Ophthalmology 115, 1347-1351, 1351 e1341-1342.
[14] Cheung CY, Chen D, Wong TY, Tham YC, Wu R, Zheng Y, Cheng CY, Saw SM, Baskaran M, Leung CK, Aung T (2011) Determinants of quantitative optic nerve measurements using spectral domain optical coherence tomography in a population-based sample of non-glaucomatous subjects. Invest Ophthalmol Vis Sci 52, 9629-9635.
[15] Cheung CY, Chan N, Leung CK (2012) Retinal nerve fiber layer imaging with spectral-domain optical coherence tomography: impact of signal strength on analysis of the RNFL map. Asia-Pacific J Ophthalmol 1, 19-23.
[16] Wu Z, Vazeen M, Varma R, Chopra V, Walsh AC, LaBree LD, Sadda SR (2007) Factors associated with variability in retinal nerve fiber layer thickness measurements obtained by optical coherence tomography. Ophthalmology 114, 1505-1512.
[17] Vizzeri G, Bowd C, Medeiros FA, Weinreb RN, Zangwill LM (2009) Effect of signal strength and improper alignment on the variability of stratus optical coherence tomography retinal nerve fiber layer thickness measurements. Am J Ophthalmol 148, 249-255 e241.
[18] Wu Z, Huang J, Dustin L, Sadda SR (2009) Signal strength is an important determinant of accuracy of nerve fiber layer thickness measurement by optical coherence tomography. J Glaucoma 18, 213-216.
[19] Sung KR, Wollstein G, Schuman JS, Bilonick RA, Ishikawa H, Townsend KA, Kagemann L, Gabriele ML, Advanced Imaging in Glaucoma Study G (2009) Scan quality effect on glaucoma discrimination by glaucoma imaging devices. Br J Ophthalmol 93, 1580-1584.
[20] Samarawickrama C, Pai A, Huynh SC, Burlutsky G, Wong TY, Mitchell P (2010) Influence of OCT signal strength on macular, optic nerve head, and retinal nerve fiber layer parameters. Invest Ophthalmol Vis Sci 51, 4471-4475
[21] Kim JH, Kim NR, Kim H, Lee ES, Seong GJ, Kim CY (2011) Effect of signal strength on reproducibility of circumpapillary retinal nerve fiber layer thickness measurement and its classification by spectral-domain optical coherence tomography. Jpn J Ophthalmol 55, 220-227.
[22] Rao HL, Kumar AU, Babu JG, Kumar A, Senthil S, Garudadri CS (2011) Predictors of normal optic nerve head, retinal nerve fiber layer, and macular parameters measured by spectral domain optical coherence tomography. Invest Ophthalmol Vis Sci 52, 1103-1110.
[23] Moreno-Ramos T, Benito-Leon J, Villarejo A, Bermejo-Pareja F (2013) Retinal nerve fiber layer thinning in dementia associated with Parkinson's disease, dementia with Lewy bodies, and Alzheimer's disease. J Alzheimers Dis 34, 659-664.
[24] Kromer R, Serbecic N, Hausner L, Froelich L, Aboul-Enein F, Beutelspacher SC (2014) Detection of retinal nerve fiber layer defects in Alzheimer's disease using SD-OCT. Front Psychiatry 5, 22.

Comments

We thank Drs. Ong and Cheung for their interest in our recently published paper [1]. We are evidently aware of the importance of image quality in the interpretation of optical coherence tomography (OCT) results. Indeed, as described in the Methods section, low quality OCT scans was an exclusion criterion for both patients and controls.

Heidelberg Spectralis optical coherence tomography (SD-OCT) (Heidelberg Engineering, Heidelberg, Germany, Spectralis software version 5.1.6.0, Eye Explorer Software 1.7.1.0) uses a scanning superluminescence diode to emit a scan beam with a wavelength of 870 nm to provide up to 40,000 A scans per second with a depth resolution of 7 μm and a transversal resolution of 14 μm. The instrument combines OCT technology with a confocal Scanning Laser Ophthalmoscope (Heidelberg Engineering, Heidelberg, Germany), which provides a reference infrared fundus image [2].

An internal nasal fixation light was used to center the disc in the incorporated circle-shaped scanning area. For OCT scanning, the SD-OCT device offers a particular characteristic of ‘freezing’ the retinal infrared fundus image, utilizing a real-time averaging algorithm (ART mode). Therefore, the circular scan can be moved within this stabilized optic disc image, in order to obtain exact scan centering and to adjust the number of frames (B scans) to average and enhance image quality for each image scan. In our study, this was achieved using a minimum of 90 to100 frames for each scan of the same scanning location during the peripapillary scanning process, thereby optimizing the signal-to-noise ratio and image quality significantly.

In cases where there was low image quality and the OCT software failed to detect the nerve fiber layer (failed automatic segmentation), obtained scans were eliminated, and measurements were repeated until excellent quality was reached.

Criteria for determining scan quality included 1) a properly aligned and a clear fundus image allowing visibility before and during image acquisition, 2) an image quality value (Q), which is a relative measure of signal strength, of 20 units (as shown by the internal signal strength control during the scanning process and in the image information window after the scan acquisition) or more with the activated ART mode, and 3) clear, well defined visualization of all retinal layers including the RPE, and the RNFL visible with no irregularities and a continuous scan pattern.

The OCT RNFL thickness scans were performed repeatedly by a single operator in order to obtain at least three high-quality scans. To optimize differentiation of retinal layers, scans were acquired in the high-resolution acquisition mode. By means of active eye tracking (TrueTrac), each peripapillary OCT scan was registered and locked to a reference image utilized by the ART mode. The incorporated True Trac technology of the instrument reduces image artifacts due to eye movements.

We thank Drs Ong and Cheung as they highlighted the critical importance of image quality in the interpretation of OCT measurements and gave us the opportunity to publish further details on the methodology used in our study.

Magda Gharbiya1 and Carlo De Lena2
1Department of Ophthalmology, Sapienza University of Rome, Umberto I University Hospital, Rome, Italy
2Department of Neurology, Sapienza University of Rome, Umberto I University Hospital, Rome, Italy

References
[1] Gharbiya M, Trebbastoni A, Parisi F, Manganiello S, Cruciani F, D'Antonio F, De Vico U, Imbriano L, Campanelli A, De Lena C (2014) Choroidal thinning as a new finding in Alzheimer's disease: evidence from enhanced depth imaging spectral domain optical coherence tomography. J Alzheimers Dis 40, 907-917.
[2] Serbecic N, Beutelspacher SC, Aboul-Enein FC, Kircher K, Reitner A, Schmidt-Erfurth U (2011) Reproducibility of high-resolution optical coherence tomography measurements of the nerve fibre layer with the new Heidelberg Spectralis optical coherence tomography. Br J Ophthalmol 95, 804-810.