J Neural Transm DOI 10.1007/s00702-015-1483-4
NEUROLOGY AND PRECLINICAL NEUROLOGICAL STUDIES - ORIGINAL ARTICLE
Retinal involvement in amyotrophic lateral sclerosis: a study with optical coherence tomography and diffusion tensor imaging Annemarie Hu¨bers1 • Hans Peter Mu¨ller1 • Jens Dreyhaupt2 • Kathrin Bo¨hm1 Florian Lauda1 • Hayrettin Tumani1 • Jan Kassubek1 • Albert C. Ludolph1 • Elmar H. Pinkhardt1
•
Received: 24 July 2015 / Accepted: 6 November 2015 Springer-Verlag Wien 2015
Abstract Although motor neuron degeneration is the predominant feature in ALS, recent data point to a more widespread pathology also comprising non-motor symptoms. Retinal thinning has been reported in a variety of neurodegenerative conditions. Yet, studies of retinal involvement in ALS are sparse and results are heterogeneous. We studied retinal alterations in ALS using a systematic approach combining Optical Coherence Tomography (OCT), Diffusion Tensor Imaging (DTI) and clinical phenotyping. We hypothesized that selective changes of specific retinal layers may be a reflection of overall neurodegeneration as measured by DTI. Spectral domain OCT images were analyzed to calculate the average thickness of retinal layers in 71 ALS patients and 20 controls. In 30 patients, the region of interest (ROI) based fractional anisotrophy (FA) was measured in the corticospinal tract (CST), as this region is preferentially affected by motor neuron degeneration. Clinical data were collected for correlation analysis. Patients showed a significant thinning of the inner nuclear layer (INL; p = 0.04) and the retinal nerve fibre layer (RNFL; p = 0.004) compared to controls. We saw significant correlations between retinal thickness and FA values of the CST in patients (p = 0.005). No significant correlation between clinical parameters and retinal involvement was observed. Our study provides evidence for a retinal involvement in ALS. Interestingly, ALS patients show a reduction in FA of the & Elmar H. Pinkhardt [email protected] 1
Department of Neurology, Ulm University, Oberer Eselsberg 45, 89081 Ulm, Germany
2
Institute of Epidemiology and Medical Biometry, Ulm University, Schwabstraße 13, 89075 Ulm, Germany
CST, which is correlated to retinal thinning. We conclude that retinal involvement is in fact associated to overall neurodegeneration and may be regarded as a potential technical biomarker in ALS. Keywords Amyotrophic lateral sclerosis Neurodegeneration Technical biomarker Optical coherence tomography Retina
Introduction Amyotrophic lateral sclerosis (ALS) is characterized by rapidly progressive neurodegeneration with involvement of the upper and lower motor neurons, leading to death with a mean survival time of approximately 3 years (Kiernan et al. 2011). Yet, neuropathology as well as the clinical phenotype of ALS goes beyond motoneuron degeneration, comprising widespread areas of the extramotor brain. As in many neurodegenerative diseases, ALS shows a characteristic spreading pattern of its underlying pathology across specific brain regions with disease progression that can be described in four stages (Brettschneider et al. 2013). In search of in vivo markers for early diagnosis and disease progression, numerous imaging-techniques have been conducted, including functional and structural magnetic resonance imaging (MRI) and positron emission tomography (PET) (Chio et al. 2014) as well as neuropsychological examinations (Lule´ et al. 2014). Kassubek et al. (2014) could retrace the neuropathological disease stages of ALS by hands of tract of interest based diffusion tensor imaging (DTI). These results mirror the proposed neuropathological propagation pattern of ALS, supporting in vivo the evidence of the progressive expansion of WM damage from the motor to extramotor networks (Agosta et al. 2015).
123
A. Hu¨bers et al.
Although ocular involvement is not seen to be a dominant feature of ALS, there is now growing evidence of an involvement of the visual pathway, comprising neuroophthalmological abnormalities such as decreased high and low contrast visual acuity (Sharma et al. 2011; Moss et al. 2012). In a histopathological examination, Fawzi et al. (2014) reported two ALS patients with C9orf72 mutations showing cone bipolar cell involvement within the inner nuclear layer (INL) of the retina. Optical coherence tomography (OCT) has been proven to detect subtle retinal changes in neurodegenerative diseases like Parkinson’s Disease (PD), other neurodegenerative disorders with parkinsonism, such as multisystem atrophy (MSA) and progressive supranuclear palsy (PSP) as well as Alzheimer’s Disease (AD) (Kesler et al. 2011; Schneider et al. 2014; Mu¨ller et al. 2014). Hence, retinal pathology and especially selective changes of retinal layers may be regarded as a reflection of the underlying neurodegenerative processes. Since the retina and optic nerve extend from the diencephalon, they are considered to be part of the central nervous system (London et al. 2013). According to the literature available, data about in vivo imaging of the retina in ALS by OCT are sparse and the results are heterogeneous. Roth et al. (2013) did not find a significant difference between ALS patients and healthy controls in any of the examined OCT measures. Moreover, OCT parameters showed no correlation with clinical measures of disease severity. In a cohort of 24 ALS patients, Ringelstein et al. (2014) demonstrated a subtle reduction in the macular thickness and the retinal nerve fiber layer (RNFL) as well as a marked thinning of the inner nuclear layer (INL), though there was no correlation of OCT findings with clinical motor deficits. We hypothesized that the examination of the single retinal layer average perifoveal thickness is suitable to detect specific retinal changes in ALS. In a multimodal approach with OCT and diffusion tensor imaging (DTI) analysis of fractional anisotropy (FA) (Mu¨ller and Kassubek 2013), we studied in a cohort of 71 ALS patients if there are correlations to (a) the clinical phenotype and (b) to structural white
matter alterations of regions that are related to the disease pathology. We chose the corticospinal tract (CST) as it is mainly and early affected in the course of the disease and shows typical alterations in ALS patients as revealed by DTI (Agosta et al. 2010a; Mu¨ller and Kassubek 2013).
Materials and methods Subjects and clinical characterization Detailed clinical characteristics and demographic features of all participants are summarized in Table 1. All patients were recruited at the Department of Neurology at the University Hospital of Ulm, Germany, in a prospective manner between April 2012 until February 2014. OCT and neurological assessments were obtained from 71 patients with clinically definite or probable ALS and 20 age- and gender-matched healthy individuals. 50 patients showed a spinal and 21 patients a bulbar onset of the disease. All patients and controls were asked to also participate in the imaging part of the study. 30 patients and 18 controls agreed to undergo additional MRI-DTI analysis. The interval between OCT and MRI examinations was not more than 2 weeks for all patients and controls. All participating patients and controls provided written informed consent for the OCT and MRI protocol according to institutional guidelines. The study was approved by the Ethical Committee of the University of Ulm (application number 36/13) and performed in accordance with the ethical standards laid out by the Declaration of Helsinki. ALS was diagnosed by board-certified neurologists specialized in motor neuron disease if patients met the revised El Escorial diagnostic criteria (Brooks et al. 2000). Exclusion criteria were other neurodegenerative disorders, including dementia, autoimmune inflammatory diseases of the central nervous system and any patient or control with any known eye disease (e.g., history of retinopathy, glaucoma, or high myopia). Subjects with a history of diabetes were excluded from the study as well.
Table 1 Subject demographics and clinical characterization Patients with ALS
Patients with spinal onset
Patients with bulbar onset
Healthy controls
p-value
Number
71
50
21
20
NA
Sex (male:female)
35:36
21:29
14:7
10:10
0.576
Age (years) Disease duration (months)
61 (28–84) 12 (2–98)
59 (30–84) 12 (4–98)
69 (28–73) 11 (2–79)
64 (53–80) NA
0.265 NA
Age of onset (years)
59 (22–83)
58 (29–83)
68 (22–72)
NA
NA
ALSFRS-Rc
40 (16–48)
40 (16–48)
40 (32–46)
NA
NA
Data are shown as median (interquartile range), min–max. a Fisher’s exact test, b Mann-Whitney-U-test, c ALSFRS-R, revised ALS Functional Rating Scale (maximum score 48) (Cedarbaum et al. 1999), NA not applicable
123
Retinal involvement in amyotrophic lateral sclerosis: a study with optical coherence…
OCT acquisition and postprocessing
outer plexiform layer (OPL), outer nuclear layer plus the photoreceptors (ONL?PR) and for the whole retina (whole retina thickness, WRT). The outer border of the ONL ? PR was set to the retinal pigment epithelium (RPE).
Horizontal, transfoveal images were obtained from each participant with a Heidelberg Spectralis HD OCT (Heidelberg Engineering GmbH, Heidelberg, Germany; Software version 5.1.1.6). By hands of TruTrack image alignment eye tracking, perifoveal volumetric retinal scans (25 single vertical axial scans, scanning area: 6 mm, centered at the fovea) were obtained. Scans for volumetric measurements were averaged from 13 images (Automatic Real Time, ART). Scans that did not pass the OSCAR-IB criteria for retinal OCT quality assessment were excluded from the analysis (Tewarie et al. 2012). Data analysis was conducted by the SOCIA (Semi-automatic Optic Coherence Image Analysis) software based on the tensor imaging and fiber tracking (TIFT) software (Schneider et al. 2014; Muller et al. 2007). A semi-automatic algorithm applies recognition of gray-level thresholds to separate the retinal layers in a single high-resolution horizontal scan through the middle of the fovea (Schneider et al. 2014). To correct for measurement errors caused by off-center measuring, each image had to be realigned horizontally and cut by approximately 25 % to a length of 4.5 mm (i.e., 2.25 mm each in the temporal and nasal direction, Fig. 1b). The analysis software is device independent and can be used with any existing OCT machine if a high-resolution bitmap export of the OCT scan is possible. Average thickness was calculated for RNFL, ganglion cell layer plus inner plexiform layer (GCL?IPL), INL,
Thirty patients and 18 control subjects underwent MRI on a 1.5 T clinical scanner (Magnetom Symphony, Siemens Medical). The DTI study protocol consisted of 2 9 31 gradient directions, including two b = 0 gradient directions (64 slices, 64 9 64 pixels). The slice thickness was 3.0 mm, in-plane pixel size was 3.3 mm 9 3.3 mm. The echo time and repetition time were 28 and 3080 ms, respectively; b was 1000 s/mm2. The DTI analysis software Tensor Imaging and Fiber Tracking (TIFT) (Muller et al. 2007) was used for postprocessing and statistical analysis. In a first step, nonlinear spatial normalization to the Montreal Neurological Institute (MNI) stereotaxic standard space (Brett et al. 2002) was performed using study-specific templates. For details of the normalization procedure, see Mu¨ller and Kassubek (2013). The whole cross-sectional group comparison analysis followed the procedure described in detail previously (Mu¨ller and Kassubek 2013). FA maps were calculated from MNI-normalized DTI data, and a Gaussian smoothing filter of 8 mm FWHM was applied to the individual normalized FA maps. Next, voxelwise statistical comparison
B
RNFL GCL + IPL INL OPL ONL+PR
ganglion cell layer, IPL: inner plexiform layer, INL: inner nuclear layer, OPL: outer plexiform layer, ONL+PR: outer nuclear layer + photoreceptors
Fig. 1 a Placement of spherical regions of interest (ROI) for averaged fractional anisotropy (FA)-value calculation (MNI ? -12/-20/-2, ROI-radius was 5 mm). ROIs were visualized on a background that shows significant differences in comparisons at the group level between ALS-patients and controls (whole brain-based
whole retinal thickness
A
Magnetic resonance imaging
positive p
NEUROLOGY AND PRECLINICAL NEUROLOGICAL STUDIES - ORIGINAL ARTICLE
Retinal involvement in amyotrophic lateral sclerosis: a study with optical coherence tomography and diffusion tensor imaging Annemarie Hu¨bers1 • Hans Peter Mu¨ller1 • Jens Dreyhaupt2 • Kathrin Bo¨hm1 Florian Lauda1 • Hayrettin Tumani1 • Jan Kassubek1 • Albert C. Ludolph1 • Elmar H. Pinkhardt1
•
Received: 24 July 2015 / Accepted: 6 November 2015 Springer-Verlag Wien 2015
Abstract Although motor neuron degeneration is the predominant feature in ALS, recent data point to a more widespread pathology also comprising non-motor symptoms. Retinal thinning has been reported in a variety of neurodegenerative conditions. Yet, studies of retinal involvement in ALS are sparse and results are heterogeneous. We studied retinal alterations in ALS using a systematic approach combining Optical Coherence Tomography (OCT), Diffusion Tensor Imaging (DTI) and clinical phenotyping. We hypothesized that selective changes of specific retinal layers may be a reflection of overall neurodegeneration as measured by DTI. Spectral domain OCT images were analyzed to calculate the average thickness of retinal layers in 71 ALS patients and 20 controls. In 30 patients, the region of interest (ROI) based fractional anisotrophy (FA) was measured in the corticospinal tract (CST), as this region is preferentially affected by motor neuron degeneration. Clinical data were collected for correlation analysis. Patients showed a significant thinning of the inner nuclear layer (INL; p = 0.04) and the retinal nerve fibre layer (RNFL; p = 0.004) compared to controls. We saw significant correlations between retinal thickness and FA values of the CST in patients (p = 0.005). No significant correlation between clinical parameters and retinal involvement was observed. Our study provides evidence for a retinal involvement in ALS. Interestingly, ALS patients show a reduction in FA of the & Elmar H. Pinkhardt [email protected] 1
Department of Neurology, Ulm University, Oberer Eselsberg 45, 89081 Ulm, Germany
2
Institute of Epidemiology and Medical Biometry, Ulm University, Schwabstraße 13, 89075 Ulm, Germany
CST, which is correlated to retinal thinning. We conclude that retinal involvement is in fact associated to overall neurodegeneration and may be regarded as a potential technical biomarker in ALS. Keywords Amyotrophic lateral sclerosis Neurodegeneration Technical biomarker Optical coherence tomography Retina
Introduction Amyotrophic lateral sclerosis (ALS) is characterized by rapidly progressive neurodegeneration with involvement of the upper and lower motor neurons, leading to death with a mean survival time of approximately 3 years (Kiernan et al. 2011). Yet, neuropathology as well as the clinical phenotype of ALS goes beyond motoneuron degeneration, comprising widespread areas of the extramotor brain. As in many neurodegenerative diseases, ALS shows a characteristic spreading pattern of its underlying pathology across specific brain regions with disease progression that can be described in four stages (Brettschneider et al. 2013). In search of in vivo markers for early diagnosis and disease progression, numerous imaging-techniques have been conducted, including functional and structural magnetic resonance imaging (MRI) and positron emission tomography (PET) (Chio et al. 2014) as well as neuropsychological examinations (Lule´ et al. 2014). Kassubek et al. (2014) could retrace the neuropathological disease stages of ALS by hands of tract of interest based diffusion tensor imaging (DTI). These results mirror the proposed neuropathological propagation pattern of ALS, supporting in vivo the evidence of the progressive expansion of WM damage from the motor to extramotor networks (Agosta et al. 2015).
123
A. Hu¨bers et al.
Although ocular involvement is not seen to be a dominant feature of ALS, there is now growing evidence of an involvement of the visual pathway, comprising neuroophthalmological abnormalities such as decreased high and low contrast visual acuity (Sharma et al. 2011; Moss et al. 2012). In a histopathological examination, Fawzi et al. (2014) reported two ALS patients with C9orf72 mutations showing cone bipolar cell involvement within the inner nuclear layer (INL) of the retina. Optical coherence tomography (OCT) has been proven to detect subtle retinal changes in neurodegenerative diseases like Parkinson’s Disease (PD), other neurodegenerative disorders with parkinsonism, such as multisystem atrophy (MSA) and progressive supranuclear palsy (PSP) as well as Alzheimer’s Disease (AD) (Kesler et al. 2011; Schneider et al. 2014; Mu¨ller et al. 2014). Hence, retinal pathology and especially selective changes of retinal layers may be regarded as a reflection of the underlying neurodegenerative processes. Since the retina and optic nerve extend from the diencephalon, they are considered to be part of the central nervous system (London et al. 2013). According to the literature available, data about in vivo imaging of the retina in ALS by OCT are sparse and the results are heterogeneous. Roth et al. (2013) did not find a significant difference between ALS patients and healthy controls in any of the examined OCT measures. Moreover, OCT parameters showed no correlation with clinical measures of disease severity. In a cohort of 24 ALS patients, Ringelstein et al. (2014) demonstrated a subtle reduction in the macular thickness and the retinal nerve fiber layer (RNFL) as well as a marked thinning of the inner nuclear layer (INL), though there was no correlation of OCT findings with clinical motor deficits. We hypothesized that the examination of the single retinal layer average perifoveal thickness is suitable to detect specific retinal changes in ALS. In a multimodal approach with OCT and diffusion tensor imaging (DTI) analysis of fractional anisotropy (FA) (Mu¨ller and Kassubek 2013), we studied in a cohort of 71 ALS patients if there are correlations to (a) the clinical phenotype and (b) to structural white
matter alterations of regions that are related to the disease pathology. We chose the corticospinal tract (CST) as it is mainly and early affected in the course of the disease and shows typical alterations in ALS patients as revealed by DTI (Agosta et al. 2010a; Mu¨ller and Kassubek 2013).
Materials and methods Subjects and clinical characterization Detailed clinical characteristics and demographic features of all participants are summarized in Table 1. All patients were recruited at the Department of Neurology at the University Hospital of Ulm, Germany, in a prospective manner between April 2012 until February 2014. OCT and neurological assessments were obtained from 71 patients with clinically definite or probable ALS and 20 age- and gender-matched healthy individuals. 50 patients showed a spinal and 21 patients a bulbar onset of the disease. All patients and controls were asked to also participate in the imaging part of the study. 30 patients and 18 controls agreed to undergo additional MRI-DTI analysis. The interval between OCT and MRI examinations was not more than 2 weeks for all patients and controls. All participating patients and controls provided written informed consent for the OCT and MRI protocol according to institutional guidelines. The study was approved by the Ethical Committee of the University of Ulm (application number 36/13) and performed in accordance with the ethical standards laid out by the Declaration of Helsinki. ALS was diagnosed by board-certified neurologists specialized in motor neuron disease if patients met the revised El Escorial diagnostic criteria (Brooks et al. 2000). Exclusion criteria were other neurodegenerative disorders, including dementia, autoimmune inflammatory diseases of the central nervous system and any patient or control with any known eye disease (e.g., history of retinopathy, glaucoma, or high myopia). Subjects with a history of diabetes were excluded from the study as well.
Table 1 Subject demographics and clinical characterization Patients with ALS
Patients with spinal onset
Patients with bulbar onset
Healthy controls
p-value
Number
71
50
21
20
NA
Sex (male:female)
35:36
21:29
14:7
10:10
0.576
Age (years) Disease duration (months)
61 (28–84) 12 (2–98)
59 (30–84) 12 (4–98)
69 (28–73) 11 (2–79)
64 (53–80) NA
0.265 NA
Age of onset (years)
59 (22–83)
58 (29–83)
68 (22–72)
NA
NA
ALSFRS-Rc
40 (16–48)
40 (16–48)
40 (32–46)
NA
NA
Data are shown as median (interquartile range), min–max. a Fisher’s exact test, b Mann-Whitney-U-test, c ALSFRS-R, revised ALS Functional Rating Scale (maximum score 48) (Cedarbaum et al. 1999), NA not applicable
123
Retinal involvement in amyotrophic lateral sclerosis: a study with optical coherence…
OCT acquisition and postprocessing
outer plexiform layer (OPL), outer nuclear layer plus the photoreceptors (ONL?PR) and for the whole retina (whole retina thickness, WRT). The outer border of the ONL ? PR was set to the retinal pigment epithelium (RPE).
Horizontal, transfoveal images were obtained from each participant with a Heidelberg Spectralis HD OCT (Heidelberg Engineering GmbH, Heidelberg, Germany; Software version 5.1.1.6). By hands of TruTrack image alignment eye tracking, perifoveal volumetric retinal scans (25 single vertical axial scans, scanning area: 6 mm, centered at the fovea) were obtained. Scans for volumetric measurements were averaged from 13 images (Automatic Real Time, ART). Scans that did not pass the OSCAR-IB criteria for retinal OCT quality assessment were excluded from the analysis (Tewarie et al. 2012). Data analysis was conducted by the SOCIA (Semi-automatic Optic Coherence Image Analysis) software based on the tensor imaging and fiber tracking (TIFT) software (Schneider et al. 2014; Muller et al. 2007). A semi-automatic algorithm applies recognition of gray-level thresholds to separate the retinal layers in a single high-resolution horizontal scan through the middle of the fovea (Schneider et al. 2014). To correct for measurement errors caused by off-center measuring, each image had to be realigned horizontally and cut by approximately 25 % to a length of 4.5 mm (i.e., 2.25 mm each in the temporal and nasal direction, Fig. 1b). The analysis software is device independent and can be used with any existing OCT machine if a high-resolution bitmap export of the OCT scan is possible. Average thickness was calculated for RNFL, ganglion cell layer plus inner plexiform layer (GCL?IPL), INL,
Thirty patients and 18 control subjects underwent MRI on a 1.5 T clinical scanner (Magnetom Symphony, Siemens Medical). The DTI study protocol consisted of 2 9 31 gradient directions, including two b = 0 gradient directions (64 slices, 64 9 64 pixels). The slice thickness was 3.0 mm, in-plane pixel size was 3.3 mm 9 3.3 mm. The echo time and repetition time were 28 and 3080 ms, respectively; b was 1000 s/mm2. The DTI analysis software Tensor Imaging and Fiber Tracking (TIFT) (Muller et al. 2007) was used for postprocessing and statistical analysis. In a first step, nonlinear spatial normalization to the Montreal Neurological Institute (MNI) stereotaxic standard space (Brett et al. 2002) was performed using study-specific templates. For details of the normalization procedure, see Mu¨ller and Kassubek (2013). The whole cross-sectional group comparison analysis followed the procedure described in detail previously (Mu¨ller and Kassubek 2013). FA maps were calculated from MNI-normalized DTI data, and a Gaussian smoothing filter of 8 mm FWHM was applied to the individual normalized FA maps. Next, voxelwise statistical comparison
B
RNFL GCL + IPL INL OPL ONL+PR
ganglion cell layer, IPL: inner plexiform layer, INL: inner nuclear layer, OPL: outer plexiform layer, ONL+PR: outer nuclear layer + photoreceptors
Fig. 1 a Placement of spherical regions of interest (ROI) for averaged fractional anisotropy (FA)-value calculation (MNI ? -12/-20/-2, ROI-radius was 5 mm). ROIs were visualized on a background that shows significant differences in comparisons at the group level between ALS-patients and controls (whole brain-based
whole retinal thickness
A
Magnetic resonance imaging
positive p
