Diffusion Tensor Imaging and Neurite Orientation Dispersion and Density Imaging Assessment of Optic Pathway Function in Patients With Anterior Visual Pathway Compression

In patients with sellar or parasellar tumors, it is crucial to evaluate visual field impairment in the pre-operative stage and to predict visual field improvement after the surgery. The purpose of this study was to investigate the associations of diffusion tensor imaging (DTI) and neurite orientation dispersion and density imaging (NODDI) parameters in the optic radiations with preoperative and post-operative visual field impairment.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Diffusion Tensor Imaging and Neurite Orientation Dispersion and Density Imaging Assessment of Optic Pathway Function in Patients With Anterior Visual Pathway Compression Koung Mi Kang, MD, Eun-Jung Choi, MSc, Woojin Jung, PhD, Jongho Lee, PhD, Seung Hong Choi, MD, PhD, Yong Hwy Kim, MD, PhD Background: In patients with sellar or parasellar tumors, it is crucial to evaluate visual field impairment in the preoperative stage and to predict visual field improvement after the surgery. The purpose of this study was to investigate the associations of diffusion tensor imaging (DTI) and neurite orientation dispersion and density imaging (NODDI) parameters in the optic radiations with preoperative and postoperative visual field impairment. Methods: This prospective study included 81 participants with sellar or parasellar tumors. Multishell diffusion imaging and a visual field impairment score (VFIS) were acquired before and after the surgery. The multishell diffusionweighted imaging was acquired to measure the neurite density and neurite orientation dispersion, as well as the diffusivity. DTI parameters were fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity, and radial diffusivity (RD). NODDI provided intracellular volume fraction (Vic), the orientation dispersion index, and isotropic volume fraction (Viso). The associations of DTI and NODDI parameters in the Department of Radiology (KMK, SHC), Seoul National University Hospital, Jongno-gu, Republic of Korea; Department of Electrical and Computer Engineering (E-JC), Laboratory for Imaging Science and Technology, Seoul National University, Gwanak-gu, Republic of Korea; AIRS Medical (WJ), Seoul, Republic of Korea; Department of Electrical and Computer Engineering (JL), Laboratory for Imaging Science and Technology, INMC, IOER, Seoul National University, Gwanak-gu, Republic of Korea and Department of Neurosurgery (YHK), Seoul National University Hospital, Jongno-gu, Republic of Korea. Supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science, ICT and Future Planning (2017R1A2B2008412), MSIT (NRF-2018R1A4A1025891), the Bio and Medical Technology Development Program MSIP (2015M3A9A7029740), and SEPRI at Seoul National University. The authors report no conflicts of interest. K. M. Kang and E.-J. Choi contributed equally to this work. Address correspondence to Yong Hwy Kim, MD, PhD, Department of Neurosurgery, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea; E-mail: kimyh96@ gmail.com e192 optic radiations with VFIS were investigated, adjusting for age, tumor height, and symptom duration. Results: Among 162 optic radiations, 117 were functionally impaired in the preoperative stage. FA and Vic had significant negative correlations, whereas MD and RD had significant positive correlations with the VFIS (all P , 0.001). In the preoperative stage, lower FA (P = 0.001; odds ratio = 0.750) and Vic (P = 0.003; OR = 0.827) and higher MD (P = 0.007; OR = 1.244) and RD (P , 0.001; OR = 1.361) were significantly associated with the presence of visual field impairment. For the degree of postoperative improvement, preoperative lower Vic (P = 0.034; OR = 0.910) and higher MD (P = 0.037; OR = 1.103) and RD (P = 0.047; OR = 1.090) were significantly associated with more postoperative improvement. Conclusions: DTI and NODDI parameters in the optic radiations were correlated with VFIS and associated with preoperative visual field impairments and postoperative improvement. It may help in predicting visual field improvement after the surgery in patients with sellar or parasellar tumors. Journal of Neuro-Ophthalmology 2022;42:e192–e202 doi: 10.1097/WNO.0000000000001309 © 2021 by North American Neuro-Ophthalmology Society V isual field defects are one of the most important clinical morbidities of sellar or parasellar tumors compressing the optic chiasm. Hence, it is important to evaluate visual field impairment in the preoperative stage and to predict visual field improvement after the surgery. Several previous studies have reported that diffusion tensor imaging (DTI) could assess injury in the anterior visual pathways that compressed pituitary adenoma (1), the impact of surgery on the optic chiasm and nerve in patients with intrasellar or parasellar mass lesions (2), and visual outcome after the surgery in patients with suprasellar tumors (3). However, those studies almost exclusively focused on the anterior visual pathway. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Recently, changes in DTI parameters in the optic radiation were associated with visual impairment in patients with anterior visual pathway lesions. The fractional anisotropy (FA) of the optic radiation significantly correlated with visual impairment in patients with optic pathway glioma (1,2) and reduced retinal nerve fiber layer thickness (1). In this regard, anterograde transsynaptic degeneration from disruptions in visual input is a likely mechanism of these ‟downstream” effects from compression of the optic chiasm (4). DTI is a well-established technique to measure water diffusion in white matter and provides quantitative parameters; FA, mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) (3). However, as it depicts only a single fiber population within a voxel, the parameters are inherently nonspecific. For example, a decrease in FA can be caused by a decrease in neurite density, an increase in the dispersion of neurite orientation distribution, and other tissue microstructural changes (5). On the other hand, neurite orientation dispersion and density imaging (NODDI) assumes a 3-compartment model, including intracellular, extracellular, and cerebrospinal fluid spaces. For NODDI scan, multiple high b-values with diffusion gradient directions are required, such as b = 700 sec/mm2 with 30directions and b = 2000 sec/mm2 with 60 directions (5). NODDI estimates intracellular volume fraction (Vic) reflecting neurite density, the orientation dispersion index (ODI) indicating neurite orientation dispersion, and isotropic volume fraction (Viso) (5). We hypothesized that the compression of the anterior visual pathway was associated with microstructural changes in the optic radiations. The purpose was to investigate whether DTI and NODDI parameters in the optic radiations were associated with visual field impairments in the preoperative stage and visual field improvement after the surgery. METHODS Our institutional review board approved this prospective study. All participants provided written informed consent. Study Participants Ninety-seven participants with sellar and parasellar tumors, scheduled for trans-sphenoidal tumor resection, were recruited from July 2017 through October 2018. The inclusion criteria were older than 18 years and no other known visual impairment and intracranial pathology, including significant trauma. MRI scanning and ophthalmologic examination were performed before and 79–122 days after the surgery (mean = 95.8 days). The exclusion criteria included inadequate MR image quality (n = 14) and declining to participate in the study (n = 2). Finally, 81 participants (40 men and 41 women; mean age of 48.8 ± 15.4 years, ranging from 19 to 77 years) were included in this study. Symptom duration was based on the patient’s history. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Visual Field Impairment Score To evaluate the exact visual field dominated by each optic radiation, a novel quantitative scoring system, a visual field impairment score (VFIS), was blindly rated by a neurosurgeon (Y.H.K. with 15 years of experience) without MRI data. The VFIS expresses the visual field defect area reflecting the function of each optic tract based on the results of the Goldmann or Humphrey perimetry (Fig. 1). The score ranged from zero to 8, with zero representing a normal visual field and 8 indicating that all parts of the unilateral optic tract had visual field defects. For example, 5 points of right VFIS means that there are 5 of 8 defects in the left visual field, which corresponds to the function of the right optic radiation. Because each patient’s right and left optic radiations were analyzed separately, our study included a total of 162 optic radiations from 81 participants. Imaging Acquisition and Processing Multishell diffusion-weighted imaging was performed using a 3T Skyra scanner (Siemens, Germany) with a 64-channel head coil. Two-dimensional echo planar imaging was acquired with the following parameters: repetition time = 3,000 milliseconds, echo time = 105 milliseconds, field of view = 240 · 240 mm2, voxel size = 1.5 · 1.5 mm2, slice thickness = 4 mm, b = 300 s/mm2 with 9 diffusion directions, b = 700 s/mm2 with 36 directions, b = 2000 s/mm2 with 72 directions, and total acquisition time = 6 min 42 seconds. The parameter b = 0 s/mm2 was included in each scan. An additional opposite phase encoding direction scan with b = 0 s/mm2 was acquired. DTI (FA, MD, AD, and RD) processing was performed by FMRIB Software Library version 5.0.8 (FMIRB analysis group, United Kingdom). The processing steps included brain extraction (6), eddy current correction (7), susceptibility-induced field correction (8), and DTI tensor fitting. NODDI parameters were processed using Accelerated Microstructure Imaging via Convex Optimization (9). The preoperative and postoperative MD images were linearly registered using FMRIB’s linear image registration tool. The transformation matrix was applied to the other postoperative images. The Medical Imaging Interaction Toolkit workbench (German Cancer Research Center, Germany) was used for manually delineating optic radiations (E.J.C. with 9 years of experience in neuroimaging analysis). The slice was selected under the second or third slice from the lower boundary of the corpus callosum on the sagittal view. The bilateral optic radiations in the axial view clearly appeared in the selected slices of both preoperative and postoperative FA maps. The registered postoperative images showed subtle deformations compared with the preoperative images due to tumor resection; thus, the optic radiation masks in the postoperative FA maps were adjusted based on those of the preoperative images (Fig. 2). The mask was applied to the other quantitative maps. e193 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Visual field impairment score (VFIS). A. Each optic tract and visual cortex is composed of and received the neural fibers from both eyes corresponding to the contralateral visual field. B. Authors divided the visual field of each eye into 8 pieces to resemble the visual field defect pattern found in the intracranial visual pathway damage (dot line) and the defects of each pieced visual field was scored to 1 of VFIS. Therefore, the homonymous hemianopsia was scored to 8 of VFIS of contralateral visual tract. Tumor size was measured the longest diameter on the sagittal plane of contrast-enhanced T1 weighted images. Statistical Analysis Intraobserver reliability for the optic radiations segmentation was assessed using the intraclass correlation coefficient. This coefficient was measured from preoperative and postoperative data of 30 randomly selected participants (60 optic radiations) at 4-month intervals. Intraobserver reliability for the measurement of VFIS was assessed using the intraclass correlation coefficient. This coefficient was measured from preoperative and postoperative data of 162 VFIS of all participants. FIG. 2. Representative DTI and NODDI parameter maps. ROI was drawn on each side of the optic radiation on the FA map. AD, axial diffusivity; DTI, diffusion tensor imaging; FA, fractional anisotropy; MD, mean diffusivity; NODDI, neurite orientation dispersion and density imaging; ODI, orientation dispersion index; RD, radial diffusivity; Vic, intracellular volume fraction; Viso, isotropic volume fraction. e194 Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Statistical analyses were performed using SPSS 25.0 (IBM Corp). A P value of ,0.05 was considered statistically significant. TABLE 1. Clinical characteristics Participants Sex (M/F) Age (yrs, mean ± SD) Diagnosis Functioning pituitary adenoma Nonfunctioning pituitary adenoma Acromegaly Prolactinoma Craniopharyngioma (recurred) TS meningioma (recurred) Rathke cleft cyst Epidermoid cyst Preoperative VFIS Postoperative VFIS Symptom duration, mo Tumor size, cm 40/41 48.8 ± 15.3 60 54 4 2 10 (1) 6 (1) 3 2 2.6 ± 2.4 0.8 ± 1.6 6±8 2.5 ± 0.9 Data for VFIS, symptom duration and tumor size are the means ± SD. VFIS, visual field impairment score. The age, tumor size, and symptom duration were adjusted in partial correlation and logistic regression analyses (10–12). Partial correlations were used to investigate the association of the DTI and NODDI parameters with the VFIS. DVFIS, which represented changes in visual fields after the surgery, was calculated as the postoperative VFIS subtracted from the preoperative VFIS. Paired t tests were performed to compare the preoperative and postoperative data. Binary logistic regressions were used to investigate the associations of the DTI and NODDI parameters with preoperative visual field impairment. The association of the preoperative DTI and NODDI parameters and the degree of visual field improvement after the surgery were evaluated using ordinal logistic regression with 3 ordinal outcomes (no improvement or worsening after the surgery, DVFIS = 0–4; mild improvement after the surgery, DVFIS = 24 w 21; and marked improvement after the surgery, DVFIS = 28 w 25). RESULTS The clinical characteristics of the participants are described in Table 1. Among 162 optic radiations, 117 optic radiations were associated with visual field impairment in the preoperative stage. After the surgery, the VFIS decreased in 100 optic radiations (improved), was unchanged in 14, and increased in 3 after the surgery (worsened) (Fig. 3). The overall intraobserver agreement was excellent. The preoperative intraclass correlation coefficient values were 0.987 for FA (95% confidence interval [CI], 0.979–0,992), 0.986 for MD (95% CI, 0.977–0.992), 0.969 for AD (95% CI, 0.948–0.981), 0.993 for RD (95% CI, 0.988–0.996), 0.994 for Vic (95% CI, 0.990–0.996), 0.892 for Viso (95% CI, 0.819–0.935), and 0.946 for ODI (95% CI, 0.910– 0.968). The postoperative intraclass correlation coefficient values were 0.985 for FA (95% CI, 0.975–0.991), 0.975 for MD (95% CI, 0.959–0.985), 0.978 for AD (95% CI, 0.964–0.987), 0.979 for RD (95% CI, 0.965–0.988), 0.980 for Vic (95% CI, 0.966–0.988), 0.964 for Viso (95% CI, 0.940–0.978), and 0.942 for ODI (95% CI, 0.903–0.965). The preoperative intraclass correlation coefficient values were 0.9995 (95% CI, 0.9993–0.9996) for the preoperative VFIS and 0.9989 (95% CI, 0.9985–0.9992) for the postoperative VFIS. Diffusion Tensor Imaging and Neurite Orientation Dispersion and Density Imaging Parameters and Visual Field Impairment Score in the Preoperative and Postoperative Stages The VFIS significantly correlated with the FA, MD, RD, Vic, and ODI values, adjusting for age, tumor size, and symptom duration. Specifically, the preoperative VFIS had significant FIG. 3. Flow chart of the study design. VFIS, visual field impairment score. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 e195 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 4. Scatterplots and partial correlation results between preoperative DTI and NODDI parameters and VFIS. The partial correlation coefficient (r) was adjusted for age, tumor size, and symptom duration. The units of MD, AD, and RD are square micrometer per second. AD, axial diffusivity; DTI, diffusion tensor imaging; FA, fractional anisotropy; MD, mean diffusivity; NODDI, neurite orientation dispersion and density imaging; ODI, orientation dispersion index; RD, radial diffusivity; VFIS, visual field impairment score; Vic, intracellular volume fraction; Viso, isotropic volume fraction. negative correlations with preoperative FA and Vic and significant positive correlations with MD, RD, and ODI (r = 20.423 and P , 0.001 for FA, r = 20.336 and P , 0.001 for Vic, r = e196 0.289 and P , 0.001 for MD, r = 0.437 and P , 0.001 for RD, and r = 0.201 and P = 0.011 for ODI; Fig. 4). Regarding postoperative changes, a significant negative correlation between Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 5. Scatterplots and partial correlation results between DDTI and DNODDI parameters and DVFIS. The partial correlation coefficient (r) was adjusted for age, tumor size, and symptom duration. The units for MD, AD, and RD are square micrometer per second. AD, axial diffusivity; DTI, diffusion tensor imaging; FA, fractional anisotropy; MD, mean diffusivity; NODDI, neurite orientation dispersion and density imaging; ODI, orientation dispersion index; RD, radial diffusivity; VFIS, visual field impairment score; Vic, intracellular volume fraction; Viso, isotropic volume fraction. DFA and DVFIS and a significant positive correlation between DRD and DVFIS were observed (r = 20.167 and P = 0.036 for DFA, r = 0.203 and P = 0.01 for DRD; Fig. 5). Additional analyses were performed for the optic radiations with preoperative visual field impairment (n = Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 117) to investigate the correlation between the preoperative DTI and NODDI parameters and the DVFIS. The DVFIS had significant positive correlations with the preoperative FA and Vic and significant negative correlations with MD and RD (r = 0.197 and P = 0.036 for FA, r = 0.253 and P = e197 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 6. Scatterplots and partial correlation results between preoperative DTI and NODDI parameters and DVFIS in the optic radiations with preoperative visual field impairment. The partial correlation coefficient (r) was adjusted for age, tumor size, and symptom duration. The units for MD, AD, and RD are square micrometer per second. AD, axial diffusivity; DTI, diffusion tensor imaging; FA, fractional anisotropy; MD, mean diffusivity; NODDI, neurite orientation dispersion and density imaging; ODI, orientation dispersion index; RD, radial diffusivity; VFIS, visual field impairment score; Vic, intracellular volume fraction; Viso, isotropic volume fraction. 0.007 for Vic, r = 20.233 and P = 0.013 for MD, and r = 20.264 and P = 0.005 for RD; Fig. 6). Table 2 shows the results of comparison between preoperative and postoperative data in optic radiations with e198 preoperative visual field impairment (n = 117) and optic radiations with normal visual fields (n = 45). In the optic radiations with visual field impairment, the FA and Vic significantly increased, whereas the MD and RD values Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 2. Comparison between preoperative and postoperative data Optic Radiations With Preoperative Visual Field Impairment (n = 117) Preoperative FA MD AD RD Vic Viso ODI 0.638 0.706 1.315 0.401 0.473 0.038 0.046 ± ± ± ± ± ± ± Postoperative 0.052 0.046 0.095 0.049 0.050 0.015 0.012 0.645 0.700 1.312 0.395 0.481 0.029 0.036 ± ± ± ± ± ± ± 0.051 0.042 0.091 0.046 0.049 0.013 0.022 Optic Radiations With Normal Visual Field (n = 45) Cohen’s d P Preoperative 0.136 0.136 0.032 0.126 0.162 0.641 0.564 0.023 0.024 0.573 0.011 0.003 ,0.001 ,0.001 0.703 0.672 1.339 0.339 0.529 0.035 0.040 ± ± ± ± ± ± ± 0.041 0.036 0.077 0.041 0.050 0.012 0.006 Postoperative 0.699 0.670 1.305 0.342 0.532 0.030 0.035 ± ± ± ± ± ± ± 0.052 0.035 0.121 0.044 0.044 0.013 0.022 Cohen’s d P 0.085 0.056 0.335 0.071 0.064 0.400 0.310 0.340 0.641 0.038 0.429 0.578 0.019 0.152 Data are the means ± SD. The mean values of MD, AD, and RD are given in units of square micrometer per sec. AD, axial diffusivity; FA, fractional anisotropy; MD, mean diffusivity; RD, radial diffusivity; ODI, orientation dispersion index; Vic, intracellular volume fraction; Viso, isotropic volume fraction; VFIS, visual field impairment score. significantly decreased after the surgery (P = 0.023 for FA, P = 0.003 for Vic, P = 0.024 for MD, P = 0.011 for RD, and P , 0.001 for ODI and Viso). However, in the optic radiations with normal visual fields, most of the DTI and NODDI parameters, except for AD and Viso, did not significantly change after the surgery (P . 0.05). Preoperative Visual Field Impairment and Postoperative Visual Field Improvement In 117 optic radiations with preoperative visual field impairment (VFIS = 1–8) and 45 optic radiations with normal visual fields (VFIS = 0), lower FA and Vic, as well as higher MD and RD, were significantly associated with the presence of visual field impairment in the preoperative stage (FA: odds radio [OR] = 0.750; 95% CI, 0.636–0.884; P = 0.001; Vic: OR = 0.827; 95% CI, 0.728–0.939; P = 0.003; MD: OR = 1.244; 95% CI, 1.060–1.459; P = 0.007; RD: OR = 1.361; 95% CI, 1.146–1.615; P , 0.001; Table 3). The degree of DVFIS for 117 optic radiations with preoperative visual field impairment was as follows: n = 17 with no improvement or worse after the surgery, n = 87 with mild improvement, and n = 13 with marked improvement. As a result, preoperative lower V ic and higher MD and RD were significantly associated with DVFIS (Vic: OR = 0.910; 95% CI, 0.834–0.993; P = 0.034; MD: OR = 1.103; 95% CI, 1.006–1.209; P = 0.037; RD: OR = 1.090; 95% CI, 1.001–1.188; P = 0.047; Table 4). DISCUSSION We investigated the correlations between the DTI and NODDI parameters in the optic radiation and VFIS using preoperative and postoperative data from patients with sellar and parasellar tumors. Several preoperative DTI and NODDI parameters in the optic radiations were significantly correlated with the preoperative VFIS and postoperative DVFIS. We identified that lower FA and Vic, as well as higher MD and RD, were significantly correlated with preoperative VFIS, even after adjusting for age, tumor size, and symptom TABLE 3. DTI and NODDI parameters and preoperative visual field impairment Optic Radiations With Preoperative Optic Radiations With Preoperative Normal Visual Field (n = 45) Visual Field Impairment (n = 117) OR (Per 0.01 Unit) 95% CI for OR FA MD AD RD Vic Viso ODI 0.703 0.672 1.339 0.339 0.529 0.035 0.040 ± ± ± ± ± ± ± 0.041 0.036 0.077 0.041 0.050 0.012 0.006 0.638 0.706 1.315 0.401 0.473 0.038 0.046 ± ± ± ± ± ± ± 0.052 0.046 0.095 0.049 0.050 0.015 0.012 0.75 1.24 0.98 1.36 0.83 1.42 1.55 0.64–0.88 1.06–1.46 0.93–1.05 1.15–1.62 0.73–0.94 0.93–2.16 0.78–3.08 P 0.001 0.007 0.6 ,0.001 0.003 0.10 0.21 Binary logistic regression classified 117 optic radiations with preoperative visual field impairment (VFIS = 1–8) and 45 optic radiations with normal visual field (VFIS = 0). As the values of the DTI and NODDI parameters were too small, OR and 95% CI were presented in 0.01 unit change, adjusted for age, tumor size and symptom duration. AD, axial diffusivity; AUC, area under the curve; CI, confidence intervals; FA, fractional anisotropy; MD, mean diffusivity; ODI, orientation dispersion index; OR, odds ratio; RD, radial diffusivity; VFIS, visual field impairment score; Vic, intracellular volume fraction; Viso, isotropic volume fraction. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 e199 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 4. DTI and NODDI parameters and postoperative visual field improvement No Improvement (n = 17) FA MD AD RD Vic Viso ODI 0.636 0.695 1.293 0.396 0.488 0.037 0.051 ± ± ± ± ± ± ± 0.065 0.050 0.117 0.057 0.054 0.017 0.021 Mild Improvement (n = 87) 0.643 0.704 1.317 0.397 0.474 0.038 0.044 ± ± ± ± ± ± ± 0.049 0.040 0.090 0.043 0.046 0.014 0.010 Marked Improvement (n = 13) 0.611 0.735 1.331 0.437 0.442 0.042 0.048 ± ± ± ± ± ± ± 0.044 0.070 0.100 0.066 0.060 0.016 0.011 OR (Per 0.01 Unit) 95% CI for OR P 0.95 1.10 1.02 1.09 0.91 1.14 0.87 0.88–1.03 1.01–1.21 0.98–1.07 1.00–1.19 0.83–0.99 0.86–1.51 0.62–1.22 0.24 0.04 0.36 0.047 0.03 0.38 0.43 Ordinal logistic regression was performed using 3 ordinal outcomes (no improvement or worse after the surgery, DVFIS = 0–4; mild improvement after the surgery, DVFIS = 24 w 21; and marked improvement after the surgery, DVFIS = 28 w 25). As the values of the DTI and NODDI parameters were too small, OR and 95% CI were presented in 0.01 unit change, adjusted for age, tumor size and symptom duration. AD, axial diffusivity; AUC, area under the curve; CI, confidence intervals; FA, fractional anisotropy; ODI, orientation dispersion index; OR, odds ratio; MD, mean diffusivity; RD, radial diffusivity; Vic, intracellular volume fraction; Viso, isotropic volume fraction; VFIS, visual field impairment score. duration. As a change in RD without a change in AD implies demyelination (13–16) and Vic indicates neurite density (5), we presumed that anterior visual pathway compression might cause decreased neurite density and some demyelination in the optic radiation. In addition, considering the reversibility of visual field defect after the surgery, edema or gliosis might be one of contributing factors. Our results were similar to a few previous studies that reported transsynaptic degeneration in sellar or parasellar lesions (1,17). Changes of DTI and NODDI parameters in the optic radiation may reflect transsynaptic degeneration, which has been reported as a mechanism of secondary pathology in the posterior visual pathways in conditions that primarily affect the anterior visual pathway, such as glaucoma (18,19) and optic neuritis (20,21). In regards to postoperative changes, preoperative MD, RD, and Vic showed significant correlations with DVFIS (r . 0.2). We noticed that the greater the changes in the preoperative DTI and NODDI parameters were, the more likely the postoperative improvement was. Specifically, lower Vic and higher MD and RD in the preoperative stage were significantly associated with visual field improvement after the surgery. In addition, similar to a previous study that investigated DTI parameters in optic tracts (16), we observed that increased RD with no change in AD in the preoperative stage was significantly associated with visual improvement after the surgery. These results support the notion that myelination and neurite density in the optic radiations are associated with visual field improvement after the surgery. Most DDTI and DNODDI parameters and preoperative FA and AD showed weak correlations with DVFIS (r , 0.2). We presumed that these results were caused by the time difference between anatomical decompression and neural restoration. Although functional improvement occurred postoperatively, neural restoration seemed to not be fully stabilized at the time of the follow-up examination in our study (79–122 days after the surgery). According to a e200 study using optical coherent tomography (22), patients with parachiasmal meningioma had visual improvement in 6 months and continued improving 1 year after the surgery. Therefore, further study with long-term follow-up examinations can accurately show the ability of DTI and NODDI parameters to predict surgical outcomes. In NODDI parameters, V ic indicating neurite density was significantly associated with both preoperative visual field impairment and postoperative visual field improvement. Regarding ODI, we could expect that it would not be significantly changed by anterior visual pathway compression because the optic radiation is a directional structure. Postoperative visual recovery has been known to be influenced by age, gender, and tumor size (10–12), and all of our results were robust after adjusting for these variables. Some limitations need to be mentioned. First, although our results demonstrated a uniform pattern regarding the NODDI parameters, fixed diffusion coefficients of 3-compartment NODDI model can be incorrect (23,24). Second, Long echo time (TE) (105 milliseconds) used in DTI and NODDI could limit our ability to measure myelin water. This might also explain weak correlations of DDTI and DNODDI parameters and preoperative FA and AD to change in VFIS because observation on the changes in the myelin associated with water is limited with DTI and NODDI data. Further research is warranted using a more dedicated method such as myelin water imaging (25). Third, we drew ROI in the undersampled optic radiations using 4-mm axial plane instead of coronal plane. As optic radiations are shown to be narrow in-plane and wide along inferior–superior view, the optic radiations could be undersampled in our data, which focused on high inplane resolution rather than thin slice thickness. It may explain weak correlations in our results. Also, we could not measure DTI and NODDI parameters in the optic chiasm, nerve, or tract. If we could measure the parameters in both anterior and posterior visual pathways, it would be valuable work to Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution compare each quantitative parameters in anterior and posterior visual pathways and to investigate their associations. However, as we mentioned in the introduction section, several previous studies have already reported relationship between pituitary adenomas and the consequent degree of visual impairment in the optic nerves and chiasm (26–29). In addition, measuring diffusion in optic radiations was easier and less likely to have trouble with postsurgical changes in the sella. Finally, there are no histological studies in animal models that have shown decreased neurite density in the optic radiation due to anterior visual pathway compression. Therefore, further NODDI/histological studies are needed to confirm our results. CONCLUSIONS In conclusion, the values of DTI and NODDI parameters in the optic radiations were significantly correlated with VFIS, and significantly associated with preoperative visual field impairment and postoperative visual field improvement in patients with sellar and parasellar tumors. These findings support the idea of anterograde trans-synaptic degeneration as a mechanism of secondary impairment in the posterior visual pathways by the anterior visual pathway compression. Clinically, preoperative DTI and NODDI parameters in the optic radiations might help evaluate preoperative visual field impairment in patients with poor communication and predict the postoperative visual outcome. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: S. H. Choi, J. Lee, and Y. H. Kim; b. Acquisition of data: Y. H. Kim; c. Analysis and interpretation of data: E-J. Choi, K. M. Kang, and W. Jung. Category 2: a. Drafting the manuscript: E-J. Choi, K. M. Kang, W. Jung, J. Lee, and S. H. Choi; b. Revising it for intellectual content: E-J. Choi, K. M. Kang, and Y. H. Kim. Category 3: a. Final approval of the completed manuscript: Y. H. Kim. REFERENCES 1. Phal PM, Steward C, Nichols AD, Kokkinos C, Desmond PM, Danesh-Meyer H, Sufaro YZ, Kaye AH, Moffat BA. Assessment of optic pathway structure and function in patients with compression of the optic chiasm: a correlation with optical coherence tomography. Invest Ophthalmol Vis Sci. 2016;57:3884–3890. 2. de Blank PMK, Berman JI, Liu GT, Roberts TPL, Fisher MJ. Fractional anisotropy of the optic radiations is associated with visual acuity loss in optic pathway gliomas of neurofibromatosis type 1. Neuro Oncol. 2013;15:1088–1095. 3. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophys J. 1994;66:259–267. 4. Gabilondo I, Martínez‐Lapiscina EH, Martínez‐Heras E, Fraga‐ Pumar E, Llufriu S, Ortiz S, Bullich S, Sepulveda M, Falcon C, Berenguer J. Trans‐synaptic axonal degeneration in the visual pathway in multiple sclerosis. Ann Neurol. 2014;75:98–107. 5. Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. Neuroimage. 2012;61:1000–1016. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 6. Smith SM. Fast robust automated brain extraction. Hum Brain Mapp. 2002;17:143–155. 7. Andersson JL, Sotiropoulos SN. An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage. 2016;125:1063–1078. 8. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE. Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage. 2004;23:S208–S19. 9. Daducci A, Canales-Rodriguez EJ, Zhang H, Dyrby TB, Alexander DC, Thiran JP. Accelerated microstructure imaging via Convex optimization (AMICO) from diffusion MRI data. Neuroimage. 2015;105:32–44. 10. Lee S, Kim SJ, Yu YS, Kim YH, Paek SH, Kim DG, Jung HW. Prognostic factors for visual recovery after transsphenoidal pituitary adenectomy. Br J Neurosurg. 2013;27:425–429. 11. Barzaghi LR, Medone M, Losa M, Bianchi S, Giovanelli M, Mortini P. Prognostic factors of visual field improvement after trans-sphenoidal approach for pituitary macroadenomas: review of the literature and analysis by quantitative method. Neurosurg Rev. 2012;35:369–379. 12. Müslüman AM, Cansever T, Yılmaz A, Kanat A, Oba E, Çavuşo glu H, Şirino glu D, Aydın Y. Surgical results of large and giant pituitary adenomas with special consideration of ophthalmologic outcomes. World Neurosurg. 2011;76:141–148. 13. Klawiter EC, Schmidt RE, Trinkaus K, Liang HF, Budde MD, Naismith RT, Song SK, Cross AH, Benzinger TL. Radial diffusivity predicts demyelination in ex vivo multiple sclerosis spinal cords. Neuroimage. 2011;55:1454–1460. 14. Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage. 2002;17:1429–1436. 15. Song SK, Yoshino J, Le TQ, Lin SJ, Sun SW, Cross AH, Armstrong RC. Demyelination increases radial diffusivity in corpus callosum of mouse brain. Neuroimage. 2005;26:132–140. 16. Paul DA, Gaffin-Cahn E, Hintz EB, Adeclat GJ, Zhu T, Williams ZR, Vates GE, Mahon BZ. White matter changes linked to visual recovery after nerve decompression. Sci Transl Med. 2014;6:266ra173. 17. Rutland JW, Padormo F, Yim CK, Yao A, Arrighi-Allisan A, Huang K-H, Lin H-M, Chelnis J, Delman BN, Shrivastava RK. Quantitative assessment of secondary white matter injury in the visual pathway by pituitary adenomas: a multimodal study at 7-Tesla MRI. J Neurosurg. 2019;1:1–10. 18. Bogorodzki P, Pia̧tkowska-Janko E, Szaflik J, Szaflik JP, Gacek M, Grieb P. Mapping cortical thickness of the patients with unilateral end-stage open angle glaucoma on planar cerebral cortex maps. PLoS One. 2014;9:e93682. 19. Gupta N, Ang LC, De Tilly LN, Bidaisee L, Yücel Y. Human glaucoma and neural degeneration in intracranial optic nerve, lateral geniculate nucleus, and visual cortex. Br J Ophthalmol. 2006;90:674–678. 20. Tur C, Goodkin O, Altmann DR, Jenkins TM, Miszkiel K, Mirigliani A, Fini C, Gandini Wheeler-Kingshott CA, Thompson AJ, Ciccarelli O. Longitudinal evidence for anterograde transsynaptic degeneration after optic neuritis. Brain. 2016;139:816–828. 21. Puthenparampil M, Federle L, Poggiali D, Miante S, Signori A, Pilotto E, Rinaldi F, Perini P, Sormani MP, Midena E. Transsynaptic degeneration in the optic pathway. A study in clinically isolated syndrome and early relapsing-remitting multiple sclerosis with or without optic neuritis. PLoS One. 2017;12:e0183957. 22. Park HH, Oh MC, Kim EH, Kim CY, Kim SH, Lee KS, Chang JH. Use of optical coherence tomography to predict visual outcome in parachiasmal meningioma. J Neurosurg. 2015;123:1489–1499. 23. Jelescu IO, Veraart J, Fieremans E, Novikov DS. Degeneracy in model parameter estimation for multi‐compartmental diffusion in neuronal tissue. NMR Biomed. 2016;29:33–47. 24. Kamiya K, Hori M, Aoki S. NODDI in clinical research. J Neurosci Methods. 2020;346:108908. e201 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 25. Lee J, Hyun JW, Lee J, Choi EJ, Shin HG, Min K, Nam Y, Kim HJ, Oh SH. So you want to image myelin using MRI: an overview and practical guide for myelin water imaging. J Magn Reson Imaging. 2020;53:360–373. 26. Yamada H, Yamamoto A, Okada T, Kanagaki M, Fushimi Y, Porter DA, Tanji M, Hojo M, Miyamoto S, Togashi K. Diffusion tensor imaging of the optic chiasm in patients with intra-or parasellar tumor using readout-segmented echo-planar. Magn Reson Imaging. 2016;34:654–661. 27. Lilja Y, Gustafsson O, Ljungberg M, Starck G, Lindblom B, Skoglund T, Bergquist H, Jakobsson KE, Nilsson D. Visual pathway impairment by pituitary adenomas: quantitative e202 diagnostics by diffusion tensor imaging. J Neurosurg. 2017;127:569–579. 28. Anik I, Anik Y, Cabuk B, Caklili M, Pirhan D, Ozturk O, Cirak M, Ceylan S. Visual outcome of an endoscopic endonasal transsphenoidal approach in pituitary macroadenomas: quantitative assessment with diffusion tensor imaging early and long-term results. World Neurosurg. 2018;112:e691–e701. 29. Metwali H, Giordano M, Kniese K, Fahlbusch R. Prognostic significance of intraoperative change in the fractional anisotropy and the volume of the optic chiasma during resection of suprasellar tumors. J Neurosurg. 2018;128:1479–1485. Kang et al: J Neuro-Ophthalmol 2022; 42: e192-e202 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.