Dataset Information

Deep learning-based relapse prediction of neuromyelitis optica spectrum disorder with anti-aquaporin-4 antibody.

ABSTRACT:

Objective

We previously identified the independent predictors of recurrent relapse in neuromyelitis optica spectrum disorder (NMOSD) with anti-aquaporin-4 antibody (AQP4-ab) and designed a nomogram to estimate the 1- and 2-year relapse-free probability, using the Cox proportional hazard (Cox-PH) model, assuming that the risk of relapse had a linear correlation with clinical variables. However, whether the linear assumption fits real disease tragedy is unknown. We aimed to employ deep learning and machine learning to develop a novel prediction model of relapse in patients with NMOSD and compare the performance with the conventional Cox-PH model.

Methods

This retrospective cohort study included patients with NMOSD with AQP4-ab in 10 study centers. In this study, 1,135 treatment episodes from 358 patients in Huashan Hospital were employed as the training set while 213 treatment episodes from 92 patients in nine other research centers as the validation set. We compared five models with added variables of gender, AQP4-ab titer, previous attack under the same therapy, EDSS score at treatment initiation, maintenance therapy, age at treatment initiation, disease duration, the phenotype of the most recent attack, and annualized relapse rate (ARR) of the most recent year by concordance index (C-index): conventional Cox-PH, random survival forest (RSF), LogisticHazard, DeepHit, and DeepSurv.

Results

When including all variables, RSF outperformed the C-index in the training set (0.739), followed by DeepHit (0.737), LogisticHazard (0.722), DeepSurv (0.698), and Cox-PH (0.679) models. As for the validation set, the C-index of LogisticHazard outperformed the other models (0.718), followed by DeepHit (0.704), DeepSurv (0.698), RSF (0.685), and Cox-PH (0.651) models. Maintenance therapy was calculated to be the most important variable for relapse prediction.

Conclusion

This study confirmed the superiority of deep learning to design a prediction model of relapse in patients with AQP4-ab-positive NMOSD, with the LogisticHazard model showing the best predictive power in validation.

SUBMITTER: Wang L

PROVIDER: S-EPMC9389264 | biostudies-literature | 2022

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep learning-based relapse prediction of neuromyelitis optica spectrum disorder with anti-aquaporin-4 antibody.

Wang Liang L Du Lei L Li Qinying Q Li Fang F Wang Bei B Zhao Yuanqi Y Meng Qiang Q Li Wenyu W Pan Juyuan J Xia Junhui J Wu Shitao S Yang Jie J Li Heng H Ma Jianhua J ZhangBao Jingzi J Huang Wenjuan W Chang Xuechun X Tan Hongmei H Yu Jian J Zhou Lei L Lu Chuanzhen C Wang Min M Dong Qiang Q Lu Jiahong J Zhao Chongbo C Quan Chao C

Frontiers in neurology 20220805

<h4>Objective</h4>We previously identified the independent predictors of recurrent relapse in neuromyelitis optica spectrum disorder (NMOSD) with anti-aquaporin-4 antibody (AQP4-ab) and designed a nomogram to estimate the 1- and 2-year relapse-free probability, using the Cox proportional hazard (Cox-PH) model, assuming that the risk of relapse had a linear correlation with clinical variables. However, whether the linear assumption fits real disease tragedy is unknown. We aimed to employ deep lea ...[more]

PMID: 35989911

Similar Datasets

Project description:BackgroundRecognizing the predictors of disease relapses in patients with anti-aquaporin-4 antibody (AQP4-ab)-positive neuromyelitis optica spectrum disorder (NMOSD) is essential for individualized treatment strategy. We aimed to identify the factors that predicted relapses among patients with AQP4-ab-positive NMOSD, develop outcome prediction models, and validate them in a multicenter validation cohort.MethodsBetween January 2015 and December 2020, 820 patients with NMOSD were registered at Huashan Hospital. We retrospectively reviewed their medical records, and included 358 AQP4-ab-positive patients with 1135 treatment episodes. Univariate and multivariate analyses were used to explore the predictors of relapse, severe visual or motor disability during follow-up. A model predicting the 1- and 2-year relapse-free probability was developed and validated in an external validation cohort of 92 patients with 213 treatment episodes.ResultsLower serum AQP4-ab titer (< 1:100), higher Expanded Disability Status Scale (EDSS) score at onset (≥ 2.5), and use of intravenous methylprednisolone (IVMP) at the first attack predicted an overall lower annualized relapse rate. Older age (> 48 years), optic neuritis at onset, and higher onset EDSS score (≥ 2.5) significantly increased the risk for blindness, while IVMP at the first attack and maintenance therapy reduced the risk for blindness. Myelitis at onset increased the possibility of motor disability (EDSS ≥ 6.0), severe motor disability or death (EDSS ≥ 8.0), while maintenance therapy reduced these possibilities. Anderson and Gill model identified that the risk factors predicting recurrent relapses under certain treatment status were female gender, high AQP4-ab titer (≥ 1:100), previous attack under same therapy, lower EDSS score at treatment initiation (< 2.5), and no maintenance therapy or oral prednisone lasting less than 6 months. A nomogram using the above factors showed good discrimination and calibration abilities. The concordance indexes in the primary and validation cohort were 0.66 and 0.65, respectively.ConclusionThis study reports the demographic, clinical and therapeutic predictors of relapse, and severe visual or motor disability in NMOSD. Early identification of patients at risk of unfavorable outcomes is of paramount importance to inform treatment decisions.

Project description:Objective Neuromyelitis optica and its spectrum disorders (NMOSD) are inflammatory demyelinating diseases (IDD) with a specific biomarker, aquaporin-4-IgG. Prior NMO/NMOSD epidemiological studies are limited by lack of aquaporin-4-IgG seroprevalence assessment, absence of population-based USA studies and under-representation of blacks. To overcome these limitations, we sought to compare NMO/NMOSD seroepidemiology across two ethnically divergent populations. Methods We performed a population-based comparative study of the incidence (2003-2011) and prevalence (on December 31, 2011) of NMO/NMOSD and aquaporin-4-IgG seroincidence and seroprevalence (sera collected in 80-84% of IDD) among patients with IDD diagnosis in Olmsted County, USA (82% white [Caucasian]) and Martinique (90% black [Afro-Caribbean]). Aquaporin-4-IgG was measured by M1-isoform-fluorescent-activated-cell-sorting assays. Results The age and sex adjusted incidence (7.3 vs 0.7/1,000,000 person-years [p<0.01]) and prevalence (10 vs 3.9/100,000[p=0.01]) in Martinique exceeded that in Olmsted County. The AQP4-IgG age and sex-adjusted seroincidence (6.5 vs 0.7/1,000,000 person-years [p<0.01]) and seroprevalence (7.9 vs 3.3/100,000[p=0.04]) were also higher in Martinique than Olmsted County. The ethnicity-specific prevalence was similar in Martinique and Olmsted County: 11.5 and 13/100,000 in blacks, and 6.1 and 4.0/100,000 in whites, respectively. NMO/NMOSD represented a higher proportion of IDD in Martinique than Olmsted County (16% vs 1.4%; p<0.01). The onset age (median, 35-37 years) and female:male distribution (8-9:1) were similar across both populations; 60% of prevalent cases were either blind in one eye, dependent on a gait aid or both. Interpretation This study reports the highest prevalence of NMO/NMOSD in any population (10/100,000 in Martinique), estimates it affects 16,000-17,000 in the USA (higher than previous predictions) and demonstrates it disproportionately affects blacks. This article is protected by copyright. All rights reserved.

Project description:ObjectiveTo assess the pharmacokinetics and pharmacodynamics of the long-acting terminal complement 5 (C5) inhibitor ravulizumab in adults with anti-aquaporin-4 antibody-positive (AQP4+) neuromyelitis optica spectrum disorder (NMOSD) in the phase 3, open-label CHAMPION-NMOSD trial (NCT04201262).MethodsPatients aged 18 years or older received a weight-based intravenous loading dose of ravulizumab (2,400-3,000 mg) on day 1, followed by weight-based maintenance doses (3,000-3,600 mg) on day 15 and once every 8 weeks thereafter. Pharmacokinetic assessments were maximum observed concentration (Cmax, assessed at the end of the infusion) and concentration at the end of the dosing interval (Ctrough, assessed before dosing) for ravulizumab. Pharmacodynamic assessment was time-matched observed free C5 concentration in serum up to 50 weeks.ResultsThe pharmacokinetic/pharmacodynamic analysis included 58 patients treated with ravulizumab. Serum ravulizumab concentrations at or above the therapeutic threshold (175 μg/mL) were achieved in all patients after administration of the first dose and maintained for 50 weeks. At week 50, the mean (standard deviation) Cmax (n = 51) and Ctrough (n = 52) were 1,887.6 (411.38) and 764.4 (217.68) μg/mL, respectively. Immediate and complete terminal complement inhibition (free C5 serum concentrations < 0.5 μg/mL) was achieved by the end of the first ravulizumab infusion and sustained throughout the treatment period. No treatment-emergent antibodies to ravulizumab were observed. No impact on ravulizumab pharmacokinetics was seen for age, sex, race, hematocrit, hemoglobin, markers of renal and liver impairment, or medications commonly used by patients with NMOSD. Body weight and BMI were significant covariates of ravulizumab pharmacokinetics.ConclusionsSerum ravulizumab concentrations were maintained above the therapeutic threshold in all patients through 50 weeks of treatment. Ravulizumab achieved immediate and complete terminal complement inhibition that was sustained throughout the treatment period in adults with AQP4+ NMOSD.

Project description:Fatigue is frequently reported by patients with multiple sclerosis, aquaporin-4-antibody neuromyelitis optica spectrum disorder and myelin-oligodendrocyte-glycoprotein antibody disease; thus they could share a similar pathophysiological mechanism. In this cross-sectional cohort study, we assessed the association of fatigue with resting-state functional MRI, diffusion and structural imaging measures across these three disorders. Sixteen patients with multiple sclerosis, 17 with aquaporin-4-antibody neuromyelitis optica spectrum disorder and 17 with myelin-oligodendrocyte-glycoprotein antibody disease assessed, outside of relapses, at the Oxford Neuromyelitis Optica Service underwent Modified Fatigue Impact Scale, Hospital Anxiety and Depression Scale and Expanded Disability Status Scale scoring. A 3T brain and spinal cord MRI was used to derive cortical, deep grey and white matter volumetrics, lesions volume, fractional anisotropy, brain functional connectivity metrics, cervical spinal cord cross-sectional area, spinal cord magnetic transfer ratio and average functional connectivity between the ventral and the dorsal horns of the cervical cord. Linear relationships between MRI measures and total-, cognitive- and physical-fatigue scores were assessed. All analyses were adjusted for correlated clinical regressors. No significant differences in baseline clinical characteristics, fatigue, depression and anxiety questionnaires and disability measures were seen across the three diseases, except for older age in patients with aquaporin-4-antibody neuromyelitis optica spectrum disorder (P = 0.0005). In the total cohort, median total-fatigue score was 35.5 (range 3-72), and 42% of patients were clinically fatigued. A positive correlation existed between the total-fatigue score and functional connectivity of the executive/fronto-temporal network in the in left middle temporal gyrus (P = 0.033) and between the physical-fatigue score and functional connectivity of the sensory-motor network (P = 0.032) in both pre- and post-central gyri. A negative relationship was found between the total-fatigue score and functional connectivity of the salience network (P = 0.023) and of the left fronto-parietal network (P = 0.026) in the right supramarginal gyrus and left superior parietal lobe. No clear relationship between fatigue subscores and the average functional connectivity of the spinal cord was found. Cognitive-fatigue scores were positively associated with white matter lesion volume (P = 0.018) and negatively associated with white matter fractional anisotropy (P = 0.032). Structural, diffusion and functional connectivity alterations were not influenced by the disease group. Functional and structural imaging metrics associated with fatigue relate to brain rather than spinal cord abnormalities. Salience and sensory-motor networks alterations in relation to fatigue might indicate a disconnection between the perception of the interior body state and activity and the actual behavioural responses and performances (reversible or irreversible). Future research should focus on functional rehabilitative strategies.

Dataset Information

Deep learning-based relapse prediction of neuromyelitis optica spectrum disorder with anti-aquaporin-4 antibody.

Objective

Methods

Results

Conclusion

Publications

Deep learning-based relapse prediction of neuromyelitis optica spectrum disorder with anti-aquaporin-4 antibody.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets