Development of an Automated MRI-Based Diagnostic Protocol for Amyotrophic Lateral Sclerosis Using Disease-Specific Pathognomonic Features: A Quantitative Disease-State Classification Study

Biomarkers in Motor Neuron Disease: A State of the Art Review

Motor neuron disease can be viewed as an umbrella term describing a heterogeneous group of conditions, all of which are relentlessly progressive and ultimately fatal. The average life expectancy is 2 years, but with a broad range of months to decades. Biomarker research deepens disease understanding through exploration of pathophysiological mechanisms which, in turn, highlights targets for novel therapies. It also allows differentiation of the disease population into sub-groups, which serves two general purposes: (a) provides clinicians with information to better guide their patients in terms of disease progression, and (b) guides clinical trial design so that an intervention may be shown to be effective if population variation is controlled for. Biomarkers also have the potential to provide monitoring during clinical trials to ensure target engagement. This review highlights biomarkers that have emerged from the fields of systemic measurements including biochemistry (blood, cerebrospinal fluid, and urine analysis); imaging and electrophysiology, and gives examples of how a combinatorial approach may yield the best results. We emphasize the importance of systematic sample collection and analysis, and the need to correlate biomarker findings with detailed phenotype and genotype data.

Machine Learning in Amyotrophic Lateral Sclerosis: Achievements, Pitfalls, and Future Directions

Background: Amyotrophic Lateral Sclerosis (ALS) is a relentlessly progressive neurodegenerative condition with limited therapeutic options at present. Survival from symptom onset ranges from 3 to 5 years depending on genetic, demographic, and phenotypic factors. Despite tireless research efforts, the core etiology of the disease remains elusive and drug development efforts are confounded by the lack of accurate monitoring markers. Disease heterogeneity, late-stage recruitment into pharmaceutical trials, and inclusion of phenotypically admixed patient cohorts are some of the key barriers to successful clinical trials. Machine Learning (ML) models and large international data sets offer unprecedented opportunities to appraise candidate diagnostic, monitoring, and prognostic markers. Accurate patient stratification into well-defined prognostic categories is another aspiration of emerging classification and staging systems. Methods: The objective of this paper is the comprehensive, systematic, and critical review of ML initiatives in ALS to date and their potential in research, clinical, and pharmacological applications. The focus of this review is to provide a dual, clinical-mathematical perspective on recent advances and future directions of the field. Another objective of the paper is the frank discussion of the pitfalls and drawbacks of specific models, highlighting the shortcomings of existing studies and to provide methodological recommendations for future study designs. Results: Despite considerable sample size limitations, ML techniques have already been successfully applied to ALS data sets and a number of promising diagnosis models have been proposed. Prognostic models have been tested using core clinical variables, biological, and neuroimaging data. These models also offer patient stratification opportunities for future clinical trials. Despite the enormous potential of ML in ALS research, statistical assumptions are often violated, the choice of specific statistical models is seldom justified, and the constraints of ML models are rarely enunciated. Conclusions: From a mathematical perspective, the main barrier to the development of validated diagnostic, prognostic, and monitoring indicators stem from limited sample sizes. The combination of multiple clinical, biofluid, and imaging biomarkers is likely to increase the accuracy of mathematical modeling and contribute to optimized clinical trial designs.

The changing landscape of motor neuron disease imaging: the transition from descriptive studies to precision clinical tools

PURPOSE OF REVIEW

Neuroimaging in motor neuron disease (MND) has traditionally been seen as an academic tool with limited direct relevance to individualized patient care. This has changed radically in recent years as computational imaging has emerged as a viable clinical tool with true biomarker potential. This transition is not only fuelled by technological advances but also by important conceptual developments.

RECENT FINDINGS

The natural history of MND is now evaluated by presymptomatic, postmortem and multi-timepoint longitudinal imaging studies. The anatomical spectrum of MND imaging has also been expanded from an overwhelmingly cerebral focus to innovative spinal and muscle applications. In contrast to the group-comparisons of previous studies, machine-learning and deep-learning approaches are increasingly utilized to model real-life diagnostic dilemmas and aid prognostic classification. The focus from evaluating focal structural changes has shifted to the appraisal of network integrity by connectivity-based approaches. The armamentarium of MND imaging has also been complemented by novel PET-ligands, spinal toolboxes and the availability of magnetoencephalography and high-field magnetic resonance (MR) imaging platforms.

SUMMARY

In addition to the technological and conceptual advances, collaborative multicentre research efforts have also gained considerable momentum. This opinion-piece reviews emerging trends in MND imaging and their implications to clinical care and drug development.

Dysfunction of attention switching networks in amyotrophic lateral sclerosis

OBJECTIVE

To localise and characterise changes in cognitive networks in Amyotrophic Lateral Sclerosis (ALS) using source analysis of mismatch negativity (MMN) waveforms.

RATIONALE

The MMN waveform has an increased average delay in ALS. MMN has been attributed to change detection and involuntary attention switching. This therefore indicates pathological impairment of the neural network components which generate these functions. Source localisation can mitigate the poor spatial resolution of sensor-level EEG analysis by associating the sensor-level signals to the contributing brain sources. The functional activity in each generating source can therefore be individually measured and investigated as a quantitative biomarker of impairment in ALS or its sub-phenotypes.

METHODS

MMN responses from 128-channel electroencephalography (EEG) recordings in 58 ALS patients and 39 healthy controls were localised to source by three separate localisation methods, including beamforming, dipole fitting and exact low resolution brain electromagnetic tomography.

RESULTS

Compared with controls, ALS patients showed significant increase in power of the left posterior parietal, central and dorsolateral prefrontal cortices (false discovery rate = 0.1). This change correlated with impaired cognitive flexibility (rho = 0.45, 0.45, 0.47, p = .042, .055, .031 respectively). ALS patients also exhibited a decrease in the power of dipoles representing activity in the inferior frontal (left: p = 5.16 × 10-6, right: p = 1.07 × 10-5) and left superior temporal gyri (p = 9.30 × 10-6). These patterns were detected across three source localisation methods. Decrease in right inferior frontal gyrus activity was a good discriminator of ALS patients from controls (AUROC = 0.77) and an excellent discriminator of C9ORF72 expansion-positive patients from controls (AUROC = 0.95).

INTERPRETATION

Source localization of evoked potentials can reliably discriminate patterns of functional network impairment in ALS and ALS subgroups during involuntary attention switching. The discriminative ability of the detected cognitive changes in specific brain regions are comparable to those of functional magnetic resonance imaging (fMRI). Source analysis of high-density EEG patterns has excellent potential to provide non-invasive, data-driven quantitative biomarkers of network disruption that could be harnessed as novel neurophysiology-based outcome measures in clinical trials.

Corticospinal tract degeneration in ALS unmasked in T1-weighted images using texture analysis

The purpose of this study was to investigate whether textures computed from T1-weighted (T1W) images of the corticospinal tract (CST) in amyotrophic lateral sclerosis (ALS) are associated with degenerative changes evaluated by diffusion tensor imaging (DTI). Nineteen patients with ALS and 14 controls were prospectively recruited and underwent T1W and diffusion-weighted magnetic resonance imaging. Three-dimensional texture maps were computed from T1W images and correlated with the DTI metrics within the CST. Significantly correlated textures were selected and compared within the CST for group differences between patients and controls using voxel-wise analysis. Textures were correlated with the patients' clinical upper motor neuron (UMN) signs and their diagnostic accuracy was evaluated. Voxel-wise analysis of textures and their diagnostic performance were then assessed in an independent cohort with 26 patients and 13 controls. Results showed that textures autocorrelation, energy, and inverse difference normalized significantly correlated with DTI metrics (p < .05) and these textures were selected for further analyses. The textures demonstrated significant voxel-wise differences between patients and controls in the centrum semiovale and the posterior limb of the internal capsule bilaterally (p < .05). Autocorrelation and energy significantly correlated with UMN burden in patients (p < .05) and classified patients and controls with 97% accuracy (100% sensitivity, 92.9% specificity). In the independent cohort, the selected textures demonstrated similar regional differences between patients and controls and classified participants with 94.9% accuracy. These results provide evidence that T1-based textures are associated with degenerative changes in the CST.

Neuroimaging in amyotrophic lateral sclerosis: current and emerging uses

INTRODUCTION

Several neuroimaging techniques have been used to define in vivo markers of pathological alterations underlying amyotrophic lateral sclerosis (ALS). Growing evidence supports the use of magnetic resonance imaging (MRI) and positron emission tomography (PET) for the non-invasive detection of central nervous system involvement in patients with ALS. Areas covered: A comprehensive overview of structural and functional neuroimaging applications in ALS is provided, focusing on motor and extra-motor involvement in the brain and the spinal cord. Implications for pathogenetic models, patient diagnosis, prognosis, monitoring, and the design of clinical trials are discussed. Expert commentary: State-of-the-art neuroimaging techniques provide fundamental instruments for the detection and quantification of upper motor neuron and extra-motor brain involvement in ALS, with relevance for both pathophysiologic investigation and clinical practice. Network-based analysis of structural and functional connectivity alterations and multimodal approaches combining several neuroimaging measures are promising tools for the development of novel diagnostic and prognostic markers to be used at the individual patient level.

Multimodal structural MRI in the diagnosis of motor neuron diseases

This prospective study developed an MRI-based method for identification of individual motor neuron disease (MND) patients and test its accuracy at the individual patient level in an independent sample compared with mimic disorders. 123 patients with amyotrophic lateral sclerosis (ALS), 44 patients with predominantly upper motor neuron disease (PUMN), 20 patients with ALS-mimic disorders, and 78 healthy controls were studied. The diagnostic accuracy of precentral cortical thickness and diffusion tensor (DT) MRI metrics of corticospinal and motor callosal tracts were assessed in a training cohort and externally proved in a validation cohort using a random forest analysis. In the training set, precentral cortical thickness showed 0.86 and 0.89 accuracy in differentiating ALS and PUMN patients from controls, while DT MRI distinguished the two groups from controls with 0.78 and 0.92 accuracy. In ALS vs controls, the combination of cortical thickness and DT MRI metrics (combined model) improved the classification pattern (0.91 accuracy). In the validation cohort, the best accuracy was reached by DT MRI (0.87 and 0.95 accuracy in ALS and PUMN vs mimic disorders). The combined model distinguished ALS and PUMN patients from mimic syndromes with 0.87 and 0.94 accuracy. A multimodal MRI approach that incorporates motor cortical and white matter alterations yields statistically significant improvement in accuracy over using each modality separately in the individual MND patient classification. DT MRI represents the most powerful tool to distinguish MND from mimic disorders.

•

Motor cortical and white matter alterations yield high accuracy in the individual MND patient classification.

•

DT MRI represents the most powerful tool to distinguish MND from mimic disorders.

•

The most pronounced damage in MND patients relative to mimic subjects was found in the motor callosal fibers.

Virtual brain biopsies in amyotrophic lateral sclerosis: Diagnostic classification based on in vivo pathological patterns

Background

Diagnostic uncertainty in ALS has serious management implications and delays recruitment into clinical trials. Emerging evidence of presymptomatic disease-burden provides the rationale to develop diagnostic applications based on the evaluation of in-vivo pathological patterns early in the disease.

Objectives

To outline and test a diagnostic classification approach based on an array of complementary imaging metrics in key disease-associated anatomical structures.

Methods

Data from 75 ALS patients and 75 healthy controls were randomly allocated in a ‘training’ and ‘validation’ cohort. Spatial masks were created for anatomical foci which best discriminate patients from controls in the ‘training sample’. In a virtual ‘brain biopsy’, data was then retrieved from these key disease-associated brain regions. White matter diffusivity indices, grey matter T1-signal intensity values and basal ganglia volumes were evaluated as predictor variables in a canonical discriminant function.

Results

Following predictor variable selection, a classification specificity of 85.5% and sensitivity of 89.1% was achieved in the training sample and 90% specificity and 90% sensitivity in the validation sample.

Discussion

This study evaluates disease-associated imaging measures in a dummy diagnostic application. Although larger samples will be required for robust validation, the study confirms the potential of multimodal quantitative imaging in future clinical applications.

•

Reliable diagnostic, monitoring and prognostic biomarkers are urgently in ALS.

•

Accurate diagnostic classification may be achieved based on MRI metrics.

•

Basal ganglia, grey and white matter indices were integrated in a diagnostic model.

•

85.5% specificity and 89.1% sensitivity were achieved in the training sample.

•

90% specificity and 90% sensitivity were achieved in the validation sample.