Botha, Hugo

Abstract Nerve conduction F-wave studies contain crucial information about subclinical motor dysfunction that can be used to diagnose patients with amyotrophic lateral sclerosis (ALS). However, F-wave responses are highly variable in morphology, making waveform interpretation challenging. Artificial intelligence techniques can extract time–frequency features to provide new insights into ALS diagnosis and prognosis. A retrospective analysis was performed on F-wave responses from 46 802 patients. Discrete wavelet transforms were applied to time-series waveform responses after stimulating ulnar, median, fibular and tibial nerves. Wavelet coefficient statistics, onset age, sex and body mass index were features for training a Gradient Boosting Machine model on 40 095 patients (5329 diagnosed with motor neuron disease). Model performance was tested on responses from 689 ALS patients meeting Gold Coast criteria and 689 age- and sex-matched controls. An exploratory analysis examined model performance on cohorts of patients with inclusion body myositis, cervical radiculopathy, lumbar radiculopathy or peripheral neuropathy, which can mimic ALS symptoms. Factors affecting survival were estimated through Cox proportional hazards regression. The model trained using wavelet features on the full waveform had 90% recall, 87% precision and 88% accuracy. Similar model performance was measured using features from only the M-wave or F-wave. Classification probabilities for ALS patients were statistically different from the diagnoses mimicking ALS symptoms (P < 0.001, ANOVA, Tukey’s post hoc test). Higher model classification probabilities of ALS, older age at onset, and family history of ALS alone or with frontotemporal dementia were factors decreasing survival. Longer diagnostic delay and upper limb onset site were factors increasing survival. Model scores two standard deviations below the mean had 4 months increased survival (two standard deviations below had 3 months decreased survival). Artificial intelligence techniques extracted important information from F-wave responses to estimate a patient’s likelihood of ALS and their survival risks. Although the model can make predictions at a specific decision threshold as presented here, the true strength of such a model lies in its ability to provide probabilities about whether a patient is likely to have ALS in comparison to other mimicking diagnoses, such as inclusion body myositis, cervical or lumbar radiculopathy or peripheral neuropathy. These probabilities provide clinicians with additional information that they can use to make the final diagnosis with greater confidence and precision. Integrating such a model into the clinical workflow could help clinicians to diagnose ALS sooner and manage treatment based on estimated survival, which might improve outcomes and the quality of life of patients.

Importance Although 18 F-fluorodeoxyglucose positron emission tomography (FDG-PET) is a well-established cross-sectional biomarker of brain metabolism in dementia with Lewy bodies (DLB), the longitudinal change in FDG-PET has not been characterized. Objective To investigate longitudinal FDG-PET in prodromal DLB and DLB, including a subsample with autopsy data, and report estimated sample sizes for a hypothetical clinical trial in DLB. Design, Setting, and Participants Longitudinal case-control study with mean (SD) follow-up of 3.8 (2.3) years. Cases were recruited consecutively between 2007 and 2022 at a referral center and among the population. Patients with probable DLB or mild cognitive impairment with Lewy bodies (MCI-LB) were included. Individuals without cognitive impairment were included from a population-based cohort balanced on age and sex for comparison. All participants completed at least 1 follow-up assessment by design. Exposure Patients with MCI-LB and DLB. Main Outcomes and Measures Rate of change in FDG-PET was assessed as standardized uptake value ratios (SUVr). Clinical progression was assessed with the Clinical Dementia Rating Sum of Boxes (CDR-SB) score. Results Thirty-five patients with probable DLB, 37 patients with MCI-LB, and 100 individuals without cognitive impairment were included. The mean (SD) age of the DLB and MCI-LB groups combined (n = 72) was 69.6 (8.2) years; 66 patients (92%) were men and 6 (8%) were women. At follow-up, 18 participants (49%) with MCI-LB had progressed to probable DLB. Patients with MCI-LB had a faster decline in FDG-SUVr, compared with that of participants without cognitive impairment, in the posterior cingulate, occipital, parietal, temporal, and lateral frontal cortices. The same regions showed greater metabolic decline in patients with DLB than in participants without cognitive impairment, with the addition of anterior-middle cingulate, insula, and medial frontal orbital cortices. Rates of change in FDG-PET in these brain regions were combined into a region of interest (ROI) labeled longitudinal FDG-PET LB meta-ROI. The rate of change in FDG-SUVr in the meta-ROI correlated with the rate of change in CDR-SB, and sample size estimates were reported for potential clinical trials in DLB. Findings were confirmed in the subsample with neuropathologic confirmation (n = 20). Conclusions and Relevance This study found that brain hypometabolism begins to evolve during the prodromal stages of DLB with changes paralleling symptomatic progression. These data may inform clinical practice and trials planning to use FDG-PET for biologic staging, monitoring disease progression, and potentially assessing treatment response.

AbstractBackgroundEvidence for abnormal amyloid‐β (Aβ) plaque accumulation is necessary prior to initiating anti‐amyloid therapy in early symptomatic Alzheimer’s disease (AD). While the clinical trials for lecanemab and related drugs utilized positron emission tomography (PET) to demonstrate brain amyloidosis, current appropriate use recommendations for clinical practice consider PET or cerebrospinal fluid (CSF) biomarkers as satisfactory for this purpose. Here, we present four clinical cases where CSF biomarker results were discordant from amyloid PET, with the potential to result in erroneous treatment targeting.MethodPatients were seen at a tertiary care subspecialty clinic in consideration for lecanemab therapy. All patients had office examination, neuropsychological assessment, structural magnetic resonance imaging (MRI), amyloid PET (with either 11C‐Pittsburgh compound B or 18F‐florbetapir), CSF AD biomarkers (Roche Elecsys p‐tau181/Aβ42 and Fujirebio Lumipulse Aβ42/Aβ40 ratios), and apolipoprotein E (APOE) allele testing. Amyloid PET interpretations were made by consensus visual review from experienced readers.ResultAll four patients had clinical diagnoses of mild cognitive impairment or mild dementia (Table 1). Two patients had positive CSF p‐tau181/Aβ42 and Aβ42/40 ratios with negative amyloid PET (indicating sparse to no neuritic plaques). Case 1 displayed aphasia (characterized by anomia and deep dysgraphia) and imaging findings possibly compatible with semantic dementia. Case 2 displayed a gait abnormality and imaging signs of abnormal CSF dynamics. Two other patients had negative CSF p‐tau181/Aβ42 measures with positive amyloid PET (indicating moderate to frequent neuritic plaques). In those instances, the clinical syndrome and imaging studies were strongly supportive of AD and the CSF Aβ42/40 ratio was “likely positive” (Table 1). For one patient (Case 4), tau PET and Lumipulse plasma p‐tau217 concentration were both positive.ConclusionClinicians prescribing anti‐amyloid therapies should be aware of the possibility for CSF biomarkers to be discordant from amyloid PET status in select scenarios. These cases highlight the value of considering PET imaging in the presence of clinically atypical features, concern for mixed etiology, or strong clinical suspicion for AD despite negative CSF biomarkers. Larger series are needed to identify broader patterns contributing to biomarker discordance by assay, modality (PET versus CSF/plasma measures), and medical/neurologic comorbidities.

AbstractBackgroundAs new anti‐amyloid immunotherapies emerge for Alzheimer’s disease (AD), it is clear that early diagnosis of AD pathology is crucial for treatment success. This can be challenging in atypical presentations of AD and, together with our reliance on CSF or PET scans, can, at times, lead to delayed diagnosis. Here, we further explore the possible role of plasma tau phosphorylated at threonine 217 (P‐tau217) for the detection of primary AD or AD co‐pathology when frontotemporal dementia spectrum disorders are the main clinical presentation.MethodParticipants were recruited through ALLFTD, a North American multisite research consortium for the longitudinal assessment of frontotemporal lobar degeneration. After excluding cases with known FTD‐causing mutations, 573 participants with sporadic disease, mild cognitive/behavioral impairment, or healthy familial controls evaluated between 2014 and 2023 were included in this analysis (45.8% female, median age 66 years). 39 of those cases underwent neuropathological evaluation. Plasma P‐tau217 was measured with the Eli Lilly electrochemiluminescence‐based assay. P‐tau217 concentrations were compared by phenotype, disease severity, genotype, and, when available, neuropathological diagnosis with non‐parametric tests. Its diagnostic performance was tested with ROC curves. Associations between plasma P‐tau217 concentrations and clinical scales of global cognition, memory, executive function, motor function, and social cognition with linear regressions and non‐parametric tests.ResultPlasma P‐tau217 concentrations were higher in logopenic variant primary progressive aphasia (lvPPA) and amnestic dementia (AmDem), compared to participants with normal cognition or other FTD phenotypes. P‐tau217 concentrations were higher in APOE e4 carriers, compared to non‐carriers, regardless of phenotype, but were not affected by disease severity. P‐tau217 concentrations showed associations with clinical measures of global cognition, memory, and executive function, but had no associations with measures of motor function or social cognition. Plasma P‐tau217 concentrations were elevated in high and intermediate ADNC scoring, or more advanced Braak stages, even when AD was not the primary pathology.ConclusionEven when FTD is suspected, high plasma p‐tau 217 concentrations are strongly associated with underlying AD as primary pathology or contributing co‐pathology. Plasma P‐tau217 could be a tool to support the development of disease‐modifying therapies for atypical Alzheimer’s disease presentations or for FTD cases with Alzheimer’s disease co‐pathology.

AbstractBackgroundMany proposed clinical decision support systems (CDSS) require multiple disparate data elements as input, which makes implementation difficult, and furthermore have a black‐box nature leading to low interpretability. Fluorodeoxyglucose Positron Emission Tomography (FDG‐PET) is an established modality for the diagnosis of dementia, and a CDSS that uses only an FDG‐PET image to produce a reliable and understandable result would ease both of these challenges to clinical application.MethodA deep variational autoencoder (VAE) was used to extract a latent representation of each image through prior training from FDG‐PET brain images (n=2000). This unsupervised VAE has a novel graph convolutional architecture that makes it applicable to masked template space images. A logistic regression model was used to classify each image. A parametric study of the latent space revealed the imaging features that were used in the logistic regression model to differentiate each class. A separate dataset of participants labeled with their clinical diagnosis (n=1239) was used to assess the logistic regression model’s diagnostic accuracy of cognitively unimpaired (CU) (n=679) and between common dementia subtypes: Alzheimer’s disease (AD) (n=310), Lewy body dementia (DLB) (n=151), and behavioral variant frontotemporal dementia (bvFTD) (n=99). The logistic regression classifier was evaluated using receiver‐operator characteristic (ROC‐AUC) curves.ResultROC‐AUC curves for CU and each dementia subtype are illustrated in Figure 1a. A k nearest neighbors model was used to develop a graphical representation of the latent space where nodes are images and edges are drawn between nearest neighbors, illustrated in Figure 1b. The graph of the latent space demonstrates a comparison between dementia subtypes and CU images.ConclusionIn this study, we developed a proof‐of‐concept VAE that can achieve differential diagnosis with 84.8% balanced accuracy. The model architecture provides visual interpretability, and shows that the hippocampi, parietal lobes, and cerebellum, were useful for distinguishing AD, the frontal lobe, posterior cingulate cortex, and hippocampi were useful for distinguishing DLB, and prefrontal pole, subcortical structures, and occipital lobe were useful for distinguishing bvFTD. These regions of metabolism align with clinical pathology and metabolic patterns.

AbstractBackgroundParkinson’s disease and dementia with Lewy bodies are characterized by abnormal iron deposition in the substantia nigra, which can be measured with quantitative susceptibility mapping (QSM) on MRI. However, neuropathologic validation of the increased QSM in the substantia nigra associated with Lewy body disease (LBD) is lacking. We compared substantia nigra QSM on antemortem MRI between patients with and without Lewy‐related pathology which was confirmed by autopsy in all cases.MethodWe studied cases with antemortem QSM who underwent autopsy from the Mayo Clinic Alzheimer’s Disease Research Center and Mayo Clinic Study of Aging (n=54). Immunohistochemistry was performed with rabbit NACP antibody against the amino acids 98‐115 of the α‐synuclein protein to detect Lewy‐related pathology. After neuropathologic evaluation, cases were classified as LBD‐present (n=13) if they had Lewy‐related pathology and LBD‐absent (n=42) if they did not have Lewy‐related pathology. Alzheimer’s disease co‐pathology was present in 9 of the LBD‐present cohort (70%). QSM was calculated for the whole substantia nigra and the two sub‐regions: pars compacta and pars reticulata. Non‐parametric Wilcoxon rank‐sum test was used to compare substantia nigra QSM values between LBD‐present and LBD‐absent cases. Area under the curve (AUC) analyses tested the accuracy of QSM to distinguish the two groups.ResultsLBD‐present group had higher QSM values in the substantia nigra than LBD‐absent group. Furthermore, sub‐regional analyses showed that the LBD‐present group had higher QSM in the substantia nigra pars compacta (p=0.008) but not in the pars reticulata than the LBD‐absent group (Figure 1). QSM of the substantia nigra pars compacta distinguished LBD‐present and LBD‐absent cases with good accuracy (AUC=0.74; p=0.006; Figure 2).ConclusionThis study provides neuropathologic insight into the relevance of QSM of substantia nigra pars compacta as an in‐vivo biomarker of autopsy‐confirmed LBD. These findings indicate that imaging evidence of abnormal iron deposition in the substantia nigra may serve as a biomarker of Lewy‐related pathology in patients with primary LBD (i.e., Parkinson disease, dementia with Lewy bodies) and patients with potential LBD co‐pathology (e.g., patients with Alzheimer disease).

AbstractBackgroundDementia with Lewy bodies (DLB) frequently co‐occurs with Alzheimer’s disease (AD) pathologies, exacerbating disease progression. Biophysical models of diffusion imaging data, such as Neurite Orientation Dispersion and Density Imaging (NODDI), may reveal novel insights into the neurobiological substrates of AD on cortical and white matter microstructural injury.MethodA cohort of 57 DLB patients on the DLB spectrum (33 clinically probable DLB and 14 prodromal DLB) and 57 cognitively unimpaired (CU) controls underwent NODDI and PET imaging with [18F]‐Flortaucipir and [11C]‐PiB (Table 1). NODDI parameters were derived from white matter tracts and grey matter regions using the JHU “Eve” and Mayo Clinic Adult Lifespan Template atlases, respectively. Conditional logistic models were used to compare these parameters between the two groups, accounting for the 1:1 matching on age and sex. The associations of tau and amyloid‐β with NODDI parameters across the white matter tracts were examined with linear regressions. A region‐by‐region approach was adopted to evaluate the relationships of tau and amyloid‐β with grey matter NODDI, adjusting for age, APOE, and local grey matter volumes.ResultDLB spectrum patients exhibited extensive white matter abnormalities relative to CU, including lower fractional anisotropy, higher mean diffusivity (MD), and lower tissue‐weighted neurite density (tNDI). In grey matter, these deficits were paralleled with higher MD and lower tNDI and ODI, primarily in the temporoparietal and posterior cortices. Regression models revealed significant associations of elevated [18F]‐Flortaucipir uptake with deficits in temporal and limbic white matter tracts and a temporal‐lobe pattern of higher MD and lower tNDI in the cortex (Figure 1). No associations were observed with [11C]‐PiB in the DLB spectrum.ConclusionThese findings demonstrated an in vivo relationship between tau deposition and neurite deficits that spatially aligned with the canonical AD “cortical signature” of neurodegeneration in both white matter tracts and grey matter regions in DLB. Our study underscores the detrimental impact of AD pathology in DLB and provides a rationale for further investigations into the underlying mechanisms of AD in DLB. Future longitudinal studies tracking the progression of these changes could inform therapeutic strategies for DLB in the presence of AD pathologies

AbstractBackgroundSleep’s crucial role in maintaining brain health is increasingly recognized, particularly due to the rising prevalence of neurodegenerative diseases. It not only supports cognitive function but also aids in clearing brain metabolic waste through cerebrospinal fluid (CSF) dynamics. During sleep, especially in the non‐rapid eye movement (NREM) phase, CSF flow increases, essential for removing neurotoxic substances like amyloid‐beta proteins. This study explores the link between sleep respiratory patterns and CSF flow promotion.MethodElectroencephalography (EEG), and functional magnetic resonance imaging (fMRI) were simultaneously acquired during sleep from Gu, et al. (n = 29) (Gu et al. 2022). Respiration, measured via a respiratory belt, calculated respiratory volume (RV) as the standard deviation within a 5 second sliding window. EEG data processed delta band power (1‐4Hz) averaged across channels. Parcellated fMRI data using the brain template followed by the calculation of the averaged gray matter (GM) blood‐oxygen‐level‐dependent (BOLD) signal using the GM mask. CSF inflow was manually extracted from the fMRI’s bottom slice at the foramen magnum level. Subsequently, all signals were temporally aligned to the peak of the CSF inflow signal.ResultAt the time of CSF inflow (time = 0 sec. in the Figure 1A‐D), we observed an earlier increase in RV and EEG delta power, followed by changes in the GM BOLD signal (Figure 1). EEG and RV changes preceded the BOLD signal changes, which, in turn, preceded changes in CSF flow. Sensitivity to detect RV involvement during MRI varied with the placement of the respiratory belt, being least sensitive on the chest and more on the abdomen.ConclusionThese findings raise an interesting question on the role respiratory patterns during sleep, particularly in the NREM phase. Speculatively, RV change may influence CSF circulation during NREM sleep (Figure 1E). This enhancement in CSF flow is crucial for the clearance of metabolic waste from the brain, potentially reducing the risk of neurodegenerative diseases. Our findings indicate that respiratory volume may serve as a significant marker for CSF circulation during sleep.

Abstract Background Many proposed clinical decision support systems (CDSS) require multiple disparate data elements as input, which makes implementation difficult, and furthermore have a black‐box nature leading to low interpretability. Fluorodeoxyglucose Positron Emission Tomography (FDG‐PET) is an established modality for the diagnosis of dementia, and a CDSS that uses only an FDG‐PET image to produce a reliable and understandable result would ease both of these challenges to clinical application. Method A deep variational autoencoder (VAE) was used to extract a latent representation of each image through prior training from FDG‐PET brain images (n=2000). This unsupervised VAE has a novel graph convolutional architecture that makes it applicable to masked template space images. A logistic regression model was used to classify each image. A parametric study of the latent space revealed the imaging features that were used in the logistic regression model to differentiate each class. A separate dataset of participants labeled with their clinical diagnosis (n=1239) was used to assess the logistic regression model’s diagnostic accuracy of cognitively unimpaired (CU) (n=679) and between common dementia subtypes: Alzheimer’s disease (AD) (n=310), Lewy body dementia (DLB) (n=151), and behavioral variant frontotemporal dementia (bvFTD) (n=99). The logistic regression classifier was evaluated using receiver‐operator characteristic (ROC‐AUC) curves. Result ROC‐AUC curves for CU and each dementia subtype are illustrated in Figure 1a. A k nearest neighbors model was used to develop a graphical representation of the latent space where nodes are images and edges are drawn between nearest neighbors, illustrated in Figure 1b. The graph of the latent space demonstrates a comparison between dementia subtypes and CU images. Conclusion In this study, we developed a proof‐of‐concept VAE that can achieve differential diagnosis with 84.8% balanced accuracy. The model architecture provides visual interpretability, and shows that the hippocampi, parietal lobes, and cerebellum, were useful for distinguishing AD, the frontal lobe, posterior cingulate cortex, and hippocampi were useful for distinguishing DLB, and prefrontal pole, subcortical structures, and occipital lobe were useful for distinguishing bvFTD. These regions of metabolism align with clinical pathology and metabolic patterns.