Koga, Shunsuke

Abstract Gross brain findings help guide differential diagnosis, but their diagnostic value across major neurodegenerative diseases remains incompletely characterized. We evaluated whether gross descriptions in autopsy reports could classify cases by final neuropathologic diagnosis. We analyzed 5,613 autopsy cases from the Mayo Clinic Brain Bank collected between 1998 and 2023, including seven diagnostic categories: Alzheimer disease (AD), Lewy body disease (LBD), combined AD and LBD (AD-LBD), frontotemporal lobar degeneration, corticobasal degeneration, progressive supranuclear palsy (PSP), and multiple system atrophy (MSA). A fine-tuned large language model (LLM) converted narrative gross descriptions into semi-quantitative scores for 39 features. Extraction accuracy was 0.95 in 200 manually annotated feature-level test examples. A CatBoost classifier used the extracted scores as input, whereas a second fine-tuned LLM used standardized textual descriptions of the same features. Both classifiers also included age at death, sex, and brain weight. Performance was evaluated in the same held-out test set of 562 cases. CatBoost achieved an accuracy of 0.73, a Cohen’s kappa of 0.65, and a macro-average area under the receiver operating characteristic curve of 0.92. The text-based LLM achieved an accuracy of 0.75 and a kappa of 0.68. Macro-average sensitivity was 0.66 for both models. PSP sensitivity was 0.92 with CatBoost and 0.93 with text-based LLM. Corresponding sensitivities were 0.87 and 0.92 for MSA, but only 0.21 and 0.03 for AD-LBD. Feature attribution analysis identified subthalamic nucleus atrophy and putaminal abnormalities as contributors to PSP and MSA predictions, respectively. Narrative gross descriptions contained diagnostic information, particularly for disorders with distinctive macroscopic patterns, whereas combined AD-LBD remained difficult to distinguish.

Objective Lewy body disease (LBD) is a complex neurodegenerative disorder characterized by the accumulation of misfolded α‐synuclein in the brain. Neuroinflammation has long been implicated in LBD pathogenesis, and recent genetic studies in Parkinson's disease (a clinical manifestation of LBD) have shown consistent association with the human leukocyte antigen (HLA) gene complex. Here, we assessed whether variation in HLA alleles influences neuropathological burden in a neuropathologically‐defined series of LBD cases. Methods We conducted a comprehensive analysis of HLA allelic variants in a cohort of 539 LBD cases of European descent from the Mayo Clinic brain bank. High‐resolution whole‐genome sequencing was used, and the HLA alleles of each sample were called using the HLA*LA tool and assessed for association with neuropathological outcomes. Results Our analysis identified 1 significant ( P < 3.43 × 10 −5 ) and 4 suggestive ( P < 0.001) associations between certain HLA alleles and specific neuropathological outcomes in LBD, suggesting a potential role for HLA‐mediated immune mechanisms in disease progression and subtype differentiation. Specifically, HLA‐DPB1*06:01 was significantly associated with lower Lewy body counts in the parahippocampal gyrus ( P = 3.30 × 10 −5 ), with weaker and suggestive associations observed in the middle frontal ( P = 1.80 × 10 −4 ) and inferior parietal gyrus ( P = 6.33 × 10 −4 ). Additionally, although only suggestive, HLA‐DRB1*11:01 correlated with a lower Thal amyloid phase ( P = 1.56 x 10 −4 ), and HLA‐B*15:01 correlated with an increased risk of diffuse LBD ( P = 7.58 x 10 −4 ). Interpretation This study provides a detailed evaluation of the relationship between HLA alleles and LBD pathology, highlighting the importance of immune‐related genetic factors in the etiology of LBD. ANN NEUROL 2026;99:492–501

Abstract Studies assessing genetic associations with neuropathological features in Lewy body disease (LBD) have been limited to candidate gene investigations, and therefore, information is lacking regarding the genetic architecture of the neuropathology of LBD. In the current study, we examined a large series of neuropathologically confirmed LBD cases (n = 980 in the discovery series, n = 503 in the replication series) and performed genome-wide association studies of 11 different neuropathological outcome measures. The 11 neuropathological outcomes included Braak neurofibrillary tangle (NFT) stage, Thal amyloid phase, LBD subtype, Lewy body (LB) counts in five different brain regions, dorsolateral and ventromedial putaminal tyrosine hydroxylase immunoreactivity and neuronal loss in the ventrolateral part of the substantia nigra. Associations between variants and outcomes were assessed using regression models appropriate for the nature of the given neuropathological outcome and that were adjusted for age at death, sex and top principal components of genetic data. In the discovery series, APOE rs429358 (i.e. APOE ε4) was associated with a greater severity of each of Braak NFT stage [odds ratio (OR) = 3.07, P = 2.34 × 10−32], Thal amyloid phase (OR = 3.57, P = 3.28 × 10−29) and LBD subtype (OR = 1.78, P = 9.85 × 10−9), with similar findings observed in the independent replication series (Braak NFT stage, OR = 2.30, P = 2.70 × 10−11; Thal amyloid phase, OR = 3.17, P = 6.39 × 10−18; LBD subtype, OR = 2.68, P = 3.85 × 10−10). In the subgroup of cases with lower levels of Alzheimer's disease pathology (Braak NFT stage ≤ III and Thal amyloid phase ≤2), there was a strong association between APOE rs429358 and LBD subtype even when adjusting for Braak stage and Thal phase in the discovery series (n = 218, OR = 2.47, P = 0.007) and the replication series (n = 141, OR = 3.60, P = 0.006). Although additional genome-wide significant associations were identified in the discovery series between LINC01581/MCTP2 rs547411734 and lower middle frontal LB counts, between TLE3 rs3743309 and lower cingulate LB counts, and between GRIN2A/ATF7IP2 rs1097915 and lower parahippocampal LB counts, these findings were not observed in the replication series. Our results indicate that the APOE ε4 allele is the most prominent genetic determinant of severity of neuropathology in LBD. These findings represent a key step forward in our understanding of genetic drivers of neuropathological features in LBD. Future studies utilizing meta-analytical approaches will be important to more precisely assess other associations that were not quite genome-wide significant in the discovery series.

Aggregated α-synuclein (α-SYN) proteins, encoded by the SNCA gene, are hallmarks of Lewy body disease (LBD), affecting multiple brain regions. However, the specific mechanisms underlying α-SYN pathology in cortical neurons, crucial for LBD-associated dementia, remain unclear. Here, we recapitulated α-SYN pathologies in human induced pluripotent stem cells (iPSCs)–derived cortical organoids generated from patients with LBD with SNCA gene triplication. Single-cell RNA sequencing, combined with functional and molecular validation, identified synaptic and mitochondrial dysfunction in excitatory neurons exhibiting high expression of the SNCA gene, aligning with observations in the cortex of autopsy-confirmed LBD human brains. Furthermore, we screened 1280 Food and Drug Administration–approved drugs and identified four candidates (entacapone, tolcapone, phenazopyridine hydrochloride, and zalcitabine) that inhibited α-SYN seeding activity in real-time quaking-induced conversion assays with human brains, reduced α-SYN aggregation, and alleviated mitochondrial dysfunction in SNCA triplication organoids and excitatory neurons. Our findings establish human cortical LBD models and suggest potential therapeutic drugs targeting α-SYN aggregation for LBD.

AbstractAimsRecent advances in artificial intelligence, particularly with large language models like GPT‐4Vision (GPT‐4V)—a derivative feature of ChatGPT—have expanded the potential for medical image interpretation. This study evaluates the accuracy of GPT‐4V in image classification tasks of histopathological images and compares its performance with a traditional convolutional neural network (CNN).MethodsWe utilised 1520 images, including haematoxylin and eosin staining and tau immunohistochemistry, from patients with various neurodegenerative diseases, such as Alzheimer's disease (AD), progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD). We assessed GPT‐4V's performance using multi‐step prompts to determine how textual context influences image interpretation. We also employed few‐shot learning to enhance improvements in GPT‐4V's diagnostic performance in classifying three specific tau lesions—astrocytic plaques, neuritic plaques and tufted astrocytes—and compared the outcomes with the CNN model YOLOv8.ResultsGPT‐4V accurately recognised staining techniques and tissue origin but struggled with specific lesion identification. The interpretation of images was notably influenced by the provided textual context, which sometimes led to diagnostic inaccuracies. For instance, when presented with images of the motor cortex, the diagnosis shifted inappropriately from AD to CBD or PSP. However, few‐shot learning markedly improved GPT‐4V's diagnostic capabilities, enhancing accuracy from 40% in zero‐shot learning to 90% with 20‐shot learning, matching the performance of YOLOv8, which required 100‐shot learning to achieve the same accuracy.ConclusionsAlthough GPT‐4V faces challenges in independently interpreting histopathological images, few‐shot learning significantly improves its performance. This approach is especially promising for neuropathology, where acquiring extensive labelled datasets is often challenging.

AbstractAmyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder characterized by upper and lower motor neuron signs. There are, however, cases where upper motor neurons (UMNs) are predominantly affected, leading to clinical presentations of UMN‐dominant ALS or primary lateral sclerosis. Furthermore, cases exhibiting an UMN‐predominant pattern of motor neuron disease (MND) presenting with corticobasal syndrome (CBS) have been sparsely reported. This study aims to clarify the clinicopathological features of patients with UMN‐predominant MND. We reviewed 24 patients with UMN‐predominant MND with TDP‐43 pathology in the presence or absence of frontotemporal lobar degeneration. Additionally, we reviewed the medical records of patients with pathologically‐confirmed corticobasal degeneration (CBD) who received a final clinical diagnosis of CBS (n = 10) and patients with pathologically‐confirmed progressive supranuclear palsy (PSP) who received a final clinical diagnosis of PSP syndrome (n = 10). Of 24 UMN‐predominant MND patients, 20 had a clinical diagnosis of an atypical parkinsonian disorder, including CBS (n = 11) and PSP syndrome (n = 8). Only two patients had antemortem diagnoses of motor neuron disease. UMN‐predominant MND patients with CBS less frequently exhibited apraxia than those with CBD, and they were less likely to meet clinical criteria for possible or probable CBS. Similarly, UMN‐predominant MND patients with PSP syndrome less often met clinical criteria for probable PSP than PSP patients with PSP syndrome. Our findings suggest that UMN‐predominant MND can mimic atypical parkinsonism, and should be considered in the differential diagnosis of CBS and PSP syndrome, in particular when criteria are not met.

Abstract Introduction: Recent advancements in artificial intelligence, particularly with large language models like GPT-4Vision (GPT-4V)—a derivative feature of ChatGPT—have expanded the potential for medical image interpretation. This study evaluates the accuracy of GPT-4V in image classification tasks of histopathological images and compares its performance with traditional convolutional neural network (CNN). Methods: We utilized 1515 images, including hematoxylin and eosin staining and tau immunohistochemistry, from patients with various neurodegenerative diseases, such as Alzheimer’s disease (AD), progressive supranuclear palsy (PSP), and corticobasal degeneration (CBD). We assessed GPT-4V’s performance using multi-step prompts to determine how textual context influences image interpretation. We also employed few-shot learning to enhance improvements in GPT-4V’s diagnostic accuracy in classifying three specific tau lesions—astrocytic plaques, neuritic plaques, and tufted astrocytes—and compared the outcomes with the CNN model YOLOv8. Results: GPT-4V accurately recognized staining techniques and tissue origin but struggled with specific lesion identification. The interpretation of images was notably influenced by the provided textual context, which sometimes led to diagnostic inaccuracies. For instance, when presented with images of the motor cortex, the diagnosis shifted inappropriately from AD to CBD or PSP. However, few-shot learning markedly improved GPT-4V’s diagnostic capabilities, enhancing accuracy from 40% in zero-shot learning to 90% with 20-shot learning, matching the performance of YOLOv8, which required 100-shot learning to achieve the same accuracy. Conclusions: Although GPT-4V faces challenges in independently interpreting histopathological images, few-shot learning significantly improves its accuracy. This approach is especially promising for neuropathology, where acquiring extensive labeled datasets is often challenging.

AbstractBackgroundMixed pathology is common at autopsy for a number of age‐associated neurodegenerative disorders; however, the frequency of comorbid pathologies in multiple system atrophy (MSA) and their clinical correlations are poorly understood.ObjectiveWe determined the frequency of comorbid pathologic processes in autopsy‐confirmed MSA and assessed their clinical correlates.MethodsThis study included 160 neuropathologically established MSA from the Mayo Clinic brain bank. Clinical information, including age at onset or death, clinical subtype, initial symptoms, antemortem clinical diagnosis, and cognitive dysfunction was collected. We assessed comorbid pathologies including Alzheimer's disease neuropathologic change, Lewy‐related pathology, argyrophilic grain disease, age‐related τ astrogliopathy, transactive DNA‐binding protein 43 pathology, cerebral amyloid angiopathy, and cerebrovascular small vessel disease and examined their clinical impact.ResultsThe majority of MSA patients (62%) had no significant comorbid pathologies. There was a positive correlation between age at onset or death with the number of comorbid pathologies; however, even in the highest quartile group (average age at death 78 ± 6 years), the average number of comorbid pathologies was <2. Logistic regression analysis revealed that none of the assessed variables, including sex, age at onset, and the presence or absence of each comorbid pathology, were significantly associated with cognitive dysfunction.ConclusionsThe majority of MSA patients do not have comorbid pathologies, even in advanced age, indicating that MSA is unique among neurodegenerative disorders in this regard. There was minimal clinical impact of comorbid pathologies in MSA. These findings warrant focusing on α‐synuclein for the treatment strategy for MSA. © 2023 International Parkinson and Movement Disorder Society.