Citations

Machine-learning (ML) models have the potential to transform health care by enabling more personalized and data-driven clinical decision making. However, their successful implementation in clinical practice requires careful consideration of factors beyond predictive accuracy. We provide an overview of essential considerations for developing clinically applicable ML models, including methods for assessing and improving calibration, selecting appropriate decision thresholds, enhancing model explainability, identifying and mitigating bias, as well as methods for robust validation. We also discuss strategies for improving accessibility to ML models and performing real-world testing. Highlights This tutorial provides clinicians with a comprehensive guide to implementing machine-learning classification models in clinical practice. Key areas covered include model calibration, threshold selection, explainability, bias mitigation, validation, and real-world testing, all of which are essential for the clinical deployment of machine-learning models. Following these guidance can help clinicians bridge the gap between machine-learning model development and real-world application and enhance patient care outcomes.

Abstract Brain-derived extracellular vesicles (BDEVs) carry tau filaments and promote tau transmission in Alzheimer’s disease (AD). However, how APOE ε4 allele, a key genetic risk factor for AD, may change BDEV molecular structures thereby facilitate disease progression is poorly understood. Here we report comprehensive analyses of BDEVs isolated from human E3/3 and E4/4 AD brains with a biological multi-omics approach. E4/4 BDEVs significantly enhanced tau propagation in aged human MAPT (Tau) knock-in and APPNL-G-F: Tau double knock-in mouse brains in vivo and increased neuronal uptake and excitability in induced pluripotent stem cell-derived neurons (iNeurons) compared to E3/3 BDEVs in vitro. Notably, correlation analysis of BDEV-lipidome and proteome exhibited synergistic enrichment in unsaturated free fatty acid (FFA)18:2, a precursor of inflammatory w6 FFA, and neural cell adhesion molecule 1 (NCAM1). Treatment of iNeurons with FFA 18:2 induces NCAM1 expression, recruits tau into EVs and enhance their tau seeding activity, which is blocked by NCAM1 antibody in vitro . Finally, intracerebroventricular injection of NCAM1 antibody significantly alleviated pathological tau accumulation and glial inflammation in PS19 tauopathy mouse brains, which had previously been reported to exhibit increased level of 18:2 FFA. This highlights novel pathological mechanism in tau transfer mediated by E4/4 BDEVs and emphasizes strong therapeutic potential of targeting EV molecules in AD progression.

Importance High-quality evidence regarding suppressive valacyclovir treatment in herpes zoster ophthalmicus (HZO) is necessary to guide care. Objective To determine whether suppressive valacyclovir compared with placebo delays the occurrence of new or worsening stromal keratitis (SK), endothelial keratitis (EK), iritis, or dendriform epithelial keratitis (DEK) during 12 months of treatment and if treatment benefit persisted at 18 months (secondary end point). Design, Setting, and Participants The Zoster Eye Disease Study (ZEDS) was a randomized clinical trial conducted in 95 sites from November 2017 to June 2024. Immunocompetent, nonpregnant adults with a history of an HZO rash, documented active keratitis or iritis within 1 year, and an estimated glomerular filtration rate of 45 mL/min/1.73 m 2 or greater were eligible. After determined to be eligible, participants were randomized in 4 strata: age at onset (<60 years vs ≥60 years) and disease duration (<6 months vs ≥6 months). Interventions A total of 12 months of double-masked daily valacyclovir, 1000 mg, or placebo. Main Outcomes and Measures The primary outcome was time to first occurrence within 12 months of new or worsening SK, EK, iritis, or DEK. Results A total of 527 participants (median [IQR] age, 60 [50-68] years; 266 female [50.5%]; 266 in the valacyclovir group; 261 in the placebo group) were randomized in 4 strata; 481 completed 12 months, and 460 completed 18 months. Data were analyzed by intention to treat. At 12 months, primary end points occurred in 86 participants (33%) assigned to placebo and 74 (28%) assigned to valacyclovir, and at 18 months in 104 participants (40%) assigned to placebo and 86 (32%) assigned to valacyclovir. The hazard ratio (HR) of the primary end point at 12 months was 0.77 for participants taking valacyclovir vs placebo (HR, 0.77; adjusted 95% CI, 0.56-1.05; P = .09) and 0.73 at the secondary end point at 18 months (HR, 0.73; adjusted 95% CI, 0.55-0.97; P = .03). There was a reduction of multiple other secondary end points at 12 months (HR, 0.70; 95% CI, 0.52-0.95; P = .02) and 18 months (HR, 0.72; 95% CI, 0.55-0.95; P = .02). Conclusions and Relevance Although the primary outcome did not show a benefit of suppressive valacyclovir treatment, secondary study outcomes showed treatment superiority at the 18-month end point and reduced number of multiple episodes of keratitis or iritis at both 12 and 18 months. These results support consideration of 1 year of suppressive valacyclovir treatment for HZO. Trial Registration ClinicalTrials.gov Identifier: NCT03134196

Importance Evidence regarding suppressive valacyclovir treatment on postherpetic neuralgia is necessary to guide care. Objective To test the hypothesis that suppressive treatment with 1000 mg/d of oral valacyclovir for 12 months reduces the prevalence, severity, and duration of postherpetic neuralgia compared with placebo at 12 and 18 months in participants with herpes zoster ophthalmicus (HZO). Design, Setting, and Participants Multicenter, placebo-controlled randomized clinical trial including 527 immunocompetent, nonpregnant adults with history of HZO rash, documented keratitis, or iritis within 1 year and an estimated glomerular filtration rate of 45 mL/min/1.73 m 2 or greater. The study was conducted at 95 participating sites (in Canada, New Zealand, and the US) from November 2017 to June 2024 and participant visits occurred every 3 months. Intervention Treatment with 1000 mg/d of valacyclovir or placebo for 12 months. Main Outcomes and Measures Prevalence of postherpetic neuralgia, severity as determined by pain score (a score of ≥3 on a scale of 1-10), pain duration (≥3 months after HZO onset), and total daily dose of pain medication. Results Of the 527 participants (490 completed 12 months of treatment and 460 completed 18 months), 73 (14%) had postherpetic neuralgia and were analyzed by age at HZO onset (<60 years or ≥60 years) and disease duration (recent [<6 months] or chronic [≥6 months]). Of the 73 participants with postherpetic neuralgia (34 in the valacyclovir group and 39 in the placebo group), the mean age was 62.4 years (SD, 13.6 years), 59% were female, 5% were Black or African American, and 10% were Hispanic. The prevalence of postherpetic neuralgia at 12 months was not reduced by valacyclovir (12/32 [38%]) compared with placebo (14/35 [40%]) (between-group difference, 2.5% [95% CI, −20.8% to 25.8%]; P >.99). The participants who were younger than 60 years at HZO onset and had a chronic disease duration had lower pain scores in the valacyclovir group (mean score, 0.3 [SD, 0.9]) vs the placebo group (mean score, 0.8 [SD, 1.9]) at 12 months ( P = .045) and at 18 months (mean score, 0.2 [SD, 0.9] vs 1.0 [SD, 2.3], respectively; P = .02). There was a decrease in pain duration in the valacyclovir group at 18 months (mean, 13.6 [SD, 11.4] months) vs the placebo group (mean, 18.7 [SD, 29.5] months) (linear mixed-effects model between-group difference, −3.39 months [95% CI, −6.73 to −0.04 months]; P = .046). The total daily dose of neuropathic pain medication was lower in the valacyclovir group (mean, 271.4 [SD, 593.8] mg/d) vs the placebo group (mean, 363.4 [SD, 592.2] mg/d) at 12 months (linear mixed-effects model P = .006) and at 18 months (mean, 209.0 [SD, 412.8] mg/d vs 286.2 [SD, 577.9] mg/d, respectively; linear mixed-effects model P = .01). Conclusions and Relevance One year of suppressive treatment with valacyclovir was associated with a lower dosage of neuropathic pain medication. Participants in the valacyclovir group, who were younger at HZO onset and had a chronic disease duration, had lower pain scores. These secondary outcomes support consideration of 1 year of suppressive treatment with valacyclovir to reduce dosage of pain medications and pain due to HZO. Trial Registration ClinicalTrials.gov Identifier: NCT03134196

Central nervous system (CNS) tumors are the leading cause of cancer-related mortality in children, with prognosis remaining dismal for some of these malignancies. Though the past two decades have seen advancements in surgery, radiation, and targeted therapy, major unresolved hurdles continue to undermine the therapeutic efficacy. These include challenges in suboptimal drug delivery through the blood–brain barrier (BBB), marked intra-tumoral molecular heterogeneity, and the elusive tumor microenvironment. Drug repurposing or re-tasking FDA-approved drugs with evidence of penetration into the CNS, using newer methods of intracranial drug delivery facilitating optimal drug exposure, has been an area of intense research. This could be a valuable tool, as most of these agents have already gone through the lengthy process of drug development and the evaluation of safety risks and the optimal pharmacokinetic profile. They can now be used and tested in clinics with an accelerated and different approach. Conclusions: The next-generation therapeutic strategy should prioritize repurposing oncologic and non-oncologic drugs that have been used for other indication, and have demonstrated robust preclinical activity against pediatric brain tumors. In combination with novel drug delivery techniques, these drugs could hold significant therapeutic promise in pediatric neurooncology.

Abstract Recent clinical trials on slowing dementia progression have led to renewed focus on finding safer, more effective treatments. One approach to identify plausible candidates is to assess whether existing medications for other conditions may affect dementia risk. We conducted a systematic review to identify studies adopting a data‐driven approach to investigate the association between a wide range of prescribed medications and dementia risk. We included 14 studies using administrative or medical records data for more than 130 million individuals and 1 million dementia cases. Despite inconsistencies in identifying specific drugs that may modify Alzheimer's or dementia risk, some themes emerged for drug classes with biological plausibility. Antimicrobials, vaccinations, and anti‐inflammatories were associated with reduced risk, while diabetes drugs, vitamins and supplements, and antipsychotics were associated with increased risk. We found conflicting evidence for antihypertensives and antidepressants. Drug repurposing for use in dementia is an urgent priority. Our findings offer a basis for prioritizing candidates and exploring underlying mechanisms. Highlights We present a systematic review of studies reporting association between drugs prescribed for other conditions and risk of dementia including 139 million people and 1 million cases of dementia. Our work supports some previously reported associations, for example, showing decreased risk of dementia with drugs to treat inflammatory disease and increased risk with antipsychotic treatment. Antimicrobial treatment was perhaps more surprisingly associated with decreased risk, supportive of recent increased interest in this potential therapeutic avenue. Our work should help prioritize drugs for entry into adaptive platform trials in Alzheimer's disease and will serve as a useful resource for those investigating drugs or classes of drugs and risk of dementia.