Citations

Abstract Artificial intelligence is spreading through public health agencies and healthcare systems faster than workforces are trained to govern it. We describe RHAMI (Responsible Health AI Readiness and Maturity Index), the first human-centered automated composite index for health organizations to score AI readiness and maturity as one measurable system of science, economics and ethics, and the physician certification-to-Master’s pathway built to teach what it measures—both designed jointly by frontline public health professionals, clinicians, ethicists, engineers, leaders, and policy-makers in low- and middle-income countries and high-income countries.

ABSTRACT Background Coeliac disease is due to an acquired, lifelong gluten‐specific immune response resulting in a gluten‐dependent chronic immune‐mediated enteropathy associated with gastrointestinal and extraintestinal morbidity. Strictly avoiding dietary gluten is the only available management but has never been subjected to rigorous efficacy assessment now required for a pharmaceutical agent. Gluten‐free diet (GFD) is partially and inconsistently effective, but it places a high burden on patients, families, the food industry, and regulators and is defined inconsistently around the world, highlighting the need for effective and convenient adjunctive pharmacological therapies. Aims To summarize key patient, scientific, and regulatory considerations for coeliac disease drug development based on a multi‐stakeholder meeting focused on accelerating clinical trials. Methods A pre‐competitive meeting convened by the Beyond Celiac Coalition (May 2024) included patients with coeliac disease, academic experts, pharmaceutical industry representatives, and participants from the US Food and Drug Administration. Moderated panel addressed patient priorities, symptom and histologic endpoints, biopsy and histology methods, gluten challenge design, clinical trial duration, and regulatory expectations. Themes were synthesized qualitatively with focus on trial design implications and unresolved questions. Results Participants described high symptom burden and psychosocial impact despite gluten‐free diet adherence underscoring unmet therapeutic need in coeliac disease. Regulatory perspectives emphasized the need for clinically meaningful improvement and evidence of intestinal benefit, framed as symptom and histology co‐primary endpoints in clinical trials. These expectations were discussed as most applicable to trial populations with both clinically meaningful symptoms and active histologic injury at baseline; for other phenotypes, it was emphasized that endpoint strategy, the role of gluten exposure, and trial duration should be tailored case by case to the population and mechanism of action. Practical challenges identified included endpoint selection and definition, discordance between symptoms and histology, selecting fit‐for‐purpose symptom measures; standardizing biopsy acquisition and central histology reads to address patchiness and variability, pre‐specifying definitions of meaningful histologic change given ongoing uncertainty about healing, and trial feasibility. Controlled gluten exposure was discussed as a tool to improve interpretability of efficacy outcomes, with careful consideration of dose, duration, patient burden and safety. Discussion addressed the ideal length of the double‐blind period, including a shorter 24‐week double‐blind study with an open‐label extension versus a 52‐week double‐blind design. Trial duration should balance scientific rigour, patient retention, and the need for longer‐term safety and durability data, with flexibility to tailor the approach to context. Conclusions Progress in coeliac disease drug development will require patient‐centred, scientifically robust but feasible trial designs with appropriate endpoints, justified use of gluten exposure, consideration of improved biomarkers to reduce patient burden and early regulatory engagement. Continued multi‐stakeholder collaboration, including dedicated consideration of paediatric drug development, is essential to advance effective therapies.

Introduction Cognitive symptoms such as ‘brain fog’ are frequent and disabling in patients with SLE. We aimed to develop and validate the Lupus Brain Fog Severity Scale (LBFSS), the first patient-reported outcome measure (PROM) for cognitive symptoms in SLE. Methods The LBFSS was developed using a multi-step international approach involving healthcare professionals and patients with SLE. Patients provided free-text descriptions of brain fog, generating candidate domains through thematic analysis (step 1). Domains were rated for relevance (step 2) and those retained converted into self-assessable items (step 3). The pilot LBFSS was iteratively refined (step 4). The finalised LBFSS was validated (step 5) in a large international SLE cohort. Psychometric evaluation included internal consistency, exploratory factor analysis and assessment of construct, convergent, known-groups and criterion-related validity. Results In step 1, 147 SLE participants across 39 countries generated 21 domains for cognitive symptoms, of which 13 were retained in consensus (step 2), converted into self-assessable items (step 3) and iteratively refined using patient and physician feedback. In step 5, validation of the LBFSS instrument in 378 patients from 36 countries demonstrated high internal consistency (Cronbach’s α=0.95) and a unidimensional structure explaining 62.2% of variance. LBFSS scores were significantly higher in participants reporting brain fog (p<0.0001), confirming known-group validity. Construct validity was supported by significant correlations with functional impact and symptom severity (ρ=0.46–0.65, p<0.0001), strong convergent validity with the Perceived Deficits Questionnaire-20 (ρ=0.84, p<0.0001) while the discriminative performance for presence versus absence of self-reported brain fog was high (area under the curve=0.85). Conclusion The LBFSS is a new, valid and feasible PROM to assess brain fog and cognitive symptoms in SLE.

We highlight the complex management considerations in a patient with cerebral amyloid angiopathy who developed venous thromboembolism, explicitly navigating the balance between risks of thromboembolism and bleeding with anticoagulation therapy. This case highlights a man in his 70s who presented to the emergency department with an acute right frontal intracerebral haemorrhage secondary to cerebral amyloid angiopathy with an acute segmental pulmonary embolism managed with an inferior vena cava (IVC) filter. During hospitalisation, the patient developed a large expansile thrombus extending inferior from the IVC filter to the right popliteal vein. This necessitated the initiation of anticoagulation therapy, beginning with unfractionated heparin, transitioning to enoxaparin and ultimately to apixaban. He was treated with 3 months of anticoagulation with no haemorrhagic complications. The patient’s successful outcome underscored the potential for safe and effective anticoagulation management in this complex clinical scenario.

Poor diet is the leading preventable risk factor for death worldwide, associated with over 10 million premature deaths and USD 8 trillion related costs every year. Artificial intelligence or AI is rapidly emerging as the most historically disruptive, innovatively dynamic, rapidly scaled, cost-efficient, and economically productive technology (which is increasingly providing transformative countermeasures to these negative health trends, especially in low- and middle-income countries (LMICs) and underserved communities which bear the greatest burden from them). Yet widespread confusion persists among healthcare systems and policymakers on how to best identify, integrate, and evolve the safe, trusted, effective, affordable, and equitable AI solutions that are right for their communities, especially in public health nutrition. We therefore provide here the first known global, comprehensive, and actionable narrative review of the state of the art of AI-accelerated nutrition assessment and healthy eating for healthcare systems, generated by the first automated end-to-end empirical index for responsible health AI readiness and maturity: the Responsible Health AI readiness and Maturity Index (RHAMI). The index is built and the analysis and review conducted by a multi-national team spanning the Global North and South, consisting of front-line clinicians, ethicists, engineers, executives, administrators, public health practitioners, and policymakers. RHAMI analysis identified the top-performing healthcare systems and their nutrition AI, along with leading use cases including multimodal edge AI nutrition assessments as ambient intelligence, the strategic scaling of practical embedded precision nutrition platforms, and sovereign swarm agentic AI social networks for sustainable healthy diets. This index-based review is meant to facilitate standardized, continuous, automated, and real-time multi-disciplinary and multi-dimensional strategic planning, implementation, and optimization of AI capabilities and functionalities worldwide, aligned with healthcare systems’ strategic objectives, practical constraints, and local cultural values. The ultimate strategic objectives of the RHAMI’s application for AI-accelerated public health nutrition are to improve population health, financial efficiency, and societal equity through the global cooperation of the public and private sectors stretching across the Global North and South.

Abstract Background: For LS-DLBCL, clinical trials (FLYER, LYSA/GOELAMS 02-03, LYSA LNH09-1B, S1001) have shown that 4 cycles of rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone (RCHOP4) without radiation (RT) leads to durable remissions in &gt;90% of cases. However, only patients (pts) with non-bulky disease and/or international prognostic index (IPI) of 0 and complete metabolic response (CMR) on interim positron emission tomography (PET) scan, if performed, were eligible for RCHOP4 alone in these studies. Outcomes in trial-ineligible LS-DLBCL and the appropriate real-world candidates for RCHOP4 are still undefined. Methods: We conducted a multicenter retrospective study of adult pts with stage I/II DLBCL of any tumor bulk, IPI and with or without B-symptoms diagnosed after 2011 and treated with RCHOP4 +/- 2 cycles R from 21 U.S. centers. Pts with PMBCL, PTLD or receiving non-standard dose RCHOP for cycle (C) 1 or planned consolidative RT were excluded. The primary endpoint was PFS from C1 by Kaplan-Meier method, not counting indolent lymphoma relapse. Event free survival (EFS) was time to next therapy without progression (including unplanned RT for non-CR), progression or death. A competing risks analysis of non-relapse mortality (NRM) vs. DLBCL progression was performed. Results: The characteristics of the 428 pts were: median age 60 (18-88), 60% male, 11% Hispanic and 4% non-Hispanic Black, 62% stage I, 15% elevated LDH, 3% ECOG PS&gt;1, 5% tumor bulk &gt;7cm, 9% had B-symptoms and median Charlson Comorbidity Index (CCI) was 4 (2-12). IPI was 0, 1 and 2-3 in 42%, 46% and 12% and stage-modified IPI (smIPI) was 0, 1, 2 and 3-4 in 25%, 48%, 21% and 6% of pts. 66% had nodal (59% head/neck, 11% pelvic, 9% multiple above diaphragm, 7% abdominal) and 46% extranodal sites (37% head/neck, 20% gastric/intestinal, 9% spleen, 8% skin/soft tissue); 6% had transformed/concurrent FL/MZL or grade 3B FL, 33% (138/408) were non-GCB cell of origin (COO), 25% (90/357) double expressor (DEL) and 0.5% (2/379) double-hit. 13% had completely resected disease, 97% staging PET and 47% bone marrow biopsy. Median diagnosis to treatment interval was 30 days (IQR 20-43), 6% had dose reductions after C1, 77% had an interim PET (iPET) after C2 (33%) or C3 (67%), 15% received 2 additional cycles R and 4% CNS prophylaxis. CMR rates by iPET were 83% after C2 and 94% after C3, and 95% by end-of-treatment PET. iPET PMR was more likely if smIPI&gt;2 or bulky disease. Median follow-up time was 2.7 years. There were 32 progression events: 14 occurred at the primary site and 17 at distant sites (1 unknown), and 12 occurred after 2 years. Another 11 pts received unplanned RT after RCHOP4 for non-CR. There were 24 deaths, 12 without prior progression. Cause of death was DLBCL in 7 and non-DLBCL in 17 cases (6 infection, 2 ILD, 2 heart failure, 2 second cancer, 1 seizure, 4 unknown). EFS, PFS and OS rates were 88%, 90% and 96% at 3 years. By univariable Cox regression, Black race, elevated LDH, smIPI&gt;2, DEL, increasing CCI, multiple subdiaphragmatic nodal sites and iPET PMR were associated with worse PFS. Non-significant variables of note included COO, extranodal sites, complete resection, missing bone marrow biopsy, missing iPET and additional cycles R. By multivariable analysis (including age&gt;60, stage, ECOG PS&gt;1, elevated LDH, bulky disease, CCI, B-symptoms and DEL), ECOG PS&gt;1 (HR 3.95 p=0.050), CCI (HR 1.29 p=0.003) and DEL (HR 2.02 p=0.069) were associated with worse PFS. Cumulative incidence of DLBCL progression and NRM were 8% and 2% at 3 years. Competing risk analysis by univariable models found that age&gt;60, ECOG PS&gt;1, smIPI&gt;2, elevated LDH, CCI and bulky disease were associated with NRM, while DEL and B-symptoms were associated with DLBCL progression.Conclusion: Presented is the largest retrospective cohort of LS-DLBCL treated uniformly with RCHOP4 to date. Despite modest follow-up time and lack of an “intention-to-treat“ population, this study shows that abbreviated RCHOP4 leads to excellent outcomes in the real-world, albeit with continuous risk of late relapse. smIPI risk factors were associated with worse PFS, though this may in part be driven by risk of non-lymphoma death. As pts with advanced age, poor performance status or comorbidities may benefit from reduced treatment exposure, RCHOP4 should still be considered for these and other “trial-ineligible” pts, particularly if other high-risk features (iPET PMR, DEL or B-symptoms) are absent.

Abstract Bioinformatics methods can be used to quantify mitochondrial DNA copy number from whole genome sequencing (WGS) data. We evaluated mitochondrial DNA copy number from human brain-derived WGS data using the fastMitoCalc tool. 341 Parkinson’s Disease cerebellum samples were compared with 74 age-matched controls from the North American Brain Expression Consortium. Parkinson’s Disease cerebellum had significantly higher mitochondrial DNA copy number compared with controls (P = 4.15e–7), and this effect was reproducible in four of five brain banks when analysis was restricted to each resource that contributed Parkinson’s Disease samples to this genetic dataset. Follow-on analyses of 128 Parkinson’s Disease cerebellum samples and 33 controls that had paired neuropathology data and clinical scores demonstrated a significant increase in mitochondrial DNA copy number with Unified Staging System for Lewy Body disorders stages and Unified Parkinson’s Disease Rating Scale (off meds) motor scores. Analysis of Lewy Body scores from ten brain regions showed cerebellum mitochondrial DNA copy number increased upon pathological infestation of α-synuclein aggregates in the brainstem and limbic system but did not increase after late-stage neocortical involvement. This genetics dataset supports previous observations of cerebellum activation in Parkinson’s Disease and suggests mitochondrial DNA copy number may increase to support this regional activation as a compensatory mechanism to pathology or motor symptoms.

Abstract Purpose Ultra‐low field MRI scanners have the potential to improve health care delivery, both through improved access in areas where there are few MRI scanners and allowing more frequent monitoring of disease progression and treatment response. This may be particularly true in white matter disorders, including leukodystrophies and multiple sclerosis, in which frequent myelin‐sensitive imaging, such as magnetization transfer (MT) imaging, might improve clinical care and patient outcomes. Methods We implemented an on‐resonance approach to MT imaging on a commercial point‐of‐care 64 mT scanner using a non‐balanced steady‐state free precession sequence. Phantom and in vivo experiments were used to evaluate and optimize the sequence sensitivity and reproducibility, and to demonstrate in vivo performance and inter‐site reproducibility. Results From phantom experiments, T 1 and T 2 effects were determined to have a negligible effect on the differential MT weighting. MT ratio (MTR) values in white matter were 23.1 ± 1.0% from 10 healthy volunteers, with an average reproducibility coefficient of variation of 1.04%. Normal‐appearing white matter MTR values in a multiple sclerosis participant (21.5 ± 6.2%) were lower, but with a similar spread of values, compared to an age‐matched healthy volunteer (23.3 ± 6.2%). Conclusion An on‐resonance MT imaging approach was developed at 64 mT that can be performed in as little as 4 min. A semi‐quantitative myelin‐sensitive imaging biomarker at this field strength is available for assessing both myelination and demyelination.

Interest in the potential applications of artificial intelligence in medicine, anesthesiology, and the world at large has never been higher. The Anesthesia Research Council steering committee formed an anesthesiologist artificial intelligence expert workgroup charged with evaluating the current state of artificial intelligence in anesthesiology, providing examples of future artificial intelligence applications and identifying barriers to artificial intelligence progress. The workgroup’s findings are summarized here, starting with a brief introduction to artificial intelligence for clinicians, followed by overviews of current and anticipated artificial intelligence–focused research and applications in anesthesiology. Anesthesiology’s progress in artificial intelligence is compared to that of other medical specialties, and barriers to artificial intelligence development and implementation in our specialty are discussed. The workgroup’s recommendations address stakeholders in policymaking, research, development, implementation, training, and use of artificial intelligence–based tools for perioperative care.

Interest in the potential applications of artificial intelligence in medicine, anesthesiology, and the world at large has never been higher. The Anesthesia Research Council steering committee formed an anesthesiologist artificial intelligence expert workgroup charged with evaluating the current state of artificial intelligence in anesthesiology, providing examples of future artificial intelligence applications and identifying barriers to artificial intelligence progress. The workgroup’s findings are summarized here, starting with a brief introduction to artificial intelligence for clinicians, followed by overviews of current and anticipated artificial intelligence–focused research and applications in anesthesiology. Anesthesiology’s progress in artificial intelligence is compared to that of other medical specialties, and barriers to artificial intelligence development and implementation in our specialty are discussed. The workgroup’s recommendations address stakeholders in policymaking, research, development, implementation, training, and use of artificial intelligence–based tools for perioperative care.