Myasoedova, Elena

Objectives Rheumatoid factor (RF) and anticitrullinated protein antibodies (ACPA) are linked to disease activity, severity and cardiovascular risk in rheumatoid arthritis (RA). Since RA-related inflammation associates with cardiovascular disease, we investigated whether RF and ACPA status influence the relationship between RA activity and event risk. Methods We assessed 3952 cardiovascular disease-free patients with prevalent RA registered in an international observational consortium between 1985 and 2012. Main outcome was a first major adverse cardiovascular event (MACE) including myocardial infarction, stroke and cardiovascular death. Follow-up accrued from enrolment until first MACE or censoring. Multivariable Cox models stratified by centre risk evaluated disease activity, RF, ACPA and their interactions. Results We recorded 184 first MACE events over 22 981 patient-years. Disease activity associated with MACE (HR 1.18 (95% CI 1.03 to 1.35)), whereas RF (HR 1.26 (0.82 to 1.94)) and ACPA (HR 1.25 (0.85 to 1.83)) did not reach statistical significance, though CIs were wide and a clinically meaningful association cannot be excluded. A three-way interaction between disease activity, RF and ACPA on MACE (p for interaction=0.044) was observed. RA activity associated with MACE in seronegative and double seropositive (p≤0.047), but not single RF-positive or ACPA-positive patients. Among RF-negative participants, RA activity was linked to MACE in ACPA-negative but not positive ones (p for interaction=0.003). Among ACPA-positive patients, disease activity was linked to MACE in RF-positive but not RF-negative ones (p for interaction=0.040). Conclusions Disease activity may better discriminate inflammation-driven cardiovascular risk in seronegative and double seropositive than single RF-positive or ACPA-positive patients. Risk protection through lowering disease activity may be serostatus dependent.

Abstract Scientific foundation models learn high-dimensional representations from diverse data modalities, yet what they encode and how to extract that knowledge remain open questions. Here we show that probing the internal representations of Evo 2, a 7-billion-parameter genomic foundation model, enables accurate and interpretable genetic variant effect prediction. We introduce a covariance-based probe that captures second-order structure from Evo 2 sequence embeddings to predict variant pathogenicity across variant types and functional consequences, matching or exceeding specialized predictors within their domains. To ground these predictions in known biological mechanisms, we train a complementary panel of probes on existing annotations to detect which genomic properties are disrupted by a variant. This categorized evidence is then integrated with each variant’s genomic context through a language model to generate variant-specific mechanistic hypotheses. Our pathogenicity predictions correlate with experimental measures of variant function, clinical penetrance, and biobank disease associations while the mechanistic hypotheses are consistent with expert reviews, known mechanism classes, and downstream molecular readouts. We release pathogenicity scores, disruption profiles, and contextualized interpretations for 4.2 million variants from the ClinVar database as an open resource through the Evo Variant Effect Explorer (EVEE). More broadly, this structured probing approach offers a general framework for interrogating foundation models across scientific disciplines and grounding their outputs in existing domain concepts.

Objectives Female sex and anticitrullinated protein antibodies (ACPA) are associated with higher disease activity in rheumatoid arthritis (RA). Since disease-related inflammation is linked to cardiovascular risk, we explored whether sex and ACPA influenced the association between disease activity at study entry and cardiovascular risk in established RA. Methods We evaluated 4008 patients with prevalent RA from an international observational cohort enrolled between 1985 and 2012. Outcomes included major adverse cardiovascular events (MACE: cardiovascular death, myocardial infarction and stroke) and ischaemic cardiovascular events (iCVE: MACE, angina, revascularisation, transient ischaemic attack and peripheral arterial disease). Follow-up accrued from enrolment until the first event or censoring. Multivariable Cox models stratified by centre risk evaluated disease activity, sex, ACPA and their interactions. Results We documented 193 MACE and 299 iCVE. Disease activity and sex were associated with MACE (all p≤0.017) and iCVE (p≤0.005) but ACPA was only associated with MACE (p=0.043). A three-way interaction on MACE (p=0.034) but not iCVE was noted. Among ACPA-negative patients, disease activity was associated with MACE in males (HR 1.57 (95% CI 1.14 to 2.16)) but not females (p-for-interaction=0.022). Among ACPA-positive patients, neither the disease activity x sex interaction (p=0.929), nor main effect of disease activity on MACE (p=0.124) was significant, but male sex was (HR 1.61 (95% CI 1.15 to 2.27)). Among females, neither disease activity x ACPA interaction (p=0.523) nor disease activity (p=0.319) was significant for MACE, but ACPA was (HR 1.57 (95% CI 1.02 to 2.42)). Conclusions The effect of disease activity at enrolment on cardiovascular risk in prevalent RA varies across patient groups with different sex and ACPA characteristics.

Objective The objective of this study was to investigate the association between dual seropositive, single seropositive, and seronegative rheumatoid arthritis (RA) with radiographic erosions, disease flares, and death. Methods We performed a retrospective, population‐based study of residents in Southern Minnesota with incident RA who fulfilled criteria for RA in 2003 to 2019. Radiographic erosions and flares were evaluated within one year of incident RA. All‐cause mortality was obtained from medical records and death certificates. Cox models adjusted for age, sex, smoking status, year of incident RA, and comorbidities were used. Results The study included 1,373 patients with RA. At RA incidence, 37% were dual seropositive, 13% seropositive for anti–cyclic citrullinated protein (anti‐CCP) only, 12% seropositive for rheumatoid factor (RF) only, and 38% seronegative. The highest proportion of radiographic erosions before or within one year of RA incidence was in the dual seropositive (31%) and lowest in those seropositive for anti‐CCP only (13%). Flares occurred in 69% of the dual seropositive and 51% of the seropositive for anti‐CCP only within the first year of RA incidence. Those seropositive for RF only had over a two‐fold increase in mortality rates compared with the seronegatives (adjusted hazard ratio [aHR] 2.18, 95% confidence interval [CI] 1.47–3.24). In contrast, the dual seropositives had only a >50% increase in mortality rates (aHR 1.66, 95% CI 1.21–2.28), and those seropositive only for anti‐CCP had a >30% increase in mortality rates (aHR 1.32, 95% CI 0.83–2.11). Conclusion Patients with RA seropositive for RF only have an increase in mortality rates compared with patients who are dual seronegative. RF positivity could be an indicator of inflammation that requires pharmacologic treatment to decrease mortality rates.

Objective The purpose of this study was to determine the cumulative incidence of diagnosis switching, drug‐free remission, and initiation of a biologic or targeted synthetic disease‐modifying antirheumatic drug (b/tsDMARD) in individuals with seronegative rheumatoid arthritis (RA). Methods Adult residents of Olmsted County, Minnesota, with incident seronegative (rheumatoid factor negative/anti–cyclic citrullinated peptide antibody negative) RA meeting the 1987 and/or 2010 American College of Rheumatology classification criteria were included. Data were collected from January 1, 2005, to December 31, 2023, through manual chart review. Drug‐free remission was defined as a period of at least six months where the individual was no longer taking treatment for RA and did not have evidence of active inflammatory arthritis upon evaluation by a rheumatologist. We calculated the 10‐year cumulative incidence of a change in diagnosis, adjusting for the competing risk of death, and of drug‐free remission and initiation of a b/tsDMARD, adjusting for the competing risks of death or change in diagnosis. Results A total of 176 individuals with seronegative RA (68% women) were included. The 10‐year cumulative incidence of a change in diagnosis was 12.8% (95% confidence interval [CI] 8.7%–18.9%). The most common change in diagnosis was to spondyloarthritis (4%). The 10‐year cumulative incidence of drug‐free remission and initiation of a b/tsDMARD was 26.6% (95% CI 20.7%–34.2%) and 19.9% (95% CI 14.7%–26.9%), respectively. Over a median follow‐up of 11.8 years, 49 individuals entered drug‐free remission. Conclusion After initial diagnosis of seronegative RA, about 13% of individuals had a change in diagnosis and a quarter experienced drug‐free remission within 10 years.

Abstract The advent of high-throughput sequencing has led to an exponential increase in genomic data, highlighting the need for efficient and accurate models to analyze and interpret this information. In this study, we introduce a novel, exomic foundational model that leverages a combination of the human reference genome and multispecies data to improve variant detection and interpretation. Our model utilizes a short-range transformer architecture and is trained on a large dataset of human exomic sequences derived from the Tapestry study. Through a series of ablation studies and scaling experiments, we demonstrate the effectiveness of our model in predicting next token accuracy and identifying clinically pathogenic variants. We also show that our model outperforms existing models in a range of downstream tasks, including variant effect prediction and disease state identification. In fact, our largest sequence transformer for range adaptive nucleotide decoding variant (1B parameters) surpassed previous benchmarks, demonstrating a mean accuracy of 0.880 [an 8.2% improvement over the original nucleotide transformer (NT) and a 7% improvement over NT-v2]. Furthermore, we construct a unique exomic ClinVar dataset to evaluate the model’s performance on pathogenicity and disease states. Our results highlight the potential of this model to improve our understanding of the human exome and its role in disease. The model and its applications have significant implications for genomics-based diagnosis and personalized medicine, including tailored therapeutic development.

Objective Cognitive impairment is prevalent in rheumatoid arthritis (RA), yet risk factors are not well understood. We explored associations between clinical characteristics and cognitive impairment in an RA cohort. Methods Data were from a longitudinal RA cohort at Veterans Affairs Puget Sound Health Care System. Cognition was evaluated using the Saint Louis University Mental Status (SLUMS) examination. Demographics and health factors, objectively measured physical function, participant‐reported symptoms, RA disease characteristics, and comorbidities were evaluated. Univariable linear regressions explored the association between clinical factors and SLUMS scores. Those with P < 0.1 in the univariable models were evaluated in separate multivariable linear regressions controlling for age, sex, and years of education. Results A total of 145 participants with RA were included, with a mean age of 64.5 (SD 11.6) years, and were predominantly male (74%). Using SLUMS, 42 participants (29%) had normal cognition, 83 (57%) had mild cognitive impairment, and 20 (14%) had dementia. Physical performance (aβ: 0.35, confidence interval [CI] 0.11 to 0.59), self‐reported exhaustion (aβ: −2.22, CI −3.44 to −0.99), pain (aβ: −0.25, CI −0.50 to −0.00), disability (aβ: −1.79, CI −3.16 to −0.42), and trouble falling asleep (aβ: −2.57, CI −4.00 to −1.14) were all independently associated with a lower SLUMS score (all P < 0.05). Conclusion Cognitive impairment was prevalent in our cohort of veterans with RA and was associated with several modifiable clinical factors. Future longitudinal studies are needed to determine the directionality of these associations and evaluate interventions for modifiable risk factors that may mitigate cognitive dysfunction. image

Objectives To examine associations between multimorbidity, the social determinants of health (SDoHs) and rheumatoid arthritis (RA) flare and remission. Methods All patients aged ≥18 years in Olmsted County, Minnesota, with incident RA in 1999–2014 were identified. Using a list of 55 chronic medical conditions, multimorbidity was defined as the presence of ≥2 conditions and substantial multimorbidity as ≥5 conditions. The Area Deprivation Index and Social Vulnerability Index (SVI) were used as proxies for adverse SDoH burden. Flare and remission were defined using Outcome Measures in Rheumatoid Arthritis Clinical Trials definitions. Mixed effects models were used to assess associations between flare/remission and multimorbidity, adverse SDoH burden and other patient characteristics. Results This study included 659 patients with incident RA. Multimorbidity and substantial multimorbidity predicted 29% (OR:1.29, 95% CI:1.04 to 1.59) and 26% (OR:1.26, 95% CI:1.03 to 1.53) higher odds of flare, respectively. Both were associated with 34% (OR:0.66, 95% CI:0.49 to 0.90) and 33% (OR:0.67, 95% CI:0.51 to 0.90) lower odds of remission, respectively. SVI predicted 8% lower odds of remission for every 0.1 increase above 0.3 (OR:0.92, 95% CI:0.85 to 0.99). Flare was also associated with female sex, smoking, younger age and shorter disease duration, but not seropositivity. Remission was also associated with male sex, never smoking, older age and longer disease duration, but not seropositivity. Conclusions Multimorbidity predicts higher odds of RA flare and lower odds of remission. Adverse SDoH burden predicts lower odds of remission. These findings have the potential to inform disease prognostication and clinician interventions.

Abstract Nearly half of premature deaths can be prevented with healthy lifestyle behaviours. However, most older adults do not adhere to established guidelines to promote a healthy lifestyle. We propose a tailored framework of lifestyle medicine that acknowledges the complex contexts and care needs of older adults. The proposed framework of lifestyle medicine for older adults integrates their lived experiences (e.g. social determinants of health, built environment), the 5 M framework of geriatric medicine (i.e. Mind, Mobility, Medications, Multicomplexity, Matters Most) and the six pillars of lifestyle medicine (i.e. nutrition/diet, physical activity/exercise, restorative sleep, stress management, avoidance of risky substances, positive social connections) through a novel transdisciplinary approach. Guided by the updated framework, key gaps and opportunities are presented to advance lifestyle medicine practice, research and policy with the goal of promoting sustained behaviour change and improving the healthspan of older adults.