Titilola Falasinnu

Background: Pain drives disability and medication use in autoimmune rheumatic diseases (ARDs), even when inflammation is controlled. The extent of pain‐related polypharmacy across ARDs and chronic overlapping pain conditions (COPCs) remains poorly characterized. We aimed to quantify first‐year pain medication burden following ARD diagnosis and examine variations by disease, COPC status, and time. Methods: Using the Merative MarketScan Commercial Claims database, we identified adults newly diagnosed with rheumatoid arthritis, systemic lupus erythematosus, ankylosing spondylitis, psoriatic arthritis, Sjögren's disease, or systemic sclerosis (2008–2021). We analyzed pain‐related medications across 10 pharmacologic categories during the first year postdiagnosis. Polypharmacy measures included counts of distinct pain medications, medication categories, ≥ 5 and ≥ 10 medication thresholds, and the Medication Quantification Scale. Findings: Among 149 742 adults with newly diagnosed ARDs, 57.6% had at least one COPC. During the first year after diagnosis, patients filled a mean of 9.0 distinct pain medications and 47.9% met criteria for ≥ 5 medications; 30.5% met criteria for ≥ 10 medications. Ankylosing spondylitis demonstrated the highest medication burden (mean 11.6 medications; 57.5% with ≥ 5 medications; 37.9% with ≥ 10 medications). Patients with fibromyalgia had consistently higher medication burden across ARDs and were the only COPC group in whom prescribing increased from 2015 onward. Overall medication burden rose from 2008 to 2014–2015 and declined thereafter; however, absolute levels remained high. Interpretation: Pain‐related polypharmacy is common early after ARD diagnosis and varies substantially by disease and pain phenotype. Despite modest declines from 2015 onward, medication burden remains high, underscoring the need for phenotype‐informed, nonpharmacologic, and deprescribing strategies in rheumatology.

Background: Nociplasticity is common in systemic lupus erythematosus (SLE), yet its longitudinal trajectory remain poorly characterized. Methods: Patients with SLE in the FORWARD Databank were classified as Minimal, Type 1, Type 2, or Mixed using the Polysymptomatic Distress Scale (PSD≥8) and the Systemic Lupus Activity Questionnaire (SLAQ) inflammatory domain score (≥2). Cross-sectional analyses (N=372) compared pain, function, disability, depression, organ damage, and medication use. Longitudinal analyses (n=301; median 3.7 years) characterized phenotype transitions using continuous-time Markov models and identified latent trajectory subgroups using joint group-based trajectory modeling (GBTM). Findings: At baseline, 29% were Minimal, 12% Type 1, 13% Type 2, and 47% Mixed. Functional impairment increased stepwise: SF-36 PCS worsened from 49.7 in Minimal to 30.0 in Mixed; PROMIS Pain Interference from 46.2 to 63.6 (both p<0.001). BILD organ damage was highest in Mixed (3.9 vs 2.6–2.9; p<0.001). Depression rose from 3% to 43% and opioid use reached 44% in Mixed. Longitudinally, Minimal and Mixed were persistent (mean duration 2.0 and 1.8 years; one-year retention 70%), while Type 1 and Type 2 were transient (∼0.5 years; retention 18% and 28%). Exit trajectories were asymmetric: Type 1 moved preferentially to Minimal (49% of exits), whereas Type 2 moved to Mixed (65%; p<0.001). Population-average PSD was nearly flat (+0.014 SD/year, p=0.07); opioid use declined to near zero in Minimal and Type 1 but remained high in Type 2 and Mixed. Joint GBTM identified four severity classes along a Minimal-to-Mixed diagonal, indicating that the two axes scale together longitudinally and that pure Type 1 and Type 2 are transient. Interpretation: Nociplastic phenotypes in SLE are persistent, severity-stratified, and associated with functional impairment, depression, cumulative damage, and sustained opioid burden. Type 1 and Type 2 are transient states with divergent exits: the former resolving, the latter progressing. Funding: Lupus Research Alliance Research in context: Evidence before this study: The Type 1 and Type 2 systemic lupus erythematosus (SLE) framework distinguishes inflammatory manifestations from symptoms associated with nociplastic pain and has provided a clinically meaningful approach to understanding the heterogeneity of SLE. Patient-reported instruments, including the polysymptomatic distress (PSD) scale combined with inflammatory symptom assessment, have enabled classification of these phenotypes in both research and clinical settings. A search of PubMed, Embase, and Web of Science (from database inception to June 01, 2026) using the terms “systemic lupus erythematosus”, “Type 1 SLE”, “Type 2 SLE”, “nociplastic pain”, “nociplasticity”, “polysymptomatic distress”, “trajectory”, and “transition”, together with screening of reference lists of relevant articles, identified studies describing cross-sectional phenotype distributions, intermittent and persistent Type 2 symptoms, and molecular differences between phenotypes. However, whether Type 1 and Type 2 phenotypes represent stable disease states or transient manifestations of a patient’s underlying longitudinal symptom trajectory is unknown. Added value of this study: This study analyzed prospectively collected patient-reported data from 301 SLE patients enrolled in the nationwide FORWARD registry over a median follow-up of 3.7 years to characterize the longitudinal behavior of Type 1 and Type 2 phenotypes. By combining continuous-time Markov modelling with group-based trajectory modelling, we distinguished threshold-defined phenotype transitions from underlying symptom trajectories. We found that Minimal and Mixed phenotypes were highly stable over time, whereas Type 1 and Type 2 phenotypes represented transient states with distinct transition patterns. Data-driven trajectories formed a stable severity continuum from Minimal to Mixed rather than persistent Type 1 and Type 2 classes. Higher nociplastic burden was also associated with sustained impairment in physical function, depression, medication burden, and opioid use. Implication of all the available evidence: These findings suggest that nociplastic symptom burden represents a persistent and clinically important component of SLE that is not fully captured by inflammatory disease activity alone. Routine assessment of nociplastic symptoms alongside inflammatory activity could improve longitudinal patient stratification and identify individuals who may benefit from combined immunomodulatory and symptom-directed management. Future studies should determine whether phenotype-informed treatment strategies improve long-term patient outcomes.

Objective: This study aims to characterize the burden and distinct patterns of chronic overlapping pain conditions (COPCs) multimorbidity in adults with autoimmune rheumatic diseases (ARDs). Methods: We analyzed 2008 to 2021 data from the Merative MarketScan Commercial Claims and Encounters Database, identifying 149,742 patients with ARDs (ankylosing spondylitis, psoriatic arthritis, rheumatoid arthritis, Sjögren disease, systemic lupus erythematosus, and systemic sclerosis) using International Classification of Diseases, Ninth/Tenth Revision, Clinical Modification codes. We assessed COPC (eg, fibromyalgia, chronic low back pain, migraine, irritable bowel syndrome, chronic fatigue syndrome, temporomandibular disorders, and pelvic pain conditions) prevalence, multimorbidity patterns, and pairwise co‐occurrence frequencies, calculating observed/expected ratios and odds ratios (ORs). Results: Of 149,742 patients with ARDs, 57.5% had one or more COPCs, with 22.7% experiencing multiple conditions. Chronic low back pain (41.4%) and fibromyalgia (21.1%) were most prevalent. Women and individuals aged 31 to 50 years had the highest COPC prevalence. Many COPC pairs co‐occurred more frequently than expected by chance (eg, fibromyalgia and chronic low back pain: observed/expected ratio 1.35; migraine and chronic low back pain: observed/expected ratio 1.42). Strongly associated dyads included urologic chronic pelvic pain and vulvodynia (OR 10.46). Conclusion: A substantial burden and distinct multimorbidity patterns of COPCs exist among patients with ARDs, suggesting shared pathophysiologic mechanisms. Routine COPC screening and targeted nonopioid multimodal interventions, especially for common dyads, are crucial for improved pain management and reducing opioid‐related risks in this population.

Background: Polypharmacy is common in autoimmune rheumatic diseases (ARDs) and increases adverse drug events (ADEs), but comparative evidence across diseases is limited. We aimed to quantify ADE burden and identify medications associated with ADE risk across six ARDs, and to examine shared and disease-specific patterns across diseases. Methods: We conducted a retrospective cohort study at a tertiary medical center (2010–2024). Adults with ankylosing spondylitis (AS), psoriatic arthritis (PsA), rheumatoid arthritis (RA), Sjögren’s disease (SjD), systemic lupus erythematosus (SLE), or systemic sclerosis (SSc) were identified using diagnostic codes. ADEs were ascertained using validated case definitions. Medications were mapped to Anatomical Therapeutic Chemical classes; active exposure was defined within 30 days before the index date. Polypharmacy was defined as ≥5 concurrent medications (minor 5–10; major >10). Within each ARD, nested case–control analyses matched on encounter type (1:4) were performed, and adjusted odds ratios (aORs) were estimated using conditional logistic regression. Findings: Among 10,578 patients, 3,154 (29.8%) experienced at least one ADE. ADE burden varied across diseases, with the highest prevalence observed in SSc (35.9%). Polypharmacy was common (57.3% minor, 39.4% major) and medication burden was consistently higher in ADE cases across encounter types (eg, SLE outpatient median 12 vs 6; inpatient 20 vs 10; emergency 17 vs 8). Across ARDs, the strongest associations with ADEs were observed for supportive and symptom-directed therapies (acid suppressors, pain adjuncts, and sedative-hypnotic/psychotropic medications), whereas conventional disease-modifying antirheumatic drugs (DMARDs) showed weaker associations. Disease-specific signatures included gastrointestinal agents in SSc (metoclopramide aOR 12.32), antibiotics and respiratory agents in AS (ciprofloxacin aOR 13.71, fluticasone aOR 8.88). Interpretation: ADEs affect nearly one-third of ARD patients and increase with medication burden. Risk concentrates in supportive and symptom-directed therapies rather than DMARDs, with both shared and disease-specific patterns. Optimizing prescribing, particularly for pain management and corticosteroid use, can reduce medication-related harm.

Background: Systemic lupus erythematosus (SLE) presents substantial life challenges, reflecting the complex and heterogeneous nature of the disease. Pain is common and often one of the earliest manifestations. Skill-based behavioral pain treatments are often inaccessible to patients with SLE, and no SLE-specific evidence exists to guide patient and clinician treatment choices. To fill a pressing evidence gap for SLE pain treatment, the Lupus PROGRESS (Pain Relief With Online Groups That Empower Skill-Based Symptom Reduction) trial will compare a lower-burden evidence-based behavioral pain treatment with online multisession cognitive behavioral therapy (CBT; the gold standard behavioral treatment for chronic pain) while also enhancing diversity and representativeness among patient advisors and study participants. Objective: Lupus PROGRESS is a national, pragmatic, randomized noninferiority trial comparing the effectiveness of 2 evidence-based behavioral pain treatments (online 8-session CBT vs 1 session the Empowered Relief program; Stanford University) in adults with SLE and chronic pain in the United States for reducing pain intensity and pain interference at 3 months after treatment (multiple primary end points). Methods: In total, 150 adults with SLE and pain (≥3 months, intensity of ≥3/10) will be enrolled from academic medicine, community health network, and insurance payer study sites. Participants are randomized 1:1 to either 1-session Empowered Relief or 8-session CBT, with both treatments delivered online to groups. Mixed-model repeated measures and intention-to-treat analytic approaches will be used to test comparative effectiveness aims from baseline to 3 months after treatment (the primary end point). Multiple primary outcomes are pain intensity and pain interference; priority secondary outcomes are sleep disturbance, pain catastrophizing, pain bothersomeness, and anxiety; and other secondary outcomes are satisfaction with roles and responsibilities and global impression of change, anger, fatigue, and depression. Durability of effects will be tested at 6 months after treatment. Empowered Relief is hypothesized to have superior treatment completion and adherence and lower treatment burden ratings than CBT. In exploratory analyses, we will compare health care use through medical record and claims data 3 months prior to enrollment to the last 3 study months, hypothesizing reduced noninferiority in reduced use. The study also incorporates a multilevel patient engagement infrastructure to improve accessibility, recruitment, and study relevance. Results: As of May 2026, we have enrolled 130 participants. Enrollment is anticipated to end in November 2026, with treatments completed by April 2027 and full analyses completed by February 2028. Conclusions: Lupus PROGRESS will provide real-world pragmatic evidence on whether a 1-session online group skill-based treatment is noninferior to 8-session CBT. Results from this patient-centered study will inform more accessible, equitable, and patient-centered chronic pain care for people living with lupus. Trial Registration: ClinicalTrials.gov NCT05612750; https://clinicaltrials.gov/study/NCT05612750 International Registered Report Identifier (IRRID): DERR1-10.2196/100104

BACKGROUND Systemic lupus erythematosus (SLE) presents substantial life challenges, reflecting the complex and heterogeneous nature of the disease. Pain is common and often one of the earliest manifestations of SLE. Skills-based behavioral pain treatments are often inaccessible to patients with SLE, and no SLE-specific evidence exists to guide patient and clinician treatment choices. To fill a pressing evidence gap for SLE pain treatment, “Lupus PROGRESS” is comparing a lower-burden evidence-based behavioral pain treatment with online multi-session cognitive behavioral therapy (CBT; the gold-standard behavioral treatment for chronic pain), while also enhancing diversity and representativeness among patient advisors and the study participants. OBJECTIVE Lupus PROGRESS is a national, pragmatic, randomized, non-inferiority trial comparing the effectiveness of two evidence-based behavioral pain treatments, online 8-session cognitive behavioral therapy (CBT; the gold-standard behavioral treatment for chronic pain) vs. 1-session Empowered Relief in U.S. adults with SLE and chronic pain, for reducing pain intensity and pain interference at 3 months post-treatment (multi-primary endpoints). METHODS One hundred fifty adults with SLE and pain (≥3 months, intensity ≥3/10) will be enrolled from academic medicine, community health network, and insurance payer study sites. Participants are randomized 1:1 to either 1-session Empowered Relief (ER) or 8-session CBT, with both treatments delivered online to groups. Mixed model repeated measures and intention to treat analytic approaches will be used to test comparative effectiveness aims from baseline to 3 months post-treatment (primary endpoint). Multi-primary outcomes are change in pain intensity and pain interference; priority secondary outcomes are sleep disturbance, pain catastrophizing, pain bothersomeness and anxiety; and other secondary outcomes are satisfaction with roles and responsibilities, global impression of change, anger, fatigue and depression. Durability of effects will be tested at 6 months post-treatment. ER is hypothesized to have superior treatment completion/adherence and lower treatment burden ratings than CBT. In exploratory analyses we will compare healthcare utilization using medical record and claims data 3 months prior to enrollment to last 3 study months, hypothesizing reduced non-inferiority in reduced utilization. The study also incorporates a multi-level patient engagement infrastructure to improve accessibility, recruitment, and study relevance. RESULTS As of May 2026, we have enrolled 130 participants. Enrollment is anticipated to end September 2026, with treatments completed by February 2027, and full analyses completed by November 2027. CONCLUSIONS Lupus PROGRESS will provide real-world pragmatic evidence on whether a 1-session online group skills-based treatment is non-inferior to 8-session CBT Results from this patient-centered study will inform more accessible, equitable, and patient-centered chronic pain care for people living with lupus. CLINICALTRIAL ClinicalTrials.gov: NCT 05612750 (https://clinicaltrials.gov/study/NCT05612750)

Key messages: What is already known on this topic Cannabis use is increasingly documented in EHR narrative text, but structured fields do not capture use status or symptom related motivations, limiting research on pain self-management strategies in autoimmune rheumatic diseases. What this study adds We developed a natural language processing pipeline using large language models to identify cannabis use status (4-class) and reasons for use (6-class), with best performance from fine-tuned GatorTron for status and an LLM for reasons. How this study might affect research, practice or policy This scalable approach can support real-world evidence studies on symptom management, medication use, and outcomes among ARD populations, and it provides a methodological template for extracting under-documented patient behaviors from narrative notes. Objective: The primary objective is to develop and evaluate a scalable and reproducible natural language processing (NLP) approach using large language models (LLM) to identify cannabis use status and reasons for cannabis use among patients with autoimmune rheumatic diseases (ARDs) from unstructured electronic health record (EHR) clinical notes. Methods and Analysis: We conducted a retrospective study using EHR clinical notes from patients with ARDs (2015–2024). Notes were screened for cannabis-related mentions using fuzzy string matching against a curated keyword lexicon with a similarity threshold of 90, extracting 50-word context windows (±25 words). Two domain experts annotated 886 randomly sampled snippets across four classes: (1) not a true cannabis mention/uncertain, (2) denial of use, (3) positive past use, and (4) positive current use. Using these annotations, we compared multiple LLM prompting strategies (zero-shot to few-shot; temperature tuning) and a fine-tuned clinical model (GatorTron 345M). For “reason for use,” 1,027 snippets were annotated into six categories: pain, nausea, sleep, anxiety/stress/mood, appetite, and not mentioned/unknown. Models were evaluated on a held-out validation set using accuracy, F1, recall, and precision. We then aggregated snippet-level predictions to patient level to describe temporal trends and subgroup differences. Results: For cannabis use status classification, the fine-tuned GatorTron model achieved the highest performance (accuracy 0.90; F1 0.91; recall 0.90; precision 0.90). For the reason of cannabis use classification, GPT-OSS-20B achieved the highest performance (accuracy 0.90; F1 0.90; recall 0.90; precision 0.92). Patient-level analyses characterized trends in documented cannabis use from 2015–2024 and compared clinical characteristics between current users and patients denying use. Conclusion: High-precision extraction of cannabis use status and reasons for use from EHR notes is feasible using a combination of fine-tuned clinical language models and LLM-based classifiers. This approach enables scalable measurement of patient-reported symptom self-management strategies in ARDs, supporting observational research and potential clinical decision support.