COMPARATIVE SAFETY OF COMMON IMMUNOSUPPRESSANTS IN SLE: A CLINICAL TRIAL EMULATION ON INFECTION RISK WITH ADJUSTMENT FOR GLUCOCORTICOID DOSE
Bibliographic record
Abstract
O053 / #469 Topic: AS24 - SLE-Treatment ABSTRACT CONCURRENT SESSION 09: SLE THERAPY – REVISITING OLD DRUGS AND UNLOCKING HIDDEN POTENTIAL OF NEW MEDICATIONS 24-05-2025 10:40 AM - 11:40 AM Background/Purpose Serious infections are among the most common causes of death in SLE. We performed a clinical trial emulation to assess the comparative safety of 4 commonly used immunosuppressants for treating SLE. Methods Using the OptumLabs Data Warehouse, we identified adults ≥18 years with SLE, defined as ≥3 ICD9/10 codes separated by ≥30 but ≤365 days, who initiated treatment with methotrexate, azathioprine, belimumab, or mycophenolate mofetil (MMF) between March 1, 2011, and September 30, 2023. We required a ≥6-month washout period during which patients could not have been treated with any of the drugs being compared. We excluded those with a prior history of lupus nephritis or organ transplant. The primary outcome was hospitalizations for infections. We emulated randomization using inverse probability weighting and stabilized weights for the 4 study groups, balancing baseline covariates, including glucocorticoid use and dose, hydroxychloroquine use, and disease severity (per the Garris algorithm; details in Table 1). Patients were followed for up to 3 years, death, or disenrollment. We applied 2 censoring approaches: intention-to-treat (ITT) and per-protocol (PP). In the ITT analysis, patients were followed based on their initial medication; in the PP analysis, patients were censored if they stopped or switched drugs. We first conducted analyses using inverse probability weighting with only baseline covariates, then further adjusted for post-baseline prednisone doses using a Marginal Structural Model. We repeated the analysis using traumatic injuries as a falsification outcome. Table 1. Baseline Characteristics of SLE Patients on Immunosuppressants After Weighting Demographics, comorbidities, SLE severity, and medication use for patients initiating azathioprine, belimumab, methotrexate, or MMF, with standardized mean differences (SMD) for group balance. Results After weighting, 738 patients initiated belimumab, 2,462 methotrexate, 1,113 MMF, and 1,459 azathioprine. The mean age was 48 years, and more than 90% were women. Most patients were on background therapy with hydroxychloroquine (66.0% to 70.8%) and glucocorticoids (72.6% to 74.3%), and around 50% had moderate SLE severity based on the Garris algorithm. Baseline covariates were balanced across groups after weighting (Table 1). Hospitalization rates for infections were highest with MMF, followed by azathioprine, methotrexate, and belimumab. In the baseline-only ITT analysis MMF showed significantly higher infection risk compared to belimumab (HR: 1.55, 95% CI 1.07-2.25, p = 0.02) and methotrexate (HR: 1.32, 95% CI 1.04-1.68, p = 0.02). The PP analysis confirmed higher infection risk with MMF compared to belimumab (HR: 2.27, 95% CI 1.11-4.62, p = 0.02). After adjusting for post-baseline prednisone doses, no significant differences were observed among the 4 drugs (Table 2). There were no differences in the cumulative incidence of traumatic injuries across all drugs, supporting internal validity. Table 2. Risk or Infection-Related Hospitalizations by Immunosuppressant in Non-Renal SLE Hazard ratios (HR) and 95% confidence interval (CI) for infection-relatad hospitalizations, with intention to treat (ITT) and per protocol (PP) analyses adjusted for baseline covariates and post-baseline prednisone doses. Significant p-values are bolded. Conclusions In this clinical trial emulation of comparative safety, adjusting for prednisone dose eliminated differences in infection-related hospitalization risk among the 4 studied immunosuppressants. These findings suggest that glucocorticoid dose, rather than the choice of immunosuppressant, may be the primary driver of serious infections in SLE
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.101 | 0.091 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.004 | 0.011 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.002 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.004 | 0.003 |
| Insufficient payload (model declined to judge) | 0.008 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".