Data reproducibility and single-center bias are concerns in preclinical research and compromise translation from animal to human. Multicenter preclinical randomized controlled trials (pRCT) may reduce the gap between experimental studies and RCT and improve the predictability of results, for example Jak1/2 inhibition in lupus nephritis. To evaluate this, we conducted the first pRCT in the kidney domain at two Spanish and two German academic sites. Eligible MRL/MpJ-Faslpr mice (female, age13-14 weeks, stress scores of less than two and no visible tumor or signs of infection) were equally randomized to either oral treatment with the Jak1/2 inhibitor baricitinib or vehicle for four weeks. Central blinded histology analysis was performed at an independent fifth site. The primary endpoint was the urinary protein/creatinine ratio. Baricitinib treatment did not significantly affect proteinuria, histological markers of activity and chronicity, or the glomerular filtration rate but significantly improved plasma autoantibody levels and lymphadenopathy. Data heterogeneity was noted across the different centers referring in part to phenotype differences between MRL/MpJ-Faslpr mice bred at different sites, mimicking well patient phenotype diversity in lupus trials. Multicenter pRCT can overcome single-center bias at the cost of increasing variability and reducing effect size. Thus, our pRCT predicts a low effect size of baricitinib treatment on human lupus nephritis in heterogeneous study populations.
Read full abstract