J. Williams,1 M. H. Hutmacher,2 M. Zierhut,1 J. Becker,1 B. Gumbiner,1 G. Spencer-Green,1 L. Melia,1 D. Yin,1 R. Li,1 X. Meng1; 1Pfizer, San Diego, CA, 2Ann Arbor Pharmacometrics Group, Ann Arbor, MI

BACKGROUND: The objective was to perform comparative assessment of clinical response data to detect potential differences between PF-05280586 and rituximab sourced from the US (ritux-US) and EU (ritux-EU). No meaningful difference would be concluded if the difference in area under effect curve (AUEC) between test and reference was within the reference range established using the data from ritux-US and ritux-EU observed in this study.
METHODS: This was a randomized, double-blinded PK similarity trial in active RA patients on methotrexate, refractory to TNF therapy. Patients received 1,000 mg PF-05280586 (n=71), ritux-EU (n=72) or ritux-US (n=71) by IV infusion on Day 1 and 15. Longitudinal nonlinear mixed effects models were fit to the PK, DAS28 and ACR response data up to 24 weeks. AUEC for each PD endpoint was simulated using model parameters and their uncertainty.
RESULTS: Observed baseline imbalances and group-to-group variation was accounted for by covariate and fixed effects in each model, which adequately recapitulated response data from this study. As an example, the figure shows the point estimate and uncertainty for the DAS28 assessment.
CONCLUSION: The mean differences between PF-05280586 and ritux-EU or ritux-US were within the boundary established by the ritux-EU and ritux-US data for ACR20/50/70 or DAS28.