A Longitudinal Tumor Growth Inhibition Model Based on Serum M-Protein Levels in Patients With Multiple Myeloma Treated by Dexamethasone
Fredrik Jonsson (1), Laurent Claret (1), Robert Knight (2), Marta Olesnyskyj (2), Christian Jacques (2), Vincent S. Rajkumar (3), Rene Bruno (1)
(1) Pharsight, a CertaraTM company, St. Louis, MI (2) Celgene corporation, Summit, NJ (3) Mayo Clinic Cancer Center, Rochester, MN
Objectives: The aim of this study was to develop and asses a tumor growth inhibition model based on changes in serum M-Protein, a tumor marker, in patients with multiple myeloma.
Methods: M-Protein measurements with time were pooled from 346 patients included in the dexamethasone arms of two pivotal phase 3 registration studies of lenalidomide plus dexamethasone vs. dexamethasone (MM009 (1), MM010 (2)). We developed a longitudinal exposure-response tumor growth inhibition model of drug effect on tumor growth dynamics (3) based on M-Protein level (taken as a marker of tumor size). The predictive performance of the model was evaluated using a posterior predictive check (PPC) based on 500 simulated replicates of the studies.
Results: The model is composed of sub-models for tumor growth dynamics (KL), drug effect (KD) and drug resistance (l). Patient-specific random effects are implemented on all of the model parameters. Drug effect is driven by drug dose over time in a virtual biophase compartment (KP). Data did not support estimation of KP and optimal value (KP: 20 week-1) was determined by log-likelihood profiling. No clinically relevant covariate effects were identified.
KL (wk -1)
KD (wk -1 per mg dexamethasone)
Lambda (wk -1)
Sigma1 (additive residual error, g/l)
Sigma2 (proportional residual error)
The model is qualified to simulate relative change from baseline of M-Protein level at the end of cycle 2 (week 8). Observed results (25th, 50th, and 75th percentiles) were consistent with the predictive distributions of the model.
Conclusions: This model enables the use of the change in M-Protein level as a continuous longitudinal biomarker for drug effect in multiple myeloma studies. This model will be part of a modeling framework to simulate expected survival of new investigational treatments and to support end-of-Phase 2 decisions and design of Phase 3 studies (4).
 Dimopoulos M, Spencer A, Attal M, Prince HM, Harousseau JL, Dmoszynska A, San Miguel J, Hellmann A, Facon T, Foà R, Corso A, Masliak Z, Olesnyckyj M, Yu Z, Patin J, Zeldis JB, Knight RD, Multiple Myeloma (010) Study Investigators. Lenalidomide plus dexamethasone for relapsed or refractory multiple myeloma. N Engl J Med, 357, 2123-2132, 2007.
 Weber DM, Chen C, Niesvizky R, Wang M, Belch A, Stadtmauer EA, Siegel D, Borrello I, Rajkumar SV, Chanan-Khan AA, Lonial S, Yu Z, Patin J, Olesnyckyj M, Zeldis JB, Knight RD, Multiple Myeloma (009) Study Investigators. Lenalidomide plus dexamethasone for relapsed multiple myeloma in North America. N Engl J Med, 357, 2133-2142, 2007.
 Claret L, Girard P, Hoff PM, Van Cutsem E, Zuideveld KP, Jorga K, Fagerberg J, Bruno R. Model-based prediction of Phase III overall survival in colorectal cancer based on Phase II tumor dynamics. J. Clin. Oncol., 27, 4103-4108, 2009.
 Bruno R., Claret L. On the use of change in tumor size to predict survival in clinical oncology studies: Toward a new paradigm to design and evaluate Phase II studies. Clin. Pharmacol. Ther., Invited Commentary, 86, 136-138, 2009.