E. Panoilia (1), E. Schindler (2), A. Brekkan (2). B. Bender (2). E. Samantas (3), G. Aravantinos (4), H. Kalofonos (5), C. Christodoulou (6), G.P. Patrinos (1), L.E. Friberg (2), G. Sivolapenko (1)
(1) Department of Pharmacy, School of Health Sciences, University of Patras, Patras, Greece; (2) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden; (3) 3rd Department of Medical Oncology, “Agii Anargiri” Cancer Hospital, Kalyftaki Nea Kifissia, Greece; (4) 2nd Department of Medical Oncology, “Agii Anargiri” Cancer Hospital, Kalyftaki Nea Kifissia, Greece; (5) Division of Oncology, Department of Medicine, University Hospital of Patras, Rion, Greece; (6) 2nd Department of Medical Oncology, “Metropolitan” Hospital, Athens, Greece
Objectives: To develop a population PK model for bevacizumab in patients with metastatic colorectal cancer (mCRC) stage IV, to evaluate the effect of VEGF polymorphisms on bevacizumab clearance (CL) and to characterize the in vivo interaction with its endogenous soluble ligand, VEGF165, so as to gain more insight into the underlying PKPD relationship.
Methods: PK and ligand data were collected from 19 adult patients with mCRC stage IV. Bevacizumab was given either at a dose of 5 and 10 mg/kg every 2 weeks, or 7.5 mg/kg every 3 weeks, by intravenous infusion together with one of the following combinations: 5-FU/Leucovorin/Irinotecan, 5-FU/Leucovorin/ Oxaliplatin, Capecitabine/Irinotecan. Bevacizumab peak and trough concentrations (86 samples) as well as pre- and post-dose VEGF165 levels (93 samples) were measured in serum by using sandwich ELISA [1,2]. VEGF polymorphisms (-2578C>A, -1154G>A, -634G>C) were identified by polymerase chain reaction (PCR) and DNA sequencing. The population PK model for bevacizumab was developed by using non-linear mixed-effects modeling implemented in NONMEM 7.3. The impact of demographic characteristics and genetics was explored on the underlying PK relationship by using a stepwise covariate model procedure (SCM). An antibody-ligand model to describe the VEGF165 profiles observed following bevacizumab administration is under development.
Results: A two-compartment model with first-order elimination best described bevacizumab concentration changes over time. The estimated CL was 0.17 L/day, the central volume of distribution (V1) was 3.1 L, the inter-compartmental clearance (Q) was 0.36 L/day and the peripheral volume (V2) was 2.6 L. An exponential distribution characterized the inter-individual variability in CL and V1 (23% and 15%) whereas the residual variability (24%) was explained by a proportional error model. Body weight was allometrically included in all PK parameters in the final model. None of the other available covariates were statistically significant. No direct relationship between VEGF polymorphisms and bevacizumab CL was identified.
Conclusions: The final population PK model adequately described the peak and trough concentrations of bevacizumab in patients with mCRC. PK parameters were consistent with those from studies on patients with solid tumors [3,4]. The antibody-ligand model is anticipated to further elucidate the role of VEGF polymorphisms in the PKPD relationship of bevacizumab.
References:
[1] Ternant D., Cézé N., Lecomte T., Degenne D., Duveau A., Watier H., Dorval E., Paintaud G. An enzyme-linked immunosorbent assay to study bevacizumab pharmacokinetics. Ther Drug Monit (2010) 32: 647-652.
[2] Quantikine® human VEGF immunoassay (2011). R&D Systems Inc, Minneapolis, USA.
[3] Gaudreault J., Greig G., Cosson V., Gupta M., Jumbe N., Hooker A.C. Population pharmacokinetics of bevacizumab. J Clin Oncol (2008) 26: (May 20 suppl; abstr 14570).
[4] Lu J., Bruno R., Eppler S., Novotny W., Lum B., Gaudreault J. Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol (2008) 62: 779-786.
Reference: PAGE 23 () Abstr 3129 [www.page-meeting.org/?abstract=3129]
Poster: Drug/Disease modeling - Oncology