Matthias Frei (1), Alexander Solms (1) and Bart Ploeger (1)
(1) Bayer AG, Berlin, Germany
Objectives:
Concomitant treatment with radium-223 and paclitaxel is a potential treatment for cancer patients with bone metastases; the agents have potentially synergistic mechanisms of action. Both agents are myelosupressive, and the increased risk for myelosuppression of the combination needs to be assessed. Objective of this analysis was to evaluate the potential interaction of concomitant paclitaxel and radium-223 treatment on the myelosuppression with a PK/PD modeling approach. Data were obtained with an individual cross-over design in a phase Ib study, enabling to separate the myelosuppressive effect of the 2 drugs, and allowing a small sample size.
Methods:
Neutrophil data were obtained from an open-label, multicenter, nonrandomized phase Ib study in cancer patients with bone metastases. Patients received up to 7 cycles of chemotherapy, with 90 mg/m2 paclitaxel IV per week in a 3-week-on / 1-week-off regimen, administered as per local standard of care. Starting with the 2nd cycle, subjects received up to 6 cycles 55 kBq/kg radium-223 IV, 1 injection every 4 weeks. Main observation period was the first 12 weeks with neutrophil observations obtained twice per week.
A previously developed semimechanistic PK/PD model[1] was used, describing the time course of paclitaxel-induced myelosuppression. This model had been adapted for monotherapy with radium-223.
The paclitaxel and radium-223 models were combined to evaluate the myelosuppression under concomitant treatment with radium-223 and paclitaxel in the current phase Ib study.
Models were implemented in NONMEM 7.3.
Results:
There were 12 patients included in the analysis, all of which completed treatment cycles 1 – 3. Following refinement of parameter values, the previously developed model adequately described the time course of paclitaxel-induced myelosuppression in cycle 1. In the combined radium-223 and paclitaxel model, the drug effect was implemented as Bliss Independence model[2]. The model adequately described the drug-induced myelosuppression in cycles with combined treatment. Model parameter estimates were consistent with previously reported results from paclitaxel studies[1]. The study design proved to be suitable for the separation of the myelosuppressive effect of the 2 drugs with a small number of subjects. Simulations of a typical patient, treated according to the analyzed study design, showed that after cycle 3 of treatment, the additional effect of radium-223 results in an additional decrease of the absolute neutrophil count at the nadir by ~10% of the baseline value when co-administered with paclitaxel, versus paclitaxel alone.
Conclusion:
A combined radium-223 and paclitaxel myelosuppression model – with the drug effect implemented as Bliss Independence model – described the neutrophil time course adequately well in cancer patients with bone lesions. The study design proved to be suitable for the separation of the myelosuppressive effect of the 2 drugs with a small number of subjects. Simulations of a typical patient, treated according to the analyzed study design, showed that after cycle 3 of treatment, the additional effect of radium-223 results in an additional decrease of the absolute neutrophil count at the nadir by ~10% of the baseline value when co-administered with paclitaxel, versus paclitaxel alone.
References:
[1] Friberg et al. J Clin Oncol. 2002; 20: 4713-4721
[2] Foucquir J & Guedj M. 2015; 3(3): 1-11
Reference: PAGE 27 (2018) Abstr 8535 [www.page-meeting.org/?abstract=8535]
Poster: Drug/Disease Modelling - Oncology