Abstract 4302: Analysis of the dose and schedule dependence of tumor kill in nonclinical tumour models after treatment with the WEE1 inhibitor AZD1775

• Published in: Cancer research (Chicago, Ill.) Ill.), 2018-07, Vol.78 (13_Supplement), p.4302-4302
• Main Authors: Yates, James William Thomas, Cadogan, Elaine, Hare, Jennifer I., Hughes, Adina M., Polanska, Urszula M., O'Connor, Mark J., Critchlow, Susan E.
• Format: Article
• Language: English
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2018-4302

AZD1775 is a highly selective, small-molecule inhibitor of WEE1 being developed to treat patients with advanced solid tumors, as monotherapy and in combination with olaparib (Lynparza). Previously a mathematical model was developed using data from patient-derived explant (PDX) and xenografted models with a range of sensitivities to AZD1775. This mathematical model could describe the dose and schedule dependency of pharmacokinetics, pCDK1 reduction in tumor and anti-tumor activity. This was for a dose range of 30mg/kg-120mg/kg dosed p.o. on a range of schedules from 3 days on 4 days off to 5 days on 9 days off. The model was then used to rank the potential effectiveness of each dosing regimen by calculating the fraction of tumor killed per week at doses resulting in drug exposure comparable to that observed in the clinic. This calculation was performed by integrating over time the rate of tumor kill predicted by the model. This analysis was complimented with a log cell kill (LCK) analysis using post treatment regrowth data to estimate in a more empirical manner the fraction of tumor killed over the treatment period. Specifically, if TC and TT are the times it takes the controls and treated tumors to grow to a prescribed volume and DT is the doubling time of control tumors then LCK=(TT-TC)/(2.3xDT). The analysis demonstrated that across the data set there was a consistent trend of increased LCK with dose level and number of days dosing in a week. The LCK values for each regimen were normalized by the total number of doses administered, to give an LCK per dose. There appeared to be a consistent LCK per dose level across the dose range considered. Interestingly, there was a greater than linear increase of LCK with increasing dose level. This was consistent with the observation that higher doses with shorter durations of dosing, were at least as active as more chronically administered lower doses. In the TNBC HBCx17 (Xentech) model, over a four-week period 60mg/kg dosed 5 days per week results in an LCK of 0.5 (70% killed) whereas 90mg/kg dosed 3 days per week has an LCK of 0.75 (83% killed). The same relationship was derived from the model simulated fraction tumor kill: higher doses generated significantly larger proportions of tumor kill, thus requiring shorter periods of dosing for the same net effect. The analysis drew greater differentiation between regimen than could be achieved by a tumor growth inhibition (TGI) analysis: where regressions were observed