In Phase II trials, an olaparib capsule dose that inhibits PARP (100 mg bd) was clinically inferior to the MTD (400 mg bd). Using an in vivo model, we investigated this discrepancy and the effect of dose and schedule on tumour progression.
PK, PD and antitumour activity were assessed using olaparib (100, 50, 25, 10, 2.5 mg/kg qd) in a breast cancer patient-derived xenograft (PDX) with mutated BRCA2 and TP53. We studied the relationship between PARP1 inhibition and DNA single-strand breaks (SSBs) by creating a mathematical model. We also compared PK/PD model parameters with free minimum olaparib steady-state plasma concentration (Cmin,ss) in seven Phase I–III clinical studies.
Only 100 and 50 mg/kg olaparib doses caused tumour regression, yet all but the lowest led to acute PAR inhibition. The differentiating factor was time over PAR IC95; simulations using the mathematical model at steady state predicted a threshold where >95% reduction in PAR caused >20-fold increase in DNA SSBs. Overlaying clinical data, we saw that 400 mg bd capsule and 300 mg bd tablet doses provide continuous PAR inhibition (Cmin,ss>IC95 upper confidence level); 100 mg bd capsule does not. Dosing schedule assessment showed PDX regression with continuous 100 mg/kg olaparib (0/10 tumours progressed by day 100). By contrast, continuous dosing at 50 mg/kg or intermittent dosing (1 wk on/1 wk off) at 100 mg/kg resulted in 2/10 and 10/10 progressions, respectively. Treatment withdrawal upon regression led to regrowth in all cases; re-challenge at continuous 100 mg/kg was effective (9/10 regressions), 50 mg/kg was not (0/10). In a maintenance setting, switching to lower or intermittent dosing resulted in tumour response more similar to that of continuous 100 mg/kg (1/10, 2/10 and 0/10 progressions, respectively).
Discrepancy between biologically and clinically effective olaparib doses is explained by a need for continuous >95% PAR inhibition. Both the 400 mg bd capsule and 300 mg bd tablet dose, used in the ongoing clinical programme, provide this. Dosing schedule data suggest continuous PAR inhibition is also needed for long-term responses. These data highlight important principles for treatment of patients with PARP inhibitor monotherapy.
Clinical trial identification
Legal entity responsible for the study
M. O\'Connor, E. Cadogan, A. Hughes, M. Learoyd, H. Xu, J. Li, J. Yates: Employee of AstraZeneca and owns stock with AstraZeneca. E. Leo: Employee of AstraZeneca.