Interrogating two schedules of the AKT inhibitor MK-2206 in patients with advanced solid tumors incorporating novel pharmacodynamic and functional imaging biomarkers

Abstract

Purpose: Multiple cancers harbor genetic aberrations that impact AKT signaling. MK-2206 is a potent pan-AKT inhibitor with a maximum tolerated dose (MTD) previously established at 60 mg on alternate days (QOD). Due to a long half-life (60-80 hours), a weekly (QW) MK-2206 schedule was pursued to compare intermittent QW and continuous QOD dosing.

Experimental design: Patients with advanced cancers were enrolled in a QW dose-escalation phase I study to investigate the safety and pharmacokinetic-pharmacodynamic profiles of tumor and platelet-rich plasma (PRP). The QOD MTD of MK-2206 was also assessed in patients with ovarian and castration-resistant prostate cancers and patients with advanced cancers undergoing multiparametric functional magnetic resonance imaging (MRI) studies, including dynamic contrast-enhanced MRI, diffusion-weighted imaging, magnetic resonance spectroscopy, and intrinsic susceptibility-weighted MRI.

Results: A total of 71 patients were enrolled; 38 patients had 60 mg MK-2206 QOD, whereas 33 received MK-2206 at 90, 135, 150, 200, 250, and 300 mg QW. The QW MK-2206 MTD was established at 200 mg following dose-limiting rash at 250 and 300 mg. QW dosing appeared to be similarly tolerated to QOD, with toxicities including rash, gastrointestinal symptoms, fatigue, and hyperglycemia. Significant AKT pathway blockade was observed with both continuous QOD and intermittent QW dosing of MK-2206 in serially obtained tumor and PRP specimens. The functional imaging studies demonstrated that complex multiparametric MRI protocols may be effectively implemented in a phase I trial.

Conclusions: Treatment with MK-2206 safely results in significant AKT pathway blockade in QOD and QW schedules. The intermittent dose of 200 mg QW is currently used in phase II MK-2206 monotherapy and combination studies (NCT00670488).

Figure 1. Pharmacokinetic profile of MK-2206 in QW schedule

A: Mean plasma MK-2206 concentrations (nM) following the first dose of Cycle 1. B: MK-2206 plasma AUC_0-168hr (nM*hr) following the first dose of Cycle 1. C: MK-2206 plasma C_max (nM) following the first dose of Cycle 1.

Figure 2A. Pharmacodynamic profile of MK-2206 in tumor and platelet-rich plasma (PRP) in a QW schedule

The analysis of paired tumor biopsies from 5 patients receiving 200mg QW of MK-2206 (Figure 2A1) showed a significant suppression of pSer473 AKT, when compared with paired pre-treatment samples. Overall, the pSer473 AKT signal decreased to 50.0% (range 37.5-60.3%; p=0.0003) of baseline levels on MK-2206 treatment, confirming target modulation. The PD effects of MK-2206 treatment on pSer473 AKT (n=11) and pSer9 GSK3β (n=11) phosphorylation at the 200mg MTD were also assessed sequentially in PRP specimens at multiple timepoints (Figures 2A2 – 2A3). The mean pSer473 AKT signal post-MK-2206 was 19.8% at 24h (p<0.0001), 30.6% at 48h (p<0.0001), 51.7% at 96h (p=0.0015) and 97.4% at 168h (p=0.92) of baseline levels (Figure 2A2), while mean pSer9 GSK3β signal was 65.0% at 24h (p<0.0001), 82.3% at 48h (p=0.012), 91.4% at 96h (p=0.56) and 117.2% at 168h (p=0.37) of baseline levels (Figure 2A3). * p<0.05; **p<0.01; ***p<0.001 Paired t test compared to baseline. Points represent the levels of PD biomarkers as a percent of the baseline levels for individual patients and orange lines represent mean of all patients at that time point.

Figure 2B. Pharmacodynamic profile of MK-2206 in platelet-rich plasma (PRP) and tumor in a QOD schedule

Paired tumor biopsies were also obtained for biomarker analysis by MSD^® electrochemiluminescence assays from 3 patients receiving 60mg QOD of MK-2206. All 3 patients showed a significant suppression of pSer473 AKT, when compared with paired pre-treatment samples (Figure 2B1). The mean pSer473 AKT levels decreased to 50.1% (range 45.3-56.2%; p=0.004) of baseline levels on MK-2206 treatment, confirming target modulation. PD effects in PRP were also assessed serially at multiple timepoints from 29 patients receiving the MTD of 60mg QOD of MK-2206 (Figures 2B1-2B4). The pSer473 AKT signal decreased significantly to mean of 36.3% at 3h (p<0.0001), 27.9% at 6h (p<0.0001), 44.0% at 24h (p<0.0001) and 70.6% at 48h (p=0.045) of baseline levels (Figure 2B2). Inhibition of pSer473 AKT was associated with significant decreases in the phosphorylation of downstream substrates pSer9 GSK3β to mean levels of 76.8% at 3h (p=0.009), 67.5% at 6h (p=0.0051), 102.9% at 24h (p=0.87) and 9116.1% at 48h (p=0.51) of baseline levels (Figure 2B3); and pThr246 PRAS40 to mean levels of 79.8% at 3h (p=0.005), 79.7% at 6h (p=0.0059), 86.9% at 24h (p=0.030) and 91.4% at 48h (p=0.42) of baseline levels (Figure 2B4). * p<0.05; **p<0.01; ***p<0.001 Paired t-test compared to baseline. Points represent the levels of PD biomarkers as a percent of the baseline levels for individual patients and orange lines represent mean of all patients at that timepoint.

Figure 3. Best responder

A 43-year-old female with ER/PR positive, HER2 negative metastatic breast cancer treated with 150mg QW of MK-2206. PD studies demonstrated maximal suppression of pSer473 AKT and its substrate pSer9 GSK3β at 24h, followed by a gradual recovery of phosphorylation signals from 48h in PRP. Points represent the levels of PD biomarkers as a percent of the baseline levels (A). Molecular characterization of circulating nucleic acids demonstrated a PIK3CA exon 20 mutation and low Ki67 proliferation. The patient had previously received treatments including cyclophosphamide, doxorubicin and radiotherapy. Following 8 weeks of treatment with MK-2206 150mg QW, there was a 22% reduction in RECIST measurements of her target lesions in the liver metastases, celiac axis and para-aortic lymph nodes (B-C). Post treatment MRI demonstrated intratumoral necrosis in liver and bone lesions, as evidence by the high signal on anatomical T2 weighted images (D), and the high ADC values on DWI-MRI (overall lymph node mass ADC was 170 × 10⁻⁵ mm²/s; the mean ADC in the areas of necrosis was greater than 200 × 10⁻⁵ mm²/s) (E). Consistent with these findings, CA15-3 levels decreased by 36% (6853U/ml to 4399U/ml) and the patient remained on treatment for a total of 24 weeks before developing disease progression.