Marizomib Overcomes Compensatory Hyperactivation of Trypsin-like and Caspase-like Subunits to Provide Pan-Proteasome Subunit Inhibition in Patients with Multiple Myeloma and Solid Tumors

Background: Proteasome inhibitors (PIs) are a highly active drug class in multiple myeloma (MM), but development of resistance is commonly observed. Although all clinical-stage PIs are effective inhibitors of chymotrypsin-like (CT-L) proteasome subunit activity, a possible mechanism of resistance is compensatory hyperactivation of the caspase-like (C-L) and trypsin-like (T-L) subunits. Marizomib (MRZ) is a novel, irreversible, second-generation PI under development for the treatment of MM and malignant glioma. MRZ potently inhibits the 3 subunits of the 20S proteasome with a specificity and activity distinct from that of bortezomib and carfilzomib and their oral analogs ixazomib and oprozomib.

Methods: In the clinical study NPI-0052-102, we evaluated the pharmacodynamic (PD) activity of MRZ against all three proteasome subunits in whole blood after single, or repeated administration, via two schedules in patients (pts) with previously treated advanced malignancies (solid tumors or hematologic). Pts received MRZ via Schedule A - weekly dosing for 3 doses in 4-week cycles by IV infusion over 1 to 10 min, or Schedule B - twice-weekly dosing for 4 doses in 3-week cycles by IV infusion over 10 min to 2 hr. Blood samples were collected on Days 1 and 15 (Schedule A) or 1 and 11 (Schedule B), pelleted by centrifugation and frozen within 48 hrs. Pellets were lysed and the activity of all 3 proteasomal subunits was assayed using specific fluorogenic substrates.

Results: Partial or complete inhibition of all three proteasome subunits was achieved with both once-weekly and twice-weekly MRZ dosing. The rank order of sensitivity, CT-L > T-L > C-L, was as expected from the biochemical and cellular potencies of the drug. For CT-L activity, proteasome inhibition was dose dependent, with both the initial effect (on Cycle 1, Day 1 - C1D1) and the peak effect of proteasome inhibition increasing in a dose-dependent manner. CT-L inhibition was modest at the lowest dose levels (≤ 0.15 mg/m²), reaching moderate levels after repeat dosing. At intermediate dose levels of 0.3 to 0.55 mg/m², CT-L inhibition was 31% to 75% on C1D1, rising to 70% to 93% after repeat dosing. At doses of 0.7 mg/m² and above, inhibition of CT-L activity was usually complete within the first cycle. By contrast, C-L and T-L activities were unchanged or increased in the first cycle, suggesting compensatory hyperactivation in response to effective blockade of CT-L activity. Importantly, however, this response was overcome by repeat dosing with MRZ, and inhibition of T-L and C-L activity was noted across all dose levels. For C-L activity, treatment with MRZ at the recommended Phase 2 dose (RP2D) of 0.5 mg/m² with twice-weekly dosing resulted in inhibition of up to 39% by Cycle 2 and was maintained, when tested, through Cycles 4 and 6. Treatment at the RP2D of 0.7 mg/m² for once-weekly dosing resulted in C-L inhibition of 14% to 37% by the end of the first cycle, rising to 31% to 50% by the end of Cycle 2. Blockade of T-L activity was more robust after multiple cycles of MRZ therapy. Although inhibition of the T-L subunit was absent on C1D1 in patients receiving 0.5 to 0.55 mg/m², inhibition of up to 80% was achieved by Cycle 2 and maintained for the duration of treatment. At the once-weekly RP2D of 0.7 mg/m², T-L inhibition of 29% to 56% was achieved by the end of the first cycle, rising to 64% to 78% by the end of Cycle 2.

Conclusions: The PD activity of MRZ against all three proteolytic subunits was assessed in patients with MM, solid tumors, and advanced lymphomas. At the twice-weekly and once-weekly RP2Ds, complete inhibition of CT-L activity was observed within 1-2 cycles of therapy, but accompanied by compensatory hyperactivation of the C-L and T-L subunits. This phenomenon in red cells suggests that it may be due to allosteric interactions within the catalytic core of the 20S proteasome rather than de novo synthesis of additional proteasomes. Ongoing MRZ therapy was able to overcome this adaptive response, resulting in robust pan-subunit proteasome inhibition within 2-6 cycles, most probably due to the cumulative effect of multiple exposures to MRZ. Due to their reversible binding mode (bortezomib, ixazomib) or monospecificity for the CT-L site (carfilzomib, oprozomib), other clinical PIs are predicted to lack this capability. This unique property of MRZ may explain in part its clinical activity in patients with MM resistant to both bortezomib and carfilzomib.