Clinical activity of carfilzomib correlates with inhibition of multiple proteasome subunits: application of a novel pharmacodynamic assay

Abstract

While proteasome inhibition is a validated therapeutic approach for multiple myeloma (MM), inhibition of individual constitutive proteasome (c20S) and immunoproteasome (i20S) subunits has not been fully explored owing to a lack of effective tools. We utilized the novel proteasome constitutive/immunoproteasome subunit enzyme-linked immunosorbent (ProCISE) assay to quantify proteasome subunit occupancy in samples from five phase I/II and II trials before and after treatment with the proteasome inhibitor carfilzomib. Following the first carfilzomib dose (15-56 mg/m(2) ), dose-dependent inhibition of c20S and i20S chymotrypsin-like active sites was observed [whole blood: ≥67%; peripheral blood mononuclear cells (PBMCs): ≥75%]. A similar inhibition profile was observed in bone marrow-derived CD138(+) tumour cells. Carfilzomib-induced proteasome inhibition was durable, with minimal recovery in PBMCs after 24 h but near-complete recovery between cycles. Importantly, the ProCISE assay can be used to quantify occupancy of individual c20S and i20S subunits. We observed a relationship between MM patient response (n = 29), carfilzomib dose and occupancy of multiple i20S subunits, where greater occupancy was associated with an increased likelihood of achieving a clinical response at higher doses. ProCISE represents a new tool for measuring proteasome inhibitor activity in clinical trials and relating drug action to patient outcomes.

Keywords: molecular analysis; multiple myeloma; myeloma therapy; pharmacology; trials.

© 2016 The Authors. British Journal of Haematology published by John Wiley & Sons Ltd.

Figures

Figure 1

Carfilzomib administration results in potent inhibition of multiple proteasome subunits. (A) ProCISE assay schematic. PABP is added to purified 20S proteasomes (activated with 0∙03% SDS‐AB) or cell lysates and incubated for 2 h. Denaturation with guanidine and capture on beads is then done simultaneously for 1 h followed by five washes in PBS + 1% BSA + 0∙05% Tween. Primary antibodies for specific proteasome active site subunits are then added overnight, followed by another five washes. A secondary HRP antibody is then added for 2 h, followed by five washes. A chemiluminescent substrate is then added for 5–10 min, and luminescence is determined. (B) Proteasome subunit levels were measured by ProCISE in 62 whole blood and 54 PBMC samples from MM patients and 27 whole blood and 24 PBMC samples from ST patients. CT‐L activity was measured in the same samples using the LLVY‐AMC substrate. Data are presented as box and whisker plots with subunit concentration (ProCISE) or specific activity (LLVY‐AMC). Mean specific activity [AMC/protein (μM/μg)] and subunit concentrations [subunit/protein (ng/μg)] are represented by a horizontal line, and error bars represent minimum and maximum values. (C) Whole blood and PBMC samples taken 1 h after the first administration of carfilzomib in MM and ST patients (doses: 15–45 mg/m2) were analysed for proteasome subunit activity by LLVY‐AMC and occupancy by ProCISE. Values were normalized to predose values, and data are presented as mean (±SEM) relative activity or occupancy. The dotted line represents the LLOQ for whole blood. (D) MM patients were treated with carfilzomib 15 or 20 mg/m2, dexamethasone 40 mg and lenalidomide 10, 15, or 25 mg on day 1. Whole blood and PBMC were isolated predose and 1 h after the first dose of carfilzomib dose was administered. Samples were analysed for proteasome activity and subunit occupancy, as described in panel A. Values are normalized to predose values, and data are presented as mean (±SEM) specific activity or occupancy (N = 2–8 per cohort). LLVY‐AMC data are reproduced from Badros et al (2013) with permission from Nature Publishing Group. (E) MM patients with varying degrees of renal function were treated with 15 mg/m2 of carfilzomib on day 1. Whole blood and PBMC were isolated predose 1 h after carfilzomib dose administration. Samples were analysed for proteasome activity and subunit occupancy as described in panel B. Values are normalized to predose values, and data are presented as mean (±SEM) relative activity or occupancy (N = 2–7 per cohort). 1°, primary; 2°, secondary; Ab, antibody; AMC, aminomethylcoumarin; BSA, bovine serum albumin; CFZ, carfilzomib; CT‐L, chymotrypsin‐like; D, day; HRP, horseradish peroxidase; LLOQ, lower limit of quantitation; LLVY, Leu‐Leu‐Val‐Tyr‐7‐amino‐4‐methylcoumarin; LMP, low‐molecular mass polypeptide; MECL, multicatalytic endopeptidase complex‐like; MM, multiple myeloma; PABP, proteasome active site‐binding probe; PBMC, peripheral blood mononuclear cell; PBS, phosphate buffer solution; SDS‐AB, sodium dodecyl sulfate in assay buffer; SEM, standard error of the mean; ST, solid tumour.

Figure 2

Repeat administration of carfilzomib results in sustained proteasome inhibition. (A) Patients received carfilzomib at 15, 20, 36 or 45 mg/m2 on days 1, 2, 8, D9, 15 and 16 of a 28‐day cycle. Whole blood and PBMC samples were taken prior to and 1 h after administration on days 1, 2 and 8 of cycle 1 and day 1 of cycle 2. Samples were analysed for proteasome activity and subunit occupancy as described in Fig 1 and were normalized to cycle 1, day 1 predose values. Data are presented as mean (±SEM) relative specific activity or occupancy. (B) Bone marrow samples from MM patients were taken during baseline screening and/or on cycle 1, day 2, 1 to 4 h after the second dose of carfilzomib. CD138+ cells and samples were analysed for proteasome activity and subunit occupancy. Data are presented as mean (±SEM) relative activity or subunit concentration. *P < 0∙001; **P < 0∙01; ***P < 0∙05 by a t‐test. C, cycle; CT‐L, chymotrypsin‐like; D, day; AMC, aminomethylcoumarin; LMP, low‐molecular mass polypeptide; MECL, multicatalytic endopeptidase complex‐like; MM, multiple myeloma; PBMC, peripheral blood mononuclear cell; SEM, standard error of the mean.

Figure 3

Immunoproteasome subunit inhibition is correlated with activity in multiple myeloma. (A) Proteasome inhibition in PBMC samples was measured as described in Fig 1. Patients were stratified according to carfilzomib dose: either 20/56 mg/m2 from PX‐171‐007 or 15 and 15/20 mg/m2 from PX‐171‐005. Values (relative to cycle 1, day 1 predose) are presented as mean (±SEM) relative subunit concentration or specific activity. The shaded area represents the time points in which the dose was stepped up to 56 mg/m2 for PX‐171‐007 patients or 20 mg/m2 for some PX‐171‐005 patients. *P < 0∙05; **P < 0∙01; ***P < 0∙001 by a t‐test. (B) Data from panel A at cycle 1, day 8 1‐h time point. #P < 0∙05; ##P < 0∙001; ###P < 0∙01; by a t‐test. C, cycle; CT‐L, chymotrypsin‐like; D, day; LMP, low‐molecular mass polypeptide; MECL, multicatalytic endopeptidase complex‐like; PBMC, peripheral blood mononuclear cell; SEM, standard error of the mean.