The introduction of new therapies has led to improved survival of patients with multiple myeloma (MM), even those with relapsed and/or refractory (R/R) disease. The proteasome inhibitors (PIs) are of particular interest in this context because they act on the ubiquitin-proteasome system, which is responsible for regulation and degradation of the majority of intracellular proteins.[1,2] The constitutive proteasomes and immunoproteasomes each contain three catalytic sites, which are characterized according to their substrate specificity as the chymotrypsin-like (CT-L; β5 and low molecular weight polypeptide [LMP]-7), the caspase-like (C-L; β1 and LMP-2), and the trypsin-like (T-L; β2 and multicatalytic endopeptidase complex–like [MECL]-1) sites.[1,2] These are targeted to different degrees by different agents. Inhibition of the proteasome leads to disruption of the cell cycle, activation of apoptosis pathways, and ultimately cell death. In MM cells and other cells, PIs have also been shown to target the unfolded protein response, a signaling pathway allowing the appropriate folding of proteins. Based on the demonstrated efficacy of bortezomib (the first of these agents to receive regulatory approval), PIs have become a clinically validated therapeutic option. This article reviews current advances in proteasome inhibition in the management of MM.
First-Generation Proteasome Inhibitors
The first PI, bortezomib (Velcade), received approval in 2003 for the treatment of R/R MM and is currently the only PI approved for therapeutic use for any indication. Bortezomib is a synthetic, boronate-based dipeptide PI that induces apoptosis in tumor cells; the apoptosis is believed to result from inhibition of the degradation of regulatory and proapoptotic proteins. Bortezomib acts more selectively than natural peptide aldehyde PIs and forms a slowly reversible tetrahedral intermediate with the 20S proteasome (Figure). It is a potent, single-agent, intravenous therapy that targets predominantly the CT-L activities of both constitutive proteasomes and immunoproteasomes. Bortezomib acts in a dose-dependent manner, reaching 74% target inhibition at a dose of 1.38 mg/m2. It is usually administered on days 1, 4, 8, and 11 of a 3-week cycle because consecutive daily dosing has been associated with severe toxicity in animals. As a single agent, bortezomib has demonstrated excellent activity in a range of hematologic malignancies, including MM, Waldenstrm macroglobulinemia, follicular non-Hodgkin lymphoma, and mantle-cell lymphoma.[6,7] It has also shown promise when administered in combination with other agents for both R/R MM and other indications.[8,9]
The toxicity profile of bortezomib is well documented. It can be associated with peripheral neuropathy, which can limit long-term administration, as well as with myelosuppression, diarrhea, and fatigue.[10-12] The use of weekly doses of intravenous bortezomib or subcutaneous bortezomib has had an impact on reducing peripheral neuropathy associated with bortezomib, but these regimens are not always effective and long-term administration can sometimes be a challenge. These challenges have prompted the search for next-generation PIs—building on the activity of bortezomib while improving the safety profile, overcoming drug resistance, and developing more convenient oral agents.
Next-Generation Proteasome Inhibitors
The most advanced of the next-generation PIs is carfilzomib, which differs structurally and mechanistically from bortezomib.[2,15] Carfilzomib is a tetrapeptide bearing an epoxyketone moiety, and it irreversibly inhibits CT-L activity,[16-18] thereby eliciting more sustained inhibition than the reversible PI bortezomib. Carfilzomib also exhibits greater target specificity than bortezomib, which may explain the ability of carfilzomib to overcome resistance to bortezomib both in vitro and in vivo. Carfilzomib inhibits proliferation and induces apoptosis in bortezomib-resistant MM cell lines and in primary MM cells from patients with clinically established resistance to bortezomib and other conventional agents. In two parallel phase I studies, carfilzomib was well tolerated when administered using highly intensive daily dosage regimens (5 consecutive days of a 14-day schedule in PX-171-001, or 2 consecutive days for 3 consecutive weeks of a 4-week schedule in PX-171-002), which suggests that proteasome inhibition sustained for > 48 hours is tolerable.[19,20]
Carfilzomib has been evaluated in phase II studies in heavily pretreated patients with relapsed and refractory MM (PX-171-003) who have prognostically unfavorable cytogenetics and significant comorbidities, and in those with preexisting peripheral neuropathy.[21,22] Carfilzomib has also been evaluated in patients with relapsed MM who have been less heavily pretreated (a median of two prior treatment regimens) or not previously exposed to bortezomib (PX-171-004).
In study PX-171-003 (A1), intravenous carfilzomib was given on days 1, 2, 8, 9, 15, and 16 of every 28 days for up to 12 cycles, in doses of 20 mg/m2, escalating to 27 mg/m2 beginning with cycle 2. The overall response rate (ORR; partial response or better, using International Myeloma Working Group Uniform Response Criteria) was 24% and was largely the same for the subpopulation of patients with active peripheral neuropathy at the time of study entry and for the subpopulation of patients with unfavorable cytogenetic characteristics.[21,25] The median duration of response was 7.8 months.
In the PX-171-004 study of high-risk but bortezomib-naive cohorts of patients with relapsed MM, a higher ORR of 48% was reported. Likewise, there is evidence that responses are sustained during long-term treatment.[28,29] The most commonly reported adverse events were fatigue, nausea, anemia, and dyspnea; the most common grade ≥ 3 adverse events were anemia, lymphopenia, pneumonia, neutropenia, and thrombocytopenia.
The incidence of newly emergent peripheral neuropathy was shown to be low in clinical trials of carfilzomib despite high rates at baseline in heavily pretreated patients. Worsening neuropathy during treatment with carfilzomib is also uncommon, suggesting that carfilzomib may be an effective treatment for patients who have neuropathy due to prior treatment involving other agents.
An orally bioavailable analog of carfilzomib, ONX 0912(formerly PR-047),  is currently in phase I clinical trials in patients with advanced refractory or recurrent solid tumors. Similar to carfilzomib, ONX 0912 is an irreversible inhibitor of the β5 subunit of the proteasome. ONX 0912 was shown to induce apoptosis in MM cells resistant to bortezomib in vitro. In xenograft models, it significantly reduced tumor growth and prolonged survival, and it elicited a response comparable to that seen with carfilzomib. Anti-MM activity was shown to be enhanced when ONX 0912 was used in combination with bortezomib. In rodents and dogs, repeated dosing of ONX 0912 was well tolerated.
NPI-0052 (salinosporamide A, marizomib) is a natural β-lactone compound derived from the marine bacterium Salinospora tropica. Similar to carfilzomib, it is an irreversible inhibitor of the β5 subunit, but unlike carfilzomib, which preferentially inhibits CT-L activity, and bortezomib, which inhibits CT-L activity and C-L activity (albeit to a much lesser extent), NPI-0052 inhibits all three catalytic activities of the proteasome (CT-L, T-L, and C-L). Preclinical research suggested an improved therapeutic ratio and significant activity in hematologic and solid tumor models. This agent is now being tested in clinical trials for several indications, including MM. NPI-0052 may be administered either orally or intravenously.
In a phase I trial, patients with MM were treated weekly for 3 weeks in 4-week cycles at doses ranging from 0.025 mg/m2 to 0.7 mg/m2. At the maximum tolerated dose (MTD) of 0.7 mg/m2, inhibition of CT-L activity in whole blood was 73% and 99% at days 1 and 15, respectively. Eight patients with R/R MM remained in the study for 6 to 15 months (three patients were in the study for > 1 year) with stable disease and no significant toxicity. Two of these patients had disease that was refractory to bortezomib.[37,39]
Overall, toxicities of NPI-0052 are comparable to those of bortezomib, although they notably do not include peripheral neuropathy. Renal insufficiency was reported in one patient, and dose-limiting toxicities consisting of transient “hallucinations” and dizziness or unsteady gait were observed at doses of 0.9 mg/m2. NPI-0052 produces dose-dependent effects and exhibits a toxicity profile that compares favorably to that of bortezomib while providing greater levels of proteasome inhibition. NPI-0052 has also been found to inhibit brain proteasome activity in a mouse model, suggesting the potential for use of this PI in the treatment of central nervous system tumors.
CEP-18770 is an orally active peptide boronic acid PI that reversibly inhibits CT-L activity of the proteasome. Antitumor activity, survival benefits, and complete tumor regressions have been demonstrated in myeloma xenograft models. In addition, CEP-18770 exhibited a more favorable cytotoxicity profile toward normal human bone marrow progenitors and bone marrow–derived stromal cells compared with bortezomib. A phase I study in patients with solid tumors or non-Hodgkin lymphoma has recently been completed, and trials in R/R MM using intravenous administration of single-agent CEP-18770 and CEP-18770 in combination with lenalidomide (Revlimid) and dexamethasone are ongoing.
MLN9708 is a second-generation, reversible, peptide boronic acid analog of bortezomib that is in development. MLN9708 is orally bioavailable and is immediately hydrolyzed to its active form, MLN2238, when exposed to aqueous solutions.[43,44] It binds preferentially to the CT-L (β5) active site of the proteasome, albeit with a dissociation half-life that is approximately six times shorter than that of bortezomib. The pharmacodynamic response of MLN2238, including inhibition of the 20S proteasome, was superior to that of bortezomib in mice bearing human lymphoma (WSU-DLCL2) or human prostate (CWR22) tumor xenografts. These differences translated into improved antitumor activity, particularly in CWR22 xenografts where MLN2238 showed greater efficacy at 0.5 MTD compared with 0.5 MTD bortezomib. Oral administration of MLN9708 likewise resulted in antitumor activity in the CWR22 xenograft model. Improved efficacy of MLN2238 compared with bortezomib was also seen in two lymphoma models—one in tumor-bearing mice and the other in a murine model of systemic lymphoma. In the first preclinical in vivo study of PIs using a transgenic mouse model of human MM with spontaneous development of plasma cell neoplasms, MLN9708 demonstrated potent antitumor activity. Phase I studies with MLN9708, alone and in combination with other agents, are ongoing in patients with newly diagnosed and refractory MM, lymphomas, and multiple nonhematologic malignancies. Oral administration of MLN9708 is under investigation in patients with R/R MM.