Therapeutic Landscape of Carfilzomib and Other Modulators of the Ubiquitin-Proteasome Pathway

For both normal and malignant cells, the degradation of intracellular proteins must be carefully controlled to adjust the levels of important regulatory proteins and rapidly eliminate damaged or misfolded proteins before their toxic aggregates compromise the cell function or survival.^1,2 The ubiquitin-proteasome pathway degrades most intracellular proteins. This complex system (Fig 1) identifies proteins intended for degradation and attaches to them chains of ubiquitin (Ub) molecules³ through a sequential system of Ub-activating enzymes, Ub-conjugating enzymes, and Ub ligases.^4,5 Ubiquitinated proteins are then recognized by the 26S proteasome⁷ and selectively digested by its distinct (chymotrypsin-like, trypsin-like, and caspase-like) proteolytic activities.^8–10

Fig 1.

Overview of the ubiquitin-proteasome pathway and its therapeutic implications. Most intracellular proteins are degraded by the 26S proteasome, a large complex (> 60 known subunits) that selectively digests proteins with covalently attached ubiquitin (Ub) chains.³ The 26S proteasome comprises two 19S regulatory complexes flanking a hollow cylindrical core particle (termed 20S proteasome). Proteins are marked for degradation by a complex enzymatic system upstream of the 26S proteasome: one of two Ub-activating enzymes (E1s) uses the energy from adenosine triphosphate (ATP) hydrolysis (to adenosine diphosphate [ADP] and inorganic phosphate) to transfer Ub to one of 40 Ub-conjugating enzymes (E2s), which interact with one of approximately 600 Ub ligases (E3s).^4,5 The latter covalently attach Ub chains to specific lysine (Lys) residue(s) of different sets of protein substrates.^4,6 The 19S proteasome complexes recognize (through their Rpn13 or S5a/Rpn10 subunits,⁷ not depicted in this figure) these ubiquitinated substrates, disassemble the Ub chains (which are then recycled), unfold the target proteins, and translocate them to the 20S proteasome chamber.^8–10 The 20S proteasome chamber (shown in cross-section in the right-hand panel) comprises three types of proteolytic subunits, β5, β2, and β1: each subunit cleaves proteins preferentially after large hydrophobic, basic, or acidic residues (chymotrypsin-like, trypsin-like, and caspase-like activities, respectively). Most tissues express this canonical constitutive 20S proteasome. Cells of the immune system also express (particularly when exposed to certain proinflammatory cytokines) the immunoproteasome, a variant form with different catalytic subunits (β1i, β2i, and β5i) and often associated with 11S regulatory complexes, to optimize presentation of antigenic peptides through major histocompatibility complex class I molecules. The proteasome inhibitors bortezomib and carfilzomib both bind to and inhibit the chymotrypsin-like activity of the β5 subunit.

This complex degradative network and its substrate proteins influence diverse aspects of cancer biology, thus creating opportunities for therapeutic interventions. The first agent targeting this cascade was bortezomib (formerly known as PS-341), an inhibitor of the chymotrypsin-like activity of the proteasome. Bortezomib has pronounced clinical activity in multiple myeloma (MM)^11–13 and other plasma cell dyscrasias (amyloidosis^14–16 and Waldenström macroglobulinemia^17,18), is also active in mantle-cell lymphoma,^19,20 but has limited, if any, activity in most other hematologic malignancies or solid tumors. Its complex molecular sequelae include suppression of antiapoptotic molecules, such as nuclear factor kappa B, Bcl-2 family members, and caspase inhibitors,^21–24 and sensitization of MM cells to diverse established^22,23 or investigational²⁴ agents. Bortezomib thus emerged as a key component of diverse anti-MM combination regimens.^25,26 Eventually, patients become resistant to bortezomib or intolerant to its main dose-limiting toxicity, namely sensory peripheral neuropathy.²⁷ To overcome these limitations, second-generation proteasome inhibitors were developed. One of them, carfilzomib, received accelerated US Food and Drug Administration approval in 2012 for treatment of patients with MM who had relapsed from and were refractory to bortezomib and at least one thalidomide derivative.²⁸ In the accompanying article, Papadopoulos et al²⁹ report results of a phase I trial of carfilzomib infusion over 30 minutes. Compared with prior studies with shorter infusion time (2 to 10 minutes),³⁰ this trial delivered higher doses (maximum-tolerated dose of 56 mg/m²) and had a higher overall response rate (50% in patients with relapsed and refractory MM, including those who were resistant to bortezomib). The safety profile for carfilzomib included thrombocytopenia (similar to bortezomib), likely because constitutive proteasome activity in platelets is required to degrade Bax and preserve their normal life span.³¹ In contrast to historical experience with bortezomib, but consistent with prior carfilzomib studies,²⁸ peripheral neuropathy was not observed, but cardiopulmonary adverse effects (eg, dyspnea, hypoxemia, pulmonary hypertension) and serum creatinine elevations were noted.

Bortezomib and carfilzomib can be administered without catastrophic clinical toxicities, likely because their clinically achievable concentrations do not completely abrogate the chymotrypsin-like activity,^11,32,33 and also spare other proteolytic (trypsin-like and caspase-like)^34,35 activities of the proteasome. Overall protein degradation is thus only modestly (< 40%) suppressed in either normal or tumor cells. Normal cells can conceivably tolerate this perturbation, but malignant plasma cells may not be able to, because they depend on higher levels of proteasome activity for a process termed endoplasmic reticulum (ER) –associated degradation^8,34,36: misfolded or unassembled proteins in the ER lumen must undergo retrograde transport to the cytoplasm to be degraded by the proteasome and prevent ER stress and apoptosis. In plasma cell dyscrasias, the proteasome capacity (availability of active proteasome particles) is apparently close to being saturated by the increased proteasome load (ie, the amount of misfolded or unassembled proteins such as immunoglobulins). Indeed, these plasma cells produce large quantities of immunoglobulins, but their assembly has an appreciable error rate (hence, the free immunoglobulin light chains detected in sera of patients with plasma cell dyscrasias). This high proteasome load for a given proteasome capacity may explain in part why proteasome inhibitors are more active in plasma cell dyscrasias, compared with most other hematologic malignancies or solid tumors, whereas differences in this relationship of proteasome load versus capacity among patients with MM have been proposed to account for the heterogeneity of clinical responses to bortezomib.^33,37,38

Carfilzomib inhibits the β5 proteasome subunit by forming with it an irreversible adduct through two covalent bonds,³⁹ conceivably allowing more sustained and more specific inhibition than the single reversible adduct formed by bortezomib. For instance, bortezomib, but not carfilzomib, is proposed to inhibit not only the β5 but also the neuroprotective molecule Htra2/Omi^40,41 and other serine proteases (eg, cathepsin G, cathepsin A),^42–44 some of which are proposed to contribute to renal injury.⁴⁵ These differences could explain the more frequent peripheral sensory neuropathy observed with bortezomib and the increase in serum creatinine levels often observed with carfilzomib.^29,46 Proteasome dysfunction and accumulation of cardiotoxic misfolded proteins^47–53 have been linked with different forms of cardiac dysfunction (eg, cardiomyopathies): the cardiopulmonary adverse events that occur with carfilzomib thus merit mechanistic dissection to identify possible predisposing factors and determine whether irreversible proteasome inhibition could be one of them.

The irreversible β5 inhibition in carfilzomib-treated cells means that proteasome capacity cannot be restored before new proteasomes are synthesized.^28,54,55 Such delayed recovery is proposed to account for observations that carfilzomib can be active in some cases in which bortezomib is not (eg, 18.6% response rate with carfilzomib in a phase II study³⁰ of patients with MM who had progressed from a bortezomib-containing last line of therapy). This observation is concordant with preclinical data that some MM cells can be resistant to one of these two proteasome inhibitors, but sensitive to the other (eg, bortezomib-resistant, carfilzomib-sensitive MM cells).^56,57 Interestingly, another second-generation proteasome inhibitor MLN2238 and its clinically administered prodrug ixazomib (MLN9708) are also active preclinically in MM cells⁵⁸ and clinically in patients who were bortezomib-resistant,⁵⁹ although MLN2238 binds the β5 subunit reversibly and with faster kinetics of release than bortezomib itself.⁶⁰ Therefore, the relationship of clinical activity with reversible versus irreversible inhibition of chymotrypsin-like activity is likely to be more complex and influenced by other pharmacodynamic and pharmacokinetic parameters. Still, the response rates reported by Papadopoulos et al²⁹ can be considered promising, compared with historical data from single-agent trials of bortezomib,^13,61 lenalidomide,⁶² and pomalidomide^63,64 in relapsed and refractory MM. Ultimately, results from randomized trials of carfizomib-containing versus bortezomib-containing regimens (eg, ENDEAVOR [NCT01568866]; Phase 3 Study With Carfilzomib and Dexamethasone Versus Velcade and Dexamethasone for Relapsed Multiple Myeloma Patients [ENDEAVOR]) should shed light on the differential safety and efficacy profiles of these agents. It will also be important to determine the depth and durability of responses to bortezomib in patients with carfilzomib-resistant, bortezomib-naive MM.

Recent studies of second-generation proteasome inhibitors, including the report by Papadopoulos et al²⁹ and other carfilzomib trials, highlight the promising clinical activity of these agents and the potential to improve their efficacy and hopefully their safety through modified infusion rates (eg, in the study by Papadopoulos et al²⁹) and more broadly, through optimized dosing, schedules, and combinations with other established or investigational anti-MM agents (eg, Berenson et al,⁶⁵ Niesvizky et al,⁶⁶ Moreau et al,⁶⁷ and Wang et al,⁶⁸ and Stewart et al^68a). Recent reports indicate that thalidomide derivatives endow the E3 ligase Cereblon with the ability to ubiquitinate IKZF1 and IKZF3, two important transcription factors for MM cells.^69,70 Active research is also exploring the selective therapeutic targeting of other enzymes that regulate the ubiquitination state of substrate proteins, including other E3 ubiquitin ligases (eg, Ooi et al⁷¹) or deubiquitinases (eg, Wang et al⁷² and Tian et al⁷³), and more specific inhibitors of immunoproteasome^74,75 are being evaluated in lymphoid neoplasias. These studies and the ongoing progress in the clinical development of second-generation proteasome inhibitors have further validated the overall concept that the regulation of protein degradation provides promising targets for therapeutic interventions in MM and beyond.

AUTHOR'S DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Disclosures provided by the authors are available with this article at www.jco.org.

AUTHOR'S DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Therapeutic Landscape of Carfilzomib and Other Modulators of the Ubiquitin-Proteasome Pathway

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Constantine S. Mitsiades

Honoraria: Millenium, Celgene

Research Funding: Johnson and Johnson, Amgen

Patents, Royalties, Other Intellectual Property: Previously submitted patent application on methods for treating cancer using proteasome inhibitors