A Phase Ib/II Study of Oprozomib in Patients with Advanced Multiple Myeloma and Waldenström Macroglobulinemia

Proteasome inhibitors (PI), including carfilzomib, bortezomib, and ixazomib, have become validated strategies for treating patients with multiple myeloma. Bortezomib has demonstrated efficacy in patients with multiple myeloma, but is associated with peripheral neuropathy and must be administered intravenously or subcutaneously, limiting its use in some patients (1–5). Given the challenges of intravenous administration, subcutaneous treatment with bortezomib is now the preferred administration route (6). Carfilzomib, an epoxyketone PI, is approved in the United States and other countries for the treatment of relapsed or refractory multiple myeloma (7). In the ENDEAVOR trial, carfilzomib demonstrated superiority over bortezomib in patients with relapsed or refractory multiple myeloma (8). One challenge associated with carfilzomib treatment is the need for twice-weekly intravenous administration (7, 9). Once-weekly carfilzomib has recently been approved for the treatment of relapsed or refractory multiple myeloma based on results from the phase III A.R.R.O.W. trial, which demonstrated prolonged progression-free survival (PFS) with once-weekly carfilzomib at 70 mg/m² versus twice-weekly carfilzomib at 27 mg/m² (7, 10).

Oral administration of PIs can enhance accessibility and feasibility of multiple myeloma treatment. Ixazomib is the only oral PI that is currently approved in the United States for the treatment of multiple myeloma (11). In the phase III TOURMALINE-MM1 trial, the addition of ixazomib to lenalidomide and dexamethasone significantly improved PFS compared with lenalidomide-dexamethasone (12). In a subgroup analysis of TOURMALINE-MM1, however, patients with 1 prior therapy and transplant did not have a PFS improvement with ixazomib-lenalidomide-dexamethasone (13).

Because of the limitations and challenges of the 3 PIs currently approved for multiple myeloma treatment, there is a need for additional PIs that are effective, tolerable, and convenient. Oprozomib is an oral tripeptide epoxyketone PI that is structurally analogous to carfilzomib and binds selectively and irreversibly to the proteasome (14). Oprozomib has shown antitumor activity similar to carfilzomib in preclinical models (14–16). Oprozomib inhibited chymotrypsin-like (CT-L) activity in human multiple myeloma cell lines, and a significant decrease in multiple myeloma cell viability occurred in response to oprozomib treatment in in vivo mouse models (15). Oprozomib also inhibited CT-L activity and induced toxicity in primary Waldenström macroglobulinemia cells from previously treated patients with Waldenström macroglobulinemia (16). In a phase I study of patients with refractory or recurrent solid tumors, single-agent oprozomib had pharmacokinetic and pharmacodynamic properties similar to carfilzomib, with proteasome inhibition of ≥80% in the majority of patients (17). Herein, we describe final results from a phase Ib/II study investigating single-agent oprozomib in patients with relapsed multiple myeloma and Waldenström macroglobulinemia.

This was a multicenter, open-label, phase Ib/II study (NCT01416428). In phase Ib, the primary objectives were to determine the MTD of single-agent oprozomib and to evaluate the safety/tolerability of oprozomib in patients with hematologic malignancies. In phase II, the primary objective was to determine the overall response rate (ORR), and the secondary objectives were to determine the duration of response and clinical benefit rate in patients with multiple myeloma, major response rates in patients with Waldenström macroglobulinemia, and PFS, and to define oprozomib pharmacokinetics and assess safety. Exploratory endpoints included evaluation of oprozomib pharmacodynamics. Dates of enrollment were from November 11, 2011, to January 22, 2014, for phase Ib and from October 11, 2013, to June 26, 2015, for phase II.

The phase Ib dose escalation plan followed a standard 3 + 3 study design. In accordance with the original protocol, oprozomib was administered in split dosing as a powder-in-capsule (PIC; 30 or 120 mg); the first dose was followed by a second dose 4 to 6 hours later for 5 consecutive days (days 1–5) of a 14-day treatment cycle (i.e., 5/14 treatment schedule). The starting dose was 120 mg/day, based on interim results of a phase I study of oprozomib capsules in patients with refractory solid tumors (17). Subsequent cohorts were escalated by 30 mg increments up to 210 mg. A protocol amendment in July 2012 enabled administration of oprozomib tablets (60, 90, or 120 mg; modified-release tablet formulation) once daily on the 5/14 treatment schedule or on days 1, 2, 8, and 9 (i.e., 2/7 treatment schedule) of a 14-day cycle. The starting dose was amended to 150 mg/day, based on the safety results from the PIC cohorts at the time the tablet was introduced. Oprozomib doses were subsequently escalated by 30-mg increments in each cohort, up to 330 mg/day (Fig. 1).

In the phase II portion of the study, patients on the 5/14 schedule received oprozomib at a dose of 240 mg/day (MTD, see Results). A subsequent protocol amendment to test alternate regimens and lower doses allowed patients to enroll in a step-up dosing scheme where patients received 150 mg/day in cycle 1, stepped-up to a target dose of 180 mg/day thereafter (i.e., 150/180 mg/day). Per the protocol amendment, patients receiving oprozomib on the 2/7 schedule in the phase II portion received oprozomib in a step-up dosing scheme of 240/300 mg/day. Oprozomib was administered to patients in the fasted state (≥2 hours since the previous meal and fasting for ≥2 hours after dosing) in cycle 1, and after a meal in cycle 2 and subsequent cycles.

Several treatments were required during the administration of oprozomib to mitigate potential side effects. Among these, aprepitant or fosaprepitant were recommended as additional antinausea agents to be used in the event of persistent nausea. For patients developing any-grade diarrhea, loperamide was recommended to be administered at the first sign of symptoms. For patients with persistent diarrhea despite the use of loperamide, diphenoxylate and atropine were strongly recommended, although other antidiarrheal agents were permitted as necessary. Following a protocol amendment in June 2013, valacyclovir (or equivalent antiviral medication) was required to prevent herpes zoster infection. Oral hydration of 1.5–2 L/24 hours was to be instituted for all patients 24 to 48 hours prior to the start of each cycle and was to be continued every day of oprozomib dosing under the June 2013 protocol amendment. All patients were required to receive premedication with a 5-hydroxytryptamine-3 inhibitor (such as ondansetron or granisetron) prior to each dose of oprozomib, with additional doses administered as needed throughout the day during the cycle. Low-dose dexamethasone (4 mg) was also given ≥30 minutes prior to oprozomib dosing. Oprozomib treatment was to be continued until disease progression or unacceptable toxicity.

Adult patients (≥18 years of age) with hematologic malignancies (excluding acute leukemia or myelodysplastic syndrome) who relapsed after receiving ≥1 line of therapy were eligible to enter the phase Ib portion of the study. The phase II portion only enrolled patients with multiple myeloma or Waldenström macroglobulinemia. Other inclusion criteria included Eastern Cooperative Oncology Group performance status 0–2; adequate cardiovascular function without New York Heart Association (NYHA) class III or IV congestive heart failure, symptomatic ischemia, conduction abnormalities uncontrolled by conventional intervention, or myocardial infarction in the previous 6 months prior to first dose of study drug; and adequate renal function, with calculated or measured creatinine clearance rate of ≥30 mL/minute (Cockcroft and Gault formula). Exclusion criteria included history of previous clinically significant gastrointestinal bleed in the last 6 months prior to the first dose of study drug. Additional inclusion and exclusion criteria are listed in Supplementary Methods.

The study was conducted in accordance with the Declaration of Helsinki. Institutional review board/independent ethics committee approval was obtained at each site and all patients provided written, informed consent.

The incidence, nature, severity, and relatedness of adverse events (AE) were graded in accordance with the NCI Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 4.03.

For each dosing schedule, the MTD was defined as the highest dose at which a DLT was observed in fewer than 2 of 6 patients (≥6 patients must have been treated at the MTD for a minimum of 1 cycle to establish this dose as tolerated). Response to oprozomib in patients with multiple myeloma was assessed according to International Myeloma Working Group Uniform Response Criteria (18), and response in patients with Waldenström macroglobulinemia was evaluated according to guidelines of the Sixth International Workshop on Waldenström Macroglobulinemia (19).

For pharmacokinetic evaluation, blood samples were drawn at predose and at time points up to 24 hours postdose. Pharmacokinetic analyses included the measurement of maximum plasma concentration (C_max), time of maximum plasma concentration (t_max), and area under the plasma concentration–time curve (AUC). To assess oprozomib pharmacodynamics, blood samples were drawn. A fluorogenic substrate assay or ELISA was used to measure proteasome activity in red blood cells. Proteasome subunit occupancy was determined via a novel proteasome constitutive/immunoproteasome subunit ELISA (20).

The safety-evaluable population included all patients who received ≥1 dose of oprozomib. All safety analyses were performed using this population, defined by cohort or schedule, as appropriate. Safety was assessed through summaries of treatment exposure, DLTs, and AEs. Extent of exposure to the study treatment was summarized using descriptive statistics. All AEs occurring on or after treatment on cycle 1 day 1 were summarized by system organ class and preferred term using the Medical Dictionary for Regulatory Activities (MedDRA), and the NCI-CTCAE version 4.03. The response-evaluable population included patients in the safety population who had ≥1 postbaseline disease assessment or who discontinued because of an AE before first postbaseline disease assessment. For patients with multiple myeloma, ORR was defined as the rate of patients with at least partial response; for patients with Waldenström macroglobulinemia, ORR was defined as the rate of patients with at least minimal response. Clinical benefit rate for multiple myeloma was defined as ORR plus minimal response, and major response rate for Waldenström macroglobulinemia was defined as partial response or better. The Clopper–Pearson method was used to estimate the 95% confidence interval for each ORR. PFS analyses were performed by the Kaplan–Meier method.

Data were collected, analyzed, and interpreted by the authors and the sponsor. All authors had full access to the primary clinical trial data and agreed to be accountable for the accuracy and integrity of the data and analyses.

Qualified researchers may request data from Amgen clinical studies. Complete details are available at the following: http://www.amgen.com/datasharing.

The phase Ib portion of the study enrolled 71 patients with multiple myeloma or Waldenström macroglobulinemia (Supplementary Fig. S1A). Patient baseline and disease characteristics are shown in Table 1. Among patients in the phase Ib portion, the median patient ages were 62.0 years (2/7 schedule) and 63.0 years (5/14 schedule), and patients had received a median of 3 prior regimens of therapy (range, 1–9; Table 1). In phase Ib, all patients with multiple myeloma on the 2/7 schedule (n = 21) received oprozomib tablets; in the 5/14 schedule, 20 patients received oprozomib tablets and 11 patients received oprozomib PIC. All patients with Waldenström macroglobulinemia on both 2/7 (n = 8) and 5/14 (n = 11) schedules in phase Ib received oprozomib tablets.

The phase II portion of the study enrolled 135 patients, including 102 patients with multiple myeloma and 33 with Waldenström macroglobulinemia (Supplementary Fig. S1B). Median patient ages for patients with multiple myeloma were 65.0 years (2/7, 240/300 mg/day), 64.5 years (5/14, 150/180 mg/day), and 63.0 years (5/14, 240 mg/day). For patients with Waldenström macroglobulinemia, median patient ages were 65.0 years (2/7, 240/300 mg/day) and 62.0 years (5/14, 240 mg/day; Table 1). Patients with multiple myeloma had received a median of 4 (range, 1–12) prior regimens of therapy in the 2/7 schedule, 240/300-mg/day cohort, 5 prior regimens (range, 1–14) in the 5/14, 240-mg/day cohort, and 3 prior regimens (range, 1–10) in the 5/14, 150/180-mg/day cohort. Patients with Waldenström macroglobulinemia had received a median of 2 (range, 1–8) prior regimens in the 2/7 schedule, 240/300-mg/day cohort and 3 (range, 1–7) prior regimens in the 5/14, 240-mg/day cohort. In phase II, all multiple myeloma patients on both 2/7 (n = 41) and 5/14 (n = 61) schedules received oprozomib tablets. All patients with Waldenström macroglobulinemia on both 2/7 (n = 15) and 5/14 (n = 17) schedules in phase II received oprozomib tablets.

Median oprozomib treatment duration for patients with multiple myeloma was 11.4 (range, 0–88) weeks in the 2/7 schedule, 240/300-mg/day cohort, 5.4 (range, 1–49) weeks in the 5/14, 240-mg/day cohort, and 10.1 (range, 3–83) weeks in the 5/14, 150/180-mg/day cohort. For patients with Waldenström macroglobulinemia, median oprozomib treatment durations were 34.6 (range, 0–91) weeks (2/7 schedule, 240/300-mg/day) and 8.1 (range, 1–128) weeks (5/14 schedule, 240-mg/day). The maximum number of cycles patients with multiple myeloma received oprozomib was 44 for the 2/7, 240/300-mg/day schedule, 42 for the 5/14, 150/180-mg/day schedule, and 25 for the 5/14, 240-mg/day schedule. The maximum number of cycles patients with Waldenström macroglobulinemia received oprozomib was 44 for the 2/7, 240/300-mg/day schedule and 61 for the 5/14, 240-mg/day schedule.

In phase Ib, 3 patients on the 2/7 schedule reported DLTs: diarrhea (multiple myeloma; 330 mg/day; n = 1), thrombocytopenia (multiple myeloma; 330 mg/day; n = 1), and hypotension (multiple myeloma; 300 mg/day; n = 1; Supplementary Table S1). Four patients on the 5/14 schedule reported DLTs: renal failure (multiple myeloma; 180 mg/day; n = 1), tumor lysis syndrome (multiple myeloma; 240 mg/day; n = 1), tumor lysis syndrome (Waldenström macroglobulinemia; 240 mg/day; n = 1), and vomiting (Waldenström macroglobulinemia; 270 mg/day; n = 1). The MTD of single-agent oprozomib was therefore determined to be 300 mg/day for the 2/7 schedule and 240 mg/day for the 5/14 schedule. The most commonly observed treatment-emergent AEs in the phase Ib portion of the study were nausea, diarrhea, vomiting, fatigue, and constipation (Supplementary Table S2). Peripheral neuropathy and rash were infrequent, and no treatment-emergent grade ≥3 peripheral neuropathy was observed in the phase 1b portion of the study. Five patients (17.2%) with multiple myeloma or Waldenström macroglobulinemia on the 2/7 schedule discontinued treatment due to AEs, compared with 8 patients (19.0%) on the 5/14 schedule, respectively (Table 2). No patients in the phase Ib portion died within 30 days after receiving their last dose of oprozomib.

On the basis of MTD, efficacy, and safety data from the phase Ib portion of the study, patients in the phase II portion were initially treated on the 5/14 schedule at the MTD of 240 mg/day. In the phase II portion of the study, 2 patients with multiple myeloma who were treated in this cohort died on study due to gastrointestinal hemorrhage, possibly related to oprozomib toxicity (1 patient also died due to plasma cell myeloma in the context of disease progression). The study protocol was amended to address the gastrointestinal hemorrhage observed in the 2 patients, as well as the higher than desired gastrointestinal toxicity observed in the initial phase II cohorts on the 5/14, 240-mg/day schedule. Patients were subsequently enrolled in the 2/7 and 5/14 step-up dosing schedules (240/300 mg/day and 150/180 mg/day, respectively). All patients reported at least 1 AE of any grade. Rates of grade ≥3 AEs were 90.2% (multiple myeloma, 2/7 schedule, 240/300 mg/day), 64.7% (multiple myeloma, 5/14 schedule, 150/180 mg/day), 85.2% (multiple myeloma, 5/14 schedule, 240 mg/day), 53.3% (Waldenström macroglobulinemia, 2/7 schedule, 240/300 mg/day), and 47.1% (Waldenström macroglobulinemia, 5/14 schedule, 240 mg/day). Across all cohorts, the most common grade ≥3 AEs included diarrhea, anemia, thrombocytopenia, fatigue, nausea, and vomiting (Table 3,Table 3). Rates of these grade ≥3 AEs were generally higher among patients with multiple myeloma in the 5/14, 240-mg/day schedule compared with the 2/7, 240/300-mg/day schedule; among patients with Waldenström macroglobulinemia, rates of these AEs were generally similar between the 2/7 and 5/14 dosing schedules. Among patients with multiple myeloma, rates of treatment discontinuation due to AEs were 39.0%, 11.8%, and 44.4% in the 2/7, 240/300-mg/day; 5/14, 150/180-mg/day; and 5/14, 240-mg/day cohorts, respectively. Among patients with Waldenström macroglobulinemia, rates of treatment discontinuation due to AEs were 13.3% and 47.1% on the 2/7 and 5/14 schedules, respectively. No deaths occurred on the 2/7 schedule or in the 5/14 step-up dosing cohorts ≤30 days after the date of the last dose. All patients received antiemetic medication; rates of antidiarrheal medication administration by cohort ranged from 35% to 93%.

The data cutoff for the final analysis of this study was August 8, 2016. In phase Ib, 26 patients on the 2/7 schedule and 39 patients on the 5/14 schedule were eligible for response evaluation (Supplementary Table S3). On the 2/7 schedule, the ORRs for patients with multiple myeloma and Waldenström macroglobulinemia were 27.8% and 37.5%, respectively. On the 5/14 schedule, the ORRs for patients with multiple myeloma and Waldenström macroglobulinemia were 46.4% and 81.8%, respectively.

In phase II, among response-eligible patients with multiple myeloma, ORRs were 41.0%, 28.1%, and 25.0% in the 2/7, 240/300-mg/day (n = 39); 5/14, 150/180-mg/day (n = 32); and 5/14, 240-mg/day (n = 24) cohorts, respectively (Table 4). Among patients with multiple myeloma who were bortezomib-refractory, ORRs were 31.0% and 17.9% on the 2/7 (n = 29) and 5/14 (n = 39) schedules, respectively (Supplementary Table S4). Among patients with multiple myeloma who were carfilzomib-refractory, ORRs were 14.3% and 9.5% on the 2/7 (n = 14) and 5/14 (n = 21) schedules, respectively. For response-eligible patients with Waldenström macroglobulinemia, ORRs were 71.4% and 47.1% on the 2/7 (n = 14) and 5/14 (n = 17) schedules, respectively (Table 4). The ORR in 10 rituximab-refractory Waldenström macroglobulinemia patients on the 2/7 schedule was 70.0%; the ORR in 16 rituximab-refractory Waldenström macroglobulinemia patients on the 5/14 schedule was 62.5% (Supplementary Table S4). Among patients with multiple myeloma, Kaplan–Meier median PFS was 6.1 months in the 2/7, 240/300-mg/day schedule; 3.7 months in the 5/14, 150/180-mg/day schedule; and 3.8 months in the 5/14, 240-mg/day schedule (Fig. 2A). Among patients with Waldenström macroglobulinemia, Kaplan–Meier median PFS was 17.1 months in the 2/7, 240/300-mg/day schedule; and 21.9 months in the 5/14, 240-mg/day schedule (Fig. 2B).

Oprozomib was rapidly absorbed following administration of capsule and tablet in patients with multiple myeloma and tablet in patients with Waldenström macroglobulinemia. Median t_max ranged from 0.4 to 2 hours for the capsule (Supplementary Table S5) and 0.8 to 2 hours for the tablet in patients with multiple myeloma, and 0.5 to 2 hours for the tablet in patients with Waldenström macroglobulinemia (Supplementary Table S6). Oprozomib was rapidly eliminated with a mean terminal elimination half-life of 0.52 to 1.8 hours after tablet administration for patients with multiple myeloma and 0.67 to 2.5 hours after tablet administration for patients with Waldenström macroglobulinemia (Supplementary Table S6). Half-life was not well characterized for capsules in patients with multiple myeloma, with terminal elimination half-life of 5.2 hours observed in a single patient (Supplementary Table S5).

There was, in general, a dose-related increase in oprozomib exposure (as assessed by AUC) following administration of oprozomib in capsules twice daily (separated by 4–6 hours) to patients with multiple myeloma over the total daily dose range of 120 to 210 mg (Supplementary Table S5). Furthermore, a dose-related increase in oprozomib exposure (as assessed by C_max and AUC_last) was observed following administration of oprozomib tablets once daily over the dose range of 150 to 330 mg in patients with multiple myeloma and over the dose range of 150 to 300 mg in patients with Waldenström macroglobulinemia (Supplementary Table S6). Lower exposures were observed for the 330 mg dose level compared with the 300 mg dose level for patients with Waldenström macroglobulinemia, although only a small number of subjects were evaluated (2 subjects for each dose level). High intersubject variability was observed by percent of coefficient of variation (CV%) up to 130% for oprozomib in capsules and up to 157% for oprozomib tablets across the cohorts for AUC_last and C_max. These results should be interpreted with caution considering the observed intersubject variability and small number of subjects evaluated.

Potent proteasome inhibition was observed in whole blood 8 hours after administration of the first dose of oprozomib tablets in the phase Ib portion of the study (Supplementary Fig. S2A). At the MTD on both the 2/7 and 5/14 schedules, proteasome inhibition of approximately ≥70% was observed at 4 hours postdose (Supplementary Fig. S2B). Data obtained from the evaluation of the pharmacodynamics of the oprozomib split-dose PIC formulation are presented in Supplementary Fig. S3.

In this phase Ib/II study, the first to evaluate safety and efficacy outcomes of single-agent oprozomib in patients with multiple myeloma and Waldenström macroglobulinemia, different formulations and dosing schedules of oprozomib were evaluated. In the phase Ib portion, the MTD was determined to be 300 mg/day for the 2/7 schedule and 240 mg/day for the 5/14 schedule. However, 2 patients treated at the 5/14 schedule MTD in the phase II portion died due to gastrointestinal hemorrhage, and subsequent patients were enrolled on step-up dosing schemes. The step-up dosing regimen addressed the higher than desired gastrointestinal toxicity and discontinuation rates observed in the early phase II cohorts in which patients were administered oprozomib 240 mg/day on the 5/14 schedule. In our study, step-up dosing appeared to be better tolerated than continuous dosing, possibly because step-up dosing allowed oprozomib levels to rise gradually. The most common AEs of any grade observed with single-agent oprozomib were gastrointestinal toxicities. Steps were taken during the study to improve the gastrointestinal tolerability of oprozomib, including protocol amendments that required the use of concomitant antidiarrheal medication and premedication with a 5-hydroxytryptamine-3 inhibitor. Fatigue and hematologic toxicities, including anemia and thrombocytopenia, were also observed in ≥30% of patients. When comparing the efficacy and safety between the different treatment dosing schedules in phase II, we found that ORRs were higher in the 2/7 schedule than in the 5/14 schedule in both multiple myeloma and Waldenström macroglobulinemia patients. In phase II, rates of grade ≥3 AEs were lower in the 5/14, 150/180-mg/day schedule than in the 2/7, 240/300-mg/day and the 5/14, 150/180-mg/day schedules. In patients with Waldenström macroglobulinemia, rates of grade ≥3 AEs were similar between the 2/7 and 5/14 schedules.

The pharmacokinetic and pharmacodynamic profile of oprozomib had previously been examined in a phase I study of patients with solid tumors, which demonstrated that oprozomib had oral bioavailability and exhibited dose-dependent inhibition of the proteasome (17). In this study, although high interpatient variability was observed, pharmacokinetic and pharmacodynamic parameters were generally similar to those observed with carfilzomib, with more patients achieving approximately 80% proteasome inhibition at 4 hours postdose compared with 1 hour postdose when single-agent oprozomib was administered at the MTD.

Despite the similarities in pharmacokinetics/pharmacodynamics between oprozomib and carfilzomib, the AE profiles of these agents differ to some extent. Carfilzomib has been associated with cardiac and renal events in phase II and III studies (8, 9, 21). In our study, the most common grade ≥3 AEs associated with oprozomib were gastrointestinal and hematologic in nature, and cardiac failure or hypertension were infrequent. The differences in cardiac toxicity between oprozomib and carfilzomib could be related to variations in C_max, while differences in gastrointestinal toxicity could be related to the route of administration (oral vs. intravenous). In two phase I studies of single-agent oral ixazomib in RR multiple myeloma, the most common treatment-related AEs were similar to those observed with oprozomib in our study, and included thrombocytopenia, diarrhea, nausea, and vomiting (22, 23). However, the gastrointestinal AE rates for ixazomib were lower than the rates seen in our study with oprozomib (22, 23), underscoring the need for an improved oprozomib formulation with reduced gastrointestinal toxicity.

The study underwent several changes resulting from protocol amendments, including changes to the formulation of oprozomib. Initially, enrolled patients received oprozomib twice-daily as a split-dose PIC. Because of gastrointestinal toxicities observed in patients receiving the PIC formulation, the protocol was amended to administer oprozomib as a once-daily modified-release tablet. Pharmacokinetic and pharmacodynamic properties were found to be similar between the split-dose PIC and modified-release tablet formulations in preclinical models (24) and in preliminary results from the current study (25). We also note that 2 patients with chronic lymphocytic leukemia and 1 each with follicular and mantle cell lymphoma were enrolled into the phase I portion of this study, but due to the small sample sizes of these cohorts they have not been characterized here.

Oprozomib showed promising antitumor activity in patients with either multiple myeloma or Waldenström macroglobulinemia. With the caveat that cross-trial comparisons should be interpreted with caution, the ORRs (defined as partial response or better) observed in the phase II portion of this study for patients with multiple myeloma (25.0%–41.0%) were similar to or higher than that of phase I–III trials of other PIs administered as single agents in the relapsed setting (2, 3, 22, 23, 26). In two phase I studies of ixazomib in patients with RR multiple myeloma, ORR was 15% to 18% (22, 23), and a phase II study of intravenously administered carfilzomib in RR multiple myeloma patients demonstrated an ORR of 23.7% (26). For intravenously administered bortezomib in RR multiple myeloma patients, the ORR was 27% in a phase II trial (2) and 38% in a phase III trial (3). For Waldenström macroglobulinemia, the only currently approved treatment is the Bruton tyrosine kinase inhibitor ibrutinib. In the phase II study that led to ibrutinib approval, the ORR was 90.5% and the major response rate (partial response or better) was 73% (27). In our study, the phase II ORRs for patients with Waldenström macroglobulinemia were encouraging, albeit somewhat lower (71.4% and 47.1% on the 2/7 and 5/14 schedules, respectively) than those of ibrutinib, possibly because treatment discontinuation rates were higher in our study than in the ibrutinib study (27). Higher ORRs with oprozomib are likely to be observed in future studies with larger Waldenström macroglobulinemia patient populations, and by improving drug formulations to prevent AE-related discontinuations.

In conclusion, this study shows as proof of principle substantial single-agent oprozomib activity in patients with relapsed multiple myeloma or Waldenström macroglobulinemia, with durable responses overall and responses observed in those refractory to bortezomib or carfilzomib. Despite the efficacy observed with oprozomib as a single agent, there are challenges associated with tolerability. In particular, gastrointestinal events (primarily diarrhea) were among the most common grade ≥3 AEs with single-agent oprozomib. Because of these tolerability limitations, the oprozomib modified-release tablet formulation used in this study is not planned for future studies; rather, oprozomib safety and efficacy are currently being evaluated with immediate-release and gastroretentive tablet formulations and optimized schedule and dosing administration. These studies aim for at least comparable activity and improved single-agent tolerability, particularly gastrointestinal tolerability. For example, in the phase Ib study INTREPID-1 (NCT02939183), the two new formulations of oprozomib plus pomalidomide and dexamethasone are being evaluated in patients with relapsed refractory multiple myeloma. In addition, steroids (e.g., dexamethasone) at doses higher than recommended in this protocol have been shown to reduce gastrointestinal toxicities (28), justifying further development of oprozomib combinations containing dexamethasone. Consequently, other studies of oprozomib and dexamethasone with or without other agents are ongoing, including a phase Ib/II study of oprozomib and dexamethasone plus lenalidomide or oral cyclophosphamide in patients with newly diagnosed multiple myeloma (NCT01881789), a phase Ib/II study of oprozomib and dexamethasone in patients with relapsed and/or refractory multiple myeloma (NCT01832727; ref. 29), and a phase Ib study of oprozomib with pomalidomide and dexamethasone in patients with primary refractory or relapsed and refractory multiple myeloma (NCT01999335).

R. Vij is a consultant/advisory board member for Amgen, Celgene, Bristol-Myers Squibb, Takeda, Janssen, Karypopharm, and Sanofi and reports receiving commercial research grants from Takeda, Celgene, and Bristol-Myers Squibb. D. Siegel reports receiving speakers bureau honoraria from and is a consultant/advisory board member for Amgen. J. Kaufman is a consultant/advisory board member for Celgene, Takeda, Bristol-Myers Squibb, Janssen, Karyopharm, TG Therapeutics, and Incyte. N. Raje is a consultant/advisory board member for Amgen, Celgene, Janssen, Takeda, and Bristol-Myers Squibb. A. Jakubowiak is a consultant/advisory board member for Amgen, Celgene, AbbVie, Bristol-Myers Squibb, Janssen, Juno, Karyopharm, SkylindeDX, and Takeda. M.R. Savona has ownership interest (including patents) in Karyopharm, is a consultant/advisory board member for Celgene, Karyopharm, Incyte, TG Therapeutics, Takeda, Amgen, and Astex, and reports receiving commercial research grants from Incyte, Takeda, TG Therapeutics, Sunesis, and Astex. J.G. Berdeja is a consultant/advisory board member for Takeda, Bristol-Myers Squibb, Karyopharm, CRISPR Therapeutics, Kite Pharma, and Servier and reports receiving commercial research grants from AbbVie, Amgen, Bluebird, Bristol-Myers Squibb, Celgene, Genentech, Glenmark, Janssen, Novartis, Poseida, Sanofi, Takeda, and Teva. No potential conflicts of interest were disclosed by the other authors.

Conception and design: I.M. Ghobrial, D. Siegel, A. Badros, M.R. Savona, M. Obreja

Development of methodology: I.M. Ghobrial, D. Siegel, M. Obreja

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): I.M. Ghobrial, R. Vij, D. Siegel, A. Badros, J. Kaufman, N. Raje, A. Jakubowiak, M.R. Savona, J.G. Berdeja

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): I.M. Ghobrial, R. Vij, D. Siegel, A. Badros, J. Kaufman, A. Jakubowiak, M.R. Savona, M. Obreja, J.G. Berdeja

Writing, review, and/or revision of the manuscript: I.M. Ghobrial, R. Vij, D. Siegel, A. Badros, J. Kaufman, N. Raje, A. Jakubowiak, M.R. Savona, M. Obreja, J.G. Berdeja

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Obreja

Study supervision: I.M. Ghobrial, A. Badros, M.R. Savona, M. Obreja

The authors wish to thank all the patients, families, caregivers, research nurses, study coordinators, and support staff who contributed to this study. Medical writing and editorial assistance was provided by Andrew Gomes and Sachi Yim of BlueMomentum, an Ashfield Company, part of UDG Healthcare PLC, and funded by Amgen Inc. This study was supported by Amgen Inc.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.