DLTs were graded using the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 and included the below, if judged by the investigator/medical monitor to be possibly related to study treatment: Non-hematologic DLTs were any ≥ Grade 3 toxicity, except: Grade 3 asymptomatic electrolyte abnormalities (except hypophosphatemia > 24 hours); Grade 3 nausea, vomiting and diarrhea < 3 days; Grade 3 fatigue < 14 days; ≥ Grade 3 hyperglycemia/toxicity due to dexamethasone; ≥ Grade 3 rash due to pomalidomide. Hematologic DLTs included: Grade 4 neutropenia if absolute neutrophil count < 0.5 x 10^9/L ≥ 7 days; febrile neutropenia; Grade 4 thrombocytopenia ≥ 7 days; Grade ≥ 3 with ≥ Grade 2 bleeding/requiring platelet transfusion.

DLT-evaluable participants received the following during the 28-day DLT window: all planned doses of oprozomib, a minimum of 17 of 21 planned doses of pomalidomide, and a minimum of 6 of 8 planned doses of dexamethasone. One participant enrolled in arm Oprozomib IR 200 mg/day + pomalidomide + dexamethasone received oprozomib IR 150 mg + pomalidomide + dexamethasone, and this participant was included in the Oprozomib IR 150 mg + pomalidomide + dexamethasone arm.

The MTD was the dose with the highest posterior probability of having a DLT rate within the target toxicity interval (15% to 25%), while the posterior probability of excessive/unacceptable toxicity (>25% to 100%) is <40%.

DLTs were graded using CTCAE version 4.03 and included the below, if judged by the investigator/medical monitor to be possibly related to study treatment: Non-hematologic DLTs were any ≥ Grade 3 toxicity, except: Grade 3 asymptomatic electrolyte abnormalities (except hypophosphatemia > 24 hours); Grade 3 nausea, vomiting and diarrhea < 3 days; Grade 3 fatigue < 14 days; ≥ Grade 3 hyperglycemia/toxicity due to dexamethasone; ≥ Grade 3 rash due to pomalidomide. Hematologic DLTs included: Grade 4 neutropenia if absolute neutrophil count < 0.5 x 10^9/L ≥ 7 days; febrile neutropenia; Grade 4 thrombocytopenia ≥ 7 days; Grade ≥ 3 with ≥ Grade 2 bleeding/requiring platelet transfusion.

DLT-evaluable participants in Part 2 who received the following during the 28-day DLT window: all planned doses of oprozomib, a minimum of 17 of 21 planned doses of pomalidomide, and a minimum of 6 of 8 planned doses of dexamethasone.

An adverse event (AE) is defined as any untoward medical occurrence in a clinical trial participant. TEAEs were any AE that started on or after receiving the first dose of investigational product and up to and including 30 days after the last dose of investigational product or the end of study date, whichever is earlier. Treatment-related TEAEs were those considered related to study treatment by the investigator. Any clinically significant changes in electrocardiograms, vital signs, physical examination with a neurological assessment and clinical laboratory tests were recorded as TEAEs.

Participants in the safety analysis set who received any amount of oprozomib, pomalidomide and/or dexamethasone in Parts 1 and 2.

TEAEs were any AE that started on or after receiving the first dose of investigational product and up to and including 30 days after the last dose of investigational product or the end of study date, whichever is earlier. Treatment-related TEAEs were those considered related to study treatment by the investigator. Serious TEAEs were any AE meeting at least 1 of the following criteria: fatal; life-threatening; required in-patient hospitalization or prolongation of existing hospitalization; resulted in persistent or significant disability/incapacity; congenital anomaly/birth defect; other medically important serious event.

The safety analysis set for roll-over participants included all roll-over participants who received any amount of oprozomib, pomalidomide and/or dexamethasone. Participants enrolled in this study from other Amgen studies were included in a single roll-over arm, as pre-specified in the SAP.

The mean Cmax of oprozomib is presented following dosing on cycle 1 day 8 and cycle 1 day 22 for the IR and GR oprozomib formulations.

Participants in the PK analysis set with data available at each time point. The PK analysis set included participants for whom at least 1 PK parameter could be adequately estimated, excluding roll-over participants.

The median Tmax of oprozomib is presented following dosing on cycle 1 day 8 and cycle 1 day 22 for the IR and GR oprozomib formulations.

Participants in the PK analysis set with data available at each time point. The PK analysis set included participants for whom at least 1 PK parameter could be adequately estimated, excluding roll-over participants.

The mean AUClast of oprozomib is presented following dosing on cycle 1 day 8 and cycle 1 day 22 for the IR and GR oprozomib formulations.

Participants in the PK analysis set with data available at each time point. The PK analysis set included participants for whom at least 1 PK parameter could be adequately estimated, excluding roll-over participants.

BOR was the best response per IMWG-URC from best to worst: stringent complete response (sCR; complete response [CR], normal serum free light chain ratio, no clonal cells in bone marrow [BM]), CR (negative immunofixation on serum and urine, soft tissue plasmacytomas disappearance, < 5% plasma cells in BM), very good partial response (VGPR; serum and urine M-protein detectable by immunofixation or ≥ 90% reduction in serum M-protein [urine M-protein level < 100 mg/24-h]), partial response (PR; ≥ 50% reduction of serum M-protein and 24-h urinary M-protein by ≥ 90% or < 200 mg/24-h), minimal response (MR; 25-49% reduction of serum M-protein and 50-89% in 24-h urinary M-protein, exceeding 200 mg/24-h), stable disease (SD; not CR, VGPR, PR or progressive disease [PD]), and PD (≥ 25% increase in serum or urine M-component, development of new or increased size of existing bone lesions or soft tissue plasmacytomas, hypercalcemia attributed to the plasma cell proliferative disorder).

The efficacy analysis set included all participants, excluding roll-over participants, who were included in the safety analysis set, and had a baseline disease assessment and at least 1 post-baseline disease assessment.

The ORR was defined as the percentage of participants with a BOR of sCR, CR, VGPR, and PR per the IMWG-URC. Complete response (CR): Negative immunofixation on serum and urine, soft tissue plasmacytomas disappearance, < 5% plasma cells in bone marrow (BM). Stringent CR (sCR): CR and normal serum free light chain ratio and no clonal cells in BM. Very Good Partial Response (VGPR): Serum and urine M-protein detectable by immunofixation or ≥ 90% reduction in serum M-protein (urine M-protein level < 100 mg/24-h). PR: ≥ 50% reduction of serum M-protein and 24-h urinary M-protein by ≥ 90% or to < 200 mg/24-h. The 95% confidence intervals (CIs) were estimated using the Clopper-Pearson method.

The efficacy analysis set included all participants, excluding roll-over participants, who were included in the safety analysis set, and had a baseline disease assessment and at least 1 post-baseline disease assessment.

PFS was defined as the number of months from a participant's first dose of study treatment to the earlier of disease progression or death due to any cause. The number of participants who had a PFS event or who were censored are presented. PFS events were an assessment of progressive disease according to the IMWG-URC or death due to any cause. PFS data was censored: participants alive and no documented disease progression at time of analysis were censored at date of last disease assessment; participants alive with no disease assessment were censored at the first study dose date; participants alive without documented disease progression, with withdrawn consent were censored at date of last disease assessment before consent withdrawal; and for participants who started anti-cancer therapy other than study treatment prior to documentation of disease progression were censored at date of last disease assessment prior to starting new therapy.

The efficacy analysis set included all participants, excluding roll-over participants, who were included in the safety analysis set, and had a baseline disease assessment and at least 1 post-baseline disease assessment.

PFS was defined as the number of months from a participant's first dose of study treatment to the earlier of disease progression or death due to any cause. Median PFS was estimated using the Kaplan-Meier method. 95% CIs were estimated using the method by Klein and Moeschberger (1997) with log-log transformation. PFS data was censored: participants alive and no documented disease progression at time of analysis were censored at date of last disease assessment; participants alive with no disease assessment were censored at the first study dose date; participants alive without documented disease progression, with withdrawn consent were censored at date of last disease assessment before consent withdrawal; and for participants who started anti-cancer therapy other than study treatment prior to documentation of disease progression were censored at date of last disease assessment prior to starting new therapy.

The efficacy analysis set included all participants, excluding roll-over participants, who were included in the safety analysis set, and had a baseline disease assessment and at least 1 post-baseline disease assessment.

DOR, presented in months, was defined as the number of days between the date of the first tumor assessment indicating an objective response (PR or better) through to the subsequent date of progression or death due to any cause, or where applicable date of censoring (date of first progressive disease assessment or death or date of censoring - date of the first objective response result + 1/30.4). Median PFS was estimated using the Kaplan-Meier method. 95% confidence intervals were estimated using the method by Klein and Moeschberger (1997) with log-log transformation.

The efficacy analysis set included all participants, excluding roll-over participants, who were included in the safety analysis set, and had a baseline disease assessment and at least 1 post-baseline disease assessment. Participants who did not achieve an objective response were excluded from the analysis of DOR. Participants who responded and had not progressed while on study were censored at the date of assessment of the last evaluable tumor assessment.