Background: Carfilzomib is an epoxyketone proteasome inhibitor (PI) that binds selectively and irreversibly to the constitutive proteasome and immunoproteasome. In addition to its direct cytotoxicity, PIs possess bone anabolic effect, by facilitating osteoblast differentiation, which in turn, inhibits multiple myeloma (MM) growth. However, the relative contribution of carfilzomib on direct tumor killing and bone formation has been difficult to discern when it is used in combination with other agents such as steroids, chemotherapeutics, or immunomodulatory agents (iMiDs); most of which increase bone loss.

Objective: We explored the effect of single agent carfilzomib on bone metabolism and disease response in patients with relapsed MM.

Methods: We conducted a phase II study of single-agent carfilzomib at 56 mg/m2 day 1, 2, 8, 9, 15, 16 of a 28 day cycle in patients with relapsed MM. Serum markers of bone resorption including carboxyterminal cross-linked telopeptide of type I collagen (ICTP) and tartrate resistant acid phosphatase 5b (TRACP 5b) and bone formation markers including N-terminal type 1 procollagen (P1NP) and osteocalcin (OC) were analyzed at baseline and 6 months after therapy, and compared using paired t-tests. Imaging using 18F-NaF PET-CT scans to monitor bone response and 8F-FDG PET-CT to measure localized tumor burden were done at baseline and 6 months post treatment. For each patient, the three sites with the highest baseline SUVs and an iliac crest (representing an uninvolved site) were selected for quantitative comparisons using paired t-test.

Results: Ten patient were accrued. Average age was 70.1 +/- 6.79 (mean +/- SD) years. Median prior lines of therapy was 2. 9 were evaluable for response. Overall response rate was 77.8%: VGPR (1/9: 11%), partial response (6/9: 67%), stable disease (2/9: 22%). Median PFS was 11.5 months. Overall survival was 21.9 months. Common treatment adverse events (AE) were comparable to previous reports including diarrhea (40%), fever (40%), dyspnea (30%) and fatigue (30%). Grade 3/4 AEs, occurring in 10% each included neutropenia, thrombocytopenia, dyspnea, fatigue, vomiting, ataxia, heart failure and pulmonary edema.

Serum ICTP and TRACP were elevated at baseline in 80% of subjects; means (+/- SD) of serum ICTP (14.4 +/- 3.1 ug/L) and TRACP (8.2 +/- 4.2 U/L), respectively. Both markers declined significantly after therapy (means +/- SD of serum ICTP 8.4 +/- 3.1 ug/L and TRACP (5.2 +/- 4.2 U/L); p<0.01), suggesting osteoclast suppression. Serum PINP and OC varied among subjects; 5 patients with low levels and 4 normal levels and 1 with elevated level: mean (+/- SD) of PINP 25 +/- 23 ug/L and OC (18 +/- 12 ug/L). Neither post treatment mean PINP nor OC differed significantly from baseline overall (p=0.16). However, all subjects who achieve PR or better had a more than 25% increase in PINP and OC, suggesting increased new bone formation.

8F-FDG and 18F-NaF were positive in all subjects. Number of active sites and SUVs vary largely among subjects. 18F-NaF images mirrored the 8F-FDG signals, appearing as rings around the tumor sites, suggesting a bone reaction to the localized tumors. 6 of 9 subjects had post treatment PET/CTs for paired comparison. Post treatment 8F-FDG SUV of representative sites decreased significantly, while 18F-NaF SUVs varied; 58% of lesions had a decline in SUV, while 32% were unchanged and 10% were brighter, suggesting heterogeneous bone response. 18F-NaF SUVs of iliac sites decreased post treatment, suggesting a lower bone turnover overall. CT scans showed increased sclerosis at the rim of lytic lesions without significant change in lesion size, at this early time point.

Conclusion: Carfilzomib at 56 mg/m2 dosing strongly inhibits bone resorption and demonstrates bone anabolic effects, in addition to cytotoxicity, in relapsed MM. Single agent use allows bone response analysis without the confounding effects of other anti-MM agents used in combination therapy. Response rates and side effects are favorable to prior reports. Serum markers of bone remodeling are useful for an assessment of overall bone metabolism while the 18F-NaF PET/CT allows assessment of both local and systemic changes but lacks specificity. Studies of osteoblast differentiation from subject's bone marrow stromal cells pre and post treatment are ongoing. Whether carfilzomib can heal bone lesions requires longer follow up.


Suvannasankha: Celgene: Research Funding; Bristol-Myers Squibb: Research Funding; Takeda: Research Funding; Janssen Pharmaceuticals: Research Funding. Abonour: Celgene: Other: Steering committee, Research Funding; Prothena: Research Funding; Takeda: Other: Steering committee, Research Funding. Roodman: PharmaMar: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.