Introduction: Similar to most other anti-cancer drugs, carfilzomib is administered at a defined dose per m2. Inspired by preclinical data showing that extrahepatic peptidases play an important role in the metabolism of carfilzomib, using samples from a prospective clinical trial, we conducted an investigation focusing on concentrations of albumin in peripheral blood and its effects on plasma protease-mediated metabolism of carfilzomib.

Methods: Baseline laboratory values, pharmacokinetics data, and response data were collected from the clinical trial. Minimal residual disease was determined by multi-parametric flow cytometry utilizing, 8-color flow panel and analyzes ≥ 3 x 106 events (sensitivity 1 x 10-5). To study the rate of metabolism in plasma, apheresis samples were obtained from the NIH blood bank, lyophilized albumin was added to plasma with and without a non-specific protease inhibitor cocktail (1X). Carfilzomib concentrations were obtained at several time points over the course of 24 hours by mass spectrometry via previously published methods.

Results: Patients with baseline serum albumin >4.0 g/dL had a near 4-fold greater clearance (1576 L/h vs 401.1 L/h; n=9 vs n=34) that was further increased in patients with normal white blood cell count. Consistent with the increase in clearance, 80% of patients (29/33) with albumin <4.0 g/dL became minimum residual disease (MRD) negative whereas only 44.4% of patients (4/9) with albumin >4.0 g/dL had a similar outcome. Therefore, individuals with baseline albumin >4.0 g/dL are at significantly greater risk for not responding to the combination of carfilzomib, lenalidomide, and dexamethasone [OR (95%CI) = 7.3 (1.4-36.7); P=0.0049]. Median progression-free survival was also longer in patients with albumin <4.0 g/dL vs. albumin >4.0 g/dL (unreached vs. 18.8 months; P=0.0002). A trend between albumin and lymphopenia was also detected (P=0.17). The addition of 1.0 g/dL albumin increased the ex vivo proteolytic rate of carfilzomib in apheresis plasma from healthy individuals (74% vs 57.2% unreacted carfilzomib after 24 hours at 37°C), and the addition of a non-specific protease inhibitor reduced carfilzomib metabolism in plasma

Conclusions: Although they have lower grade myeloma, patients with albumin >4.0 g/dL, surprisingly, are at greater risk at having worse clinical outcomes to carfilzomib therapy; this finding appears to depend upon the ability of albumin to directly potentiate plasma protease-mediated metabolism of carfilzomib. Therefore, dose adjustments or concomitant therapy with protease inhibitors may be warranted in such patients. Efforts are underway to identify specific proteases that metabolize carfilzomib and determine potential inhibitors that may be clinically useful.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.