Carfilzomib Dosing Patterns and Survival in Patients with Relapsed and Refractory Multiple Myeloma: An Analysis from US Community Oncology Practices

Introduction

While novel agents have improved survival over the last decade, multiple myeloma (MM) remains incurable.  Carfilzomib (CFZ), a second-generation proteasome inhibitor, was approved in July 2012 by the US Food and Drug Administration and had a label change in July 2015. During this study's data period, the label recommended Cycle 1 dose at 20 mg/m²/day and if tolerated increase Cycle 2 dose and subsequent cycles doses to 27 mg/m²/day.  The purpose of this study is to assess baseline characteristics, CFZ dosing patterns and survival among MM patients in a US community oncology setting.

Methods

A retrospective study of MM patients from US Oncology Network practices that fully implemented McKesson Specialty Health's iKnowMed (iKM) oncology-specific electronic health records database was conducted on patients whose first treatment of CFZ (index) occurred between Jul-2012 and Nov-2014.  Patients were eligible if they had a documented initial MM diagnosis date and had their first CFZ cycle documented in the database.  Additionally, patients were required, before Dec-2014, to have either another visit to the practice post-index or a record of death and not have participated in interventional clinical trials during the previous 6 years.  Data on eligible patients were collected up to March 2015.  The death event was defined by the Social Security Death Index, supplemented by iKM; patients without the event were censored at the date of last observed visit.  To adjust for clinical practice variations, a 10% variability was allowed for the recommended daily dose levels of 20 mg/m² and 27 mg/m².  A subgroup was defined for patients with a 2^nd cycle: “escalators” if they received 20 mg/m²/day doses throughout Cycle 1 and increased to 27 mg/m²/day on the first dose of Cycle 2; “non-escalators” if they received only 20 mg/m² doses throughout Cycle 1 and did not increase to 27 mg/m² on the first dose of Cycle 2; receiving any dose not equal to 20 or 27 mg/m² were classified into “other”.  Survival after index was estimated using the Kaplan-Meier method with 95% confidence intervals (CI).  A multivariable Cox proportional hazards (PH) model was conducted to evaluate the impact of escalation on survival accounting for selected baseline demographic and clinical characteristics.

Results

The cohort of 718 CFZ patients were identified with a median (interquartile range [IQR]) age of 68 (61-75) years at index, 57% (n=409) were male, and 12% (n=87) were Black and 77% (n=551) were Caucasian.  At initial MM diagnosis, 19%, 27% and 42% were ISS Stage I, II, and III, respectively.  Median (IQR) time from MM diagnosis to index was 3.6 (1.9-5.8) years. At index, 66% of patients had an ECOG performance status of 0-1, 21% of 2, and 2% of 3+; 54% (n=369) had moderate to severe renal impairment (eGFR<60 mL/min per 1.73 m²).  Ninety percent (n=644) of patients started CFZ at 20 mg/m², 4% (n=27) at 27 mg/m² and 4% (n=25) at 15 mg/m².  Patients had a median (IQR) of 4 (2-7) cycles of CFZ and 45% (n=321) escalated to ≥27 mg/m².  Among these 321 patients, median (IQR) time to first escalation was 30 (28-56) days with 60% escalating in Cycle 2.  The subgroup defined in “Methods” included 605 patients: 148 (24%) escalators, 342 (57%) non-escalators, and 115 (19%) other.  Median (95% CI) duration from index to death was 21 (17.5-23.2) months.  Unadjusted overall survival (OS) was significantly lower among non-escalators compared to escalators (log-rank p=0.024) [Figure 1].  Survival rates (95% CI) for non-escalators and escalators were 68% (62-74%) and 75% (66-82%) at year 1 and 42% (33-50%) and 61% (49-71%) at year 2, respectively.  Within the multivariable Cox model, escalators showed a 33% significantly lower risk of death compared to non-escalators (HR=0.67, p=0.03) while also accounting for race, sex, age group, renal function per EGFR, and MM chain type.  Other significant variables in this model were: EGFR < 15 and 15-29 vs 30-59 ml/min per 1.73m² (HR=2.79, p<0.01; HR=1.64, p=0.04, respectively) and lambda vs kappa light chain (HR=1.55, p=0.03).   

Conclusions

These results indicate escalation of CFZ at first dose of Cycle 2 is associated with better survival than dosing at 20 mg/m² in Cycle 1 but not escalating at the start of Cycle 2.  More research is needed to assess factors that impact physician decision-making on dose escalation to better inform physicians to improve the quality of multiple myeloma care.