Renal Response in Carfilzomib (K)- Versus Bortezomib (V)-Treated Patients (Pts) with Relapsed or Refractory Multiple Myeloma (RRMM) in the Real-World Practice Setting

Introduction: Renal impairment (RI) is a classic clinicopathological feature of MM, is associated with poor prognosis and shorter survival (Eleutherakis-Papaiakovou Leuk Lymphoma 2007), and can complicate drug dosing and limit treatment options. In clinical trials, the proteasome inhibitors K and V have demonstrated efficacy in RRMM pts with RI, with superior efficacy observed in K-treated pts (Dimopoulos Blood 2017). K and V have also been observed to improve renal function in RRMM pts (Dimopoulos Clin Lymphoma Myeloma 2009; Dimopoulos Blood Advances 2017). However, few studies have investigated the real-world effectiveness of K- and V-based regimens in renal rescue. This study aimed to describe and compare renal response rates in RRMM pts treated with K + dexamethasone (Kd) and V + dexamethasone (Vd) in oncology clinics.

Methods: The Oncology Services Comprehensive Electronic Records (OSCER) database (Lau Clin Epidemiol 2011) contains electronic medical records (EMR) from community- and hospital-affiliated oncology clinics in the United States. MM pts aged ≥18 who entered an OSCER clinic Jan 2012‒Feb 2018; initiated Kd or Vd treatment as line 2 (2L), 3L, or 4L; and had baseline renal impairment (Modification of Diet in Renal Disease Study estimated glomerular filtration rate [eGFR] <50 mL/min from the most recent baseline serum creatinine measurement) were included. Index date was the date of Kd or Vd initiation. The baseline period was 60 d prior to index to 30 d post-index. Follow-up was 30 d after index date to 60 d after completion of the same line. Primary outcome was renal response during follow-up. Renal response was defined according to the International Myeloma Working Group criteria (Dimopoulos J Clin Oncol 2016): complete response (RCR; baseline eGFR <50 mL/min and best post-treatment eGFR ≥60 mL/min), partial response (baseline eGFR <15 mL/min and best post-treatment eGFR 30‒59 mL/min), minor response (baseline eGFR <15 mL/min and best post-treatment eGFR 15‒29 mL/min, or baseline eGFR 15‒29 mL/min and best post-treatment eGFR 30‒59 mL/min). Renal response rates in Kd- and Vd-treated pts were evaluated using the Kaplan-Meier method and log-rank test. Incidence rate ratios (IRR) and 95% confidence intervals (CIs) were calculated for renal overall response (ROR) and RCR using multivariate Cox proportional hazard models adjusted for baseline covariates (demographics, Eastern Cooperative Oncology Group Performance Status, International Staging System stage, baseline renal function, V use in 1L, time from MM diagnosis to line initiation, use of intravenous bisphosphonates, and baseline serum calcium, serum free light chain ratio, lactate dehydrogenase, and whole blood hemoglobin measurements). Missing baseline values were estimated by multiple imputations.

Results: 543 Kd-treated and 1005 Vd-treated pts were included. Baseline characteristics were similar between cohorts, with no difference in baseline eGFR stages. More Kd pts were <65 yrs of age at MM diagnosis. Median time from diagnosis to line initiation was longer in Kd pts (Kd: 17.1 mos; Vd 16.1 mos). V-based regimens were the most common therapies in 1L. Median (interquartile range) treatment duration was 4.4 (2.3-8.3) mos for Kd and 4.3 (2.1-8.3) mos for Vd. Pts treated with Kd at 2L were more likely than those treated with Vd to achieve both ROR (51.4% [178/346] vs 39.6% [327/825]; log-rank test p<0.0001) and RCR (26.6% [92/346] vs 22.2% [183/825]; log-rank test p=0.0229). This pattern persisted when 3L and 4L pts were included, (ROR: 48.4% [263/543] vs 39.8% [400/1005]; log-rank p<0.0001; RCR: 26% [141/543] vs 22.1% [222/1005]; log-rank p=0.0084) (Figures 1A and 1B). Following adjustment for potential baseline confounders, 2L Kd pts were 45% more likely to achieve ROR (IRR, 95% CI: 1.45, 1.18‒1.78) and 68% more likely to achieve RCR (1.68, 1.24‒2.28) vs 2L Vd pts (Figure 2). Kd pts in 2‒4L were 36% more likely to achieve ROR (1.36, 1.15‒1.62) and 52% more likely to achieve RCR (1.52, 1.19‒1.94).

Conclusions: Using one of the most complete oncology EMR databases available, we found that RRMM pts with baseline RI treated with Kd had higher rates of overall and complete renal response, and were more likely to have a renal response, compared with pts treated with Vd. A limitation of the study is that we could not account for all concomitant medications or comorbidities that affect renal function, as some may not be captured in OSCER.

View largeDownload slide

View largeDownload slide

Close modal