Renal Function Measures Improve on Lenalidomide and Dexamethasone and Compete with Patient Characteristics to Predict Lenalidomide Dose Density and Hematologic Toxicity: An E4A03 Analysis

Abstract 1882 Introduction: Lenalidomide (L) with dexamethasone (D) has demonstrated activity in untreated symptomatic multiple myeloma (MM) (Rajkumar et al. Lancet Oncol 2010; 11: 29–37). Neutropenia may occur, requiring dose reductions, interruptions, or G-CSF support in up to 20% of patients. Renal clearance accounts for 67% of L elimination, and renal insufficiency is associated with an increased risk of cytopenias. Product information recommends the use of the Cockcroft-Gault (C-G) equation for creatinine clearance (CrCL) estimation for initial dosing; however, other estimates including the Modification of Diet in Renal Disease (MDRD) method may be more accurate. We evaluated the impact of renal function on L dosing, cytopenias, and response in patients on the Eastern Cooperative Group (ECOG) E4A03 study. Methods: All patients (n=445) enrolled on E4A03 were included in this analysis. Clinical data at baseline and after 4 cycles was reviewed. Renal function over the first 4 treatment cycles was estimated using both last reported on cycle 4 and last ever reported serum creatinine (SCr) as the former method resulted in a loss of 70 more patients. Patients were classified by CrCL (mL/min) stage: (1) ≥90 (2) 60–89 (3) 30–59 (4) 15–29 (5) <15. Patients were further categorized into CrCL groups collapsing stages 3–5 while maintaining division of normal status. Demographic and disease data of age, ECOG performance status (PS), sex, race, ISS stage, and complete blood counts were included. Percent of planned L dose and dose interruptions were calculated. Descriptive statistics were used to assess measures at each timepoint. Associations between baseline MDRD, C-G, SCr and patient characteristics were assessed by chi-square testing, and logistic regression models of L dose modification, cytopenias, and response status with renal function measures were created. Results: Mean SCr improved in 361/445 (81%) by 0.23 mg/dL following treatment initiation. Mean MDRD CrCL improved in 264/445 (59%) by 24.3 mL/min. Because only initial weight was captured, C-G estimates over time were not available. ISS stage II/III was associated with baseline elevated SCr (> 1.5 mg/dL) and low CrCL (p<0.001). Dose modification of L occurred in 44% of patients over 4 cycles (Table 1). In adjusted analyses, C-G predicted L dose modifications, p<0.001 [stage 3–5 vs 1 OR 2.09 (1.23-3.55)] along with ECOG PS. Similarly, in adjusted analyses, grade 2–5 cytopenias were better predicted by C-G stage, p<0.001 [stage 3–5 vs 1 OR 5.08 (2.18-11.84)], over MDRD stage, p=0.025 [stage 3–5 vs 1 OR 2.66 (1.07-6.61)]. No association between baseline or change in renal function and ≥ VGPR after 4 cycles was noted. Conclusion: Renal function improvement in patients receiving LD warrants regular CrCL monitoring and L dose increases in patients on reduced doses without cytopenias. Baseline MDRD is not superior to C-G for CrCL estimation for predicting L dosing and cytopenia development. Optimizing L exposure through pharmacokinetic modeling or other methods may further refine dosing. Disclosures: Lonial: Millenium: Consultancy; Celgene: Consultancy; Novartis: Consultancy; Bristol Myers Squibb: Consultancy; Onyx: Consultancy; Merck: Consultancy.