The Chronic Kidney Disease-Epidemiology Collaboration-Cystatin C (CKD-EPI-CysC) Equation Has an Independent Prognostic Value for Overall Survival in Newly-Diagnosed Patients with Symptomatic Multiple Myeloma: Comparison with 3 Other Equations for GFR Estimation

Abstract 1849

Renal impairment (RI) is a frequent complication in multiple myeloma (MM). The estimation of glomerular filtration rate (GFR) is based on equations that use serum creatinine (sCr) as a marker of RI (i.e. MDRD or CKD-EPI). The IMWG has recommended the use of the MDRD formula for the estimation of GFR in MM patients with stabilized sCr, while the classification of RI is based on the 5 stages of the KDIGO classification (Dimopoulos et al, JCO 2010;28:4976–84). However, the equations based on sCr are imprecise and thus novel markers of renal injury have been used in patients with renal damage, including cystatin-C (CysC). CysC is considered as a more sensitive marker of GFR than sCr. Recently, the CKD-EPI investigators have reported that a combined sCr-CysC (CKD-EPI-sCr-CysC) equation correlated better with GFR than equations based on either of these markers alone (CKD-EPI or CKD-EPI-CysC; Inker et al, NEJM 2012;367:20–9). Although cysC has been reported by our group to be elevated in MM patients, the CKD-EPI equations and their value on MM patients' survival have never been evaluated.

Therefore, we studied 220 newly-diagnosed, previously untreated, symptomatic MM patients. The median age was 69 years (range: 36–94 years) and 16% had sCr ≥2 mg/dl. Serum CysC was measured on the BN ProSpec analyser using a latex particle-enhanced nephelometric immunoassay (Dade Behring-Siemens Healthcare Diagnostics, Liederbach, Germany). Serum CysC was increased in MM patients compared to 52 age- and gender-matched controls [median: 1.07 mg/l vs. 0.72 mg/l, p<0.0001]. The median values for eGFR calculated by the MDRD, CKD-EPI, CKD-EPI-CysC and CKD-EPI-sCr-CysC equations were 63.45 ml/min/1.73m², 68.13 ml/min/1.73m², 68.11 ml/min/1.73 m², and 64.87 ml/min/1.73 m², respectively (p<0.01).

Patients were divided in the 5 CKD stages of KDIGO classification, according to eGFR (stage 1: eGFR >90 ml/min/1.73 m²; stage 2: 60–89 ml/min/1.73m²; stage 3: 30–59 ml/min/1.73 m²; stage 4: 15–29 ml/min/1.73 m²; stage 5: <15 ml/min/1.73 m² or on dialysis). For each studied equation, the number of patients with RI stage 3–5 (i.e. eGFR <60 ml/min/1.73²) was 39.5% for MDRD vs. 42.2% for CKD-EPI vs. 43.1% for CKD-EPI-CysC vs. 45% for CKD-EPI-sCr-CysC (p<0.01; see also the table). Concordance for CKD stage allocation for the 4 equations of estimating eGFR was 97% for MDRD vs. CKD-EPI, 60% for MDRD vs. CKD-EPI-CysC and 84% for MDRD vs. CKD-EPI-sCr-CysC. A significant correlation was found between ISS stage and all studied equations (p<0.01 for all).

The median overall survival (OS) for all patients was 52 months. In the univariate analysis per CKD stage, the 4 equations could predict for OS (the higher CKD stage had poorer survival) with the following significance: MDRD (p=0.057), CKD-EPI (p=0.01), CKD-EPI-CysC (p<0.0001), and CKD-EPI-sCr-CysC (p=0.006). When we tested the 4 equations as continuous variables, all had prognostic value for OS but the CKD-EPI-CysC had the strongest prognostic value (p<0.0001 and Wald=24.0 vs. p<0.0001 and Wald=19.7 for CKD-EPI-sCr-CysC, p=0.003 and Wald=8.9 for CKD-EPI and p=0.005 and Wald=7.7 for MDRD). In the multivariate analysis, that included ISS stage, LDH ≥300 U/l and eGFR for each different equation (as a continuous variable) only eGFR that included CysC but not sCr (CKD-EPI-CysC and not CKD-EPI-sCr-CysC) had independent significance (p=0.013) along with high LDH (p=0.029).

Our data suggest that the CKD-EPI-sCr-CysC equation for the estimation of GFR detects more MM patients with stage 3–5 RI than the MDRD, CKD-EPI or CKD-EPI-CysC equations. However, CKD-EPI-CysC was the only equation that could predict for OS, possibly due to the very strong correlation of CysC with ISS (as myeloma cells produce CysC also). The confirmation of these data will lead to the broader use of equations based on CysC (CKD-EPI-CysC with or without sCr) for the evaluation of RI in patients with MM, as it has been suggested for patients with several other renal disorders.