Effect of Lenalidomide (LEN) Exposure on AML-Free Survival and Overall Survival in LEN-Treated Patients (Pts) with IPSS Low- or Intermediate-1-Risk Myelodysplastic Syndromes (MDS) with Del(5q)

Introduction: The efficacy and safety of LEN in red blood cell (RBC) transfusion-dependent pts with International Prognostic Scoring System (IPSS)-defined Low- or Intermediate-1-risk MDS and del(5q) was assessed in 2 large pivotal trials; protocol-defined dose modifications due to adverse events were required in the majority of LEN-treated pts (MDS-003 [List A, et al. N Engl J Med. 2006;355:1456-65] and MDS-004 [Fenaux P, et al. Blood. 2011;118:3765-76]). Previously we have shown that achievement of cytogenetic response (CyR) with LEN is associated with LEN dose and with achieving RBC transfusion-independence and improved acute myeloid leukemia (AML)-free survival. The current analysis evaluates the impact of actual LEN exposure, including induction-type dosing in Cycle 1 and subsequent dose reductions for LEN-associated cytopenias, and CyR on AML-free survival and overall survival (OS) in lower-risk MDS pts with del(5q) treated in the MDS-003 and MDS-004 studies.

Methods: This analysis includes pooled data from pts who were treated with LEN in MDS-003 and MDS-004. Pts received LEN at one of the following starting doses and schedules: 5 mg/day, days 1–28 (MDS-004); 10 mg/day, days 1–21 (MDS-003 and MDS-004); or 10 mg/day, days 1–28 (MDS-003); all given in 28-day cycles. LEN dose reductions related to adverse events (NCI CTCAE v3.0) of grade ≥ 3 (MDS-003) or grade 4 thrombocytopenia or neutropenia (MDS-004) were predefined in the study protocols. Full dose modification guidelines for MDS-004 have been published previously (Fenaux P, et al. Blood. 2011;118:3765-76). Dose interruptions were not considered. AML-free survival and OS were estimated by the Kaplan–Meier method with differences evaluated by the log-rank test. Univariate and multivariable Cox proportional hazards models were used to identify predictive factors for AML-free survival and OS. CyR and LEN dose reduction were analyzed as time-varying covariates.

Results: Among the 286 LEN-treated pts from MDS-003 and MDS-004, median total doses in Cycle 1 were 245, 210, 280, and 140 mg in the 10 mg (MDS-003: 10 mg × 28 days), 10 mg (MDS-003: 10 mg × 21 days), 10 mg (MDS-004: 10 mg × 21 days), and 5 mg (MDS-004: 5 mg × 28 days) treatment groups, respectively. Median total times on LEN were 364, 385, 510, and 273 days, respectively, and median times to first dose reduction were 53, 63, 54, and 63 days across the treatment groups, respectively. AML-free survival and OS were significantly longer in pts who received > 210 mg versus ≤ 210 mg in Cycle 1 of therapy (log-rank P = 0.0005 and P = 0.0002, respectively). In multivariable analyses, higher total LEN dose in Cycle 1, analyzed as a continuous variable, was significantly associated with improved AML-free survival (hazard ratio [HR] 0.97; P = 0.033) and OS (HR 0.97; P = 0.036) (Table). Sensitivity analyses showed total dose in Cycles 1–3 to be significant, but total dose in Cycle 1 was a better predictor. In the same model, the occurrence of LEN dose reduction was significantly associated with improved AML-free survival (HR 0.54; P < 0.001) and OS (HR 0.56; P = 0.001) (Table) and was strongly associated with longer duration on study (P < 0.001).

Conclusions: In this pooled analysis of LEN-treated pts with MDS and del(5q) from the pivotal MDS-003 and MDS-004 trials, the total cumulative LEN dose actually received in Cycle 1 was a significant predictor of AML-free survival and OS; this is likely mediated through clone suppression, evidenced by the association with CyR. Long-term outcomes were significantly improved in LEN-treated pts who had received early cumulative exposure using 10 mg daily and longer duration through dose modifications. These findings support the use of LEN 10 mg as the recommended induction starting dose, and the importance of dose reductions as required to maximize treatment duration and optimize response and survival outcomes in these pts with lower-risk MDS with del(5q).