Abstract
Background: Lenalidomide (LEN) is effective in MDS patients (pts) with or without deletion (del) 5q cytogenetic abnormalities. Common toxicities include neutropenia and thrombocytopenia. Both occurrence of cytopenias and response to LEN is more common in pts with the del 5q abnormality. This study analyzes whether development of treatment-related cytopenias is associated with response to LEN in lower-risk MDS pts.
Methods: Transfusion-dependent, low/int-1-risk MDS pts were enrolled in the MDS-003 (del 5q pts) and MDS-002 (non-del 5q pts) studies. Pts were treated with 10 mg LEN (daily or 21/28 days). Baseline thrombocytopenia was defined as a platelet (plt) count <150,000/mm3; neutropenia as an absolute neutrophil count (ANC) <2000/mm3 (grade 1–4 using the CTC v2.0). Cytopenias were assessed within the first 8 weeks of LEN therapy, and given functional definitions based on frequency tables. Response was assessed using International Working Group criteria.
Results: Of 147 evaluable pts in MDS-003, 59 (40%) had thrombocytopenia, 59 (40%) neutropenia, and 84 (57%) neutropenia and/or thrombocytopenia according to baseline labs. Of 210 evaluable pts in MDS-002, 69 (33%) had thrombocytopenia and 81 (39%) neutropenia at baseline. For both studies, median age was 71 and 72 years and MDS duration was 2.5 and 2.2 years, respectively. RBC transfusion independence (TI) was achieved by 99 pts (67%) in MDS-003 (List et al. NEJM 2006) and 56 pts (26%) in MDS-002. For pts with del 5q, development of thrombocytopenia correlated with TI, regardless of baseline plt count (p=0.005). Comparing pts who had a ≥50% drop vs those who did not, TI was achieved in 76% vs 47% of pts without baseline thrombocytopenia and in 67% vs 38% of pts with thrombocytopenia, respectively. Similar results held for pts without baseline neutropenia: 82% whose ANC fell ≥75% achieved TI, compared to 56% whose ANC fell <75% (p=0.018). In pts with baseline neutropenia, ANC drop did not correlate with TI (p=0.75). In pts with any baseline cytopenia, those whose ANCs fell by ≥75% and/or plt by ≥50% were more likely to achieve TI than those whose counts did not drop substantially, controlling for baseline cytopenias (71% vs. 60%, p=0.024). In multivariate analyses, both a treatment-related ANC drop ≥75% (odds ratio [OR]=2.68, p=0.04) and a plt drop ≥50% (OR=2.79, p=0.05) remained associated with TI in MDS-003. Neither was associated with duration of TI response, though there was a trend with drop in ANC (hazard ratio=2.04, p=0.06). In contrast, for pts without del 5q (MDS-002), no correlation exists between TI and drop in plt count (p=0.36 for patients without and p=0.16 for those with baseline thrombocytopenia), drop in ANC (p=0.43 for those without and p=0.44 for those with baseline neutropenia), or development of either cytopenia. No correlation with TI could be established in MDS-002 for drops of 25%, 50%, or 75% within 4, 8, or 16 weeks of therapy, in both univariate and multivariate analyses.
Conclusions: In MDS pts with del 5q, treatment-related thrombocytopenia, and neutropenia in those with normal baseline ANCs, correlate with response to LEN, supporting the link between suppression of the del 5q clone and erythroid response. This correlation was not observed in non-del 5q MDS pts, indicating alternate mechanisms of action of LEN.
High-dose regimens of hypomethylating agents promote transfusion independence in IPSS lower-risk myelodysplastic syndromes: a meta-analysis of prospective studies