A Methylation Profile of Tumor Suppressor Genes Predicts a Poorer Survival in Patients with Refractory Anemia with Ringed Sideroblasts.

Abstract 2780

Poster Board II-756

Patients with refractory anemia with ring sideroblasts (RARS) are considered to have good prognosis and anemia is usually managed with best supportive care and erythroid stimulating agents. Nowadays, no specific molecular marker related to outcome and disease progression has been identified. Several genes involved in cell cycle and apoptosis that may become inactivated by aberrant hypermethylation have been identified in patients with myelodysplastic syndromes (MDS) but the significance of a methylation profile (studying several genes at the same time) in RARS is unknown, mainly because the number of patients with RARS in published reports is rather low. To assess the implication of aberrant methylation in RARS, we studied the methylation status of 25 sequences of a set of tumor suppressor genes in 40 patients (median age 70 yr, 25 male gender) with RARS according to FAB criteria [WHO classification, RARS in 22 patients (55%); refractory citopenia with multilineage dysplasia and ring sideroblasts, 18 (45%)] by means of methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA) assay. The MS-MLPA procedure (SALSA MLPA kit ME001, MRC-Holland, Amsterdam, The Netherlands) has been developed for detecting in a semi-quantitative manner changes in the methylation status of 25 tumor suppressor genes in a single experiment. Briefly, approximately 50 ng of DNA were denatured and hybridized with MLPA probes. Subsequently, the probes were ligated in half of every sample, whereas for the rest of the sample, ligation was combined with an endonuclease HhaI digestion resulting in ligation of the methylated sequences only. PCR was performed as described by the manufacturer, and then separated by capillary gel electrophoresis and quantified using the Genemapper software (ABI 310, Applied Biosystems, Foster City, CA). Quantification of the methylation status was done by dividing the peak area with the combined areas of the control probes lacking the target sequence of the HhaI. Finally, the relative peak area of each target probe from the digested sample was compared with those obtained from the undigested sample. Aberrant methylation was scored when the calculated methylation percentage was >10%. To validate the MS-MLPA method the methylation status of CDKN2B was also analyzed by methylation-specific PCR (MSP). Actuarial curves of OS were built by Kaplan-Meier method and differences between curves compared with log-rank tests. Small numbers precluded the use of multivariate methodology. The MS-MLPA analysis detected methylation of at least one gene in 17 patients (42.5%). The genes aberrantly methylated were CDKN2B (n = 8, 20%), RASSF1 (n = 7, 17%), RARB (n = 3, 7.5%), CDH13 (n = 3, 7.5%) and FHIT (n = 2; 5%). Of the 17 patients, five (12%) presented methylation in two genes (RASSF1-FHIT in 2, RASSF1-RARB in 1, RASSF1-CDH13 in 1, and CDKN2B-CDH13 in 1, who was the only patient who progressed to AML). The presence of aberrant gene methylation was not significantly associated with any clinical or biological characteristic or WHO morphological subtype. Patients with aberrant gene methylation had a significantly shorter overall survival (OS) than patients without methylated genes (median OS, 72 mo vs 114 mo, respectively; P = 0.03). Patients with hemoglobin level below 10 g/dL and platelet count below 150 ×10⁹/L also had a significantly poorer OS (P= 0.01 and P= 0.02, respectively). As the majority of probes used in the MS-MPLA method detect methylation in only one CpG pair, the results of CDKN2B methylation were validated by MSP, which yielded the same methylation results than with MS-MPLA methodology. The 8 patients with methylated CDKN2B show a trend for a shorter survival than the remaining 32 with unmethylated CDKN2B (median 72 mo vs 114 mo, P = 0.08). The results of this study indicate that aberrant methylation of several tumor suppressor genes is observed in a substantial proportion of patients with RARS. As occurs in patients with high-risk MDS, our results suggest that aberrant gene methylation confers a poor prognosis in RARS, but these data and their potential independent value require confirmation in larger series that employ multivariate methods. Finally, these findings provide a strong rationale for the use of hypomethylating agents (e.g., azacitidine or decitabine) in patients with RARS.

This work has been partially supported by ISCIII grants RD06/0020/0031 and CA08/00141.