standard test for what remains a clinical dilemma, distinguishing AA from hMDS. This is in part, we believe, because CD34 numbers, although statistically different between the disease groups, showed considerable overlap in several studies. Rather than attempting to 'reinvent the wheel,' we believe that our results build upon past studies by providing clinical outcomes that improve the ability to classify patients with either disease state. When doing this, we find no overlap in marrow CD34 numbers between AA and hMDS, and thus believe quantifying marrow CD34 numbers should be further tested as a standard approach for separating the two disorders. Furthermore, these findings also suggest that patients with normal cytogenetics and low CD34 þ percentages are at a low risk of disease progression and may be managed in a conservative manner in the absence of life-threatening cytopenias. WH

Although aberrant promoter hypermethylation of O6-methylguanine-DNA methyltransferase (MGMT) is a favorable prognostic marker in patients with diffuse large B-cell lymphoma (DLBCL), MGMT protein expression has not been thoroughly examined. The aim of this study was to evaluate the clinical implication of MGMT protein expression and its correlation with promoter hypermethylation of the gene. We investigated MGMT protein expression by immunohistochemical analysis of 63 DLBCL patients who received cyclophosphamide as part of multidrug regimens. In addition, promoter methylation of the MGMT gene was analyzed by a methylation-specific polymerase chain reaction assay, and correlations with chemotherapeutic effect and prognosis were statistically evaluated. Immunohistochemical assay results for MGMT protein were negative in 30.2% of patients with newly diagnosed DLBCL. Immunostaining results were closely correlated with the methylation status of the promoter. Promoter DNA methylation of the gene was not detected in 34 (81.0%) of 42 tumor samples determined to be MGMT-positive DLBCL by immunostaining and was detected in 15 (88.2%) of 17 cases of MGMT-negative DLBCL. Overall survival (OS) and disease-free survival (DFS) rates were significantly higher in MGMT-negative patients than in MGMT-positive patients (5-year OS, 81.3% versus 56.6% [P = .0375]; 5-year DFS, 66.3% versus 39.9% [P = .0121]). The combined rate for complete response (CR) plus unconfirmed CR was significantly higher in MGMT-negative patients (15/19, 79.0%) than in MGMT-positive patients (25/44, 56.8%) (P = .0488). A multivariate analysis showed that absence of MGMT expression was an independent prognostic factor for OS (relative risk, 4.09; P = .0258). Lack of MGMT protein expression is associated with aberrant promoter DNA methylation and appears to be a useful marker for predicting the survival of DLBCL patients.

P53 mutation has been reported in various solid tumors, acute leukemia and myelodysplastic syndrome (MDS), but the diagnostic significance of p53 in MDS remains to be determined. The purpose of the present paper was to examine p53 mutation and immunostaining of the same patients, because there have been few reports of simultaneous analysis of these markers. Seven p53 mutations were observed among 37 MDS and 11 cases of overt leukemia transformed from MDS (MDS-OL). Mutated p53 mainly observed in high-risk MDS had more intense p53 staining than in MDS with wild-type p53 overexpression. Aplastic anemia (AA) produced no p53 staining. The percentage of p53 staining in MDS (71%) was higher than that of mutated p53 (11%) but did not reach 100% of MDS cases studied, therefore the authors attempted to differentiate MDS, especially refractory anemia (RA) and AA, using a combination of p53 immunostaining, hemoglobin F (HbF) immunostaining and chromosome abnormality, because HbF of erythroblasts was reportedly observed in MDS RA but not in AA. Most MDS/MDS-OL (47/48) had at least one positive marker. Among 11 AA cases, only two were positive for HbF. The present results suggest that the combination of these three markers is useful to discriminate MDS from AA.