Numerous studies have demonstrated a significant increase in apoptosis in patients having MDS using different methodologies.^7,8,9,10 The frequency and percentage of cells showing apoptosis in these patients vary depending on the methodology used and cell population analyzed. Apoptotic cell death can be measured by various methods including terminal dUTP nick-end labeling (TUNEL), caspase, annexin V and mitochondrial membrane potential. Each assay method measures a different component or event of apoptotic process. In this study, we used mitochondrial membrane potential measurement and annexin V assay as representative for early and advanced apoptosis.^11,12,13 Mitochondria has been shown to play a pivotal role in apoptosis in a number of ways, including the release of caspase activators, changes in electron transport and loss of mitochondrial transmembrane potential.^12,13 DePsipher is a dye that enters the mitochondria, polymerizes when the mitochondrial membrane potential is intact, and emits orange fluorescence; when the mitochondrial membrane potential is disturbed, the dye does not polymerize and emits green fluorescence.¹⁴ These two fluorescence colors can be analyzed by flow cytometry, and this assay can be used to detect early apoptosis. Annexin V allows identification of cell surface changes that occur during the apoptotic process. The binding of annexin V to phosphatidylserine on the damaged cell surface can be measured by flow cytometry. It is important to distinguish apoptotic cells from damaged cells as a result of non-apoptotic cell injury. In this study, we used PI in the four-color staining incubation mix to exclude cells that have lost membrane integrity, such as necrotic cells that may not reflect apoptotic process. Relatively lower estimates of apoptotic cells in this study compared to other studies that did not incorporated PI in the assay may be due to the exclusion of PI-positive cells in our annexin V assay. Furthermore, some studies⁷ used frozen samples and annexin V is significantly higher in frozen samples as compared with fresh samples.

Both mitochondrial potential and annexin V assays showed higher degree of apoptosis in RAEB-T as compared to AML. The possibility that the increase in apoptosis in MDS and RAEB-T cases could be due to increased percentage of mature and maturing cells should be considered and for this reason, we analyzed apoptosis in CD34⁺ cells separately from mature neutrophils. Higher apoptosis was observed in both CD34⁺ cells, as well as in CD34^- cells by annexin V assay. Similarly, higher apoptosis was present in both polymorphonuclear cells and mononuclear cells assessed by mitochondrial potential assay. Thus, apoptosis is not merely a function of the percentage of blasts in the bone marrow. Our findings are consistent with those of Raza et al^4,5 who demonstrated a high degree of apoptosis in bone marrow biopsies from patients with MDS.

We observed particularly increased apoptosis in CMML using mitochondrial potential. Overall, there was significant correlation between mitochondrial potential and annexin V in detecting apoptosis in CMML, as well as in all other subgroups. The disturbance in the mitochondrial potential is the earliest change in the apoptotic pathway. In contrast, annexin V positivity is a late phenomenon. The relatively higher apoptosis in CMML, which is detected by mitochondrial potential may reflect difference in the stage of apoptosis and how apoptotic cells are cleared.

Measurement of apoptosis is particularly influenced by the method of processing, time to processing, and other physiologic factors. Although we have made every effort to process samples in a similar fashion as soon as possible and many samples were analyzed in duplicate, we cannot rule out the possibility of variation due to the methodology of the assays. The high number of cases analyzed and the consistent results obtained using two different methodologies, support our conclusions.

Most published studies did not separate RAEB-T from other subgroups of MDS in their analysis. Specifically, one study that grouped RAEB-T with secondary AML (MDS-AML) showed a significantly lesser degree of apoptosis in patients having RAEB-T/MDS-AML compared with those having RA/RARS and RAEB.⁹ Our group recently reported a significantly higher degree of apoptosis in patients having RAEB-T compared with those having AML as assessed according to their caspase 3 activity (P = 0.0001).¹⁰ In that study, the caspase 3 activity in RAEB-T patients did not differ significantly from that in patients having other MDS. In addition, RAEB-T was distinguished from AML with respect to several laboratory and clinical parameters. In particular, RAEB-T patients had a higher proliferating rate and tend to have a lower platelet count and bone marrow cellularity as compared with AML patients.¹⁰

The data presented here support the concept that MDS disease is a disease of ineffective hematopoiesis and the reason for the ineffective hematopoiesis is increased cell death in bone marrow. In contrast, AML is a disease of proliferation of immature cells (blasts). Both diseases are clonal. In MDS, leukemic cells are capable of differentiating while in AML, usually leukemic cells do not differentiate. There is certainly some overlap between MDS and AML in the level of apoptosis. Patients classified as RAEB-T may include those having AML who were detected at an early stage, and similarly some AML patients with a blast count higher than 30% may have higher degree of apoptosis, therefore biologically closer to MDS. Our data suggest that perhaps neither 20% nor 30% blasts is a magic cut-off point to distinguish AML from MDS, but overall RAEB-T is biologically closer to MDS than AML. Irrespective of the cut-off point, therapy should take into consideration the biological and cellular abnormalities in AML and MDS and exploit these abnormalities either to kill the leukemic cells or correct the environment that allow these cells to dominate. Currently, outcome of RAEB-T patients may not be significantly different from that of AML, however, as we develop new therapies that specifically target the biological and molecular abnormalities in leukemic cells, we may start seeing differences between AML and MDS. Elimination of the RAEB-T category by lumping it in with AML does not achieve any goal and does not advance our understanding or management of AML nor MDS.