For many years, there were few attractive treatment options for patients diagnosed with a form of myelodysplastic syndrome (MDS) beyond supportive and palliative care or, for a small number of younger patients with good performance status and a suitable human leukocyte antigen (HLA)-matched donor, allogeneic hematopoietic stem cell transplantation.[1-4] Chemotherapy regimens similar to those administered to patients with acute myeloid leukemia (AML) are generally too toxic for routine use in the typical older patient with MDS (median age at diagnosis: ~65 years), and many patients never recover hematopoiesis after such aggressive cytoreductive therapy. Furthermore, MDS patients who achieve a complete hematologic remission after antileukemic chemotherapy usually relapse within a few months. A number of pilot studies were conducted in the 1980s and 1990s to try to improve this grim situation (eg, with retinoic acid, amifostine(Drug information on amifostine) [Ethyol]), but most such clinical trials showed little benefit.
In the past 3 years, the US Food and Drug Administration (FDA) has approved three disease-modifying therapies specifically for use in patients with MDS.[5,6] Additionally, several investigators have described their experiences in the MDS setting with drugs approved for other indications. While it is questionable whether a substantial improvement in the natural history of the disease has truly been achieved, it is clear that some patients benefit from these therapies.[5,7] The expanding repertoire of available treatments for MDS means that choosing the right therapy for a newly diagnosed patient is becoming increasingly challenging. Here, we discuss the agents most commonly used for MDS therapy at present, and outline a strategy for management based on an assessment of an individual patient's risk for cytopenia-related complications, disease progression, and death. For patients who are disinclined to pursue newer therapies or cannot afford them, good palliative and supportive care is still a reasonable option.
Accurate Assessment of Risk
The conditions grouped together as MDS are as diverse as the patients suffering from them. Some patients with this diagnosis progress rapidly to AML or die from cytopenia-related complications within just a few months of diagnosis, while others do well for many years with a simple strategy of observation and "watchful waiting." In view of this heterogeneity, predicting the risk the disease confers to an individual patient is critically important, especially since all of the therapies currently used for MDS carry the potential for serious adverse effects. Given that the only curative therapy for MDS remains stem cell transplantation, it is important to ensure that the adverse effects of any other proposed treatmentwhere the goal is not cureare unlikely to be greater than the morbidity of the disease itself. Unfortunately, the current prognostic methods for MDS are rather blunt tools, and accurate risk assessment remains a long way off.
Most clinicians have become comfortable using the simple 1997 International Prognostic Scoring System (IPSS) for assessment of the likelihood of AML transformation or death in de novo MDS (Table 1). The IPSS takes three key factors into accountthe proportion of undifferentiated blasts in the patient's blood and marrow (as measured by the manual aspirate differential, not by flow cytometry, which tends to overestimate blast numbers), the cytogenetic risk profile, and the number of cytopenias that the patient hasin order to come up with a four-category prognostic assessment. However, while the IPSS was an important step forward, it is not perfect: As illustrated in Figure 1, any given IPSS category is associated with a broad range of clinical outcomes.
The 2001 MDS classification system of the World Health Organization (WHO) divides MDS into eight categories (Table 3). Although intended as a clinicopathologic classification system rather than a rigorous prognostic tool, the WHO schema does help clinicians perform risk assessment. For instance, forms of MDS that are limited to erythroid dysplasia are now recognized as carrying a better prognosis than forms where marrow dysplasia involves multiple cell lineages; also, patients with excess blasts (> 5%, and especially > 10%) in their bone marrow have a worse outlook than those with a normal blast count. The best outcomes are for patients with classic 5q- syndrome, who enjoy a median survival on the order of 8 to 10 years. A recent proposal to integrate the IPSS with the WHO classification into a "World Health Organization Prognostic Scoring System (WPSS)" (Table 4 and Figure 2) is currently being independently evaluated.
It is important to note that the IPSS is based on 816 patients with de novo MDS. Patients with known secondary MDS (ie, prior exposure to chemotherapy with alkylating agents or topoisomerase inhibitors, therapeutic or accidental exposure to ionizing radiation, or both) have a very high risk of AML progression and should probably be considered similarly to the highest IPSS risk group. Additionally, there are many cytogenetic and molecular genetic risk categories that are not accounted for by the IPSS, because many of the less common MDS-associated karyotypes were represented by only a few patients in the 1997 analysis, resulting in inadequate statistical power to detect the difference between these small groups in outcome.[16-18] However, small cohort studies have suggested that some patients with certain uncommon but recurrent karyotypic changes (eg, abnormalities of chromosome 17p that involve the TP53 gene) do particularly poorly.[16,19,20] These factors need to be considered when deciding on a particular therapeutic course for an individual patient.