Introduction and context

The myelodysplastic syndromes (MDSs) are a group of clonal hematological stem cell disorders characterized by defects in maturation resulting in cytopenias and a varied risk of transformation to acute myeloid leukemia (AML). Approximately 7,000-12,000 cases are diagnosed annually in the United States [1]. The majority of patients with MDS are older than 50 and the prognosis is quite variable, with some patients never requiring any intervention, and others needing treatment similar to patients with acute leukemia [2,3]. For many years, the French-American-British (FAB) classification distinguished five types of MDS [4]. The World Health Organization (WHO) proposed revisions in 1999 that included changing the definition of AML from 30% blasts to 20%, recognizing that the percentage of blasts is prognostic, stating that unilineage and multilineage dysplasia are separate entities, and recognizing the 5q^- syndrome [5]. The WHO system has again been recently revised; there are now eight types of MDS (Table 1).

While classification systems aid in diagnosis and communication, they are less helpful for predicting prognosis, which is essential for a disease with multiple therapeutic options. The International Prognostic Scoring System (IPSS) was based on information from 816 patients with MDS. In the IPSS, the number of cytopenias, percentage of blasts in the bone marrow, and cytogenetic abnormalities were used to define four risk groups (low, intermediate-1, intermediate-2, and high), which predicted overall survival and risk of AML transformation (Table 2) [6]. The IPSS has subsequently become the most widely used method to predict prognosis and assist in deciding which patients should receive aggressive treatment.

Clinical manifestations of MDS are usually related to cytopenias. The goals of treatment are to alleviate symptoms, improve quality of life, and prolong survival. Therapeutic options range from supportive care to aggressive therapy. Patients who are elderly or with significant comorbidities are usually supported with transfusions, antibiotics, and growth factors. Recently, three new therapies have been approved for treatment. Lenalidomide, an analog of thalidomide, showed a remarkable response in initial trials in patients with the 5q^- syndrome [7]. Studies in low-risk patients without loss of 5q have also shown a significant response [8,9]. For higher risk patients, 5-azacitidine and decitabine, which act through DNA demethylation, have been approved for use. Both are given safely to outpatients, but can induce significant cytopenias, especially during the first few cycles [10-13]. Even with these new agents, hematopoietic stem cell transplantation (HSCT) remains the only curative therapy for MDS. However, HSCT historically possesses significant morbidity and mortality, and given the heterogeneous prognosis of MDS, it was not routinely offered to many patients.

Recent advances

Traditionally, only patients younger than 50 years of age with an HLA-matched donor were eligible for HSCT, however, improvements in prognostication, reduced intensity conditioning regimens, and experience with alternative donors have allowed physicians to consider many formerly ineligible patients as potential candidates for HSCT. Using registry data, investigators attempted to define the best time to perform HSCT for patients with MDS. Their findings showed that for patients with low and intermediate-1 IPSS risk disease, a strategy of delayed transplantation led to improved survival with the optimal time for HSCT being the development of a new cytogenetic abnormality, the appearance of a cytopenia, or the progression to a higher risk IPSS group. For patients with intermediate-2 and high IPSS disease, immediate transplantation led to maximal overall survival [14].

Reduced intensity conditioning (RIC) regimens were developed in the hope of decreasing acute transplant-related mortality (TRM) while allowing patients with advanced age or comorbidities to undergo HSCT. The therapeutic mechanism of RIC-HSCT is an immunological graft-versus-malignancy effect. Several different RIC regimens have been published showing reasonable toxicity in older patients [15-17]. There have also been retrospective analyses comparing myeloablative and RIC regimens in patients with MDS/AML with results suggesting that while TRM is indeed lower with RIC, relapse rates are higher, and overall and disease-free survival appear equivalent [18-21]. To date, no prospective randomized trials comparing conventional and reduced intensity transplants have been performed, however, it appears RIC-HSCT can offer approximately a 30-40% chance of long-term remission with an associated TRM between 15-40% for patients with MDS.

While the optimal HSCT donor remains a matched relative, alternative donors, such as unrelated donors (URDs) and umbilical cord blood (UCB), have become reasonable options. With the emergence of DNA-based HLA typing methods, outcomes of URD transplants have significantly improved, mainly due to decreased TRM. A few series of URD-HSCT in MDS have illustrated encouraging results that approach the outcomes of those with related donors [22-25]. In many centers, if a related or unrelated donor is not available, UCB has emerged as the preferred donor source. Several encouraging series of UCB-HSCT have been reported, including our own series, which shows a 14% 100-day TRM and a 67% 1-year disease-free survival in various hematological malignancies [26-29]. However, the experience with UCB-HSCT is early, and we eagerly await longer follow-up.

Implications for clinical practice

Even with the introduction of several new therapies, allogeneic HSCT remains the only curative treatment for patients with MDS. Historically, the majority of patients with MDS were not offered HSCT. Several recent advances have allowed physicians to consider HSCT for an increasing number of patients with MDS. Improved prognostic algorithms and analyses have allowed better prediction of disease outlook to permit aggressive therapies to be tailored to high-risk patients. Recognition of the graft-versus-malignancy effect in MDS and introduction of reduced intensity regimens have enabled many older patients to be considered for HSCT. In addition, the growing experience with unrelated donors and the use of umbilical cord blood has obviated the requirement for a matched sibling. Even with this progress, questions still remain. Future prognostic systems will no doubt incorporate genetic and molecular data in combination with the clinical information used in the IPSS. Moreover, more experience with reduced intensity regimens and alternative donors needs to be gained before they become standard of care.