Infections are a significant cause of morbidity and mortality in patients with MDS accounting for approximately 38% of all deaths in untreated cases, more so than that resulting from MDS progression to acute myeloid leukemia (AML). Although there is accumulating data on the risk factors and types of infections in MDS patients, interpretation of the available data has been challenging. This stems from several reasons and includes data reporting from different sources, year of study, and use of novel treatment regimens in the last decade, and transplant-related immune defects in those undergoing bone marrow transplant (BMT).
Most of the infection data in MDS comes from retrospective studies that report findings based on review of patient’s medical records and hence, are often prone to selection or information bias and inaccurate record keeping leading to erroneous inclusion or exclusion of cases or misclassification of infections. The increasing use of DNA hypomethylating drugs (5-azacytidine and decitabine) and lenalidomide in recent years for treating MDS has affected the natural history of the disease but in doing so has likely added to the risk of infections in these patients by transiently worsening the pre-existing low white blood cell (WBC) counts.
However, this increased risk of infections with newer therapies is still substantially lower when compared to the era when all the higher-risk MDS patients would be treated like AML patients with aggressive chemotherapy, commonly resulting in profound suppression of immunity. The situation is even worse in higher-risk MDS patients undergoing BMT as these patients due to their need for continued immune suppression to prevent graft rejection puts them at a very high risk for infections. Hence, cautious interpretation is warranted when reviewing data from pre- and post-DNA hypomethylating era and in the context of BMT.
While qualitative and quantitative defects in neutrophils (a kind of WBC) is considered to be the most important predisposing factor to infections, several other immune defects involving B cells, T cells and NK cells have been described. A wide spectrum of infections including bacterial, fungal, and viral infections, as well as atypical infections has been reported in MDS patients but by far, bacterial infections are the most common. However, in only 30% of these patients, a proper laboratory identification of causative pathogen is possible. Elderly MDS patients with coexisting medical conditions such as diabetes, respiratory conditions and liver disease and particularly those with greater transfusion needs are most vulnerable to infections. Available data from two randomized clinical trials are unclear about the risks of infections with use of hypomethylating therapy – while azacytidine group had lower rates of neutropenic (low neutrophils) fevers when compared to the patients on supportive care or low-dose cytarabine, those receiving decitabine had higher infection rates compared to those on supportive care. Contrary to prior findings, a recent retrospective series reported transfusion requirements prior to starting azacytidine therapy and platelet counts < 20,000/μl to be significantly associated with occurrence of infections whereas advancing age and neutropenia did not increase risk.
Lenalidomide therapy can be a double-edged sword. While profound neutropenia seen in approximately two-third of patients on lenalidomide predicted for eventual response to treatment, septic deaths associated with treatment-related neutropenia have been reported. The risk and severity of infections in MDS patients is even greater after allogeneic BMT, mostly due to the relative older age of patients, prior pre-transplant infections, and frequent use of antimicrobials leading to emergence of drug resistance and iron overload from chronic transfusions. One retrospective study showed an association between pre-transplant neutropenia and increased 3-year mortality from gram-positive bacterial and fungal infections in transplant patients.
Currently, no evidence-based guidelines exist for preventing or treating infections in MDS patients. Efforts at improving the neutrophil counts have been tried in neutropenic MDS patients with growth factors such as G-CSF or GM-CSF with results inconclusive at best. G-CSF or GM-CSF decreased the duration of neutropenia but had no significant effect on infection rate or overall survival even when tested in patients receiving chemotherapy. One study, in contrast, showed higher infection rates with use of GM-CSF. Similarly, the role of anti-bacterial and anti-fungal prophylaxis in MDS patients remains largely unknown due to lack of data. Iron lowering agents too have failed to show a reduction in the risk of infections.
At present, neutropenic fevers in MDS patients are mostly managed empirically with intravenous broad-spectrum antimicrobials, with choice largely determined by local patterns of infection and drug resistance, previous history of infections, and severity of the patient’s co-existing medical conditions. In this situation, treatment duration should be limited to the time when the counts are at the lowest. In select situations, use of G-CSF or GM-CSF can be justified in neutropenic patients with severe infections but should generally be avoided in those MDS patients with excess blasts (> 5% immature blood cells in bone marrow) due to concerns for progression to AML.