Does cytomegalovirus viral load in stem-cell transplant recipients matter?

Abstract

The question of cytomegalovirus (CMV) viral load on outcome after allogeneic stem-cell transplantation has been discussed since the development of quantitative techniques for assessing CMV DNA more than 15 years ago. Emery and colleagues did groundbreaking work in this area that was published in 2000 showing that initial viral load and the rate of viral load increase were predictive of subsequent development of CMV disease in diff erent types of transplant recipients. However, most patients were not given pre-emptive antiviral therapy. Ljungman and colleagues studied recipients of stem-cell transplants who were receiving PCR-based pre-emptive antiviral therapy and were unable to show an eff ect of pretreatment viral load on the risk for CMV disease but that the rate of decrease in viral load predicted development of CMV disease during later CMV replication episodes. Other studies have also addressed this question, providing support for viral load on outcome development of CMV disease, but data for hard endpoints such as survival are lacking. Why is this question important? New drugs against CMV are urgently needed because presently licensed antiviral drugs have important side-eff ects. Until now, regulatory authorities have required hard clinical endpoints in pivotal studies such as survival or development of CMV disease. However, due to developments in the fi eld of CMV management (eg, in pre-emptive therapy) the frequency of these complications has decreased to levels that make it very diffi cult to do such trials because of the large number of patients required. Establishment of a surrogate endpoint such as viral load for studies of CMV would therefore be of major importance, such as is the case for HIV and HCV. A report in The Lancet Haematology by Margaret Green and colleagues provides important information in this regard. First, they used a standardised assay using the international standard to analyse their material. Second, they noted a clear viral-load dependent negative eff ect on outcome measured by hard endpoints such as overall and non-relapse mortality. Third, they identifi ed that commencement of antiviral therapy when CMV replication has started might be too late, because even low viral loads had a negative eff ect on outcome. The fi ndings of this study therefore support the use of viral load as an important study endpoint in future clinical trials and that prevention of CMV replication rather than pre-emptive therapy should be the goal when managing recipients of stem-cell transplants. The limitation of this study is its retrospective design. However, prospective studies of CMV prophylaxis are ongoing and verifi cation of the fi ndings from Green and colleagues’ retrospective study might be possible through analysis of the data from these studies.