Where is the start line?

In this issue of Blood, Mielcarek and colleagues report on the results of a retrospective analysis of the outcomes of initial acute GVHD therapy.

Despite advances in the management of complications related to hematopoietic cell transplantation, treatment of acute graft-versus-host disease (GVHD) remains suboptimal. Corticosteroids are the primary front-line therapy for acute GVHD at a standard dose of 2 mg/kg per day prednisone, with response rates of about 50%.^1-3  In this issue of Blood, Mielcarek et al report on the outcomes of initial acute GVHD therapy among 773 transplantation patients at the Fred Hutchinson Cancer Research Center from 2000 to 2005.⁴  Patients were treated with either the standard dose of 2 mg/kg per day prednisone or low-dose prednisone (1 mg/kg per day) at the discretion of the attending physician. By day 100 after initiating therapy, patients treated in the low-dose group received a mean cumulative dose of 44 mg/kg compared with 87 mg/kg in the standard-dose group. Adjusted outcomes between the 2 groups were not statistically different. In multivariate analysis, treatment with low-dose steroids was associated with a reduction in prolonged hospitalization and a trend toward lower risk for invasive fungal infections. The authors conclude that initial treatment with low-dose prednisone did not compromise acute GVHD control or mortality and was associated with decreased toxicity.

This article addresses an important question regarding the preferred initial steroid dose for patients with acute GVHD. It is clearly desirable to use as low a dose of steroids as possible to reduce the risk of toxicity. However, this goal must be balanced with the need to attain early and durable control of acute GVHD. This single center large study provides implied uniformity of supportive care as a strength. However, it is important to note that these conclusions must be restricted to patients with grade I-II acute GVHD, as less than 3% of patients treated with low-dose prednisone had severe (grades III-IV) acute GVHD. In addition, this analysis was restricted to adults. Although age has not been a consistent factor in the response to acute GVHD therapy, future trials should include children.

Inherent in the retrospective design of this study is the presence of potentially confounding variables. The starting steroid dose was chosen at the discretion of the attending physician and as the authors discuss, there must certainly have been a bias toward giving the higher dose steroids to patients with more fulminant GVHD. This is suggested by the fact that only 2.6% of the low-dose steroid patients had severe GVHD compared with 16.3% patients in the standard-dose steroid group.

The 2 treatment groups also differed in a number of important clinical variables. A higher proportion of the low-dose patients had more frequent gut and only limited or no skin GVHD compared with the standard-dose group. The high incidence of gut GVHD in low-dose steroid recipients led to greater use of oral beclomethasone dipropionate, an agent that has been shown to incur a steroid sparing effect and survival advantage in an earlier study.⁵ 

Notably, despite being associated with less toxicity, the use of low-dose steroids was not associated with less transplantation-related mortality or overall mortality. Although relapse did not differ between all patients in the 2 steroid groups, in a subgroup analysis, risk of relapse was 1.5-fold higher in low-dose steroid recipients than in standard-dose recipients. The groups were balanced with respect to disease risk; however, future prospective trials are needed to determine the impact of steroid dose on this outcome.

In the race to improve outcomes in the treatment of acute GVHD, Mielcarek et al provide compelling evidence that, at least for adult patients with grades I-II acute GVHD, the start line may have been moved back to 1 mg/kg per day. Future prospective studies are warranted to determine the optimal starting dose and to carefully evaluate its consequences for both adults and children and especially for those with severe GVHD. In addition, further identification of patient and graft characteristics predictive of response will lead to a more tailored approach to acute GVHD therapy in the future.