my personal edition > bone marrow and pbsc transplantation > news


  E-Mail this DGDispatch to a colleague

DGDispatch

Acute Graft Versus Host Disease Activity Index Can Be Used to Predict Non-Relapse Mortality: Presented at ASBMT

By Alan McClelland

HONOLULU, HI -- February 21, 2006 -- Researchers have developed an evidence based real-time scoring system for quantitative assessment of acute graft versus host disease (GvHD) activity.

According to George McDonald, MD, Fred Hutchinson Cancer Research Center in Seattle, Washington, United States, presented the acute GVHD activity index model at the 2006 Blood and Marrow Transplantation Tandem meetings (ASBMT).

Current scoring systems for acute GvHD are inadequate due to retrospective assignment of scores, failure to take efficacy of treatment into account and variability between observers, Dr. McDonald explained during his presentation on February 17^th.

An improved grading system would help determine prognosis in individual patients and quantify the burden of GvHD over time, he said.

Dr. McDonald and colleagues developed the system by examining and scoring the severity of GvHD symptoms and abnormalities, use of immunosuppressive drugs, and performance status in 386 patients with chronic myelogenous leukemia (CML) who underwent allogeneic hematopoietic stem cell transplantation.

The researchers scored patients at 10 day intervals from the onset of GvHD until day 100 and used data from 191 randomly selected patients to generate an activity index that they say predicts non-relapse mortality at day 200.

Parameters that they found contribute to the predictive value of the model include liver dysfunction (serum bilirubin levels), upper gastrointestinal tract abnormalities (reduced caloric intake, nausea, vomiting), immunosuppressive therapy and performance status.

The investigators tested the accuracy of the model in predicting day 200 non-relapse related mortality against an independent data set of 193 patients and obtained good predictive values comparable to those obtained with the initial data set used to develop the model (area under receiver operator curve of 0.87), Dr. McDonald said.

Advantages of the activity index are that it is evidence based, easy to calculate, can be used in real time and has good predictive value for non-relapse related mortality, he said.

"This score puts an objective number on a subjective clinical observation," Dr. McDonald said. As such, it should be a useful tool to help clinicians determine when to initiate more aggressive therapy for GvHD and might also be of use for stratification of patients according to risk of mortality in therapeutic trials, he added.


[Presentation title: Determining Prognosis for Patients with Acute GVHD in Real Time; Development and Testing of an Acute GVHD Activity Index. Abstract 29]

E-Mail this DGDispatch to a colleague

To print, use this version