Early prediction of relapse after allogenic transplant for acute myelogenous leukemia

Shaji Kumar, Michael G. Chen, Dennis A. Gastineau, Morie A. Gertz, David J. Inwards, Martha Q. Lacy, Ayalew Tefferi, Mark R. Litzow

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Background: Allogenic BMT is potentially curative in patients with relapsed AML and in those with poor prognostic factors. However, relapse is an important cause of transplant failure with nearly 20 -30% relapsing within a year.There are no reliable markers to predict relapse. Lymphocyte recovery at four weeks post-transplant has been shown to predict relapse in patients undergoing BMT for AML. Also, patients who develop acute GVHD are less likely to relapse after allogenic BMT.Hypothesis: Graft-versusLeukemia effect is more likely to be active against minimal residual disease during the early post-transplant period. Hence, a scoring system taking into consideration time to development of acute GVHD and lymphocyte recovery at 30 days post-transplant will enable early prediction of relapse. Methods and materials: A retrospective review of patients with AML who underwent matched related allogenic BMT at our institution between 1982 and 1999 was done to examine this hypothesis. Eighty-five transplants done over the 18-year period were examined. Only patients surviving up to 30 days post transplant were included for analysis. Two risk factors were used to form a scoring system, absolute lymphocyte count (ALC) at day 30 post-transplant <175/nL and absence of the development of any grade of acute GVHD by day 30. Patients were divided into three groups depending on the presence of these risk factors. Group I had both the risk factors present, group II had either one present and group III had neither of the risk factors present. Results: Twenty-four relapses occurred among the 74 eligible transplants. There were 7, 28 and 39 patients each in group I, II and III respectively. 4,10 and 10 patients each in the groups I, II and III relapsed. Patients in group I were 4.61 (95% CI, 1.41, 15.05) times more likely to relapse compared to those on group III, and those in group II were 1.50 (95% CI, 0.62, 3.60) times more likely to relapse. The relapse free survival as estimated by the Kaplan Meier product limit method demonstrated significant difference in the relapse survival between the three groups. The log rank test for comparison of the curves yielded a p value of 0.024.Conclusion:We propose that the above criteria be used to identify patients at high risk for relapse after allogenic BMT for AML. Patients who have a poor lymphocyte recovery (< 175/iL ALC) at day 30, and those who fail to develop acute GVHD by day 30 should have immunosuppression tapered rapidly and should be considered for other measures for prevention of relapse including donor lymphocyte infusion. These results should be validated using larger patient groups.

Original languageEnglish (US)
Pages (from-to)197a
Issue number11 PART I
StatePublished - 2000

ASJC Scopus subject areas

  • Biochemistry
  • Immunology
  • Hematology
  • Cell Biology

Fingerprint Dive into the research topics of 'Early prediction of relapse after allogenic transplant for acute myelogenous leukemia'. Together they form a unique fingerprint.

Cite this