Background. Hematologic relapse is seen in 20–40% of adult acute lymphoblastic leukemia (ALL) patients undergoing an allogeneic hematopoietic stem cell transplant (HSCT) and is usually non responsive to donor lymphocyte infusions (DLI). This has suggested a lack of graft vs leukemia effect (GvL) in ALL patients. It is currently possible to monitor minimal residual disease (MRD) post-HSCT and this may allow to identify patients at high risk of hematological relapse.

Objectives. To monitor MRD in B-ALL patients post-HSCT and to treat MRD positive patients with DLI.

Patients and Methods. MRD was evaluated on bone marrow samples using nested polymerase chain reaction (PCR) and real time PCR (RT-PCR) assay to analyze respectively IgH VDJ and BCR-ABL gene rearrangement. Molecular positivity/relapse was defined with two consecutive positive PCR assays. MRD monitoring was performed in 28 patients grafted from an HLA-identical sibling (n=19), family mismatched related donor (n=2) or matched unrelated donor (n=7). Median follow-up was 47 months (range 5–163). Sixteen patients (57%) were in early disease phase at the time of transplant.

Results. We identified 3 groups of patients. A) 14 patients (50%) had no evidence of MRD after HSCT. B) 6 (21%) had a positive MRD and received escalating dose DLI, within 60 days and C) 8 (29%) had a positive MRD, but did not received DLI, because not available or because of an early hematological relapse. Median time from HSCT to molecular relapse was 139 days (range 46–1048). The median follow up was 1385 days (144 – 4877). Median number of infused CD3+ cells was 0.6x10^7/Kg of recipient body weight (range 0.01–7). Hematological relapse was seen in 0%, 0% and 88% respectively of group A,B,C (p=0.0001) and disease-free survival (DFS) was 100% , 67% , 12% (p=0.0001).

Conclusions. This study shows that (1) MRD monitoring of ALL patients post-transplant identifies patients at high risk of hematologic relapse and (2) that treatment with DLI, on the basis of MRD positivity, significantly reduces the risk of leukemia relapse and may improve DFS. These data also confirm the existence of a GvL effect, previously shown in this disease (

Weiden et.al
NEJM
,
300
,
1068
;
1979
).

Author notes

Corresponding author

Sign in via your Institution