Treatment of Minimal Residual Disease (MRD) with Donor Lymphocyte Infusions (DLI) in Acute Leukemia Patients Undergoing an Allogeneic Hemopoietic Stem Cell Transplants (HSCT).

Background. Hematologic relapse after an allogeneic HSCT for acute leukaemia is associated with poor outcome: in our Unit, 5 year survival for acute lymphoblastic leukaemia (ALL) relapsing after an allograft is 10% and for acute myeloid leukaemia (AML) this figure is 15%. Most relapses (79%) occur within the first year after transplant.

Aim of the study. Prospectively assess minimal residual disease (MRD) after allogeneic HSCT to test (a) incidence of MRD positivity, (b) predictive value on hematologic relapse and (c) whether MRD positivity can be used to initiate treatment with donor lymphocyte infusions (DLI).

Patients and methods. Seventy three patients with ALL (n=29) or AML (n=44) received an allogeneic HSCT, most of them after a conventional (CY-TBI) conditioning regimen. Re-arrangement for JH regions and fusion genes (bcr/abl) (RQ-PCR) were used to monitor MRD in ALL, whereas Wilms Tumour 1 (WT1) expression (RQ-PCR) was used in AML. Monitoring was done on bone marrow samples taken pre-transplant, and on day+30, +60, +90, +120, +150, +180. At the time of a positive MRD patients would be eligible for DLI in escalating doses, starting at 10^6/kg CD3+ cells for matched related donors, increasing by ½ log every 30 days up to a maximum dose of 5x10^7/kg CD3+ cells. For unrelated and 1 antigen family mismatched transplants the dose of DLI was reduced by 1 log.

Number of MRD+ patients post-transplant. Overall 30 patients of 73 patients (41%) were MRD positive after transplant, 66% for ALL and 25% for AML.

Predictive value of MRD positivity on relapse. Hematologic relapse occurred in 15 of 30 MRD+ patients and in 2 of 43 MRD negative patients (p=0.0001). For ALL patients these figures were 10/19 vs 2/10 (p=0.09) and for AML it was 5/11 vs 0/33 (p=0.0004).

Preventing relapse with DLI in MRD+ patients. Of the 30 MRD positive patients 16 were given no DLI because the donor was not available or unsuitable for second donation at the time of MRD positivity: 14 patients relapsed (87%) and 3 (19%) survive. Fourteen patients were given DLI, one relapsed (7%) and 12 survive (86%). The difference of hematologic relapse in MRD positive patients receiving DLI (8%) or not receiving DLI (57%) is highly significant (p=0.0001). This particularly true for ALL patients receiving DLI (n=10) (1 relapse and 8 survivors) as compared to ALL patients not receiving DLI (n=9) (9 relapses and 1 survivor).

Conclusions. This study shows that (a) MRD positivity is frequent after allogeneic HSCT, (b) MRD positivity predicts hematologic relapse and (c) treatment of MRD with DLI protects against hematologic relapse, and improves survival. This study also confirms our previous data on the efficacy of DLI in acute lymphoblastic leukemia, when the tumour burden is low.