Poor-Risk Acute Leukemia Patients with An EBMT Low-Risk Score and An 8/8 Matched Unrelated Donor Show Excellent Survival After Hematopoietic Stem Cell Transplantation.

Allogeneic hematopoietic stem cell transplantation (alloSCT) from volunteer unrelated donors (URD) may be associated with a higher non-relapse mortality (NRM) and worse outcome as compared to alloSCT using HLA-identical sibling donors. However, many parameters next to donor type define NRM. The impact on outcome of allele-matching for HLA-A, -B, -C and -DRB1 between donor and recipient has clearly been demonstrated. The prognostic impact of the EBMT risk score, that takes into account age, stage of disease, time from diagnosis to transplantation, donor type and donor-recipient gender combination, has recently been validated in a variety of hematological malignancies including acute leukemia and myelodysplastic syndrome (MDS). We evaluated the relative prognostic value of high-resolution HLA matching and the EBMT risk score for patients with poor-risk acute leukemia and MDS who received an URD transplant.

Between 1987 and 2006, 327 patients (≥16y) with poor-risk acute leukemia and MDS underwent URD alloSCT in the Netherlands. Patients were in 1^st complete remission (CR1, n=129), 2^nd CR (CR2, n=91), beyond CR2 or not in remission (n=107). The leukemia-risk was considered to be poor if patients had adverse cytogenetics or were not in CR1. The majority of the grafts was T-cell depleted (94%). High-resolution typing of HLA-A, -B, -C, and -DRB1 alleles was available for analysis in 270 donor-recipient pairs and had in part been performed retrospectively.

We evaluated the impact of high-resolution matching for HLA-A, -B, -C and -DRB1 on progression free survival (PFS) and overall survival (OS). Patients who were fully matched (8/8) with their donors (n=170) hadsignificantly superior PFS (40+/-4% vs 26+/-5%, hazard ratio (HR)=0.68; 95%CI 0.50–0.92, p=0.01) and OS (39+/-4% vs 29+/-5%, HR=0.70; 95%CI 0.51-0.96, p=0.03), compared to patients with mismatched (≤7/8) donors (n=100). Superior OS in the 8/8 group appeared to be due to a lower NRM (24+/-4% vs 39+/-5%, HR=0.54; 95%CI 0.35-0.85, p=0.008), while the relapse mortality rate was identical in both groups (37+/-4% vs 32+/-5%). Patients with EBMT risk scores of 1-2 (n=71), 3 (n=77), 4 (n=76) and 5-7 (n=103) had a predicted 5 year OS of 52%, 41% (HR=1.57; 95%CI 0.98-2.52), 29% (HR=2.07; 95%CI 1.32-3.26) and 19% (HR=2.69; 95%CI 1.76-4.11), respectively (p<0.001). Relapse mortality rate and NRM increased with increasing EBMT risk score. As shown in the table, the impact of allele-matching on OS was most evident in the EBMT low-risk group. EBMT low-risk (1-2) patients with 8/8 donors showed excellent 5 year OS compared to EBMT low-risk patients with ≤7/8 donors (73+/-8% vs 35+/-12%). The favorable impact of a fully matched donor was absent in patients with higher EBMT risk scores.

Both the EBMT risk score and the degree of allele-matching independently predicted outcome after URD alloSCT. The predictive value of allele-matching was especially evident in EBMT low-risk patients, while patients with the highest EBMT risk scores (>4) had a dismal outcome, despite allele-matching. These results emphasize the importance of incorporating age, disease stage, donor-recipient gender combination and time interval from diagnosis to transplantation (EBMT risk score parameters) as well as high-resolution HLA-typing in the risk assessment prior to URD alloSCT. As excellent OS was noted in well matched EBMT low-risk patients, our data underscore the importance of an immediate search for an unrelated donor in poor-risk leukemia patients in CR1 below the age of 40, who should then receive their alloSCT as early consolidation therapy following induction chemotherapy.

No relevant conflicts of interest to declare.