Molecular monitoring of engraftment and leukemia relapse after allogeneic hematopoietic stem cell transplantation

Abstract: Relapse of the underlying disease, graft-versus-host disease (GVHD) and infections are the main complications after allogeneic hematopoietic stem cell transplantation (HSCT). A lower tumor burden increases the likelihood of cure after both chemotherapy and HSCT. The term minimal residual disease (MRD) refers to detection of leukemic cells below the threshold of standard morphological assessment. The aims of this thesis were to evaluate whether mixed chimerism (MC) analysis - i.e., the detection of recipient-derived hematopoietic cells - was sensitive and specific enough for early prediction of leukemia- relapse after HSCT. We also wanted to evaluate the effect of MC on GVHD incidence after HSCT and engraftment kinetics after less intense conditioning regimens. By using cell separation, we could demonstrate an increased sensitivity of chimerism analysis in patients with acute leukemia of almost two orders of magnitude. Twelve patients having pre-B-ALL were prospectively analyzed with MC analysis particularly focused on the B-cell lineage combined with MRD analysis by patient-specific leukemia rearrangement analysis to ensure a relevant level of detection of MC. Four of five patients with MC, detected in the leukemia- affected cell lineage, relapsed. In contrast, none of the seven patients with donor chimerism (DC) relapsed (p=0.01). Mixed chimerism was detected between 2.5 and 4 months before overt relapse in three patients who relapsed. In all five patients with MC in the B-cell lineage, the leukemia-specific clone was detected in the corresponding samples. We also prospectively evaluated the clinical impact of MC in the leukemia-affected cell lineage in 30 patients with AML and MDS after HSCT. Twelve patients relapsed after HSCT. Mixed chimerism in the leukemia-affected cell lineage was detected in 14 patients of whom 10 relapsed vs. 2/16 DC patients (p<0.01). The four patients with MC and continuous complete remission showed only MC in bone marrow (BM). By contrast, all 8 patients with MC detected in peripheral blood relapsed vs. 4/22 DC patients (p<0.001). In this study, MC was detected median 66 (23-332) days before hernatological relapse. Six patients with B-CLL were studied to see whether leukemic-derived recipient cells were present in the B- cell fraction after HSCT. To show whether detectable recipient B-cells belonged to the leukemic clone, we compared the findings of chimerism analysis to those of leukemia-specific V(D)J rearrangement analysis. All patients were MRD positive after HSCT. Patients with a high tumor burden before grafting had the most persistent (range 15-38 months) MRD after HSCT. Detection of MC in the leukemia-affected B-cell lineage correlated with detection of the leukemia-specific clone in all (25/25) samples analyzed. This suggests that MC analysis of the B-cell lineage is sufficient for monitoring patients with B-CLL after HSCT. A durable GVL response is probably needed for this patient group. In 102 patients, those with T-cell MC had 5% cumulative incidence of acute GVHD grades II-IV compared to 52% for T-cell DC patients (p<0.001). Patient characteristics were similar in the T-cell MC and DC groups. In multivariate analysis T-cell DC was the most significant risk factor for acute GVHD grades II-IV. We analyzed the engraftment kinetics after non-myeloablative HSCT in 30 patients. The various preparative regimens were well tolerated and hernatological recovery was rapid. All patients engrafted and showed various degrees of MC after HSCT. At the time of acute GVHD, 18/22 patients had detectable T- cell MC and 6/15 at time of disease response. All patients became T-cell DC after onset of acute GVHD. The development of DC varied between the various cell lineages with T-cell DC lagging behind myeloid and B-cell DC. High recipient T-cell levels (>50%) one month after HSCT were associated with impending graft rejection. Therefore, alloreactive responses may occur after nonmyeloablative HSCT despite T-cell MC. In conclusion, this thesis shows the feasibility of lineage-specific chimerism analysis for early prediction of relapse, monitoring of engraftment kinetics and GVHD after HSCT.