Clinical Relevance of Minimal Residual Disease and Chimerism for the Detection of Relapse after Hematopoietic Stem Cell Transplantation

The clinical relevance of both minimal residual disease (MRD) and chimerism after hematopoietic stem cell transplantation (HSCT) for the detection of impending relapse has not yet been extensively studied. We investigated MRD in 119 consecutive children (median age, 12 years) with ALL (n=56), AML (n=36), MDS (n=20), or CML (n=7) who underwent bone marrow (n=80) or peripheral blood stem cell (n=39; T-cell depleted: n=18) transplantation in a single center. Thirteen patients received autologous HSCT and 106 patients underwent allogeneic HSCT. The donor was HLA-matched unrelated in 58 patients and HLA-identical related in 48 patients. The Wilms’ tumor gene (WT1) expression was used for the detection of MRD because WT1 gene is overexpressed in the vast majority of patients with leukemia. In contrast, WT1 gene is not expressed in normal peripheral blood mononuclear cells (PBMCs) and only weakly expressed in normal bone marrow (BM) cells. We performed a quantitative reverse transcriptase-polymerase chain reaction (PCR) to examine the level of WT1 gene expression. For the present study, we followed up MRD in both BM and PB after HSCT. In the 119 patients, a median of 11 analyses (range, 3–27 analyses) covering a median period of 708 days (range, 29–3101 days) was performed. Analysis of 25 paired BM and PB samples revealed that the level of MRD in BM was on average 9 times higher and paralleled that in PB. All 85 patients with continuous normal WT1 expression levels in BM and continuous undetectable WT1 expression levels in PB remained in complete remission without relapse at a median of 1884 days (range, 58–5905 days) after HSCT. In contrast, all 32 patients who suffered from hematological relapse after a median of 152 days (range, 28–1104 days) presented with high levels of WT1 gene expression in BM and PB (P < .001). In 19 patients, we observed a gradual or rapid increase of WT1 expression levels at a median of 31 days (range, 12–119 days) before hematological relapse. In 14 of these 19 patients, DNA was available at the time of increase of WT1 expression level to perform analysis of hematopoietic chimerism using a semi-quantitative short-tandem-repeat PCR. Interestingly, 10 of the 14 patients (71%) revealed a complete donor chimerism, whereas only 4 of the 14 patients (29%) showed a mixed chimerism. In two patients, we diagnosed a molecular relapse using WT1 gene expression at 161 and 360 days after transplantation. At the time of molecular relapse, both children revealed a complete donor chimerism. In both patients, molecular remission was achieved by withdrawal of cyclosporin A and by donor lymphocyte transfusion, respectively. Both children are alive and well without relapse at 10 and 9 years after transplantation. In conclusion, quantitative analysis of WT1 gene expression is a valuable tool for monitoring of MRD in patients with ALL, AML, CML, and MDS after HSCT. Increasing levels of WT1 gene expression strongly predict hematological relapse. Therefore, this approach is very useful for early diagnosis and treatment of molecular relapse after HSCT. MRD measurement using WT1 gene expression is more sensitive for the detection of impending relapse than the analysis of hematopoietic chimerism.