Is Next-Generation Sequencing the way to go for Residual Disease Monitoring in Acute Lymphoblastic Leukemia?
Haematopoietic stem cell transplantation (HSCT) in the therapy of paediatric acute lymphoblastic leukaemia (ALL) is reserved for the patients who have failed conventional chemotherapy. A high-risk disease (eg T-cell immunophenotype, presence of t(9;22) or t(4;11)), poor response to the initial treatment and relapse are the most common indications for HSCT. Despite the intensity of HSCT treatment, relapse of the disease occurs in approximately 40% of patients. The recurrence of the disease and transplant-related mortality (TRM) are the major factors with a negative influence on the outcome after HSCT in paediatric patients with ALL. A high emphasis is placed on searching for prognostic markers to distinguish the patients who will benefit from HSCT from those who will profit from different therapeutic modalities or further modifications of HSCT. Three major reports have recently presented the importance of minimal residual disease (MRD) level prior to HSCT as a prognostic marker in paediatric ALL patients. Knechtli et al have used a semiquantitative PCR and oligoprobe blotting-based analysis of T-cell receptor (TCR) and immunoglobulin (Ig) gene rearrangements to categorise MRD levels 1–11 weeks before HSCT. In this study, patients were divided into three categories according to the MRD level prior to HSCT: negative, lowand high-level positive. A similar approach was used in the study published by Bader et al. In the report published by van der Velden et al, a real-time quantitative PCR (RQ-PCR) approach was used to assess the level of MRD. Indisputable influence of the MRD level prior to HSCT on the outcome was presented in each report. An international Pre-BMT MRD Study Group was formed as a part of European Study Group on MRD in ALL (ESG-MRDALL) to discuss the possibilities of a combined clinical trial with stratification based on the pre-HSCTMRD levels, and therapy modifications for patients in particular MRD groups. This group involves centres from the Czech Republic, Denmark, Germany, the Netherlands and the UK. In addition, the group aimed to standardise laboratory techniques. The Pre-BMT MRD Study Group aimed to use the RQ-PCR approach to identify leukaemic cells by clonal rearrangements of Ig and TCR genes. Standardised techniques together with quality control rounds ensure identical procedures and comparable results among all participating centres. We present here an analysis of the data from the three above-mentioned studies together with unpublished data of the Pre-BMT MRD Study Group. The cohort consisted of a total of 140 patients who underwent HSCT between December 1988 and March 2002 in one of the following centres: University Hospital Motol, Prague, Czech Republic; University Children’s Hospital, Tübingen, Germany; University Medical Centres, Leiden and Utrecht, the Netherlands, and Royal Hospital for Sick Children, Bristol, UK. Despite the fact that preparative regimens and preand post-transplant treatment varied among the centres, the log-rank test statistics found no significant difference in the survival among the patients from the different centres (Mantel method P1⁄4 0.736). Consequently, it was statistically correct to pool the data. A multivariate analysis of the whole group with respect to sex, age at presentation, presence of chromosomal aberrations, T-cell depletion (TCD) of the graft, remission status and MRD level was performed to find independent factors influencing the event-free survival (EFS). Delay between obtaining the sample and HSCT was also taken into account. Immunophenotype of the leukaemic cells and type of the donor could not be included in the multivariate analysis as this method of analysis was not robust enough to manage these multiple variables. Therefore, univariate analysis of these two variables was performed. A total of 85 patients were transplanted in second or higher remission. A multivariate analysis of the above described covariates with the duration of first complete remission added was also performed for these patients. The Cox proportional hazards estimation model with stratification on the transplant centre was used for the multivariate and univariate analysis. Since the 5-year EFS in the high-level MRD group (MRDX10 ) varied among the centres (range 0–41.2%) and also the distribution of T-cell depleted vs nondepleted transplantations (23.5–79.0% of TCD grafts) was unequal, stratification on the transplant centre was maintained in the analysis unless stated otherwise. The unpublished data of the Pre-BMT-MRD Study Group were considered as one group. The level of significance was assessed as Po0.05. The results of the statistical analysis are summarised in Table 1. MRD proved to be a highly significant (Po0.001) independent factor to influence EFS of this group. Also, MLL gene rearrangements (present in seven patients) and pro-B immunophenotype (present in five patients) were significant negative prognostic factors. The analysis of patients transplanted in second (CR2) or higher complete remission revealed that the duration of first complete remission (CR1) was a significant independent factor. The 5-year EFS of the group with negative, lowand high-level positive-MRD was 75.2, 40.7 and 21.4%, respectively (Mantel Po0.001) (Figure 1). When the lowand highlevel MRD-positive groups were combined, the EFS of the MRD-positive group was 29.8% (Mantel Po0.001). Each Bone Marrow Transplantation (2003) 32, 849–851 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00