Measurable residual disease and clonal evolution in acute myeloid leukemia with focus on NPM1-mutations

Sammanfattning: Acute myeloid leukemia (AML) is the most common form of acute leukemia in adults, with mutations in the NPM1 gene occurring in almost one third of all cases. The ability to detect residual leukemia below the resolution of conventional microscopy is crucial for evaluation of relapse risk after therapy. In principle, this can be achieved by measuring residual disease (MRD) with two different approaches, both used routinely in everyday hematological practice and in this thesis: multicolor flow cytometry (MFC) and molecular techniques. The latter include methods such as reverse transcription quantitative PCR (RT-qPCR), quantitative PCR (qPCR), droplet digital PCR (ddPCR) and next-generation sequencing (NGS). NPM1 mutations are ideal targets for molecular MRD and the level of NPM1-MRD, as determined by quantification of RNA transcripts, is currently considered the most relevant prognostic factor after first-line treatment. Instead, to explore the clinical relevance of genomic DNA-based molecular MRD methods, this thesis has focused on targeting NPM1 mutations, but also other AML-associated mutations, to decipher patterns of clonal evolution in AML before, during and after treatment.First, a qPCR-based protocol for quantification of the NPM1 type A mutation was refined, validated, and shown to be more sensitive than MFC for determination of MRD. This study was followed by an extensive comparison of DNA- and RNA-based methods for MRD assessment. The DNA-based methods proved highly accurate with respect to RNA thresholds of importance for treatment response. In addition, although RT-qPCR was more sensitive, it failed to detect leukemic transcripts in about 10% of samples with clear-cut NPM1-mutated DNA. Hence, DNA-based MRD techniques can add important information with respect to residual leukemia, of possible clinical relevance for MRD assessment. Next, several mutations in addition to NPM1 were targeted with ddPCR and monitored in follow-up samples after treatment. This strategy revealed several patterns of clonal evolution in relapsing AML. In one pattern, all monitored mutations reappeared at relapse regardless of the number of subclones. In other relapses, a subclone different from the original leukemia was responsible for the recurrence. Finally, in some patients, the leukemia relapsed from persistent clonal hematopoiesis despite complete morphological and immunophenotypical remission. To explore the mutational landscape and clonal evolution in elderly patients, who are often treated outside clinical trials, a cohort of patients older than 75 years with de novo AML with mutated NPM1 was analysed. The results indicate that the mutational pattern may differ between younger and older patients, with more TET2 and SRSF2 mutations but fewer DNMT3A mutations in the elderly.In conclusion, this thesis shows that DNA-based methods are more sensitive than MFC for determination of MRD and that they may complement RT-qPCR, with possible consequences for risk assessment of patients treated for NPM1-mutated AML. Targeting several mutations with ddPCR or other DNA-based techniques may be relevant for accurate and complete MRD assessment in the personalised follow-up of most AML patients. Finally, the mutational landscape seems to differ between younger and elderly AML-patients, with possible implications for risk stratification, and ultimately, treatment.