Myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) are hematological malignancies characterized by clonal hematopoietic stem cell abnormalities. Current diagnostic methods rely heavily on bone marrow (BM) biopsies, which are invasive and may not fully capture the disease's heterogeneity. Circulating tumor DNA (ctDNA), fragments of tumor DNA circulating in the bloodstream, has emerged as a promising minimally invasive biomarker for cancer diagnosis and monitoring. Several studies have shown excellent correlations between BM DNA and plasma-derived ctDNA in MDS and AML, suggesting ctDNA's potential for reflecting disease burden and response to treatment. Dynamic ctDNA monitoring has shown promise in predicting relapse after allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, comprehensive studies evaluating ctDNA's utility for monitoring the entire disease course in MDS and AML are limited. This study aimed to investigate the feasibility and clinical utility of serial ctDNA monitoring using targeted next-generation sequencing (NGS) to assess treatment response, track clonal evolution, and predict survival outcomes in patients with MDS and AML. The researchers hypothesized that ctDNA levels and changes in ctDNA mutations would correlate with disease progression, treatment response, and overall survival, offering a valuable tool for personalized medicine in these challenging malignancies.

Previous research has demonstrated the strong correlation between bone marrow-derived DNA and ctDNA in MDS and AML. Studies have utilized targeted NGS of ctDNA to monitor disease burden and predict relapse post-allo-HSCT. However, the application of serial ctDNA monitoring throughout the entire disease course remains under-investigated. Existing literature highlights the potential of ctDNA as a valuable tool in understanding clonal evolution, but larger, longitudinal studies are needed to confirm these findings and establish clinical guidelines. The review of previous literature informed the design of this study to specifically focus on a longitudinal assessment of ctDNA to monitor treatment response, track clonal evolution, and predict survival, thus filling a gap in the current literature.

This study involved 35 adult patients with MDS and AML. Twenty-seven patients had pre- and post-treatment plasma-derived ctDNA and paired baseline BM DNA NGS assessments. Four patients had only baseline assessments, and four had only pre- and/or post-treatment ctDNA assessments. Targeted NGS was performed to analyze 46 mutated genes. Concordance analysis compared mutations and variant allele frequencies (VAFs) between BM DNA and plasma ctDNA. Dynamic ctDNA analysis tracked changes in VAFs and ctDNA concentrations before and after treatment. Patients were categorized into complete response (CR), CR with incomplete hematologic recovery (CRi)/morphologic CR (mCR)/morphologic leukemia-free state (mLFS)/stable disease (SD), and progressive disease (PD) groups. Survival analyses, including progression-free survival (PFS) and overall survival (OS), were performed. Statistical analyses assessed correlations between ctDNA metrics (concentration and VAF) and clinical outcomes. Clonal evolution was analyzed in refractory/relapsed patients by tracking mutations and VAF changes over time. Flow cytometry data were used for comparison in some cases.

A high concordance (R = 0.854, p < 0.001) was observed between BM DNA and plasma ctDNA VAFs for concordant mutations. Pre-treatment mean ctDNA concentrations showed a strong correlation (R = 0.618, p < 0.001) with baseline BM blasts, reflecting tumor burden. Patients achieving CR showed a dramatic decrease in post-treatment VAFs (median VAF: 16.7% pre-treatment vs. 0% post-treatment, p < 0.001). Persistent VAFs were observed in CRi/mCR/mLFS/SD patients, while VAFs increased in PD patients. Post-treatment ctDNA positivity was significantly associated with shorter PFS (p < 0.001) and OS (p < 0.001). Increased mean ctDNA VAF and concentration from pre- to post-treatment predicted decreased PFS and OS (p < 0.001 for both). A model based on mean ctDNA concentration change demonstrated better discrimination of PFS (C-index: 0.79, AUC: 0.84) compared to the model based on mean ctDNA VAF. Analysis of clonal evolution revealed both linear and branched patterns in refractory/relapsed MDS and AML, highlighting ctDNA's potential for early detection of relapse (approximately 4 months earlier than flow cytometry).

This study provides strong evidence for the clinical utility of serial ctDNA monitoring in MDS and AML. The findings demonstrate that ctDNA is a reliable biomarker reflecting tumor burden, treatment response, and survival outcomes. The strong correlation between ctDNA and bone marrow blasts validates its use as a minimally invasive alternative to BM biopsies. The superior performance of the ctDNA concentration-based model over the VAF-based model in predicting PFS suggests that the overall ctDNA level may be a more robust prognostic indicator than individual mutation changes. The observation of different clonal evolution patterns highlights the potential for personalized medicine strategies based on ctDNA monitoring. The ability of ctDNA to detect relapse earlier than flow cytometry underscores its importance in timely intervention.

Serial ctDNA monitoring is a valuable minimally invasive tool for assessing treatment response, predicting survival, and tracking clonal evolution in MDS and AML. CtDNA concentration changes appear to be a particularly strong prognostic biomarker. Further prospective studies are warranted to validate these findings and integrate ctDNA monitoring into routine clinical practice for improved patient management. Future research might explore the use of ctDNA for early detection and personalized treatment strategies in these diseases.

The relatively small sample size of this study limits the generalizability of the findings. The study included a heterogeneous patient population with varying disease stages and treatment regimens. Further investigation is needed to assess the impact of different treatment modalities on ctDNA dynamics and to validate these findings in larger, more diverse cohorts. Long-term follow-up is needed to confirm the long-term prognostic value of ctDNA monitoring.