Molecular insights from ctdna dynamics: Monitoring MRD and tracking divergent progression modes and clonal evolution patterns Free

Background: Both baseline clinical and biological parameters, including IPI, and end-of-treatment PET (EOT-PET), which assesses treatment response, serve as critical prognostic factor for DLBCL. Circulating tumor DNA (ctDNA) has been validated as a reliable parameter for genotyping and minimal residual disease (MRD) detection in DLBCL. It remains intriguing whether EOT-MRD analysis could complement both baseline IPI and EOT-PET for prognostic stratification in DLBCL patients (pts).

Genetic subtype-guided chemoimmunotherapy is promising in DLBCL, but focused on induction therapy. Hence, the second purpose is to analyze evolutionary dynamics and progression modes in different genetic subtype of DLBCL using ctDNA detection, which may chart future directions for subtype-guided therapeutic development.

Methods:

Between April 2021 and July 2023, we enrolled 164 DLBCL pts from our center undergoing R-CHOP therapy for ctDNA profiling at 3 pre-defined milestones (baseline, EOT and progression disease (PD) plasma samples), using a 475 gene lymphoma-specific sequencing panel. By February 2025, the median follow-up had reached 28.4 months (range: 19.9-46.7 months). EOT-MRD positivity (EOT-MRD(+)) is defined by meeting either criterion: (1) Detection of any mutation identified in the matched baseline tissue sample, or (2) Presence of at least one reported driver mutation (Reddy A, Cell, 2017) in plasma ctDNA analysis at EOT.

Results:

After frontline therapy, 62.8% (103/164) achieved MRD clearance at EOT, while 37.2% (61/164) remained EOT-MRD (+). Among the 131 pts achieving complete remission (CR), 36 (27.5%) pts retained EOT-MRD(+), while subsequent PD was observed in 33.3% (12/36) of the CR pts. Among the 9 pts with partial remission (PR), 4 pts were identified as EOT-MRD(+), of whom 3 pts (75%) subsequently progressed, and 87.5% of the PD/SD pts were identified as EOT-MRD(+).

EOT-MRD(+) indicated treatment response and was significantly associated with progression free survival (PFS) and overall survival (OS), while IPI is a classic prognostic model widely used in clinical practice, along with pretreatment ctDNA burden, indicates baseline tumor burden. When incorporating these 3 factors into multivariate COX regression model, IPI score 3-5 and EOT-MRD(+) ultimately retained as independent determinants for OS and PFS. Furthermore, no matter in CR or non-CR pts, EOT-MRD(+) pts still exhibited significantly worse PFS than EOT-MRD(-) pts. Moreover, the combination of IPI and MRD (IPI-M) surpassed IPI alone in survival stratification in whole group, and enhanced PFS stratification in CR-achieving pts (IPI-M=1 vs IPI-M=0, HR 2.953, P=0.040; IPI-M=2 vs IPI-M=0, HR 9.806, P<0.001).

Furthermore, ctDNA served as convenient approach to track divergent progression modes and clonal evolution. Firstly, 77.1% (37/48) of pts acquired new gene alterations (GAs) at PD, most were enriched in cell cycle regulation, p53 pathway, PI3K/AKT signaling pathway, and epigenetic regulation. Secondly, in different progression modes, primary refractory pts exhibited a significantly higher proportion of PRE-PD shared mutations, while relapsed pts showed a greater proportion of PD-specific and lower proportion of shared mutations.

With regard to different genetic subtypes, TP53 disruption pts showed highest prevalence of primary refractory rather than relapsed cases. However, MCD subtype exhibited a persistent high risk of relapse despite reaching CR, while the clinical problem for BN2 pts is primary refractory and high rate of relapse. EOT-MRDclearance was unstable for MCD subtype with high risk of relapse, while achieving EOT-MRD(-) indicated favorable prognosis for ST2, BN2 and Other subtype. Moreover, different subtypes exhibited divergent clone patterns. MCD harbored more PRE-specific GAs yet fewer shared GAs in PRE-PD matched plasma, while BN2 subtype exhibited dominance in shared GAs, suggesting a preference for “linear” evolution. However, TP53 disruption exhibited disorder of co-occurring GAs except for TP53 mutation, indicating the essential role of TP53 mutation persistence in progression.

Conclusions: Our study demonstrated the essential role of ctDNA in monitoring MRD and tracking clonal evolution dynamics. EOT-MRD can complement both baseline IPI and EOT-PET for prognostic stratification. Besides, ctDNA-based tracking of clonal evolution and progression modes provides the rationale for subtype-specific personalized therapy.