Plasma Concentrations and Cancer-Associated Mutations in Cell-Free Circulating DNA of Treatment-Naive Follicular Lymphoma for Improved Non-Invasive Diagnosis and Prognosis

Abstract

Follicular lymphoma (FL) is the second most frequent non-Hodgkin lymphoma accounting for 10-20% of all lymphomas in western countries. As a clinically heterogeneous cancer, FL occasionally undergoes histological transformation to more aggressive B cell lymphoma types that are associated with poor prognosis. Here we evaluated the potential of circulating cell-free DNA (cfDNA) to improve the diagnosis and prognosis of follicular lymphoma patients. Twenty well-characterized FL cases (13 symptomatic and 7 asymptomatic) were prospectively included in this study. Plasma cfDNA, formalin-fixed paraffin-embedded (FFPE) tumor tissue DNA, and patient-matched granulocyte genomic DNA samples were obtained from 20 treatment-naive FL cases. Ultra-deep targeted next-generation sequencing was performed with these DNA samples by using a custom-designed platform including exons and exon-intron boundaries of 110 FL related genes. Using a strict computational bioinformatics pipeline, we identified 91 somatic variants in 31 genes in treatment-naive FL cases. Selected variants were cross-validated by using PCR-Sanger sequencing. We observed higher concentrations of cfDNA and a higher overlap of somatic variants present both in cfDNA and tumor tissue DNA in symptomatic FL cases compared to asymptomatic ones. Variants known to be associated with FL pathogenesis such as STAT6 p.D419 or EZH2 p.Y646 were observed in patient-matched cfDNA and tumor tissue samples. Consistent with previous observations, high Ki-67 staining, elevated LDH levels, FDG PET/CT positivity were associated with poor survival. High plasma cfDNA concentrations or the presence of BCL2 mutations in cfDNA showed significant association with poor survival in treatment-naive patients. BCL2 mutation evaluations in cfDNA improved the prognostic utility of previously established variables. In addition, we observed that a FL patient who had progressive disease contained histological transformation-associated gene (i.e. B2M and BTG1) mutations only in cfDNA. Pre-treatment concentrations and genotype of plasma cfDNA may be used as a liquid biopsy to improve diagnosis, risk stratification, and prediction of histological transformation. Targeted therapies related to oncogenic mutations may be applied based on cfDNA genotyping results. However, the results of this study need to be validated in a larger cohort of FL patients as the analyses conducted in this study have an exploratory nature.

Whole Transcriptome Sequencing Reveals Cancer-Related, Prognostically Significant Transcripts and Tumor-Infiltrating Immunocytes in Mantle Cell Lymphoma

“…To evaluate the consistency between WTS and qRT-PCR expression values through calculation of the Pearson correlation coefficients, RStudio software version 4.0.2 (RStudio Team, Boston, MA, USA, https://www.rstudio.com/, accessed on 23 July 2022) was used, as described previously [48]. The WTS data values used in linear correlation analyses were initially obtained by dividing normalized counts per million (CPM) values of CCND1 or SNHG5 to the CPM values of the RPS13 housekeeping gene for each sample after differential expression (DE) analysis of MCL tumor samples (n = 32) and reactive tonsil B-cell subset samples (n = 5).…”

“…For differential transcript expression analyses with DESeq2, FDR-adjusted p values were obtained through correction of the Wald test p values with multiple testing by applying the Benjamini and Hochberg method [ 40 ]. The significance of the MCL survival differences between two patient groups dichotomized based on the expression of transcripts or immunocyte ratios was evaluated by calculating p values based on the log-rank test [ 48 ]. To evaluate the consistency between expression values obtained with whole transcriptome sequencing and qRT-PCR for selected transcripts, the p values were calculated via Pearson correlation testing for paired expressions [ 49 ].…”

The impact of preanalytical variables on the analysis of cell-free DNA from blood and urine samples

“…qPCR-based techniques are also commonly used to quantify cfDNA.With their high sensitivity ( Pan et al, 2017 ), accuracy ( Leung et al, 2021 ), and low false positive rate ( Yin et al, 2022 ), these methods can be used to measure trace nucleic acids effectively ( Pan et al, 2017 ) and analyze cfDNA for known mutations ( Hatipoglu et al, 2022 ). By detecting housekeeping genes ( Aucamp et al, 2016 ) or noncoding repetitive sequences ( Hussein et al, 2019 ) in cfDNA and fitting the standard curve with a reference substance ( Tang et al, 2020 ), absolute cfDNA concentrations can be quantified using PCR-based methods.…”

Technical and biological variations in the purification of extrachromosomal circular DNA (eccDNA) and the finding of more eccDNA in the plasma of lung adenocarcinoma patients compared with healthy donors

“…However, the field of cfDNA is still developing, and cfDNA assays often show insufficient sensitivity and specificity for many cancers, especially for those in the early stages (8,11). Linear DNA is unstable (T1/2 = 114 min, (12)) and is only found at low concentrations in plasma, causing high variability of cfDNA concentration and content among patients and samples (13)(14)(15)(16)(17)(18)(19). This reduces linear plasma DNA's applicability as a broad cancer marker despite its relevance as a direct indicator of tumor constituents.…”