Figure 4.
Noninvasive genotyping and genomic analysis of ctDNA at diagnosis. (A) Dot plots of somatic mutations according to their variant allele frequencies (VAF) in concurrent pretreatment plasma samples (y-axis) and diagnostic tumor biopsies (x-axis) in 2 patients. Case#1 with high concordance, Case#2 with spatial heterogeneity demonstrated with both plasma- and tumor-specific mutations in addition to shared mutations. Colors of dots indicate mutations in driver genes (red) and immunoglobulin targets (blue). (B) Dot and line graph tracking somatic reporter mutations in a patient with relapse tissue biopsy sequenced. Purple color denotes reporter mutations in the diagnostic tissue biopsy (driver genes labeled) that are subclonal or not detected in concurrent pretreatment plasma sample and not present in the consecutive relapse tissue. Green color and text annotations denote drivers shared by the diagnostic and relapse biopsies. Gray dots denote other indifferent reporter mutations between the samples. (C) Oncoprint of the coding driver mutation landscape according to pretreatment ctDNA concentration. Columns represent individual patients, and rows represent different clinical variables or driver genes. Genes mutated in ≥15% of the patients included, and the percentages are indicated. Asterisks in clinical factors denote a positive association with pretreatment ctDNA concentration. (D) Box plot demonstrating the difference in pretreatment ctDNA concentration between TP53-mutated and WT patients. (E) Kaplan-Meier curve for overall survival demonstrating the poor outcomes of patients with TP53 mutations in the ctDNA at diagnosis. (F) Box plot demonstrating the different abundance of mutations with RCH and TW hypermutation signatures (y-axis, post-GC signature, #RCH:#TW ratio) between GCB and non-GCB lymphomas according to Hans’ algorithm. (G) Oncoprint with patients arranged according to post-GC signature with driver genes associated with the signature. Driver genes with Wilcoxon rank-sum P values < .05 for association are shown. FDR, false discovery rate. (H) Forest plots of multivariable Cox-proportional hazard model hazard ratios, 95% coincidence intervals, and P values for progression-free survival (PFS). Post-GC signature (median cutoff, model #1) or CD79B mutation (model #2) and pretreatment ctDNA burden. (I) Oncoprint of genomic targets and genes most affected by somatic phased events (rows). Top annotation bar plot; sum of variants detected in-phase per patient (columns). (J) Box plot demonstrating the number of detected BCL2 mutations (y-axis) per patient according to BCL2 translocation status in the diagnostic biopsy according to FISH (x-axis). Break-apart probe detection. (K) Dot plot of somatic mutations and their VAF (y-axis) in pretreatment plasma sample of Case#3 with multiple subclonal BCL10 C-terminal truncating mutations (black dots and text annotation). Other driver events are annotated with text and green dots. Gray dots denote other somatic events. (L) Lollipop plot showing the location and VAFs of BCL10 targeting nonphased subclonal mutations. Below, a screenshot from interactive genomics viewer (IGV) showing the mutually exclusive nature of reads containing the mutations.

Noninvasive genotyping and genomic analysis of ctDNA at diagnosis. (A) Dot plots of somatic mutations according to their variant allele frequencies (VAF) in concurrent pretreatment plasma samples (y-axis) and diagnostic tumor biopsies (x-axis) in 2 patients. Case#1 with high concordance, Case#2 with spatial heterogeneity demonstrated with both plasma- and tumor-specific mutations in addition to shared mutations. Colors of dots indicate mutations in driver genes (red) and immunoglobulin targets (blue). (B) Dot and line graph tracking somatic reporter mutations in a patient with relapse tissue biopsy sequenced. Purple color denotes reporter mutations in the diagnostic tissue biopsy (driver genes labeled) that are subclonal or not detected in concurrent pretreatment plasma sample and not present in the consecutive relapse tissue. Green color and text annotations denote drivers shared by the diagnostic and relapse biopsies. Gray dots denote other indifferent reporter mutations between the samples. (C) Oncoprint of the coding driver mutation landscape according to pretreatment ctDNA concentration. Columns represent individual patients, and rows represent different clinical variables or driver genes. Genes mutated in ≥15% of the patients included, and the percentages are indicated. Asterisks in clinical factors denote a positive association with pretreatment ctDNA concentration. (D) Box plot demonstrating the difference in pretreatment ctDNA concentration between TP53-mutated and WT patients. (E) Kaplan-Meier curve for overall survival demonstrating the poor outcomes of patients with TP53 mutations in the ctDNA at diagnosis. (F) Box plot demonstrating the different abundance of mutations with RCH and TW hypermutation signatures (y-axis, post-GC signature, #RCH:#TW ratio) between GCB and non-GCB lymphomas according to Hans’ algorithm. (G) Oncoprint with patients arranged according to post-GC signature with driver genes associated with the signature. Driver genes with Wilcoxon rank-sum P values < .05 for association are shown. FDR, false discovery rate. (H) Forest plots of multivariable Cox-proportional hazard model hazard ratios, 95% coincidence intervals, and P values for progression-free survival (PFS). Post-GC signature (median cutoff, model #1) or CD79B mutation (model #2) and pretreatment ctDNA burden. (I) Oncoprint of genomic targets and genes most affected by somatic phased events (rows). Top annotation bar plot; sum of variants detected in-phase per patient (columns). (J) Box plot demonstrating the number of detected BCL2 mutations (y-axis) per patient according to BCL2 translocation status in the diagnostic biopsy according to FISH (x-axis). Break-apart probe detection. (K) Dot plot of somatic mutations and their VAF (y-axis) in pretreatment plasma sample of Case#3 with multiple subclonal BCL10 C-terminal truncating mutations (black dots and text annotation). Other driver events are annotated with text and green dots. Gray dots denote other somatic events. (L) Lollipop plot showing the location and VAFs of BCL10 targeting nonphased subclonal mutations. Below, a screenshot from interactive genomics viewer (IGV) showing the mutually exclusive nature of reads containing the mutations.

Close Modal

or Create an Account

Close Modal
Close Modal