Genomic landscape of spontaneously regressing CLL tumors. WES data initially analyzed using standard bioinformatics criteria, subsequently underwent further selection to minimize artifacts and focus on variants of potential functional significance. The selected variants were predicted to be truncating or frameshift, missense mutations predicted to be deleterious, variant allele frequency ≥ 20% and could be validated by Sanger or RNA-seq. (A) The genomic landscape of 18 spontaneously regressed sorted CD19⁺CD5⁺ CLL tumors is displayed. The CNV data of each spontaneous regression case is combined with their respective SNV data obtained from WES. Different types of genomic events are represented by different colors, with the color code displayed adjacent to the table. The frequency of each genomic event is represented in the bar chart beneath the table. (B-D) Fish plots of somatic variants in sorted CD19⁺CD5⁺ CLL cells from (B) CLL05, CLL06, and CLL19 showing clonal equilibrium during the course of spontaneous regression, (C) CLL20 showing clonal equilibrium followed by clonal evolution during CLL relapse after spontaneous regression, and (D) CLL02 showing clonal fluctuation during regression. T0, T1, and T2 represent the diagnostic, regression, and relapse time points, respectively.