COLD SPRING HARBOR, NY (GenomeWeb News) – Researchers from the University of Oxford and Illumina have used whole-genome sequencing to follow the changes in mutation patterns during treatment, remission, and relapse for two individuals with B-cell chronic lymphocytic leukemia, attendees at the Biology of Genomes meeting heard yesterday.
University of Oxford hematologist Anna Schuh presented data for pre- and post-treatment and matched normal genomes for two individuals with B-CLL. Both patients were treated with several rounds of purine analog and second-generation monoclonal antibody treatments and neither carried TP53 mutations or complex genomic rearrangements that have been previously tied to poor B-CLL outcomes, Schuh explained.
In an effort to find mutations that might explain B-CLL relapse events or act as markers for diagnoses and treatment response, the researchers sequenced tumor genomes to an average of 30 times coverage with the Illumina HiSeq2000. Matched normal samples were sequenced by a similar approach using DNA from buccal or saliva samples, Schuh explained, since blood samples contain tumor cells.
Overall, the team detected between 4,000 and 6,000 apparent somatic mutations or small insertions and deletions per tumor sample, though fewer than half of these changes fell within protein-coding genes.
When they focused in on mutations predicted to alter protein sequences, the team found between 15 and 17 non-synonymous or nonsense mutations in tumor samples from one of the patients and 18 to 20 such mutations in tumors from the other patient.
By doing targeted deep sequencing of these mutations — generating some 10,000 to 50,000 fold coverage for each — the researchers were able to glean information about mutation frequency in the tumors at the various time points.
They then classified these mutations into different groups depending on whether they disappeared after treatment, remained following various treatments, became more common after treatment, or only appeared in late stages of the disease.
In general, Schuh noted, so-called Class III and IV mutations, which emerged or became more frequent after treatment, tended to fall in genes from known cancer pathways, while the Class I and II mutations, which either disappeared or stayed fairly steady after treatments, often occurred in genes involved in processes such as innate immunity or apoptosis.
Among the mutations that cropped up in relapse tumors late in disease progression were alterations to MEK1, a kinase coding gene implicated in some solid tumors, she explained, hinting that MEK1 inhibitors might hold promise for treating some forms of the disease.
The team is currently doing similar sequential analyses on tumor samples for four more B-CLL patients, Schuh said. They have also done array-based copy number analyses of pre-treatment and relapse samples for nearly three dozen B-CLL patients.