Deep Sequencing the Tet2 Gene in 360 Patients with Myeloid Neoplasms Provides a Comprehensive and Quantitative Mutation Map and Reveals Low Level Mutant Clones.

Abstract 733

Tet2 is a novel candidate tumor suppressor and mutations in this gene have been found in 24% of myeloid neoplasms following conventional Sanger sequencing. Using a massively parallel pyrosequencing methodology (Roche GS FLX) we sequenced the entire Tet2 gene coding region in 360 patients with myeloid malignancies (MDS 279; CMML 32; AML 10; MDS/MPD 10; 5q syndrome 8; others 21). We also utilized high density SNP6 arrays to screen for chromosomal aberrations including those at the Tet2 locus itself. Initially we found 106 mutations in Tet2 corresponding to 75 patients (21%), where the average sequence coverage was more than 100-fold for >90% of the gene. 25 patients had more than one mutant Tet2 clone. 83 mutations (in 17% of patients) were found in 30% to 95% of sequenced clones, where 50% or 100% could be indicative of heterozygous and homozygous mutations that would be found in all sample bone marrow cells respectively. The remaining 23 mutations were found in ≤30% of sequence clones, applying a lower threshold of 17% to avoid false positives, and were not reliably picked up by confirmatory Sanger sequencing. The 106 mutations were distributed as 44 nonsense or InsDel mutations throughout the gene, 59 nonsynonomous mutations within (23) and outside (36) the two known conserved domains, and 3 spice site mutations.

Tet2 mutations were not correlated with survival outcome (probability of survival after 3 and 5 years from diagnosis 40% and 38% [p=0.3 log rank]for mutants vs non-mutants respectively). Patients with small (≤30%) or multiple mutant clones did not deviate from the overall survival trend. Furthermore, no difference in survival was seen when patients were analysed according to their WHO subgroups or IPSS scores. CMML featured as the only subgroup with a relatively high Tet2 mutation rate (53% of patients). Chromosomal aberrations linked to the Tet2 locus, including UPD or micro-deletion, were detected in 20-25% of patients, but this was not strongly correlated to detectable mutations in the Tet2 gene itself (24% coincidence), perhaps indicating that several forms of mutation or instability at this region are linked to disease. Interestingly, high percentage mutant clones which are indicative of homozygosity and thus loss of heterozygosity (LOH), were found as InsDel mutations in Tet2 for the majority (85%), suggesting a common defect in DNA replication or repair.

Following up on additional smaller Tet2 mutant clones, we lowered our threshold for mutation detection and after further analysis and removal of experimental background, 42 additional candidate mutations of clone size between 8% and 17% were found. Taken as a whole, Tet2 gene mutations were found in 30% of patients but this may rise considerably if deeper sequencing screens for even smaller mutant clones yield further positive results.

We therefore show via highly sensitive ‘Deep sequencing' on a relatively large patient cohort a comprehensive and highly quantitative mutation map across the Tet2 gene to allow increased prognostic comparisons. We also show that mutation figures rise significantly when smaller mutant clones are considered that cannot be detected via Sanger sequencing. In addition to this, if larger chromosomal aberrations such as UPD are added to the evolving number of mutant Tet2 patients, then Tet2 abnormalities may in fact be present in a majority of patients with myeloid neoplasms.