Invariant phenotype and molecular association of biallelic TET2 mutant myeloid neoplasia

Abstract

Somatic TET2 mutations (TET2 ^MT) are frequent in myeloid neoplasia (MN), particularly chronic myelomonocytic leukemia (CMML). TET2 ^MT includes mostly loss-of-function/hypomorphic hits. Impaired TET2 activity skews differentiation of hematopoietic stem cells toward proliferating myeloid precursors. This study was prompted by the observation of frequent biallelic TET2 gene inactivations (biTET2 ⁱ ) in CMML. We speculated that biTET2 ⁱ might be associated with distinct clinicohematological features. We analyzed TET2 ^MT in 1045 patients with MN. Of 82 biTET2 ⁱ cases, 66 were biTET2 ^MT, 13 were hemizygous TET2 ^MT, and 3 were homozygous TET2 ^MT (uniparental disomy); the remaining patients (denoted biTET2 ^- hereafter) were either monoallelic TET2 ^MT (n = 96) or wild-type TET2 (n = 823). Truncation mutations were found in 83% of biTET2 ⁱ vs 65% of biTET2 ^- cases (P = .02). TET2 hits were founder lesions in 72% of biTET2 ⁱ vs 38% of biTET2 ^- cases (P < .0001). In biTET2 ⁱ , significantly concurrent hits included SRSF2 ^MT (33%; P < .0001) and KRAS/NRAS ^MT (16%; P = .03) as compared with biTET2 ^- When the first TET2 hit was ancestral in biTET2 ⁱ , the most common subsequent hits affected a second TET2 ^MT, followed by SRSF2 ^MT, ASXL1 ^MT, RAS ^MT, and DNMT3A ^MT BiTET2 ⁱ patients without any monocytosis showed an absence of SRSF2 ^MT BiTET2 ⁱ patients were older and had monocytosis, CMML, normal karyotypes, and lower-risk disease compared with biTET2 ^- patients. Hence, while a second TET2 hit occurred frequently, biTET2 ⁱ did not portend faster progression but rather determined monocytic differentiation, consistent with its prevalence in CMML. Additionally, biTET2 ⁱ showed lower odds of cytopenias and marrow blasts (≥5%) and higher odds of myeloid dysplasia and marrow hypercellularity. Thus, biTET2 ⁱ might represent an auxiliary assessment tool in MN.

Graphical abstract

Figure 1.

TET2 gene mutation classification, type, and clinical characteristics. (A) Scatterplot of the VAFs of patients with TET2^MT. The VAF of first-hit TET2^MT was plotted on the x-axis and that of the second-hit TET2^MT, if present, on the y-axis. Patients were categorized into 5 groups as explained in the text. The red oval corresponds to biTET2^MT cases, the gray bar to undetermined cases, the blue oval to biclonal TET2^MT cases, the light green oval to monoTET2^MT cases, and the yellow oval to hemihomozygous TET2^MT (UPD) cases. (B) Percentages of different types of TET2^MT in biTET2ⁱ cases and significance of truncating mutations vs monoTET2^MT. Fisher’s exact test was used for analysis. (C) Bar graphs showing the distribution of biTET2ⁱ and monoTET2^MT cases per diagnosis and cytogenetics. The bar columns indicate percentages. (D) Pie chart showing the percentage of cases per configuration. (E) Pie charts of biTET2ⁱ, monoTET2^MT, and TET2^WT respectively representing the percentage of cases per classification. CMML (+) indicates cases with CMML at the time of presentation, CMML (−) indicates no CMML diagnosis, monocytosis (+) indicates the presence of monocytosis, and monocytosis (−) indicates the absence of monocytosis.

Figure 2.

Clonal architecture and hierarchy of TET2^MTin biTET2ⁱ. (A) Plot showing dominant, codominant, and secondary mutations in the 82 biTET2ⁱ patients. Mutated gene names, cytogenetics, and diagnosis are color coded as indicated. For this presentation, only genes mutated ≥3 times among the biTET2ⁱ population are represented. Each column represents 1 patient, and each row corresponds to 1 gene or family of genes. (B) Pie chart displaying the percentage of first-hit TET2 occurring as dominant (ancestral), codominant (ancestral), and subclonal (secondary) in the biTET2ⁱ population. (C-E) The bar graphs show the percentages of the corresponding dominant genes to the secondary/subclonal first-hit TET2 gene (C), secondary clones to the dominant first-hit TET2 gene (D), and codominant genes to the codominant first-hit TET2 gene (E). (F) Frequency (in percentage) of mutations in selected genes in the population. Ten genes that are frequently mutated in myeloid neoplasms were selected. Columns are color coded per TET2^MT configuration (TET2^WT, monoTET2^MT, and biTET2ⁱ).

Figure 3.

Significance of concurrent gene mutations in biTET2ⁱand correlation per disease subtype. (A) Forest plot showing the OR of associated gene mutations in biTET2ⁱ vs biTET2⁻ cases. As indicated, red squares correspond to significant cases, while red stars correspond to highly significant cases. Fisher’s exact test was used to test significance. (B) Frequency (in percentage) of mutations in selected genes per disease subtype in biTET2ⁱ vs biTET2⁻. Significance was tested via Fisher’s exact test.

Figure 4.

Univariate analysis for baseline, clinical, and genomic features in biTET2ⁱvs wild-type. Univariate analysis showing the significant results for baseline (older age, lower risk, MDS, MDS/MPN, CMML, and normal cytogenetics), clinical (leukopenia, neutropenia, monocytosis, pancytopenia, BM blasts, BM hypercellularity, and myeloid dysplasia), and genomic (KRAS/NRAS^MT, SRSF2^MT, and TP53^MT) features in biTET2ⁱ vs TET2^WT.