Landscape of TP53 Mutations in MPN

Background

The Myeloproliferative Neoplasms, including Essential Thrombocythemia (ET), Polycythemia Vera (PV), and Myelofibrosis (MF) are stem cell disorders which carry the risk of progression to accelerated phase or blast-phase (MPN-AP/BP). Among the risk factors for transformation to MPN-AP/BP are TP53 mutations. TP53 mutations are a risk factor for progression to MPN-AP/BP in chronic-phase myeloproliferative neoplasms (MPN) and indeed 30% of patients with MPN-LT bear TP53 mutations. A recent study indicated that TP53 mutations may persist at low levels for years without an immediate risk of progression in some chronic-phase MPN patients [Kubesova et al. Leukemia 2017]. These observations raise the question as to whether the types of TP53 mutations, their frequency, or co-occurring variants differ between MPN subtypes. TP53 mutations are characterized by frameshift, missense, nonsense or silent mutations. Mutations in TP53 have traditionally been considered functionally equivalent in many prognostic studies, but an increased understanding of the effects of distinct mutations on TP53 activity has led to the recognition of the distinct functional significance of these different mutations. The understanding of the landscape of TP53 mutations in MPN and the association between different TP53 mutations and mutational burdens among MPN subtypes may allow more accurate prognostic evaluation and personalized treatment for patients. We therefore sought to assess the landscape of TP53 mutations in MPN subtypes.

Methods

We performed Next-Generation Sequencing (NGS) on a cohort of 651 samples derived from 439 MPN patients with ET, PV, MF and MPN-AP/BP. The data were from 3 targeted panels with 576, 585 and 156 genes (with 91, 241 and 319 samples; median coverage 500x). Samples were obtained from Memorial Sloan Kettering Cancer Center and the Myeloproliferative Neoplasms Research Consortium. Putative oncogenic mutations were called using a combination of 4 variant callers and by comparison to established cancer databases. The final variants were manually reviewed to guarantee high quality of downstream analysis.

Results

We identified somatic mutations in 428/439 patients (157 MPN-AP/BP, 140 MF, 67 ET and 64 PV). Of these, 55 patients (~13%) had at least one TP53 mutation with variant allele frequency (VAF) >= 2% (median TP53 VAF= 28%). In total, 68 TP53 somatic mutations were identified. Majority of these patients had a single TP53 mutation (46/55) and some had multiple mutations (7 had two and 2 had four mutations). Mutations were enriched in MPN-AP/BP with 45/68 (66%) occurring in this group followed by 10/68 (15%) in ET, 9/68 (13%) in MF and 4/68 (6%) in PV (panel A). Missense mutations were the most common type of TP53 variants and constituted ~80-90% of all mutations in ET, MF and MPN-AP/BP. Most stop-gains and frameshifts were observed in MF or MPN-AP/BP group. Furthermore, 92% of TP53 mutations are localized on DNA-binding domain (exons 5 to 8; panel B). The latter observation is consistent with results from other human cancers and highlights the role of these mutations in reducing TP53 DNA binding affinity. About 8% of mutations occur in the tetramerization domain of TP53, which is also critical for DNA binding, as well as protein-protein interactions and post-translational modification. We did not identify a significant association between a specific TP53 mutation type and any particular MPN subtype. However, we identified a significant association between TP53 VAF and MPN subtype (panel C); TP53 VAF was significantly higher in MPN-AP/BP compared to ET (p =0.0001) and MF (p =0.016).

Conclusion

TP53 mutations have important prognostic significance in patients with MPN. However, subgroups of patients with TP53 mutant chronic-phase disease have been observed to have relatively stable clinical course. Our data demonstrate that the spectrum of TP53 mutations, in terms of type and location within the gene, does not appear to differ between ET, PV, MF, or MPN-AP/BP. However, we observe significant difference with regard to the VAF of TP53 mutations, with an association of higher VAF and MPN-AP/BP state. This data argues that the VAF may be an important consideration in assessing the prognostic impact of a TP53 mutation identified in an MPN patient. Data regarding copy number variations in this cohort, co-occurring mutations, and the impact of TP53 mutations on treatment outcomes will be presented.

Figure

View largeDownload slide

Figure

View largeDownload slide

Close modal