Page 110 - 2021_09-Haematologica-web

E. Oppliger Leibundgut et al.
of genes in MPN, and clonality has been demonstrated.7-9 In ET, mutations in ASXL1, TET2 and DNMT3A genes are most frequent and are all involved in epigenetic regula- tion. Less frequent recurrent mutations are detected in EZH2, TP53, IDH1, IDH2 and CBL, as well as in genes of the splicing machinery, such as SF3B1, SRSF2, U2AF1 and ZRSF2.
So-called “adverse mutations” in SF3B1, SRSF2, U2AF1, TP53, IDH2 and EZH2 have been found to have negative effects on overall and myelofibrosis-free survival in ET, and TP53 mutations predict leukemic transformation.8-10,14 Furthermore, ASXL1 mutations have been identified as a genetic risk factor for transformation to myelofibrosis in ET patients, as they are most frequently found in post-ET myelofibrosis.11 Subsequently, genomic data were integrat- ed in prognostic models to predict patients’ outcomes.13,14
An influence of additional non-driver mutations on treatment response in MPN has been reported for interfer- on-α, ruxolitinib and imetelstat. In patients with CALR-mutated ET treated with interferon-α, the presence of additional mutations in ASXL1, TET2, IDH2 and TP53 correlated with a poorer molecular response.15 TET2- mutated clones were resistant to interferon-α therapy in JAK2-mutated patients with polycythemia vera.16 Resistance to ruxolitinib was reported in patients with myelofibrosis carrying three or more mutations.17 Furthermore, in the first clinical trial with imetelstat in myelofibrosis patients, treatment response was reported to be negatively influenced by ASXL1 mutations and favorably impacted by SF3B1 and U2AF1 mutations.18
In the present study, we assessed a panel of genes fre- quently mutated in MPN by next-generation sequencing at study entry and during treatment with imetelstat, and investigated the dynamics of additional mutations in ET patients and their association with hematologic and molecular response.
Methods
Patients and response criteria
A total of 18 patients with ET diagnosed according to the World Health Organization (WHO) 2008 criteria were treated with ime- telstat in a phase II study.4 The study was approved by the insti- tutional review board at each participating site. All patients pro- vided written informed consent. Diagnoses were re-evaluated according to the WHO 2016 classification.19 Sequential blood sam- ples were taken at baseline and at up to eight time-points during treatment with imetelstat through cycle 26 (28-day cycles), with approximately 12 weeks between samples. Mutational analysis was performed on all collected samples.
Clinical and hematologic responses were assessed according to the European LeukemiaNet criteria.20 Molecular responses of phe- notypic driver mutations were defined as follows: a major molec- ular response (MMR) was achieved when the mutant allele bur- den reduction was >50% from baseline value, and a partial molec- ular response (PMR) was present when a 25% to 49% reduction of the mutant allele burden was observed.
Genetic analysis
DNA was extracted from granulocytes or leukocytes from peripheral blood samples. The molecular response of JAK2 V617F, CALR and MPL mutations was assessed using allele-specific real- time polymerase chain reaction, sequencing and fragment length analysis, respectively, as previously described.4
Targeted next-generation sequencing of all relevant exons and adjacent intronic sequences of 15 recurrently mutated genes (ASXL1, CBL, DNMT3A, EZH2, IDH1, IDH2, JAK2, MPL, SOCS1, TET2, TP53, SF3B1, SRSF2, U2AF1 and ZRSR2) was performed using Ion TorrentTM semiconductor chip technology on the Ion Personal Genome Machine® PGMTM (Thermo Fisher Scientific Inc.). Genes were covered by two custom-designed amplicon libraries comprising 511 and 307 amplicons. In addition, a com- mercial panel for TP53 was used for confirmation of TP53 variants (Ion AmpliSeqTM Community Panel TP53, Thermo Fisher). For each primer pool, 10 ng of DNA were processed using the AmpliSeqTM chemistry for selective amplification of target sequences and library preparation according to the manufacturer’s instructions. Libraries were diluted and combined according to the Ion PGM chip size to obtain a minimum coverage of 500x for all amplicons. Templates were prepared on the Ion ChefTM and sequencing was performed on the Ion PGM instrument. Variants were called using IonTorrent VariantCaller v4.3 software based on the human reference genome (GRCh37/hg19). Analysis of TP53 was performed according to the manufacturer’s instruction. Annotation was done using the Mutalyzer, dbSNP, COSMIC, ClinVar, UniProt and IARC TP53 databases and the functional in silico prediction algorithms PolyPhen-2 and SIFT.21
Fragment analysis was used to screen for insertions and dele- tions in ASXL1 exon 12 (NG_027868.1), which are frequently missed by next-generation sequencing. Primers were designed according to Pratcorona et al.22 with small adaptations, and analy- sis was performed on a 3130 Genetic Analyzer using peak scanner software (Thermo Fisher). Sequences were confirmed by Sanger sequencing.
The limits of detection for real-time polymerase chain reaction analysis of JAK2 and MPL mutations were 0.5%, whereas those for CALR and ASXL1, determined by fragment analysis, and all variants detected by next-generation sequencing were set at 2%.
Validation of genetic variants
All novel variants were confirmed by Sanger sequencing. For ASXL1 and TET2 analysis, published primers were used,23,24 and primers for other genes were designed using Oligo7 and Primer3 software.25,26 Low-level variants (<10%) were confirmed by a sec- ond round of next-generation sequencing analysis.
Statistics
Categorical patients’ characteristics were summarized by fre- quencies and percentages and continuous characteristics by medi- ans, means and an unpaired Student t-test. The efficiency of ime- telstat treatment was analyzed by a paired Student t-test compar- ing percentages of mutant allele burdens before treatment and at best response. Smooth estimates of allele burdens over time were generated using running medians and smoothing splines. Standard errors and confidence intervals were computed by bootstrap.
Results
Characteristics of patients and phenotypic driver mutations
Of 18 patients with ET enrolled in the study, nine (50%) were refractory and 14 (78%) were intolerant of at least one prior therapy. Thirteen patients had received more than one prior therapy and the median time since diagno- sis was 7.2 years (range, 0.3-24.9) (Table 1). The median age of patients at study entry was 59.5 years (range, 21-83) (Online Supplementary Table S1). Upon treatment with ime- telstat, all patients had a hematologic response, with 16 patients achieving complete hematologic responses.4
2398
haematologica | 2021; 106(9)