# TE TN

## Patients and Methods

This study is a retrospective study, using some biologic samples collected in the context of the care and/or diagnosis, and conserved into two declared bio-collections.

***Patients’ inclusion and non-inclusion criteria***
Patients from Angers and Brest university hospitals where included following the inclusion criteria : 1) patients triple-negative for JAK2V617F, exon 9 CALR and MPL515L/K, 2) patient with a diagnosis of essential thrombocytemia according to the 2016 revision to the World Health Organization classification of myeloid neoplasms or diagnosis of chronic thrombocytosis, 3) > 18 years old, 4) patients followed in clinical haematology department, 5) written informed consent. And the non inclusion criterion was : patients with reactive thrombocytosis, including : 1) iron deficiency, 2) splenectomy, 3) recent surgery, 4) infection, 5) inflammation, 6) cancer.

***Ethical considerations***

The samples used in the present study come from 2 bio-collections conserved in the Angers and Brest biologic resource centers. These collections were created in accordance with the law n° 2004-800 of 2004 relating to bioethics and n° 2012-300 of 2012 relating to research on the human person. This collection received a favorable opinion of the Committee for the Protection of Persons (CPP) of Angers West II and Brest. Prior to the inclusion of the biological collection, the patients signed a consent. In addition, patients’ clinical data collected in the present study received authorization from the “Commission Nationale Informatique et Libertés” (CNIL) on 04/29/2020 (n° ar20-0040v0). When Sanger sequencing analysis were performed, the patients signed a consent for investigate the germline status of variants detected after NGS analysis.

***NGS analysis***

Target next generation sequencing (NGS) used in the present study, use a custom targeted panel of 24 genes (list in supplemental data) was applied on DNA extracted from peripheral blood. DNA libraries, built with the Sureselect® target enrichment protocol (Agilent Technologies, Santa Clara, CA, USA), were paired-end with a NextSeq 500® Instrument (Illumina, San Diego, CA, USA). Data analysis used an in-house pipeline including trimmed reads alignment to the GRCh34 human reference genome, tumor variants detection by differents variants callers (GATK HaplotypeCaller, VarScan ), variant annotations with public databases (gnomAD, COSMIC, dbSNP, ClinVar) and in silico predictors in the case of unknown variants (SIFT, PolyPhen-2), using updated versions whenever available. Mutations were considered significant if they reached a good quality score, a minimum variant allele frequency of at least 2% and a minimum of 20 reads supporting the variant for hotspots or 50 reads for non-hotspot variants.

Variants were classified as followed :
*Type A variant* (pathologic variant) including non-sens or frameshift variants, or variants already described in myeloid neoplasms (COSMIC data base), as somatic variant ; or mutation with a functional effect demonstrated (including splicing).
*Type B variant* (likely pathologic) : Mutation not classified as pathogenic with a VAF <40% or >60% (likely somatic)
*Variant of Undetermined Significance* (VUS) : Mutation not classified as pathogenic or likely pathogenic with a VAF between 40 and 60 %.

## Results

137 patients from two centers (nA = 65 , nB = 72), diagnosed between January 1980 and December 2020, were included in our cohort.

***Demographic data***

Median age at diagnosis was 51 years old (18 - 86 years old) and 73% were females.
131 out of 137 patients underwent a bone marrow biopsy. Local histopathology was in favor of myeloproliferative neoplasm for 80 patients, not in favor for 46 patients, and not contributive for 5 patients. Clinical and laboratory features of the two groups, MPN vs not MPN, are displayed in Table 1. Thrombocytosis, leucocytosis and PNN *(polynucleosis ??)* were significantly different between the two groups (p=0.0008, p=0.0149 and p=0.0113 respectively). We observed no differences regarding history of thrombosis nor splenomegaly.


***Analysis of bone marrow biopsy***

118 biopsies were re-analysed by an expert group (GEBOM) leading to a reclassification of 23 patients. 12 biopsies initially in favor of a MPN were considered as not in favor ; whereas 6 biopsies initially not in favor were considered as in favor of a MPN. Two patients, diagnosed before 2016 with a BM biopsy in favor of essential thrombocytemia, were reclassified as premyelofibrosis.
*This reclassification had no impact on clinical and laboratory features.*

58% of patients (n=79) were treated with a cytoreductive agent : hydroxycarbamide (50 patients, 36%), interferon (17 patients, 12%), anagrelide (6 patients, 4%), pipobroman (5 patients, 4%) and melphalan (1 patient, 1%).

***Next generation sequencing results***

About molecular characterization, 56% (77 patients) of our cohort had no mutations. The remaining 60 patients had, in majority, 1 or 2 sequence variants (56/60). When retreiving variants with an allelic frequency between 40 and 60%, 33 patients ...

The most frequent sequence variants involved MPL, ASXL1, SH2B3, JAK2, TET2 and DNMT3A gene.

Mutations of MPL gene are identified in 14 patients, of which one constitutionnal variant (MPL K39N) and 3 likely constitutionnal considering the VAF, and 10 pathologic or likely pathologic variants .
Variant of ASXL1 gene are identified in 10 patients, nevertheless, for only 4 patients the mutation are considered as pathologic or likely pathologic.
Mutations of SH2B3 gene are identified in 9 patients, 3 variants of SH2B3 are considered as pathologic or likely pathologic.
Constitutionnal or likely constitutionnal mutations of JAK2 gene are identified in 8 patients. Pathologic variants of DNMT3A gene are detected in 6 patients.
Variant of TET2 gene are identified in 7 patients, of wich 5 pathologic or likely pathologic variants.

*(Graphique avec nombre de variants par patients et proportion des différents variants ?)*
Je suis d'accord avec toi, il sera difficile de s'y soustrère, meme si au final c'est sur qu'avoir 3 variants vus ça a moins de csq qu'un seule variant reconnu patho.

NGS confirmed 27 diagnosis, reclassified 22 patients and was non informative for 86 patients. For some patients, we couldn't analyze nails in order to confirm or infirm ... hence limiting our ...

***Impact of mutations detected by NGS on diagnosis***

Before NGS analysis, 76/131 patients are classified with ET or with NMP.

For 66% of the patients included, NGS was not contributed (87/131 patients), and no mutation was detected in 77 patients (59%). A Clonal marker was detected in 27 patients (15 patients with ET and 12 patients with thrombocytosis), inducing a reclassification of 13/131 (9%) patients to NMP. In 26/131 (20%) cases, NGS analysis confirm the diagnosis of NMP.

For 10 patients the NGS showed a constitutional mutation, suggesting a constitutionnal thrombocytosis, in large part because a variant of JAK2 were detected, and considered as germinal variant, suggesting a diagnosis compatible with e constitutional thrombocytosis.

impact du reclassement : graphiques de Damien

#titre
**gras**