TP53 ABERRATIONS Methodical considerations

Transcription

1 TP53 ABERRATIONS Methodical considerations Sarka Pavlova University Hospital and Masaryk University, Brno, Czech republic Belgrade March 16-17, 2018

2 TP53 gene in CLL: modes of inactivation mutation(s) without 17p deletion FISH neg 17p deletion without mutation FISH pos mutation & 17p deletion monoallelic mutation N=228 mutation/cn-loh N=52 (SNP array) Examine not only del(17p) by FISH but also TP53 mutation Karla Plevova

3 TP53 gene in CLL: KEEP IN MIND Despite some codons are mutated more frequently (hotspots), deleterious variants can occur IN ANY PART OF CODING SEQUENCE Leroy etal., Human Mutation 2014 Pospisilova and Pettitt, 2010 In CLL, TP53 mutations are (1) frequently subclonal, (2) clinically significant even if the second allele remains intact KNOW DETECTION LIMIT OF YOUR METHOD AND CANCER CELL FRACTION Landau et al., Cell 2013 SangerSeq detection limit Mutations can NEWLY OCCUR IN RELAPSE Pre-therapy Relapse 1 Pre-therapy Relapse 1 Relapse 2-n Malcikova et al., Leukemia 2015

4 Gene analysis: technical view Part I From clinician to lab: SAMPLING WHEN to sample? WHAT type of tissue? INFORMATION which should be provided WHICH CELLS to enrich for? DNA or RNA? WHICH GENE PARTS to analyze? METHOD? SANGER SEQ: protocol and analysis? NGS: what to consider? CORRECT VARIANT DESCRIPTION INTERPRETATION: does the variant affect function? Keep communicating Part III From lab to clinician: REPORTING clear and precise message from lab to clinician ALWAYS KEEP GENERAL STANDARDS OF LABORATORY PRACTICE: strictly avoid sample swap/cross-contamination at any phase and include appropriate controls

5 Part I From clinician to lab 1. WHEN to analyze TP53 during disease course? IF DECIDING ABOUT TREATMENT = before first therapy and before every subsequent therapy Not necessary if the information does not impact patient s management: mutation documented previously p53-independent therapy cannot be given (comorbidities, limited access)? if targeted therapy is given irrespective of TP53 status 2. WHAT type of tissue? PERIPHERAL BLOOD Optionally - if cancer cell fraction (CCF) is low in PB (SLL/CLL) BONE MARROW LYMPH NODE: fresh (frozen) FFPE not suitable unless specific methodology is established

6 WHICH CELLS to enrich for? MONONUCLEAR FRACTION works for majority of samples density gradient centrifugation using appropriate media, i.e. Ficoll or Histopaque 1077 Optionally for <60-70% lymphocytes: CD19+ separation is recommended e.g. RosetteSep or MACS, magnetic-activated cell sorting or FACS use sensitive method (deep NGS) and recalculate VAF CLL cells = Cancer cell fraction (CCF)

7 CD 19 CD 19 WHICH CELLS to enrich for? Cancer cell fraction (CCF) Mutation VAF CLL: WBC 14.7x10 9 /L lymphocytes 59% MONONUCLEAR CELLS (gradient centrifugation) B-CELLS (non B-cell depletion by RosetteSep) LEUKOCYTES Cancer Cell Fraction (CCF): CLL = 20% 1 % normal B-cells CLL 35% 2 % normal B-cells CLL >95% 21 % MNC cells CD 5 43 % CD5+ T-cells 1 % MNC cells p.y220c gdna 8% p.y220c gdna 21% CD 5 <1 % CD5+ T-cells CONSIDER CANCER CELL FRACTION either during sample processing or result interpretation (esp. CLL/SLL) INFORMATION ON BLOOD CELL COUNT parameters (at least WBC) provided with the sample may be helpful

8 NUCLEID ACID isolation and storage DNA of standard quality DNA storage (deep NGS): in TE buffer with low EDTA (0.1mM) at -20 C for longer periods avoid unnecessary dilution For NGS, low concentration or poor quality low amount of template molecules in the sample false negative and false positive results

9 HOW: which gene parts should be covered OPTIMUM: WHOLE CODING REGION = EXONS 2-11 Leroy etal., Human Mutation 2014; Pospisilova and Pettitt, 2010

10 HOW: which gene parts should be covered OPTIMUM: WHOLE CODING REGION = EXONS 2-11 Thierry Soussi, The TP53 Web Page Leroy etal., Human Mutation 2014; Pospisilova and Pettitt, 2010

11 HOW: which gene parts should be covered minimum: exons 4-10 Thierry Soussi, The TP53 Web Page ALWAYS INCLUDE SPLICE SITES intronic +/- 2 bp from intron/exon boundaries Splice site variants are pathogenic

12 METHOD OF ANALYSIS TP53 Sanger sequencing Sanger sequenci ng and NGS 28% Sanger seq, I am not NGS and sure other 1% 6% Next generation Sequencing - NGS Other only 2% NGS only 25% Sanger seq only 38% ERIC/Gilead survey Easy and cheap to establish for any lab using capillary sequencing Running few samples per month Fast Subclonal variants with 10-20% VAF may escape Incorrect quantification for some variants Easily detects all 10-20% mutations if properly established Quantitative High-throughput method large cohorts More genes/diseases at once in a gene panel Research on clonal evolution minor mutations <<10% can be caught Takes longer High-throughput = expensive or slow for low sample numbers or if not combined with other genes/diseases If low-vaf variants desired, higher coverage must be applied for all genes in the panel

13 NGS IN RESEARCH ON CLONAL EVOLUTION 01/2014 Before therapy 11/2015 Complete remission 06/2016 Progression R-Bendamustine 4x del(17p) neg p.g105d 2.5% VAF cell separation CLL 15% 96% TP53 mutations 70% VAF p.g105d 61% VAF + 6 other mutations 2.9%-0.5% del(17p) neg, cn-loh 80% TP53 mutations 76% VAF p.g105d 68% VAF + 6 other mutations 2.1%-0.5% 02/2014 Before therapy FCR 4x efekt PR 11/2016 Progression Minor clones may evolve into prevalent refractory clone under therapy pressure (but not always do) Landau et al., 2013; Rossi et al., 2014; Malcikova et al. Ibrutinib? del(17p) neg c.128delt 1.5% del(17p) neg c.128delt 4.2% p.r248q 1.5% Convergent mutated subclones occur in parallel Jehtwa et al., 2013; Malcikova et al Jitka Malcikova, Jana Kotaskova

14 TP53 Next Generation Sequencing Standalone assay or gene panel Amplicon-based or capture-based Unique molecular identifiers (UMI) may be helpful Minimum limit of detection Sanger seq, i.e. 10% or less Corresponding coverage at all analyzed positions (to reach at least 10 reads per variant) and DNA/allelic input (6 pg per human cell) FALSE NEGATIVE RESULTS Low/non-uniform coverage Low DNA input or quality sampling effect Inefficient capture low DNA input Variant excluded/not recognized by bioinformatics pipeline (suboptimal alignment or variant calling) FALSE POSITIVE RESULTS Background noise are higher than your expectations Sequencing errors PCR errors Non-proof reading polymerase DNA from FFPE False variant created or not excluded during bioinformatics pipeline library preparation sequencing bioinformatics pipeline manual inspection and interpretation

15 Sanger sequencing of TP53 gene SEQUENCING PROTOCOL: you may start at IARC p53 WEBSITE ALWAYS SEQUENCE BOTH FORWARD AND REVERSE STRAND IARC p53 website protocol: adaptable, can be modified number of PCR cycles some primers span rare polymorphic sites some primers very close to exon/intron boundary

16 Sanger sequencing of TP53 gene Carefully go through chromatogram not to overlook subclonal variants Free web-based software GLASS available via ERIC website Do not trust software completely and check the sequence manually

17 Variant interpretation: does the variant affect function? CHECK FUNCTIONAL IMPACT OF THE MUTATION IN GENE-SPECIFIC DATABASE (do not use dbsnp) TP53 web site/seshat tool TP53 IARC database p53.iarc.fr IF YOU HAVE DOUBTS (RARE or FUNCTIONAL VARIANT), REPEAT THE ANALYSIS FROM PCR TO EXCLUDE ANALYTICAL ERRORS

18 Variant interpretation: does the variant affect function? SEVERAL WELL-DESCRIBED SNPs / NEUTRAL GERMLINE EXONIC VARIANTS OCCUR IN HUMAN POPULATION:

19 Variant interpretation: does the variant affect function? MOST OTHER VARIANTS DETECTED BY SANGER ARE DELETERIOUS BUT, VARIANTS THAT PROBABLY DOES NOT AFFECT FUNCTION MAY RARELY OCCUR OFTEN GERMLINE - RARE SNPs! c.704a>g: p.asn235ser c.847c>t: p.arg283cys our cases: 4/4 germline no personal or family history of Li-Fraumeni syndrome NEVER REPORT THESE VARIANTS AS PATHOGENIC Ask for non-tumor DNA (buccal swab, T-cells etc.) WT R283C MUT R175H RGC p21 bax RGC p21 bax RGC p21 bax Adapted from Jagosova et al. Int J Oncol 2012

20 Part III From lab to clinician: REPORTING Reporting -clear and precise message from lab to clinician REPORT TEMPLATE see ERIC webpage and ERIC Recommendations on TP53 Analysis GIVE CLEAR RESULT - It is preferred not to include common polymorphisms in the report to physicians. REPORT VARIANTS DETECTABLE BY SANGER SEQUENCING AND VARIANTS PRESENT IN >10% VAF IF TESTED BY NGS Reporting variants between 5-10% VAF is acceptable only if explicitly stating that the clinical impact of minor subclonal mutations has not been conclusively documented.! Low VAF in sample with low CCF 5% VAF in sample with 20% CCF corresponds to 25% in sample with 100% purity

21 Part III From lab to clinician: REPORTING Reporting: variant description Use HGVS NOMENCLATURE: Report the cdna and protein level including reference sequence (LRG_321 NM_00546) NGS: pipelines do not often describe dup, ins and dels correctly, always inspect NGS: notice if appropriate reference sequence is used

22 Final note: Publishing and databases Publishing and scientific reporting in databases Do not change patient and sample IDs in various papers duplicities in databases Include genomic coordinates and reference genome List all variants, including synonymous and benign variants

23 THANK YOU FOR ATTENTION Jitka Malčíková Šárka Pospíšilová Michael Doubek Jana Kotašková Yvona Brychtová Lenka Juračková Anna Panovská Marcela Ženatová Jiří Mayer Lenka Kociánová clinical department Renata Hurníková staff Zuzana Bučková Karla Plevová Jana Šmardová Martin Trbušek Boris Tichý Nikola Tom Karol Pál cytogenetic and flowcytometry laboratory TP53 aberrations: Methodical considerations