
d821bfe4c7a9c743d32f872321a4302e.ppt
- Количество слайдов: 17
A Multicentre Technology Assessment of the Abbott Fragile X Assay CMGS Spring Meeting 3 rd April 2008 - Liverpool
Outline of test – key features • Abbott Fragile X kit (part no: 6 L 4301) – Analyte specific reagent (ASR) – Accurate allele sizing (<71+/-1; 71230+/-3) – Amplification and detection of large expansions (up to 645 repeats) – X specific/FMR allele ratio – potential to differentiate between hetero/homozygosity – Gender determination – Reduction in Southern Blotting
Testing workflow 10 -25 ng DNA Genemapper 5 -70 repeats (1 hr) 17 u. L PCR (4 hrs) 2% agarose gel (2 hrs) Genemapper 70 -250 repeats (2 hrs)
Aims of study • Test kit performance – Accuracy of allele sizing – Differentiation between hetero and homozygosity in females – Detection of large expansions/full mutations – Detection of mosaicism – Ease of use in diagnostic setting – Reproducibility
Design of study • 13 laboratories (10 UKGTN – 3 Eurogentest) – 8 ‘Testing labs’ used kit & provided samples – 5 ‘Sample labs’ provided samples • 577 samples analysed – 6 reference control samples • All 8 testing centres • Test for consistency and robustness – 196 retrospective samples • Analysed blind and unblind • Full range of genotypes – 375 prospective samples • Analysed alongside routine samples • Typical spread of genotypes in normal use
Results - reliability • Variability between centres • ~1/12 failure rate
Results - sizing Analysis of 6 sequenced alleles from reference control samples by 8 centres Range 23 to 73 repeats Slight tendency to overestimate (+0. 21 to +0. 93 bp) Significant differences between centres (ANOVA - F 35 = 20. 31; P = 8. 05 x 10 -9)
Results – sizing precision Precision of allele sizing +/-1. 96 standard deviations (SD) Precision within +/-1 repeat up to 73 repeats
Results – determination of hetero/homozygosity 30, FM 30, 30 X FMR-1 Abbott Molecular suggested TR/X ratio ranges X FMR-1
Variability in TR/X ratios – reference control samples Centre 05 TR/X = 0. 12 Centre 08 TR/X = 1. 25
Variability in TR/X ratios – prospective samples Significant overlap between TR/X ratio of homozygotes and heterozygotes at all centres TR/X too unreliable to be used diagnostically
Results – large expansions • 57/58 (98. 3%) of full mutation males detected on blind analysis • 48/54 (88. 9%) of full mutation females detected on blind analysis Visible most consistently on raw data (beyond largest size standard!)
Results - mosaicism Mosaicism consistently represented between centres However kit only detects size mosaicism NOT methylation mosaicism
Results – mosaicism • Concordance between in house genotype and kit low • 6/11 male mosaics identified • 2/3 female mosaics detected • 5 further female mosaics identified on blind testing
Results – mosaicism Male sample genotyped in house as Normal/Intermediate (N/I) mosaic Abbott genotype Intermediate (I) Close inspection of data showed a low level Normal (N) allele of correct size Is the ‘in house’ PCR assay selectively amplifying the normal allele more strongly? May account for some of the non-concordance between mosaicism reported on in house and Abbott testing
Conclusions • Accurately sizes alleles through critical Normal – Small premutation range • Routinely amplifies majority of full mutations (but not all) • TR/X ratio too variable to be used diagnostically to determine hetero/homozygosity • Size mosaicism only detected – may not correspond with ‘in house’ PCR/Southern data • Superior to ‘in house’ PCR alone -useful for urgent cases/PNDs • Use would not significantly reduce the Southern blotting workload • Full report available online www. ngrl. org. uk
Acknowledgments • Yogen Patel • Co-authors – D Barton, PA van Bunderen, J Duncan, J Dunlop, S Man, J Mac. Pherson, G Monaghan, J Mc. Luskey, G Norbury, H Powell, V Race, M Sweeney, E Thompson, R Treacy, MM Weiss, N Williams, HE White, B Wymer • Participating Laboratories – Birmingham, Cambridge, Dublin, Edinburgh, Glasgow, GOS, Leiden, Leuven, Newcastle, NGRL(Wessex), Oxford, Sheffield • Abbott Molecular – Jonathan Bradshaw & John Norton
d821bfe4c7a9c743d32f872321a4302e.ppt