A Comparative Study of Five Technologically Diverse CFTR Testing Platforms
2007; Elsevier BV; Volume: 9; Issue: 3 Linguagem: Inglês
10.2353/jmoldx.2007.060163
ISSN1943-7811
AutoresMonique Johnson, Marvin J. Yoshitomi, C. Sue Richards,
Tópico(s)Biosensors and Analytical Detection
ResumoMultiple cystic fibrosis (CF) testing platforms, using diverse and rapidly evolving technologies, are available to clinical laboratories commercially or for evaluation. Considerations when choosing a CF platform may include: sensitivity, specificity, accuracy, signal discrimination, ability to genotype, ability to reflex test, no calls/repeat rate, composition of mutation panel, hands-on time, start-to-finish time, integration into laboratory workflow, data analysis methods, flexibility regarding custom test design, and required instrumentation. Mindful of these considerations, we evaluated five technologically diverse CF platforms: 1) eSensor, an electronic detection assay system; 2) InPlex, a signal amplification methodology using a microfluidics card; 3) oligonucleotide ligation assay, an electrophoretic-based separation of amplicon-derived ligation-generated products; and two liquid bead arrays; 4) Signature, a direct hybridization assay using allele-specific capture probes; and 5) Tag-It, an assay using allele-specific primer extension and a universal microarray. A core of 150 samples, focusing on mutations in the American College of Medical Genetics/American College of Obstetricians and Gynecologists mutation panel, was tested throughout several runs for each platform. All of the platforms performed comparably in respect to sensitivity, specificity, and no-call rate. As our results indicate, consideration of all of the parameters evaluated may be useful when selecting the most appropriate platform for the specific setting. Multiple cystic fibrosis (CF) testing platforms, using diverse and rapidly evolving technologies, are available to clinical laboratories commercially or for evaluation. Considerations when choosing a CF platform may include: sensitivity, specificity, accuracy, signal discrimination, ability to genotype, ability to reflex test, no calls/repeat rate, composition of mutation panel, hands-on time, start-to-finish time, integration into laboratory workflow, data analysis methods, flexibility regarding custom test design, and required instrumentation. Mindful of these considerations, we evaluated five technologically diverse CF platforms: 1) eSensor, an electronic detection assay system; 2) InPlex, a signal amplification methodology using a microfluidics card; 3) oligonucleotide ligation assay, an electrophoretic-based separation of amplicon-derived ligation-generated products; and two liquid bead arrays; 4) Signature, a direct hybridization assay using allele-specific capture probes; and 5) Tag-It, an assay using allele-specific primer extension and a universal microarray. A core of 150 samples, focusing on mutations in the American College of Medical Genetics/American College of Obstetricians and Gynecologists mutation panel, was tested throughout several runs for each platform. All of the platforms performed comparably in respect to sensitivity, specificity, and no-call rate. As our results indicate, consideration of all of the parameters evaluated may be useful when selecting the most appropriate platform for the specific setting. Cystic fibrosis (CF), caused by mutations in the cystic fibrosis transmembrane conductance regulator gene (CFTR), is one of the most common autosomal recessive diseases in the North American Caucasian population with an incidence of 1 in 2500 to 3300 live births.1Rommens JM Iannuzzi MC Kerem B-S Drumm ML Melmer G Dean M Rozmahel R Cole JL Kennedy D Hidaka N Zsiga M Buchwald M Riordan JR Tsui LC Collins FS Identification of the cystic fibrosis gene: chromosome walking and jumping.Science. 1989; 245: 1059-1065Crossref PubMed Scopus (2526) Google Scholar2Riordan JR Rommens JM Kerem B-S Alon N Rozmahel R Grzelczak Z Zielenski J Lok S Plavsic N Chou JL Drumm ML Iannuzzi MC Collins FS Tsui LC Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA.Science. 1989; 245: 1066-1073Crossref PubMed Scopus (5910) Google Scholar3Kerem B-S Rommens JM Buchanan JA Markiewicz D Cox TK Chakravarti A Buchwalk M Tsui LC Identification of the cystic fibrosis gene: genetic analysis.Science. 1989; 245: 1073-1080Crossref PubMed Scopus (3201) Google Scholar4Palomaki GE Haddow JE Bradley LA FitzSimmons SC Updated assessment of cystic fibrosis mutation frequencies in non-Hispanic Caucasians.Genet Med. 2002; 4: 90-94Crossref PubMed Scopus (33) Google Scholar The carrier frequency in non-Hispanic Caucasians is ∼1 in 25 to 30, in individuals of Ashkenazi Jewish descent 1 in 29, in Hispanics 1 in 46, in Africans 1 in 65, and in Asians 1 in 90.5Farrell PM Fost N Prenatal screening for cystic fibrosis: where are we now?.J Pediatr. 2002; 141: 758-768Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar6Grody WW Cutting GR Klinger KW Richards CS Watson MS Desnick RJ Laboratory standards and guidelines for population-based cystic fibrosis carrier screening.Genet Med. 2001; 3: 149-154Abstract Full Text Full Text PDF PubMed Scopus (391) Google Scholar7Watson MS Cutting GR Desnick RJ Driscoll DA Klinger K Mennuti M Palomaki GE Popovich BW Pratt VM Rohlfs EM Strom CM Richards CS Witt DR Grody WW Cystic fibrosis population carrier screening: 2004 revision of American College of Medical Genetics mutation panel.Genet Med. 2004; 6: 387-391Crossref PubMed Scopus (361) Google Scholar In 2001, the American College of Medical Genetics (ACMG) and the American College of Obstetricians and Gynecologists (ACOG) recommended that CF carrier screening be offered to reproductive couples.6Grody WW Cutting GR Klinger KW Richards CS Watson MS Desnick RJ Laboratory standards and guidelines for population-based cystic fibrosis carrier screening.Genet Med. 2001; 3: 149-154Abstract Full Text Full Text PDF PubMed Scopus (391) Google Scholar Of the 1300 CFTR sequence variants identified (), a screening panel of 25 CFTR mutations was initially selected based on inclusion criterion of mutations having a threshold of 0.1% frequency in the general US population.8Richards CS Bradley LA Amos J Allitto B Grody WW Maddalena A McGinnis MJ Prior TW Popovich BW Watson MS Standards and guidelines for CFTR mutation testing.Genet Med. 2002; 4: 379-391Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar In addition, they recommended reflex testing for R117H of the 5/7/9T polymorphic alleles in intron 8, as well as reflex testing for individuals homozygous for ΔF508, ie, testing for interference from benign variants I506V, I507V, and F508C. This panel was modified in 2004 and currently includes 23 CFTR mutations as well as the R117H and ΔF508 reflex tests.7Watson MS Cutting GR Desnick RJ Driscoll DA Klinger K Mennuti M Palomaki GE Popovich BW Pratt VM Rohlfs EM Strom CM Richards CS Witt DR Grody WW Cystic fibrosis population carrier screening: 2004 revision of American College of Medical Genetics mutation panel.Genet Med. 2004; 6: 387-391Crossref PubMed Scopus (361) Google Scholar Using this panel, the resulting carrier detection rate, for Ashkenazi Jews, Northern European Caucasians, Hispanics, African Americans, and Asians is 94 to 97%, 90%, 72%, 64 to 69%, and 49%, respectively.6Grody WW Cutting GR Klinger KW Richards CS Watson MS Desnick RJ Laboratory standards and guidelines for population-based cystic fibrosis carrier screening.Genet Med. 2001; 3: 149-154Abstract Full Text Full Text PDF PubMed Scopus (391) Google Scholar7Watson MS Cutting GR Desnick RJ Driscoll DA Klinger K Mennuti M Palomaki GE Popovich BW Pratt VM Rohlfs EM Strom CM Richards CS Witt DR Grody WW Cystic fibrosis population carrier screening: 2004 revision of American College of Medical Genetics mutation panel.Genet Med. 2004; 6: 387-391Crossref PubMed Scopus (361) Google Scholar8Richards CS Bradley LA Amos J Allitto B Grody WW Maddalena A McGinnis MJ Prior TW Popovich BW Watson MS Standards and guidelines for CFTR mutation testing.Genet Med. 2002; 4: 379-391Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar Frequency and ethnic distribution of additional CFTR mutations in the US population, with prevalence of AR553XΔF508/2789 + 5G>AR553X/G551DR347P/ΔF508ΔF508/R1162XΔF508/G85E621 + 1G>T711 + 1G>T/ΔF508ΔF508/711 + 1G>T621 + 1G>T/W1282XΔF508/G551D1898 + 1G>A/R117HΔF508/R117H621 + 1G>T/G542XΔF508/R347P621 + 1G>T/3120 + 1G>A3849 + 10kbC>TΔF508/G542XR117H2789 + 5G>AΔI507R117H/ΔF5082789 + 5G>A/ΔF508R117H/1898 + 1G>AG542X3659delCG542X/G542X1717 − 1G>AG542X/621 + 1G>T1717 − 1G>A/ΔF508I148TG542X/ΔF508G551DA455E/ΔF5083120 + 1G>A/621 + 1G>TG551D/ΔF508R560TNormal (100 archived samples)G551D/R553XW1282XR1162XW1282X/621 + 1G>TR1162X/ΔF508 Open table in a new tab We evaluated each platform with respect to sensitivity, specificity, accuracy, signal discrimination, ability to genotype, ability to reflex test according to ACMG recommendations, no calls/repeat rate, number of and specific mutations (relating to ethnicity) in the available panels, hands-on time, start-to-finish time, laboratory work flow (including DNA extraction methods), data analysis methods, availability of additional tests, flexibility in respect to custom test design, and required instrumentation. Table 2 summarizes our evaluation of the five platforms.Table 2Summary of Platform EvaluationeSensorInPlexOLASignatureTag-It% Concordance100100100100100*Data reported here were generated by a Tm Bioscience technician performing the assay in our laboratory. Modifications to the research use-only protocol (RUO) were made by the Tm technician. The RUO was the only available protocol at the time this study was performed. These modifications were made in an effort to minimize the previously high no-call rate and shorten the start-to-finish time and hands-on time. The modifications are as follows: 1) Exo-Sap denaturation was run 30 seconds instead of 15 minutes; 2) multiplex ASPE annealing was performed at 56°C instead of 52°C; 3) ASPE extension was run for 30 seconds instead of 1 minute; and 4) bead hybridization at 37°C was run 30 minutes instead of 1 hour.% No calls0.70.70.70.70.0*Data reported here were generated by a Tm Bioscience technician performing the assay in our laboratory. Modifications to the research use-only protocol (RUO) were made by the Tm technician. The RUO was the only available protocol at the time this study was performed. These modifications were made in an effort to minimize the previously high no-call rate and shorten the start-to-finish time and hands-on time. The modifications are as follows: 1) Exo-Sap denaturation was run 30 seconds instead of 15 minutes; 2) multiplex ASPE annealing was performed at 56°C instead of 52°C; 3) ASPE extension was run for 30 seconds instead of 1 minute; and 4) bead hybridization at 37°C was run 30 minutes instead of 1 hour.Signal to noise ratioNA†The concepts of signal and noise do not apply to these assays.NA†The concepts of signal and noise do not apply to these assays.NA†The concepts of signal and noise do not apply to these assays.10:1‡Signal to noise ratio was calculated as follows: for Tag-It, signal is the median fluorescence intensity (MFI) generated by an allele in the sample and noise is the MFI generated by allele in the no template control; for Signature, signal is the allele ratio (mutant signal/mutant signal + wild-type signal) and noise is the standard deviation of the allele ratio.20:1 to 100:1‡Signal to noise ratio was calculated as follows: for Tag-It, signal is the median fluorescence intensity (MFI) generated by an allele in the sample and noise is the MFI generated by allele in the no template control; for Signature, signal is the allele ratio (mutant signal/mutant signal + wild-type signal) and noise is the standard deviation of the allele ratio.Mutations§The following platforms were able to resolve a G551D/R553X compound heterozygote: InPlex, TagIt, Signature, and eSensor. The study evaluated only the ACMG/ACOG panel of 23 mutations (2184delA was not represented). Mutations in italics are unique to that IVD/ASR.Total: 23 ACMG/ACOG 23Total: 42 ACMG/ACOG 23 and V520F, 3876delA, 394delTT, R347H, I148T, 1078delT, 3905insT, S549N, Y122X, Y1092X S549R(T>G), 2183AA>G, S549R(A>C), D1152H, 3849 + 4A>G,E60X, Q493X, D1270N, Y1092X(C>G)Total: 32 ACMG/ACOG 23 and V520F, 3876delA, 394delTT, R347H, I148T, 1078delT, 3905insT, S549N/RTotal: 23 ACMG/ACOG 23Total: 40 ACMG/ACOG 23 and V520F, 3876delA, 394delTT, R347H, I148T, 1078delT, 3905insT, S549N, Y122X, S549R(T>G), 2183AA>G, Y1092X, 2307insA, A559T, 1898 + 5G>T, M1101K, S1255XReflex tests: are poly-T reflex tests masked or run separately?Masked¶Software for unmasking reflex test results was not available at time of evaluation; poly-T reflex test results were not evaluated.MaskedSeparatelySeparatelyMaskedDoes the assay detect interfering benign variants (I506V, I507V, F508C) in the case of unexpected ΔF508 homozygosity?NA∥The polymorphisms do not interfere with panel mutation genotyping in the eSensor method (information provided by the manufacturer). We did not test this aspect of the platform.NA**Because of the Invader chemistry, ie, the specificity with which the cleavage products are produced, the I506V and I507V polymorphisms do not interfere with genotyping (information provided by the manufacturer). We did not test this aspect of the platform. The F508C variant is assayed during the initial run with results masked.YesYesYesInput DNA range (ng)Information provided by manufacturer.10 to 60025 to 3501 to 1510 to 10002 to 200Extraction methods usedMagNa Pure LC DNA isolation kit 1- and Gentra Generation capture column kit-extracted samples were tested on each platform. No other extraction methods were tested.MP,GMP,GMP,GMP,GMP,GStart to finish time (hours)Time determined was averaged throughout several runs for 24 samples/run, excluding DNA extraction. Time calculations were based on use of a PTC-200 DNA engine thermal cycler (MJ Research/Bio-Rad). Tag-It start to finish time reflects protocol modifications indicated in *, ie, ASPE extension was run for 30 seconds instead of 1 minute; bead hybridization at 37°C was run 30 minutes instead of 1 hour.6 to 73.5 to 46 to 75 to 66.5 to 8Hands-on time (hours)Time determined was averaged throughout several runs for 24 samples/run, excluding DNA extraction. Time calculations were based on use of a PTC-200 DNA engine thermal cycler (MJ Research/Bio-Rad). Tag-It start to finish time reflects protocol modifications indicated in *, ie, ASPE extension was run for 30 seconds instead of 1 minute; bead hybridization at 37°C was run 30 minutes instead of 1 hour.2.5<11.511.5 to 2.5Number of sample transfers32224Ease of protocolAs determined by our laboratory based on number of steps in the protocol, tolerances within those steps, and number of sample transfers.21223Open platformFacilitates custom test development.NoNoYesYesYesRequired instrumentation specific to assayeSensor 4800GeNios or GeniosFL fluorometer, card bucket and clips, card sealerABI Prism 3100/3130 genetic analyzerLuminex 100xMAP systemLuminex 100xMAP systemIVD or ASRIVD***eSensor is FDA-cleared for carrier testing.ASRASRASRIVDNA, not applicable; ACMG, American College of Medical Genetics; ACOG, American College of Obstetrician and Gynecologists; MP, MagNa Pure LC DNA isolation kit 1; G, Gentra Generation capture column kit; IVD, In Vitro Diagnostic; ASR, analyte-specific reagent.* Data reported here were generated by a Tm Bioscience technician performing the assay in our laboratory. Modifications to the research use-only protocol (RUO) were made by the Tm technician. The RUO was the only available protocol at the time this study was performed. These modifications were made in an effort to minimize the previously high no-call rate and shorten the start-to-finish time and hands-on time. The modifications are as follows: 1) Exo-Sap denaturation was run 30 seconds instead of 15 minutes; 2) multiplex ASPE annealing was performed at 56°C instead of 52°C; 3) ASPE extension was run for 30 seconds instead of 1 minute; and 4) bead hybridization at 37°C was run 30 minutes instead of 1 hour.† The concepts of signal and noise do not apply to these assays.‡ Signal to noise ratio was calculated as follows: for Tag-It, signal is the median fluorescence intensity (MFI) generated by an allele in the sample and noise is the MFI generated by allele in the no template control; for Signature, signal is the allele ratio (mutant signal/mutant signal + wild-type signal) and noise is the standard deviation of the allele ratio.§ The following platforms were able to resolve a G551D/R553X compound heterozygote: InPlex, TagIt, Signature, and eSensor. The study evaluated only the ACMG/ACOG panel of 23 mutations (2184delA was not represented). Mutations in italics are unique to that IVD/ASR.¶ Software for unmasking reflex test results was not available at time of evaluation; poly-T reflex test results were not evaluated.∥ The polymorphisms do not interfere with panel mutation genotyping in the eSensor method (information provided by the manufacturer). We did not test this aspect of the platform.** Because of the Invader chemistry, ie, the specificity with which the cleavage products are produced, the I506V and I507V polymorphisms do not interfere with genotyping (information provided by the manufacturer). We did not test this aspect of the platform. The F508C variant is assayed during the initial run with results masked.†† Information provided by manufacturer.‡‡ MagNa Pure LC DNA isolation kit 1- and Gentra Generation capture column kit-extracted samples were tested on each platform. No other extraction methods were tested.§§ Time determined was averaged throughout several runs for 24 samples/run, excluding DNA extraction. Time calculations were based on use of a PTC-200 DNA engine thermal cycler (MJ Research/Bio-Rad). Tag-It start to finish time reflects protocol modifications indicated in *, ie, ASPE extension was run for 30 seconds instead of 1 minute; bead hybridization at 37°C was run 30 minutes instead of 1 hour.¶¶ As determined by our laboratory based on number of steps in the protocol, tolerances within those steps, and number of sample transfers.∥∥ Facilitates custom test development.*** eSensor is FDA-cleared for carrier testing. Open table in a new tab NA, not applicable; ACMG, American College of Medical Genetics; ACOG, American College of Obstetrician and Gynecologists; MP, MagNa Pure LC DNA isolation kit 1; G, Gentra Generation capture column kit; IVD, In Vitro Diagnostic; ASR, analyte-specific reagent. All platforms evaluated demonstrated excellent specificity and sensitivity (100% concordance) and acceptable no call rates (all ≤0.7%). The following platforms were able to resolve a G551D/R553D compound heterozygote: InPlex, Tag-It, Signature, and eSensor. Note that because all platforms were challenged with the same set of 150 DNA samples, only normal samples and samples with genotypes representing the ACMG/ACOG panel of 23 were tested; eSensor and Signature test only the ACMG/ACOG panel of 23. Both the OLA and Signature platforms require poly-T reflex tests to be run separately, whereas the InPlex, Tag-It, and eSensor platforms allow the poly-T reflex test to be run concurrently with other mutations and are masked until user chooses to unmask that data. As to the detection, via reflex testing, of ΔF508 interfering benign polymorphic variants (I506V, I507V, and F508C), Tag-It, Signature, and OLA require a separate run, whereas eSensor does not require reflex testing for these variants, and InPlex does not require reflex testing for I506V and I507V because, according to the manufacturers, these polymorphisms do not interfere with the hybridization of the wild-type probe on these platforms. We did not test this aspect of either assay. The InPlex probe set does include probes for the identification of F508C and are part of the initial run. The F508C result is masked in the absence of ΔF508. The start-to-finish time and hands-on time are in respect to processing of 24 samples (excluding DNA extrac-tion). We found InPlex to require the least time in both categories. Here, the reported times are based on use of a PTC-200 DNA engine thermal cycler (MJ Research/Bio-Rad). All platforms tested performed equally well using either MagNa Pure LC DNA Isolation Kit 1 or Gentra Generation capture column kit for DNA isolation from whole blood. The DNA input range, as stated by the manufacturer, is quite broad with the exception of the OLA platform. None of the platforms required DNA quantitation, ie, expected yields as described by the manufacturers of the DNA isolation kits were sufficient to use as estimates of DNA concentration. We determined ease of protocol based on number of steps, tolerances within those steps, and number of sample transfers. We found the InPlex platform to be of greatest ease, Tag-It to be the most complex, and the remaining three platforms falling in between these. The signal:noise ratio was not calculable for three of the panels, and therefore this factor was unfortunately not comparable among all platforms. All of the platforms require specialized instrumentation. With the exception of the eSensor, additional tests can be run using the same instrumentation. In addition, three platforms, Tag-It, Signature, and OLA. are considered to be open platforms, ie, enable development of cu
Referência(s)