Clinical Chemistry Siemens Point of Care - Urinalysis
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 53: 1084-1091, 2007. First published April 26, 2007; 10.1373/clinchem.2006.076497
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2006.076497v1
53/6/1084    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cronin, M.
Right arrow Articles by Watson, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cronin, M.
Right arrow Articles by Watson, D.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Cancer Diagnostics (since 2002)
(Clinical Chemistry. 2007;53:1084-1091.)
© 2007 American Association for Clinical Chemistry, Inc.

Cancer Diagnostics

Analytical Validation of the Oncotype DX Genomic Diagnostic Test for Recurrence Prognosis and Therapeutic Response Prediction in Node-Negative, Estrogen Receptor–Positive Breast Cancer

Maureen Cronina, Chithra Sangli, Mei-Lan Liu, Mylan Pho, Debjani Dutta, Anhthu Nguyen, Jennie Jeong, Jenny Wu, Kim Clark Langone and Drew Watson

Genomic Health, Inc., Redwood City, CA.

aAddress correspondence to this author at: Genomic Health, Inc., 301 Penobscot Dr., Redwood City, CA 94063. Fax 650-556-1132; e-mail mcronin{at}genomichealth.com.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Oncotype DXTM is a clinically validated, high-complexity, multianalyte reverse transcription–PCR genomic test that predicts the likelihood of breast cancer recurrence in early-stage, node-negative, estrogen receptor–positive breast cancer. The Recurrence ScoreTM (RS) provides a more accurate, reproducible measure of breast cancer aggressiveness and therapeutic responsiveness than standard measures. Individualized patient management requires strict performance criteria for clinical laboratory tests. We therefore investigated the analytical performance of the assay.

Methods: Assays used a pooled RNA sample from fixed paraffin-embedded tissues to evaluate the analytical performance of a 21-gene panel with respect to amplification efficiency, precision, linearity, and dynamic range, as well as limits of detection and quantification. Performance variables were estimated from assays carried out with sample dilutions. In addition, individual patient samples were used to test the optimized assay for reproducibility and sources of imprecision.

Results: Assay results defined acceptable operational performance ranges, including an estimated maximum deviation from linearity of <1 cycle threshold (CT) units over a ≥2000-fold range of RNA concentrations, with a mean quantification bias of 0.3% and CVs of 3.2%–5.7%. An analysis of study design showed that assay imprecision contributed by instrument, operator, reagent, and day-to-day baseline variation was low, with SDs of <0.5 CT.

Conclusion: The analytical and operational performance specifications defined for the Oncotype DX assay allow the reporting of quantitative RS values for individual patients with an SD within 2 RS units on a 100-unit scale.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Women with early-stage breast cancer have many treatment options, but diagnostic methods derived from traditional anatomic-pathology assessment of tumor are of limited value for predicting treatment outcomes (1). It is clear that some women are undertreated for their disease and others are overtreated. Emerging molecular genomic methods, such as molecular characterization of tumor tissue, are advancing clinical oncology diagnostics and improving patient outcomes by providing a rational basis for individualizing oncology treatment (2)(3).

Molecular-biomarker tests must be rigorously validated both analytically and clinically. Tests that simply dichotomize patient populations into broad categories do not translate easily into clinically useful tools. Clinically relevant tests provide precise, quantitative results that can be used to gauge the probabilities of success with available alternative courses of disease treatment (4). We have adapted methods commonly used to validate single-analyte laboratory tests to validate a complex multianalyte genomic diagnostic test, the Oncotype DX Breast Cancer Assay.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
gene selection for the oncotype dx assay
Review of published breast cancer biomarkers studies led to the identification of 250 gene candidates, which were evaluated in 3 clinical-association studies to find biomarkers related to the risk of breast cancer recurrence (5)(6)(7). Statistical analysis and modeling of these study results led to the selection of the 21 genes constituting the Oncotype DX assay panel (Table 1 ; for primers and probes for all 21 gene expression assays, see Table 1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol53/issue6). We chose 16 cancer-related genes for their consistent univariate association with clinical outcome in all 3 clinical association studies, and we selected 5 reference genes (ACTB, GAPDH, GUSB, RPLP0, and TFRC)1 for their consistently low variation in expression and lack of association with clinical outcome in the same clinical studies. To compensate for sample variation in extracted RNA attributable to tissue-fixation processes, specimen age, and other variables unrelated to gene expression and patient outcome, we used the mean expression of the 5 reference genes to normalize the expression of the 16 cancer-related genes before calculating the Recurrence ScoreTM (RS)2 (8)(9)(10)(11)(12).


View this table:
[in this window]
[in a new window]

  		Table 1. Oncotype DX assay panel of 16 cancer-related genes and 5 reference genes.1

samples
We defined the histopathologic characteristics of samples acceptable for Oncotype DX assay analysis during 3 preliminary clinical-association studies. Guidelines for selecting fixed paraffin-embedded tissue (FPET) samples for the Oncotype DX assay are consistent with those for selecting optimal tissue blocks for standard immunohistochemistry assays. Essentially, the block containing the greatest amount of invasive breast carcinoma that is morphologically consistent with the submitting diagnosis and having the least amount of noninvasive mammary epithelium is selected. Samples with metabolically active nontumor elements constituting >50% of the tissue have those elements dissected out before sample extraction.

The goal of analytical validation was to provide an assay process with a fully optimized and documented standard operating procedure. Our studies focused on characterizing individual genes to define assay amplification efficiencies, linearity, dynamic range, reproducibility, and limits of detection and quantification. To measure the realistic limits of assay performance uncomplicated by biological variability contributed by individual samples, we created a single pooled test sample that represented a range of high and low expression values across the 21-gene panel (details in the online Data Supplement).

The final set of analytical validation studies tested the performance limits of the RS with individual patient samples to evaluate comprehensive assay performance in a way that realistically reflected patient sample testing.

fpet extraction
RNA was extracted from 3 sections (10 µm thick) of each FPET block. Paraffin was removed by xylene extraction followed by ethanol washing. RNA was isolated from deparaffinized tissue with the MasterPureTM Purification Kit (Epicentre Biotechnologies) with DNase I treatment, as previously described (13).

rna quantification
Extracted RNA was quantitated with the RiboGreen® fluorescence method (Molecular Probes/Invitrogen) as described previously (13).

residual genomic dna
Extracts were evaluated for residual genomic DNA with a TaqMan® quantitative PCR assay for ß-actin DNA (13). Samples exceeding the DNA threshold were retreated with DNase I. Nine of the Oncotype DX gene primer sets (for BAG1, CCNB1, SCUBE2, GRB7, ERBB2, MKI67, MYBL2, PGR, and TFRC genes) span intron sequences and so do not detect genomic DNA; the 12 remaining Oncotype DX genes were tested for sensitivity to genomic DNA added into an FPET RNA pool. A threshold was set at 3 SDs below the lowest value of genomic DNA detected for this assay to function as a reverse transcription (RT) negative control.

rt
RT of the purified FPET RNA was carried out as previously described with the OmniscriptTM RT Kit (Qiagen) and combined random hexamer and gene-specific priming (13).

pcr
Quantitative RT-PCR analysis was done in 384-well plates in a 10-µL volume with cDNA equivalent to 2 ng RNA. The only exception was in the linearity study, in which sample input varied over RNA equivalents between 2–10 and 23 ng/reaction. Plate cycling was carried out with ABI PRISM® 7900HT instruments (Applied Biosystems) according to the manufacturer’s instructions, as described previously (details in the online Data Supplement) (13).

instrumentation and reagent calibration
We calibrated liquid-handling robots to an independent calibration system. The performance of the ABI PRISM 7900HT instruments was assessed by calculating a within-plate CV with a standardized ribonuclease P assay (Applied Biosystems).

data reference normalization and rs calculation
To make RT-PCR measurements comparable for clinical interpretation, the Oncotype DX assay normalizes gene expression measurements to the mean expression of 5 reference genes (ACTB, GAPDH, GUSB, RPLP0, and TFRC). We tested combinations of these genes for normalization performance before selecting the mean of all 5 genes for our standard reference normalization method. For each sample, normalized expression measurements are calculated as the mean cycle threshold (CT) for the 5 reference genes minus the mean CT of triplicate measurements for each individual gene. Normalized expression measurements are scaled from 0 to 15 units, where 1 unit reflects an ~2-fold change in RNA quantity.

After normalization, a single quantitative RS is calculated with expression values for 16 cancer-related genes, as has been described previously (14). The RS, scaled from 0 through 100, expresses the likelihood of distant breast cancer recurrence and is specific to the tumor of each individual patient.

measurement of assay amplification efficiencies
Amplification efficiency for each gene was approximated from RNA serial-dilution experiments recognizing the relationship: amplification efficiency {approx} (2–1/Slope – 1) x 100%, where the slope is estimated from the simple linear regression of CT measurements vs log2 RNA concentration (15)(16). Each of 15 sample dilutions underwent RT-PCR analysis twice with triplicate assays. Mean CT scores from both runs were averaged to assign a final expression value for each RNA concentration measurement used in the linear regression analysis.

measurement of assay linearity
For each gene in the assay panel, the range of linear assay response was assessed by comparing the proportionality of CT values relative to the input RNA concentrations calculated for each of the 15 pooled test sample dilutions used to measure amplification efficiency (see the online Data Supplement for sample dilution details). For this analysis, the polynomial method originally proposed by Krouwer et al. (17) and recommended Clinical and Laboratory Standards Institute guidelines (18) were applied. Specifically, orthogonal polynomial regression was used to obtain coefficients and associated tests of significance for the 1st (linear), 2nd (quadratic), and 3rd (cubic) order polynomials. Given that the log of the variance in CT measurement is proportional to the mean CT, error variance was modeled by use of a log-linear variance model. Model fits were obtained by use of the PROC Mixed procedure in SAS version 8.02.

In accordance with CLSI guidelines, any degree of nonlinearity in signal response was assessed by examining the SE of the regression and selecting the higher-order (nonlinear) polynomial model with the best fit. Specifically, at each input RNA concentration, the deviation from linearity (DL) was calculated as follows:

Formula
where p(xi) is the value of the best-fitting polynomial at RNA concentration xi. Consequently, DLi is a measure of the difference between the nonlinear model and the best-fit straight line at each of the calculated RNA concentrations.

measurement of quantitative bias and precision
Assessments of analytical bias are typically obtained by determining how much observed gene expression measurements differ from expected expression values derived from standard reference RNAs and definitive methods of analysis. Universal standard reference RNAs are not available for the 21 genes, nor are there universally accepted definitive analysis methods. Consequently, the quantitative bias and imprecision of predicted RNA concentrations relative to the calculated input RNA amounts in the series of pooled sample dilutions was estimated for each of the 21 genes. Specifically, for every CT measurement, an inverse prediction of RNA concentration was derived from the best fitting (nonlinear) polynomial calibration model derived during the linearity analysis (19). An estimate of the quantitative bias of the assay is given by the mean percent bias in prediction at each RNA concentration k (k = 1 ... 15); namely, by:

Formula
where xk is the calculated input RNA amount for the kth level and xijk is the predicted RNA amount obtained from inverse prediction for the ith plate (i = 1, 2), jth well (j = 1, 2, 3), and kth RNA concentration value.

As a measure of the quantitative precision of the assay, for each RNA concentration level, ANOVA was used to separate the total variability in the difference between individual predicted and calculated RNA concentrations into components of variance due to plate and well within plate, treating plates and wells within plates as random. For this purpose, a random effects model of the following form was applied to the individual differences between predicted and calculated RNA concentrations: Ydiff ij = µ + {alpha}j + {epsilon}i(j), where Ydiff ij is the difference between the predicted and actual RNA concentrations for the ith (i = 1, 2, 3) observation on the jth plate (j = 1, 2), µ is the overall mean difference, {alpha}j~iid N(0,{sigma}2{alpha}) is the effect of the jth plate (i = 1, 2), and {epsilon}i(j)~iid N(0,{sigma}2{epsilon}) is random error. We further assume that {alpha}s, ßs, and {epsilon}s distribute independently and that the distribution of error terms vary as a function of calculated RNA concentration level (e.g., variance increases near the limit of quantification of the assay). This information was used to obtain an estimate of within-plate CV in predicted RNA concentration at the kth RNA concentration level, namely:

Formula
where xk is the calculated input RNA concentration for the kth level and Formula2{epsilon} is the estimate of error variance derived from the ANOVA. Limit of detection and limit of quantification values were calculated for each gene as well (details in the online Data Supplement).

measurement of assay reproducibility
Reproducibility in RS and in the measurement of the expression of individual genes ensures that results remain comparable for patients over time and for different submitting pathology laboratories. Individual gene and RS reproducibility were measured by performing repeat analyses across multiple days, operators, RT-PCR plates, 7900HT instruments, and liquid-handling robots. Two operators obtained replicate CT measurements on 2 aliquots of a single RNA sample over the course of 5 days with 3 7900HT instruments and 2 liquid-handling robots. The study design has a G-efficiency of >50% (20), allowing estimation of all main fixed effects, including 7900HT instrument, liquid-handling robot, and operator. All plates within days were processed in randomized order.

Mixed-effect ANOVA was used to divide calculated total variability in observed CT measurement and RS into components of variance due to days, plates within days, and within plates by treating operator, 7900HT instrument, and liquid-handling robot as fixed effects. For each of the 21 Oncotype DX genes and RS, restricted maximum likelihood estimates of the components of assay variance were obtained.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
amplification efficiency
Amplification efficiencies for individual genes must be similar for Oncotype DX test reference normalization and RS calculation procedures to be valid (15)(16). In an optimized PCR mixture without inhibitors or reagent limitations, amplification efficiency is an inherent characteristic of the primers and probes, so under a standard set of conditions, assay sets can be directly compared for relative amplification efficiencies. Although individual patient samples may affect amplification efficiency, the influence is on all of the transcripts, leaving the relationships among the genes unaffected. Amplification efficiencies for the 16 cancer-related genes range from ~75% to 112%, whereas mean efficiency was ~88% for the reference genes (Table 2 ). Some values exceeded 100% because of inherent assay error throughout the cumulative analysis of the sample-dilution series. The 95% lower confidence bounds on the amplification efficiencies reflect model error as well as inherent assay error. The results were within the experimental error expected for this study design and were consistent among all 21 gene assays.


View this table:
[in this window]
[in a new window]

  		Table 2. Summary of amplification efficiency results for all 21 genes in the Oncotype DX assay.

assay linearity and dynamic range
For each gene, assay linearity was assessed by comparing the proportionality of CT measurements with respect to calculated RNA concentrations over the series of sample dilutions used to measure amplification efficiency. The quadratic and cubic polynomial terms were nonsignificant (P >0.05) for 6 genes (ACTB, BAG1, CD68, ESR1, RPLP0, and TFRC), indicating linear performance over the entire RNA range tested (2–10 to 8 ng). Another 6 genes (CCNB1, CTSL2, GSTM1, GUSB, MKI67, and PGR) were linear over more restricted RNA concentration ranges. For the remaining genes, the deviation from the linear model was estimated to be within the prespecified acceptance criterion of 1 CT unit. Based on the prespecified CLSI criterion, each of the 21 genes in the breast cancer–recurrence gene panel had an estimated maximum deviation from linearity of 1 CT over at least an 11-log2 (>2000-fold) concentration range (Table 3 ).


View this table:
[in this window]
[in a new window]

  		Table 3. Assay linearity and dynamic range results for each of the 21 genes in the Oncotype DX assay.

At the highest RNA concentration tested in the series, 8 ng RNA/reaction, CT values for the 21 genes varied over an ~8-fold range (256-fold concentration range). Consequently, the lowest-expressing genes (GUSB and CTSL2) were studied over a more limited concentration range than high-expressing genes (Table 3Up ). Low-expressing genes in this pooled sample would be expected to demonstrate linear performances similar to highly expressed genes at higher RNA concentrations. A linear assay response was seen for all genes over the expression range typically experienced in clinical samples.

assay quantitative bias and precision
The quantitative nature of the Oncotype DX RS depends directly on analytical accuracy and precision of measurement for each component of the 21-gene assay. Analytical accuracy could not be assessed in the absence of standard reference materials for each of the 21 analytes, so quantitative bias was measured by comparing predicted RNA concentrations for each gene to expected RNA concentrations calculated for a range of sample dilutions. At every CT measurement, an inverse prediction of RNA concentration was derived from the best-fitting polynomial calibration model derived from the linearity study (19). Assay bias is expressed as mean percentage deviation from the calculated value in prediction at each RNA concentration. At the 2-ng/well value used for the Oncotype DX assay, the estimated deviation from the expected value in the predicted RNA concentration ranged from –10% to 6%, with an estimated mean deviation from the expected value of –0.3% for the 16 cancer-related genes. For the reference genes, the mean percentage deviation from the expected value was 0.7%, indicating >99% mean quantitative correctness at this assay condition (Table 4 ).


View this table:
[in this window]
[in a new window]

  		Table 4. Quantitative bias and imprecision of predicted RNA concentration for all 21 genes in the Oncotype DX assay at the standard assay RNA input of 2 ng/reaction.

Each gene was analyzed for analytical imprecision of measurement. ANOVA was used at each RNA concentration to estimate the total variance in predicted RNA concentration derived from the inverse calibration model vs the actual RNA concentration. This information was used to estimate the CV in the predicted RNA concentration at each known RNA concentration. For the standard reaction, the imprecision of measurement for the 16 cancer-related genes on the RNA concentration scale had a CV of 5.7%, whereas the reference genes had a CV of 3.2%. All values were well within the prespecified acceptance limit of 20% (Table 4Up ).

day-to-day reproducibility
SDs in CT measurements varied from ~0.06 to 0.15 CT units for each of the 21 genes, and the upper bounds on 2-sided 95% CIs for the CVs were all within 10%, indicating a high degree of precision and reproducibility in the assay (Table 5 ). These SD and CV values are for the estimates of total variance. The between-day SD values were close to 0 for all 21 genes. A maximum SD of 0.15 at a CT of 30 translates to a CV of 0.5%. At this level of precision it is possible to reliably distinguish a 15% change in expression for specific genes.


View this table:
[in this window]
[in a new window]

  		Table 5. Analytical reproducibility for normalized expression measurements and restricted maximum likelihood (REML) estimates of the variance components for all 21 genes and RS in the Oncotype DX assay.

Additionally, pairwise differences in (least-squares) mean CT values between operators, liquid-handling robots, and 7900HT instruments were calculated. The largest differences between operators, as well as between liquid-handling robots and 7900HT instruments, were <0.5 CT units for each of the 21 Oncotype DX genes (data not shown).

monitoring and controlling assay performance in the clinical reference laboratory
Clinical validation of the Oncotype DX Breast Cancer Assay was conducted after validating assay analytical performance with quality-control measures established from the results of analytical validation studies. The entire assay process, including process controls with associated performance-acceptance limits, was documented as a series of standard operating procedures that provide the basis for the current reference laboratory operation.

A standard RNA control sample is assayed at least once per batch of patient samples (≤46 samples), and PCR controls are run in every assay plate to verify that the process and reagents continue to perform within specified ranges. RT-PCR failures, identified by analyzing the amplification curve from every assay well, are excluded from analysis. Expression values are assigned when at least 2 of 3 assay wells provide acceptable RT-PCR results. All 21 genes must have an expression value assigned for an RS to be calculated and reported.

Process monitoring in our clinical reference laboratory shows that RS reproducibility remains very high. Repeat testing with deidentified patient samples shows a cumulative SD of <2 RS units on a 100-unit scale, which represents all sources of process variation (Table 6 ).


View this table:
[in this window]
[in a new window]

  		Table 6. SD of the normalized expression measurement for the Oncotype DX assay, representing cumulative sources of process imprecision shown for each gene and the RS.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The Oncotype DX RS, developed to quantify likelihood of distant breast cancer recurrence for individual patients, has been clinically validated in multiple studies (14)(21)(22). Each clinical study successfully met prospectively defined statistical endpoints after analytical validation studies allowed standardization of the underlying RT-PCR–based laboratory test (4)(23).

This genomic diagnostic test is highly complex compared with more traditional clinical chemistry tests. Because it requires 21 quantitative measurements rather than the measurement of a single analyte, the assay does not conform to standard assay-validation formats. Nevertheless, successful analytical validation was critical to our goal of reporting quantitative results for individual patients. We defined a validation process analogous to validation methods for single analyte clinical assays to characterize assay amplification efficiency, linearity, quantification limitations, dynamic range, analytical precision, and reproducibility performance of the RT-PCR process that underlies the Oncotype DX assay for individual genes and for the assay as a whole as reflected by the RS (24).

Diagnostic gene expression assays fall into 2 principal categories (25). Both types test multiple analytes but diverge from one another in the quantitative analytical resolution they are able to achieve. One type of diagnostic test yields gene expression profiles for thousands or tens of thousands of candidate analytes by means of technologies such as hybridization microarrays, bead arrays, and protein mass spectrometry (26)(27)(28)(29)(30)(31)(32). Initially large analyte sets ultimately may be reduced to smaller sets that classify patients nearly as well as the original large test panel. Results obtained thus far indicate the challenge in validating this type of test for clinical utility, because such complex sets of targets—each with a relatively low precision and correctness of measurement—lend themselves to overfitting in clinical-development and -validation studies (33)(34). Reproducibly measuring large panels is difficult, and such measurements typically produce high CVs. This imprecision erodes the quality of patient classifications, thereby limiting the value of these assays in clinical decision-making (35)(36)(37).

The Oncotype DX assay is a prototype for an alternative type of genomic diagnostic test. From hundreds of candidate genes quantitatively evaluated for association with the clinical outcomes of individual patients during a series of clinical-development studies, we selected an optimal set of RT-PCR assays for high overall performance in predicting the outcome of individual patients. In contrast to tests that dichotomize patients into general populations by matching classification profiles, the output from this test is individualized; a quantitative value places a patient precisely within a defined continuum of clinical outcomes on the basis of the results of clinical-validation studies.

Maintaining consistent, predictable RT-PCR assay performance for individual genes throughout clinical validation allowed assay-performance effects to be reliably differentiated from true patient variability. Consequently, each biomarker was repeatedly confirmed as important in predicting patient outcome while a foundation was concurrently established for continuous monitoring of assay quality in the clinical reference laboratory setting. Because key assay-performance metrics are quantitatively monitored, patients can be assured that their clinical reports are reliable and remain consistent with clinical-validation experience. This approach to analytical validation for a high-complexity genomic diagnostic test has been demonstrated to be useful by more than 3 years of successful operation in the clinical reference laboratory setting. Routine monitoring of the ongoing performance of the test shows that it continues to perform within the originally defined analytical conditions and to give reliable results of breast cancer risk to patients and their physicians.


   Acknowledgments
 
We gratefully acknowledge assistance from Jay Snable, Julie Ballard, and Heidi Cheng in carrying out these studies; Jerry Hurst for helpful discussions; Stephanie Butler and Raymond Chao for assistance with documentation; and Joff Baker and Steve Shak for their critical review of the manuscript.

Grant/funding support: None declared.

Financial disclosures: All authors are employees of Genomic Health, Inc.


   Footnotes
 
1 Human genes: BAG1, BCL2-associated athanogene; BCL2, B-cell CLL/lymphoma 2; CCNB1, cyclin B1; CD68, CD68 molecule; SCUBE2, signal peptide, CUB domain, EGF-like 2; CTSL2, cathepsin L2; ESR1, estrogen receptor 1; GRB7, growth factor receptor-bound protein 7; GSTM1, glutathione S-transferase M1; ERBB2, v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma derived oncogene homolog (avian); MKI67, antigen identified by monoclonal antibody Ki-67; MYBL2, v-myb myeloblastosis viral oncogene homolog (avian)-like 2; PGR, progesterone receptor; AURKA, aurora kinase A; MMP11, matrix metallopeptidase 11 (stromelysin 3); BIRC5, baculoviral IAP repeat-containing 5 (survivin); ACTB, actin, beta; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GUSB, glucuronidase, beta; RPLP0, ribosomal protein, large, P0; TFRC, transferrin receptor (p90, CD71).

2 Nonstandard abbreviations: RS, Recurrence Score; FPET, fixed paraffin-embedded tissue; RT, reverse transcription; CT, cycle threshold.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Hayes DF. Prognostic and predictive factors for breast cancer: translating technology to oncology [Editorial]. J Clin Oncol 2005;23:1596-1597.[Free Full Text]
  2. Abramovitz M, Leyland-Jones B. A systems approach to clinical oncology: focus on breast cancer. Proteome Sci 2006;4:5.[CrossRef][Medline] [Order article via Infotrieve]
  3. Duffy MJ. Predictive markers in breast and other cancers: a review. Clin Chem 2005;51:494-503.[Abstract/Free Full Text]
  4. Simon R. Roadmap for developing and validating therapeutically relevant genomic classifiers. J Clin Oncol 2005;23:7332-7341.[Abstract/Free Full Text]
  5. Cobleigh MA, Tabesh B, Bitterman P, Baker J, Cronin M, Liu ML, et al. Tumor gene expression and prognosis in breast cancer patients with 10 or more positive lymph nodes. Clin Cancer Res 2005;11:8623-8631.[Abstract/Free Full Text]
  6. Esteban J, Baker J, Cronin M, Liu M-L, Llamas MG, Walker MG, et al. Tumor gene expression and prognosis in breast cancer: multi-gene RT-PCR assay of paraffin-embedded tissue [Abstract]. Proc Am Soc Clin Oncol 2003;22:850.
  7. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, et al. Multi-gene RT-PCR assay for predicting recurrence in node negative breast cancer patients: NSABP studies B-20 and B-14 [Abstract]. Breast Cancer Res Treat 2003;82:A16.
  8. Andersen CL, Jensen JL, Orntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res 2004;64:5245-5250.[Abstract/Free Full Text]
  9. Simon R. Development and validation of therapeutically relevant multi-gene biomarker classifiers. J Natl Cancer Inst 2005;97:866-867.[Free Full Text]
  10. Szabo A, Perou CM, Karaca M, Perreard L, Quackenbush JF, Bernard PS. Statistical modeling for selecting housekeeper genes. Genome Biol 2004;5:R59.[CrossRef][Medline] [Order article via Infotrieve]
  11. Thellin O, Zorzi W, Lakaye B, De Borman B, Coumans B, Hennen G, et al. Housekeeping genes as internal standards: use and limits. J Biotechnol 1999;75:291-295.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  12. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002;3:RESEARCH0034..
  13. Cronin M, Pho M, Dutta D, Stephans JC, Shak S, Kiefer MC, et al. Measurement of gene expression in archival paraffin-embedded tissues: development and performance of a 92-gene reverse transcriptase-polymerase chain reaction assay. Am J Pathol 2004;164:35-42.[Abstract/Free Full Text]
  14. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med 2004;351:2817-2826.[Abstract/Free Full Text]
  15. . Applied Biosystems. User bulletin #2: relative quantitation of gene expression 2001:36pp Applied Biosystems Foster City. .
  16. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29:e45.[Abstract/Free Full Text]
  17. Krouwer JS, Schlain B. A method to quantify deviations from assay linearity [published erratum appears in Clin Chem 1994;40:1781]. Clin Chem 1993;39:1689-1693.[Abstract]
  18. . NCCLS. Evaluation of the linearity of quantitative measurement procedures: a statistical approach 2003:47pp NCCLS Approved guideline NCCLS document EP6-A (ISBN 1-56238-498-8). Wayne, PA. .
  19. Brown P. Measurement, Regression and Calibration 1993:210pp Oxford University Press Oxford, United Kingdom. .
  20. Atkinson AC, Donev AN. Optimum Experimental Designs 1992:528pp Oxford University Press New York. .
  21. Habel LA, Shak S, Jacobs MK, Capra A, Alexander C, Pho M, et al. A population-based study of tumor gene expression and risk of breast cancer death among lymph node-negative patients. Breast Cancer Res 2006;8:R25.[CrossRef][Medline] [Order article via Infotrieve]
  22. Paik S, Tang G, Shak S, Kim C, Baker J, Kim W, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol 2006;24:3726-3734.[Abstract/Free Full Text]
  23. Ransohoff DF. Rules of evidence for cancer molecular-marker discovery and validation. Nat Rev Cancer 2004;4:309-314.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  24. Bustin SA. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 2000;25:169-193.[Abstract]
  25. van’t Veer LJ, Paik S, Hayes DF. Gene expression profiling of breast cancer: a new tumor marker. J Clin Oncol 2005;23:1631-1635.[Free Full Text]
  26. Huang E, Cheng SH, Dressman H, Pittman J, Tsou MH, Horng CF, et al. Gene expression predictors of breast cancer outcomes. Lancet 2003;361:1590-1596.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  27. Ma XJ, Salunga R, Tuggle JT, Gaudet J, Enright E, McQuary P, et al. Gene expression profiles of human breast cancer progression. Proc Natl Acad Sci U S A 2003;100:5974-5979.[Abstract/Free Full Text]
  28. Perou CM, Jeffrey SS, van de Rijn M, Rees CA, Eisen MB, Ross DT, et al. Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc Natl Acad Sci U S A 1999;96:9212-9217.[Abstract/Free Full Text]
  29. Ramaswamy S, Ross KN, Lander ES, Golub TR. A molecular signature of metastasis in primary solid tumors. Nat Genet 2003;33:49-54.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  30. van de Vijver MJ, He YD, van’t Veer LJ, Dai H, Hart AA, Voskuil DW, et al. A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 2002;347:1999-2009.[Abstract/Free Full Text]
  31. Foekens JA, Atkins D, Zhang Y, Sweep FC, Harbeck N, Paradiso A, et al. Multicenter validation of a gene expression-based prognostic signature in lymph node-negative primary breast cancer. J Clin Oncol 2006;24:1665-1671.[Abstract/Free Full Text]
  32. Chemnitz JM, Driesen J, Classen S, Riley JL, Debey S, Beyer M, et al. Prostaglandin E2 impairs CD4+ T cell activation by inhibition of lck: implications in Hodgkin’s lymphoma. Cancer Res 2006;66:1114-1122.[Abstract/Free Full Text]
  33. Brenton JD, Carey LA, Ahmed AA, Caldas C. Molecular classification and molecular forecasting of breast cancer: ready for clinical application?. J Clin Oncol 2005;23:7350-7360.[Abstract/Free Full Text]
  34. Simon R, Radmacher MD, Dobbin K, McShane LM. Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification. J Natl Cancer Inst 2003;95:14-18.[Free Full Text]
  35. Daly TM, Dumaual CM, Dotson CA, Farmen MW, Kadam SK, Hockett RD. Precision profiling and components of variability analysis for Affymetrix microarray assays run in a clinical context. J Mol Diagn 2005;7:404-412.[Abstract/Free Full Text]
  36. Stec J, Wang J, Coombes K, Ayers M, Hoersch S, Gold DL, et al. Comparison of the predictive accuracy of DNA array-based multigene classifiers across cDNA arrays and Affymetrix GeneChips. J Mol Diagn 2005;7:357-367.[Abstract/Free Full Text]
  37. Ma C, Lyons-Weiler M, Liang W, LaFramboise W, Gilbertson JR, Becich MJ, et al. In vitro transcription amplification and labeling methods contribute to the variability of gene expression profiling with DNA microarrays. J Mol Diagn 2006;8:183-192.[Abstract/Free Full Text]



The following articles in journals at HighWire Press have cited this article:


Home page
 JCOHome page
S. S. Badve, F. L. Baehner, R. P. Gray, B. H. Childs, T. Maddala, M.-L. Liu, S. C. Rowley, S. Shak, E. D. Perez, L. J. Shulman, et al.
Estrogen- and Progesterone-Receptor Status in ECOG 2197: Comparison of Immunohistochemistry by Local and Central Laboratories and Quantitative Reverse Transcription Polymerase Chain Reaction by Central Laboratory
J. Clin. Oncol., May 20, 2008; 26(15): 2473 - 2481.
[Abstract] [Full Text] [PDF]

Home page
 The OncologistHome page
J. S. Ross, C. Hatzis, W. F. Symmans, L. Pusztai, and G. N. Hortobagyi
Commercialized Multigene Predictors of Clinical Outcome for Breast Cancer
Oncologist, May 1, 2008; 13(5): 477 - 493.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
F. Weber and C. Eng
Update on the Molecular Diagnosis of Endocrine Tumors: Toward -omics-Based Personalized Healthcare?
J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1097 - 1104.
[Abstract] [Full Text] [PDF]

Home page
 ANN INTERN MEDHome page
L. Marchionni, R. F. Wilson, A. C. Wolff, S. Marinopoulos, G. Parmigiani, E. B. Bass, and S. N. Goodman
Systematic Review: Gene Expression Profiling Assays in Early-Stage Breast Cancer
Ann Intern Med, March 4, 2008; 148(5): 358 - 369.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
J. A. Sparano and S. Paik
Development of the 21-Gene Assay and Its Application in Clinical Practice and Clinical Trials
J. Clin. Oncol., February 10, 2008; 26(5): 721 - 728.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
S. K. Lau, P. C. Boutros, M. Pintilie, F. H. Blackhall, C.-Q. Zhu, D. Strumpf, M. R. Johnston, G. Darling, S. Keshavjee, T. K. Waddell, et al.
Three-Gene Prognostic Classifier for Early-Stage Non Small-Cell Lung Cancer
J. Clin. Oncol., December 10, 2007; 25(35): 5562 - 5569.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2006.076497v1
53/6/1084    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cronin, M.
Right arrow Articles by Watson, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cronin, M.
Right arrow Articles by Watson, D.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Cancer Diagnostics (since 2002)

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS