Presentation is loading. Please wait.

Presentation is loading. Please wait.

The role of the Oncotype DX® assay in breast cancer management

Similar presentations


Presentation on theme: "The role of the Oncotype DX® assay in breast cancer management"— Presentation transcript:

1 The role of the Oncotype DX® assay in breast cancer management
GHI10006_0511

2 Objectives Brief overview of the Oncotype DX® assay and reports
Review assay development strategy and supporting studies Technical feasibility studies Gene discovery and refinement studies Analytical validation studies Review clinical validation studies in women with breast cancer Prognostic studies Predictive studies Discuss future development plans

3 Case study presentation
A 55-year-old post-menopausal woman presents with an infiltrating ductal carcinoma Tumor size 1.0 cm ER/PR IHC positive HER2 IHC negative Sentinel lymph node negative Excellent overall health How should this patient be evaluated for treatment? What is her risk of disease recurrence? How likely is she to benefit from hormonal or chemotherapy? 3

4 Breast cancer treatment in the United States (2009)
Approximately 110,000 women with ER+, lymph node-negative breast cancer are diagnosed annually in the United States This represents ~50% of newly diagnosed patients today Many women are offered chemotherapy, but all do not receive substantial benefit Main point: Many women diagnosed with breast cancer are offered chemotherapy, but all do not benefit equally from this treatment. A robust assay that could characterize an individual’s risk of recurrence and benefit from treatment would enable physicians to better tailor therapeutic intervention. This type of assay can help inform decisions about patient management. For example, patients at high risk of recurrence and/or with large chemotherapy benefit may decide, together with their physicians, to receive appropriate chemotherapy, while those at low risk and/or with little chemotherapy benefit may decide, together with their physicians, to avoid the unwanted toxicities associated with chemotherapy. Better identification of disease markers is needed to help make therapeutic decisions

5 Prognostic & predictive markers utilized in breast cancer management
Prognostic (recurrence risk) Axillary node status Histologic type/grade Tumor size Patient age Lymphatic/Vascular invasion ER/PR status HER2 neu status Oncotype DX® test Predictive (treatment benefit) ER/PR status HER2 neu status Oncotype DX® test Main point: A host of factors are utilized in the management of patients with breast cancer. Many of these factors are prognostic; some are predictive. Some are both predictive and prognostic. Prognosis refers to the risk of disease recurrence if untreated. Prediction refers how well a patient will respond to a given therapy. ER, HER2, and the Oncotype DX® assay test are examples of factors that are both prognostic and predictive. Cianfrocca and Goldstein. Oncologist. 2004;9(6): ; Lonning PE. Ann Oncol. 2007;18(suppl 8):viii3-viii7. These markers can be used to estimate the risk of disease recurrence These markers can be used to predict treatment benefit Cianfrocca and Goldstein. Oncologist. 2004;9(6): ; Lonning PE. Ann Oncol. 2007;18(suppl 8):viii3-viii7.

6 The Oncotype DX® assay Quantitatively predicts the likelihood of breast cancer recurrence in women with newly diagnosed, early stage, ER+ invasive breast cancer Assesses the likely benefit from both hormonal therapy and chemotherapy Is recommended by both ASCO and NCCN clinical practice guidelines Main point: The Oncotype DX® assay quantitatively predicts the likelihood of recurrence and assesses the likely benefit from both hormonal therapy and chemotherapy in women with newly diagnosed, early stage invasive breast cancer. It is the only multi-parameter gene expression assay to show clinical utility in breast cancer, meaning that it has been shown to alter treatment decisions in a significant number of cases. It is the only multivariate genomic classifier recommended in both ASCO and NCCN clinical practice guidelines. Harris L, et al. J Clin Oncol. 2007;33(25): NCCN Clinical Practice Guidelines in Oncology. Breast Cancer. Version Available at: Accessed December 8, 2008. Harris L, et al. J Clin Oncol. 2007;33(25): NCCN, National Comprehensive Cancer Network Adapted from NCCN Practice Guidelines in Oncology – v

7 Oncotype DX® Report Samples
Oncotype DX report provides valuable information on: Clinical prognosis Predicted chemotherapy benefit Quantitative data on ER / PR / HER2 Node positive report contains an additional page with prognosis and predicted chemo benefit information specific to node-positive patients Main point: After processing the patients tumor block in our lab, the final product for the patient’s chart is the Oncotype DX® report. Page 1 contains prognostic information – the Recurrence Score® result as a measure of risk of 10-year distant recurrence, validated in the NSABP B14 study. Page 2 contains predictive information concerning chemotherapy benefit, validated in the NSABP B20 study. Page 3 contains both predictive and prognostic information for node-positive patients, as validated in the SWOG 8814 study. Page 4 contains quantitative ER, PR and HER2 single gene results.

8 Oncotype DX® Technology Development Overview
Technical Feasibility 2001 Gene Discovery & Refinement 2002 Analytical Validation 2002 Clinical Validation (prognostic) Main point: This slide shows the development overview for the Oncotype DX® assay. 2004 Clinical Validation (predictive) 2005 8

9 Oncotype DX® Technology Development Overview
Technical Feasibility Gene Discovery & Refinement Analytical Validation Main point: The first step in development process was technical feasibility. The development of a genomic classifier that would be widely adopted required the use of a robust method of gene expression analysis that could be done on formalin fixed paraffin embedded tissue. Clinical Validation (prognostic) Clinical Validation (predictive)

10 Purpose of technical feasibility studies
Technical feasibility studies were designed to assess: RNA yield and the quality of RNA after extraction from FPET tissues Gene expression differences and similarities between whole section and enriched tumor tissue sections To establish criteria for manual microdissection Gene expression heterogeneity within breast tumor tissues Assess within block and between block gene expression heterogeneity Selection of reference genes (important for normalization of pre-analytical factors) Delay to fixation, duration of fixation, fixative Main point: The purposes of the technical feasibility studies are summarized here. The purpose of technical feasibility studies are to: Assess the RNA yield and RNA quality obtained from the tumor tissue to be studied Assess for differences between whole sections and enriched tumor tissue and establish metrics for manual microdissection Assess for tumor heterogeneity, both within FPET block and between FPET block Select reference genes (account for all sources of pre-analytical variability such as delay to fixation, choice of fixative, duration of fixation)

11 Importance of manual microdissection
Example from study of 16 breast cancer blocks for ER expression Most cases show minimal differences in ER expression between WS and ET Some tumors contain significant amounts of non-tumor elements (e.g., biopsy cavities, skin, smooth muscle) which require manual microdissection Thus, if < 50% invasive carcinoma, manual microdissection is always performed 14 ER 12 r = 0.73, p = 0.001 10 ER Whole Section 8 6 4 Main point: This study highlights the importance of manual microdissection with regards to ER expression. The 16 breast cancer samples were both assayed as a whole section and then manually microdissected to enrich for invasive tumor. The figure shows the correlation between the whole section and the manually microdissected tumor. On the y-axis is quantitative ER from whole tissue sections. On the x-axis is quantitative ER from the manually microdissected tumor (enriched for tumor). This is a log base two scale where each unit corresponds to an approximate doubling of mRNA. The correlation is high between the 16 unique cases of whole section and enriched tumor and most of the cases fall on the line of identity; however, two samples that contained less than 50% invasive carcinoma (as compared to the other metabolically active elements, e.g., benign breast epithelium) show an increased ER measurement in the enriched tumor section (yellow dot). This helped to establish criteria for microdissection: in cases containing less than 50% invasive carcinoma manual microdissection is performed to enrich for assayable tumor. 2 2 4 6 8 10 12 14 ER Enriched Tumor ER = estrogen receptor Differences in non-tumor tissue may impact single gene assessment

12 Evaluation of tumor gene expression heterogeneity
Example of the differences in gene expression within & among 3 FPET blocks from two patients 3 Blocks from spatially distinct tumor regions ER/PR/HER2 IHC Status H&E PR+ ER+ HER2– Block Main point: A study was performed to evaluate the heterogeneity of tumor gene expression within and among tumor blocks obtained from patients. Two breast tumors with 3 spatially distinct tumor blocks were studied to assess for tumor gene expression heterogeneity both between and among FPET tumor blocks (N=2). This slide shows three H&E sections from the 3 FPET blocks from one of the patient tumors. Per study design, for each patient, the 3 FPET blocks were each sectioned at 5 different levels to obtain 35 microns of tissue (one for H&E staining and 6 × 5 micron unstained sections for analysis). RT-PCR for 92 genes was then performed on the RNA obtained from each of the 15 levels (3 blocks step sectioned ×5). The next slide shows two representative genes, ER and HER2, for the two patients. Poster presented at: United States-Canadian Academy of Pathology 93rd Annual Meeting; March, 2004; Vancouver, British Columbia. The three FPET blocks were step sectioned at five different levels Quantitative RT-PCR was performed on all 15 samples Poster presented at: United States-Canadian Academy of Pathology 93rd Annual Meeting; March, 2004; Vancouver, British Columbia.

13 Importance of standardized quantitative measurement using RT-PCR: minimal gene expression heterogeneity within & among tumor blocks Expression by RT-PCR (relative to reference genes) Expression by RT-PCR (relative to reference genes) Main point: The impact of tumor heterogeneity is minimal in tumor whole sections. First ER on the left: On the y-axis is the quantitative expression of ER measured by quantitative RT-PCR with the normalized log base two scale. The prespecified cut point for positivity of ER is 6.5 (cutpoint established by IHC data obtained from 3 previous published studies). On the x-axis are the two patients – each with 3 different tumor blocks. Plotted are the mean point estimates for gene expression as well as error bars reflecting the standard deviation from the 5 step sections obtained from each FPET block. The within block variability for the measurement of a single gene is less than 0.5 expression units, while the between block variability is less than 1 expression unit indicating that the impact of tumor heterogeneity is minimal when tumor whole sections are studied. Also note that in none of the cases did tumor heterogeneity result in a change of positive or negative status of a single gene result. HER2 is on the right: Notice the prespecified cut point for positivity for HER2 is 11.5 (cutpoint established by IHC & FISH data obtained from 3 previous published studies). Poster presented at: United States-Canadian Academy of Pathology 93rd Annual Meeting; March, 2004; Vancouver, British Columbia. Reproducibility Within block expression: standard deviation < 0.5 normalized expression units Among block expression: standard deviation < 1.0 normalized expression units Poster presented at: United States-Canadian Academy of Pathology 93rd Annual Meeting; March, 2004; Vancouver, British Columbia.

14 Oncotype DX® Technology Development Overview
Technical Feasibility Gene Discovery & Refinement Analytical Validation Clinical Validation (prognostic) Clinical Validation (predictive)

15 Oncotype DX® gene panel was developed from clinical trial evidence
250 cancer-related genes were selected Genes were analyzed for expression and relapse-free interval correlations across 3 independent studies of 447 breast cancer patients Study site N Node status ER status Treatment NSABP B-20, Pittsburgh, PA 233 N– ER+ Tamoxifen (100%) Rush University, Chicago, IL 78 ≥ 10 positive nodes ER+/– Tamoxifen (54%) Chemotherapy (80%) Providence St. Joseph’s Hospital, Burbank, CA 136 N+/– Tamoxifen (41%) Chemotherapy (39%) Main point: The Oncotype DX® gene panel was developed based on clinical trial evidence. In the first step of the gene selection process, 250 cancer-related candidate genes were selected from microarray data1-4, genomic databases, cancer literature, and molecular and cell biology experiments. The 250 cancer-related genes were selected from a variety of sources, including the scientific literature, microarray data and genomic databases were analyzed in a total of 447 patients from three separate studies: The National Surgical Adjuvant Breast and Bowel Project (NSABP) B-20 study: A multicenter study in which tissue was analyzed from 233 patients in a homogeneous patient cohort characterized by having histologically negative nodes, estrogen receptor–positive tumors, and treatment with tamoxifen and no other intervention.5 Rush Presbyterian-St. Luke’s Hospital: A single-site study in which tissue was analyzed from 78 patients, all characterized by having >10 positive nodes and treatment predominantly with chemotherapy and/or tamoxifen.6 Providence St. Joseph’s Hospital: A single-site study in which tissue was analyzed from 136 patients who were either node positive or negative, ER positive or negative, and treated with tamoxifen or chemotherapy.7 The expression levels of these 250 genes where correlated with relapse-free interval using univariate models across 3 independent studies of 447 breast cancer patients with known clinical outcome data. In addition to the 5 references genes, obtained from these three studies, the 16 cancer related genes which correlated most strongly with relapse-free interval, consistently across the study populations, were selected and were used in the development of the 21-gene profile and an algorithm for calculating a Recurrence Score® result. Gruvberger S, Ringner M, Chen Y, et al. Cancer Res. 2001; 61(16): Ramaswamy S, Ross KN, Lander ES, Golub TR. Nat Genet. 2003;33(1):49-54. Sorlie T, Perou CM, Tibshirani R, et al. Proc Natl Acad Sci U S A. 2001;98(19): van't Veer LJ, Dai H, van de Vijver MJ, et al. Nature. 2002; 415(6871): Paik S, Shak S, Tang G, et al. SABCS Abstract #16. Cobleigh MA, Bitterman P, Baker J, et al. Proceedings of ASCO Abstract #3415. Esteban J, Baker J, Cronin M, et al. Proceedings of ASCO Abstract #3416. From these studies, 21 genes were selected Paik et al. SABCS Abstract #16. Cobleigh et al. Clin Cancer Res. 2005;11(24 Pt 1): Esteban et al. Proc ASCO Abstract #3416. 15

16 Oncotype DX® Recurrence Score® result: calculated from 21 different genes
16 Cancer related genes Estrogen Proliferation HER2 Invasion Others ER PR Bcl2 SCUBE2 Ki-67 STK15 Survivin Cyclin B1 MYBL2 GRB7 HER2 Stromelysin 3 Cathepsin L2 CD68 GSTM1 BAG1 Main point: The final gene set used for the Oncotype DX® assay includes the 16 cancer genes identified in the clinical trials. 5 genes are in the proliferation group, 2 in the HER2 group, 4 in the estrogen receptor group, 2 in the invasion group, and 3 are unaligned. Some of the genes are well known in the breast cancer literature; others are relatively new. The 5 reference genes are used for normalizing the expression of the cancer-related genes. As was previously stated, it is important to note that there are other genes linked to breast cancer (e.g., the 250 candidate genes from which the 16 genes were selected). The 16 genes presented in this slide were selected for the Oncotype DX assay based on the three clinical trials, which demonstrated a consistent statistical link between these genes and distant breast cancer recurrence and the most robust predictive power across the three studies. Additional information on some of the 16 cancer related genes: MYBL2: The protein encoded by this gene, a member of the MYB family of transcription factor genes, is a nuclear protein involved in cell cycle progression. The encoded protein is phosphorylated by cyclin A/cyclin-dependent kinase 2 during the S-phase of the cell cycle and possesses both activator and repressor activities. It has been shown to activate the cell division cycle 2, cyclin D1, and insulin-like growth factor-binding protein 5 genes. Transcript variants may exist for this gene, but their full-length natures have not been determined. STK15: The protein encoded by this gene is a cell cycle-regulated kinase that appears to be involved in microtubule formation and/or stabilization at the spindle pole during chromosome segregation. The encoded protein is found at the centrosome in interphase cells and at the spindle poles in mitosis. This gene may play a role in tumor development and progression. A processed pseudogene of this gene has been found on chromosome 1, and an unprocessed pseudogene has been found on chromosome 10. Multiple transcript variants encoding the same protein have been found for this gene. Scube2 can modulate the long-range action of Bmp-dependent Hedgehog signaling in the neural tube and somites. STMY3 (MMP11): Proteins of the matrix metalloproteinase (MMP) family are involved in the breakdown of extracellular matrix in normal physiological processes, such as embryonic development, reproduction, and tissue remodeling, as well as in disease processes, such as arthritis and metastasis. Most MMP's are secreted as inactive proproteins which are activated when cleaved by extracellular proteinases. However, the enzyme encoded by this gene is activated intracellularly by furin within the constitutive secretory pathway. Also in contrast to other MMP's, this enzyme cleaves alpha 1-proteinase inhibitor but weakly degrades structural proteins of the extracellular matrix CL2: The protein encoded by this gene, a member of the peptidase C1 family, is a lysosomal cysteine proteinase that may play an important role in corneal physiology. This gene is expressed in colorectal and breast carcinomas but not in normal colon, mammary gland, or peritumoral tissues, suggesting a possible role for this gene in tumor processes. GSTM1: Cytosolic and membrane-bound forms of glutathione S-transferase are encoded by two distinct supergene families. At present, eight distinct classes of the soluble cytoplasmic mammalian glutathione S-transferases have been identified: alpha, kappa, mu, omega, pi, sigma, theta and zeta. This gene encodes a glutathione S-transferase that belongs to the mu class. The mu class of enzymes functions in the detoxification of electrophilic compounds, including carcinogens, therapeutic drugs, environmental toxins and products of oxidative stress, by conjugation with glutathione. The genes encoding the mu class of enzymes are organized in a gene cluster on chromosome 1p13.3 and are known to be highly polymorphic. These genetic variations can change an individual's susceptibility to carcinogens and toxins as well as affect the toxicity and efficacy of certain drugs. Null mutations of this class mu gene have been linked with an increase in a number of cancers, likely due to an increased susceptibility to environmental toxins and carcinogens. Multiple protein isoforms are encoded by transcript variants of this gene. BAG1: The oncogene BCL2 is a membrane protein that blocks a step in a pathway leading to apoptosis or programmed cell death. The protein encoded by this gene binds to BCL2 and is referred to as BCL2-associated athanogene. It enhances the anti-apoptotic effects of BCL2 and represents a link between growth factor receptors and anti-apoptotic mechanisms. At least three protein isoforms are encoded by this mRNA through the use of alternative translation initiation sites, including a non-AUG site Paik et al. N Engl J Med. 2004;351: 5 Reference genes Beta-actin GAPDH RPLPO GUS TFRC Paik et al. N Engl J Med. 2004;351:

17 Oncotype DX® Recurrence Score® result calculation and risk categories
× HER2 Group Score – 0.34 × Estrogen Group Score × Proliferation Group Score × Invasion Group Score × CD68 – 0.08 × GSTM1 – 0.07 × BAG1 Risk group Low risk Recurrence Score < 18 Intermediate risk High risk ≥ 31 Main point: The Recurrence Score® result is calculated from the expression results for each of the 16 cancer-related genes by the equation shown in this slide. The Recurrence Score result ranges from 0 to 100. Expression for each individual gene is measured on a scale from 0 to 15, where an increase of 1 unit generally reflects a 2-fold increase in RNA quantity. Although the coefficients for each gene or gene group influence the Recurrence Score result, the quantitative expression for each gene can have a dominant effect. For example, there is a 200-fold range of expression of ER in the quantitative RT-PCR assay. For individual tumors, the expression of any one gene can affect the Recurrence Score result to a large degree. Cutoff points for Recurrence Score risk groups were defined prior to the initiation of the validation study: A low-risk group with a Recurrence Score value of < 18 An intermediate-risk group with a Recurrence Score value between 18 and 30 A high-risk group with a Recurrence Score value of  31 Paik et al. N Engl J Med. 2004;351: Paik et al. N Engl J Med. 2004;351:

18 The Oncotype DX® Recurrence Score® result is a continuous predictor of recurrence risk
What is the 10-year probability of distant recurrence for a patient with a Recurrence Score of 30? Lower Risk Higher Risk Distant recurrence at 10 years Recurrence Score 30 = 20% risk of distant recurrence at 10 years Main point: The Oncotype DX® Recurrence Score® is a continuous predictor of recurrence risk. The likelihood of distant recurrence at 10 years increases continuously with increases in Recurrence Score value. Patients with a Recurrence Score result of 3 have a much lower risk of distant recurrence than patients with a Recurrence Score result of 17. The use of the Recurrence Score result as a continuous predictor provides the most accurate and precise estimates for the likelihood of distant recurrence at 10 years In the example presented in this slide, the patient has a Recurrence Score result of 30; therefore, the risk of a distant recurrence at 10 years for this patient is 20%. Paik S, Shak S, Tang G, et al. Use of a RT-PCR assay to predict the likelihood of breast cancer recurrence in node-negative, estrogen receptor–positive, tamoxifen-treated patients. In preparation. Recurrence Score Dotted lines represent 95% CI

19 Oncotype DX® Technology Development Overview
Technical Feasibility Gene Discovery & Refinement Analytical Validation Main point: The analytical validation of the Oncotype DX® technology required the construction of a lab, the development of SOPs and lab procedures, and the hiring and training of laboratory technicians. Clinical Validation (prognostic) Clinical Validation (predictive)

20 The Oncotype DX® assay is analytically validated
Analytical validation is the assessment of assay performance characteristics and the optimal conditions to generate accuracy, precision and reproducibility Elements of analytic validation Analytical sensitivity (limits of detection and quantitation) Assay precision and linear dynamic range Analytical reproducibility PCR amplification efficiency Sample and reagent stability Reagent calibration Instrument validation and calibration Main point: The Oncotype DX® assay has been analytically validated to give accurate, precise, and reproducible results across different reagent lots, operators, days of the week, and patient samples. Analytical validation is the assessment of assay performance characteristics and the optimal conditions to generate accuracy, precision and reproducibility. Chau CH, et al. Clin Cancer Res. 2008;14(19): Chau CH, et al. Clin Cancer Res. 2008;14(19):

21 Oncotype DX® uses RT-PCR technology
RT-PCR provides > 65,000-fold range of measurement Maximizes ability to discriminate the full range of gene expression differences among individual samples RT-PCR reactions can be repeated with high quantitative precision Provides required reliability for individualized reporting RT-PCR works well with RNA from formalin-fixed paraffin-embedded tissue Reporter Quencher Forward Primer R Probe Q Polymerization Reverse Primer R Q Strand displacement and cleavage of probe Main point: The Oncotype DX® assay uses real time PCR methodology to determine the expression levels of ER and HER2. Quantitative real time PCR measures gene expression by assessing mRNA levels. RNA is extracted from sections of formalin-fixed paraffin-embedded tissue. It is reverse transcribed into cDNA and quantified in real time based on fluorescene given off by the DNA probe sets. PCR is a mature technology that is in widespread clinical use. Cronin M, et al. Am J Pathol. 2004;164:35-42. Q R Polymerization completion and signal detection Cronin M, et al. Am J Pathol. 2004;164:35-42. 21

22 Oncotype DX® assay process
Standardized RT-PCR Optimized for the small RNA fragments extracted from fixed paraffin embedded tissue (FPET) Optimized to be robust with regard to sources of pre-analytic variability such as Delay to fixation Duration of fixation Fixative type Sample age Main point: RNA extracted from formalin-fixed paraffin embedded tissue is short, ranging from base pairs long. This figure presents RNA size analysis using microcapillary electrophoresis.1 The internal lanes represent profiles of RNA extracted from archival breast tumor FPET blocks and frozen section from the same tumor (1/30 of each sample, 1 L, was analyzed). The external lanes contain different sets of molecular weight RNA markers. The electropheresis results clearly indicate that formalin fixation does not destroy RNA, as originally thought, but yields RNA in small fragments base pairs long.1 Importantly, because this fragmentation is random and does not occur with any particular bias, the relative ratio among the genes remains constant, enabling quantitative analysis of RNA expression.1 Cronin M, Pho M, Dutta D, 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. Cronin et al. Am J Pathol. 2004;164:35-42

23 Normalization accounts for all sources of pre-analytic variability
Delays to fixation, duration of fixation, different fixatives and sample age can affect RNA quality Reference normalization compensates for these differences in sample processing and sample age Hollandes Main point: The use of the 5 reference or normalization genes allows for normalization of all these pre-analytic factors. This is an example of the ability of the 5 reference genes to normalize for fixative differences, in this example the difference between Hollandes and Formalin. Formalin

24 Oncotype DX® assay process steps
PRE-ANALYTIC Pathology review of the FPET sample by a Board Certified Anatomic Pathologist with Breast Expertise Determine whether manual microdissection for tumor enrichment is required (~40% of submissions are microdissected for tumor enrichment) ANALYTIC RNA extraction and quantitation (RiboGreen® method) qPCR test for residual genomic DNA Reverse transcription TaqMan PCR Data quality control POST-ANALYTIC Calculation of Recurrence Score® result Report preparation and approval

25 Pre-analytic processing: all FPET blocks are bar-coded before entering histology
Main point: All specimens, FFPE blocks, slides and tubes are bar-coded and tracked by our LIMS system from the moment they enter the GHI laboratory and they are tracked throughout the process all the way through to report generation.

26 Pre-analytic processing: all tumors assessed by surgical pathologists with breast expertise
Pathology review to assess: Is tumor present? Is there sufficient tumor? Main point: Even in this new era of molecular pathology, the role of the anatomic pathologist with strong microscopy skills is paramount. Each case is reviewed by our board certified pathologist to assess for the presence of sufficient invasive tumor and records information such as tumor subtype. A minimum of 2 mm of invasive tumor is necessary. In cases containing less than 50% invasive tumor, or if there is skin, skeletal muscle or biopsy cavity, manual macrodissection is always necessary. When necessary, the pathologist marks these cases for dissection

27 Patient samples are barcode tracked from submission to report
Main point: All samples are barcode-tracked by computer.

28 Automation is central to laboratory processes

29 Patient sample tracking: LIMS bar-coding integrates reagents and robots for tracking and process control Patient FPET sample RNA extraction Reverse transcription cDNA plates Reverse primer pool 96-well Plate 96-well Plate 96-well Plate QPCR master mix 96-well Plate SARP GEMTools Material manager Primers Oligo plate assembly Pool 96-well plate 384-well plate Main point: Bar-coding of specimens and use of automation allows for an efficient and accurate process. After the specimen is reviewed by the pathologist and passed into the analytical lab it undergoes RNA extraction. The RNA is then reversed transcribed into cDNA. The samples, master mix and primer probe sets are then laid out on 384 well plates to perform the analytical portion of the assay. Each of the 21 genes for each sample is run in triplicate. QPCR reaction plate Probe Assay detection system Material manager Tecan Robotics Plate layout template and assembly for primers and samples

30 Patient report delivery: automated output and delivery
Print FedEx Fax Web Report distribution service Main point: Delivery of the final patient report is automated and allows for the report to be viewed in multiple different formats including almost instantaneous reporting by fax, or viewing on our online portal. A paper copy of the report is also sent by express delivery. PDF report w/ electronic signature approval

31 Oncotype DX® Technology Development Overview
Technical Feasibility Gene Discovery & Refinement Analytical Validation Main point: Clinical validation of the assay requires that the test satisfy the predefined primary objective with a prespecified plan of analysis. Up until this point, we have identified a technically feasible method, but we still must prove that the test will measure what we think it will measure. The test must be clinically validated in the appropriate patient cohorts – prospective, randomized, multi-center trials such as those sponsored by one of the major cancer cooperative groups, the NSABP, ECOG, or SWOG. We must be able to demonstrate that the test satisfies the predefined primary objective with a prespecified plan of analysis. Dr. Richard Simon of the NIH points out that tests should be developed in one set of patients and validated in another. Dr. Simon goes on to point out that to be truly meaningful, a test should be validated in two different studies. Clinical Validation (prognostic) Clinical Validation (predictive)

32 Clinical validation of Oncotype DX® breast cancer assay in node-negative disease
Main point: The clinical validation of the Oncotype DX® breast cancer assay was performed in ER positive, node negative patients is the NSABP B-14 study, a landmark study which served to establish the efficacy of tamoxifen.

33 Oncotype DX® clinical validation: NSABP B-14
Objective: Prospectively validate Recurrence Score® result as predictor of distant recurrence in node-negative, ER+ patients Multicenter study with prespecified 21-gene assay, algorithm, endpoints, analysis plan Placebo—not eligible Randomized Tamoxifen—eligible Registered Tamoxifen—eligible Main point: A study was performed to clinically validate the prespecified 21-gene RT-PCR assay and Recurrence Score® algorithm as a predictor of the prospectively defined primary endpoint of distant recurrence-free survival in node-negative, estrogen receptor–positive patients treated with tamoxifen from the large multicenter NSABP Study B-14.1 The study protocol also defined the prespecified endpoints and analysis plan. The laboratory was blinded to the clinical outcomes. NSABP B-14 trial (original): 2800 N-, ER+ women were randomized 1:1 to tamoxifen or placebo in double- blind fashion between 1982 and An additional 1235 patients were registered to tamoxifen in the 10-month period following closure of the trial in 1988, making 2617 tamoxifen-treated women available.1 Paik S, Shak S, Tang G, et al. Use of an RT-PCR assay to predict the likelihood of breast cancer recurrence in node-negative, estrogen receptor–positive, tamoxifen-treated patients. In preparation. Paik S, et al. N Engl J Med. 2004;351:

34 Oncotype DX® clinical validation: NSABP B-14, Distant Recurrence
Distant recurrence over time 100% 10-Year rate of recurrence = 6.8%* 95% CI: 4.0%, 9.6% 90% 80% 10-Year rate of recurrence = 14.3% 95% CI: 8.3%, 20.3% 70% 60% 10-Year rate of recurrence = 30.5%* 95% CI: 23.6%, 37.4% Proportion without distant recurrence 50% 40% All Patients, n = 668 RS < 18, n = 338 30% Main point: The Oncotype DX® assay is clinically validated to predict the risk of distant recurrence at 10 years in patients with ER+, node-negative breast cancer treated with tamoxifen. The Recurrence Score® result was calculated for each patient: 51% of the patient population fell into the low-risk group (n = 338), 22% fell into the intermediate-risk group (n = 149), and 27% fell into the high-risk group (n = 181).1 This graph demonstrates the difference in distant relapse-free survival (DRFS) over time for the different risk categories. The DRFS for the high- and low-risk groups were statistically significantly different; the 10-year distant relapse-free survival for the low-risk category was 93% compared to 69% for the high-risk category.1 Paik S, Shak S, Tang G, et al. Use of an RT-PCR assay to predict the likelihood of breast cancer recurrence in node-negative, estrogen receptor–positive, tamoxifen-treated patients. In preparation. 20% RS 18-30, n = 149 10% RS ≥ 31, n = 181 P < 0.001 0% 2 4 6 8 10 12 14 16 RS, Recurrence Score® result Years *10-Year distant recurrence comparison between low- and high-risk groups: P < 0.001 34 Paik S, et al. N Engl J Med. 2004;351:

35 Analysis without Recurrence Score Analysis with Recurrence Score
Multivariate Cox proportional hazards regression of age, tumor size, tumor grade and Recurrence Score® value in relation to likelihood of distant recurrence (NSABP B-14) Variable Age at surgery Clinical tumor size Tumor grade Moderately differentiated Poorly differentiated HER2 amplification Estrogen-receptor protein 50-99 fmol/mg fmol/mg > 200 fmol/mg Recurrence Score Analysis without Recurrence Score P value HR 0.1 0.7 0.13 1.35 0.04 <0.001 1.87 5.14 0.89 1.04 0.23 0.38 0.9 0.71 0.78 0.97 - Analysis with Recurrence Score P value HR 0.22 0.76 0.38 1.19 0.15 < 0.001 1.55 3.34 0.06 0.51 0.32 0.72 0.94 0.75 0.9 1.02 2.81 Main point: Only the Recurrence Score® result and poor tumor grade are independently associated with recurrence. Shown here is a multivariate Cox proportional hazards analysis examining all the standard measures alone (including age, tumor size, tumor grade, HER2, and ER expression), and then the standard measures in combination with the Oncotype DX® Recurrence Score® result. Paik et al. N Engl J Med. 2004;351: Paik S, et al. N Engl J Med. 2004;351:

36 Oncotype DX® assay NSABP B-14 subgroup analysis

37 Oncotype DX® NSABP B-14: Recurrence Score® result subgroups by tumor grade
All patients N = 668 Well Moderate Main point: In all subgroups of tumor grade, the Recurrence Score® result provides important additional information about the likelihood of recurrence. This forest plot shows that for tumor grade, as determined by Pathologist A, the Recurrence Score result provides additional information that predicts the likelihood of recurrence. For tumors characterized as well differentiated, some have high Recurrence Score results and higher rates of distant recurrence. For tumors characterized as moderately differentiated, the Recurrence Score result provides additional information about the likelihood of recurrence. For tumors characterized as poorly differentiated, these tumors tend to do worse, but again the Recurrence Score result provides important additional information. These results indicate that the genomics of the tumor contain information that is not provided by standard measures such as tumor grade. They also indicate that the information provided by the standard measures should not be ignored. Paik S, et al. N Engl J Med. 2004;351: All patients Low risk (RS < 18) Int risk (RS 18-30) High risk (RS ≥ 31) Poor 20% % % 80% % % Distant recurrence-free at 10 years RS, Recurrence Score result Paik S, et al. N Engl J Med. 2004;351:

38 Oncotype DX® clinical validation: conclusions, NSABP B-14
Oncotype DX® Recurrence Score® result is validated as a predictor of recurrence in node-negative, ER+ patients Oncotype DX Recurrence Score performance exceeds standard measures (patient age, tumor size, and tumor grade) Oncotype DX Recurrence Score result (based on tumor gene expression) more accurately quantifies the risk of distant recurrence than do the NCCN guidelines (based on patient age, tumor size, and tumor grade) Main point: The Oncotype DX® Recurrence Score® result is validated as a predictor of recurrence in node-negative, ER+ patients. Using the prospectively defined 21-gene expression assay and endpoints, the Recurrence Score result predicted the likelihood of distant recurrence in node-negative, ER+ tamoxifen-treated breast cancer patients. The Recurrence Score result predicts the likelihood of distant recurrence with performance exceeding standard measures such as age, tumor size, and tumor grade in prognostic power and reproducibility. Overall, 50% of patients are reclassified by the Recurrence Score result when compared to NCCN or St. Gallen criteria (not shown). The Recurrence Score result (based on tumor gene expression) more accurately quantifies the risk of distant recurrence than do the NCCN guidelines (based on age, tumor size, and tumor grade). Although all patients enrolled in B-14 were later treated with tamoxifen, which has implications for interpretation of Recurrence Score results for individual patients, the tissue that was archived for analysis was taken from tumors before any systemic therapy was administered. Paik et al. N Engl J Med. 2004;351: Paik et al. N Engl J Med. 2004;351:

39 Oncotype DX® clinical validation: the Kaiser Permanente study
Study design Matched case-control Study population (N = 4964) Kaiser Permanente patients < 75 years old in 14 Northern California hospitals diagnosed with node-negative breast cancer between , no adjuvant chemotherapy Cases: Deaths from BC (n = 220) Controls: Randomly selected, matched on age, race, diagnosis year, KP facility, tamoxifen (n = 570) Data sources Cancer registry, medical records, archived diagnostic slides, and tumor blocks Main point: The Kaiser Permanente study was performed using a case control study design. Of the population of almost 5000 patients with node-negative breast cancer, 220 cases with death due to breast cancer and 570 matched controls were studied. Of those 220 cases, 55 were treated with tamoxifen, and 150 tamoxifen-treated matched controls were studied. This was a good test of the robustness of the assay, as it involved largely a community hospital–based patient population from multiple sites and many treating physicians. (albeit the study was conducted in one region of the country with physicians practicing in the same large multi-specialty group). Habel LA, et al. Breast Cancer Res. 2006;8(3):R25. Habel LA, et al. Breast Cancer Res. 2006;8(3):R25.

40 The Kaiser Permanente study: risk of BC death at 10 years: ER+, Tam-treated patients
Risk classification (Recurrence Score® group) 10-Year absolute risk1 Kaiser 10-Year absolute risk1 NSABP B-14 Low 2.8% 3.1% Intermediate 10.7% 12.2% High 15.5% 27.0% 1Based on methods by Langholz and Borgan, Biometrics 1997;53: The Recurrence Score result has now been shown to be strongly associated with risk of breast cancer-specific mortality among node-negative, ER+, Tam-treated patients participating in a clinical trial and among similar patients from the community setting. Results from our study suggest that combining Recurrence Score result, tumor grade, and tumor size provides better risk classification than any one of these factors alone. Main point: Results indicate that the Oncotype DX® assay Recurrence Score® result is predictive of breast cancer survival in the Kaiser Permanente population. Note the very low rate of breast cancer death, 2.8%, in the low-risk group at 10 years. This result is similar to the result in NSABP B-14, where 3% of low-risk patients died from breast cancer at 10 years. Also, as in NSABP B-14, more than half of the Kaiser Permanente population is in the low-risk group with a Recurrence Score <18. No formal statistical analyses were performed to compare the results of Kaiser and NSABP B-14. The primary reason was the difference in study designs (randomized trial versus case-cohort design) and the resulting differences in analysis approaches (Cox proportional hazards model versus conditional logistic regression modeling). The Kaiser study further validates the Oncotype DX breast cancer assay by showing that Recurrence Score result is strongly associated with risk of breast cancer death among LN-negative, ER+ patients in a community setting and confirms the data from the NSABP B-14 study. The Recurrence Score result should be used in conjunction with traditional factors to best make treatment decisions for patients. Habel LA, et al. Breast Cancer Res. 2006;8(3):R25. Habel LA, et al. Breast Cancer Res. 2006;8(3):R25.

41 Oncotype DX® Technology Development Overview
Technical Feasibility Gene Discovery & Refinement Analytical Validation Main point: The Oncotype DX® assay was validated as a predictive tool to help determine response to chemotherapy using another prospective multicenter trial from the NSABP, the B-20 study. Clinical Validation (prognostic) Clinical Validation (predictive)

42 Oncotype DX® clinical validation: NSABP B-20
Objective: To determine the relationship between Recurrence Score® value and chemotherapy benefit in node-negative, ER+ patients Multicenter study with prespecified 21-gene assay, algorithm, endpoints, analysis plan Tam + MF Randomized Tam + CMF Main point: The objective of this additional study of the NSABP B-20 patients was to determine the magnitude of the chemotherapy benefit (with MF or CMF) as a function of the 21-gene Oncotype DX® assay. Patients who were randomized in NSABP B-20 to tamoxifen or to tamoxifen plus either CMF or MF chemotherapy were eligible. The primary analysis was prespecified to compare the tamoxifen-treated patients with both chemotherapy arms combined. In secondary prespecified analyses, similar results were seen when the patients treated with CMF or with MF were examined separately. The analysis by NSABP shows the B-20 study subjects included in this study were similar to all B-20 patients in the cohort and the loss of cases was principally due to blocks never being collected. Fisher B, Dignam J, Wolmark N, et al. J Natl Cancer Inst.1997;89: Paik S, Shak S, Tang G. Expression of the 21 genes in the Oncotype DX® assay and prediction of clinical benefit from tamoxifen in NSABP study B-14 and chemotherapy in NSABP study B-20. SABCS Abstract #24. Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol. 2006;24: Tam Paik S, et al. J Clin Oncol. 2006;24:

43 High Recurrence Score® result correlates with greater benefit from chemotherapy (NSABP B-20)
1.0 0.9 Patients with high RS 28% absolute benefit from tamoxifen + chemotherapy 0.8 0.7 0.6 Proportion without distant recurrence 0.5 N Events All patients Tamoxifen + chemotherapy Tamoxifen 424 227 33 31 P = 0.02 N Events 4.4% absolute benefit from tamoxifen + chemotherapy 0.4 0.3 RS < 18 Tamoxifen + chemotherapy Tamoxifen 218 135 8 4 P = 0.61 Main point: The Recurrence Score® result shows that not all patients benefit equally from chemotherapy. These results recapitulate the seminal findings from the original B-20 study (Fisher B, Dignam J, Wolmark N, et al. J Natl Cancer Inst.1997;89: ), which supported the widespread use of adjuvant cytotoxic therapy for this population. The absolute benefit of chemotherapy at 10 years is 4.4%. Is this small overall benefit due to a little benefit in many or most of the patients, or is this small benefit due to a large benefit in a subset of the patients? To address this question, the magnitude of the chemotherapy benefit was examined for each of the Recurrence Score risk groups, showing that not all women benefit equally from chemotherapy. It appears that much of the benefit associated with CMF therapy in the B-20 study was a function of the risk reduction experienced by this high-risk cohort, which represents approximately 25% of the population. The high-risk patients have a large benefit (28% absolute) from the addition of chemotherapy. Patients in the low-risk group (Recurrence Score value <18) have a very low likelihood of distant recurrence. The difference between the tamoxifen alone and tamoxifen plus chemotherapy groups is not statistically significant, so the benefit of chemotherapy appears to be minimal, if any. Patients in the intermediate-risk group (Recurrence Score value 18-30), as expected, had a higher risk of distant recurrence than those in the low-risk group. The benefit of chemotherapy in the intermediate-risk patients does not appear to be large. Paik S, et al. J Clin Oncol. 2006;24: 0.2 RS 18-30 Tamoxifen + chemotherapy Tamoxifen 89 45 9 4 P = 0.39 0.1 RS ≥ 31 Tamoxifen + chemotherapy Tamoxifen 117 47 13 18 P < 0.001 2 4 6 8 10 12 Years RS, Recurrence Score result Paik S, et al. J Clin Oncol. 2006;24:

44 Recurrence Score® result can add prognostic discrimination not always provided by traditional prognostic factors Age 44% of patients < 40 years old had low Recurrence Score results (ie, there is a large fraction of younger patients for whom chemotherapy benefit may be minimal) Tumor size 46% of patients with large tumors (> 4 cm) had low Recurrence Score results Some patients with small tumors (< 1 cm) had intermediate or high Recurrence Score results Tumor grade Assessment by local pathologists revealed that, even for poorly differentiated tumors, 36% of patients had low Recurrence Score result Approximately 20% of poorly differentiated tumors still had a low Recurrence Score result Main point: The Recurrence Score® result can add prognostic discrimination not always provided by traditional prognostic factors. In the B-20 study, 44% of patients under 40 years old had low Recurrence Score® results. While overall, younger patients do worse and are probably more likely to benefit from chemotherapy, there is a large fraction of younger patients for whom the Recurrence Score result is low and benefits of chemotherapy may be minimal. If one looks at tumor size in the B-20 study, there are a number of patients with small tumors who have intermediate and high Recurrence Score results. In addition, if one looks at patients with large tumors, >4 cm, 46% of these patients had low Recurrence Score results. Tumor grade in the B-20 study as assessed by local pathologists (reflective of the kind of grading normally provided in community practice) showed that even for poorly differentiated tumors, 36% of these patients had low Recurrence Score results. While grading done at a central lab is more highly correlated with outcomes, 19% of poorly differentiated tumors still had a low Recurrence Score result. Paik S, et al. J Clin Oncol. 2006;24: Paik S, et al. J Clin Oncol. 2006;24:

45 NSABP B-20: Many younger patients have low Recurrence Score® results
41% % % % 14% % % % 44% % % % N= N= N= N=196 In the B-20 study, 44% of patients under 40 years old had low Recurrence Score® results . While overall, younger patients do worse and are probably more likely to benefit from chemotherapy, there is a large fraction of younger patients for whom the Recurrence Score result is low and benefits of chemotherapy may be minimal. The P values are derived from a Cochran-Mantel-Haenszel chi-square test for determining whether there is a relationship between Recurrence Score result and age stratified by RS risk group. Paik S, et al. J Clin Oncol. 2006;24: Paik S, et al. J Clin Oncol. 2006;24:

46 NSABP B-20: Many small tumors have Intermediate to High Recurrence Score® values
33% 20% % % 21% 46% N= N= N= N=24 Recurrence Score ≤1 cm cm cm >4 cm Clinical tumor size 20 40 60 80 100 P=0.001 64% % % 16% % % If one looks at tumor size in the B-20 study, there are a number of patients with small tumors who have intermediate and high Recurrence Score® results. In addition, if one looks at patients with large tumors, >4 cm, 46% of these patients had low Recurrence Score results. The P values are derived from a Cochran-Mantel-Haenszel Cochran-Mantel-Haenszel chi-square test for determining whether there is a relationship between Recurrence Score result and clinical tumor size stratified by RS risk group. Paik S, et al. J Clin Oncol. 2006;24: Paik S, et al. J Clin Oncol. 2006;24:

47 NSABP B-20: Significant proportion of high-grade tumors have low Recurrence Score® values
12% % % 16% % % 73% % % N= N= N=163 P<0.001 5% % % 12% % % 83% % % N= N= N=190 For the left figure, grade for B-20 reflects original assessment by local pathologists – not NSABP central lab as was reported for B-20. This reflects the kind of grading normally provided in community practice. These figures show that even for the poorly differentiated tumors, 36% of these patients had low Recurrence Score® results. In this example, while grading done at a central lab is more highly correlated with outcomes, 19% of poorly differentiated tumors still had a low Recurrence Score result. The P values are derived from a Cochran-Mantel-Haenszel chi-square test for determining whether there is a relationship between Recurrence Score result and grade stratified by RS risk group. Paik S, et al. J Clin Oncol. 2006;24: Paik S, et al. J Clin Oncol. 2006;24:

48 Tamoxifen benefit and the Oncotype DX® assay
NSABP B-14 tamoxifen benefit study in node-negative, ER+ patients Objective: determine whether the Oncotype DX assay provides information on Prognosis (likelihood of recurrence) Response to tamoxifen (change in likelihood of recurrence with tamoxifen) Both Placebo-eligible Randomized Tamoxifen-eligible Main point: A study was performed to clinically validate the prespecified 21-gene RT-PCR assay and Recurrence Score® algorithm as a predictor of the prospectively defined primary endpoint of distant recurrence-free survival in node-negative, ER+ patients treated with tamoxifen from the large multicenter NSABP Study B-14.1 The study protocol also defined the prespecified endpoints and analysis plan. The laboratory was blinded to the clinical outcomes. NSABP B-14 trial (original): 2828 node-negative, ER+ patients were randomized 1:1 to tamoxifen or placebo in double-blind fashion between 1982 and An additional 1235 patients were registered to tamoxifen in the 10-month period following closure of the trial in 1988, yielding 2617 clinically eligible tamoxifen-treated patients. Patients who were treated with tamoxifen in the randomized portion or the registered portion of the trial were eligible for the Genomic Health study. 1. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351: Paik S, et al. N Engl J Med. 2004;351:

49 B-14 overall benefit of tamoxifen
All patients (N = 645) 1.0 0.9 0.8 0.7 0.6 Proportion without distant recurrence 0.5 0.4 Main point: Shown here is the overall benefit of tamoxifen in the 645 evaluable patients. Paik S, et al. ASCO 2004; Abstract 510. 0.3 0.2 0.1 Placebo Tamoxifen 0.0 2 4 6 8 10 12 14 16 Years Paik S, et al. ASCO 2004; Abstract 510.

50 B-14 benefit of tamoxifen by Recurrence Score® risk category
Distant recurrence-free interval RS < 18 Years 2 4 6 8 14 16 0.0 0.2 0.4 0.6 0.8 1.0 12 10 N Placebo 171 Tamoxifen 142 P = 0.039 RS 18-30 Years 2 4 6 8 14 16 0.0 0.2 0.4 0.6 0.8 1.0 12 10 P = 0.02 N Placebo 85 Tamoxifen 69 RS ≥31* Years 2 4 6 8 14 16 0.0 0.2 0.4 0.6 0.8 1.0 12 10 P = 0.82 N Placebo 99 Tamoxifen 79 Main point: The Recurrence Score® result identifies patients for whom tamoxifen alone may be appropriate therapy. The magnitude of the benefit of tamoxifen was then examined for each of the Recurrence Score groups. The patients in the low-risk group and in the intermediate-risk group as defined by the Recurrence Score result had a large benefit from tamoxifen. The apparent benefit from tamoxifen in patients in the high-risk group, as defined by a Recurrence Score value 31, was smaller. The p-value for the interaction between tamoxifen treatment and the Recurrence Score result is These results suggest that the Recurrence Score result captures both prognosis and responsiveness to tamoxifen. In addition, none of the individual genes alone can capture both prognosis and responsiveness to tamoxifen as successfully as the Recurrence Score result (based on multiple genes). The results should not be used to indicate that tamoxifen should not be given to the high-risk group. First, we cannot exclude a small but clinically important benefit of tamoxifen with regard to DRFS. Second, tamoxifen has benefits in these patients with regard to prevention of contralateral breast cancer or second primaries. Nevertheless, these results indicate that one cannot expect large benefits from tamoxifen alone in the high-risk group. Although this study did not evaluate aromatase inhibitors (AI), to the extent one accepts the recent data showing superiority over tamoxifen for these agents, it is possible that the rate of distant recurrence in the low-risk group may fall even further with AI therapy. Paik S, et al. ASCO 2004; Abstract 510. Interaction P = 0.06 *Results should not be used to indicate that tamoxifen should not be given to the high-risk group RS, Recurrence Score result Paik et al. ASCO 2004; Abstract 510.

51 10-Year absolute risk BC death (%) (95% CI)
The Oncotype DX® assay identifies patients for whom tamoxifen alone may be appropriate therapy Recurrence Score® Result < ≥ 31 NO SYSTEMIC TREATMENT Tamoxifen benefit Tamoxifen benefit Main point: The Kaplan-Meier estimates of the risk of death from breast cancer at 10 years indicate that systemic treatment with tamoxifen is more beneficial in low- and intermediate-risk patients. Shown here are the Kaplan-Meier estimates of the risk of breast cancer death at 10 years for the NSABP B-14 study. The results on top are for patients with no systemic therapy. The results at the bottom are for patients treated with hormonal therapy. The largest benefits of tamoxifen are observed in low- and intermediate-risk patients. One should not expect a large benefit of tamoxifen for patients in the high-risk group. Paik S, et al. ASCO. 2005; Abstract 510. Tamoxifen 10-Year absolute risk BC death (%) (95% CI) Paik S, et al. ASCO. 2005; Abstract 510.

52 10-Year absolute risk BC death (%) (95% CI)
The Oncotype DX® assay can predict benefit from chemotherapy and tamoxifen Recurrence Score® Result < ≥ 31 Tamoxifen Chemotherapy benefit Tamoxifen + CHEMO Main point: The Kaplan-Meier estimates of the risk of death from breast cancer at 10 years indicate that the addition of chemotherapy to tamoxifen is more beneficial in high-risk patients. Shown here are the Kaplan-Meier estimates of the risk of breast cancer death at 10 years for the NSABP B-20 study. The results on top are for patients treated with hormonal therapy. The results at the bottom are for patients treated with hormonal therapy plus chemotherapy. The largest benefits of chemotherapy are observed in the high-risk patients. There was a 28% absolute benefit observed in patients with Recurrence Score value ≥ 31. These data demonstrate that not all patients benefit equally from adjuvant chemotherapy. Adapted from Paik S, et al. J Clin Oncol. 2006;24:3726. 10-Year absolute risk BC death (%) (95% CI) Adapted from Paik S, et al. J Clin Oncol. 2006;24:3726.

53 ASCO guidelines on the use of tumor markers in breast cancer
Oncotype DX® testing can be used to determine prognosis in newly diagnosed patients with node-negative, estrogen receptor-positive breast cancer who will receive tamoxifen Conclusions may not be generalizable to hormonal therapies other then tamoxifen or to other chemotherapy regimens To predict risk of recurrence in patients considering treatment with tamoxifen To identify patients who are predicted to obtain the most therapeutic benefit from adjuvant tamoxifen and may not require adjuvant chemotherapy Patients with high Recurrence Score® result appear to achieve relatively more benefit from adjuvant chemotherapy (specifically CMF) than tamoxifen Main point: The ASCO guidelines state that Oncotype DX® testing can be used to determine prognosis in newly diagnosed patients with node-negative, ER+ breast cancer who will receive tamoxifen. The Oncotype DX® assay is now included in the ASCO guidelines for the use of tumor markers in breast cancer to: Predict risk of recurrence in patients considering treatment with tamoxifen Identify patients who are predicted to obtain the most therapeutic benefit from adjuvant tamoxifen and may not require adjuvant chemotherapy Identify Patients with high recurrence who appear to achieve relatively more benefit from adjuvant chemotherapy (specifically CMF) than tamoxifen Harris L, et al. J Clin Oncol. 2007;33(25): Harris L, et al. J Clin Oncol. 2007;33(25):

54 pT1, pT2, or pT3 and pN1mi (≤ 2mm axillary node metastasis)
NCCN® guidelines include Oncotype DX® testing in the treatment-decision pathway for node-negative and micrometastatic disease HR+, HER2– disease pT1, pT2, or pT3 and pN1mi (≤ 2mm axillary node metastasis) Tumor > 0.5 cm Consider Oncotype DX (Category 2A) No test Adjuvant endocrine therapy ± adjuvant chemotherapy (Category 1) RS < 18 Adjuvant endocrine therapy (Category 2B) RS 18-30 Adjuvant endocrine therapy ± adjuvant chemotherapy (Category 2B) Main point: The National Comprehensive Cancer Network (NCCN) Practice Guidelines in Oncology for breast cancer suggests the use of the Oncotype DX® assay in patients with HR+, HER2–, node-negative or micrometastasis-involved disease. The option of using a gene-based assay of tumor tissue, namely the Oncotype DX assay, to help guide chemotherapy treatment decisions is now included within the systemic adjuvant treatment decision pathway for patients with node-negative or pN1mi (micrometastasis: mm), hormone receptor-positive, HER2-negative tumors that are > 0.5 cm. NCCN Categories of Evidence and Consensus: Category 1: Based upon high-level evidence, there is uniform NCCN consensus that the intervention is appropriate. Category 2A: Based upon lower-level evidence, there is uniform NCCN consensus that the intervention is appropriate. Category 2B: Based upon lower-level evidence, there is NCCN consensus that the intervention is appropriate. Category 3: Based upon any level of evidence, there is major NCCN disagreement that the intervention is appropriate. All recommendations are category 2A unless otherwise noted. (http://www.nccn.org/professionals/physician_gls/categories_of_consensus.asp) Please note that the levels of evidence for each of the actual adjuvant treatment decisions relate only to experts' opinions regarding these therapies, and not to the Recurrence Score. National Comprehensive Cancer Network, Inc. Available at: Adapted from NCCN Practice Guidelines in Oncology-v RS ≥ 31 Adjuvant endocrine therapy + adjuvant chemotherapy (Category 2B) RS, Recurrence Score® result NCCN is a trademark of the National Comprehensive Cancer Network. NCCN does not endorse any therapy or product. Adapted from NCCN Practice Guidelines in Oncology – v

55 Oncotype DX® testing in node-positive disease
Main point: The Oncotype DX® assay was first validated and adopted in ER-positive, node-negative breast cancer. Can the Oncotype DX assay identify patients who are at low, intermediate and high risk of recurrence and predict differential chemotherapy benefit in node-positive patients – a group that has traditionally been thought to be at high risk and to unequivocally require chemotherapy?

56 Oncotype DX® clinical validation in node-positive patients (ECOG trial 2197)
Operable breast cancer 0-3 positive nodes T.1cm if node negative N = 2885 eligible patients 776 samples with genomic data, including Recurrence Score® results Paraffin blocks with cancer cells occupying < 5% of the section area excluded Manual micro-dissection RNA extraction AC Doxorubicin 60 mg/m2 Cyclophosphamide 600 mg/m2 Every 3 weeks × 4 cycles AT Doxorubicin 60 mg/m2 Docetaxel 60 mg/m2 Every 3 weeks × 4 cycles Tamoxifen × 5 years If HR-positive (amended to allow Als) Plus RT if indicated Tamoxifen × 5 years If HR-positive (amended to allow Als) Plus RT if indicated Main point: The prognostic utility of the Oncotype DX® assay in node-positive patients was validated in several studies, including the E2197 study. Goldstein, L et al. for the ECOG reported the results of E2197, which compared 4 cycles of a standard doxorubicin-cyclophosphamide (AC) combination given every 3 weeks with 4 cycles of doxorubicin plus docetaxel (AT) in patients with 0-3 positive lymph nodes. The trial was powered to detect a 25% reduction in the disease-free survival (DFS) hazard rate (from an anticipated 5-year DFS of 78% for the AC arm to 83% for the AT arm). All HR+ patients received tamoxifen 20 mg daily (or an AI) for 5 years following completion of chemotherapy. The treatment arms were well balanced with regard to median age (51 years), proportion of lymph node-negative disease (65%), and proportion of estrogen receptor (ER)-positive disease (64%). The results indicated no advantage for the AT combination. Although subgroup analysis demonstrated a trend toward a lower recurrence rate for patients with ER-negative disease, use of the AT regimen as given in this study and in this population was not recommended. It is noteworthy that the DFS of 84% in the AC control arm was higher than the 78% anticipated for this arm based on the historical data, consistent with other reports indicating that the prognosis for patients with early-stage disease continues to improve. Goldstein LJ, et al. ASCO Abstract 526. No difference between arms Median follow-up 76 months 96.8% reported follow-up until death or for at least 5 years Goldstein LJ, et al. ASCO Abstract 526.

57 5-Year event rates by nodal status & Recurrence Score result
Patients with 1-3 positive nodes and low Recurrence Score® result do well without chemotherapy* 5-Year event rates by nodal status & Recurrence Score result RS Nodes RFI (%) DFS (%) OS (%) < 18 Negative 96 93 95 Positive 91 97 18-30 86 87 77 ≥ 31 80 92 75 61 72 *Including micrometastases (pN1mi) Main point: Patients with a low Recurrence Score® result have a very good RFI even if they have positive nodes. It is not clear whether this is due to the administration of AC or AT chemotherapy in this group or whether the prognosis of this group, without chemotherapy, is favorable. The fact that all groups received chemotherapy precludes this determination. Conversely, the patients with a high Recurrence Score results do more poorly, especially if they have positive nodes. This suggests that if one has a high Recurrence Score result and positive nodes, conventional AC or AT chemotherapy (along with hormone treatment) is not sufficient adjuvant therapy. This also suggests that the Recurrence Score result is capturing biology above and beyond that captured by nodal status (at least in the 0-3 node group). It is not clear whether the group with a low Recurrence Score result and positive nodes did well because of chemotherapy or because their prognosis without chemotherapy was good. All patients were treated with chemotherapy in this study. Other analyses are underway to help answer this question. Goldstein LJ, et al. ASCO Abstract 526. Low Recurrence Score results (< 18) in patients with 1-3 positive axillary nodes may eventually be used to select individuals for a short course of chemotherapy plus hormonal therapy Elevated Recurrence Score results (≥ 18) may eventually be used to select individuals for participation in clinical trials evaluating novel treatment strategies or for more aggressive chemotherapy regimens RS, Recurrence Score result Goldstein LJ, et al. ASCO Abstract 526.

58 Oncotype DX® clinical validation in node-positive patients (SWOG 8814 sub-analysis)
Postmenopausal, node-positive, ER-positive breast cancer N = 1477 Patients with samples (n = 666) RT-PCR obtained (n = 601) Tamoxifen alone (n = 148) CAF + T (n = 243) CAF  T (n = 219) Sample for primary analysis = 367 (40% of parent trial) Tamoxifen × 5 yrs n = 361 CAF × 6 + tamoxifen n = 550 CAF × 6  tamoxifen n = 566 Main point: The prognostic utility of the Oncotype DX® assay in node-positive patients was validated in several studies, including the SWOG 8814 study. The sequential CAF –tamoxifen arm turned out to have superior disease-free survival and overall survival over 10 years Thus comparisons in this genomic study utilized this sequential arm rather than the concurrent arm, since it represents the most effective way to give CAF and tamoxifen Albain KS, et al. Lancet Oncol. 2009; [Epub ahead of print]. Superior disease-free survival and overall survival over 10 years 58 Albain KS, et al. Lancet Oncol. 2010;11(1):55-65.

59 Recurrence Score® result is prognostic for node-positive patients (tamoxifen arm)
DFS by risk group (tamoxifen-alone arm) OS by risk group (tamoxifen-alone arm) 1.00 0.75 0.50 0.25 0.00 1.00 0.75 0.50 0.25 0.00 Stratified log-rank P = at 10 years Stratified log-rank P = at 10 years RS < 18 (n = 55) RS (n = 46) RS ≥ 31 (n = 47) RS < 18 (n = 55) RS (n = 46) RS ≥ 31 (n = 47) Main point: The Recurrence Score® result was shown to be prognostic for patients with node-positive disease. Among these node positive patients, the Recurrence Score result was able to separate out a group with a better and a group with a worse prognosis. These patients were treated with tamoxifen alone and thus these results are prognostic in the sense that they do not include any chemotherapy effect (though tamoxifen effect is included). There was a wide gap between the 10-year DFS (disease-free survival) in those patients on this study with a low Recurrence Score result (40% 10-year DFS) and those with a high Recurrence Score result (57% 10-year DFS) Here, DFS was used, as opposed to DRFS (distant recurrence-free survival) which was used in the NSABP B-14 and B-20 analyses with node negative patients. Thus, the event rate would be expected to be higher here. Albain KS, et al. Lancet Oncol. 2009; [Epub ahead of print]. Years since registration Years since registration 10-Year DFS: 60%, 49%, 43% 10-Year OS: 77%, 68%, 51% RS, Recurrence Score result Albain KS, et al. Lancet Oncol. 2010;11(1):55-65.

60 DFS BY TREATMENT & RS GROUP
High Recurrence Score® result predictive of chemotherapy benefit in node-positive patients DFS BY TREATMENT & RS GROUP RS < 18 RS 18-30 RS ≥ 31 1.00 0.75 0.50 0.25 0.00 Stratified log-rank P = 0.97 at 10 years Stratified log-rank P = 0.48 at 10 years Stratified log-rank P = at 10 years Main point: Prediction involves asking whether the Recurrence Score® result can determine a group of patients in this study who may or may not benefit from an intervention, in this case CAF chemotherapy (given prior to tamoxifen). If the Recurrence Score result is low, there appears to be little if any benefit of sequential CAF—T therapy over tamoxifen alone. If the Recurrence Score result is high, there appears to be a strongly statistically significant benefit of sequential CAF—T therapy over tamoxifen alone. Albain KS, et al. Lancet Oncol. 2009; [Epub ahead of print]. CAF  T (n = 91, 26 events) Tam (n = 55, 15 events) CAF  T (n = 46, 22 events) Tam (n = 57, 20 events) CAF  T (n = 47, 26 events) Tam (n = 71, 28 events) 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 Years since registration Years since registration Years since registration No benefit to CAF over time if low or intermediate RS Strong benefit if high RS RS, Recurrence Score result Albain KS, et al. Lancet Oncol. 2010;11(1):55-65.

61 Risk of Distant Recurrence Using Oncotype DX® Assay in Postmenopausal Primary Breast Cancer Patients Treated with Anastrozole or Tamoxifen: a TransATAC Study Dowsett M et al on behalf of the ATAC Trialists’ Group San Antonio Breast Cancer Symposium. 2008; Abstract 53.

62 Tamoxifen + Anastrozole
Study overview ATAC study population (N = 9366) Tamoxifen Anastrozole Tamoxifen + Anastrozole (combination arm not examined) Primary Analysis: To determine whether Oncotype DX® assay significantly adds to a proportional hazards model for time to distant recurrence (age, tumor size, grade, treatment) in node-negative, HR+, patients with no adjuvant chemotherapy Main point: The primary analysis of this study sought to determine whether the addition of the continuous Recurrence Score® result added significantly to a statistical model that included age, tumor size, grade and treatment (tamoxifen or anastrozole). Note that the tamoxifen and anastrozole arms were combined for the initial analysis, per the prospectively-specified analysis plan. They were subsequently broken out for separate analysis. Since there was no added benefit to the addition of anastrozole to tamoxifen, this arm was discontinued early in the original ATAC trial and these patients were offered the opportunity to switch to anastrozole or tamoxifen monotherapy. Tumor blocks were not collected for patients in the combination arm. Note that there were 1308 samples left with reportable Recurrence Score® result that were ER+ and/or PgR+ The “TransATAC” study eliminated from the ATAC study those patients that received the combination of tamoxifen and anastrozole or were ER-/PgR-. Another large group of patients were not included in this analysis because blocks were either not recoverable or insufficient tumor was present in the blocks for analysis. Several secondary analyses were also performed. This included looking at the Recurrence Score result both as a continuous and as a categorical variable. Note that a comparison with Adjuvant! Online was also performed. Dowsett M, et al. SABCS 2008; abstract 53. Secondary analyses: Determine whether the relationship between continuous Recurrence Score® result and time to distant recurrence differs by nodal status or treatment arm Determine the relationship of predefined Recurrence Score groups with time to distant recurrence by nodal status and treatment arm Evaluate whether Recurrence Score result adds to the Adjuvant! Online estimate of risk Dowsett M, et al. SABCS 2008; abstract 53.

63 Number of evaluable patients and distant events by nodal status
Node negative Node positive Node unknown Total All 890 363 55 1308 Adjuvant chemo –9 –55 –1 –65 HR negative –4 –0 Didn’t start T or A –5 –2 –8 Evaluable patients 872 (71%) 306 (25%) 53 (4%) 1231 (100%) Number of distant events 72 74 6 152 Main point: Of the 1308 patients enrolled in the TransATAC study, 1231 were evaluable, including 306 (25%) patients with node-positive disease. Dowsett M, et al. SABCS 2008; abstract 53. Distributions of the clinical variables in the 1231 evaluable (non-N Am) patients were similar to those in the 2929 ATAC (non-N Am) patients who were not included in this study Dowsett M, et al. SABCS 2008; abstract 53.

64 Primary analysis: time to distant recurrence and Recurrence Score® value adjusted for clinical covariates (node-negative patients, both treatment arms) Variable HR (95% CI)* P value Recurrence Score / 50* 5.25 (2.84, 9.73) < 0.001 Tumor Size: > 2 vs ≤ 2 cm 2.78 (1.70, 4.57) Central grade Moderate vs Well Poor vs Well 1.70 (0.75, 3.86) 2.06 (0.82, 5.17) 0.270 Multivariate analysis adjusted for treatment arm and patient age *Hazard Ratio for a 50-point increment in Recurrence Score value Main point: The Recurrence Score® result as a continuous variable had a highly significant hazard ratio of 5.25 in this multivariate analysis. This analysis is for node negative patients with both tamoxifen alone and anastrozole alone treatment arms analyzed. Tumor size had a statistically significant hazard ratio of 2.78 in this analysis. Dowsett M, et al. SABCS 2008; abstract 53. Multivariate analysis confirms that the Oncotype DX® Recurrence Score result as a continuous variable is a highly significant predictor of time to distant recurrence Dowsett M, et al. SABCS 2008; abstract 53.

65 Time to distant recurrence by Recurrence Score® group
Node negative (n = 872) (both treatment arms) Node positive (n = 306) (both treatment arms) 1.0 96% 1.0 0.9 88% 0.9 0.8 0.8 83% Log-rank P < 0.001 75% Log-rank P < 0.001 0.7 0.7 72% 0.6 0.6 Proportion distant recurrence-free 0.5 Proportion distant recurrence-free 0.5 51% 0.4 0.4 0.3 0.3 N (%) Events Low (59%) Int (26%) High 130 (15%) N (%) Events Low (52%) Int (31%) High (17%) 0.2 0.2 0.1 0.1 Main point: The Recurrence Score® result is prognostic for patients with node-negative or node-positive disease (both tamoxifen and anastrozole arms combined). Node Negative For the node negative group, the Recurrence Score® result, as a categorical variable, is highly prognostic when the tamoxifen and the anastrozole arms (individually treated) are combined. A low Recurrence Score result gives a 96% proportion distant recurrence-free at nine years, while a high Recurrence Score gives only a 75% proportion distant recurrence-free at nine years. Node Positive There was a clearly better prognosis for the low Recurrence Score group with positive nodes compared with the high Recurrence Score group with positive nodes. The former had a 83% proportion distant recurrence-free at nine years, while the latter had only a 51% proportion distant recurrence-free at nine years. Dowsett M, et al. SABCS 2008; abstract 53. 0.0 0.0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Years Years RS group HR* (95% CI) High vs Low 5.2 ( ) Int vs Low 2.5 ( ) RS group HR* (95% CI) High vs Low 2.7 ( ) Int vs Low 1.8 ( ) *Hazard ratio for RS group adjusted for tumor size, grade, age and treatment Dowsett M, et al. SABCS 2008; abstract 53. RS, Recurrence Score result

66 Percent with distant recurrence at 9 years (node negative)
No. of events All patients, low RS (n = 513) 20 All patients, int. RS (n = 229) 24 All patients, high RS (n = 130) 28 All patients (n = 872) 72 Tamoxifen, low RS (n = 245) 8 Tamoxifen, int. RS (n = 117) 12 Tamoxifen, high RS (n = 70) 21 All tamoxifen (n = 432) 41 Main point: There is no striking difference in recurrence between anastrozole and tamoxifen, especially in the low and intermediate Recurrence Score® groups. Dowsett M, et al. SABCS 2008; abstract 53. Anastrozole, low RS (n = 268) 12 Anastrozole, int. RS (n = 112) 12 Anastrozole, high RS (n = 60) 7 All anastrozole (n = 440) 31 10 20 30 40 50 60 70 RS, Recurrence Score® result Percent with distant recurrence at 9 years Dowsett et al., SABCS 2008, Abstract # 53

67 Percent with distant recurrence at 9 years (node positive)
No. of events All patients, low RS (n = 160) 25 All patients, int. RS (n = 94) 25 All patients, high RS (n = 52) 24 All patients (n = 306) 74 Tamoxifen, low RS (n = 79) 11 Tamoxifen, int. RS (n = 47) 13 Tamoxifen, high RS (n = 26) 11 All tamoxifen (n = 152) 35 Main point: There is no striking difference in recurrence between anastrozole and tamoxifen, especially in the low and intermediate Recurrence Score® groups. Dowsett M, et al. SABCS 2008; abstract 53. Anastrozole, low RS (n = 81) 14 Anastrozole, int. RS (n = 47) 12 Anastrozole, high RS (n = 26) 13 All anastrozole (n = 154) 29 10 20 30 40 50 60 70 RS, Recurrence Score® result Percent with distant recurrence at 9 years Dowsett et al., SABCS 2008, Abstract # 53

68 9-Year risk of distant recurrence (%) Recurrence Score value
Rate of distant recurrence increases with the number of positive nodes for all Recurrence Score® results 100 95% CI Mean ≥ 4 Positive nodes 90 80 70 60 9-Year risk of distant recurrence (%) 50 1-3 Positive nodes 40 Main point: The risk of distant recurrence increases with the number of positive nodes for the Recurrence Score® result. The 9-year risk of distant recurrence increases with the number of positive nodes. The 9-year risk of distant recurrence is low for patients with 1-3 positive nodes with Recurrence Score values in the lower range. Patients with 1-3 positive nodes experienced a much lower risk of distant recurrence, compared with patients with ≥ 4 positive nodes. Dowsett M, et al. SABCS. 2008; Abstract 53. 30 Node negative 20 10 5 10 15 20 25 30 35 40 45 50 Recurrence Score value Dowsett M, et al. SABCS 2008; abstract 53.

69 ATAC conclusions Confirms performance of Oncotype DX® Recurrence Score® result in postmenopausal HR+ patients treated with tamoxifen in a large contemporary population Demonstrates for the first time that the Oncotype DX Recurrence Score result is an independent predictor of distant recurrence in node negative and node positive HR+ patients treated with anastrozole The established relationship between Oncotype DX Recurrence Score result and distant recurrence for tamoxifen may be applied for anastrozole with adjustment for the lower risk of distant recurrence with the aromatase inhibitor Main point: The authors’ conclusions are shown here. This study confirms performance of Oncotype DX® Recurrence Score® result in postmenopausal HR+ patients treated with tamoxifen in a large contemporary population. This study demonstrates for the first time that the Oncotype DX Recurrence Score result is an independent predictor of distant recurrence in node negative and node positive HR+ patients treated with anastrozole The established relationship between the Oncotype DX Recurrence Score result and distant recurrence for tamoxifen may be applied for anastrozole with adjustment for the lower risk of distant recurrence with the aromatase inhibitor Dowsett M, et al. SABCS 2008; abstract 53. Dowsett M, et al. SABCS 2008; abstract 53.

70 Reproducible clinical validation essential in changing standard of care: more than 4000 patients studied in 12 trials Study Type No. Pts Nodal status Providence Exploratory 136 Neg Rush* 78 Pos NSABP B-20 233 NSABP B-14* Prospective 668 MD Anderson* 149 Kaiser Permanente* Prospective case-control 790 cases/controls NSABP B-14 Prospective placebo vs Tam 645 Milan* 89 Neg/Pos NSABP B-20* Prospective Tam vs Tam+Chemo 651 ECOG 2197* Exploratory and prospective 776 SWOG 8814 367 ATAC Prospective Tam vs AI 1231 Main point: Although this assay has emerged for clinical use over the last few years, there is evidence supporting its utility in multiple studies involving over 4000 patients. References from top to bottom for the studies shown including both published studies and abstracts Esteban J, Baker J, Cronin M, et al. Tumor gene expression and prognosis in breast cancer: Multi-gene RT-PCR assay of paraffin-embedded tissue. Proc Am Soc Clin Oncol 22: 2003 (abstr 3416). Cobleigh MA, Tabesh B, Bitterman P, et al. Tumor gene expression and prognosis in breast cancer patients with 10 or more positive lymph nodes. Clin Cancer Res. 2005;11(24 Pt 1): Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351(27): Esteva FJ, Sahin AA, Cristofanilli M, et al. Prognostic role of a multigene reverse transcriptase-PCR assay in patients with node-negative breast cancer not receiving adjuvant systemic therapy. Clin Cancer Res. 2005;11(9): Habel L, Shak S, Jacobs 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;May 31;8(3):R25 [Epub ahead of print]. Gianni L, Zambetti M, Clark K, et al. Gene expression profiles in paraffin-embedded core biopsy tissue predict response to chemotherapy in women with locally advanced breast cancer. J Clin Oncol.2005;23(29): Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol. 2006;24: L. J. Goldstein, R. Gray, B. H. Childs, et al. Prognostic Utility of 21-Gene Assay in Hormone Receptor (HR) Positive Operable Breast Cancer and 0-3 Positive Axillary Nodes Treated with Adjuvant Chemohormonal Therapy (CHT): An Analysis of Intergroup Trial E2197. Journal of Clinical Oncology, 2007 ASCO Annual Meeting Proceedings Part I. Vol 25, No. 18S (June 20 Supplement), 2007: 526 *Published studies

71 TAILORx study Trial Assigning Individualized Options for Treatment
Primary objective is to determine whether adjuvant hormonal therapy is not inferior to adjuvant chemohormonal therapy for patients with a Recurrence Score® result 11-25 Correlates with 10-20% risk of distance recurrence at 10 years Potential implications Reduce chemotherapy overtreatment in those likely to be treated with hormonal therapy alone Reduce inadequate treatment by identifying individuals who derive great benefit from chemotherapy Evaluate benefit of chemotherapy where uncertainty still exists about its utility Main point: The TAILORx study – Trial Assigning IndividuaLized Options for Treatment (Rx) – opened in April PACCT-1 is the first study that is part of the NCI PACCT (Program for the Assessment of Clinical Cancer Tests) program. This trial is sponsored by the National Cancer Institute and administered by ECOG. All the North American breast cancer cooperative groups are participating in this study. The PACCT program was established by the NCI in 2000 to integrate clinical cancer tests into clinical practice. One of the first questions that the NCI wanted to address with this program was the overtreatment of ER+, node-negative women. They wanted to utilize tests that would determine the risk of recurrence and tailor treatment to these patients. Other tests were considered, such as Ki67, uPA/PAI-1; however, the Oncotype DX® breast cancer assay was selected as the most appropriate marker for the first PACCT trial based on the number and quality of the trials that tested the prognostic and predictive value of this assay. The premise of the trial is that the Oncotype DX assay is able to identify women who are expected to benefit from chemotherapy and women who are expected to have little or no benefit. The potential implications are that many women are currently being overtreated with chemotherapy and could benefit from hormonal therapy alone without loss of treatment benefit and other patients are getting inadequate treatment and could benefit from chemotherapy. In this trial, patients will be treated based upon their risk of recurrence and likelihood of chemotherapy benefit.  There are some women for whom uncertainty still exists about the likelihood of benefit from chemotherapy in addition to hormonal therapy, and these women will be randomized to determine whether they could benefit from hormonal therapy alone without loss of treatment benefit. This slide shows how the Recurrence Score® result is reported, and the intermediate group has a score between 18 and 30. The low-risk, intermediate-risk, and high-risk cutoffs were prospectively defined for the purposes of the validation trial, but these do not represent true cutoffs in risk of recurrence among the 3 groups. For the TAILORx trial, the middle-range cutoffs are 11 and 25 because these scores correlate roughly with a 10%-20% risk of distance recurrence at 10 years when looking at the upper 95% confidence intervals. These cutoffs were selected by the TAILORx study designers to define stratification groups in whom, in addition to hormone therapy, (a) chemotherapy does not improve outcomes (Recurrence Score value < 11); (b) chemotherapy does improve outcomes (Recurrence Score value > 25); and (c) there is a legitimate question whether chemotherapy improves outcomes (Recurrence Score value between 11 and 25).

72 TAILORx schema Trial Assigning Individualized Options for Treatment
Recurrence Score® result ≤ 10 Hormone therapy registry Patients with node-negative, hormone-positive breast cancer Oncotype DX® assay Recurrence Score result 11-25* Randomize to either hormone therapy or chemotherapy + hormone therapy Register specimen banking Recurrence Score result > 25 Chemotherapy + hormone therapy Main point: The study schema of the TAILORx trial is shown here. The eligible patients for this trial are node-negative, ER+ and are candidates for chemotherapy (ie, patients who do not have comorbid conditions that would preclude them from receiving chemotherapy and who are willing to take it if recommended). The fact that the Breast Cancer Intergroup is stratifying patients for the TAILORx trial by the Oncotype DX® assay demonstrates that this assay is widely accepted and validated in the study population. Treatment will be based on the results of the assay. Patients will be stratified as follows: Patients with a Recurrence Score® result below 11 will receive hormonal therapy alone. Patients with a Recurrence Score between 11 and 25 will be randomized to either hormonal therapy alone or hormonal therapy + chemotherapy. This is the primary study group. This corresponds approximately to a risk of recurrence at 10 years of 10%-20% (at upper bound of 95% CI). Patients with a Recurrence Score greater than 25 will receive chemotherapy + hormonal therapy. Since this trial has a dealer’s choice–type design, individual investigators can select the type of hormonal therapy and chemotherapy from a list included in the protocol. The groups in this trial do not correspond to the low-, intermediate-, and high-risk cutoffs found on the Oncotype DX report. The treatment groups for the TAILORx trial were selected for different purposes from those involved with the selection of cutoffs for the validation trial and the Oncotype DX report.  The treatment groups for the TAILORx study were selected to correspond with specific levels of risk of recurrence and likelihood of chemotherapy benefit.  The TAILORx investigators felt it would not be ethical to withhold chemotherapy from women who have a 20% risk of recurrence This study has a non-inferiority design. The hypothesis is that patients in this middle-range Recurrence Score risk group will do no worse with hormonal therapy alone than they would with hormonal therapy plus from chemotherapy. Note: The study does not assess the validity of the Oncotype DX assay. This trial design is predicated on the assessment that the assay is fully validated. For a patient with a Recurrence Score < 11, chemo will not be given. For a patient with a Recurrence Score > 25, chemo will be given. The risk generated from the Recurrence Score result is felt to be validated and is actionable, based on the references below. Paik S, Shak S, Tang G, et al. N Engl J Med. 2004;351(27): Paik S, Tang G, Shak S, et al. J Clin Oncol Aug 10;24(23) [Epub ahead of print May 23, 2006]. Habel L, Shak S, Jacobs M, et al. Breast Cancer Res. 2006;May 31;8(3):R25 [Epub ahead of print]. Simon R. J Clin Oncol. 2005;23(29): *Primary study group: Recurrence Score result correlates with a 10-20% risk of distance recurrence at 10 years (upper 95% CI)

73 Single-gene reporting
Quantitative hormone receptor analysis

74 Oncotype DX® assay also provides quantitative data for ER, PR, HER2
Provides additional insight into the biology of individual tumors ER score PR score Main point: Oncotype DX® testing provides quantitative RT-PCR measurement of ER PR, and HER2. Since February 2008 for ER and PR and October 2008 for HER2, Genomic Health has included Single-Gene Testing values on all Oncotype DX reports. Clinicians have long known that ER, PR, and HER2 were among the genes used in calculating the Recurrence Score® result, and they have requested a separate analysis of these gene expression levels. Genomic Health is able to provide these particular Single-Gene Testing scores based on studies confirming concordance of RT-PCR and IHC/FISH. The Single-Gene Testing values can offer additional quantitative insight into the biology of individual tumors and aid in clinical decision-making, particularly in the intermediate Recurrence Score range, where the quantitative ER score may give insight into how much benefit the patient can expect to derive from tamoxifen and aromatase inhibitors. Pathologists can use the Single-Gene Testing values as an external standard for ongoing QC of their own immunohistochemical and FISH assays. The ASCO-CAP guidelines for HER2 testing state that laboratories performing HER2 testing must show 95% concordance with a validated outside method. Similar guidelines will be forthcoming for ER testing. HER2 score

75 Continuous measurement of ER/PR is reflective of tumor biology
Overall PR range 1000-fold NSABP B-14 (N = 645) 1 2 ER– PR+ ER+ PR+ 1 1 1 9 8 (relative to reference genes; log 2) PR expression by RT-PCR 7 6 ER+ range 200-fold 5 Main point: Shown here is the distribution of quantitative ER and PR in the NSABP B-14 study. Quantitative ER expression is plotted on the x-axis. Quantitative PR expression is plotted on the y-axis. ER and PR expression are measured on a base 2 logarithmic scale from 0 to 15 (relative to 5 reference genes), where a one unit increment is associated with a 2-fold change in RNA expression The provisional cutoffs for PR and ER are indicated by the dotted lines, respectively. ER+/PR+ patients are noted in the upper right hand quadrant, ER+/PR- in the lower right hand quadrant, ER-PR- in the lower left hand quadrant and ER-PR+ in the upper left hand quadrant. Note the similar distributions observed in both studies. Note that If ER expression is low then PR expression is low; however, when ER expression is high PR expression can be low or high. In both studies ER and PR are modestly to moderately correlated, with a Person’s coefficient of 0.38 in B-14 and 0.66 in Kaiser Permanente. The overall range of expression of the ER gene as measured by RT-PCR was quite large, a range of over 3,000 fold Even for “ER positive” patients as determined by RT-PCR, the range of ER expression was over 200 fold The overall range of PR expression was over 1000 fold Reproducibility of the assay has a standard deviation of less than 0.4 units Data on file compiled from Paik S, et al. N Engl J Med. 2004;351: 4 Overall ER range fold 3 ER– PR– ER+ PR– 2 2 3 4 5 6 7 8 9 1 1 1 1 2 1 3 1 4 ER expression by RT-PCR (relative to reference genes; log 2) Reproducibility of the assay has a standard deviation of less than 0.4 units Data on file compiled from Paik S, et al. N Engl J Med. 2004;351: 75

76 Oncotype DX® clinical validation: NSABP B-14
Objective: Prospectively validate Recurrence Score® result as predictor of distant recurrence in node-negative, ER+ patients Multicenter study with prespecified 21-gene assay, algorithm, endpoints, analysis plan Placebo—not eligible Randomized Tamoxifen—eligible Registered Tamoxifen—eligible Main point: The objective of this analysis was to determine whether and to what degree quantitative ER and quantitative PR are 1) prognostic 2) predictive of tamoxifen benefit or 3) both The primary endpoint for the study was Distant Recurrence-Free Interval (DRFI) The secondary endpoints were Relapse-Free Interval (RFI) and Overall Survival (OS) For primary analysis, patients censored at time of development of contralateral breast cancer, second non-breast cancer, or death without breast cancer recurrence For this analysis, the same primary endpoint were used in the prior Paik publication – distant recurrence free interval As shown previously, the 645 pts that were evaluable for this study were similar to all B-14 patients indicating there was no selection bias relative to the original B-14 cohort. Paik S, et al. N Engl J Med. 2004;351: Paik S, et al. N Engl J Med. 2004;351:

77 Quantitative ER expression is not strongly prognostic but is predictive of tamoxifen benefit in ER+ patients 1.0 0.9 Large tamoxifen benefit in highest / mid-ER tertiles 0.8 Little tamoxifen benefit in lowest ER tertile 0.7 0.6 Quantitative ER expression in ER+, placebo-treated patients is not strongly prognostic (P = 0.54) Proportion without distant recurrence 0.5 0.4 N Main point: Quantitative ER expression is not strongly prognostic but is predictive of tamoxifen benefit in ER+ patients. First the ER+, placebo-treated patients alone were examined (solid lines). This Kaplan-Meier plot shows those patients divided into tertiles by their quantitative ER expression. There was no significant difference in time to distant recurrence among the three ER tertiles. Then, in the dotted lines, time to distant recurrence is shown for tamoxifen treated ER+ patients in the three ER tertiles. Comparing the blue and green dotted line to the blue and green solid lines, there is a large tamoxifen benefit in patients in the highest and mid ER expression tertiles. Finally, in the yellow dotted line, time to distant recurrence is shown in the tamoxifen treated ER + patients in the lowest ER tertile Comparing the yellow dotted line to the yellow solid line, there is only little benefit of tamoxifen in ER + patients in the lowest ER expression tertile. This differential benefit among cohorts indicates that quantitative ER is predictive of tamoxifen benefit Baehner FL, et al. SABCS Abstract #510. Placebo – highest ER tertile 105 Placebo – mid ER tertile 117 Placebo – lowest ER tertile 113 0.3 0.2 Tamoxifen – highest ER tertile 99 Tamoxifen – mid ER tertile 88 Tamoxifen – lowest ER tertile 94 0.1 0.0 2 4 6 8 10 12 14 16 Years Baehner FL, et al. SABCS Abstract #510.

78 Quantitative PR expression is strongly prognostic in ER+, placebo-treated patients
1.0 P value = 0.002 0.9 0.8 0.7 0.6 DRFI 0.5 0.4 Main point: Quantitative PR expression is strongly prognostic in ER+, placebo-treated patients. Shown here again in the solid lines are the Kaplan Meier plots in the ER+ placebo patients by the 3 PR expression tertiles. The Kaplan Meier plot shows ER+ placebo treated patients divided into tertiles by their quantitative PR expression. In contrast to ER, there was a significant difference in time to distant recurrence among the three PR tertiles, p value = The highest PR tertile had the longest time to distant recurrence, the lowest PR tertile had the shortest time to distant recurrence. This indicates that quantitative PR expression in ER+ placebo treated patients is strongly prognostic. Baehner FL, et al. SABCS Abstract #510. 0.3 N 0.2 Placebo – highest ER tertile 127 Placebo – mid-ER tertile 94 Placebo – lowest ER tertile 114 0.1 0.0 2 4 6 8 10 12 14 16 Years Baehner FL, et al. SABCS Abstract #510.

79 E2197 and Northern California Kaiser Permanente Studies
Concordance between RT-PCR and IHC for estrogen & progesterone receptors E2197 and Northern California Kaiser Permanente Studies

80 ECOG 2197 & Kaiser studies: methods for ER/PR assessment
Nested, case-control study N = 769 Used tissue microarrays with mm cores per tumor Contained ER+ and ER– tumors ER antibody was 1D5 (Dako Cytomation) PR antibody was 636 (Dako Cytomation) ECOG central laboratory did IHC staining Staining assessed by 2 expert anatomic breast pathologists Quantitation used Allred Score (0-8) with positive AS ≥ 3 Kaiser study Case-control study N = 607 Used fixed paraffin whole-tumor tissue sections Confined to ER+ tumors ER antibody was SP1 (LabVision) PR antibody was 636 (Dako Cytomation) Phenopath central laboratory did IHC staining Staining assessed by 2 expert anatomic pathologists Quantitation used semiquantitative scoring method: 0%, 0-25%, %, > 75% Main point: Methods used for ER/PR assessment in the E2197 and Kaiser Permanente studies are compared here. In ECOG 2197, 769 patients were studied using tissue microarrays with the tumors in 1.0 mm duplicate cores. The ER antibody used was 1D5 and PR was 636. The ECOG central laboratory performed the staining. Scoring was done by two breast pathologists in tandem using the Allred score, a semi-quantitative scoring method incorporating staining intensity and percentage ranging from 0-8, with positivity defined as greater than 2. In the Kaiser study, 607 patients were studied using whole tissue sections and the SP1 antibody. Staining was done at Phenopath laboratories and a semiquantitative scoring method was used: 0%, 0-25%, 25-75%, >75%. Badve SS. ASCO Breast Cancer Symposium Abstract #87. Baehner FL, et al. ASCO Breast Cancer Symposium Abstract #88.

81 High degree of concordance between RT-PCR and IHC for ER/PR status
ER and PR concordance between central IHC vs RT-PCR Oncotype DX® assay Central IHC vs Oncotype DX ECOG study (n = 769) concordance (95% CI) Kaiser study (n = 607) ER status 93% (91-94%) 96% (94-97%) PR status 90% (88-92%) 90% (87-92%) Hormone receptor status 93% (91-95%) 95% (93-97%) Main point: There was a high degree of concordance between RT-PCR and IHC for ER/PR status. In the ECOG study using the tissue microarrays, the 1D5 antibody for ER and the 636 antibody for PR and Allred scoring, there is 93% concordance between RtPCR for ER and 90% for PR. In the Kaiser study, using whole tissue sections, the SP1 antibody for ER and the same 636 antibody for PR and the semi quantitative scoring method, there is 96% concordance for ER and 90% for PR. Badve SS. ASCO Breast Cancer Symposium Abstract #87. Baehner FL, et al. ASCO Breast Cancer Symposium Abstract #88. ER, estrogen receptor; PR, progesterone receptor; IHC, immunohistochemistry; RT-PCR, reverse transcriptase-polymerase chain reaction; ECOG, Eastern Cooperative Oncology Group; CI, confidence interval Badve SS. ASCO Breast Cancer Symposium Abstract #87. Baehner FL, et al. ASCO Breast Cancer Symposium Abstract #88.

82 E2197: ER & PR expression by central RT-PCR & central IHC
First lets look at ER expression in the ECOG 2197 samples. ER by RT-PCR is plotted on the X axis, with the 3000 fold dynamic range and log2 based scale. Each one unit increase in ER expression represents a doubling in the amount of mRNA. ER assessed by immunohistochemistry using the ID5 antibody and quantitated by the Allred score is plotted on the Y axis. The cutpoint for ER positivity for ER by RT-PCR is 6.5 units. The cutpoint by IHC is an Allred score greater than two. There is a significant correlation coefficient of 0.85, meaning that ER expression measured by the two methods move in the same direction. Increasing ER levels by RT-PCR are associated with increasing levels by IHC and vice versa. Overall concordance is 93%. Negative agreement is shown in the lower left quadrant. Positive agreement is shown in the upper right quadrant. Note the broad range of the RT-PCR measurement. For any positive case by IHC, the ER measurement by PCR can be either high or relatively low. Even among the negative cases by IHC, the RT-PCR measurement can be either low or relatively high, with some cases approaching the 6.5 cutpoint. Fourteen percent negative cases by IHC are positive by RT-PCR, and among these discordant cases, the RT-PCR measurement ranges from 6.5 to 10.4 units Now lets look at PR. We have the same orientation of the axes. The cutpoint for PR positivity by RT-PCR is 5.5 expression units. Allred score cutpoint for PR positivity is the same as ER, greater than 2 . There is an overall concordance rate of 90% and a correlation coefficient of Fifteen percent of positive cases by IHC are negative by RT-PCR, with measurements for these cases ranging from 2.4 to 5.4 units. Badve SS. ASCO Breast Cancer Symposium Abstract #87. 14% of ER-negative cases by IHC are ER positive by RT-PCR For these discordant cases, RT-PCR measurements range from 6.5 to 10.4 units 15% of PR-positive cases by IHC are PR negative by RT-PCR For these discordant cases, RT-PCR measurements range from 2.4 to 5.4 units Badve SS. ASCO Breast Cancer Symposium Abstract #87.

83 Kaiser: ER & PR expression by central RT-PCR & central IHC
+ + - - Lets now look at ER and PR expression in the Kaiser study. Remember ER IHC was assessed at Phenopath using the SP1 antibody and quantitated using their semiquantitative scoring system. PR was assessed using the same 636 antibody as in the Kaiser study. First ER: orientation of the axes is the same as in the Kaiser study. The RT-PCR measurement on the x axis and IHC on the Y axis. Here the cutpoint for IHC positivity is a score of 1+ or above, which represents any positive staining.. The overall concordance is 96%. Note how the Phenopath semiquantitative scoring system dichotamtimizes the IHC result; most cases are either strongly positive or negative. This well known phenomenon of the SP1 antibody has been described in the literature. Of the 7 discordant cases that are positive by IHC and negative by RT-PCR, 6 (86%) are within 1 unit of the 6.5 cutoff. Of the 20 discordant pairs that are negative by IHC and positive by RT-PCR, 17 (85%) are within 1 unit of the 6.5 cutoff. Now PR: The overall concordance is 90%, which is identical to that in the Kaiser study. Of the 22 discordant pairs that are PR positive by IHC and negative by RT-PCR, 19 (86%) are within 1 unit of the 5.5 cutoff Of the 38 discordant pairs that are PR negative by IHC and positive by RT-PCR, 23 (61%) are within 1 unit of the 5.5 cutoff. Baehner FL, et al. ASCO Breast Cancer Symposium Abstract #88. Of the 7 discordant pairs ER positive by IHC but ER negative by RT-PCR, 6 (86%) are within 1 unit of the 6.5 cutoff. Of the 20 discordant pairs ER negative by IHC but ER positive by RT-PCR, 17 (85%) are within 1 unit of the 6.5 cutoff. Of the 22 discordant pairs PR positive by IHC but PR negative by RT-PCR, 19 (86%) are within 1 unit of the 5.5 cutoff . Of the 38 discordant pairs PR negative by IHC but PR positive by RT-PCR, 23 (61%) are within 1 unit of the 5.5 cutoff. Baehner FL, et al. ASCO Breast Cancer Symposium Abstract #88.

84 Concordance between RT-PCR and IHC/FISH for HER2 receptor
E2197 and Northern California Kaiser Permanente Studies

85 ECOG 2197 & Kaiser studies: methods for HER2 assessment
Used tissue microarrays HER2 assessed using HercepTest ECOG central laboratory Positive result defined by ASCO guidelines Kaiser study Used fixed paraffin whole-tumor tissue sections HER2 assessed using Vysis PathVysion HER2/neu DNA Probe Kit Phenopath central laboratory Positive result defined by ASCO guidelines Main point: Methods used for HER2 assessment in the E2197 and Kaiser Permanente studies are compared here. ECOG 2197 was used to compare central measurements of HER2 by IHC to RT-PCR. Tissue microarrays were used. Staining was performed in the ECOG central laboratory using the DAKO HercepTest. The Kaiser study was used to compare HER2 measured by FISH to RT-PCR whole sections were used. Phenopath central laboratory performed the FISH assays using the Vysis PathVysion HER2 DNA Probe Kit. Concordance in both studies was calculated using ASCO-CAP guidelines in which overall concordance is calculated excluding equivocal cases from both assays

86 HER2 distribution by RT-PCR, IHC & FISH
E2197 study *By IHC: Neg: 0 or 1+ staining, Equiv: 2+ staining or 3+ staining in < 30% cells, Pos: 3+ staining in > 30% cells **By RT-PCR: Neg: < 10.7, Equiv: ≥ 10.7-< 11.5, Pos: ≥ 11.5 Kaiser study Main point: The distribution of HER2 status by RT-PCR, IHC, and FISH are shown here. These bar charts show the distribution of positive, negative and equivocal results comparing HER2 assessment by both IHC and FISH to RT-PCR. In the ECOG IHC concordance study, IHC and RT-PCR showed similar numbers of positive cases. IHC, however showed more equivocal cases, whereas RT-PCR showed more negative cases. In the Kaiser FISH concordance study, there were again almost equal numbers of positive cases. RT-PCR had more equivocal cases, while FISH had more negative.

87 ER by RT-PCR (expression units)
Patients had continuous expression levels for HER2 and ER (E2197 study) 16 15 14 13 12 11 10 9 8 7 6 5 HER2+/HR- HER2+/HR+ + HER2+ by IHC (3+ with > 30 staining) HER2– or equivocal by IHC (0-3+ with < 30% staining) HER2 (by RT-PCR – log2) HER2-/HR- HER2-/HR+ Main point: Overall concordance excluding equivocal cases is 95%. ER expression by RT-PCR is plotted on the X axis. HER2 expression by RT-PCR is plotted on the Y axis. This is a log 2 base scale, in which each one unit increase reflects a doubling in the amount of mRNA. ER, with a 3000 fold dynamic range has a cut point if HER2, with a dynamic range of 500 fold has an equivocal range between 10.5 and Cases below 10.5 are negative, and those above 11.7 are positive. Cases that are positive by IHC are shown with a “plus sign” and negative cases are shown with a yellow dot. ER by RT-PCR (expression units) The overall range of HER2 expression in this study is approximately 1000-fold Range of HER2 expression for HER2 positive is approximately 16-fold Separate reproducibility studies indicate that standard deviation for the assay is less than 0.4 units

88 Patients had continuous expression levels for HER2 and ER (Kaiser study)
16 15 14 13 12 11 10 9 8 7 HER2+/HR- HER2+/HR+ + HER2+ by FISH HER2– by FISH Cutoff = 2.2 HER2 expression (relative to ref genes) HER2-/HR+ HER2-/HR- Main point: Overall concordance excluding equivocal cases is 97%. ER expression by RT-PCR is plotted on the X axis. HER2 expression by RT-PCR is plotted on the Y axis. This is a log 2 base scale, in which each one unit increase reflects a doubling in the amount of mRNA. ER, with a 3000 fold dynamic range has a cut point if HER2, with a dynamic range of 500 fold has an equivocal range between 10.5 and Cases below 10.5 are negative, and those above 11.7 are positive. Cases that are positive by IHC are shown with a “plus sign” and negative cases are shown with a yellow dot. ER by RT-PCR (expression units) The overall range of HER2 expression in this study is approximately 500-fold Range of HER2 expression for HER2 positive is approximately 16-fold Separate reproducibility studies indicate that standard deviation for the assay is less than 0.4 units

89 High degree of concordance between RT-PCR and FISH for HER2
HER2 concordance 2×2 Equivocal cases excluded by both assays (according to ASCO/CAP guidelines) ECOG 2197 Central IHC + Central IHC – Total Oncotype DX + 94 (78%) 4 (1%) 98 Oncotype DX – 27 (22%) 439 (99%) 466 121 443 564 CONCORDANCE 95%* [95% CI (92%,96%) Kappa 83%, 95% CI (77%,88%)] *Concordance calculated as (94+439)/564 Kaiser Study Central FISH + Central FISH – Total Oncotype DX + 55 (98%) 11 (3%) 66 Oncotype DX – 1 (2%) 408 (97%) 409 56 419 475 Main point: There is a high degree of concordance in HER2 testing between IHC/FISH and RT-PCR. Concordance rates for both IHC and FISH were calculated according to ASCO-CAP guidelines, in which equivocal cases are excluded. Concordance with IHC in ECOG 2197 was overall 95%: 78% in cases of positive agreement and 99% in cases of negative agreement. Concordance with FISH in the Kaiser Permanente study was overall 97%: 98% in cases of positive agreement and 97% in cases of negative agreement. Sparano JL, et al. ASCO Breast Cancer Symposium. 2008; Abstract 13. Baehner FL, et al. ASCO Breast Cancer Symposium. 2008; Abstract 41. Wolff AC, et al. J Clin Oncol. 2007;25: CONCORDANCE 97%* [95% CI (96%,99%), Kappa 89%, 95% CI (82%,95%)] *Concordance calculated as (55+408)/475 Sparano JL, et al. ASCO Breast Cancer Symposium. 2008; Abstract 13. Baehner FL, et al. ASCO Breast Cancer Symposium. 2008; Abstract 41. Wolff AC, et al. J Clin Oncol. 2007;25:

90 Single-Gene Testing in the Oncotype DX® assay addresses limitations with current methodologies
Both IHC and FISH are associated with variability that can affect the accuracy of test results. The impact of variability can be minimized by “normalization” strategies used in quantitative gene expression assessment as performed by quantitative RT-PCR in the Oncotype DX assay. By minimizing variability, hormone-receptor status can be more accurately reported, and treatment decisions that depend on ER or HER2 status can be made with greater confidence.

91 Single-Gene Testing in the Oncotype DX® assay addresses limitations with current methodologies
There is a high degree of overall concordance between local and central IHC and central RT-PCR for ER, PR, and hormone-receptor status. The relatively high incidence of IHC-negative hormone-receptor status positive by RT-PCR is notable and deserves further study. Quantitative RT-PCR using the Oncotype DX assay is an alternative method for determining hormone-receptor status.

92 Single-Gene Testing in the Oncotype DX® assay addresses limitations with current methodologies
For HER2 status, there is a high degree of overall concordance between central FISH for gene amplification and central RT-PCR for quantitative gene expression. For HER2 status, there is a high degree of overall concordance between central IHC for protein expression and central RT-PCR for quantitative gene expression. Quantitative RT-PCR by the Oncotype DX assay for HER2 status is an alternative to FISH. Quantitative Single-Gene Testing results may be used by local pathology laboratories as an external concordance standard for ER, PR, and HER2.

93 Prospective Multi-center Study of the Impact of Oncotype DX® Assay on Medical Oncologist and Patient Adjuvant Breast Cancer Treatment Selection Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO

94 Background A multi-center study was designed to prospectively examine whether the Oncotype DX® Recurrence Score® result can affect medical oncologist and patient adjuvant treatment selection Main point: This study was designed to prospectively examine whether the Oncotype DX® Recurrence Score® result can affect medical oncologist and patient adjuvant treatment selection. Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO

95 Methods 17 medical oncologists at 1 community and 3 academic practices participated. Each medical oncologist consecutively offered enrollment to eligible women with node-negative, ER-positive breast cancer. Each medical oncologists and consenting patient completed pre- and post-Oncotype DX® questionnaires. Medical oncologists stated their adjuvant treatment recommendation and confidence in it pre- and post-Oncotype DX testing. Patients indicated treatment choice pre- and post-Oncotype DX testing. In addition, patients completed measures for quality of life, anxiety, and decisional conflict pre and post assay. Oncotype DX results were returned to the medical oncologist and shared with patients for routine clinical care. Main point: This is the first prospective study looking at the impact of the Recurrence Score® result on patients and patient decision-making. Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO

96 Post- Recurrence Score
Change in medical oncologist treatment recommendation by Recurrence Score® result Treatment recommendation Pre- Recurrence Score Post- Recurrence Score Number (%) Mean RS Number (%) CHT 42 (47.2) 21 23 (25.8) 29 HT alone 46 (51.7) 18 60 (67.4) 16 Equipoise 1 (1.1) 19 6 (6.7) Main point: With knowledge of the Recurrence Score® result, recommendations of chemotherapy decreased from 47.2% to 25.8%. In a side by side comparison, we see that the percentage of patients offered chemotherapy pre- and post-Recurrence Score® information dropped from 47.2% to 25.8%. Also, note that there appears to be little difference in mean Recurrence Score values in those patients classified for treatment before knowledge of the Recurrence Score result was known, suggesting that doctors cannot predict well what the Recurrence Score result will be. Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO CHT, chemo and hormonal therapy; HT, hormonal therapy; equipoise defined as either chemo and hormonal therapy, hormonal therapy alone, or enrollment onto the TAILORx clinical trial; RS, Recurrence Score result Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO

97 Medical oncologist treatment recommendations changed 31.5% of the time
Medical oncologist treatment recommendation pre- to post-Oncotype DX® assay Number of cases (%) CHT  HT 20 (22.5) HT  CHT 3 (3.4) CHT or HT  Equipoise 5 (5.6) Treatment plan did not change 61 (68.5) Total 89 (100) Main point: The chemotherapy recommendation changed 31.5% of the time after the the Recurrence Score® result is known. In 22.5% of the cases, the recommendation switched from chemotherapy (plus hormone therapy) to hormone therapy alone Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO Treatment recommendation changed for 28 (31.5%) cases after results of the Oncotype DX assay were known The most common change was recommendation from CHT to HT (22.5% of cases) Lo SS, Mumby PB, Norton J, et al. J Clin Oncol. 2010;28. doi: /JCO

98 Clinical summary Oncotype DX® assay
Oncotype DX assay provides: An individualized prediction of 10-year distant recurrence risk for patients who receive 5 years of tamoxifen An individualized prediction of tamoxifen benefit An individualized prediction of chemotherapy benefit to inform adjuvant treatment decisions in women with early stage breast cancer Quantitative RT-PCR for ER/PR/HER2 is highly concordant with both IHC and FISH (using ASCO/CAP guidelines for determination of concordance). 93-96% concordant with IHC for ER 90% concordant with IHC for PR 95% and 97% concordant with IHC and FISH for HER2, respectively Oncotype DX assay is the only multi-gene expression assay recommended in both ASCO and NCCN clinical practice guidelines. The Oncotype DX® assay provides: An individualized prediction of 10-year distant recurrence risk for patients who receive 5 years of tamoxifen An individualized prediction of tamoxifen benefit An individualized prediction of chemotherapy benefit to inform adjuvant treatment decisions in women with early stage breast cancer The Oncotype DX assay has been clinically validated in >4,000 patients The Oncotype DX assay is highly concordant with both IHC and FISH testing (using ASCO/CAP guidelines for determination of concordance) 93-96% concordant with IHC for ER 90% concordant with IHC for PR 95% and 97% concordant with IHC and FISH for HER2, respectively* The Oncotype DX assay is the only proven, multi-gene expression assay recommended by ASCO and NCCN clinical practice guidelines

99 Reproducible clinical validation essential in changing standard of care: more than 4000 patients studied in 12 trials Study Type No. Pts Nodal status Providence Exploratory 136 Neg Rush* 78 Pos NSABP B-20 233 NSABP B-14* Prospective 668 MD Anderson* 149 Kaiser Permanente* Prospective case-control 790 cases/controls NSABP B-14 Prospective placebo vs Tam 645 Milan* 89 Neg/Pos NSABP B-20* Prospective Tam vs Tam+Chemo 651 ECOG 2197* Exploratory and prospective 776 SWOG 8814 367 ATAC Prospective Tam vs AI 1231 Main point: Although this assay has emerged for clinical use over the last few years, there is evidence supporting its utility in multiple studies involving over 4000 patients. References from top to bottom for the studies shown including both published studies and abstracts Esteban J, Baker J, Cronin M, et al. Tumor gene expression and prognosis in breast cancer: Multi-gene RT-PCR assay of paraffin-embedded tissue. Proc Am Soc Clin Oncol 22: 2003 (abstr 3416). Cobleigh MA, Tabesh B, Bitterman P, et al. Tumor gene expression and prognosis in breast cancer patients with 10 or more positive lymph nodes. Clin Cancer Res. 2005;11(24 Pt 1): Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351(27): Esteva FJ, Sahin AA, Cristofanilli M, et al. Prognostic role of a multigene reverse transcriptase-PCR assay in patients with node-negative breast cancer not receiving adjuvant systemic therapy. Clin Cancer Res. 2005;11(9): Habel L, Shak S, Jacobs 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;May 31;8(3):R25 [Epub ahead of print]. Gianni L, Zambetti M, Clark K, et al. Gene expression profiles in paraffin-embedded core biopsy tissue predict response to chemotherapy in women with locally advanced breast cancer. J Clin Oncol.2005;23(29): Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol. 2006;24: L. J. Goldstein, R. Gray, B. H. Childs, et al. Prognostic Utility of 21-Gene Assay in Hormone Receptor (HR) Positive Operable Breast Cancer and 0-3 Positive Axillary Nodes Treated with Adjuvant Chemohormonal Therapy (CHT): An Analysis of Intergroup Trial E2197. Journal of Clinical Oncology, 2007 ASCO Annual Meeting Proceedings Part I. Vol 25, No. 18S (June 20 Supplement), 2007: 526 *Published studies

100 Case study revisited A 55-year-old post-menopausal woman presents with an infiltrating ductal carcinoma Tumor size 1.0 cm ER/PR IHC positive HER2 IHC negative Sentinel lymph node negative Excellent overall health How should this patient be evaluated for treatment? What is her risk of disease recurrence? How likely is she to benefit from hormonal or chemotherapy?

101 Oncotype DX® Recurrence Score® result
RESULTS 11 Recurrence Score = CLINICAL EXPERIENCE Patients with a Recurrence Score of 11 in clinical validation study had an Average Rate of Distant Recurrence at 10 years of 7.4% (95% CI: 4.9%, 9.8%) Main point: The Oncotype DX® Recurrence Score® result is a continuous predictor of recurrence. The 10 year rate of distant recurrence is plotted on the Y axis. The Recurrence Score result is plotted on the X axis. The yellow line shows 10 distant recurrence rate as a continuous function of the Recurrence Score result. The 95% confidence intervals are plotted as dotted lines on either side of the line. You’ll notice that confidence intervals are tighter for lower recurrence score. When the test was first being developed, we asked oncologists what would be important in developing a tool to assess risk. They responded that we would like a tool to assess low risk patients and that low risk meant a risk of recurrence of less than 10%. So you will notice that the upper limit of the 95% confidence interval for the low risk group (recurrence score less than 18) is 10% risk. In the example here, the patient has a Recurrence Score result of 11, which corresponds to a 10 year risk of distant recurrence of 7% with 95%confidence intervals ranging from 5% - 10%.

102 Oncotype DX® assay: quantitative hormone receptor analysis
Main point: Oncotype DX® testing provides quantitative RT-PCR measurement of ER PR, and HER2. Since February 2008 for ER and PR and October 2008 for HER2, Genomic Health has included Single-Gene Testing values on all Oncotype DX reports. Clinicians have long known that ER, PR, and HER2 were among the genes used in calculating the Recurrence Score® result, and they have requested a separate analysis of these gene expression levels. Genomic Health is able to provide these particular Single-Gene Testing scores based on studies confirming concordance of RT-PCR and IHC/FISH. The Single-Gene Testing values can offer additional quantitative insight into the biology of individual tumors and aid in clinical decision-making, particularly in the intermediate Recurrence Score range, where the quantitative ER score may give insight into how much benefit the patient can expect to derive from tamoxifen and aromatase inhibitors. Pathologists can use the Single-Gene Testing values as an external standard for ongoing QC of their own immunohistochemical and FISH assays. The ASCO-CAP guidelines for HER2 testing state that laboratories performing HER2 testing must show 95% concordance with a validated outside method. Similar guidelines will be forthcoming for ER testing. The Oncotype DX assay has not been validated to predict response to anti-HER2–targeted therapy.


Download ppt "The role of the Oncotype DX® assay in breast cancer management"

Similar presentations


Ads by Google