Oncotype DX and Gene Expression Profiling

No, that's not a Lite-Brite... It's a DNA microarray!

Over recent years, interest has blossomed in the development of biomarkers and assays to help predict cancer outcomes. Today, we'll talk about a few of these for breast cancer, focusing on Oncotype DX (Genomic Health).

What is a prognostic marker?

A pathologic or molecular marker that can predict disease-free survival, disease-specific survival, or overall survival.

What is a predictive marker?

A pathologic or molecular marker that is used to predict response to therapy.

What are some prognostic markers for breast cancer?

Tumor grade
Hormone receptor status
HER2/neu status
Ki-67 (a proliferation marker)
S-Phase Fraction (a proliferation marker)
Mitotic Index
Cathepsin-D (a protease)
Lymphovascular invasion

What are the primary prognostic determinants for breast cancer?

Tumor size
Nodal Metastasis
Distant Metastasis

What are some predictive markers for breast cancer?

Estrogen Receptor Status
HER2/neu Status

What is a therapeutic approach that can be used for node-positive patients with tumors that overexpress HER2/neu?

trastuzumab (Herceptin, Genentech)

What are some therapeutic approaches can be used for patients with tumors that are estrogen-receptor postive?

Selective estrogen-receptor modulators (SERMs): tamoxifen, fulvestrant (Faslodex, AstraZeneca)
Aromatase inhibitors

What did the clinical trial NSABP B-14 examine?

NSABP B-14 demonstrated the benefit of 5-years of tamoxifen as adjuvant therapy for node-negative, estrogen-receptor-postive breast cancer in both survival and recurrence.
Patients (N0,ER+) were randomized to placebo (n=1453) or tamoxifen (n=1439).
Tamoxifen benefited women over placebo with hazard ratios for overall survival of 0.80 (95% CI, 0.71-0.91) and for recurrence-free survival of 0.58 (95% CI, 0.50-0.67) when followed for 15 years.

What did the clinical trial NSABP B-20 examine?

NSABP B-20 demonstrated the benefit of tamoxifen and chemotherapy for node-negative, estrogen-receptor-positive breast cancer by decreasing recurrences, but only a borderline effect on survival.
Patients (N0,ER+) were randomized to tamoxifen (n=788) or chemotherapy (CMF) and tamoxifen (n=789).
Tamoxifen plus chemotherapy benefited women over tamoxifen alone with hazard ratios for overall survival of 0.78 (95% CI, 0.60-1.01) and recurrence-free survival of
0.52 (95% CI, 0.39-0.68) when followed for 12 years. This benefit was seen primarily in patients under 60 years.

In Fisher et al. Lancet 2004, they reanalyzed the data for B-20 by comparing it with the placebo group of B-14. They found that there was about a 20-25% absolute benefit in recurrence-free survival with tamoxifen and chemotherapy compared to placebo.

Adding chemotherapy to tamoxifen likely benefits some patients in reducing recurrences, but probably not all, with overtreatment of possibly 80% of patients based only on the criteria of node-negative, estrogen-receptor-positive...

At the moment, many oncologists recommend giving both tamoxifen and chemotherapy for all ER+, node-negative tumors greater than 1cm... This may be overtreating the majority.

What is Oncotype DX?

A gene expression assay developed for node-negative, estrogen-positive breast tumors that uses information derived from fixed, paraffin-embedded tissue specimens to develop a recurrence prediction score.

Why is it nice that they use paraffin-embedded tissue specimens for the assay?

You, the surgeon, do not need to think about sending fresh or frozen tissue samples at the time of surgery! The multidisciplinary team can decide at a later time after reviewing the pathology if the test will help with management.

How is the assay performed?

Quantitative RT-PCR (Reverse transcriptase polymerase chain reaction)
This quantifies RNA for 21 different genes to give an idea of their relative level of expression. These levels are then used in an algorithm to calculate a "recurrence score".

What categories of genes are studied by this assay?

5 controls and 16 cancer-related genes (derived from 250 candidates)
1. Proliferation: Ki67, STK15, Survivin, cyclin B1, MYBL2
2. Invasion: Stromolysin 3, Cathepsin L2
3. HER2: GRB7, HER2
4. Estrogen: ER, PR, Bcl-2, SCUBE2
5. Other: GSTM1, CD68, BAG1
6. Controls: beta-actin, GAPDH, RPLPO, GUS, TFRC

How was the recurrence score developed?

Based on analysis using blocks from patients in the Tamoxifen arm of NSABP B-20, they divided patients into three groups:
1. Low Risk
2. Intermediate Risk
3. High Risk

How did they validate the assay?

In Paik et al. NEJM, 2004, they examined available blocks from patients in the tamoxifen arm of NSABP B-14. Based on stratification of recurrence score risk categories, they were able to determine Kaplan-Meier estimates for recurrence rates.

What were their findings for rates of distant recurrence at 10 years?

Low Risk Category (51%) - 7% (95% CI, 4-10%)
Intermediate Risk (22%) - 14% (95% CI, 8-20%)
High Risk Category (27%) - 31% (95% CI, 24-37%)

The recurrence score also correlated with overall survival and relapse-free interval.

When performing multivariate analysis for independent predictors of recurrence in B-14, what variables were significant?

Before including the recurrence score, age less than 50 years and tumor size greater than 2cm were independent predictors of recurrence.
After including the recurrence score, both variables dropped out, and only recurrence score remained as an independent predictor of recurrence.

What treatment algorithm could be developed utilizing Oncotype DX for node-negative, ER-positive patients who will receive tamoxifen, without factoring cost and access into the issue?

For low risk patients, tamoxifen only
For high risk patients, tamoxifen plus chemotherapy
For intermediate patients, tamoxifen with or without chemotherapy (i.e. depending on age, size, grade)

What is the cost of Oncotype DX?

US $3460 (as of 06 Mar 2007)
Medicare covers this 100%
A few insurances are beginning to cover this.

What are some other prognostic assays creeping into the clinical mainstream?

The following utilize cDNA microarray technology to classify breast tumors based on molecular signatures.
1. 70-gene profile - categorizes tumors into good and poor prognostic groups
2. Intrinsic-Subtype Signatures - Categorizes breast tumors into 5 subtypes: basal-like (ER/PR-,HER2-), ER-negative/HER2-positive, luminal A, luminal B, and normal-like.
3. Wound-Response Signatures - categorizes tumors based on activated or quiescent responses

What is a cDNA Microarray?

Sometimes called a gene chip.
A test that examines the expression of numerous genes probed simultaneously. Analysis of the profile can allow categorization of tumors based on the characteristic signatures of overexpressed and underexpressed genes.

What is an IVDMIA?

In Vitro Diagnostic Multivariate Index Assay
It is the US FDA's new classification for algorithm-based gene profiling assays including Oncotype DX and Mammaprint.

What is Mammaprint?

Mammaprint (Agendia) is the 70-gene-profile microarray. It recently received FDA approval on February 6, 2007, but has been marketed in Europe since 2005.

How do Mammaprint and Oncotype DX differ?

Oncotype DX uses quantitative RT-PCR on fixed paraffin-embedded specimens.
Mammaprint uses DNA microarray technology on freshly-obtained core biopsy specimens.
Oncotype DX is marketed for node-negative, ER-positive patients who will be starting tamoxifen to decide on starting additional adjuvant chemotherapy.
Mammaprint is marketed for T1/T2N0M0 patients under 60 years old to decide on type of adjuvant therapy.
Mammaprint was developed in the Netherlands and has been a focus of study in Europe. Oncotype DX has been developed and studied primarily in North America.

In the future, the EORTC's MINDACT trial (Microarray in Node-Negative Disease May Avoid Chemotherapy), which uses the Mammaprint 70-gene profile, will give more to think about...

Phew! That's all for now... SG

Sources:
Schwartz' Principles of Surgery, 8th ed.
Fisher B, Jeong JH, Bryant J, Anderson S, Dignam J, Fisher ER, Wolmark N. Treatment of lymph-node-negative, oestrogen-receptor-positive breast cancer: long-term findings from National Surgical Adjuvant Breast and Bowel Project randomized clinical trials. Lancet 2004; 364: 858-868.
Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, Baehner FL, Walker MG, Watson D, Park T, Hiller W, Fisher ER, Wickerham DL, Bryant J, Wolmark N. A Multigene Assay to Predict Recurrence of Tamoxifen-Treated, Node-Negative Breast Cancer. NEJM 2004; 351: 2817-2826.
O'Shaugnessy JA. Molecular Signatures Predict Outcomes in Breast Cancer. NEJM 2006; 355: 615-617.
Mammaprint
Cost information obtained from a phone conversation with a representative from Genomic Health, 1-866-ONCOTYPE