[Back to Breast Cancer Index]

Beginning in late 2004, one of our medical oncologists desired genomic testing of a breast cancer case. [Genomic...gene expression...evolution] By March 2005, we had sent maybe 5 cases to Genomic Health Labs (GHL) in Redwood City, California. Currently, we are probably asked to forward a case per week to that lab for the "oncotypeDX" genomic test. In early Jan. 2008, they began to also report the ER & PR components of their m-RNA panel (they report ER & PR m-RNA via RT-PCR, while our local IHC reports intranuclear protein product resulting from m-RNA); also reporting HER-2 by end of 2008. So, our values may not seem concordant.

This is not a test for oncogenic gene (DNA) mutations. The idea is that GHL tests the cancer tumor tissue for the m-RNA⁴ expression of a panel of 16 cancer genes & 5 reference genes², the test being performed by the method of RT-PCR.  The results of this expression panel are calculated through a said-to-be-clinically-validated proprietary algorithm (based on NSABP B-14 data & cases) which produces a recurrence-risk score somewhere between 0 and 100.  The purpose is to calculate the "recurrence score" which then falls into a low risk group (averaging a rate of distant recurrence at ten years of 7%), an intermediate group (averaging 14%), and a high risk group (averaging 31%). If low risk, patient might opt out of chemo. As of 2004, the charge for this panel is $3200-3500. Here are key points for the local Pathologist: 

• Excellently formalin-fixed tissue was the fixative used for optimization of the RT-PCR testing and is, therefore, the fixative of choice (By the TaqMAN process). Our Hartmann's fixative seems to work if no formalin fixed tissue available [L08-10462].
• Patient selection: Limited to node negative, stage I & II, and ER positive patients who will be treated with Tamoxifen. If marker studies show HER-2 gene (mutated, cancer-promoting gene = oncogene) amplification by FISH, the oncotypeDX will not produce a low risk result (L06-9471 & L07-648²).
• Many of the expression points in the array relate to ER & PR, proliferation, and to HER-2 status^{2, 5, 6}, hence the unfavorable impact of HER-2 "positivity", above.
• Block selection: Pick a block in which >50% of any epithelial component is invasive malignant epithelium.  (As with the old chemical means of testing ER & PR, too much benign epithelium can dilute the specificity for malignant cells. The test is trying to get the malignant cellular RNA in concentrated enough form to rise above the levels of "background" so that the "array expression pattern" can be discerned.) Pick block that does not have BX site as that tissue is loaded with proliferation m-RNA. Pick an area with as few macrophages as possible, as these present an interfering problem.
• Histological processing in the reference lab: Thick sections (10 microns) are cut and one section placed into each of the test tubes for the actual genomic testing.

RISK BRACKETS:

1. Low risk: group ave. 7% rate of distant recurrence at 10 years (95% confidence interval, 4-10%).
2. Intermediate risk: group ave. 14% rate of distant recurrence at 10 years (95% confidence interval, 8-20%).
3. High risk: group ave. 31% rate of distant recurrence at 10 years (95% confidence interval, 24-37%).

We have a number of our early cases in this Table where we intended to compare with a couple of other prognostic calculations. On 2/6/07, the US FDA approved a similar test done in Europe checking about 70 genes...MammaPrint (not yet available in USA). And there is a Mammostrat on the horizon⁵. And Clarient Lab has something similar (Insight DX) on the horizon as 2008 ended.

REFERENCES:

1. Genomic Health, Inc., Marketing handouts.
2. Personal, telephone, e-mail communications with expert persons providing or utilizing this test.
3. GHL update e-mail of 10/25/07 about ASCO tumor markers committee update issued 10/22/07.
4. Phillip Buckhaults, PhD @ South Carolina Cancer Center, Columbia, S. C.
5. David G. Hicks, MD, Director of S. P., Dept. Path. & Lab. Med. @ U. of Rochester Med. Ctr., Rochester, N. Y. & previously @ Cleveland Clinic. Molecular interests. He was a speaker at The Second International Course in Applied Immunohistochemistry and Molecular Pathology (Santa Barbara, Calif. 1/28/08-2/1/08).
6. Rick Baehner, M. D.,UCSF Anatomic Path. Dept. & Genomic health Labs; he was a speaker at The Second International Course in Applied Immunohistochemistry and Molecular Pathology (Santa Barbara, Calif. 1/28/08-2/1/08).

(posted 2 September  2005; latest update 16 January 2009)