Prostate Cancer Guidelines Lack Molecular Test Recommendations

Daniel Lin, MD

When compared with the National Comprehensive Cancer Network (NCCN) breast cancer guidelines, the corresponding prostate cancer guidelines fall short in one distinct characteristic—the lack of a molecular test in tissue-based platforms at this point in time, said Daniel Lin, MD, chief of Urologic Oncology in the department of Urology at the University of Washington in Seattle.

Specifically, the breast cancer guidelines include a marker for “hormone receptor positive or negative status, HER2 status, and a 21-gene panel,” added Lin. Sparano and researchers¹ recently published in the New England Journal of Medicine results of a prospective trial involving about 10,000 women who had hormone receptor-positive, HER2-negative breast cancer. Those patients who had a favorable gene-expression profile had very low rates of recurrence at 5 years with endocrine therapy alone.

Lin said during his presentation at the 2015 Society of Urologic Oncology² meeting in Washington DC, that evidence is emerging for potential biomarkers, “but there’s clearly a need for biomarkers to differentiate between active surveillance with potential delayed treatment versus radical curative treatments. The guideline options reveal equipoise between those choices with no molecular marker available.”

Table: Tissue-based Platforms

Source: Lin D, University of Washington, Department of Urology.

Before a biomarker is adopted and added to the guidelines, Lin said, a number of questions need to be addressed, namely: ‘Does the data support the biomarker?’ ‘Is there clinical feasibility?’ ‘Is there appropriate application?’ and ‘Does it add to the established prognostic models?’

There are many molecular genomic tools that are available for diagnosis of prostate cancer, but Lin’s discussion focused on tissue-based platforms after the diagnosis of prostate cancer. These include the Prolaris (Myriad, Inc), Oncotype DX Prostate (Genomic Health, Inc), Decipher (GenomeDx Biosciences, Inc), and ProMark (Metamark Genetics, Inc) tests (see Table). “These tests give us insight into tumor behavior that go beyond the clinical information that we already have, such as stage, grade, prostate specific antigen level, and age of patient.”

Each test provides different readouts, yields different results, and gives the patient and physician different pieces of information and different items to inform the physician how the patient might be treated.

The Prolaris test monitors cell proliferation and “we know that proliferation in multiple solid tumors is robustly associated with disease aggressiveness,” said Lin. It is also the most widely published and has the most validation cohorts based on the endpoints: prostate cancer-specific mortality in watchful waiting,3 biochemical recurrence in post radical prostatectomy cohort,^3,4 and decision-making for choice of treatment.^5,6

The Oncotype DX Prostate test uses a 17-gene multi-pathway signature that includes stromal response, androgen signaling, cell organization, and proliferation from a formalin-fixed tissue biopsy. The test endpoint is adverse pathology (upgrading or upstaging). “The test results provide the inverse, or the likelihood of favorable pathology,” Lin said. “The rationale for this test is its role in pre-therapy decision making.”

The Decipher test is a 22-gene multi-pathway signature that uses formalin-fixed tissue from radical prostatectomies. The endpoints in this test predict metastasis after radical prostatectomy and measure the response to adjuvant radiation. “Decipher does have some data that show that it can provide prognosis information, such as how aggressive a disease might be, but it seems to have information on prediction of response of treatment, which the others don’t.” Several validation studies have shown that those patients with a high Decipher score are most at risk for developing metastases helping to guide more personalized treatment strategies for men after surgery.⁷

The ProMark test uses eight biomarkers and indicates unfavorable pathology. The test provides a personalized prediction independent of clinical and pathological characteristics. When combined with existing risk stratification methods, ProMark provides additional information to support improved clinical decision-making. Unlike genomic-based tests that require pathologists to indicate the areas of tumors, ProMark technology allows for analysis of proteins—a potentially more direct reflection of biologic activity, directly from the cancerous regions of interest.

Lin said that the true value of these tests is that they provide “that one more bit of information about how the tumor might behave and how the patient compares with a group of individuals with a similar tumor profile.” From this information, the physician can glean if the patient is on the high side and the more aggressive side of the comparison group, or if the patient is on the lower side of that group, or in the next group down. “This might give us some indication whether to watch the patient, whether to treat a patient, or whether to treat the patient aggressively, such as adding hormonal therapy with radiation therapy, as opposed to treating with radiation therapy alone.

“What is missing are most studies on how these tests affect urologists’ decisions,” said Lin. To determine that, a study was undertaken that queried providers about their therapy recommendations to their patient before undergoing a genomic test. The patient took one of the genomic tests and results were reviewed by the urologist. The researchers queried the urologists again about treatment recommendations to determine if they had changed after reviewing the test results. Then, another round of questioning determined what actually happened. “The urologists reported changing their recommendation about half the time,” said Lin. “That raises the question if the wrong decision or inappropriate treatment decision had been made initially.” The other major gap in this area is the validation of active surveillance cohorts, which is the group of patients who are primarily targeted for these tests in current clinical use.

At the conclusion of his presentation, Lin said, “I would love to have a marker included in the guidelines in the coming years. There’s clearly data to support their use, they do work to predict various endpoints in multiple validation cohorts, and they add to the clinical picture. Now, we’re just waiting for the clinical utility studies to be completed, in particular in the intended-use populations.”

References

1. Sparano JA, Gray RJ, Makower DF, et al. Prospective validation of a 21-gene expression assay in breast cancer. N Engl J Med. 2015; 373:2005-2014.
2. Lin D. Currently available genomic tools for risk stratification of localized prostate cancer. Presented at 2015 Society of Urologic Oncology, December 2-4, 2015. Washington DC.
3. Cuzick J, Swanson GP, Fisher G, et al. Prognostic value of an RNA expression signature derived from cell cycle proliferation genes in patients with prostate cancer: a retrospective study. Lancet Oncol. 2011;12(3):245-255.
4. Cooperberg MR, Simko JP, Cowan JE, et al. Validation of a cell-cycle progression gene panel to improve risk stratification in a contemporary prostatectomy cohort. J Clin Oncol. 2013;31(11):1428-1434.
5. Shore N, Concepcion R, Saltzstein D, et al. Clinical utility of a biopsy-based cell cycle gene expression assay in localized prostate cancer. Curr Med Res Opin. 2014;30(4):547-553.
6. Crawford ED, Scholz MC, Kar AJ, et al. Cell cycle progression score and treatment decisions in prostate cancer: results from an ongoing registry. Curr Med Res Opin. 2014;30(6):1025-1031.
7. Den RB, Choeurng V, Howard L, et al. Validation of a genomic classifier for prediction of metastasis following postoperative salvage radiation therapy. Presented at: ASTRO 2015; October 18-21, 2015; San Antonio, Tx. Abstract #306.