Outcome prediction doesn’t reflect clinical utility
Article Type
Changed
Fri, 01/04/2019 - 12:14
Display Headline
Prolaris test eyed as predictor of prostate cancer outcomes

CHICAGO – Prostate cancer outcomes were predicted by a test that measures the expression of cell cycle progression genes, according to results from a retrospective analysis of prostate tissue samples from five patient cohorts.

The Prolaris test gives each tissue sample a cell cycle progression (CCP) score based on measures of 31 CCP genes, normalized to 15 "housekeeper" genes. A unit change in the test is defined as a doubling in CCP genes. For each unit increase in the test’s score, there was a two- to threefold increase in the risk of disease progression, Dr. Jack M. Cuzick reported at the annual meeting of the American Society of Clinical Oncology.

The CCP signature of Myriad Genetics’ Prolaris test was a highly significant predictor of outcome, said Dr. Cuzick of the Wolfson Institute of Preventive Medicine, London. In all five studies, the hazard ratio per unit change in the CCP score was similar, ranging from 1.89 to 2.92. The findings indicate that the effect size for the CCP score is robust in multiple patient cohorts and diverse clinical settings.

The test provides information for differentiating aggressive and indolent disease beyond that available from clinicopathologic variables, he said. As the natural history of prostate cancer can be variable and difficult to predict, the Prolaris test could help to match treatment more appropriately to each individual’s risk of progression.

In the study that examined the test’s predictive value, five patient groups were evaluated. Formalin-fixed tissue samples were obtained from two English patient cohorts that were conservatively managed (n = 337 and 349), two U.S. patient cohorts that underwent radical prostatectomy (366 men treated at Scott & White Hospital, Temple, Tex.; and 413 men treated at the University of California, San Francisco), and one U.S. cohort that underwent external beam radiation therapy (141 men treated at the Durham, N.C., VA Medical Center).

The cohort of conservatively managed English patients was from the late 1990s and had more than 15 years of follow up. In the 337-patient cohort diagnosed via transurethral resection of the prostate (TURP) and conservatively managed, there were 57 deaths from prostate cancer. In the 349-patient cohort diagnosed via needle biopsy and conservatively managed, there were 90 deaths from prostate cancer.

For each unit increase in the CCP score, the hazard ratio for the cohort diagnosed via TURP was 2.9 and the hazard ratio for those diagnosed via needle biopsy was 2. The CCP score was the dominant variable for predicting death from prostate cancer in univariate analysis (P = 6.1 x 10–22 after diagnosis via TURP, and P = 8.6 x 10–10 after diagnosis via needle biopsy). In both studies, the CCP score remained highly significant in multivariate analysis and was a stronger predictor of disease-specific mortality than other prognostic variables, he said.

In the U.S. prostatectomy cohorts, there were 132 biochemical recurrences (BCRs) in the first cohort and 83 BCRs in the second cohort. With each unit increase in the CCP score, there was a doubling of risk for recurrence. After prostatectomy, the CCP score predicted BCR in univariate analysis (Scott & White: P = 5.6 x 10–9; University of California: P = 2.23 x 10–6) and provided additional prognostic information in multivariate analysis (Scott & White: P = 3.3 x 10–6; University of California: P = 9.5 x10–5).

After radiation therapy, the CCP score predicted BCR in univariate (P = .0017) and multivariate (P = .034) analysis. In the 141-patient cohort that was diagnosed by needle biopsy and underwent external beam radiation, there were 19 prostate cancer deaths and more than a doubling of risk with each unit increase in CCP score.

CCP scores only modestly correlated with the Gleason score and prostate-specific antigen (PSA) value. The test adds value beyond those measures, Dr. Cuzick said.

CCP scores predict patient outcome in multiple clinical settings, provide independent information beyond clinicopathological variables, and help to further differentiate aggressive from indolent prostate cancer. With low-grade Gleason 6 cancers, the results can aid in telling who is at low risk and who needs aggressive therapy, he concluded.

The study was funded by Myriad Genetics, the maker of the Prolaris test. Dr. Cuzick received honoraria and research support from Myriad.

Body

The ability to improve clinical management by finding prostate cancer patients who would benefit from more – or less – therapy is much needed. Clinicians are concerned that many prostate cancer patients are now overtreated, but they lack reliable prognostic guides.

Cell cycle progression (CCP) scores are interesting retrospectively, but how much are they able to improve on CAPRA (Cancer of the Prostate Risk Assessment) scores for predicting prognosis?

Prostate cancer is uniquely multifocal, with most men having multiple independent foci of cancer. In the example of the conservatively managed patients, if one is looking at men with indolent disease who have low-volume disease as a single core of one or two foci, is one really going to be able to predict the biologic outcome of the cancer? Those who fail after a surveillance approach often do so early and had undersampling of their disease. So it hasn’t been proven yet that this test can predict the behavior of cancer that hasn’t been sampled.

The CCP results proved to be statistically significant, but that finding does not indicate clinical utility. It’s not known whether the novel biomarkers in this test improve on existing markers. You find yourself asking what you would do differently in a patient whose risk of progression goes from 7% to 12%.

Even if a test independently predicts outcome, that doesn’t necessarily indicate it has clinical utility. The ability to improve clinical management is key to the adoption of new prognostic tests. The real question is whether CCP results improve on the existing model. Does the test improve on CAPRA for prognosis?

Dr. Scott Tomlins is with the department of urology at the University of Michigan Health System, Ann Arbor. He was the invited discussant of the paper at the meeting. Dr. Tomlins disclosed that he is a consultant to and receives honoraria from Ventana Medical Systems/Roche. He has patents via the University of Michigan on several diagnostic genetic tests.

Meeting/Event
Author and Disclosure Information

Publications
Topics
Legacy Keywords
Prostate cancer, cell cycle progression genes, Prolaris, CCP, Dr. Jack M. Cuzick, American Society of Clinical Oncology
Sections
Author and Disclosure Information

Author and Disclosure Information

Meeting/Event
Meeting/Event
Body

The ability to improve clinical management by finding prostate cancer patients who would benefit from more – or less – therapy is much needed. Clinicians are concerned that many prostate cancer patients are now overtreated, but they lack reliable prognostic guides.

Cell cycle progression (CCP) scores are interesting retrospectively, but how much are they able to improve on CAPRA (Cancer of the Prostate Risk Assessment) scores for predicting prognosis?

Prostate cancer is uniquely multifocal, with most men having multiple independent foci of cancer. In the example of the conservatively managed patients, if one is looking at men with indolent disease who have low-volume disease as a single core of one or two foci, is one really going to be able to predict the biologic outcome of the cancer? Those who fail after a surveillance approach often do so early and had undersampling of their disease. So it hasn’t been proven yet that this test can predict the behavior of cancer that hasn’t been sampled.

The CCP results proved to be statistically significant, but that finding does not indicate clinical utility. It’s not known whether the novel biomarkers in this test improve on existing markers. You find yourself asking what you would do differently in a patient whose risk of progression goes from 7% to 12%.

Even if a test independently predicts outcome, that doesn’t necessarily indicate it has clinical utility. The ability to improve clinical management is key to the adoption of new prognostic tests. The real question is whether CCP results improve on the existing model. Does the test improve on CAPRA for prognosis?

Dr. Scott Tomlins is with the department of urology at the University of Michigan Health System, Ann Arbor. He was the invited discussant of the paper at the meeting. Dr. Tomlins disclosed that he is a consultant to and receives honoraria from Ventana Medical Systems/Roche. He has patents via the University of Michigan on several diagnostic genetic tests.

Body

The ability to improve clinical management by finding prostate cancer patients who would benefit from more – or less – therapy is much needed. Clinicians are concerned that many prostate cancer patients are now overtreated, but they lack reliable prognostic guides.

Cell cycle progression (CCP) scores are interesting retrospectively, but how much are they able to improve on CAPRA (Cancer of the Prostate Risk Assessment) scores for predicting prognosis?

Prostate cancer is uniquely multifocal, with most men having multiple independent foci of cancer. In the example of the conservatively managed patients, if one is looking at men with indolent disease who have low-volume disease as a single core of one or two foci, is one really going to be able to predict the biologic outcome of the cancer? Those who fail after a surveillance approach often do so early and had undersampling of their disease. So it hasn’t been proven yet that this test can predict the behavior of cancer that hasn’t been sampled.

The CCP results proved to be statistically significant, but that finding does not indicate clinical utility. It’s not known whether the novel biomarkers in this test improve on existing markers. You find yourself asking what you would do differently in a patient whose risk of progression goes from 7% to 12%.

Even if a test independently predicts outcome, that doesn’t necessarily indicate it has clinical utility. The ability to improve clinical management is key to the adoption of new prognostic tests. The real question is whether CCP results improve on the existing model. Does the test improve on CAPRA for prognosis?

Dr. Scott Tomlins is with the department of urology at the University of Michigan Health System, Ann Arbor. He was the invited discussant of the paper at the meeting. Dr. Tomlins disclosed that he is a consultant to and receives honoraria from Ventana Medical Systems/Roche. He has patents via the University of Michigan on several diagnostic genetic tests.

Title
Outcome prediction doesn’t reflect clinical utility
Outcome prediction doesn’t reflect clinical utility

CHICAGO – Prostate cancer outcomes were predicted by a test that measures the expression of cell cycle progression genes, according to results from a retrospective analysis of prostate tissue samples from five patient cohorts.

The Prolaris test gives each tissue sample a cell cycle progression (CCP) score based on measures of 31 CCP genes, normalized to 15 "housekeeper" genes. A unit change in the test is defined as a doubling in CCP genes. For each unit increase in the test’s score, there was a two- to threefold increase in the risk of disease progression, Dr. Jack M. Cuzick reported at the annual meeting of the American Society of Clinical Oncology.

The CCP signature of Myriad Genetics’ Prolaris test was a highly significant predictor of outcome, said Dr. Cuzick of the Wolfson Institute of Preventive Medicine, London. In all five studies, the hazard ratio per unit change in the CCP score was similar, ranging from 1.89 to 2.92. The findings indicate that the effect size for the CCP score is robust in multiple patient cohorts and diverse clinical settings.

The test provides information for differentiating aggressive and indolent disease beyond that available from clinicopathologic variables, he said. As the natural history of prostate cancer can be variable and difficult to predict, the Prolaris test could help to match treatment more appropriately to each individual’s risk of progression.

In the study that examined the test’s predictive value, five patient groups were evaluated. Formalin-fixed tissue samples were obtained from two English patient cohorts that were conservatively managed (n = 337 and 349), two U.S. patient cohorts that underwent radical prostatectomy (366 men treated at Scott & White Hospital, Temple, Tex.; and 413 men treated at the University of California, San Francisco), and one U.S. cohort that underwent external beam radiation therapy (141 men treated at the Durham, N.C., VA Medical Center).

The cohort of conservatively managed English patients was from the late 1990s and had more than 15 years of follow up. In the 337-patient cohort diagnosed via transurethral resection of the prostate (TURP) and conservatively managed, there were 57 deaths from prostate cancer. In the 349-patient cohort diagnosed via needle biopsy and conservatively managed, there were 90 deaths from prostate cancer.

For each unit increase in the CCP score, the hazard ratio for the cohort diagnosed via TURP was 2.9 and the hazard ratio for those diagnosed via needle biopsy was 2. The CCP score was the dominant variable for predicting death from prostate cancer in univariate analysis (P = 6.1 x 10–22 after diagnosis via TURP, and P = 8.6 x 10–10 after diagnosis via needle biopsy). In both studies, the CCP score remained highly significant in multivariate analysis and was a stronger predictor of disease-specific mortality than other prognostic variables, he said.

In the U.S. prostatectomy cohorts, there were 132 biochemical recurrences (BCRs) in the first cohort and 83 BCRs in the second cohort. With each unit increase in the CCP score, there was a doubling of risk for recurrence. After prostatectomy, the CCP score predicted BCR in univariate analysis (Scott & White: P = 5.6 x 10–9; University of California: P = 2.23 x 10–6) and provided additional prognostic information in multivariate analysis (Scott & White: P = 3.3 x 10–6; University of California: P = 9.5 x10–5).

After radiation therapy, the CCP score predicted BCR in univariate (P = .0017) and multivariate (P = .034) analysis. In the 141-patient cohort that was diagnosed by needle biopsy and underwent external beam radiation, there were 19 prostate cancer deaths and more than a doubling of risk with each unit increase in CCP score.

CCP scores only modestly correlated with the Gleason score and prostate-specific antigen (PSA) value. The test adds value beyond those measures, Dr. Cuzick said.

CCP scores predict patient outcome in multiple clinical settings, provide independent information beyond clinicopathological variables, and help to further differentiate aggressive from indolent prostate cancer. With low-grade Gleason 6 cancers, the results can aid in telling who is at low risk and who needs aggressive therapy, he concluded.

The study was funded by Myriad Genetics, the maker of the Prolaris test. Dr. Cuzick received honoraria and research support from Myriad.

CHICAGO – Prostate cancer outcomes were predicted by a test that measures the expression of cell cycle progression genes, according to results from a retrospective analysis of prostate tissue samples from five patient cohorts.

The Prolaris test gives each tissue sample a cell cycle progression (CCP) score based on measures of 31 CCP genes, normalized to 15 "housekeeper" genes. A unit change in the test is defined as a doubling in CCP genes. For each unit increase in the test’s score, there was a two- to threefold increase in the risk of disease progression, Dr. Jack M. Cuzick reported at the annual meeting of the American Society of Clinical Oncology.

The CCP signature of Myriad Genetics’ Prolaris test was a highly significant predictor of outcome, said Dr. Cuzick of the Wolfson Institute of Preventive Medicine, London. In all five studies, the hazard ratio per unit change in the CCP score was similar, ranging from 1.89 to 2.92. The findings indicate that the effect size for the CCP score is robust in multiple patient cohorts and diverse clinical settings.

The test provides information for differentiating aggressive and indolent disease beyond that available from clinicopathologic variables, he said. As the natural history of prostate cancer can be variable and difficult to predict, the Prolaris test could help to match treatment more appropriately to each individual’s risk of progression.

In the study that examined the test’s predictive value, five patient groups were evaluated. Formalin-fixed tissue samples were obtained from two English patient cohorts that were conservatively managed (n = 337 and 349), two U.S. patient cohorts that underwent radical prostatectomy (366 men treated at Scott & White Hospital, Temple, Tex.; and 413 men treated at the University of California, San Francisco), and one U.S. cohort that underwent external beam radiation therapy (141 men treated at the Durham, N.C., VA Medical Center).

The cohort of conservatively managed English patients was from the late 1990s and had more than 15 years of follow up. In the 337-patient cohort diagnosed via transurethral resection of the prostate (TURP) and conservatively managed, there were 57 deaths from prostate cancer. In the 349-patient cohort diagnosed via needle biopsy and conservatively managed, there were 90 deaths from prostate cancer.

For each unit increase in the CCP score, the hazard ratio for the cohort diagnosed via TURP was 2.9 and the hazard ratio for those diagnosed via needle biopsy was 2. The CCP score was the dominant variable for predicting death from prostate cancer in univariate analysis (P = 6.1 x 10–22 after diagnosis via TURP, and P = 8.6 x 10–10 after diagnosis via needle biopsy). In both studies, the CCP score remained highly significant in multivariate analysis and was a stronger predictor of disease-specific mortality than other prognostic variables, he said.

In the U.S. prostatectomy cohorts, there were 132 biochemical recurrences (BCRs) in the first cohort and 83 BCRs in the second cohort. With each unit increase in the CCP score, there was a doubling of risk for recurrence. After prostatectomy, the CCP score predicted BCR in univariate analysis (Scott & White: P = 5.6 x 10–9; University of California: P = 2.23 x 10–6) and provided additional prognostic information in multivariate analysis (Scott & White: P = 3.3 x 10–6; University of California: P = 9.5 x10–5).

After radiation therapy, the CCP score predicted BCR in univariate (P = .0017) and multivariate (P = .034) analysis. In the 141-patient cohort that was diagnosed by needle biopsy and underwent external beam radiation, there were 19 prostate cancer deaths and more than a doubling of risk with each unit increase in CCP score.

CCP scores only modestly correlated with the Gleason score and prostate-specific antigen (PSA) value. The test adds value beyond those measures, Dr. Cuzick said.

CCP scores predict patient outcome in multiple clinical settings, provide independent information beyond clinicopathological variables, and help to further differentiate aggressive from indolent prostate cancer. With low-grade Gleason 6 cancers, the results can aid in telling who is at low risk and who needs aggressive therapy, he concluded.

The study was funded by Myriad Genetics, the maker of the Prolaris test. Dr. Cuzick received honoraria and research support from Myriad.

Publications
Publications
Topics
Article Type
Display Headline
Prolaris test eyed as predictor of prostate cancer outcomes
Display Headline
Prolaris test eyed as predictor of prostate cancer outcomes
Legacy Keywords
Prostate cancer, cell cycle progression genes, Prolaris, CCP, Dr. Jack M. Cuzick, American Society of Clinical Oncology
Legacy Keywords
Prostate cancer, cell cycle progression genes, Prolaris, CCP, Dr. Jack M. Cuzick, American Society of Clinical Oncology
Sections
Article Source

AT THE ASCO ANNUAL MEETING 2013

PURLs Copyright

Inside the Article

Vitals

Major finding: In conservatively managed prostate cancer patients, the cell cycle progression score in tissue samples was the dominant variable for predicting death from prostate cancer in univariate analysis (P = 6.1 x 10–22 after diagnosis via TURP, and P = 8.6 x 10–10 after diagnosis via needle biopsy).

Data source: A retrospective study of tissue samples from more than 1,600 patients in five patient cohorts who were either managed conservatively, underwent prostatectomy, or received external beam radiotherapy.

Disclosures: The study was funded by Myriad Genetics, the maker of the Prolaris test. Dr. Cuzick received honoraria and research support from Myriad.