By Turna Ray
With the publication this week of a validation study for its Oncotype DX colon cancer test, Genomic Health is hoping that more insurers will now be willing to pay for the diagnostic.
However, the same study was unable to confirm that the test can predict which patients will benefit from treatment with chemotherapy — a finding that may negatively impact the test's adoption, according to some industry observers.
"The continuous 12-gene [recurrence score] has been validated in a prospective study for assessment of recurrence risk in patients with stage II colon cancer after surgery and provides prognostic value that complements [other markers such as] T stage and [mismatch repair protein status]," the study authors concluded in their report on the Quick and Simple and Reliable, or QUASAR, study published this week in the Journal of Clinical Oncology.
A treatment score algorithm that was also assessed in the trial, however, "was not predictive of chemotherapy benefit," the authors concluded.
The QUASAR results, along with Palmetto GBA's recent decision to cover the test for the Medicare population, "pave the way for expanded reimbursement and increased patient access to the Oncotype DX colon cancer test," said Steven Shak, Genomic Health's chief medical officer, in a statement this week.
However, some Wall Street analysts had more cautious projections. "Even with increased coverage, we ultimately see the utility of the test limited by its inability to predict the benefit of chemotherapy, which we believe is what made Oncotype DX in node-negative breast cancer a remarkably successful diagnostic," David Ferreiro, an analyst at the investment bank Oppenheimer, wrote in a research note.
In the QUASAR study, researchers led by Richard Gray of the University of Oxford analyzed fixed paraffin-embedded tumor tissue from more than 1,400 Stage II colon cancer patients treated with adjuvant fluoropyrimidine chemotherapy versus surgery alone. The researchers assessed the patients' risk of recurrence and the likelihood that they would benefit from chemotherapy based on the expression levels of 13 cancer-related genes and from five reference genes. Out of the 13 cancer-related genes in the panel, seven genes were associated with disease recurrence and six were informative of treatment benefit.
"Risk of recurrence was significantly associated with [the Oncotype DX] recurrence score," the study authors found. At three years, patients defined as being at low risk of recurrence had a 12 percent risk; those categorized as having intermediate risk of recurrence had 18 percent risk; and those deemed to be at high risk of recurrence had 22 percent risk.
The researchers also gauged cancer recurrence risk by conventional measures, such as T-stage and mismatch repair status, nodes examined, tumor grade, and lymphovascular invasion. Of these, T4 stage (when the tumor has grown through the bowel wall) and MMR status (associated with high microsatellite instability) were also independent prognostic factors of cancer recurrence.
Based on the significance of this finding, Genomic Health plans to start providing MMR immunohistochemistry testing services later this year in addition to the Oncotype DX test (PGx Reporter 11/2/2011).
The QUASAR study authors noted that the "12-gene recurrence score will have the greatest clinical utility when used as a complement to T-stage and MMR status, specifically for patients who have T3, MMR-proficient, stage II disease," which represents around 70 percent of stage II colon cancer patients.
However, when it came to validating the predictive value of the test, "there was no trend for increased benefit from chemotherapy at higher treatment scores," the researchers reported.
The study authors attribute this to the fact that the gene markers underlying the treatment score algorithm were not developed using samples from patients randomized to chemotherapy or surgery alone.
In developing the gene signatures for the recurrence and treatment scores, the researchers measured the expression levels of 761 candidate genes on samples from four studies involving patients with resected stage II and III colon cancer. Two of these studies involved patients who were treated with surgery alone, while the other two involved patients who were treated with surgery followed by adjuvant fluorouracil/folinic acid chemotherapy.
"This highlights the importance of validation of other putative markers of responsiveness to chemotherapy in large cohorts of patients randomly assigned to either chemotherapy or to a surgery-alone arm," Gray et al. noted, acknowledging that "few such cohorts, other than QUASAR, exist."
Although the treatment score with regard to FU/FA benefit wasn't validated, Genomic Health is continuing to investigate response markers associated with other kinds of chemotherapy treatments. The company said it has completed laboratory analysis of a gene identification study to predict the benefit of oxaliplatin chemotherapy in stage II and III colon cancer patients, using samples from the landmark NSABP C-07 study. The results are slated for release next year.
It is estimated that more than 150,000 people worldwide are diagnosed with stage II colon cancer annually. Based on this, Oppenheimer's Ferreiro projected that Genomic Health could potentially sell over 1,100 tests this year and more than 1,400 tests next year. He added that reporting MMR status and establishing the test's ability to predict oxaliplatin benefit in stage II colon cancer would likely bolster sales.
Genomic Health maintained that the QUASAR study in no way diminishes the clinical utility of the Oncotype DX colon cancer test, and discouraged comparisons of this test with its breast cancer assay. "Unlike breast cancer, in which robust biomarkers including ER, PR, and HER2 have been in clinical use for years, we are just beginning to see the use of biomarkers in clinical practice in colon cancer," a Genomic Health spokesperson told PGx Reporter.
"Our discussions with oncologists support the clinical utility of the quantitative individualized recurrence risk assessment using the colon cancer recurrence score in guiding patients and physicians to make more informed decisions about the use of adjuvant chemotherapy for stage II colon cancer," the spokesperson said.
Although the study authors characterized QUASAR as the first large validation study that convincingly demonstrates the ability of a prospectively defined gene expression test to gauge the risk of stage II colon cancer recurrence better than conventional prognostic methods, they also pointed out several limitations of the trial.
One drawback was that tissue samples were collected from only 68 percent of patients in the study. However, according to the study authors, this isn't an entirely bad record when compared with other recently reported trials involving prospective tissue collection.
Further, only 38 percent of patients had at least 12 nodes examined, which is a lower sampling rate than standard clinical practice. "Future studies of [the recurrence score] should include patients with stage II colon cancer who have had adequate nodal sampling, reflective of current practice," the study authors recommended.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.