Nuvera Biosciences and the University of Texas’s MD Anderson Cancer Center are developing a 200-gene predictor for gauging the likelihood that patients will survive longer following treatment with adjuvant hormonal therapy.
The company, an MD Anderson spin-off, presented study data for an assay it is calling NuvoSelect at the American Society of Clinical Oncology Breast Cancer Symposium in San Francisco Sept. 7-8.
The company’s data “has shown that the assay is much more significantly predictive of survival in the treated cohort” than current gene-based prognostics, Nandan Padukone, Nuvera CEO and president, told Pharmacogenomics Reporter this week.
Nuvera presented data from a study in which the assay predicted 10-year distant relapse-free survival for 277 patients who received tamoxifen treatment for five years, but not chemotherapy. According to the company, the 200-gene panel could not adequately predict DRFS in the 209 patients who did not receive any adjuvant treatment.
The company conducted a second validation study in which the 200-gene test predicted that 176 patients treated with tamoxifen for five years would likely see a 10-year DRFS. The test could not adequately predict DRFS for 298 patients who received no treatment.
“For patients without treatment, in statistical terms, the assay did not separate patients with high survival from low survival with statistical significance whereas it did so in the treated patients,” a Nuvera spokesperson explained to Pharmacogenomics Reporter in an e-mail.
Padukone noted that the company had spent the last two years clinically validating the 200-gene predictor. “It is a different [type of test] in that it shows the survival if they are treated with tamoxifen or any endocrine therapy,” he said. “By examining that same effect in the untreated patients we see that the effect is very small or not significant at all.”
“The investigators believe that this is the first genomic assay that provides independent prediction about the likely benefit from adjuvant hormonal therapy,” Nuvera said in a statement.
Currently, there are a variety of gene-based breast cancer tests that predict cancer recurrence, including products made by Genomic Health, Exagen, Agendia, and Quest. Among these players, Genomic Health’s Oncotype Dx is currently the only marketed test that also predicts whether ER-positive, node-negative breast cancer patients will benefit from tamoxifen treatment.
According to Padukone, NuvoSelect predicts tamoxifen benefit for patients who are ER-positive and either node positive or node negative.
“We have not compared our predictor [with OncotypeDx] head-to-head, and I don’t think that is the point necessarily,” Padukone said. Other than the difference in the indicated population, NuvoSelect predicts survival with hormonal therapy, not just treatment benefit with tamoxifen.
“The other difference, perhaps a little debatable, is that perhaps we are the first group to show this differentiation of how a predictor will do, still in a prognostic manner, but trying to predict survival in treated patients against a set of patients who are untreated completely,” Padukone said.
Nuvera plans to take NuvoSelect, an in vitro diagnostic multivariate index assay, through the US Food and Drug Administration’s regulatory process. Prior to FDA approval, the company said it plans to offer its tests for research use only by year-end at MD Anderson and a few other research centers.
Following the approval of Agendia’s MammaPrint breast cancer recurrence test in May, the FDA issued a special controls guidance that conferred de facto Class II status to all breast cancer prognostic tests and absolved device makers from going through the FDA’s more rigorous premarket approval process [see PGx Reporter 05-16-2007]. However, that means that predictive tests will be held as Class III devices and likely have to undergo premarket approval.
The FDA recently issued a second draft of its controversial IVDMIA guidance, which takes a risk-based approach toward such devices [see PGx Reporter 08-01-2007].
Nuvera plans to continue research to cut down on the number of genes, increase the predictive power, and improve the overall performance of NuvoSelect. “We’re looking into the best assay format, including microarray, that would be able to take this out on the market as quickly as possible,” Padukone said. He noted that the precedent for a microarray-based IVDMIA in breast cancer was set by Agendia’s MammaPrint. “So, in our thinking, the short-term nearest value would be perhaps to offer it to the medical community, to take the microarray format,” he added.
Tamoxifen, an orally active selective estrogen receptor modulator, was originally discovered by AstraZeneca. The drug is metabolized by the CYP2D6 gene. People with mutations in that gene may metabolize the drug too slowly to receive full benefit. An FDA advisory committee last October recommended the agency update tamoxifen’s label to include information about genetic testing [see PGx Reporter 11-15-2006].
“The investigators believe that this is the first genomic assay that provides independent prediction about the likely benefit from adjuvant hormonal therapy.”
According to Nuvera, existing genomic tests for breast cancer provide information about future risk but don’t provide information about patients’ chances of benefiting from individual treatment options. “For example, about 70 percent of breast cancers express the ER, however, anti-estrogen therapies, such as tamoxifen, only benefit half of these patients,” the company said in a statement.
Nuvera is also developing a 30-gene predictor of sensitivity to chemotherapy with MD Anderson. The company presented a second abstract at the ASCO conference describing a study comparing the 30-gene predictor with NuvoSelect and a 76-gene prognostic test developed by Erasmus University in the Netherlands and Veridex.
“We studied these three genomic predictors together in a scientific collaboration to see if they would provide the necessary information for physicians to decide on the best therapy or combination of therapies,” Lajos Pusztai, associate professor of breast medical oncology at MD Anderson, said in a statement. “Our results demonstrate that this is feasible.”
In this second study of 198 patients with node-negative breast cancer, 55 were predicted to be at relatively low risk of breast cancer recurrence, 38 percent of whom were predicted to be sensitive to chemotherapy and 29 percent were predicted to be sensitive to endocrine therapy.
Meanwhile, of 143 patients predicted to have a high risk of recurrence, 76 percent were predicted to have cancer unlikely to respond to endocrine therapy, 45 percent were predicted to be insensitive to chemotherapy, and 27 percent were predicted to be unlikely to respond to both therapies.
“Every patient perhaps needs all these three pieces of information. If you have a woman diagnosed with extremely good prognosis, you may also want to know the endocrine sensitivity and the chemo sensitivity because a fair amount of the fraction of patients reported as low risk do show sensitivity to chemotherapy or tamoxifen,” Padukone said.
“If you were to present that report to an oncologist today … they might say that although the patient’s chances are pretty good, if he or she is sensitive to one of these therapies let me … make sure [the patient] gets cured. Those three predictors complete the decision equation for breast cancer patients,” he added.
The studies were conducted by investigators from MD Anderson, Nuvera Biosciences, Institute Jules Bordet in Belgium, Erasmus University, and Veridex.