In a head-to-head comparison of two multi-gene breast cancer recurrence tests, Genomic Health's Oncotype DX and Agendia's MammaPrint yielded divergent results.
The study, led by researchers from Genomic Health and L’Institut du Sein at the Paris Breast Center, found that Oncotype DX identified a lower number of breast cancer patients as being at high risk of recurrence than did MammaPrint. The researchers presented the results at the San Antonio Breast Cancer Symposium held earlier this month.
In the study, Bruno Poulet of L'Institut du Sein and colleagues used Oncotype DX to retest 67 samples from patients with estrogen receptor-positive, HER2-negative early stage breast cancer who were previously analyzed by MammaPrint at the Paris Breast Center. These patients were predominantly deemed to be at low or at intermediate risk of recurrence when doctors assessed their prognosis just by clinical and pathological features.
"This direct comparison demonstrates that the MammaPrint and Oncotype DX tests classify a large proportion of patients differently," Poulet et al. reported. "Of note, nearly half of the patients with a high risk MammaPrint result had a low recurrence score [by Oncotype DX] indicating minimal, if any, benefit from chemotherapy."
The reasons for the discordance between the two test platforms differ depending on whether one talks to Agendia or Genomic Health. According to officials from both companies, the risk classifications from the two tests likely vary due to the fact that MammaPrint is intended to gauge recurrence risk before breast cancer patients receive systemic therapy such as chemotherapy or hormonal therapy, while Oncotype DX has been validated largely in patients after treatment with the hormonal therapy tamoxifen. Additionally, MammaPrint provides a binary result, while Oncotype DX provides a recurrence risk score between one and 100 that corresponds to high risk, intermediate risk, or low risk of recurrence.
The authors of the study acknowledged these differences, noting that the "development and validation cohorts for these genomic assays are significantly different," but argued that they carried out the comparison because "the tests are frequently believed to provide equivalent information."
Indeed, both tests are used to assess the risk of breast cancer recurrence and, subsequently, guide treatment strategy. Patients with early-stage estrogen receptor-positive breast cancer usually receive surgery to remove the tumor and then are treated with a combination of chemotherapy and hormonal therapy to reduce the chances that their cancer will return. However, patients who are at low risk of recurrence can potentially avoid the toxicities of chemotherapy, and be treated just with hormonal therapy.
The analysis by Poulet et al. revealed that Oncotype DX yielded a recurrence risk score for all 67 patient samples, while MammaPrint could not produce a test result for 10 of the samples. For the 57 samples for which both tests provided a result, Poulet and colleagues reported that Oncotype DX found only two samples to be at high risk of recurrence, while MammaPrint determined 22 patient samples to be high risk.
Oncotype DX classified 22 samples as being at intermediate risk of recurrence. These were evenly split between the MammaPrint low-risk and high-risk categories. An Oncotype DX recurrence score below 18 indicates that patients will respond well to tamoxifen, but will gain little benefit from treatment with chemotherapy. Those who receive a recurrence score higher than 31 will likely benefit from chemotherapy in addition to tamoxifen treatment. For patients in the intermediate recurrence risk category — between 18 and 31 — it is unclear whether chemotherapy is necessary and whether patients should be treated with tamoxifen alone.
The researchers point out in the abstract presented at SABCS that 45 percent of patients who were at high risk of recurrence by MammaPrint but received a low recurrence score result by Oncotype DX also had quantitative estrogen receptor expression greater than 9.5 expression units. This, the researchers said, "is associated with likely hormonal therapy benefit."
The two tests reported a similar number of low risk patients, with 33 patients deemed low risk by Oncotype DX and 35 patients categorized as low risk by MammaPrint.
"There was no clear association between tumor characteristics and MammaPrint failure or differences in risk classification," Poulet et al. wrote in the abstract.
Reasons for Discordance
Oncotype DX, a 21-gene RT-PCR assay, has been on the market as a laboratory-developed test since 2004. The gene-expression assay is indicated to be used on women with HER2-negative, estrogen receptor-positive, early-stage breast cancer that hasn't spread to the lymph nodes. Studies have shown that post-menopausal women who have hormone receptor-positive breast cancer that has spread to the lymph nodes may also derive benefit from testing with Oncotype DX.
MammaPrint, which has gained 510(k) clearance from the US Food and Drug Administration, analyzes the expression of 70 genes using a microarray platform. The test gauges which early-stage breast cancer patients are at risk of distant recurrence following surgery, independent of estrogen receptor status and prior treatment.
Oncotype DX and MammaPrint are different in terms of the underlying platform technology, in the manner they were validated, and in the way they characterize patients' risk of recurrence. Still, the tests are competitors in the breast cancer recurrence testing market. As such, Genomic Health and Agendia have divergent views on the why the recurrence risk results from the two tests are discordant, particularly with regard to those patients who are deemed to be at high risk by MammaPrint.
Genomic Health Chief Medical Officer Steve Shak believes that the head-to-head comparison shows that Oncotype DX is more useful than MammaPrint since it has been validated to gauge which patients will benefit from chemotherapy. "To date, MammaPrint has not been able to demonstrate chemotherapy benefit in large, randomized clinical trials with homogeneous and relevant populations," Shak told PGx Reporter via email. "Unlike MammaPrint, which has only been shown to be prognostic and in only non‐randomized studies with heterogeneous populations, Oncotype DX predicts chemotherapy benefit as demonstrated in multiple randomized clinical trials."
Meanwhile, Agendia officials maintain that given the small size of the study presented at SABCS, it is difficult to draw any definitive conclusions about the prognostic accuracy of either test. Still, Agendia officials asserted that MammaPrint is the better tool for assessing breast cancer recurrence, since Oncotype DX was validated in patients treated with tamoxifen, while MammaPrint is intended to be used to gauge risk of recurrence for patients ahead of systemic treatment.
Bas van der Baan, Agendia's VP of clinical affairs, suggested that MammaPrint may have found more patients to be at high risk in the L'Institut du Sein study because the test was developed to measure risk in untreated patients. "The biggest difference between the two assays is that our assay was developed in untreated patients, and gives the risk of recurrence after surgery and radiation," he said. "Oncotype DX is developed on tamoxifen-treated patients. So, their risk assessment is valid only after five years of tamoxifen."
As a result, he said, "you would indeed expect that there is a group of patients that is high risk by our assay, that have a high risk of distant metastases, if left untreated."
Counter to this, Genomic Health's Shak maintained that Oncotype DX provides both prognostic and predictive information, and therefore is an improvement over MammaPrint. “The Oncotype DX test provides information on both recurrence risk and whether there is benefit from hormonal therapy versus chemotherapy, whereas MammaPrint provides information on risk without systemic treatment," Shak said. "These results highlight the value of providing patients with a quantifiable, individualized assessment of both risk and treatment benefit, which reflects continuous tumor biology. Clearly, methods matter and all genomics tests are not equivalent."
Of course, which test provided the correct risk assessment can only be revealed by following longitudinally to track their outcomes. Since the L'Institut du Sein study is so small, no long-term follow-up is planned.
However, earlier this year, Agendia presented data from a study that showed that doctors can withhold chemotherapy for patients with a low risk of recurrence by MammaPrint without impacting their five-year survival. Data from the prospective, non-randomized Microarray Prognostics in Breast Cancer, or RASTER, study showed that the use of MammaPrint led to a 20 percent reduction in adjuvant chemotherapy use in patients whom the test determined to be at low risk of recurrence (PGx Reporter 3/28/2012).
Picking the Right Test
According to Agendia officials, the fact that MammaPrint could not analyze 10 of the 67 samples is due to the fact that the study with L'Institut du Sein was originally conducted while MammaPrint was only analyzing fresh tissue samples. The company began accepting formalin-fixed paraffin-embedded biopsy samples in January this year, which officials said significantly reduces the chance that MammaPrint won't be able to analyze the sample due to the lack of tumor material (PGx Reporter 4/25/2012).
Douglas Bradley, Agendia's executive VP of global marketing, noted that it's not unusual to get a percentage of fresh tumor samples that are insufficient for molecular analysis. "A lot of the time … they think the tumor is there, but if it's not there then we can't process the tissue," Bradley said.
"If it's FFPE, they can cut slides off of the block and see the cancer in there, so your [rate of] insufficient tumor samples goes way, way down," he said, asserting that now that Agendia's test analyzes FFPE samples, MammaPrint's rate of insufficient tumor samples is on par with Oncotype DX.
According to Agendia, the move from analyzing fresh tissue to FFPE has resulted in a doubling of revenues year over year. However, since Agendia is privately held, Bradley could not provide further details on the company's financials.
For the three months ended Sept. 30, Genomic Health's revenues jumped to $59 million from $52 million in the year-ago quarter and the company delivered more than 18,000 Oncotype DX test results during the quarter compared to approximately 16,900 results in the prior-year period. However, Genomic Health reported weaker-than-expected testing volumes for the quarter, which company officials attributed to a number of factors, among them fewer physician visits and increased competitive activity (PGx Reporter 11/14/2012).
As competitors, Agendia is highlighting the fact that MammaPrint provides a binary answer, while Genomic Health is emphasizing Oncotype DX's continuous score, which translates into more nuanced risk groupings.
"It is known that underlying biology in breast cancer is continuous," Shak said. "The Recurrence Score provides a quantitative risk assessment that allows physicians to make an individualized treatment decision based on underlying tumor biology. Presenting test results as simply high versus low withholds important patient-specific information and drastically reduces the ability to individualize treatment."
Although by now there is sufficient clinical support showing that women deemed at high risk of recurrence by Oncotype DX should be treated with chemotherapy and low-risk patients can avoid chemo and do well with just hormonal treatment, the treatment strategy for patients with intermediate scores isn't yet clear. A large clinical trial is currently ongoing to figure out whether patients in this risk category should receive chemotherapy.
Agendia's van der Baan cited this as a shortcoming of the L'Institut du Sein study. "What surprised us was in the poster, [the authors] didn't mention patients that received intermediate results at all," he said.
In the past, breast cancer experts PGx Reporter has spoken to have noted that while some doctors may want a more nuanced risk characterization from genetic testing, intermediate scores from Oncotype DX may heighten uncertainty about how to treat these patients. Other experts have further discussed the benefits of having a binary test result in busy community practices where the majority of cancer patients are treated (PGx Reporter 4/25/2012).