SAN ANTONIO — In a large cohort of patients from Ontario with ductal carcinoma in situ (DCIS), Genomic Health's Oncotype DX DCIS score accurately assessed the 10-year risk of local and invasive local recurrence for those treated with breast conservation surgery, researchers at the San Antonio Breast Cancer Symposium reported.
Approximately 50,000 US women have DCIS, a pre-invasive form of breast cancer. DCIS comprises 30 percent of newly diagnosed breast cancers and is usually treated with breast conservation surgery then radiation. Although patients at low risk of recurrence can forgo radiation treatment, risk prognosis using traditional clinical and pathological factors isn't fully reliable.
As such, biomarker-guided strategies are necessary to improve risk assessments that lead to more personalized interventions, says molecular testing firm Genomic Health, which partially funded the study along with the Canadian Cancer Society Research Institute. According to Genomic Health, this is the largest genomic study conducted in DCIS.
Using the 13-gene expression risk score, researchers led by Eileen Rakovitch from the Sunnybrook Health Sciences Center in Toronto analyzed tumor samples from an Ontario population-based trial that enrolled more than 3,300 women with DCIS between 1994 to 2003. Researchers collected tumor samples from more than 1,500 patients, out of which approximately 700 women received breast conservation surgery without radiation while close to 850 women received breast conservation surgery with radiation.
After a median follow up time of 9.4 years, out of approximately 570 patients treated with surgery with clear margins — meaning the removed tumor was surrounded by normal tissue cells — 100 patients had recurrence. Of these patients, 44 had localized DCIS, 56 had invasive disease, and one patient had both. Meanwhile, of those treated with breast conservation surgery and radiation, 86 had recurrence.
For DCIS patients treated with just surgery, 12.7 percent, 33 percent, and 27.8 percent were deemed low risk, intermediate risk, and high risk, respectively, by Oncotype DX. Patients deemed to be at low risk by the DCIS test were at significantly lower risk of recurrence than patients deemed intermediate or high risk by the test. According to Genomic Health, the study showed that "every 50-point increase in the DCIS Score was associated with a two-fold increase in the risk of developing recurrence."
For patients with invasive local recurrence, 8 percent, 20.9 percent, and 15.5 percent were classified as low risk, intermediate risk, and high risk, respectively. Of those with DCIS local recurrence, 5.4 percent, 14.1 percent, and 13.7 percent of patients were deemed low risk, intermediate risk, and high risk by Oncotype DX DCIS.
The study authors concluded that the DCIS score is associated with local and invasive local recurrence in DCIS patients treated with breast conservation surgery alone. They recommended integrating the DCIS score with established risk factors, such as age and tumor size, to identify DCIS patients treated with breast conservation surgery who are at low risk of recurrence within a decade and patients who are still at high risk despite surgery and radiation and could require aggressive interventions.
Rakovitch and her team are still analyzing whether the DCIS score predicts recurrence risk for patients treated with surgery and radiation. Additionally, they want to develop a nomogram that incorporates the DCIS score and clinical factors to guide doctors about the best treatment options for their patients.
This is the second validation study for the Oncotype DX DCIS test. Last year researchers from Genomic Health and ECOG-ACRIN published data from the first validation trial in the Journal of the National Cancer Institute showing that the DCIS score extends beyond traditional clinical and pathological prognostic factors and gauges 10-year risk of local recurrence. In that study, which analyzed samples from the ECOG E5194 trial involving low-risk DCIS patients treated with surgery, 70 percent of patients had a low DCIS score. Compared to those with intermediate or high-risk scores, those with low DCIS scores had significantly lower risk of recurrence at 10 years.
Rakovitch noted that the second validation study was conducted in a much broader population of women with diverse DCIS characteristics. She also noted that in the study, during the time period when women were recruited, tamoxifen treatment wouldn't have been common. More recent use of tamoxifen in women with estrogen receptor-positive DCIS has shown to reduce recurrence.
"The effect of tamoxifen is beyond the scope of our particular study," Rakovitch said at SABCS. "But what our study does show is that the DCIS score can provide individualized estimates of the risk of recurrence and that can help clinicians, medical oncologists, radiation oncologists, and surgeons better understand the risk and [help] women to understand their risk, and discuss in a more informed way what the potential benefit for radiation or tamoxifen might be."
At the American Society of Clinical Oncology's annual meeting in late May/early June, Genomic Health presented data from a decision impact study involving the DCIS test, which showed that the assay changed physicians' treatment plans for one out of three patients. In 84 patients for whom doctors initially prescribed radiation before Oncotype DX testing, the same doctors later recommended no radiation for 26 patients after testing. Then, for 31 patients that doctors initially thought didn't need radiation, they changed their recommendation and prescribed radiation for 10 of these patients after viewing their DCIS scores.
Following Rakovitch's presentation at SABCS, C. Kent Osborne, director of the Dan L. Duncan Cancer Center at Baylor College of Medicine, pointed out that many clinicians don't consider DCIS to be breast cancer. "People get confused about the word DCIS because it has the word cancer in it. Patients get confused and doctors get confused," Osborne said. "Many people don't even consider it a cancer. It's a precancerous lesion and there have been several discussions over the past decade about renaming it ... because it has the word cancer in it [and] we tend to treat it more aggressively."
He noted that a test like Oncotype DX DCIS may help doctors identify patients who need less aggressive therapy or no therapy at all. "This is one of the first steps to show that we can start backing off on treatment in some of these patients," Osborne said, noting that radiation therapy is expensive. Depending on the cancer type, the type of radiation and the number of treatments, radiation therapy can cost between $10,000 to $50,000 for uninsured patients, according to one estimate. Oncotype DX DCIS is list priced at around $4,000. Osborne noted that if the test can show that it can stave off unnecessary healthcare costs, payors will likely agree to reimburse it.