The advent of gene expression assays has provided predictive insight into chemotherapy benefit that can be combined with prognostic information yielded by gene expression profiling to better target patients with breast cancer who are at higher clinical risk for the use of adjuvant chemotherapy.
The advent of gene expression assays has provided predictive insight into chemotherapy benefit that can be combined with prognostic information yielded by gene expression profiling to better target patients with breast cancer who are at higher clinical risk for the use of adjuvant chemotherapy, according to Joseph A. Sparano, MD.
“The notion of precision medicine in breast cancer is not a new concept; it dates back to the 1970s when we began using ER and PR protein expression initially by ligand-binding assay and then by immunohistochemistry to identify patients who would benefit from adjuvant endocrine therapy,” said Sparano, a professor of medicine and women’s health at Albert Einstein College of Medicine, in a presentation during the 19th AnnualInternational Congress on the Future of Breast Cancer® West, a virtual program by Physician Education Resource® (PER®).1
“We then entered the second generation using gene amplification for HER2/neu via FISH testing to identify women who could benefit from adjuvant trastuzumab (Herceptin), an anti-HER2–based therapy,” added Sparano. “Fifteen years into the third generation [we use] gene expression profiles to guide the use of adjuvant chemotherapy and endocrine therapy. We're currently pretty far into the fourth generation of assays in terms of mutational profiling, which can identify individuals who could benefit from targeted therapies.”
In his presentation, Sparano, who is also an associate chairman in the Department of Oncology at Montefiore Medical Center, provided insight into advances made with gene expression profiles, the clinical utility of available assays, as well as future directions in this area.
Gene expression profiles first emerged in the breast cancer paradigm about 15 years ago based on unsupervised analyses, which indicated that breast cancer was a heterogenous disease, that there were distinct subtypes, and that prognosis could vary by subtype, according to Sparano. The unsupervised work resulted in the PAM50 assay (Prosigna), which can be used to identify the distinct breast cancer subtypes that exist.
The next generation of gene expression assays were based on supervised analyses, which evaluated genes associated with a better or worse prognosis. This research led to the development of various prognostic assays, which included the 21-Gene recurrence score assay (Oncotype DX) and the 70-Gene signature test (MammaPrint), among others.
“One important point is that there’s a lack of concordance in the prognostic classification provided by these assays,” said Sparano.
The first trial to show that a gene expression assay could provide independent prognostic information was the B14 trial, which included archival samples from a total of 668 patients with estrogen receptor (ER)–positive, node-negative breast cancer who had received treatment with tamoxifen for 5 years.
Fifty-one percent of patients fell into the low-risk group, which was defined as a recurrence score (RS) of less than 18, 22% of patients were in the intermediate-risk group (RS of 18-30), and 27% were in the high-risk group (RS of 31 or greater). The 10-year risk of distant recurrence was 7%, 14%, and 31%, respectively, for each of these groups.2
“Further data revealed a statistically significant association for RS that was independent of age and tumor size,” said Sparano. “As such, RS was not a surrogate marker for these other factors; it provided independent prognostic information.”
Results from another study, referred to as B20, went on to demonstrate prediction of benefit with these assays.3 A total of 651 patients with ER-positive, node-negative breast cancer were randomized to receive tamoxifen or tamoxifen plus chemotherapy. “For the entire cohort you see approximately a 4% improvement in distant relapse-free survival for patients who received chemotherapy,” noted Sparano. “However, there was a very large benefit for patients who had a RS of 31 or higher; the absolute benefit was in the range of about 25% in the group with the highest RS, suggesting that one can identify, using this assay, a subpopulation of patients who are deriving all of the benefit from chemotherapy.”
The prospective TAILORx trial included women with hormone receptor–positive, HER2-negative, and axillary node-negative breast cancer and they were randomized to treatment based on their RS.
Of the 10,273 women enrolled on the trial, 1629 who had a low RS of 0 to 10, were assigned to receive endocrine therapy alone (arm A). Women with a high RS of 26 to 100 were assigned to endocrine therapy plus chemotherapy (arm D). Those in the midrange who had a RS of 11 to 25 were randomized to receive either endocrine therapy plus chemotherapy in the standard arm (arm C), versus endocrine therapy alone in the experimental arm (arm B).
The study had a noninferiority design with invasive disease-free survival (iDFS) as the primary end point, and full information is expected after 835 iDFS events were reported.
“We modified the mid-range group for several reasons. The TAILORx population excluded HER2-positive disease and we know that the 21-gene assay includes a HER2 module that drives the RS up and is associated with a higher score, but we know that most HER2-positive tumors have a higher RS,” explained Sparano.5 “As such, if you use the assay in a HER2-negative population, you’ll have a different RS distribution.”
Additionally, the RS assay is used selectively in practice in situations where there is therapeutic equipoise, which is typically intermediate-grade tumors that are 1 cm to 2 cm; this results in more tumors having a score in the mid-range group, according to Sparano. “The trial really needed to be designed to address that group,” he said.
Moreover, the RS range was adjusted to preserve prediction in the highest-risk group and minimize the potential for undertreatment in the low-risk group, Sparano added.
Initial data from the low-risk group showed that at 5 years, the rate of freedom from recurrence of breast cancer at a distant site was 99.3% (95% CI, 98.7%-99.6%).4 “This information was subsequently integrated into the American Joint Committee on Cancer’s Cancer Staging Manual,” noted Sparano.
After a median of 7.5 years, results from the intent-to-treat population (arms B and C) were released and showed that the primary end point for iDFS was met (HR, 1.08; 95% CI, 0.94-1.24; P = .26), demonstrating noninferiority of endocrine therapy compared with the standard.6 Endocrine therapy alone was also found to be noninferior to chemoendocrine therapy with regard to freedom of recurrence of breast cancer at a distant site (HR, 1.10; 95% CI, 0.85-1.41; P = .48).
Investigators then examined whether any patients with a mid-range score were still deriving benefit from the chemotherapy. No benefit was observed with regard to increasing tumor size or grade, but statistically significant chemotherapy treatment interactions were observed between age, RS, and chemotherapy benefit. “Patients who had a higher RS and a higher clinical risk within this younger group seemed to derive benefit,” explained Sparano.
At 9 years, in those with a RS of 16-20, a 1.6% absolute benefit from chemotherapy was observed versus a 6.5% absolute benefit in those with RS ranging from 21 to 25, added Sparano.
An exploratory analysis looking at the impact of age and menopausal status on chemotherapy benefit in patients with a RS ranging from 16 to 25 showed that there was no benefit in older women with an increasing score. However, curves began to separate for younger women with a RS of less than 25, noted Sparano.
When examining absolute differences in 9-year distant recurrence rates by chemotherapy use in women 50 years or younger with RS of 16 to 25 stratified by RS and clinical risk, investigators noted that the estimated absolute benefit of chemotherapy in women with a RS of 16 to 20 who were not stratified by clinical risk was +1.6%. The estimated absolute chemotherapy benefit stratified by clinical risk was -0.2% in those with low clinical risk (n = 671) and a RS between 16 and 20 and +6.5% in those with high clinical risk (n = 215).7
When looking at the impact of age on chemotherapy benefit, investigators observed that women who were closer to menopause, aged 46 to 50 years, experienced the greatest benefit. “Interestingly, younger women really had no benefit, suggesting that some of the effect that was seen with chemotherapy in these younger patients who had higher RS might have been due to a castration effect,” explained Sparano.
At 9 years, a 3% distant recurrence with endocrine therapy alone was observed in patients with an RS of 0 to 10 (arm A). An overall 5% distant recurrence rate was reported in those with an RS between 11 and 25 (arms B and C). Between arms B and C, a less than 1% difference was observed for all end points. In those with a RS between 26 and 100 (Arm D), a 13% distant recurrence was observed, despite chemotherapy plus endocrine treatment.
The MINDACT trial included 6,693 patients who were assigned to a clinical risk or a genomic risk. Patients who had discordance in their clinical and genomic risks were randomized to receive either no chemotherapy or chemotherapy.8 The primary end point of the trial was distant metastasis-free survival at 5 years for those with high clinical risk and low genomic risk without chemotherapy.
“No effect with chemotherapy was observed in older women, but there was a 5% benefit from chemotherapy in younger women, which is very similar to what we saw in TAILORx,” said Sparano.
The phase 3 Plan B trial used the Oncotype DX Recurrence Score to define a genomically low-risk subset of patients with clinically high-risk pN0-1 early breast cancer for adjuvant treatment with endocrine therapy alone. A total of 3198 patients were enrolled on the trial and chemotherapy was omitted in 86.1% of eligible patients with a RS of 11 or less.
At a median follow-up of 5 years, DFS in the patients treated with endocrine therapy alone who had a RS of 11 of less was 94% versus 94% in those with an RS between 12 and 25 and 84% in those with an RS of greater than 25 (P < .001). In patients who received chemotherapy, the 5-year overall survival was 99% versus 97% versus 93%, respectively (P < .001).
“This provided a limited amount of level 1 evidence supporting the use of the Oncotype assay in patients with low-volume disease,” noted Sparano.
With all of the data yielded thus far, it is clear that gene expression assays provide prognostic information, that the 21-gene assay offers predictive information, and that the 70-gene assay provides prognostic information, according to Sparano.
“It’s important to remember that these assays are not interchangeable, and there’s a lack of concordance in risk classification which needs to be considered when deciding which assay to use and what to do with the information yielded,” concluded Sparano. “Future plans involve integration of the clinical and gene expression profile information to recalibrate existing tools in an effort to provide more refined information regarding prognosis as well as an estimation of chemotherapy benefit.”