The availability of a test that would be able to improve the risk versus benefit assessment of long-term adjuvant endocrine therapy for late recurrence in patients with estrogen receptor (ER)-positive breast cancer would be a valuable tool for enhancing the treatment decision-making of oncologists. A new study compared the ability of 3 modalities—the new Breast Cancer Index (BCI) molecular test, the 21-gene recurrence score known as Oncotype DX, and the immunohistochemical (IHC4) prognostic model—to predict early and late disease recurrence in patients with ER-positive, node-negative breast cancer (Sgroi DC, et al. Lancet Oncol. 2013;14:1067-1076).
In this prospective, comparison study, tumor samples from 665 postmenopausal women with ER-positive breast cancer were evaluated; the women were enrolled in the ATAC (Arimidex, Tamoxifen, Alone or in Combination) clinical trial. The samples had been previously tested by the 21-gene recurrence score (Oncotype DX) and the IHC4 prognostic model.
The researchers conducted a BCI analysis in matched samples using 2 BCI models—cubic (BCI-C) and linear (BCI-L)—and previously validated cutoffs. The researchers identified the BCI groups with prespecified cutoff points for each model.
The primary end point was the ability of the BCI test to predict distant recurrence over 10 years compared with that of the 21-gene recurrence score or the IHC4 model. The researchers also tested the ability of the 3 assays to predict early (0-5 years) and late (5-10 years) distant recurrence.
The primary analysis showed significant differences in the risk of distant recurrence over 10 years in the BCI-C risk groups (P <.001), with 6.8% (95% confidence interval [CI], 4.4-10.0) of patients in the low-risk group, 17.3% (95% CI, 12.0-24.7) in the intermediate group, and 22.2% (95% CI, 15.3-31.5) in the high-risk group of having distant recurrence.
The secondary analysis showed that BCI-L was a stronger predictor for the overall 10 years of distant recurrence compared with BCI-C (P <.001).
Overall, the 21-gene recurrence score was less able to predict distant recurrence than the BCI test (P = .002) as was the IHC4 model (P <.001). In a multivariable analysis, all 3 assays had significant prognostic ability for early distant recurrence; however, only the BCI-L test showed significant ability to predict late recurrence (P = .048) compared with the 21-gene recurrence score (P = .47) or the IHC4 model (P = .20).
“This information generated by BCI could be extremely valuable and have fundamental relevance to breast cancer oncologists and patients, especially those who are at 5 years postdiagnosis,” said lead investigator of the study, Dennis C. Sgroi, MD, Professor, Molecular Pathology Unit and Center for Cancer Research, Massachusetts General Hospital, Boston. “This is highly relevant to clinical management, as it has been demonstrated in previous studies that late disease recurrence is a hallmark of ER-positive breast cancer, with more than half of recurrence occurring after 5 years of adjuvant therapy.”
This new study shows that all 3 modalities are able to predict early recurrence of ER-positive breast cancer, but only the BCI molecular test is a significant prognostic test in relation to both early and late distant disease recurrence.
