New tool predicts late recurrence in breast cancer

SAN ANTONIO – A novel multigene signature known as the breast cancer index markedly outperformed the widely used Oncotype DX Recurrence Score and IHC4 tools in predicting late recurrences of estrogen receptor-positive/lymph node-negative breast cancer, Dr. Dennis C. Sgroi reported at the annual San Antonio Breast Cancer Symposium.

In a large clinical study, all three tools demonstrated significant prognostic value for early distant recurrences – that is, breast cancers recurring within 5 years of diagnosis. But only the breast cancer index (BCI) had sustained power for predicting distant recurrences arising during years 5-10. Both Oncotype DX and the IHC4 – a test based upon a tumor’s estrogen receptor, progesterone receptor, HER2, and Ki-67 status – lost their prognostic ability at about the 5-year mark, observed Dr. Sgroi of Massachusetts General Hospital, Boston.

More than half of all recurrences of estrogen receptor–positive early-stage breast cancer take place after 5 years of adjuvant tamoxifen or aromatase inhibitor therapy. Thus, the BCI fills a major need: a biomarker capable of identifying those estrogen receptor–positive breast cancer patients at increased risk for recurrence 5-10 years after their diagnosis.

The BCI includes the ratio of HOXB13-to-IL17BR gene expression and a five-gene molecular grade index shown to be predictive beyond tumor grade of distance metastasis. Dr. Sgroi and coinvestigators have previously established that these two biomarkers are complementary in their power to predict recurrences (Clin. Cancer Res. 2008;14:2601-8).

At the San Antonio symposium, Dr. Sgroi presented an analysis of the comparative prognostic performance of the BCI, the Oncotype DX recurrence score, and the IHC4 as applied to baseline tumor samples from 665 hormone receptor-positive participants in the randomized, multicenter TransATAC (Arimedex, Tamoxifen, Alone or Together) trial. The primary endpoint was distant recurrence.

The baseline BCI score differentiated three risk groups over the course of 10 years of follow-up in TransATAC: 58% of women fell into the low-risk group with a 4.2% risk of distant recurrence within 10 years, 25% had an intermediate score with an 18% risk, and 17% had a high score with a 30% risk.

In the first 5 years of follow-up, patients with a low- or intermediate-risk BCI had a distant recurrence risk of 1.3% and 5.6%, respectively. Thus, the BCI identified 83% of study participants as having, on average, a 5-year distant recurrence risk of less than 5%. In contrast, women with a high BCI score had a 5-year recurrence risk of 18%.

Among patients who remained disease free at 5 years, the 61% with a low baseline BCI score had a 3.5% rate of distant recurrence during years 5-10. Those with an intermediate or high BCI had a 13.4% rate.

In a multivariate analysis adjusted for the clinical treatment score – an algorithm based upon nodal status, tumor size and grade, age, and treatment (J. Clin. Oncol. 2011;29:4273-8) – the BCI, IHC4, and Oncotype DX displayed similar prognostic performance for distant recurrences in years 0-5. But the latter two tests lost their predictive capability in years 5-10.

Estrogen receptor–positive/node-negative patients have traditionally been considered at low risk. But, as shown in the TransATAC analysis, applying the BCI to primary tumor specimens from such patients enables physicians to identify two groups at the time of diagnosis: those who are at low risk for recurrence within the next 5 years and who are adequately treated with endocrine therapy alone; and a high-risk subgroup who should be considered for chemotherapy or some other additional therapy along with endocrine therapy, Dr. Sgroi concluded.

Moreover, the BCI score at diagnosis also permits identification of two distinct groups among patients remaining disease free at 5 years of follow up: those at low risk of late recurrence and who don’t need subsequent therapy; and those at roughly a threefold greater risk of late recurrence and are therefore candidates for further systemic adjuvant therapy.

"What that additional treatment should be is something we don’t know at this point. It might be extended adjuvant hormonal therapy alone or in combination with another therapeutic agent," he added.

In response to an audience question, Dr. Sgroi admitted that the exact biologic role of the genes incorporated in the BCI is "still a bit of a mystery," although it’s clear that they’re not solely involved in proliferation.

The TransATAC study was funded by AstraZeneca. Dr. Sgroi’s BCI work is supported by the National Institutes of Health, the U.S. Department of Defense, the Avon Foundation, and the Susan G. Komen for the Cure.

SAN ANTONIO – A novel multigene signature known as the breast cancer index markedly outperformed the widely used Oncotype DX Recurrence Score and IHC4 tools in predicting late recurrences of estrogen receptor-positive/lymph node-negative breast cancer, Dr. Dennis C. Sgroi reported at the annual San Antonio Breast Cancer Symposium.

In a large clinical study, all three tools demonstrated significant prognostic value for early distant recurrences – that is, breast cancers recurring within 5 years of diagnosis. But only the breast cancer index (BCI) had sustained power for predicting distant recurrences arising during years 5-10. Both Oncotype DX and the IHC4 – a test based upon a tumor’s estrogen receptor, progesterone receptor, HER2, and Ki-67 status – lost their prognostic ability at about the 5-year mark, observed Dr. Sgroi of Massachusetts General Hospital, Boston.

More than half of all recurrences of estrogen receptor–positive early-stage breast cancer take place after 5 years of adjuvant tamoxifen or aromatase inhibitor therapy. Thus, the BCI fills a major need: a biomarker capable of identifying those estrogen receptor–positive breast cancer patients at increased risk for recurrence 5-10 years after their diagnosis.

The BCI includes the ratio of HOXB13-to-IL17BR gene expression and a five-gene molecular grade index shown to be predictive beyond tumor grade of distance metastasis. Dr. Sgroi and coinvestigators have previously established that these two biomarkers are complementary in their power to predict recurrences (Clin. Cancer Res. 2008;14:2601-8).

At the San Antonio symposium, Dr. Sgroi presented an analysis of the comparative prognostic performance of the BCI, the Oncotype DX recurrence score, and the IHC4 as applied to baseline tumor samples from 665 hormone receptor-positive participants in the randomized, multicenter TransATAC (Arimedex, Tamoxifen, Alone or Together) trial. The primary endpoint was distant recurrence.

The baseline BCI score differentiated three risk groups over the course of 10 years of follow-up in TransATAC: 58% of women fell into the low-risk group with a 4.2% risk of distant recurrence within 10 years, 25% had an intermediate score with an 18% risk, and 17% had a high score with a 30% risk.

In the first 5 years of follow-up, patients with a low- or intermediate-risk BCI had a distant recurrence risk of 1.3% and 5.6%, respectively. Thus, the BCI identified 83% of study participants as having, on average, a 5-year distant recurrence risk of less than 5%. In contrast, women with a high BCI score had a 5-year recurrence risk of 18%.

Among patients who remained disease free at 5 years, the 61% with a low baseline BCI score had a 3.5% rate of distant recurrence during years 5-10. Those with an intermediate or high BCI had a 13.4% rate.

In a multivariate analysis adjusted for the clinical treatment score – an algorithm based upon nodal status, tumor size and grade, age, and treatment (J. Clin. Oncol. 2011;29:4273-8) – the BCI, IHC4, and Oncotype DX displayed similar prognostic performance for distant recurrences in years 0-5. But the latter two tests lost their predictive capability in years 5-10.

Estrogen receptor–positive/node-negative patients have traditionally been considered at low risk. But, as shown in the TransATAC analysis, applying the BCI to primary tumor specimens from such patients enables physicians to identify two groups at the time of diagnosis: those who are at low risk for recurrence within the next 5 years and who are adequately treated with endocrine therapy alone; and a high-risk subgroup who should be considered for chemotherapy or some other additional therapy along with endocrine therapy, Dr. Sgroi concluded.

Moreover, the BCI score at diagnosis also permits identification of two distinct groups among patients remaining disease free at 5 years of follow up: those at low risk of late recurrence and who don’t need subsequent therapy; and those at roughly a threefold greater risk of late recurrence and are therefore candidates for further systemic adjuvant therapy.

"What that additional treatment should be is something we don’t know at this point. It might be extended adjuvant hormonal therapy alone or in combination with another therapeutic agent," he added.

In response to an audience question, Dr. Sgroi admitted that the exact biologic role of the genes incorporated in the BCI is "still a bit of a mystery," although it’s clear that they’re not solely involved in proliferation.

The TransATAC study was funded by AstraZeneca. Dr. Sgroi’s BCI work is supported by the National Institutes of Health, the U.S. Department of Defense, the Avon Foundation, and the Susan G. Komen for the Cure.

SAN ANTONIO – A novel multigene signature known as the breast cancer index markedly outperformed the widely used Oncotype DX Recurrence Score and IHC4 tools in predicting late recurrences of estrogen receptor-positive/lymph node-negative breast cancer, Dr. Dennis C. Sgroi reported at the annual San Antonio Breast Cancer Symposium.

In a large clinical study, all three tools demonstrated significant prognostic value for early distant recurrences – that is, breast cancers recurring within 5 years of diagnosis. But only the breast cancer index (BCI) had sustained power for predicting distant recurrences arising during years 5-10. Both Oncotype DX and the IHC4 – a test based upon a tumor’s estrogen receptor, progesterone receptor, HER2, and Ki-67 status – lost their prognostic ability at about the 5-year mark, observed Dr. Sgroi of Massachusetts General Hospital, Boston.

More than half of all recurrences of estrogen receptor–positive early-stage breast cancer take place after 5 years of adjuvant tamoxifen or aromatase inhibitor therapy. Thus, the BCI fills a major need: a biomarker capable of identifying those estrogen receptor–positive breast cancer patients at increased risk for recurrence 5-10 years after their diagnosis.

The BCI includes the ratio of HOXB13-to-IL17BR gene expression and a five-gene molecular grade index shown to be predictive beyond tumor grade of distance metastasis. Dr. Sgroi and coinvestigators have previously established that these two biomarkers are complementary in their power to predict recurrences (Clin. Cancer Res. 2008;14:2601-8).

At the San Antonio symposium, Dr. Sgroi presented an analysis of the comparative prognostic performance of the BCI, the Oncotype DX recurrence score, and the IHC4 as applied to baseline tumor samples from 665 hormone receptor-positive participants in the randomized, multicenter TransATAC (Arimedex, Tamoxifen, Alone or Together) trial. The primary endpoint was distant recurrence.

The baseline BCI score differentiated three risk groups over the course of 10 years of follow-up in TransATAC: 58% of women fell into the low-risk group with a 4.2% risk of distant recurrence within 10 years, 25% had an intermediate score with an 18% risk, and 17% had a high score with a 30% risk.

In the first 5 years of follow-up, patients with a low- or intermediate-risk BCI had a distant recurrence risk of 1.3% and 5.6%, respectively. Thus, the BCI identified 83% of study participants as having, on average, a 5-year distant recurrence risk of less than 5%. In contrast, women with a high BCI score had a 5-year recurrence risk of 18%.

Among patients who remained disease free at 5 years, the 61% with a low baseline BCI score had a 3.5% rate of distant recurrence during years 5-10. Those with an intermediate or high BCI had a 13.4% rate.

In a multivariate analysis adjusted for the clinical treatment score – an algorithm based upon nodal status, tumor size and grade, age, and treatment (J. Clin. Oncol. 2011;29:4273-8) – the BCI, IHC4, and Oncotype DX displayed similar prognostic performance for distant recurrences in years 0-5. But the latter two tests lost their predictive capability in years 5-10.

Estrogen receptor–positive/node-negative patients have traditionally been considered at low risk. But, as shown in the TransATAC analysis, applying the BCI to primary tumor specimens from such patients enables physicians to identify two groups at the time of diagnosis: those who are at low risk for recurrence within the next 5 years and who are adequately treated with endocrine therapy alone; and a high-risk subgroup who should be considered for chemotherapy or some other additional therapy along with endocrine therapy, Dr. Sgroi concluded.

Moreover, the BCI score at diagnosis also permits identification of two distinct groups among patients remaining disease free at 5 years of follow up: those at low risk of late recurrence and who don’t need subsequent therapy; and those at roughly a threefold greater risk of late recurrence and are therefore candidates for further systemic adjuvant therapy.

"What that additional treatment should be is something we don’t know at this point. It might be extended adjuvant hormonal therapy alone or in combination with another therapeutic agent," he added.

In response to an audience question, Dr. Sgroi admitted that the exact biologic role of the genes incorporated in the BCI is "still a bit of a mystery," although it’s clear that they’re not solely involved in proliferation.

The TransATAC study was funded by AstraZeneca. Dr. Sgroi’s BCI work is supported by the National Institutes of Health, the U.S. Department of Defense, the Avon Foundation, and the Susan G. Komen for the Cure.