GLASGOW, SCOTLAND-Two thirds of women with ER-positive HER-negative breast cancer routinely assigned to receive chemotherapy on the basis of standard scoring systems have little chance of deriving any benefit from it, according to findings from a pilot study of genomic testing reported here at the European Breast Cancer Conference EBCC9 (Abstract 54).
The 67 premenopausal and postmenopausal patients participating in the trial would all normally have been recommended for chemotherapy after their breast and axillary surgery because conventional scores predicted that they could derive at least a five percent benefit in 10-year overall survival from it. But in the study, the researchers were able to intervene and avoid the use of chemotherapy by calculating the risk and response to treatment more accurately.
Use of the 21-gene recurrence score (Oncotype DX) test distinguished a minority group of 22 from among the patients who could gain up to a 30 percent benefit in overall survival from drugs which in the remaining patients might have caused more harm than good.
Nigel Bundred, MD, Professor of Surgical Oncology at University Hospital of South Manchester, said in an interview at the meeting that the genomic test outperformed conventional factors for assessing prognosis and predicting tumor behavior and was independent of other predictive criteria including Ki67 and Adjuvant! Online.
"We used to use anatomic factors-node, stage, grade-that don't relate to the biology of the tumor. But this [genomic test] is looking at the molecular biology of the genes that the tumor expresses," he said.
He pointed out that the test is licensed to predict the risk of distant recurrence and the magnitude of chemotherapy benefit: "If you are low risk-on a low recurrence score with this test-you have a higher chance of responding to hormone therapy, so you have more assurance that you have a low risk of recurrence and a better response."
The study also found that the test was accepted well by both patients and clinicians-"This is important," Bundred said. "If you're going to use a test, you have to believe in the outcome of the test and act on it."
The study's first author, Anne Armstrong, MRCP, PhD, MB, ChB, a medical oncologist at Christie Hospital, said in an interview that many patients want to avoid chemotherapy if possible and that the genomic test can help them make a decision either way: "It really helps them to feel that they should have chemotherapy if it's a high score; and obviously saves a lot of people having chemotherapy if they have a low score," she said.
Bundred noted that the test can distinguish low risk even in younger patients: "The value of the test is that it works irrespective of the age of the patient. Age by itself, as with many anatomical factors, is not a predictor of the actual behavior of the tumor," he said.
Also reported at the meeting, the International Multidisciplinary Application of Genomics in Clinical Practice (MAGIC) survey identified tumor heterogeneity as a hurdle to overcome when making treatment decisions, and one in which genomic testing should increasingly be able to help (Abstract 24).
The investigators surveyed the criteria commonly being used by clinicians as a basis for decision making about whether to recommend adjuvant chemotherapy. They found that international guidelines (mostly St. Gallen) and such tools as Adjuvant! Online were prominently used and that patient age, tumor grade, size, nodal status, and ER expression were considered important, while Ki67 and PR were considered to be least important. But it was clear in MAGIC that these factors were only weakly able to distinguish patients who would clearly benefit from chemotherapy and even less able to pick out those in whom chemotherapy could be safely withheld.
In an interview after presenting the findings, the lead investigator, Matti Aapro, MD, Dean of the Multidisciplinary Oncology Institute in Genolier, Switzerland, said that the study found that the low power of standard prognostic and predictive scoring systems had serious consequences: "We see there's a huge variation among our colleagues in this setting of uncertainty: some will give chemotherapy, others will not give chemotherapy for exactly the same patient," he said, calling for better tools for the future involving more refined genomic tests.
Also in an interview, David Cameron, MD, Professor of Oncology at Edinburgh University and Director of Cancer Service at Lothian NHS and one of the organizers of the conference, said he was optimistic about future progress with genomic testing for cancer and welcomed the change towards more refined testing.
"We're beginning to realize that not everybody needs [chemotherapy] because not everybody has a high risk of recurrence." In addition, some of the drugs commonly used don't necessarily work against an individual patient's cancer type: "We can better target treatments and treatment combinations to particular cancers rather than dishing them out to everybody," he said.