French Investigators Prospectively Test Genomically Driven Treatment in Metastatic Breast Cancer 

Get Permission

Whole-genome DNA analysis prospectively identified alterations in metastatic tumors that could be individually targeted with molecular agents, in a study presented at the 2012 European Society for Medical Oncology (ESMO) Congress by Fabrice André, MD, of Institut Gustave-Roussy, Villejuif, France.

First Study of Its Kind

The SAFIR-01 trial is the first large-scale—in fact, countrywide—study to prospectively test the entire genome from a biopsy of a metastatic lesion, then match the alteration with a targeted agent. Currently, most genetic testing looks for specific genetic alterations in a limited number of genes, and thus, many treatment opportunities may be lost, said Dr. André.

The study used whole-genome (comparative genomic hybridization) arrays to look at genetic copy numbers (gains and losses), and identified “hot spot” mutations, specifically PIK3CA and Akt mutations, which are common in breast cancer.

The investigators were able to perform whole-genome analysis on 251 of 393 patients, of whom 172 (69%) were found to have genetic alterations. While the majority of these involved “druggable” genes that can be targeted by an anticancer drug, about 20% were rare and sometimes unexpected alterations.

The common occurrence of these low-frequency alterations “highlights the need for whole-genome approaches,” Dr. André said.

In the study, patients with metastatic disease undergo biopsy, but their disease is stable at the time of genetic testing. Therefore, they do not receive mutation-guided therapy until they develop progressive disease. To date, only 26 patients have shown disease progression and have been matched with treatment that has included 13 different targeted regimens (single agents or combinations)—often PI3K and Akt inhibitors. Evidence of activity has been noted in eight patients (response or stable disease ≥ 4 months).

At a press briefing, co-investigator Monica Arnedos, MD, also of Institut Gustave-Roussy, explained, “Most patients still have no need for treatment, as their disease is controlled, but we expect within 3 years to have treated 80 or more patients.”

Listener Concerns

Upon the conclusion of Dr. Andre’s presentation, questions from the audience indicated that many oncologists are not yet comfortable with exploring genetic alterations in metastatic lesions.

Dr. André noted that 18 patients in the study “could not go to biopsy,” and among “several” adverse events were liver hematomas and pneumothorax. “I prefer to have the final analysis before reporting on these,” he said. “It is becoming common to recommend biopsy at the time of relapse, but I do think we should be comfortable with the adverse event rate. I would like to see a large epidemiologic study of adverse events associated with these biopsies.”

Peter Ellis, MBBS, PhD, of McMaster University in Canada suggested there may be another risk associated with genomic-driven treatment. “It is likely that the genetic and molecular abnormalities are very complex, and the assumption that we can implement a targeted approach because we can identify these abnormalities may put patients at risk of passing up standard treatments,” he said. “In some ways, it feels like we are on a rollercoaster that is a little out of control.”

Neoadjuvant Setting

In a separate study led by Dr. André, the investigators reported outcomes using whole-genome expression arrays in the neoadjuvant setting, the aim being to match the neoadjuvant chemotherapy to the genetic profile of the tumor.2

Results of the phase III trial, REMAGUS04, showed no increase in pathologic complete responses after genetically driven treatment, compared with standard treatment. However, patients with a positive probe were more likely to have disappearance of tumor. Moreover, the study proved that whole-genome expression array testing can be accomplished in the clinical setting; 61% of patients were successfully tested and started on targeted treatment, reported Jean-Yves Pierga, MD, of the Institut Curie, Paris.

“This is the first prospective trial showing that whole-genome expression array is feasible in the context of daily practice within 15 days,” he said. “Gene-expression arrays could be a solution in the future, an all-in-one assay for personalized medicine.”

Could Affect Drug Approval

ESMO press briefing moderator Christoph Zielinski, MD, of the Medical University of Vienna, Austria, commented that the findings from SAFIR-01 trial suggest the field of oncology will “probably move sooner rather than later from an anatomically oriented drug approval process to one oriented toward signaling pathways.”

He noted, “Surprising compounds have emerged that we would never expect to work in certain situations, but due to unsuspected genetic alterations, they do. I think this may change the [drug] registration process.”

Meanwhile, he believes whole-genome testing should be restricted to the research setting, and not applied in the clinic. “We still have to prove this is the optimum approach, though … this is an enormously important step forward.” ■

Disclosure: Drs. André, Arnedos, Ellis, Pierga, and Zielinski reported no potential conflicts of interest.


1. André F, Bachelot T, Campone M, et al: Array CGH and DNA sequencing to personalize therapy for metastatic breast cancer: A prospective national trial (UNICANCER SAFIR-01). 2012 ESMO Congress. Abstract LBA13. Presented October 1, 2012.

2. Pierga J-Y, Asselain B, Alsafadi S, et al: A prospective randomized trial evaluating gene expression arrays to select neoadjuvant chemotherapy regimen for operable breast cancer: First report of the REMAGUS04 trial. 2012 ESMO Congress. Abstract 2450. Presented September 30, 2012.