NEW YORK (GenomeWeb) – Two letters to the editor of the New England Journal of Medicine have pointed out possible flaws or omissions in a recent paper that suggested that a gene test could guide breast cancer treatment decisions.
In August, researchers from the Microarray in Node-negative and 1 to 3 positive lymph node Disease may Avoid ChemoTherapy (MINDACT) study reported in NEJM that women with early-stage breast cancer who have a high clinical risk of disease recurrence might be able to forgo adjuvant chemotherapy if they have a low genomic risk of recurrence, as gauged using Agendia's MammaPrint assay.
"Given these findings, approximately 46 percent of women with breast cancer who are at high clinical risk might not require chemotherapy," the MINDACT researchers concluded in their paper.
In a letter appearing this week in NEJM, though, a trio of researchers from Leiden University Medical Center pointed out some potential hitches with the study. For instance, they noted that the two groups at high clinical risk of disease recurrence — one with low genomic risk and one with high genomic risk of recurrence — might have varied too much in their disease characteristics.
For instance, they noted that 90 percent of the tumors from patients at high clinical risk, but low genomic risk of recurrence were luminal and HER2 negative, and that 71 percent were grade 1 or grade 2 tumors. However, only half the tumors from patients at both high clinical and genomic risk were luminal and HER2-negative, and three-quarters were grade 3 tumors. From this, the Leiden trio calculated that 82 percent of luminal grade 1 or grade 2 tumors from patients at high clinical risk would be classified as being at low genomic risk.
The trio, which included Erik Blok, Cornelis van deVelde, and Vincent Smit, also noted that the study found, but didn't highlight, that for about 60 percent of patients — particularly ones at low clinical risk or high clinical risk with triple-negative cancers — the MammaPrint assay wasn't useful.
In addition, they suggested that a different decision-making algorithm might be better suited to guiding treatment decisions. A chi-square-based decision tree analysis that incorporates tumor characteristics such as ER, HER2, and Ki-67 status might predict MammaPrint outcome, they said.
"On the basis of the outcome of such a study, the use of MammaPrint could be restricted to patients for whom the clinicopathological risk assessment is insufficient," the Leiden trio wrote.
A Johns Hopkins Kimmel Cancer Center team recently developed a model that uses such histopathological factors to predict whether Genomic Health's Oncotype DX would place a patient at high or low risk of tumor recurrence.
The MINDACT researchers, though, countered that a decision-tree model using tumor characteristics might not be effective. They noted that in their study those measurements were not associated with the primary endpoint and, further, that variability in how those tumor characteristics are measured have been reported.
"We think that a model can be predictive only if the standardization of the characteristics guarantees minimal variability and high reproducibility," they said.
In addition, a duo from Radboud University Medical Center and the Prince of Wales Hospital in Australia noted in a separate letter that the MINDACT study didn't take patient preferences or quality-of-life evaluations into consideration. Previous studies, they said, have found that patients would pursue chemotherapy treatment even for a small increase in survival.
"[I]n the era of precision medicine, there is an urgent need to include quality-of-life evaluations and psychological outcomes as secondary end points in clinical trials exploring the utility of genomic profiling in women with early-stage breast cancer," the pair said.
The MINDACT researchers responded that they, too, consider quality of life to be an important measure, but didn’t include such an analysis in their study as it would have entailed some participants to sign four different informed consent documents.
They also said that they "agree that the final decision about chemotherapy always remains with the patient with breast cancer after her physician has informed her about potential benefits and risks." They added that that they view their study as providing information to help patients make that decision.