Carlson, Robert H.
SAN FRANCISCO—The estrogen-receptor (ER), progesterone-receptor (PR), and HER2 status of breast tumors—positive or negative—can change between evaluation of the primary tumor at diagnosis and evaluation of residual disease after neoadjuvant chemotherapy. A retrospective study by researchers at MD Anderson Cancer Center demonstrates just how common this change in biomarkers is, and also showed that status change was associated with better relapse-free survival.
Among 398 women with known ER, PR, and HER2 status in primary and residual disease, about 41 percent (162 patients) had a change in at least one biomarker, said Napa Parinyanitikul, MD, a post-doctoral student, reporting the results here at the Breast Cancer Symposium. The changes may have implications when treatment and prognosis of breast cancer are determined by tumor subtype, she said, adding that what role neoadjuvant chemotherapy might play in the change, if any, is unknown.
The data do show, though, that a change in biomarker status in these patients was associated with significantly increased relapse-free survival.
Experts commenting during the session said such changes should rarely change management decisions.
In her presentation (Abstract 48), Parinyanitikul said that breast cancer tumor heterogeneity and lack of stability in the biomarker can account for the changes.
Positive to Negative, Negative to Positive
In the study, Cox proportional models were ﬁt to determine the association of receptor status changes with outcomes after adjustment for patient and disease characteristics. Twenty-three (11%) of 211 ER-positive tumors changed to ER-negative, and 39 (21%) of 187 ER-negative tumors changed to ER-positive.
Fifty-seven (35%) of 162 PR-positive tumors changed to PR-negative, and 28 (12%) of 235 PR-negative tumors changed to PR-positive.
And, 29 (40%) of 72 HER2-positive tumors changed to HER2-negative, while 16 (46%) of 35 trastuzumab-treated tumors changed from HER2-negative to HER2-positive.
Parinyanitikul reported that at a median follow-up of 40 months, 128 women (32%) had died, and 167 (42%) had experienced a recurrence.
The five-year overall survival estimates were 73 percent for patients with any receptor change, versus 63 percent for patients without any receptor change.
Five-year relapse-free survival estimates also favored those with status changes: 63 percent for patients with any receptor change, versus 48 percent for patients without any receptor change.
The degree of change appeared to influence survival as well. Among patients with baseline ER-positive tumors, the five-year overall survival estimate was 73 percent when the absolute ER percent decrease was less than 20 percent, and 87 percent for those whose absolute ER percent decrease was more than 20 percent. The five-year relapse-free survival estimates were 59 and 71 percent, respectively.
A change in any receptor was associated with longer relapse-free survival, with a hazard ratio of 0.63, but not with overall survival (hazard ratio of 0.79).
Neoadjuvant chemotherapy regimens were anthracycline-based, taxane-based, and anthracycline-taxane-based chemotherapy. And 88 percent of patients with HR-positive disease received adjuvant endocrine therapy.
Not Cause for Concern?
Still, although treatment decisions for breast cancer are commonly made based on biomarker subtype, members of a panel that discussed this report said they were not especially concerned by its implications.
Lajos Pusztai, MD, PhD, Professor of Medicine and Co-Director of the Cancer Genetics Research Program and Director of Breast Medical Oncology Section at Yale Cancer Center, noted that ER or PR status is known to change if testing is repeated on the same specimen without any change in the actual markers.
He said he would not conclude that the changes after neoadjuvant chemotherapy seen in this study were due to the chemotherapy: “You can't conclude that it was the chemotherapy that changed the receptor status, because part of this discordance is from the technical imprecision. The clinical significance [of status change] is overblown, and although it is an interesting research question, I wouldn't recommend that we stress ourselves out because of the heterogeneity that might exist.”
Pusztai said it would be dangerous to withhold endocrine therapy in the adjuvant setting for someone whose tumor turns negative on a second assay, because one of the assays could be wrong, “and there is no way to know which is wrong.”
To be really precise, he said, a third “tie-breaker” assay would have to be done using the same method: “I would be very careful making a decision if there are conflicting results,” Pusztai said, adding that if retesting is discordant and one of the tests gives an ER-positive or HER2-positive result, then the person should receive therapy appropriate for ER-positive or HER2-positive disease.
Comments from Kathy Albain
Session moderator Kathy Albain, MD, Professor of Medicine and Director of the Breast Clinical Research Program at Loyola University Stritch School of Medicine, said that there are cases when an assay on residual tumor after neoadjuvant chemotherapy is warranted: “We've all seen the patient with a lot of bulky residual nodal disease even after the primary tumor does well with neoadjuvant therapy, and [in such cases] I'm often tempted to reassess the biomarker status on those nodes in case I have missed a clone that is ER positive or HER2 positive.”
In an e-mail exchange after the meeting, Parinyanitikul said she and her coauthors agreed with Pusztai's statement about receptor change due to technical imprecision of immunohistochemistry in a proportion of patients. However, she said, they also believe there is a component of tumor evolution and clonal selection due to heterogeneity after exposure to therapy: “This work was not completed to suggest therapy changes and run the risk of withholding effective therapy. The goal was actually the opposite, in that it may be that some of these tumor receptor changes may direct physicians to add endocrine therapy.
“We also agree with the idea of a tie-breaker assay, and therefore have suggested using samples from clinical trials without any selection bias to apply not only standard markers, but a control marker that confirm the discordance findings.”
The researchers did not want to imply that endocrine therapy should be withheld, Parinyanitikul continued. “On the contrary, maybe there are some patients who initially were HR-negative and if in the residual specimen we identify that a tumor is HR-positive, as a clinician I will discuss the addition of endocrine therapy to that patient.”
The Symposium is co-sponsored by the American Society of Breast Disease, American Society of Breast Surgeons, American Society of Clinical Oncology, American Society for Radiation Oncology, National Consortium of Breast Centers, and the Society of Surgical Oncology.