Omit Surgery in Early-Stage Breast Cancer? Not Yet

Pam Harrison

December 17, 2019

SAN ANTONIO ― It's an enticing idea in an age of breast cancer treatment de-escalation: omit surgery if a patient achieves what looks like a complete response to neoadjuvant chemotherapy.

However, three new studies indicate breast-conserving surgery cannot be skipped, because even if imaging reveals an exceptional response, subsequent biopsies miss too much residual disease for the process to be considered highly reliable and safe.

In short, for initially assessing response, imaging and biopsy are just not good enough, conclude the trio of studies presented here at the San Antonio Breast Cancer Symposium (SABCS) 2019.

The Minimal Invasive Complete Response Assessment (MICRA) trial involved 167 evaluable patients who underwent MRI both prior to receiving neoadjuvant chemotherapy and afterward.

Marie-Jeanne Vrancken Peeters, MD, Netherlands Cancer Institute, Amsterdam, and colleagues reported that 135 patients achieved a radiologic complete response (rCR), as evidenced on MRI, and that another 32 patients achieved a radiologic partial response (rPR).

However, only 59% of patients who achieved an rCR on MRI experienced a pathologic complete response (pCR), the gold standard for assessing response.

After neoadjuvant chemotherapy, eight ultrasound-guided 14-gauge (14G) core biopsy samples were taken in the preneoadjuvant chemotherapy marked tumor area. It is notable that these core biopsies are considerably smaller than vacuum-assisted biopsies that are usually performed.

"Pathology results from the biopsies and from surgical specimens were then compared," the investigators note.

For slightly more than half (53%) of patients, there was evidence of a pCR in the surgical specimen. All were correctly identified by post-treatment biopsy, for a false-positive rate of 0%.

However, post-treatment biopsy missed residual disease in 37% of patients overall; for patients with an rCR on MRI, the false-negative rate, in which residual disease was again missed on biopsy, was even higher, at 45%, she added.

"We have to conclude from our results that MRI is not accurate enough for pCR detection in this patient group and that performing eight 14G biopsies in patients with an excellent response to neoadjuvant chemotherapy is not accurate enough either to detect pCR, and we are not going to omit breast surgery after primary systemic treatment since residual disease is missed in two thirds of patients," she emphasized.

"You might think that we should just have just taken larger biopsies to improve our complete response prediction," Vrancken Peeters told Medscape Medical News. "On the other hand, it is good that we now have these results. It gives us a way to move forward."

Another expert cautioned about relying on imaging and biopsy to direct surgery decisions.

Terry Mamounas, MD, Orlando Regional Medical Center, Florida, noted that just because oncologists may not see any disease on imaging does not mean there's not microscopic residual disease remaining in the tumor bed.

Further, another issue not discussed by the three study presenters at SABCS is the fact that patients who have residual disease, particularly those with triple-negative or HER2-positive breast cancer, are candidates for additional treatment to improve outcomes. "So if you do not know that there is residual disease in the tumor bed, you can't offer them that additional therapy," he noted.

"I think people understand that we still have to do surgery," Mamounas added.

"But we were hoping that with these biopsies, there wouldn't be any residual disease in 90% to 95% of patients, so we could then omit surgery and rely on radiation to treat the tumor bed," he said.

Mamounas summarized: "We tell patients that the imaging studies showed no residual disease — which is great news — but we also discuss with them the fact that we still need to do surgery just to make sure there is a real pCR."

Image-Guided Vacuum-Assisted Biopsy

The RESPONDER trial from Germany added an additional insight: taking larger biopsy specimens via image-guided vacuum-assisted biopsy (VAB) does not on its own adequately rule out residual disease in the breast, either.

Joerg Heil, MD, University Hospital Heidelberg, and colleagues reviewed data on 398 women with stage I–III breast cancer who had achieved a partial or complete response following neoadjuvant chemotherapy and who underwent VAB.

VAB results were compared with pathologic findings in the surgical specimen.

Confirmed residual disease in the surgical specimen but not in the VAB sample was considered a false-negative finding.

For the whole cohort, VAB alone failed to detect residual tumor in 37 of 208 cases in which there was residual tumor in the surgical specimen, yielding a false-negative rate of 17.8%.

On a slightly more positive note, when investigators combined results from the VAB and breast imaging using ultrasound and mammography, residual tumor was missed in only 6.2% of biopsies.

Trimodality Imaging

In an effort to answer patients' questions as to why they still had to undergo surgery when there was no tumor left after successful neoadjuvant chemotherapy, Mark Basik, MD, Segal Cancer Center, Jewish General Hospital, Montreal, Canada, and colleagues carried out the phase 2 NRG-BR005 trial to assess the accuracy of tumor bed biopsy in predicting pCR in patients with rCR or near rCR on trimodality imaging following neoadjuvant chemotherapy .

Trimodality imaging consisted of findings from mammography, ultrasound, and MRI.

Basik explained the trial's rationale. The negative predictive value of trimodality imaging is close to 80%. If their methods could show a 10% improvement on that rate, the resulting 90% negative predictive value might allow them to rule out surgery if results of subsequent biopsy also proved to be negative.

Tumor bed needle core biopsy plus clinical examination along with trimodality imaging were all performed to see whether it would be possible to identify patients who could proceed, after neoadjuvant chemotherapy, to radiotherapy without undergoing breast-conserving surgery.

Of 98 evaluable patients, 36 patients had either invasive residual disease or ductal carcinoma in situ, which was enough patients with a non-pCR response for the investigators to conduct their primary analysis.

In the overall cohort, the sensitivity of biopsy was 50%, and the negative predictive value was 77.5%, "which did not meet our criteria of 90% or more that we wanted to achieve," Basik reported.

Among different tumor subtypes, the negative predictive value of the tri-pronged approach almost met their study criteria in HER2-positive patients, at 89.5%, he added.

However, this predictive value was lower in triple-negative disease, at 74.1%, and was lower still in hormone receptor–positive/HER2-negative breast cancer, at 46.2%.

When the investigators compared pathologic findings from biopsy to those from surgery among only those patients with invasive disease, the negative predictive value again almost met study criteria, coming in at 89.2%.

Nevertheless, the investigators were forced to conclude that the addition of image-directed tumor bed biopsy to trimodality imaging did not achieve a negative predictive value of 90% in the entire cohort and that, on the basis of their study criteria, the findings do not support breast-conserving treatment without surgery.

"A lot of women are afraid of the surgery, even lumpectomy, and if you can avoid it among patients who have a good response to neoadjuvant chemotherapy, we should really try to do it," Basik told Medscape Medical News.

"So I think we must put a lot of effort into making it happen, and I think we eventually will," he suggested.

Vrancken Peeters, Heil, Basik, and Mamounas have disclosed no relevant financial relationships.

San Antonio Breast Cancer Symposium (SABCS) 2019: Abstract GS5-06, GS5-03, and GSF-05. Presented December 13, 2019.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: