SAN ANTONIO – Immunohistochemical expression of the proliferation marker Ki67 in a pretreatment biopsy of breast cancer is valid as both a positive predictive marker and a negative prognostic marker of response to neoadjuvant therapy, said Dr. Carsten Denkert, senior pathologist and head of the translational cancer research group at the Institute of Pathology, Charité University Hospital, Berlin.
The findings are based on an analysis of data from the GeparTrio trial of the German Breast Group.
Among the eight molecular subtypes of breast cancer (hormone receptor positive/negative, HER2 positive/negative, and four subtypes created by combining these features), Ki67 was a positive predictive marker in six subtypes and a negative prognostic marker in the three hormone receptor–positive subtypes, Dr. Denkert said at the San Antonio Breast Cancer Symposium.
Across cutpoints for Ki67 positivity ranging from 10% to 45%, women with Ki67-positive tumors were significantly more likely to have a pathologic complete response (pCR) to neoadjuvant chemotherapy, but they also had significantly poorer disease-free and overall survival.
The analysis was unable to identify an optimal cutpoint for positivity, but the "results provide an explanation for the very different cutpoint results in various previous clinical studies," he said. As Ki67 is a continuous marker that measures tumor proliferation, cutpoints can be defined by the scientific community.
In the GeparTrio trial, the 1,166 patients all had tumors measuring at least 2 cm, underwent a core biopsy, received two cycles of neoadjuvant chemotherapy, underwent ultrasound assessment of response, and then received additional cycles of chemotherapy tailored to that response.
The research team led by Dr. Denkert performed immunohistochemical staining for Ki67 in the pretreatment core biopsy using the MIB-1 antibody and systematically assessed 94 cutpoints for Ki67 positivity between 0% and 100% for their association with outcomes.
Results showed that 93 of the cutpoints were significantly associated with pCR, 48 were significantly associated with disease-free survival, and 58 were significantly associated with overall survival. "So it is very difficult to decide which one is the optimal cutpoint," Dr. Denkert commented.
For further analyses, patients were split into three equal groups according to Ki67 cutpoint: low (less than or equal to 15%), intermediate (15.1%-35%), and high (greater than 35%).
In a multivariate analysis, these three groups differed significantly with respect to pCR rate (P less than .0005), median disease-free survival (P = .012), and median overall survival (P = .013).
In a discussion after the presentation, Dr. Steven Vogl of the Montefiore Medical Center, New York, referenced earlier reports from the same trial. "Getting a partial response for some of the hormone receptor–positive tumors was good in terms of prognosis, even though a pCR wasn’t achieved. Did you look at Ki67 in the non-pCR patients and see if that gave you any information as to their prognosis? That is, Ki67 at the time of the final surgical procedure."
"In fact, this was one of the aims of the large Ki67 program that we did on the core biopsies and also on the surgical tissues," Dr. Denkert replied. "So far, we have not yet been able to define Ki67 as a parameter for the effect observed with the hormone receptor–positive tumors in the GeparTrio trial. Very likely, we have to combine this parameter with other parameters, and we are currently working on this."
Although the 12th St. Gallen International Breast Cancer Conference (2011) Expert Panel has proposed that Ki67 may be able to differentiate between the luminal A and luminal B molecular subtypes of breast cancer (Ann. Oncol. 2011;22:1736-47), the optimal cutpoint of the percentage of positive cells for defining a tumor to be Ki67 positive is still debated, he noted.
Dr. Denkert disclosed that he receives research/grant support from Siemens Medical Solutions and Sividon Diagnostics; that he is a consultant for Celgene, Amgen, and Sividon Diagnostics; and that he is a shareholder in Sividon Diagnostics.