Abstract S4-5: Ki67 levels in pretherapeutic core biopsies as predictive and prognostic parameters in the neoadjuvant GeparTrio trial

Abstract

Background: Ki67 has been suggested as a marker for definition of luminal A and luminal B tumors by the 2011 St. Gallen consensus panel. However, the cutoffs for Ki67 are still under debate. In particular, it is not clear if one single cutoff is useful for prognostic and predictive information in the different molecular subtypes. It is an advantage of the neoadjuvant approach that predictive and prognostic outcome measurements can be separated in the same cohort. In this study, we evaluated a large cohort of core biopsies from the neoadjuvant GeparTrio trial to investigate the impact of pretherapeutic Ki67 levels as a predictive marker for response to neoadjuvant chemotherapy as well as a prognostic marker for progression-free and overall survival. The analysis was stratified for hormone-receptor positive and negative tumors as well as HER2 status. Methods: A total of 1166 pretherapeutic core biopsies from the neoadjuvant Gepartrio trial were evaluated for Ki67 by immunohistochemistry, a total of 200 cells were counted in each sample. Ki67 cutoffs were evaluated using web-based software Cutoff Finder (http://molpath.charite.de/cutoff/). The details of the GeparTrio study design have been described before (von Minckwitz, JNCI 2008). We compared pCR rate as well as the overall and disease free survival in the complete cohort as well as subgroups of patients based on hormone receptor and HER2 expression. Results: Using Ki67 as a continuous parameter, a wide range of cutoffs between 10% and 80% for Ki67 were predictive for pCR. For DFS and OS, a wide range of cutoffs between 10% and 45% was significant. For further analysis, the three groups of Ki67 0–15% vs. Ki67 15.1%–35% vs. Ki67 >35 were defined and were compared for different outcome parameters. The pCR rates in these three groups of Ki67 expression were 4.2%, 12.9% and 29.0% (p < 0.0005). For DFS and OS, the groups were significantly linked to prognosis in univariate and multivariate analysis. A detailed subgroup analysis was performed showing that Ki67 was significantly predictive for pCR in all molecular subgroups. However, in subgroup survival analysis, Ki67 was prognostic in luminal, but not in triple-negative tumors. Conclusion: Ki67 is a valid predictive and prognostic marker in breast cancer. This marker is significant over a wide range of different cutoffs, which explains the different results of Ki67 cutoffs in different previous studies. Therefore, the variability observed in different studies evaluating Ki67 might reflect A) the wide range of valid cutoffs B) the different clinical endpoints of the studies and C) the different contribution of the molecular subtypes in the study cohorts. Based on our results we suggest three subgroups for Ki67 (0–15% vs. 15.1–35 vs. >35%) as a reasonable approach for further standardization of this marker. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr S4-5.

Breast Cancer 2013 - Interpretation of New and Known Data

“…It has already been shown in clinical cohorts that the Ki-67 determination in punch biopsies can predict a therapy response, especially among patients with triple-negative and hormone receptor-positive breast cancer [27]. Similar analyses are now being performed in the prospective GeparTrio study of tumour material from 1101 patients [65]. In the analysis both the prognostic and the predictive relevance with regard to a pathological complete remission (pCR) were confirmed.…”

“…Bereits an Klinikkohorten konnte gezeigt werden, dass die Ki-67-Bestimmung in der Stanze ein Therapieansprechen insbesondere bei Patientinnen mit einem triple-negativen und hormonrezeptorpositiven Mammakarzinom vorhersagen kann [27]. Ähnliche Analysen wurden nun in der prospektiven GeparTrio-Studie an Tumormaterial von 1101 Patientinnen durchgeführt [65]. In der Analyse konnte die prognostische und prädiktive Relevanz in Bezug auf eine pathologische Komplettremission (pCR) bestätigt werden.…”

“…A distinction on the basis of grading is considered to be not sufficiently substantiated for clinical practice. The report of Cheang et al proposes a biological cut-off value of 14% [63], but this does not mean that this is also the best predictor for, e.g., a response to chemotherapy [27,65], which in retrospective clinical analyses lies at about 35 %. It is unclear whether the majority of committee members decided for a cutoff value of 20 % for just this reason.…”

Predictive Factors for Response to Neoadjuvant Therapy in Breast Cancer

“…The proliferation marker , Ki-67 is a valuable tool that provides predictive information regarding NAT response and prognosis across molecular subgroups [22]. Being a continuous variable, numerous cut-off points have been defined for Ki67 to try and discern those with a better from those with a worse prognosis [23,24]. An issue that remains under debate is the different performance of the marker regarding its predictive and prognostic value in different molecular subgroups.…”

Prognostische und prädiktive Faktoren beim Mammakarzinom – Welche Biomarker sind wichtig?

“…Dabei ist zu beachten, dass die Wertigkeit prognostischer/prädiktiver Faktoren davon abhängt, in welcher Mammakarzinom-Subgruppe sie bestimmt werden. Beispielsweise hat der Biomarker KI-67 bei tripelnegativen Mamakarzinomen eine starke prädiktive, aber keine prognostische Bedeutung, während derselbe Parameter bei hormonrezeptor-positiven, HER2-negativen Mammakarzinomen sowohl prognostische als auch prädiktive Wertigkeit besitzt [33]. Im Gegensatz zu den oben genannten prognostischen Faktoren ist die Datenlage hinsichtlich von Faktoren, die für ein spezifisches Therapieansprechen prädiktiv sein könnten, limitiert.…”