1656P - Breast cancer (BC) and the study of intratumoral heterogeneity through PI3K pathway-biomarkers

Date 30 September 2012
Event ESMO Congress 2012
Session Poster presentation II
Topics Breast Cancer
Translational Research
Presenter Leticia De Mattos-Arruda
Authors L. De Mattos-Arruda1, J. Cortes Castan1, C.M. Aura2, J. Gregori3, G. Caratú4, C. Saura5, M. Oliveira5, J. Tabernero5, J. Seoane6, A. Vivancos4
  • 1Medical Oncology Department, Vall d'Hebron Institute of Oncology, Vall d'Hebron University Hospital, 08035 - barcelona/ES
  • 2Molecular Pathology, Vall d'Hebron University Hospital, 08035 - Barcelona/ES
  • 3Statistics Department, Faculty of Biology, Barcelona University, 08007 - Barcelona/ES
  • 4Genomics Laboratory, Vall d'Hebron Institute of Oncology, Vall d'Hebron University Hospital, 08035 - Barcelona/ES
  • 5Medical Oncology Department, Vall d'Hebron Institute of Oncology, Vall d'Hebron University Hospital, 08035 - Barcelona/ES
  • 6Gene Expression And Cancer Laboratory, Vall d'Hebron Institute of Oncology, Vall d'Hebron University Hospital, 08035 - Barcelona/ES

Abstract

Introduction

Alterations of the PI3K pathway are BC drivers. Intratumoral heterogeneity seems to be key to deal with therapeutic resistance. In this study, we analyzed the presence, intratumor distribution and prognostic implications of PI3K-pathway biomarkers (PTEN expression and PIK3CA point mutations).

Methods

99 BC formalin-fixed, paraffin-embedded (FFPE) tumor-tissues and 20 matched plasma samples were assessed. PTEN expression was determined by immunohistochemistry (IHC); the H-score was calculated by multiplying the percentage fraction of positively stained tumor area (TA) by staining intensity (0, 1 + , 2 + , or 3+). Sorting the H-scores, with h1 as the highest and h3 as the lowest, and normalizing them to add to 100, i.e. h`1 = h1/(h1 + h2 + h3)*100, and taking the maximum value: Hm = max i (h'i) = h'1. The sample was classified as homogeneous (Ho) when either Hm ≥ 90 or Hm = 0, otherwise as heterogeneous (He). PTEN low: H-score< 50. PIK3CA and its % mutant alleles (mA) from FFPE or cell-free DNA (cfDNA) in plasma were determined by OncoCarta Panel v1.0 (Sequenom). TA and stromal area (SA) were scored by H&E; real tumor area: RTA = TA-SA. Metastatic BC patients (pts) were classified: ER + /HER2- (LUM); HER2+ (HER2); triple negative (TN).

Results

42% (41/97) of FFPE had PTEN Ho, 29.9% (29/97) PTEN low and 36% (33/96) PIK3CA mutations. For PTEN, there was a trend for less homogeneity among LUM vs. non-LUM tumors (41 vs. 59%, p = 0.067). PIK3CA was more homogeneously distributed within LUM than non-LUM as there was a significant linear correlation between mA and RTA (r = 0.77, p = 0.00008). Prognosis was better for LUM pts (N = 98, p = 0.0002), but did not differ between Ho vs. He (N = 95, p = 0.123) or PIK3CA mutant vs. non mutant pts (N = 89, p = 0.23). CfDNA among PIK3CA mutant pts had a concordance of 58% (7/12, 1st sample for advanced BC pts) and 12.5% (1/8, 2nd matched sample in 1st follow-up), reflecting changes with treatment as compared with the primary/metastatic FFPE. Updated analysis will be presented.

Conclusions

The intratumoral heterogeneity of PTEN and PIK3CA mutations were detected. In the case of LUM tumors, while PIK3CA mutations seemed to be more homogeneously distributed, reassuring its driver role, PTEN tended to be more heterogeneous. Profiling cfDNA could be essential to mirror the tumor from which they derive.

Disclosure

All authors have declared no conflicts of interest.