American Association for Cancer Research
10780432ccr150965-sup-148549_1_supp_3202730_nwvrhv.pptx (355.39 kB)

Supp Figure 1: Cutoff Finder analysis. from Integrated Analysis of PTEN and p4EBP1 Protein Expression as Predictors for pCR in HER2-Positive Breast Cancer

Download (355.39 kB)
posted on 2023-03-31, 18:11 authored by Sibylle Loibl, Silvia Darb-Esfahani, Jens Huober, Alexander Klimowicz, Jenny Furlanetto, Bianca Lederer, Arndt Hartmann, Holger Eidtmann, Berit Pfitzner, Peter A. Fasching, Katharina Tiemann, Christian Jackisch, Keyur Mehta, Gunter von Minckwitz, Michael Untch, Carsten Denkert

(A) OR were plotted against all possible cutoff points: the range of significant cutoffs was 39.4% with an optimal cutoff at 60.1 immunofluorescence expression values (vertical line). Dotted lines: 95% confidence interval (B) Waterfall plot showing correct and wrong classification of cases in accordance to pCR depending on PTEN immunofluorescence expression.






Background: The PI3K/AKT pathway and phosphatase and tensin homolog (PTEN) aberrations are common in breast cancer. We investigated the correlation between phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), PTEN, p4EBP1 (phosphorylated E4 binding protein 1), and pathologic complete response (pCR) in patients receiving neoadjuvant therapy.Experimental Design: We retrospectively evaluated PIK3CA, PTEN, and p4EBP1 protein expression in centrally HER2-positive patients (n = 181) who received epirubicin cyclophosphamide/trastuzumab followed by docetaxel/trastuzumab alone or concomitant/followed by capecitabine within the GeparQuattro study. PTEN was assessed using the automated quantitative immunofluorescence analysis and was analyzed as a dichotomic variable. p4EBP1 was assessed by immunohistochemistry and used as a continuous and dichotomic variable.Results: p4EBP1 was available from 137, PTEN from 108, and PIK3CA genotype from 83 patients. Overall, the pCR rate in PTEN-low tumors was 27.6%, and in PTEN-high tumors, it was 57.1% (P = 0.010). pCR rates were not statistically different between PIK3CA wild-type and mutant (35% vs. 22%) or p4EBP1 IRS ≤ 4 and IRS > 4 (39% vs. 33%). pCR rate was 57.1% (8/14) in PTEN-high/PIK3CA wild-type and decreased to 15.4% in PTEN-low/PIK3CA-mutant tumors (P = 0.023). In multivariable analysis adjusted for baseline parameters, PTEN independently predicted pCR in the following cohorts: overall [OR, 7.54; 95% confidence interval (CI), 2.03–28.06; P = 0.003], PIK3CA wild-type (OR, 23.81; 95% CI, 1.75–324.05; P = 0.017), p4EBP1 IRS > 4 (OR, 11.53; 95% CI, 1.84–72.24; P = 0.009), and hormone receptor–positive (OR, 40.91; 95% CI, 2.93–570.44; P = 0.006). p4EBP1 was independently predictive for pCR in PIK3CA wild-type tumors (OR, 0.14; 95% CI, 0.03–0.78; P = 0.025).Conclusions: The study showed the potential role of PIK3CA genotype, PTEN, and p4EBP in predicting pCR after anthracycline–taxane-based chemotherapy and anti-HER2 treatment. Clin Cancer Res; 22(11); 2675–83. ©2016 AACR.

Usage metrics

    Clinical Cancer Research