Scientific Reports (Feb 2021)

Baseline microvessel density predicts response to neoadjuvant bevacizumab treatment of locally advanced breast cancer

  • Kristi Krüger,
  • Laxmi Silwal-Pandit,
  • Elisabeth Wik,
  • Oddbjørn Straume,
  • Ingunn M. Stefansson,
  • Elin Borgen,
  • Øystein Garred,
  • Bjørn Naume,
  • Olav Engebraaten,
  • Lars A. Akslen

DOI
https://doi.org/10.1038/s41598-021-81914-0
Journal volume & issue
Vol. 11, no. 1
pp. 1 – 11

Abstract

Read online

Abstract A subset of breast cancer patients benefits from preoperative bevacizumab and chemotherapy, but validated predictive biomarkers are lacking. Here, we aimed to evaluate tissue-based angiogenesis markers for potential predictive value regarding response to neoadjuvant bevacizumab treatment in breast cancer. In this randomized 1:1 phase II clinical trial, 132 patients with large or locally advanced HER2-negative tumors received chemotherapy ± bevacizumab. Dual Factor VIII/Ki-67 immunohistochemical staining was performed on core needle biopsies at baseline and week 12. Microvessel density (MVD), proliferative microvessel density (pMVD; Factor VIII/Ki-67 co-expression), glomeruloid microvascular proliferation (GMP), and a gene expression angiogenesis signature score, were studied in relation to pathologic complete response (pCR), clinico-pathologic features and intrinsic molecular subtype. We found that high baseline MVD (by median) significantly predicted pCR in the bevacizumab-arm (odds ratio 4.9, P = 0.012). High pMVD, presence of GMP, and the angiogenesis signature score did not predict pCR, but were associated with basal-like (P ≤ 0.009) and triple negative phenotypes (P ≤ 0.041). pMVD and GMP did also associate with high-grade tumors (P ≤ 0.048). To conclude, high baseline MVD significantly predicted response to bevacizumab treatment. In contrast, pMVD, GMP, and the angiogenesis signature score, did not predict response, but associated with aggressive tumor features, including basal-like and triple-negative phenotypes.