The volumetric-tumour histogram-based analysis of intravoxel incoherent motion and non-Gaussian diffusion MRI: association with prognostic factors in HER2-positive breast cancer

J Transl Med. 2019 Jul 2;17(1):182. doi: 10.1186/s12967-019-1911-6.

Abstract

Background: To evaluate the imaging biomarkers of human epidermal growth factor receptor 2 (HER2) positive breast cancer in comparison to other molecular subtypes and to determine the feasibility of identifying hormone receptor (HR) status and lymph node metastasis status using volumetric-tumour histogram-based analysis through intravoxel incoherent motion (IVIM) and non-Gaussian diffusion.

Methods: This study included 145 breast cancer patients with 148 lesions between January and November in 2018. Among the 148 lesions, 74 were confirmed to be HER2-positive. The volumetric-tumour histogram-based features were extracted from the combined IVIM and non-Gaussian diffusion model. IVIM and non-Gaussian diffusion parameters obtained from images of the subjects with different molecular prognostic biomarker statuses were compared by Student's t test or the Mann-Whitney U test. The area under the curve (AUC), sensitivity, and specificity at the best cut-off point were reported. The Spearman correlation coefficient was calculated to analyse the correlations of clinical tumor nodule metastasis (TNM) stage and Ki67 with the IVIM and non-Gaussian diffusion parameters.

Results: The entropy of mean kurtosis (MK) was significantly higher in the HER2-positive group than in the HER2-negative group (p = 0.015), with an AUC of 0.629 (95% CI 0.546, 0.707), a sensitivity of 62.6%, and a specificity of 66.2%. For HR status, the MD 5th percentile was higher in the HR-positive group of HER2-positive breast cancer (p = 0.041), with an AUC of 0.643 (95% CI 0.523, 0.751), while for lymph node status, the entropy of mean diffusivity (MK) was lower in the lymph node positive group (p = 0.040), with an AUC of 0.587 (95% CI 0.504, 0.668). The clinical TNM stage and Ki67 index were correlated with several histogram parameters.

Conclusion: Volumetric-lesion histogram analysis of IVIM and the non-Gaussian diffusion model can be used to provide prognostic information about HER2-positive breast cancers and potentially contribute to individualized anti-HER2 targeted therapy plans .

Keywords: Breast cancer; HER2-positive; Histogram-based analysis; Intravoxel incoherent motion; Non-Gaussian diffusion.

Publication types

  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Algorithms*
  • Area Under Curve
  • Biomarkers, Tumor / analysis
  • Breast Neoplasms / diagnosis*
  • Breast Neoplasms / genetics
  • Breast Neoplasms / metabolism
  • Breast Neoplasms / pathology
  • Computer Graphics*
  • Diffusion
  • Diffusion Magnetic Resonance Imaging / methods*
  • Feasibility Studies
  • Female
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Image Processing, Computer-Assisted / standards
  • Middle Aged
  • Motion
  • Prognosis
  • ROC Curve
  • Receptor, ErbB-2 / genetics
  • Receptor, ErbB-2 / metabolism
  • Tumor Burden*

Substances

  • Biomarkers, Tumor
  • ERBB2 protein, human
  • Receptor, ErbB-2