Quantitative evaluation of breast cancer response to neoadjuvant chemotherapy by diffusion tensor imaging: Initial results

J Magn Reson Imaging. 2018 Apr;47(4):1080-1090. doi: 10.1002/jmri.25855. Epub 2017 Sep 13.

Abstract

Background: Diffusion tensor imaging (DTI) yields several parameters that have not been tested in response evaluation to neoadjuvant chemotherapy (NAC).

Purpose: To evaluate and compare in reference to histopathology findings the ability of DTI and dynamic contrast-enhanced (DCE) MRI to monitor response to NAC.

Study type: Retrospective.

Population: Twenty patients treated with neoadjuvant chemotherapy.

Field strength/sequence: 1.5T MRI axial, bilateral T2 -weighted, DTI, and DCE-MRI.

Assessment: A standardized blinded image analysis at pixel resolution generated color-coded maps of DTI and DCE parameters STATISTICAL TESTS: Pearson's correlation analysis and Bland-Altman plots of the DTI and DCE size changes and of the pathological final residual tumor diameter and DCE or DTI final diameter, from pre- to post-NAC. Spearman coefficient of rank correlation between the DTI and DCE size changes from pre- to post-NAC and Miller and Payne (M&P) pathological response grading. Receiver operating characteristic curve analyses to differentiate between responders to nonresponders on the basis of the DTI and DCE percent size changes and the changes in DTI parameters.

Results: DTI and DCE changes in the cancers' diameter and volume from pre- to post-NAC exhibited high and significant Pearson correlation (r = 0.82 P = 1.2 × 10-5 ). The DTI volume changes exhibited a significant Spearman coefficient rank correlation (0.68, P = 0.001) with the pathological M&P grading and differentiated between responders and nonresponders with area-under-the-curve (AUC) of 0.83 ± 0.10. A similar AUC for differentiating responders from nonresponders was exhibited by the changes in the highest diffusion coefficient (0.84 ± 0.11) and the mean diffusivity (0.83 ± 0.11). The DTI residual-tumor-diameter showed a high and significant Pearson correlation (r = 0.87 P = 1.2 × 10-6 ) to pathology tumor diameter.

Data conclusion: DTI monitors changes in cancer size and diffusion tensor parameters in response to NAC with an accuracy equivalent to that of DCE, enabling differentiation of responders from nonresponders and assessment of residual tumor size in high congruence with pathology.

Level of evidence: 1 Technical Efficacy: Stage 4 J. Magn. Reson. Imaging 2018;47:1080-1090.

Keywords: breast cancer; diffusion tensor imaging; magnetic resonance imaging; neoadjuvant therapy; response to therapy.

Publication types

  • Evaluation Study

MeSH terms

  • Adult
  • Aged
  • Breast / diagnostic imaging
  • Breast Neoplasms / diagnostic imaging*
  • Breast Neoplasms / drug therapy*
  • Contrast Media
  • Diffusion Tensor Imaging / methods*
  • Evaluation Studies as Topic
  • Female
  • Humans
  • Image Enhancement / methods
  • Magnetic Resonance Imaging
  • Middle Aged
  • Neoadjuvant Therapy / methods*
  • Retrospective Studies

Substances

  • Contrast Media