OBJECTIVES

To evaluate the diagnostic performance of post-neoadjuvant therapy (NAT) diffusion-weighted magnetic resonance imaging (DW-MRI) compared to contrast-enhanced MRI (CE-MRI) in predicting tumor response in breast cancer.

MATERIALS AND METHODS

Patients with breast cancer who were examined using 3.0-T MRI after NAT and underwent surgery between 2015 and 2018 were studied. Qualitative assessment of pathological complete response (pCR) and size of residual tumor following NAT were evaluated using CE-MRI and DW-MRI with b-values of 0 and 1000 s/mm. Apparent diffusion coefficient (ADC) cut-off for predicting pCR was estimated using Youden's Index in lesions with residual high signal intensity. Agreement of residual tumor size at CE-MRI, DW-MRI, and pathology was compared.

RESULTS

A total of 1062 patients (49.3 years ± 9.0) were evaluated. Sensitivity for predicting pCR did not show a significant difference between CE-MRI and DW-MRI (65.3% and 62.9%, p = 0.50). The HR+/HER2- subtype showed significantly lower positive predictive value and higher negative predictive value than the others. Agreement of residual tumor size between CE-MRI and pathology (intraclass correlation coefficient [ICC], 0.62; 95% confidence interval [CI]: 0.58-0.66) and between DW-MRI and pathology (ICC, 0.64; 95% CI: 0.61-0.68) did not differ significantly (p = 0.49). Residual tumor size was underestimated in the HR+/HER2- subtype and overestimated in the HR-/HER2+ subtype. The optimum ADC cut-off was 1.22 × 10 mm/s, with the area under the curve of 0.73 (95% CI: 0.68-0.78).

CONCLUSION

DW-MRI demonstrated comparable sensitivity for predicting pCR and agreement in residual invasive tumor size measurement to CE-MRI. ADC can be used as an additional tool for distinguishing residual tumors from post-treatment changes.

KEY POINTS

Question The effectiveness of diffusion-weighted MRI in predicting tumor response to NAT in breast cancer remains uncertain. Findings Diffusion-weighted MRI demonstrated similar sensitivity to CE-MRI in predicting pCR and comparable accuracy in measuring residual tumor size. Clinical relevance Diffusion-weighted MRI could potentially be used as an alternative or complementary technique to CE-MRI for evaluating tumor response to NAT in patients with breast cancer.