Whole-lesion histogram analysis of the apparent diffusion coefficient: Evaluation of the correlation with subtypes of mucinous breast carcinoma.

PURPOSE

To evaluate the utility of the whole-lesion histogram apparent diffusion coefficient (ADC) for characterizing the heterogeneity of mucinous breast carcinoma (MBC) and to determine which ADC metrics may help to best differentiate subtypes of MBC.

MATERIALS AND METHODS

This retrospective study involved 52 MBC patients, including 37 pure MBC (PMBC) and 15 mixed MBC (MMBC). The PMBC patients were subtyped into PMBC-A (20 cases) and PMBC-B (17 cases) groups. All patients underwent preoperative diffusion-weighted imaging (DWI) at 1.5T and the whole-lesion ADC assessments were generated. Histogram-derived ADC parameters were compared between PMBC vs. MMBC and PMBC-A vs. PMBC-B, and receiver operating characteristic (ROC) curve analysis was used to determine optimal histogram parameters for differentiating these groups.

RESULTS

The PMBC group exhibited significantly higher ADC values for the mean (P = 0.004), 25 (P = 0.004), 50 (P = 0.004), 75 (P = 0.006), and 90 percentiles (P = 0.013) and skewness (P = 0.021) than did the MMBC group. The 25 percentile of ADC values achieved the highest area under the curve (AUC) (0.792), with a cutoff value of 1.345 × 10 mm /s, in distinguishing PMBC and MMBC. The PMBC-A group showed significantly higher ADC values for the mean (P = 0.049), 25 (P = 0.015), and 50 (P = 0.026) percentiles and skewness (P = 0.004) than did the PMBC-B group. The 25 percentile of the ADC cutoff value (1.476 × 10 mm /s) demonstrated the best AUC (0.837) among the ADC values for distinguishing PMBC-A and PMBC-B.

CONCLUSION

Whole-lesion ADC histogram analysis enables comprehensive evaluation of an MBC in its entirety and differentiating subtypes of MBC. Thus, it may be a helpful and supportive tool for conventional MRI.

LEVEL OF EVIDENCE

4 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:391-400.