Objectives <p>This study compared the diagnostic performance of diffusion biomarkers estimated from an abbreviated diffusion-weighted imaging (DWI) protocol and assessed their potential to reduce unnecessary biopsies of benign BI-RADS 4 lesions identified on dynamic contrast-enhanced (DCE) MRI.</p> Methods <p>A retrospective study was conducted from 2019 to 2023. All patients underwent abbreviated DWI at 3 T with four <i>b</i>-values (0 s/mm<sup>2</sup>, 200 s/mm<sup>2</sup>, 800 s/mm<sup>2</sup>, and 1500 s/mm<sup>2</sup>). Regions of interest were manually placed on DWI, and biomarkers, including the apparent diffusion coefficient (ADC<sub>0–800</sub>), perfusion fraction intravoxel incoherent motion, non-Gaussian diffusion (ADC<sub>0</sub> and kurtosis [<i>K</i>]), signature index (<i>S</i>-index), and shifted ADC (sADC), were estimated. Diagnostic performance and the potential to reduce unnecessary biopsies were evaluated for each parameter.</p> Results <p>In total, 168 female patients (mean age ± standard deviation, 56.2 ± 13.5 years) with 178 BI-RADS 4 lesions on DCE MRI were analyzed. The median ADC<sub>0–800</sub>, sADC, and ADC<sub>0</sub> were significantly lower in malignant lesions, while <i>S</i>-index and <i>K</i> were significantly higher (all <i>p</i> ≤ 0.001). The diagnostic performance to reclassify lesions as benign or malignant was identical for ADC<sub>0–800</sub> (area under the curve = 0.67), sADC (0.69), <i>S</i>-index (0.69), ADC<sub>0</sub> (0.68), and <i>K</i> (0.66). Applying an ad-hoc threshold cutoff, all parameters reduced unnecessary biopsies (around 16%), while <i>K</i> resulted in a slightly higher reduction rate than ADC<sub>0–800</sub> (20.5% vs 15.9%, <i>p</i> = 0.317) without reducing sensitivity.</p> Conclusion <p>Diffusion MRI biomarkers obtained using an abbreviated DWI protocol reduced unnecessary biopsies in BI-RADS 4 lesions, with <i>K</i> performing slightly better than ADC.</p> Key Points <p><Emphasis Type="BoldItalic">Question</Emphasis> <i>MRI BI-RADS category 4 includes a substantial number of benign lesions, and reducing unnecessary biopsies remains a critical clinical concern</i>.</p> <p><Emphasis Type="BoldItalic">Findings</Emphasis> <i>The parameters from abbreviated DWI show lesion differentiation comparable to ADC and have greater potential to reduce unnecessary biopsies</i>.</p> <p><Emphasis Type="BoldItalic">Clinical relevance</Emphasis> <i>This study underscores the potential of imaging biomarkers from abbreviated DWI for assessing breast MRI BI-RADS 4 lesions. These biomarkers may be comparable or superior to standard ADC in reducing unnecessary biopsies and could aid in improving patient management decisions</i>.</p> Graphical Abstract <p></p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Reduction of biopsy rate in BI-RADS4 breast lesions: potential of an abbreviated advanced DWI protocol

  • Mariko Goto,
  • Denis Le Bihan,
  • Koji Sakai,
  • Kei Yamada

摘要

Objectives

This study compared the diagnostic performance of diffusion biomarkers estimated from an abbreviated diffusion-weighted imaging (DWI) protocol and assessed their potential to reduce unnecessary biopsies of benign BI-RADS 4 lesions identified on dynamic contrast-enhanced (DCE) MRI.

Methods

A retrospective study was conducted from 2019 to 2023. All patients underwent abbreviated DWI at 3 T with four b-values (0 s/mm2, 200 s/mm2, 800 s/mm2, and 1500 s/mm2). Regions of interest were manually placed on DWI, and biomarkers, including the apparent diffusion coefficient (ADC0–800), perfusion fraction intravoxel incoherent motion, non-Gaussian diffusion (ADC0 and kurtosis [K]), signature index (S-index), and shifted ADC (sADC), were estimated. Diagnostic performance and the potential to reduce unnecessary biopsies were evaluated for each parameter.

Results

In total, 168 female patients (mean age ± standard deviation, 56.2 ± 13.5 years) with 178 BI-RADS 4 lesions on DCE MRI were analyzed. The median ADC0–800, sADC, and ADC0 were significantly lower in malignant lesions, while S-index and K were significantly higher (all p ≤ 0.001). The diagnostic performance to reclassify lesions as benign or malignant was identical for ADC0–800 (area under the curve = 0.67), sADC (0.69), S-index (0.69), ADC0 (0.68), and K (0.66). Applying an ad-hoc threshold cutoff, all parameters reduced unnecessary biopsies (around 16%), while K resulted in a slightly higher reduction rate than ADC0–800 (20.5% vs 15.9%, p = 0.317) without reducing sensitivity.

Conclusion

Diffusion MRI biomarkers obtained using an abbreviated DWI protocol reduced unnecessary biopsies in BI-RADS 4 lesions, with K performing slightly better than ADC.

Key Points

Question MRI BI-RADS category 4 includes a substantial number of benign lesions, and reducing unnecessary biopsies remains a critical clinical concern.

Findings The parameters from abbreviated DWI show lesion differentiation comparable to ADC and have greater potential to reduce unnecessary biopsies.

Clinical relevance This study underscores the potential of imaging biomarkers from abbreviated DWI for assessing breast MRI BI-RADS 4 lesions. These biomarkers may be comparable or superior to standard ADC in reducing unnecessary biopsies and could aid in improving patient management decisions.

Graphical Abstract