Janis M. Yee1, Daniel S. Hippe1, Michael Hirano1, Bonny Chau1, Debosmita Biswas1, Anum S. Kazerouni1, Mary Lynn Bryant1, Isabella Li1, Jennifer Xiao1, Wei Huang2, Savannah C. Partridge1, and Habib Rahbar1
1Radiology, University of Washington, Seattle, WA, United States, 2Oregon Health & Science University, Portland, OR, United States
Synopsis
Mammographic calcifications remain a diagnostic dilemma with
low positive predictive value for malignancy. Our study investigated the use of
quantitative DCE and DW-MRI metrics to problem solve suspicious calcifications
prior to biopsy. In patients with a suspicious enhancing MRI correlate, malignant
lesions exhibited higher peak PE, peak SER, functional tumor volume, and Ktrans.
Basic and advanced 3D DW-MRI parameters did not yield statistically significant
discriminatory values in this small pilot study. Use of quantitative MRI
features shows potential to reduce the number of unnecessary biopsies for
suspicious mammographic calcifications among the subset that demonstrate
suspicious enhancement.
Introduction
Mammographic calcifications are associated with nearly half of clinically
occult breast cancers, with the majority linked to ductal carcinoma in situ (DCIS).1 However, the positive
predictive value (PPV) for malignancy of pure calcifications remains low,
ranging from 15-40%.2
Breast MRI has been proposed as a means to problem solve mammographically-detected
calcifications due to its high sensitivity (reported as high as 94-100%).3 However, specificity and PPV
for malignancy on the basis of clinical features alone is modest; thus, the use of
conventional MRI alone still results in many unnecessary biopsies. Prior
studies have shown that quantitative lesion analysis using a combined high
spatial and high temporal resolution acquisition can provide discriminatory
value to differentiate benign from malignant lesions. Furthermore, advanced
diffusion weighted imaging metrics, including intravoxel incoherent motion
(IVIM)-derived parameters, hold promise to provide complementary MRI signatures
to further improve the characterization of breast lesions.4-6 The purpose of our
study was to investigate if these quantitative MRI parameters could be used to
accurately predict benign or malignant pathology of suspicious mammographic calcifications
presenting as suspicious enhancement on MRI. Materials and Methods
This IRB-approved, HIPAA-compliant, prospective study
(Clinical Trials.org NCT03495011) included 76 women with 79 unique areas of previously
identified mammographic calcifications who provided informed consent prior to
undergoing a research 3T dynamic contrast-enhanced (DCE) and multi-b value (0,
10, 20, 30, 50, 70, 100, 150, 200, 400, 600, 800, 1000 s/mm2) diffusion
weighted MRI (DW-MRI) from April 2018 to March 2020. All MRIs were performed
prior to biopsy of the calcifications. DCE-MRI was performed using a hybrid
acquisition combining high temporal resolution keyhole-based (variable 6.9-30s
sampling, 23 total acquisitions) and high spatial resolution (0.8 × 0.8 mm2,
three total acquisitions, pre-contrast, 2 and 7 min post-contrast) techniques. MRIs
were prospectively interpreted by a fellowship-trained breast imaging
radiologist prior to biopsy, referencing the mammographic findings.
Calcifications with an enhancing correlate underwent quantitative 3D lesion MRI
analysis. DCE-MRI data was fit to the Kety-Tofts model to extract
pharmacokinetic parameters (Ktrans, ve, kep).
Additional features including signal enhancement ratio (SER), percent
enhancement (PE), functional tumor volume (FTV), and washout fraction were
obtained. Peak PE and peak SER were determined for
hot-spot regions producing the highest PE and SER values. DW-MRI data
was analyzed to extract apparent diffusion coefficient (ADC), pseudodiffusion
(Dp), true diffusion coefficient (Dt), and perfusion
fraction (f) for each lesion (Figure 1). Reference
standard for all lesions was histopathological evaluation of core needle
biopsies (CNB) obtained via stereotactic biopsy and surgical excision for
malignant lesions. The ability of each quantitative MRI parameter to
discriminate benign and malignant lesions was summarized using the area under
the receiver operating characteristic curve (AUC) and tested using the Wilcoxon
rank-sum test. Results
Of the
entire cohort of 79 calcifications, 22 were malignant (9 high grade DCIS, 6
non-high grade DCIS, 7 invasive carcinoma) resulting in a PPV of 28% (22/79) on
mammography alone. Twenty-three out of 79 calcifications demonstrated a
suspicious enhancing correlate on MRI, of which 15 were malignant (6 high grade
DCIS, 3 non-high grade DCIS, 6 invasive carcinoma) for a PPV of 65% (15/23) and
a negative predictive value of 88% (49/56; 3 high grade DCIS, 3 non-high grade
DCIS, and 1 invasive carcinoma demonstrated no suspicious enhancement). Included
women were mean age 56 ±10 years; 70% (16/23) had mammographically dense
breasts while 65% (15/23) had minimal/mild background parenchymal enhancement
on MRI (Table 1). Malignant lesions exhibiting suspicious enhancement
demonstrated higher peak PE (mean: 209 ±45 vs. 158 ±48, AUC=0.82, p=0.013),
peak SER (mean: 1.22 ±0.23 vs. 1.06 ±0.19, AUC=0.77, p=0.040), FTV (median: 1.34
cm3 vs. 0.23 cm3, AUC=0.93, p<0.001), and Ktrans (mean:
0.06 ±0.03 s-1 vs. 0.05 ±0.02 s-1, AUC=0.76, p=0.047)
than benign lesions (Table 2). Washout fraction, ve, and kep
were not statistically significantly different between malignant and benign
lesions (p>0.05). Of the 23 suspicious enhancing lesions, 19 (13 malignant, 6
benign) demonstrated a DW-MRI correlate free from significant artifact that
allowed for quantitation. No basic or advanced diffusion parameters were found
to provide discrimination between benign and malignant lesions, although f (mean:
0.09 ±0.03 vs. 0.17 ±0.12, AUC=0.71, p=0.18) showed potential value (Table 2).Conclusion and Discussion
Results from this prospective trial show that specific quantitative
measures of enhancing MRI correlates to mammographically suspicious calcifications
can assist with differentiating benign from malignant pathologies and thereby further
decrease unnecessary biopsies. These findings are consistent with prior reports
on the use of quantitative DCE parameters 7, 8
to discriminate among benign and malignant lesions. Although no DW-MRI feature
added significant value in this small study, IVIM perfusion fraction showed a potential
trend that could be studied in larger cohorts. Additional studies are warranted
to further investigate advanced DW-MRI parameters and their utility to problem
solve suspicious calcifications as well as serve as prognostic markers for the
subset of DCIS lesions that represent possible overdiagnosis. Acknowledgements
This study was supported by NIH/NCI grants R01CA203883, R01CA207290 and R01CA248192.References
- Stomper PC, Geradts J, Edge SB, Levine EG. Mammographic predictors of the presence and size of invasive carcinomas associated with malignant microcalcification lesions without a mass. AJR Am J Roentgenol. Dec 2003;181(6):1679-84. doi:10.2214/ajr.181.6.1811679
- Shehata M, Grimm L, Ballantyne N, et al. Ductal Carcinoma in Situ: Current Concepts in Biology, Imaging, and Treatment. Journal of Breast Imaging. 2019;1(3):166-176. doi:10.1093/jbi/wbz039
- Liberman L, Morris EA, Lee MJ-Y, et al. Breast Lesions Detected on MR Imaging: Features and Positive Predictive Value. American Journal of Roentgenology. 2002/07/01 2002;179(1):171-178. doi:10.2214/ajr.179.1.1790171
- Tao WJ, Zhang HX, Zhang LM, et al. Combined application of pharamcokinetic DCE-MRI and IVIM-DWI could improve detection efficiency in early diagnosis of ductal carcinoma in situ. J Appl Clin Med Phys. Jul 2019;20(7):142-150. doi:10.1002/acm2.126245.
- Partridge SC, Nissan N, Rahbar H, Kitsch AE, Sigmund EE. Diffusion-weighted breast MRI: Clinical applications and emerging techniques. https://doi.org/10.1002/jmri.25479. Journal of Magnetic Resonance Imaging. 2017/02/01 2017;45(2):337-355. doi:https://doi.org/10.1002/jmri.25479
- Liu C, Liang C, Liu Z, Zhang S, Huang B. Intravoxel incoherent motion (IVIM) in evaluation of breast lesions: Comparison with conventional DWI. European Journal of Radiology. 2013;82(12):e782-e789. doi:10.1016/j.ejrad.2013.08.006
- Arasu VA, Chen RCY, Newitt DN, et al. Can signal enhancement ratio (SER) reduce the number of recommended biopsies without affecting cancer yield in occult MRI-detected lesions? Academic radiology. 2011;18(6):716-721. doi:10.1016/j.acra.2011.02.008
- Jansen SA, Fan X, Karczmar GS, Abe H, Schmidt RA, Newstead GM. Differentiation between benign and malignant breast lesions detected by bilateral dynamic contrast-enhanced MRI: a sensitivity and specificity study. Magnetic resonance in medicine. 2008;59(4):747-754. doi:10.1002/mrm.21530