Assessment of the diagnostic accuracy of contrast-enhanced digital mammography in the differentiation between benign and malignant breast mass lesions

INTRODUCTION

Breast cancer is the most common cancer in women worldwide, and it is important to find an accurate and cost-effective tool for its early diagnosis ¹. Screen-film mammography had been proved for the reduction of breast cancer mortality; Since the early 2000s, It was progressively replaced by digital mammography (DM), which improved performance, especially in women under the age of 50 years and those with dense breasts ^2..Contrast-enhanced spectral mammography (CESM) utilizes a contrast agent to highlight areas of increased vascularization, such as those around and within tumors, using standard mammography equipment ³.

 Contrast enhanced digital mammography (CEDM) combines digital mammography with the benefits of contrast imaging, using  the  high  rates  of  blood  vessel  formation in tumors to distinguish between cancerous and healthy tissue ⁴. The physiologic information provided by CEDM is similar to that is provided by breast MRI, without the added time or expense of conventional breast MRI protocols ⁵. CEDM has a higher sensitivity for breast cancer detection compared with the sensitivities of full-field digital mammography (FFDM) alone and FFDM combined with ultrasound. For assessing tumor extent, its findings had a good correlation with the histopathology size, even in dense breasts ^6-15.

The purpose of this work was to evaluate the accuracy of CEDM in the diagnosis and differentiation between benign and malignant breast masses and mass like lesions.

PATIENTS AND METHODS

Patient’s demographic data

This prospective study included 100 female patients with 154 breast lumps; their ages ranged from 30 to 75 years with a mean age=±42 years who referred to radio-diagnosis departments with a clinical diagnosis of a suspicious breast lump. Informed consent was obtained from all patients. Patients who had general contraindications for contrast media or who could not tolerate the mammography exam. were excluded from this study.

Inclusion criteria

All Patients with newly developed suspicious breast lump who did not have history of surgery or radiation therapy.

Exclusion criteria

Pregnant patients and patients with general contraindications for contrast media as well as those who could not tolerate the mammography examination. 

All patients were subjected to local breast examination of both sides, renal function tests; Conventional mammography (CM), breast ultrasonography (U/S) & contrast-enhanced digital mammography (CEDM). Each detected lesion was specified according to the BI-RADS classification, localization, size (maximum diameter), and the distance between disease sites, which was used to determine whether surgery would involve conservation (multifocal disease) or mastectomy (multicentric disease).

Conventional & Dual-energy CEDM technique

Conventional & CEDM were performed to all patients by using a digital mammography unit (Senographe 2000 D full-field digital mammography Essential GE Healthcare). First, craniocaudal (CC) and mediolateral oblique (MLO) views were obtained for all patients while additional views were taken if indicated (e.g. Lateral /Spot views). A cannula is inserted on the opposite side of the affected breast & I.V. iodinated contrast agent (omnipaque, 300 mg/ml) injected prior to patient positioning in a dose(1-1.5 ml/Kg).

Ultrasonography(U/S):- Detailed breath U/S exam. was done for both  breasts and axilla by ultrasound machine (Logiq P5, GE Medical Systems) with 7.5–12 MHz using lineararray transducer. Also, all biopsies were taken under an U/S guide.

Dual-energy CEDM was performed by acquiring a pair of low- and high-energy images in quick succession during a single breast compression & after the initiation of contrast medium injection by 3 minutes, a set of bilateral imaging as followed: CC view of the unaffected breast followed by CC and MLO views of the affected breast; and then MLO view of the unaffected breast. Low-energy images were acquired at peak Kilovoltage (KV), 25-30, which is below the K-edge of iodine "33.2 KV" while high-energy images were acquired at 42-50 kV, which is above the K-edge of iodine, to perform a pair of low and high energy exposures in order to generate two subtracted images with contrast agent uptake information.

Image interpretation

All images were reviewed in consensus by two radiologists (HA and YA) with experience in breast imaging who were blinded to clinical data, and only concentrated on image findings and lesion characterization, including:- location, number, shape, borders, surrounding parenchyma, calcification if present and pattern of contrast enchantment. Focally enhanced lesions beyond normal background breast enhancement was considered as abnormal and the probability of malignancy using the BI-RADS system was applied for each lesion evaluation in all imaging techniques.

Biopsy with the histopathology results or clinical diagnosis after a conservative treatment & follow-up was taken as gold standard

Statistical analysis

Data entry and analysis was done using the program Statistical Package for the Social Sciences (SPSS). The technique results were studied using Student’s t-test & Sensitivity, specificity; negative and positive predictive values (PPV) were assessed and compared. P

RESULTS

This study was carried on 100 patients with 154 breast lesions that according to the histopathology results proved to be:-122 malignant neoplastic lesions (79.22%), with 98invasive lesions & 24 lesions were non-invasive. While the remaining 32 lesions were benign (20.78%) with 16  fibro-adenomas & 9 lesions were granulomatous mastitis, (Table 1).

Conventional mammographic & U/S findings in studied breast lesions

Twenty-one benign lesion (21/32) showed well-defined margin (BI-RAD 2) that proved to be 16 fibroadenomas (FAL), 3 chronic abscesses and 2 fibro-adenolipoma while, 10 benign lesions (BI-RAD 3-4) showed poorly defined outlines that proved to be 9  granulomatous mastitis and 1 fat necrosis. The last benign lesion showed lobulated outlines that pathologically proved to be a phylloid(BI-RAD 3). However, 59/122 malignant lesions showed lobulated margins (BI-RAD 3-4) and 47/122 lesions (BI-RAD 4-5) showed ill-defined margin and the remaining 16 lesions showed speculated outlines (BI-RAD 5). Macro-calcification were detected in 17/35 beings lesions & micro-calcification seen in 69/122 malignant lesions while mixed calcifications were seen in 7maligant lesion & no calcification detected in the remaining 61 lesions, (Table 2).

Distribution of lesions visible in CEDM

We made our study  on 100 patients, 70 have single lesion that all were visible by sono-mammography (62 malignant and 8 benign lesions),while 59/62 lesions were visible by CEDM and 3 patientswith malignant lesions were not enhanced; 13 patients each had 2 contrast-enhanced lesions (18 malignant&8 benign).  The remaining 17 patients, had more than 2 lesions (42 malignant & 16 benign lesions), (Table 3).

CEDM Imaging findings & enhancement pattern

We diagnosed 122 pathological lesions as malignant lesions, 78 lesions had heterogeneous mass enhancement while  homogenous  enhancement  was seen in 12 masses, 2 lesions showed non-mass ductal enhancement, 14 lesions showed non-mass regional heterogeneous enhancement, non-mass segmental enhancement was seen in 6 lesions, heterogeneous mass enhancement with ductal enhancement seen in 5 masses, heterogeneous mass enhancement with foci of enhancement seen in 2  masses and 3 malignant lesions showed no enhancement and were not observed in CEDM (false negative). The 32 pathologically diagnosed benign lesions showed heterogeneous enhancement in 12 (7 granulomatous mastitis, 4 fibroadenoma & 1 phylloid lesion), homogenous seen enhancement in 12 fibroadenomas, ring enhancement seen in  8 lesions, 3 chronic abscesses, 2 granulomatous mastitis, 2 fibroadenolipoma and 1 fat necrosis), (Table 4).

Lesion outlines, borders and size in CEDM in correlation with pathological correlations

Out of the 154 detected  lesions 35 diagnosed as benign lesions (32 lesions that showed  well-defined margin but 3 lesions were with indistinct outlines). However, 119/154 lesions were diagnosed as malignant lesions, 59 lesions showed lobulated margins (58 malignant lesions & 1 was phylloid according to histopathology results), and 60 lesions showed ill-defined margin (59 lesions proved to be malignant and  1 was benign fat necrosis by histopathology. As regards the lesion sizes, they were ranged from 0.4 to 7 cm that did not reflected on radiologic diagnosis, with P-value= 0.4117 (not significant), (table 5).

Sensitivity, specificity, and accuracy of CEDM versus mammography  combined with U/S for all studied lesions

CEDM was found to have higher sensitivity 97.56%, specificity 93.55) and accuracy 96.75% when compared to combined sono-mammography   results that showed sensitivity 62.30%, specificity 68.75% and accuracy 63.64% , (Table 6, Figures 1-5).

Fig 1: A 31 yes female Pt.  (a) DM: revealed a non specific asymmetric densities at the UOQ. (b) U/S: showed a para-areolar cystic lesion with thick irregular walls and non-homogenous echoes(a&b: BIRAD 3-4). (c)CEDM: revealed a small well defined lesion with ring enhancement at the UOQ & areas of non-mass segmental enhancement suggesting inflammatory process with abscess formation (BIRAD 2). Aspiration & cytology:-Abscess 

Fig 2: A 42 yes female Pt.  (a) DM:  revealed poorly defined dense lesions at UOQ (b) U/S: showed a poorly defined deep hypoechoic lesion with irregular outlines (a&b: BIRAD 3). c)CEDM: revealed non-mass heterogeneous regional enhancement with distortion of related tissue architectures echoes (BIRAD 4). Histopathology: Ductal Ca insitu  with microscopic invasion.

Fig 3: A 55 yes female Pt.  (a) DM.: revealed diffuse edema with skin thickening  and UOQ a small well defined dense lesion. (b) U/S: showed an ill-defined heterogeneous hypo-echoic lesion with intra-ductal extension (a&b: BIRAD 4). (c) CEDM: Revealed: an irregular  moderate heterogeneous enhanced lesion with speculated outlines (BIRAD 5). Histopathology: Invasive ductal Ca.

Fig 4: A 40 yes female Pt. (a)DM: reveled LOQ intermediately dense obscured lesions with clusters of mixed calcification. (b)U/S: showed multiple infiltrative hypoechoic lesions with micro-calcification and ductal extensions(a&b: multiple lesions with BIRAD 5). (c) CEDM: detected more lesions with irregular outlines and moderate/marked heterogeneous enhancement (Multicentric lesions with BIRAD 5). Histopathology: Multicentric IDC.

Fig 5: A 62  yes  female Pt.  (a) DM revealed a well circumscribed high density lesion at the LIC. (b) U/S showed a well-defined hypoechoic lesion (a&b: BIRAD 3).  (c)CEDM revealed  a well defined lesion with  irregular outlines & diffuse intense homogenous enhancement (BIRAD 5). Histopathology:  Metastatic neuro-endocrine Ca.

Table 1: Final histopathology diagnosis of studied 154 breast lesions.

Table 2: CM & U/S findings in the studied breast lesions using BI-RAD scoring.

Table 3: Distribution of lesions visible in CESM.

Table 4: Enhancement pattern of the studied 154 lesions in CEDM.

Table 5: Lesion outlines, borders and size in CEDM in correlation with pathological results.

Table 6: Sensitivity, specificity and accuracy of CEDM for all studied lesions.