Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions

Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions

+Model ARTICLE IN PRESS DIII-854; No. of Pages 11 Diagnostic and Interventional Imaging (2016) xxx, xxx—xxx ORIGINAL ARTICLE /Breast imaging Cont...

3MB Sizes 0 Downloads 37 Views

+Model

ARTICLE IN PRESS

DIII-854; No. of Pages 11

Diagnostic and Interventional Imaging (2016) xxx, xxx—xxx

ORIGINAL ARTICLE /Breast imaging

Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions L. Li a, R. Roth a, P. Germaine b, S. Ren c, M. Lee a, K. Hunter e, E. Tinney d, L. Liao f,∗ a

Department of Radiology, Cooper University Hospital, 900 Centennial Blvd., Building 1, Suite B, Voorhees, NJ 08043, USA b Department of Radiology, Cooper University Hospital, One Cooper Plaza, Camden, NJ 08013, USA c Department of Pathology, Cooper University Hospital, , One Cooper Plaza, Camden, NJ 08013, USA d Cooper University Hospital, Department of Radiology, 900 Centennial Blvd., Building 1, Suite B, Voorhees, NJ 08043, USA e Cooper Medical School of Rowan University; Statistician II, Cooper Research Institute, Cooper University Hospital, 401 Haddon Ave., Room 140, Camden, NJ 08103, USA f Cooper Breast Imaging Center, Cooper University Hospital, 900 Centennial Blvd., Building 1, Suite B, Voorhees, NJ 08043, USA

KEYWORDS Contrast-enhanced spectral mammography (CESM); Magnetic resonance imaging (MRI); Breast cancer diagnosis; Comparative studies

Abstract Objective: The purpose of this study was to retrospectively compare the diagnostic performance of contrast-enhanced spectral mammography (CESM) with that of breast magnetic resonance imaging (BMRI) in breast cancer detection using parameters, including sensitivity, positive predictive value (PPV), lesion size, morphology, lesion and background enhancement, and examination time. Materials and methods: A total of 48 women (mean age, 56 years ± 10.6 [SD]) with breast lesions detected between October 2012 and March 2014 were included. Both CESM and BMRI were performed for each patient within 30 days. The enhancement intensity of lesions and breast background parenchyma was subjectively assessed for both modalities and was quantified for comparison. Statistical significance was analyzed using paired t-test for mean size of index



Corresponding author. Cooper Breast Imaging Centers (CBIC), Anderson Cancer Center, Cooper University Hospital, Cooper Medical School of Rowan University, 900 Centennial Blvd., Building 1, Suite B, Voorhees, NJ 08043, USA. E-mail addresses: [email protected] (L. Li), [email protected] (R. Roth), [email protected] (P. Germaine), [email protected] (S. Ren), [email protected] (M. Lee), [email protected] (K. Hunter), [email protected] (E. Tinney), [email protected], [email protected] (L. Liao). http://dx.doi.org/10.1016/j.diii.2016.08.013 2211-5684/© 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

2

L. Li et al. lesions in all malignant breasts (an index lesion defined as the largest lesion in each breast), and a mean score of enhancement intensity for index lesions and breast background. PPV, sensitivity, and accuracy were calculated for both CESM and BMRI. The average duration time of CESM and MRI examinations was also compared. Results: A total of 66 lesions were identified, including 62 malignant and 4 benign lesions. Both CESM and BMRI demonstrated a sensitivity of 100% for detection of breast cancer. There was no statistically significant difference between the mean size of index lesions (P = 0.108). The enhancement intensity of breast background was significantly lower for CESM than for BMRI (P < 0.01). The mean score of enhancement intensity of index lesions on CESM was significantly less than that for BMRI (P < 0.01). The smallest lesion that was detected by both modalities measured 4 mm. CESM had a higher PPV than BMRI (P > 0.05). The average examination time for CESM was significantly shorter than that of BMRI (P < 0.01). Conclusion: CESM has similar sensitivity than BMRI in breast cancer detection, with higher PPV and less background enhancement. CESM is associate with significantly shorter exam time thus a more accessible alternative to BMRI, and has the potential to play an important tool in breast cancer detection and staging. © 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.

Screening mammography is well documented to reduce mortality of breast cancer [1]. Mammographic sensitivity for the detection of breast cancer is 75—85% [2], but can decrease to less than 50% in dense breast parenchyma [3,4]. Mammographic specificity and positive predictive value (PPV) are in the 25—45% range [5—7]. Contrast-enhanced breast magnetic resonance imaging (BMRI) is the imaging gold standard for breast cancer detection, staging, and post-treatment follow-up [6,8]. Sensitivity of BMRI for cancer detection ranges from 79 to 98% [9,10], but is limited by variable specificity of 52 to 77% [11—13], high cost, relatively long examination time, and limited availability due to a variety of reasons, including contraindications, magnet time, different locations, and insurance pre-authorization challenges [13—15]. Contrasted enhanced spectral mammography (CESM) was first approved by the FDA in 2011 (SenoBright, GE HealthCare). CESM utilizes dual energy for mammographic acquisition after intravenous administration of iodinated contrast agent. Previous studies have compared the sensitivity of CESM to that of conventional digital mammography, ultrasound, and BMRI [16—18]. CESM sensitivity has been reported as 93—100% [18,19], significantly higher compared to mammography and ultrasound alone. CESM has been shown to detect additional mammographically occult cancers and to depict more accurately the extent of disease to help guide surgical and treatment planning [20—22]. However, few studies have been published on CESM compared to BMRI for breast cancer detection, lesion size estimation, and preoperative staging [19,20]. The purpose of this study was to retrospectively compare the diagnostic performance of CESM with that of BMRI in breast cancer detection using parameters, including sensitivity, positive predictive value (PPV), lesion size, morphology, lesion and background enhancement, and examination time.

Materials and methods The study was HIPAA compliant and was approved by the Institutional Review Board (IRB) at Cooper University Hospital. All patients completed informed consent for CESM as required prior to being selected as subjects for the current study.

Inclusion and exclusion criteria Subjects selected in the study were patients who had a confirmed breast malignancy by tissue sampling and had undergone CESM and BMRI within 30 days. Patients were excluded if they did not have a cancer diagnosis. A total of 48 patients during the recruitment period of October 2012 to March 2014 fulfilled the inclusion criteria and were selected in the analysis.

Logistic of CESM and BMRI In compliance with the institutional clinical protocol, 45 CESM were performed before needle biopsy for suspicious imaging findings on either screening or diagnostic mammogram or ultrasound. Only 3 CESM were performed for staging after needle biopsy (patients referred from other institutions with outside tissue sampling). Forty BMRI examinations were performed after cancer diagnosis was made as a staging examination. Eight BMRI were performed before needle biopsy. This is because BMRI exam requires precertification while CESM does not require a pre-certification. More biopsies were performed on additional lesions identified on CESM and/or BMRI either in the same breast or contralateral breast, to assess multifocality or contralateral breast cancer. Eighteen of the 48 study patients have received additional biopsies after initial diagnosis. In total, there were 62 malignant and 4 benign lesions in 50 breasts

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

Contrast-enhanced spectral mammography versus breast magnetic resonance imaging of 48 patients (two patients were diagnosed with bilateral breast cancer).

Image analysis After patients were selected for the current study, CESM and BMRI images were retrospectively reviewed by two American Board or Radiology (ABR) certified radiologists in a consensus manner. CESM and BMRI were read separately. The two radiologists have at least two years of experience in CESM and eight years in BMRI. All study cases with known diagnosis were reviewed retrospectively. CESM was reviewed independently and at different time from BMRI to avoid bias. The largest lesion of each breast was defined as an index lesion. The size of index lesions was recorded. The level of enhancement intensity both on index lesions and breast background on CESM and BMRI were quantified using guidelines for BMRI by from American College of Radiology Brest Imaging-Reporting and Data System 5th edition (ACR BI-RADS BI-RADS® ) [23]. Breast background enhancement represents breast parenchymal enhancement (BPE). The quantification for enhancement intensity of index lesions was scaled from 0 to 3, as 0 = no enhancement, 1 = minimal enhancement, 2 = moderate enhancement, and 3 = marked enhancement. The quantification for breast parenchymal enhancement (BPE) was scaled from 0 to 3, as 0 = no enhancement, 1 = minimal enhancement (< 25% glandular enhancement), 2 = moderate enhancement (26—50% glandular enhancement), and 3 = moderate and marked enhancement (> 50% glandular enhancement). The reason to combine moderate and marked enhancement to one group is because there is few number of marked background enhancement in our study patients because the number of marked background enhancement in CESM study was small. Cancer morphology was analyzed using BMRI BI-RADS lexicon [24,25] as standard. The consistency was scaled using scores 1 and 2, 1 indicating consistent and 2 indicating inconsistent. The percentage consistency was calculated. No statistical analysis was performed on consistency. The time duration of each examination was recorded and the mean time values for CESM and BMRI were calculated and compared using statistical analysis. The cost of CESM and BMRI examinations was also reviewed using Medical Billing System Cooper University Hospital (Camden, NJ 08103, USA).

CESM protocols Sieve® 370 (Iopamidol injection 76%) from Bracco Diagnostics (Monroe Township, NJ 08831, USA) was used as CESM contrast. The admission dosage was calculated by patient’s weight at 1.5 cm3 /kg (maximum 100 cm3 ). The admission rate was 1.5—2 cm3 /s. The contrast was administered using a power injector. CESM was performed utilizing the Senographe Essential Full Field Digital Mammography system equipped with CESM (SenoBright® , GE, Chalfont StGiles, UK). Examination preparation, contraindications, and contrast administration followed the institutional standard body computer tomography (CT) protocol [26]. CESM images were acquired within ten minutes after intravenous administration of contrast. CESM high and low-energy images were obtained in CC and MLO projections of each breast.

3

CESM high-energy images were used to produce subtracted images. CESM high-energy images were not displayed for radiologist to review. CESM low-energy and subtracted images were reviewed by radiologists using GE Centricity Imagecast PACS 3.6.134.0 on NDS Dome E5 Grayscale 5 megapixel monitor.

BMRI protocol MRI contrast Multihance (gadobenate dimeglumine) was produced by Bracco Diagnostics (Monroe Township, NJ 08831, USA). The administration dosage is calculated by patient’s weight at 0.1 mmol/kg (maximum 20CC). The rate is at 1.5—2 cm3 /s. The contrast was administered using a power injector. BMRI studies were performed using a 1.5-T MR system (Symphony Vision, Siemens, Erlangen, Germany) and a dedicated seven-channel breast coil (Sentinelle Breast Coil by Invivo) after intravenously administration of contrast while patients were placed in prone position. The BMRI protocol included T1- and T2-weighted pre-contrast fat suppressed images in the transverse plane; dynamic imaging included one pre-contrast and 4 cycles of post-contrast imaging, compliant with ACR-recommendations. Reconstructed sagittal post-contrast images were also obtained [27].

Statistical analysis Paired t-test was used on applicable data to determine statistically significant differences in mean size of lesions, mean score of enhancement intensity of breast background and lesions, and mean examination duration of CESM and BMRI. PPV, sensitivity, and accuracy were also calculated in order to see which diagnostic tool was more effective. P values < 0.05 were considered statistically significant.

Results The study population consisted of 48 patients and 50 breasts (two patients had bilateral lesions). All of the 48 patients were female and the mean age was 56 years ± 10.6 (SD) (range: 35—76 years). There were 64 suspicious lesions identified by CESM and all 64 lesions were detected on BMRI (Table 1). BMRI identified two additional lesions, which were not seen on CESM. These two lesions were proven to be benign by pathology (Table 2). There were 62 true malignant lesions proven by pathology. Histopathology types of the tumor include ductal carcinoma in situ (DCIS), invasive ductal carcinoma (IDC), infiltrating lobular carcinoma (ILC), invasive tubular carcinoma (ITC), papillary carcinoma, poorly differentiated carcinoma, and fibrocystic changes. Pathology results are shown in Table 2. All 62 malignant lesions were identified on both CESM and BMRI. Sensitivity for CESM and BMRI was 100% (Table 1). CESM was found to have a higher PPV and a lower false positive rate compared to BMRI. The false positive rate for CESM was 3.1% (2/64) versus 6.1% (4/66) for BMRI. CESM had a PPV of 97% (62/64) while PPV for BMRI was 94% (62/66); the difference was not statistically significant (P = 0.62). CESM has a

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

4

L. Li et al. Table 1 Overview of the diagnostic performance of contrast-enhanced spectral mammography (CESM) and breast magnetic resonance imaging (BMRI). Diagnostic test parameter

CESM

BMRI

No. of selected patients No. of selected breasts No. of detected lesions No. of malignant lesions (TP) No. of benign lesions (FP) Sensitivity [95% CI] PPV [95% CI] Accuracy [95% CI] Size of index lesions: mean ± SD (mm) Lesion enhancement: mean ± SD Background enhancement: mean ± SD Size of smallest lesion detected (mm) Consistency on morphology for malignant lesions Test time: mean ± SD (min) Cost of each procedure (US dollar)

48 50 64 62 2 100% [95%, 100%] 96.9% [90%, 99.5%] 96.9% [90%, 99.5%] 33.6 ± 28.2 1.7 ± 0.76 1.0 ± 1.08 4 100% 10.2 ± 2.3 200

48 50 66 62 4 100% [95%, 100%] 93.9% [86%, 98%] 93.9% [86%, 98%] 34.7 ± 28.7 2.3 ± 0.70 1.4 ± 1.13 4 25.3 ± 5.1 1400

P value

1.0 0.62 0.62 0.108 < 0.01 < 0.01 N/A < 0.01

CESM: contrast-enhanced spectral mammography; BMRI: breast MRI; TP: true positive; FP: false positive; PPV: positive predictive value; CI: confidence interval; SD: standard deviation; N/A: not applicable.

slightly higher level of accuracy than BMRI, 97% vs. 94%, the difference had statistical significance (P = 0.62). The average size of index lesions was measured on CESM and BMRI. All index lesions had pathology diagnosis of breast

cancer (Table 2). The mean size on CESM was 33.6 mm ± 28.2 (SD) compared to 34.7 mm ± 28.7 (SD) on BMRI (P = 0.108). The smallest lesion detected on both CESM and BMRI was 4 mm (Fig. 1, Table 1).

Figure 1. A 50-year-old woman has had a CESM for high risk screening. A. CESM low-energy right craniocaudal (CC) view: The right breast is dense. No suspicious findings. B. CESM subtracted right CC: a 4 mm enhancing focus was identified only on CC view of CESM subtraction image (arrow). This was proven to be DCIS via an ultrasound-guided needle biopsy. C. CESM low-energy right mediolateral oblique (MLO) view: dense breast without suspicious finding. D. CESM subtracted right MLO: incidentally, a suspicious non-lesion enhancement was seen in the right posterior central breast on CESM subtracted image. This was not seen on C. E. BMRI axial post-contrast: both findings were demonstrated on BMRI: the arrow indicates known DCIS (anterior focus on the axial image). F. BMRI reconstructed sagittal post-contrast: The circle indicated a second lesion. The posterior second lesion was proven DCIS by MRI core needle biopsy. The size measurement of the first DCIS (arrows on both B and E) identified both CESM and BMRI was 4 mm. The morphology is consistent on both CESM and BMRI. Note that both lesions demonstrated lower intensity on CESM than on BMRI (mild on CESM and moderate on BMRI).

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

Contrast-enhanced spectral mammography versus breast magnetic resonance imaging The intensity of lesion enhancement on CESM was significantly lower compared to BMRI (P < 0.01), with the mean score of lesion enhancement intensity being 1.7 ± 0.76 (SD) for CESM and 2.3 ± 0.70 (SD) for BMRI (Fig. 1, Table 1). The enhancement intensity of malignant lesions varied from mild to marked on CESM. Eighty-seven percent of malignant lesions demonstrated moderate (Fig. 2) to marked enhancement on CESM (Fig. 3). Only 13% showed mild

5

enhancement, mostly on early stage low-grade breast cancer (Figs. 1 and 4). Note in Fig. 1, the size measurement of the first DCIS (arrows on both images B and E) identified on both CESM and BMRI was 4 mm. The morphology is consistent on both CESM and BMRI. Note that both lesions demonstrated lower intensity on CESM than on BMRI (mild on CESM and moderate on BMRI).

Figure 2. A 67-year-old woman underwent CESM and MRI for staging after ultrasound biopsy proven infiltrating ductal carcinoma of the right medial breast. A. CESM low-energy right CC: heterogeneously dense breast parenchyma with no highly suspicions findings. B. CESM subtracted right CC: a 2.4 cm low intensity non-mass enhancement in the medial right breast (circle). Note that there is only mild background enhancement although breast density is heterogeneously dense. C. BMRI reconstructed right sagittal post-contrast: a similar morphology and size of lesion (circle) compared to CESM. Note that unlike CESM, there is a marked background enhancement on BMRI. Marked BPE limits visualization of abnormal enhancement of malignant lesion on BMRI.

Figure 3. A 47-year-old woman exhibited a large irregular enhancing lesion in the inferior central right breast. A. CESM low-energy right CC. B. CESM subtracted right CC: a 3.5 × 2.9 cm speculated strong enhancing index lesion, consistent with diagnosis of high grade IDC. There is a second 1 cm enhancing lesion (boxes) anterior laterally to the index lesion. The second lesion was proven to be a satellite IDC. C and D. BMRI right axial and reconstructed sagittal post-contrast: both lesions were identified with the same morphology and strong enhancement.

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

6

L. Li et al.

Figure 4. A 69-year-old woman had conventional mammogram screening which showed abnormal finding. A and C. CESM low-energy left CC and MLO: a round lesion with obscured margins in the left lateral central breast posteriorly at 3:00 (arrows). B and D. CESM subtracted left CC and MLO: extremely mild enhancement of same round lesion with indistinct margins is at 3:00 O’clock (arrows). The morphology mimics a benign lesion. Core biopsy proved IDC. Note that the lesion can be identified on CESM subtracted images due to no BPE.

The intensity of background parenchymal enhancement was statistically significantly lower on CESM than on BMRI (1.0 versus 1.4; P < 0.01) (Table 1). The morphology of enhancing malignant lesions was 100% consistent between CESM and BMRI (Table 1). The imaging characteristics of enhancing malignant lesions on CESM

and BMRI were similar for both mass (Figs. 1, 5 and 6) and non-mass (Figs. 7 and 8) enhancement. A non-mass enhancement was commonly seen with DCIS and lobular carcinoma (Figs. 7 and 9). CESM was much less inexpensive with a shorter examination time than BMRI, making it an affordable and more

Figure 5. A 51-year-old woman with no family or personal history of breast cancer was called back for an asymmetry in the right medial breast on her screening mammogram. A. FFDM: right CC: circle indicates the area of asymmetry. B. CESM subtracted right CC: there was no suspicious enhancement in the right medial breast in the area of mammographic concern. Incidentally noted was a 1.3 cm avidly enhancing lesion in the lateral subareolar region (arrow). Ultrasound-guided core needle biopsy demonstrated invasive lobular carcinoma. Even retrospectively, it is difficult to identify the known malignancy on conventional mammogram. C. BMRI MIP axial post-contrast: similar size and morphology of the known malignancy compared to CESM (arrow).

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

Contrast-enhanced spectral mammography versus breast magnetic resonance imaging

7

Figure 6. A 68-year-old woman was diagnosed with bilateral breast cancer. A. FFDM left CC: an asymmetry was identified in the left lateral breast (box). B. CESM subtracted left CC: a 1.2 cm corresponding enhancing lesion (box). CESM incidentally showed an additional area of abnormal enhancement in the left medial breast (arrow), with a vague area of architectural distortion retrospectively identified in the same area on conventional mammography (A). The finding was suspicious for multicentric disease. Biopsy proved IDC of the medical breast lesion (C) BMRI reconstructed left sagittal post-contrast: Two malignant lesions were identified. A similar morphology of both lesions was seen on CESM when compared to BMRI (a mass and a distortion). BMRI demonstrated a stronger intensity on lesion enhancement than that on CESM. D and E. Ultrasound of both lesions: an irregular mass at 2:00 corresponding to the lateral mass on CESM and BMRI (boxes in A—C); an irregular mass at 11:00 representing the medial lesion on CESM and BMRI (arrows in B, C). F. The corresponding pathology showed DCIS. G. The biopsy at 11:00 revealed IDC.

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

8

L. Li et al.

Figure 7. A 63-year-old woman had microcalcifications on her right digital diagnostic mammogram. A and B. CESM subtracted left CC and MLO: a non-mass mild regional nodular enhancement in the right upper breast, which was pathology, proved for DCIS. C. BMRI axial MIP construction image: Similar enhancement is shown on BMRI. Note that the enhancement intensity is much higher on BMRI than CESM.

Table 2

Histopathological results in 66 breast lesions.

Lesion types

Number of lesions

Ductal carcinoma in situ (DCIS) Invasive ductal carcinoma (IDC) IDC + DCIS Infiltrating lobular carcinoma (ILC) Invasive tubular carcinoma (ITC) Invasive mammary carcinoma (IMC) Papillary carcinoma Poorly differentiated carcinoma Fibrocystic changes Total

5 26 10 11 2 5 2 1 4 66

accessible alternative. According to the Medical Billing at Cooper University Hospital, the cost of one CESM procedure and materials was approximately $200 versus $1400 for one BMRI. CESM has significantly shorter examination time than BMRI. On average, CESM took 10 minutes to complete versus 20 minutes needed for BMRI (P < 0.01) (Table 1).

Discussion BMRI is currently the gold standard for breast cancer detection and staging but is limited by low specificity [13—15], high cost, long duration of examination time, and limited availability. CESM is a relatively new imaging modality that provides both anatomic and functional information of breast lesions similar to MRI at a lower cost and examination duration.

Limited studies have shown that CESM sensitivity is similar to BMRI, and the present study supported this conclusion. CESM has a comparable sensitivity to BMRI on detecting invasive breast cancer that demonstrates enhancement. CESM also allows for detection of other abnormalities identified on the low-energy images (similar as digital mammography). The smallest enhancing lesion detected on post-contract images was a 4 mm DCIS. This lesion was identified on both CESM and BMRI. In addition to detecting contract enhancing lesions, the findings that could be identified on low-energy images on CESM include microcalcifications, enhancing focus, architectural distortion or spiculated lesion. These findings might be missed on both subtracted post-contrast CESM and BMRI when there is no significant lesion enhancement. This allows CESM to detect additional low-grade or in situ tumors, though more research needs to be conducted. A reported prospective study showed that CESM has a higher specificity when comparison was made to BMRI [21]. Our study supports this finding. The present study indicates that CESM may have a higher specificity than BRMI, reflected by a higher PPV. The PPV on CESM is higher than PPV on BMRI. The difference is not statistically significant. A relative small sample size might be a contributing factor to non-significance result on PPV. More research conducted on larger sample size is needed to determine PPV and specificity of CESM. The other factor which could contribute to a higher specificity on CESM compared to BMRI is that BPE on CESM was significant lower than on BMRI. With a low BPE on CESM, the current study demonstrates a lower false positive rate on CESM compared to BMRI. The lack of significance of the difference when data was analyzed statistically may be attributed to the relative small sample size. More research

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

Contrast-enhanced spectral mammography versus breast magnetic resonance imaging

9

Figure 8. A 50-year-old woman underwent CESM for further evaluation of a right breast density on screening mammogram. A. CESM lowenergy right CC: no suspicious microcalcifications. Retrospectively a small density can be appreciated in medial breast. B. CESM subtracted right CC: a minimally enhance nodule in the lower inner right breast (arrow). Biopsy revealed DCIS. C. BMRI reconstructed right sagittal post-contrast: tow lesions were identified: the superior (12:00) enhancing lesion (circle) was not seen on CESM. This lesion was proven to be a benign finding (MRI false positive). The diagnosis was fibrocystic change with focus of stromal calcification and usual ductal hyperplasia. The second lesion in inferior right breast (arrow) corresponds to CESM DCIS. The focus of DCIS demonstrates the same morphology but higher intensity of enhancement than shown on CESM. D. Pathology (× 100): a benign (false positive) finding-fibrocystic change with focus of stromal calcification Usual ductal hyperplasia corresponded to the lesion only seen on BMRI (circle on C). E. Pathology (× 100): tissue biopsy of the inferior (6:00) lesion exhibited DCIS (100 ×) corresponded to the lesion found on CESM (arrow on A and B), and the inferior lesion on BMRI (arrow on C) with the same morphology but higher intensity of enhancement than shown on CESM.

using large sample size is needed for more definitive assessment of CESM specificity. Dromain et al. found that the size of the malignant enhancing lesions correlated with the histological size of the cancer at surgery [16,19]. In the current study, a large number of patients received neoadjuvant treatment before mastectomy or lumpectomy, so final pathology results did not reflect the true initial size of the tumor. Therefore, size comparison was made between CESM and BMRI, not made with size at surgical pathology. Our study found that index lesion size was slightly larger on BMRI compared to CESM (P < 0.05). It is important to note that 95% of CESM was performed prior to biopsy, while 80% BMRI was performed

after initial needle biopsy (pre-certification constraint). Post-biopsy changes on BMRI likely accounts for some overestimation of lesion size, thought actual size of lesion at gross pathology was unable to be determined. There is a paucity of published research in CESM morphology. Using BMRI guideline for lesion morphology assessment, our study indicates that CESM has 100% consistency with BMRI on morphology features of malignant lesions. More research is needed on this topic. CEMS overcomes many of the limitations of BRMI. Based on the Medical Billing at Cooper University Hospital, the cost of one CESM is much less expensive than one BMRI. Our study indicates that CESM has significant shorter exam

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

10

L. Li et al.

Figure 9. A 46-year-old f woman underwent CESM for abnormal finding in the right breast. A and C. CESM low-energy right MLO and CC: a 3 cm clumped area of microcalcifications is in the inferior medial breast (circles). There is a second cluster of microcalcifications located posterior to ILC (box in C). B and D. CESM subtracted right MLO and CC: a 3 cm abnormal non-mass enhancement associated with suspicious microcalcifications located in anterior aspect of the right breast (circles). Biopsy was performed and this was invasive lobular carcinoma (ILC). No suspicious enhancement was identified on the poster cluster of microcalcifications. E. FFDM magnification right CC: the posterior microcalcifications demonstrated LCIS by tissue biopsy (box in C and E).

duration than BMRI so that more study slots could be offered to patients. CESM does not require a pre-certification process particularly before a cancer diagnosis is made, thus making CESM more accessible. CESM can be offered to patients who have contraindications to BMRI (pacemakers, claustrophobia, etc.). Overall, CESM appears to be a more affordable, efficient and more accessible alternative to BRMI. CESM can be easily incorporated into daily workflow as a next step for diagnosis and when conventional digital mammography and ultrasound are inconclusive for suspicious findings. The methodological limitations of the current study include being a retrospective study, relatively small sample size and lack of surgical pathology size for comparison. Due to these limitations, true specificity was unable to be determined. More research needs to be conducted on larger sample size to determine diagnostic performance of CESM. In conclusion, CESM was found to have similar sensitivity for breast cancer detection than BMRI, with higher PPV and less background enhancement. CESM is associated with significantly shorter exam time thus a more accessible alternative to MRI, and has the potential to be an important tool in breast cancer detection and staging.

Disclosure of interest The authors declare that they have no competing interest.

References [1] Loberg M, Lousdal ML, Bretthauer M, Kalager M. Benefits and harms of mammography screening. Breast Cancer Res 2015;17:63. [2] Shetty MK. Screening for breast cancer with mammography: current status and an overview. Indian J Surg Oncol 2010;1:218—23.

[3] Boyd NF, Guo H, Martin LJ, Sun L, Stone J, Fishell E, et al. Mammographic density and the risk and detection of breast cancer. N Engl J Med 2007;356:227—36. [4] Mori M, Akashi-Tanaka S, Suzuki S, Daniels MI, Watanabe C, Hirose M, et al. Diagnostic accuracy of contrastenhanced spectral mammography in comparison to conventional full-field digital mammography in a population of women with dense breasts. Breast Cancer 2016, http://dx.doi.org/10.1007/s12282-016-0681-8. [5] Kolb TM, Lichy J, Newhouse JH. Comparison of the performance of screening mammography, physical examination, and breast US and evaluation of factors that influence them: an analysis of 27,825 patient evaluations. Radiology 2002;225:165—75. [6] Leach MO, Boggis CR, Dixon AK, Easton DF, Eeles RA, Evans DG, et al. Screening with magnetic resonance imaging and mammography of a UK population at high familial risk of breast cancer: a prospective multicentre cohort study (MARIBS). Lancet 2005;365:1769—78. [7] Pisano ED, Gatsonis C, Hendrick E, Yaffe M, Baum JK, Acharyya S, et al. Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 2005;353:1773—83. [8] Schell AM, Rosenkranz K, Lewis PJ. Role of breast MRI in the preoperative evaluation of patients with newly diagnosed breast cancer. AJR Am J Roentgenol 2009;192:1438—44. [9] Berg WA. Rationale for a trial of screening breast ultrasound: American College of Radiology Imaging Network (ACRIN) 6666. AJR Am J Roentgenol 2003;180:1225—8. [10] Morris EA, Liberman L, Ballon DJ, Robson M, Abramson AF, Heerdt A, et al. MRI of occult breast carcinoma in a high-risk population. AJR Am J Roentgenol 2003;181:619—26. [11] Kim HR, Jung HK, Ko KH, Kim SJ, Lee KS. Mammography, US, and MRI for preoperative prediction of extensive intraductal component of invasive breast cancer: interobserver variability and performances. Clin Breast Cancer 2016;16:305—11. [12] Choi JS, Han BK, Ko EY, Ko ES, Hahn SY, Shin JH, et al. Comparison between two-dimensional synthetic mammography reconstructed from digital breast tomosynthesis and full-field digital mammography for the detection of T1 breast cancer. Eur Radiol 2016;26:2538—46. [13] Houssami N, Ciatto S, Macaskill P, Lord SJ, Warren RM, Dixon JM, et al. Accuracy and surgical impact of magnetic resonance imaging in breast cancer staging: systematic review and

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013

+Model DIII-854; No. of Pages 11

ARTICLE IN PRESS

Contrast-enhanced spectral mammography versus breast magnetic resonance imaging

[14]

[15]

[16]

[17]

[18]

[19]

[20]

meta-analysis in detection of multifocal and multicentric cancer. J Clin Oncol 2008;26:3248—58. Roganovic D, Djilas D, Vujnovic S, Pavic D, Stojanov D. Breast MRI, digital mammography and breast tomosynthesis: comparison of three methods for early detection of breast cancer. Bosn J Basic Med Sci 2015;15:64—8. Nekhlyudov L, Kiarsis K, Elmore JG. MRI of the breast: does the internet accurately report its beneficial uses and limitations? Breast J 2009;15:189—93. Dromain C, Balleyguier C, Muller S, Mathieu MC, Rochard F, Opolon P, et al. Evaluation of tumor angiogenesis of breast carcinoma using contrast-enhanced digital mammography. AJR Am J Roentgenol 2006;187:W528—37. Dromain C, Balleyguier C, Adler G, Garbay JR, Delaloge S. Contrast-enhanced digital mammography. Eur J Radiol 2009;69:34—42. Dromain C, Thibault F, Muller S, Rimareix F, Delaloge S, Tardivon A, et al. Dual-energy contrast-enhanced digital mammography: initial clinical results. Eur Radiol 2011;21:565—74. Fallenberg EM, Dromain C, Diekmann F, Engelken F, Krohn M, Singh JM, et al. Contrast-enhanced spectral mammography versus MRI: initial results in the detection of breast cancer and assessment of tumour size. Eur Radiol 2014;24:256—64. Lewin JM, Isaacs PK, Vance V, Larke FJ. Dual-energy contrast-enhanced digital subtraction mammography: feasibility. Radiology 2003;229:261—8.

11

[21] Jochelson MS, Dershaw DD, Sung JS, Heerdt AS, Thornton C, Moskowitz CS, et al. Bilateral contrast-enhanced dual-energy digital mammography: feasibility and comparison with conventional digital mammography and MR imaging in women with known breast carcinoma. Radiology 2013;266:743—51. [22] Badr S, Laurent N, Regis C, Boulanger L, Lemaille S, Poncelet E. Dual-energy contrast-enhanced digital mammography in routine clinical practice in 2013. Diagn Interv Imaging 2014;95:245—58. [23] Langlotz CP. ACR BI-RADS for breast imaging communication: a roadmap for the rest of radiology. J Am Coll Radiol 2009;6:861—3. [24] Ikeda DM. Updated breast MRI Lexicon. Eur J Radiol 2012;81:S63. [25] El Khoury M, Lalonde L, David J, Labelle M, Mesurolle B, Trop I. Breast imaging reporting and data system (BI-RADS) lexicon for breast MRI: interobserver variability in the description and assignment of BI-RADS category. Eur J Radiol 2015;84:71—6. [26] Antoch G, Freudenberg LS, Stattaus J, Jentzen W, Mueller SP, Debatin JF, et al. Whole-body positron emission tomographyCT: optimized CT using oral and IV contrast materials. AJR Am J Roentgenol 2002;179:1555—60. [27] Giess CS, Yeh ED, Raza S, Birdwell RL. Background parenchymal enhancement at breast MR imaging: normal patterns, diagnostic challenges, and potential for false-positive and false-negative interpretation. Radiographics 2014;34:234—47.

Please cite this article in press as: Li L, et al. Contrast-enhanced spectral mammography (CESM) versus breast magnetic resonance imaging (MRI): A retrospective comparison in 66 breast lesions. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.013