- Email: [email protected]
Use of Contrast-Enhanced Spectral Mammography for Intramammary Cancer Staging
Original Investigations
Use of Contrast-Enhanced Spectral Mammography for Intramammary Cancer Staging: Preliminary Results Katrin S. Blum, MD, Christian Rubbert, MD, Britta Mathys, MD, Gerald Antoch, MD, Svjetlana Mohrmann, MD, Silvia Obenauer, MD Rationale and Objectives: To prospectively evaluate and compare the accuracy of contrast-enhanced spectral mammography (CESM) and ultrasound (US) in size measurement of breast cancer with histologic tumor sizes as gold standard. Materials and Methods: Twenty women aged between 40–73 years (mean age, 57 10 years) with histologically proven invasive ductal/ lobular carcinomas were included in the study. Agreement between imaging tumor size (CESM and US) and histopathologic tumor size was evaluated with Bland–Altman analysis. Stereotactically guided vacuum biopsy was performed in four patients after CESM. Two independent reviewers described artifacts of CESM. Results: Motion artifacts did not occur in the study. CESM-specific artifacts caused by scattered radiation mostly occurred in oblique view of CESM. Background enhancement of breast tissue was seen in four patients. Mean difference of tumor sizes was 0.3 mm (6.34%) between CESM and histology and 2.2 mm ( 7.59%) between US and histology. Limits of agreement ranged from 18.9 to 19.48 mm for CESM and from 17.1 to 12.7 mm with US. Especially smaller tumors with a size <23 mm were measured more precisely with CESM. Enhancement of breast tissue around microcalcifications correlated with abnormalities. Conclusions: CESM is accurate in size measurements of small breast tumors. On average CESM leads to a slight overestimation of tumor size, whereas US tends to underestimate tumor size. Assessment of the breast tissue can be limited by the scattered radiation artifact and background enhancement of breast tissue. CESM seems to be helpful in the characterization of breast tissue around microcalcifications. Key Words: Contrast-enhanced spectral mammography; tumor staging; ultrasound; artifacts. ªAUR, 2014
C
ontrast-enhanced spectral mammography (CESM) is an imaging technique combining digital mammography with intravenous injection of iodinated contrast media to detect hypervascularized lesions, especially in dense breast tissue (1). Although it was developed several years ago, knowledge about the performance of this technique in clinical routine, especially in breast cancer screening, is still limited. However, a recent study proposes similar indications for CESM as for magnetic resonance imaging (MRI), namely preoperative staging, detection of occult lesions, monitoring of treatment response (2). Initial results comparing this new technique with mammography, ultrasound (US) and breast
Acad Radiol 2014; 21:1363–1369 From the Department of Diagnostic and Interventional Radiology, Medical Faculty, University Dusseldorf, Moorenstr. 5, 40225 Dusseldorf, Germany (K.S.B., C.R., B.M., G.A., S.O.); and Department of Gynecology, Medical Faculty, University Dusseldorf, Dusseldorf, Germany (S.M.). Received May 20, 2014; accepted June 24, 2014. Address correspondence to: K.S.B. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2014.06.012
MRI show a better detection of suspicious lesions with CESM compared to full-field digital mammography and the combined imaging of mammography and US (3) but a lower detection rate of hypervascularized breast lesions compared to breast MRI (4). Studies evaluating the accuracy of CESM compared to MRI in preoperative tumor staging show a similar accuracy in lesion size measurement compared to MRI (5). Preoperative staging of cancer extent in the breast is necessary to plan the optimal treatment (6). MRI is the most commonly used approach to determine the extent of the tumor in the breast and to decide which surgery should be performed and if the breast should be radiated (7,8). Besides its high cost and limited availability, one major problem of breast MRI is background enhancement of breast tissue, which decreases the detection of breast lesions and affects breast cancer staging (9). In such patients, US is a good alternative method for breast cancer staging. Studies correlating tumor size determined with imaging and histopathology describe a tendency to underestimate tumor size with US (10–13), whereas MRI tends to overestimate tumor size 1363
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
TABLE 1. Details of Patients Included in the Study
Number of Patients
Age (Mean Age; Standard Deviation)
20
40–73 (57; 10)
ACR (Artifacts) Disease Invasive cancer Invasive cancer with EIC
ACR2 2 4 (1)
ACR3 4 6
ACR 4 1 3 (2)
The majority of patients had invasive cancer with additional extensive in-situ component (EIC).
(13,14). The good performance of CESM in the detection of multifocal and multicentric cancer dissemination has been shown (4), as well as the high accuracy of tumor size measurements with CESM compared to MRI (5). The performance of CESM in preoperative tumor size measurement compared to US has not been investigated so far. For CESM a technical artifact has been reported which is called ‘‘scattered radiation artifact’’ caused by differences in breast tissue thickness from the thorax to the edge of the breast and different characteristics of mammography with low and high energy (3). Motion artifacts or background enhancement of the breast parenchyma interfering with image quality of CESM has not been reported to date. Therefore, the purpose of this study was to compare the accuracy of local tumor staging with CESM compared to US, with histopathology serving as the gold standard. In addition, CESM was assessed for artifacts potentially limiting interpretation of the CESM data set. In four patients, a proof of principle was performed to show that CESM imaging before stereotactically guided vacuum-assisted core biopsy is feasible.
MATERIAL AND METHODS The local ethics committee approved this prospective study, which was performed in accordance with the Declaration of Helsinki. Written consent of all patients was obtained before enrollment of the study.
Breast density was classified with low-energy mammography of CESM by the standardized classification system of the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS).
CESM Imaging
A Senographe Essential CESM (GE Healthcare, Solingen, Germany) was used for CESM examinations. Dual-energy CESM was performed with automated parameters acquiring a low- and high-energy mammography during one breast compression. Molybdenum (Mo) or rhodium (Rh) target and Mo or Rh filter were used depending on the breast density and compression thickness. Peak kilovoltage (kVp) values ranging from 26 to 31 were used for the acquisition of low-energy mammography, whereas high-energy mammography of CESM was acquired with 45–49 kVp, with x-ray spectrum above the k-edge of iodine (33.2 keV). Solutrast 300 (Bracco Imaging Germany, Konstanz, Germany) was used as contrast media with a weight-adapted dose of 1.5 mL/kg body weight with a flow rate of 3 mL/s, followed by an injection of 20-mL sodium chloride. Mediolateral (MLO) view and craniocaudal (CC) view of the nonaffected breast were acquired 2 minutes after the initiation of contrast media application, followed by MLO- and CC-CESM of the affected breast. The Senographe Essential CESM generated iodine-enhanced images from the lowand high-energy images. Regions of contrast media uptake were displayed with recombined image-by-image processing.
Patients Ultrasound
Women with histologically proven invasive breast cancer, older than 30 years, were enrolled in the study between February and October 2013. Carcinoma was diagnosed with mammography or US and histologically proven by biopsy, before CESM. CESM was clinically indicated for determination of tumor size. Additional MRI was performed if CESM results were not clear (n = 4). Patients with impaired renal function, pregnant or breastfeeding patients, patients with hyperthyroidism, and patients with a history of anaphylactic reaction to contrast media were excluded. CESM was not performed at the same day with other investigations in which application of contrast media was necessary (MRI and computed tomography). Neoadjuvant chemotherapy was an exclusion criterion, as well as hormone treatment or radiotherapy of the breast. 1364
An Aixplorer Supersonic ultrasound system (Supersonic Imaging, Munich, Germany) or Toshiba Aplio MX (Toshiba Medical Systems, Neuss, Germany) were used for second look US before biopsy or surgery. Three gynecologists with 5–12 years of experience in US investigated and staged both breasts. Measurements were performed in all three axes. The largest tumor diameter, which was defined in consensus, was recorded to compare lesion size with histopathologic lesion size. Investigators were blinded to measurements in CESM. Image Analysis
Contrast-Enhanced Spectral Mammography. Anonymized images were evaluated with a high-resolution workstation
Academic Radiology, Vol 21, No 11, November 2014
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
Figure 1. Low-energy mammography (a) and processed CESM (b) oblique mammography (RMLO) and craniocaudal view (RCC; c and d) of the right breast of a woman with invasive ductal breast cancer. Scattered radiation artifact is marked with arrows, appears more prominent in RMLO view, and does not affect tumor size measurement. CESM, contrast-enhanced spectral mammography; RCC, right craniocaudal; RMLO, right mediolateral.
(Barco Inc, Duluth, GA) on completion of each examination by two radiologists with 15 and 3 years of experience in reading mammographies and breast MRI and basic training in reading CESM images. Both readers were blinded to US results and past medical history. Breast density and microcalcifications were recorded by evaluation of the low-energy CESM image by the radiologists, and lesion size, artifacts, and potential additional lesions were evaluated in processed CESM image using the BI-RADS classification. Consensus-based measurements of tumor size was performed in three axes, and the maximum diameter was recorded for a later comparison with US and histopathology.
Statistical Analysis
All calculations were performed using R v3.0.2. Bland–Altman analysis was used to compare differences in lesion size measurements with CESM and US and, furthermore, to correlate tumor size measurements between CESM and the
gold standard (15). Mean differences between tumor size measurements with imaging and histopathologic results were calculated and related to the limits of agreement (LOA), which is the interval in which 95% of the calculated differences were found. The influence of breast density and additional extensive in-situ component (EIC) component on lesion size measurements of invasive cancer on was evaluated as subanalysis.
RESULTS CESM was performed in 22 patients with invasive breast cancer. Twenty patients aged between 40–73 years were included in the study, and two patients were excluded because of neoadjuvant chemotherapy before surgery. Six patients had scattered fibroglandular breast densities (ACR2), heterogeneously dense breast tissue (ACR3) was reported in ten patients, and four patients had extremely dense breast tissue (ACR4) as displayed in Table 1. Four women underwent 1365
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
Figure 2. Processed contrast-enhanced spectral mammography in (a) craniocaudal (LCC) and (b) oblique view (LMLO) in a woman with invasive ductal breast cancer with additional extensive intraductal component. The arrows mark the tumor; asterisks (*) mark enhancing foci retroareolar. Additional non-mass enhancement medial in the breast (**) is caused most likely because of hormonal proliferation. LCC, left craniocaudal; LMLO, left mediolateral.
Figure 3. (a) Low-energy mammography of contrast-enhanced spectral mammography (CESM) of a woman with histological proven cancer in the upper part of the breast with suspicious microcalcifications (*). Additional suspicious microcalcifications are detected in the anterior third of breast tissue in the upper part of the breast (arrow in the magnification of a in b). Processed CESM image (c) with enhancement of the tumor and around the additional microcalcifications (arrow in the magnification of c in d). (e) Stereotactic-guided vacuum biopsy of the additional microcalcification, which turned out to be a ductal carcinoma in situ.
stereotactic-guided vacuum-assisted core biopsy of the breast within 1 hour after CESM. Seven patients had invasive cancer and 13 patients had invasive cancer with an additional EIC (Table 1). One patient had invasive lobular carcinoma with an additional EIC component; twelve patients had invasive ductal carcinoma together with EIC. In two patients with invasive cancer, histopathologic analysis showed invasive lobular carcinoma, and five patients had an invasive ductal carcinoma.
1366
Artifacts
Motion artifacts affected none of the CESM mammographies, neither the low-energy nor the high-energy CESM. The previously described scattered radiation artifact (Fig 1), which describes brighter areas near the edges of the breast and is explained by scattered radiation, was observed in all breasts but did not lead to misinterpretation of the CESM image. This artifact was more prominent on oblique views (Fig 1b) compared to the CC view (Fig 1d).
Academic Radiology, Vol 21, No 11, November 2014
TABLE 2. Histopathologic Results of Four Patients with Suspicious Microcalcifications, Which Received CESM before Stereotactic-Guided Vacuum Biopsy
Patients Primary Cancer Type Invasive lobular cancer
Enhancement Around Microcalcifications N
N
Invasive ductal cancer
N
Invasive ductal cancer
Y
N
Invasive ductal cancer
Y
Histopathologic Finding Fibrocystic mastopathy, ductal hyperplasia, and sclerosing adenosis Fibrocystic mastopathy, ductal hyperplasia, and sclerosing adenosis Fibrocystic mastopathy, ductal hyperplasia, and chronic inflammation Atypical lobular hyperplasia (ALH)* Fibrocystic mastopathy and ductal hyperplasia DCIS*
CESM, contrast-enhanced spectral mammography; DCIS, ductal carcinoma in situ. CESM showed enhancement surrounding microcalcifications in two patients, one with DCIS and one with ALH. Both lesions were excised. *Histologic result led to therapy changes.
Background Parenchymal Enhancement of Breast Tissue
Focal enhancement or additional non-mass enhancement of the breast tissue most likely due to hormone proliferation of the breast tissue, similar to focal or non-mass enhancement of breast tissue on MRI, was seen in breasts of three patients (Table 1). Homogenous enhancement of the breast tissue occurred in one woman with fibroglandular density of the breast (ACR2), aged 58 years and two patients with extremely dense breast tissue (ACR4) with age of 42 and 43 years (Fig 2). CESM Before Stereotactically Guided VacuumAssisted Core Biopsy of the Breast
Four women underwent CESM before stereotactic-guided vacuum-assisted core biopsy of the breast to assess the extent
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
of histologically proven tumor and detect potential enhancement of additional suspect microcalcifications. Six microcalcifications detected with mammography were biopsied; two patients underwent bilateral biopsy. In two patients, enhancement of the breast tissue around microcalcifications was detected with CESM (Fig 3). Histopathologic analysis of biopsied microcalcifications with enhancement of the surrounding breast tissue was either atypical lobular hyperplasia or ductal carcinoma in situ (DCIS), which led to a different surgical therapy (Table 2). Microcalcifications without enhancement of the surrounding breast tissue showed benign histopathologic results (Table 2). No Significant Influence of Additional EIC on Tumor Size Measurements of Invasive Breast Cancers
Histologically determined tumor size ranged between 9 and 40 mm with a mean tumor size of 23 mm (8.8 mm). The mean difference between sonographically and histopathologically determined tumor size was 2.2 mm (LOA, 17.1 to 12.7 mm; Table 3). As displayed in Table 3, additional EIC led to an underestimation of tumor size with US by 2.5 mm. In comparison, invasive cancer alone was underestimated by 1.6 mm. Student t test was performed, and all results were found to be nonsignificant (P > .05). The mean difference between tumor (invasive cancer and invasive cancer with EIC) size determined with CESM and histopathology was 0.3 mm. As shown in Table 3, additional EIC led to nonsignificant overestimation of tumor size with CESM by a mean of 0.8 mm, whereas invasive cancer without additional EIC was underestimated using CESM by 0.7 mm. An additional EIC component did not significantly influence tumor size measurements when assessed with CESM (Fig 4; Table 3). Good agreement of tumor size measurements with CESM and histology were seen in smaller tumors #22 mm (Fig 4). LOA for tumor size measurements with CESM was with 18.9 and 19.5 mm much wider than measurements with US (LOA, 17.1 to 12.7 mm). The standard deviation of the mean difference was 9.6 mm for CESM and 7.5 mm for US showing a lower precision of measurements with CESM compared to US. Thus, lesion size variability as compared to the reference standard was lower for US than for CESM. However, US systematically underestimated lesion size as compared to the reference standard.
DISCUSSION This study demonstrates that CESM is accurate in determination of tumor size as compared to the reference standard and US although tumor size measurements with CESM may not be very precise. CESM is an imaging technique, which is more and more introduced into clinical routine. The duration of the examination is shorter; the detection rate of breast cancer is comparable to that of MRI. The sensitivity of detecting additional cancer of the breast is lower, compared to MRI, but 1367
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
TABLE 3. Comparison of the Differences Between Measured Tumor Size with Imaging and Histopathology Ultrasound Tumor Type Invasive cancer (n = 7) Invasive cancer + EIC (n = 13) Total (n = 20)
Mean (mm) 1.571 2.538 2.2
CESM LOA (mm) 19.48 to 14.4 12.82 to 9.67 17.1 to 12.7
Mean (mm) 0.714 0.846 0.3
LOA (mm) 11.15 to 9.72 22.06 to 23.75 18.88 to 19.48
CESM, contrast-enhanced spectral mammography; EIC, extensive in situ component; LOA, limits of agreement. t Test showed no significant under- or over-estimation of tumor size (P < .05).
Figure 4. Bland–Altman analysis of agreement of histologic tumor size with size measurements with contrast-enhanced spectral mammography (CESM) and ultrasound (US) in patients with invasive breast cancer with or without extensive intraductal component (EIC). High agreement of histologic tumor size and CESM measurements in patients with smaller tumors (#22 mm). Additional EIC component did not affect tumor size measurements in either way.
CESM has a higher specificity (4). Studies evaluating the accuracy of CESM in preoperative tumor staging show a similar accuracy in lesion size measurement compared to MRI (5). The present study is the first study comparing tumor size measurements of CESM with US in preoperative staging. Similar to other studies (13), we found an underestimation of tumor size with US, whereas tumor-sized measurement with CESM slightly overestimated the tumor size similar to measurements with MRI (14,16,17). Good agreement and precision of tumor size diameter measured with CESM and histopathologic results were seen in tumors with #22 mm size, similar to studies on MRI (18) and other studies using CESM (19). In comparison to US, CESM measured tumor size more accurately and was not affected by the density of the breast tissue, as described for conventional digital mammography (20). Similar to the study of Lobbes et al., the tumor size was highly overestimated in one patient (19). In our study, this was due to the misinterpretation of background enhancement as tumor tissue. 1368
Preoperative size assessment is important to decide the most beneficial surgical therapy for the patient. Although only a few patients were included in this study, CESM seems to be a good imaging method for preoperative tumor staging, as suggested by others (5). The similar slight overestimation of tumor size of CESM as described for MRI seems to be most likely caused by a different vascularization of the tumor, or precancerous surrounding breast tissue (21). Background enhancement of the breast tissue as described for MRI (9) seems to limit the evaluation of tumor size in CESM in a similar way. A disadvantage of CESM shown in this study is that the accuracy of tumor size measurements decreases with larger tumor size. Similar results have been reported for tumor size measurements with MRI or CESM (18,19). Lower accuracy of tumor size measurements with imaging in patients with large tumors leads to wide LOA, although the mean tumor size difference between gold standard and CESM stays very small (0.3 mm). This is displayed by the larger standard deviation of mean differences.
Academic Radiology, Vol 21, No 11, November 2014
A beneficial effect of CESM compared to MRI seems to be the evaluation of microcalcifications, which can be exactly correlated to enhancement of the breast tissue. Although we investigated only a few patients with CESM before stereotactically guided vacuum biopsy, we could show that the latter is not affected by CESM. Moreover, the present study suggests that CESM before stereotactically guided vacuum core biopsy may be beneficial to differentiate malignant from nonmalignant lesions by detection of contrast enhancement surrounding microcalcifications in malignancy. Our study has some limitations. The major limitation is the small amount of patients investigated in the study, which might be a reason for nonsignificant results. Because of the small study group, a subgroup analysis of different cancer types was not performed, which is the second limitation. Another limitation is that CESM was not performed with regard to the menstrual cycle, which might have affected the results in patients with background enhancement. In conclusion, we showed the feasibility of preoperative tumor staging of breast cancer with CESM. Albeit a small study group, we could show more accurate tumor size measurements with CESM compared to US. Similar artifacts as described for MRI, affecting tumor size measurements, were reported in the study for the first time. Preliminary results suggest a beneficial effect of CESM when differentiating benign from malignant microcalcifications of the breast. Further research is necessary and more patients have to be investigated to prove those preliminary results.
REFERENCES 1. Dromain C, Balleyguier C, Adler G, et al. Contrast-enhanced digital mammography. Eur J Radiol 2009; 69(1):34–42. 2. Lobbes MB, Smidt ML, Houwers J, et al. Contrast enhanced mammography: techniques, current results, and potential indications. Clin Radiol 2013; 68(9):935–944. 3. Dromain C, Thibault F, Muller S, et al. Dual-energy contrast-enhanced digital mammography: initial clinical results. Eur Radiol 2011; 21(3):565–574. 4. Jochelson MS, Dershaw DD, Sung JS, 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(3):743–751.
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
5. Fallenberg EM, Dromain C, Diekmann F, 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(1):256–264. 6. Litiere S, Werutsky G, Fentiman IS, et al. Breast conserving therapy versus mastectomy for stage I-II breast cancer: 20 year follow-up of the EORTC 10801 phase 3 randomised trial. Lancet Oncol 2012; 13(4):412–419. 7. Pengel KE, Loo CE, Teertstra HJ, et al. The impact of preoperative MRI on breast-conserving surgery of invasive cancer: a comparative cohort study. Breast Cancer Res Treat 2009; 116(1):161–169. 8. Tendulkar RD, Chellman-Jeffers M, Rybicki LA, et al. Preoperative breast magnetic resonance imaging in early breast cancer: implications for partial breast irradiation. Cancer 2009; 115(8):1621–1630. 9. Uematsu T, Kasami M, Watanabe J. Does the degree of background enhancement in breast MRI affect the detection and staging of breast cancer? Eur Radiol 2011; 21(11):2261–2267. 10. Hieken TJ, Harrison J, Herreros J, et al. Correlating sonography, mammography, and pathology in the assessment of breast cancer size. Am J Surg 2001; 182(4):351–354. 11. Finlayson CA, MacDermott TA. Ultrasound can estimate the pathologic size of infiltrating ductal carcinoma. Arch Surg 2000; 135(2):158–159. 12. Pritt B, Ashikaga T, Oppenheimer RG, et al. Influence of breast cancer histology on the relationship between ultrasound and pathology tumor size measurements. Mod Pathol 2004; 17(8):905–910. 13. Gruber IV, Rueckert M, Kagan KO, et al. Measurement of tumour size with mammography, sonography and magnetic resonance imaging as compared to histological tumour size in primary breast cancer. BMC Cancer 2013; 13:328. 14. Uematsu T, Yuen S, Kasami M, et al. Comparison of magnetic resonance imaging, multidetector row computed tomography, ultrasonography, and mammography for tumor extension of breast cancer. Breast Cancer Res Treat 2008; 112(3):461–474. 15. Lobbes MB, Nelemans PJ. Good correlation does not automatically imply good agreement: the trouble with comparing tumour size by breast MRI versus histopathology. Eur J Radiol 2013; 82(12):e906–e907. 16. Jiang YZ, Xia C, Peng WT, et al. Preoperative measurement of breast cancer overestimates tumor size compared to pathological measurement. PLoS ONE 2014; 9(1):e86676. 17. Schouten van der Velden AP, Boetes C, Bult P, et al. The value of magnetic resonance imaging in diagnosis and size assessment of in situ and small invasive breast carcinoma. Am J Surg 2006; 192(2):172–178. 18. Grimsby GM, Gray R, Dueck A, et al. Is there concordance of invasive breast cancer pathologic tumor size with magnetic resonance imaging? Am J Surg 2009; 198(4):500–504. 19. Lobbes MB, Lalji U, Houwers J, et al. Contrast-enhanced spectral mammography in patients referred from the breast cancer screening programme. Eur Radiol 2014; 24(7):1668–1676. 20. Fasching PA, Heusinger K, Loehberg CR, et al. Influence of mammographic density on the diagnostic accuracy of tumor size assessment and association with breast cancer tumor characteristics. Eur J Radiol 2006; 60(3):398–404. 21. Guinebretiere JM, Le Monique G, Gavoille A, et al. Angiogenesis and risk of breast cancer in women with fibrocystic disease. J Natl Cancer Inst 1994; 86(8):635–636.
1369
Use of Contrast-Enhanced Spectral Mammography for Intramammary Cancer Staging: Preliminary Results Katrin S. Blum, MD, Christian Rubbert, MD, Britta Mathys, MD, Gerald Antoch, MD, Svjetlana Mohrmann, MD, Silvia Obenauer, MD Rationale and Objectives: To prospectively evaluate and compare the accuracy of contrast-enhanced spectral mammography (CESM) and ultrasound (US) in size measurement of breast cancer with histologic tumor sizes as gold standard. Materials and Methods: Twenty women aged between 40–73 years (mean age, 57 10 years) with histologically proven invasive ductal/ lobular carcinomas were included in the study. Agreement between imaging tumor size (CESM and US) and histopathologic tumor size was evaluated with Bland–Altman analysis. Stereotactically guided vacuum biopsy was performed in four patients after CESM. Two independent reviewers described artifacts of CESM. Results: Motion artifacts did not occur in the study. CESM-specific artifacts caused by scattered radiation mostly occurred in oblique view of CESM. Background enhancement of breast tissue was seen in four patients. Mean difference of tumor sizes was 0.3 mm (6.34%) between CESM and histology and 2.2 mm ( 7.59%) between US and histology. Limits of agreement ranged from 18.9 to 19.48 mm for CESM and from 17.1 to 12.7 mm with US. Especially smaller tumors with a size <23 mm were measured more precisely with CESM. Enhancement of breast tissue around microcalcifications correlated with abnormalities. Conclusions: CESM is accurate in size measurements of small breast tumors. On average CESM leads to a slight overestimation of tumor size, whereas US tends to underestimate tumor size. Assessment of the breast tissue can be limited by the scattered radiation artifact and background enhancement of breast tissue. CESM seems to be helpful in the characterization of breast tissue around microcalcifications. Key Words: Contrast-enhanced spectral mammography; tumor staging; ultrasound; artifacts. ªAUR, 2014
C
ontrast-enhanced spectral mammography (CESM) is an imaging technique combining digital mammography with intravenous injection of iodinated contrast media to detect hypervascularized lesions, especially in dense breast tissue (1). Although it was developed several years ago, knowledge about the performance of this technique in clinical routine, especially in breast cancer screening, is still limited. However, a recent study proposes similar indications for CESM as for magnetic resonance imaging (MRI), namely preoperative staging, detection of occult lesions, monitoring of treatment response (2). Initial results comparing this new technique with mammography, ultrasound (US) and breast
Acad Radiol 2014; 21:1363–1369 From the Department of Diagnostic and Interventional Radiology, Medical Faculty, University Dusseldorf, Moorenstr. 5, 40225 Dusseldorf, Germany (K.S.B., C.R., B.M., G.A., S.O.); and Department of Gynecology, Medical Faculty, University Dusseldorf, Dusseldorf, Germany (S.M.). Received May 20, 2014; accepted June 24, 2014. Address correspondence to: K.S.B. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2014.06.012
MRI show a better detection of suspicious lesions with CESM compared to full-field digital mammography and the combined imaging of mammography and US (3) but a lower detection rate of hypervascularized breast lesions compared to breast MRI (4). Studies evaluating the accuracy of CESM compared to MRI in preoperative tumor staging show a similar accuracy in lesion size measurement compared to MRI (5). Preoperative staging of cancer extent in the breast is necessary to plan the optimal treatment (6). MRI is the most commonly used approach to determine the extent of the tumor in the breast and to decide which surgery should be performed and if the breast should be radiated (7,8). Besides its high cost and limited availability, one major problem of breast MRI is background enhancement of breast tissue, which decreases the detection of breast lesions and affects breast cancer staging (9). In such patients, US is a good alternative method for breast cancer staging. Studies correlating tumor size determined with imaging and histopathology describe a tendency to underestimate tumor size with US (10–13), whereas MRI tends to overestimate tumor size 1363
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
TABLE 1. Details of Patients Included in the Study
Number of Patients
Age (Mean Age; Standard Deviation)
20
40–73 (57; 10)
ACR (Artifacts) Disease Invasive cancer Invasive cancer with EIC
ACR2 2 4 (1)
ACR3 4 6
ACR 4 1 3 (2)
The majority of patients had invasive cancer with additional extensive in-situ component (EIC).
(13,14). The good performance of CESM in the detection of multifocal and multicentric cancer dissemination has been shown (4), as well as the high accuracy of tumor size measurements with CESM compared to MRI (5). The performance of CESM in preoperative tumor size measurement compared to US has not been investigated so far. For CESM a technical artifact has been reported which is called ‘‘scattered radiation artifact’’ caused by differences in breast tissue thickness from the thorax to the edge of the breast and different characteristics of mammography with low and high energy (3). Motion artifacts or background enhancement of the breast parenchyma interfering with image quality of CESM has not been reported to date. Therefore, the purpose of this study was to compare the accuracy of local tumor staging with CESM compared to US, with histopathology serving as the gold standard. In addition, CESM was assessed for artifacts potentially limiting interpretation of the CESM data set. In four patients, a proof of principle was performed to show that CESM imaging before stereotactically guided vacuum-assisted core biopsy is feasible.
MATERIAL AND METHODS The local ethics committee approved this prospective study, which was performed in accordance with the Declaration of Helsinki. Written consent of all patients was obtained before enrollment of the study.
Breast density was classified with low-energy mammography of CESM by the standardized classification system of the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS).
CESM Imaging
A Senographe Essential CESM (GE Healthcare, Solingen, Germany) was used for CESM examinations. Dual-energy CESM was performed with automated parameters acquiring a low- and high-energy mammography during one breast compression. Molybdenum (Mo) or rhodium (Rh) target and Mo or Rh filter were used depending on the breast density and compression thickness. Peak kilovoltage (kVp) values ranging from 26 to 31 were used for the acquisition of low-energy mammography, whereas high-energy mammography of CESM was acquired with 45–49 kVp, with x-ray spectrum above the k-edge of iodine (33.2 keV). Solutrast 300 (Bracco Imaging Germany, Konstanz, Germany) was used as contrast media with a weight-adapted dose of 1.5 mL/kg body weight with a flow rate of 3 mL/s, followed by an injection of 20-mL sodium chloride. Mediolateral (MLO) view and craniocaudal (CC) view of the nonaffected breast were acquired 2 minutes after the initiation of contrast media application, followed by MLO- and CC-CESM of the affected breast. The Senographe Essential CESM generated iodine-enhanced images from the lowand high-energy images. Regions of contrast media uptake were displayed with recombined image-by-image processing.
Patients Ultrasound
Women with histologically proven invasive breast cancer, older than 30 years, were enrolled in the study between February and October 2013. Carcinoma was diagnosed with mammography or US and histologically proven by biopsy, before CESM. CESM was clinically indicated for determination of tumor size. Additional MRI was performed if CESM results were not clear (n = 4). Patients with impaired renal function, pregnant or breastfeeding patients, patients with hyperthyroidism, and patients with a history of anaphylactic reaction to contrast media were excluded. CESM was not performed at the same day with other investigations in which application of contrast media was necessary (MRI and computed tomography). Neoadjuvant chemotherapy was an exclusion criterion, as well as hormone treatment or radiotherapy of the breast. 1364
An Aixplorer Supersonic ultrasound system (Supersonic Imaging, Munich, Germany) or Toshiba Aplio MX (Toshiba Medical Systems, Neuss, Germany) were used for second look US before biopsy or surgery. Three gynecologists with 5–12 years of experience in US investigated and staged both breasts. Measurements were performed in all three axes. The largest tumor diameter, which was defined in consensus, was recorded to compare lesion size with histopathologic lesion size. Investigators were blinded to measurements in CESM. Image Analysis
Contrast-Enhanced Spectral Mammography. Anonymized images were evaluated with a high-resolution workstation
Academic Radiology, Vol 21, No 11, November 2014
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
Figure 1. Low-energy mammography (a) and processed CESM (b) oblique mammography (RMLO) and craniocaudal view (RCC; c and d) of the right breast of a woman with invasive ductal breast cancer. Scattered radiation artifact is marked with arrows, appears more prominent in RMLO view, and does not affect tumor size measurement. CESM, contrast-enhanced spectral mammography; RCC, right craniocaudal; RMLO, right mediolateral.
(Barco Inc, Duluth, GA) on completion of each examination by two radiologists with 15 and 3 years of experience in reading mammographies and breast MRI and basic training in reading CESM images. Both readers were blinded to US results and past medical history. Breast density and microcalcifications were recorded by evaluation of the low-energy CESM image by the radiologists, and lesion size, artifacts, and potential additional lesions were evaluated in processed CESM image using the BI-RADS classification. Consensus-based measurements of tumor size was performed in three axes, and the maximum diameter was recorded for a later comparison with US and histopathology.
Statistical Analysis
All calculations were performed using R v3.0.2. Bland–Altman analysis was used to compare differences in lesion size measurements with CESM and US and, furthermore, to correlate tumor size measurements between CESM and the
gold standard (15). Mean differences between tumor size measurements with imaging and histopathologic results were calculated and related to the limits of agreement (LOA), which is the interval in which 95% of the calculated differences were found. The influence of breast density and additional extensive in-situ component (EIC) component on lesion size measurements of invasive cancer on was evaluated as subanalysis.
RESULTS CESM was performed in 22 patients with invasive breast cancer. Twenty patients aged between 40–73 years were included in the study, and two patients were excluded because of neoadjuvant chemotherapy before surgery. Six patients had scattered fibroglandular breast densities (ACR2), heterogeneously dense breast tissue (ACR3) was reported in ten patients, and four patients had extremely dense breast tissue (ACR4) as displayed in Table 1. Four women underwent 1365
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
Figure 2. Processed contrast-enhanced spectral mammography in (a) craniocaudal (LCC) and (b) oblique view (LMLO) in a woman with invasive ductal breast cancer with additional extensive intraductal component. The arrows mark the tumor; asterisks (*) mark enhancing foci retroareolar. Additional non-mass enhancement medial in the breast (**) is caused most likely because of hormonal proliferation. LCC, left craniocaudal; LMLO, left mediolateral.
Figure 3. (a) Low-energy mammography of contrast-enhanced spectral mammography (CESM) of a woman with histological proven cancer in the upper part of the breast with suspicious microcalcifications (*). Additional suspicious microcalcifications are detected in the anterior third of breast tissue in the upper part of the breast (arrow in the magnification of a in b). Processed CESM image (c) with enhancement of the tumor and around the additional microcalcifications (arrow in the magnification of c in d). (e) Stereotactic-guided vacuum biopsy of the additional microcalcification, which turned out to be a ductal carcinoma in situ.
stereotactic-guided vacuum-assisted core biopsy of the breast within 1 hour after CESM. Seven patients had invasive cancer and 13 patients had invasive cancer with an additional EIC (Table 1). One patient had invasive lobular carcinoma with an additional EIC component; twelve patients had invasive ductal carcinoma together with EIC. In two patients with invasive cancer, histopathologic analysis showed invasive lobular carcinoma, and five patients had an invasive ductal carcinoma.
1366
Artifacts
Motion artifacts affected none of the CESM mammographies, neither the low-energy nor the high-energy CESM. The previously described scattered radiation artifact (Fig 1), which describes brighter areas near the edges of the breast and is explained by scattered radiation, was observed in all breasts but did not lead to misinterpretation of the CESM image. This artifact was more prominent on oblique views (Fig 1b) compared to the CC view (Fig 1d).
Academic Radiology, Vol 21, No 11, November 2014
TABLE 2. Histopathologic Results of Four Patients with Suspicious Microcalcifications, Which Received CESM before Stereotactic-Guided Vacuum Biopsy
Patients Primary Cancer Type Invasive lobular cancer
Enhancement Around Microcalcifications N
N
Invasive ductal cancer
N
Invasive ductal cancer
Y
N
Invasive ductal cancer
Y
Histopathologic Finding Fibrocystic mastopathy, ductal hyperplasia, and sclerosing adenosis Fibrocystic mastopathy, ductal hyperplasia, and sclerosing adenosis Fibrocystic mastopathy, ductal hyperplasia, and chronic inflammation Atypical lobular hyperplasia (ALH)* Fibrocystic mastopathy and ductal hyperplasia DCIS*
CESM, contrast-enhanced spectral mammography; DCIS, ductal carcinoma in situ. CESM showed enhancement surrounding microcalcifications in two patients, one with DCIS and one with ALH. Both lesions were excised. *Histologic result led to therapy changes.
Background Parenchymal Enhancement of Breast Tissue
Focal enhancement or additional non-mass enhancement of the breast tissue most likely due to hormone proliferation of the breast tissue, similar to focal or non-mass enhancement of breast tissue on MRI, was seen in breasts of three patients (Table 1). Homogenous enhancement of the breast tissue occurred in one woman with fibroglandular density of the breast (ACR2), aged 58 years and two patients with extremely dense breast tissue (ACR4) with age of 42 and 43 years (Fig 2). CESM Before Stereotactically Guided VacuumAssisted Core Biopsy of the Breast
Four women underwent CESM before stereotactic-guided vacuum-assisted core biopsy of the breast to assess the extent
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
of histologically proven tumor and detect potential enhancement of additional suspect microcalcifications. Six microcalcifications detected with mammography were biopsied; two patients underwent bilateral biopsy. In two patients, enhancement of the breast tissue around microcalcifications was detected with CESM (Fig 3). Histopathologic analysis of biopsied microcalcifications with enhancement of the surrounding breast tissue was either atypical lobular hyperplasia or ductal carcinoma in situ (DCIS), which led to a different surgical therapy (Table 2). Microcalcifications without enhancement of the surrounding breast tissue showed benign histopathologic results (Table 2). No Significant Influence of Additional EIC on Tumor Size Measurements of Invasive Breast Cancers
Histologically determined tumor size ranged between 9 and 40 mm with a mean tumor size of 23 mm (8.8 mm). The mean difference between sonographically and histopathologically determined tumor size was 2.2 mm (LOA, 17.1 to 12.7 mm; Table 3). As displayed in Table 3, additional EIC led to an underestimation of tumor size with US by 2.5 mm. In comparison, invasive cancer alone was underestimated by 1.6 mm. Student t test was performed, and all results were found to be nonsignificant (P > .05). The mean difference between tumor (invasive cancer and invasive cancer with EIC) size determined with CESM and histopathology was 0.3 mm. As shown in Table 3, additional EIC led to nonsignificant overestimation of tumor size with CESM by a mean of 0.8 mm, whereas invasive cancer without additional EIC was underestimated using CESM by 0.7 mm. An additional EIC component did not significantly influence tumor size measurements when assessed with CESM (Fig 4; Table 3). Good agreement of tumor size measurements with CESM and histology were seen in smaller tumors #22 mm (Fig 4). LOA for tumor size measurements with CESM was with 18.9 and 19.5 mm much wider than measurements with US (LOA, 17.1 to 12.7 mm). The standard deviation of the mean difference was 9.6 mm for CESM and 7.5 mm for US showing a lower precision of measurements with CESM compared to US. Thus, lesion size variability as compared to the reference standard was lower for US than for CESM. However, US systematically underestimated lesion size as compared to the reference standard.
DISCUSSION This study demonstrates that CESM is accurate in determination of tumor size as compared to the reference standard and US although tumor size measurements with CESM may not be very precise. CESM is an imaging technique, which is more and more introduced into clinical routine. The duration of the examination is shorter; the detection rate of breast cancer is comparable to that of MRI. The sensitivity of detecting additional cancer of the breast is lower, compared to MRI, but 1367
BLUM ET AL
Academic Radiology, Vol 21, No 11, November 2014
TABLE 3. Comparison of the Differences Between Measured Tumor Size with Imaging and Histopathology Ultrasound Tumor Type Invasive cancer (n = 7) Invasive cancer + EIC (n = 13) Total (n = 20)
Mean (mm) 1.571 2.538 2.2
CESM LOA (mm) 19.48 to 14.4 12.82 to 9.67 17.1 to 12.7
Mean (mm) 0.714 0.846 0.3
LOA (mm) 11.15 to 9.72 22.06 to 23.75 18.88 to 19.48
CESM, contrast-enhanced spectral mammography; EIC, extensive in situ component; LOA, limits of agreement. t Test showed no significant under- or over-estimation of tumor size (P < .05).
Figure 4. Bland–Altman analysis of agreement of histologic tumor size with size measurements with contrast-enhanced spectral mammography (CESM) and ultrasound (US) in patients with invasive breast cancer with or without extensive intraductal component (EIC). High agreement of histologic tumor size and CESM measurements in patients with smaller tumors (#22 mm). Additional EIC component did not affect tumor size measurements in either way.
CESM has a higher specificity (4). Studies evaluating the accuracy of CESM in preoperative tumor staging show a similar accuracy in lesion size measurement compared to MRI (5). The present study is the first study comparing tumor size measurements of CESM with US in preoperative staging. Similar to other studies (13), we found an underestimation of tumor size with US, whereas tumor-sized measurement with CESM slightly overestimated the tumor size similar to measurements with MRI (14,16,17). Good agreement and precision of tumor size diameter measured with CESM and histopathologic results were seen in tumors with #22 mm size, similar to studies on MRI (18) and other studies using CESM (19). In comparison to US, CESM measured tumor size more accurately and was not affected by the density of the breast tissue, as described for conventional digital mammography (20). Similar to the study of Lobbes et al., the tumor size was highly overestimated in one patient (19). In our study, this was due to the misinterpretation of background enhancement as tumor tissue. 1368
Preoperative size assessment is important to decide the most beneficial surgical therapy for the patient. Although only a few patients were included in this study, CESM seems to be a good imaging method for preoperative tumor staging, as suggested by others (5). The similar slight overestimation of tumor size of CESM as described for MRI seems to be most likely caused by a different vascularization of the tumor, or precancerous surrounding breast tissue (21). Background enhancement of the breast tissue as described for MRI (9) seems to limit the evaluation of tumor size in CESM in a similar way. A disadvantage of CESM shown in this study is that the accuracy of tumor size measurements decreases with larger tumor size. Similar results have been reported for tumor size measurements with MRI or CESM (18,19). Lower accuracy of tumor size measurements with imaging in patients with large tumors leads to wide LOA, although the mean tumor size difference between gold standard and CESM stays very small (0.3 mm). This is displayed by the larger standard deviation of mean differences.
Academic Radiology, Vol 21, No 11, November 2014
A beneficial effect of CESM compared to MRI seems to be the evaluation of microcalcifications, which can be exactly correlated to enhancement of the breast tissue. Although we investigated only a few patients with CESM before stereotactically guided vacuum biopsy, we could show that the latter is not affected by CESM. Moreover, the present study suggests that CESM before stereotactically guided vacuum core biopsy may be beneficial to differentiate malignant from nonmalignant lesions by detection of contrast enhancement surrounding microcalcifications in malignancy. Our study has some limitations. The major limitation is the small amount of patients investigated in the study, which might be a reason for nonsignificant results. Because of the small study group, a subgroup analysis of different cancer types was not performed, which is the second limitation. Another limitation is that CESM was not performed with regard to the menstrual cycle, which might have affected the results in patients with background enhancement. In conclusion, we showed the feasibility of preoperative tumor staging of breast cancer with CESM. Albeit a small study group, we could show more accurate tumor size measurements with CESM compared to US. Similar artifacts as described for MRI, affecting tumor size measurements, were reported in the study for the first time. Preliminary results suggest a beneficial effect of CESM when differentiating benign from malignant microcalcifications of the breast. Further research is necessary and more patients have to be investigated to prove those preliminary results.
REFERENCES 1. Dromain C, Balleyguier C, Adler G, et al. Contrast-enhanced digital mammography. Eur J Radiol 2009; 69(1):34–42. 2. Lobbes MB, Smidt ML, Houwers J, et al. Contrast enhanced mammography: techniques, current results, and potential indications. Clin Radiol 2013; 68(9):935–944. 3. Dromain C, Thibault F, Muller S, et al. Dual-energy contrast-enhanced digital mammography: initial clinical results. Eur Radiol 2011; 21(3):565–574. 4. Jochelson MS, Dershaw DD, Sung JS, 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(3):743–751.
CESM IS ACCURATE IN TUMOR SIZE DETERMINATION
5. Fallenberg EM, Dromain C, Diekmann F, 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(1):256–264. 6. Litiere S, Werutsky G, Fentiman IS, et al. Breast conserving therapy versus mastectomy for stage I-II breast cancer: 20 year follow-up of the EORTC 10801 phase 3 randomised trial. Lancet Oncol 2012; 13(4):412–419. 7. Pengel KE, Loo CE, Teertstra HJ, et al. The impact of preoperative MRI on breast-conserving surgery of invasive cancer: a comparative cohort study. Breast Cancer Res Treat 2009; 116(1):161–169. 8. Tendulkar RD, Chellman-Jeffers M, Rybicki LA, et al. Preoperative breast magnetic resonance imaging in early breast cancer: implications for partial breast irradiation. Cancer 2009; 115(8):1621–1630. 9. Uematsu T, Kasami M, Watanabe J. Does the degree of background enhancement in breast MRI affect the detection and staging of breast cancer? Eur Radiol 2011; 21(11):2261–2267. 10. Hieken TJ, Harrison J, Herreros J, et al. Correlating sonography, mammography, and pathology in the assessment of breast cancer size. Am J Surg 2001; 182(4):351–354. 11. Finlayson CA, MacDermott TA. Ultrasound can estimate the pathologic size of infiltrating ductal carcinoma. Arch Surg 2000; 135(2):158–159. 12. Pritt B, Ashikaga T, Oppenheimer RG, et al. Influence of breast cancer histology on the relationship between ultrasound and pathology tumor size measurements. Mod Pathol 2004; 17(8):905–910. 13. Gruber IV, Rueckert M, Kagan KO, et al. Measurement of tumour size with mammography, sonography and magnetic resonance imaging as compared to histological tumour size in primary breast cancer. BMC Cancer 2013; 13:328. 14. Uematsu T, Yuen S, Kasami M, et al. Comparison of magnetic resonance imaging, multidetector row computed tomography, ultrasonography, and mammography for tumor extension of breast cancer. Breast Cancer Res Treat 2008; 112(3):461–474. 15. Lobbes MB, Nelemans PJ. Good correlation does not automatically imply good agreement: the trouble with comparing tumour size by breast MRI versus histopathology. Eur J Radiol 2013; 82(12):e906–e907. 16. Jiang YZ, Xia C, Peng WT, et al. Preoperative measurement of breast cancer overestimates tumor size compared to pathological measurement. PLoS ONE 2014; 9(1):e86676. 17. Schouten van der Velden AP, Boetes C, Bult P, et al. The value of magnetic resonance imaging in diagnosis and size assessment of in situ and small invasive breast carcinoma. Am J Surg 2006; 192(2):172–178. 18. Grimsby GM, Gray R, Dueck A, et al. Is there concordance of invasive breast cancer pathologic tumor size with magnetic resonance imaging? Am J Surg 2009; 198(4):500–504. 19. Lobbes MB, Lalji U, Houwers J, et al. Contrast-enhanced spectral mammography in patients referred from the breast cancer screening programme. Eur Radiol 2014; 24(7):1668–1676. 20. Fasching PA, Heusinger K, Loehberg CR, et al. Influence of mammographic density on the diagnostic accuracy of tumor size assessment and association with breast cancer tumor characteristics. Eur J Radiol 2006; 60(3):398–404. 21. Guinebretiere JM, Le Monique G, Gavoille A, et al. Angiogenesis and risk of breast cancer in women with fibrocystic disease. J Natl Cancer Inst 1994; 86(8):635–636.
1369