Breast cancer is the most common health problem among women in developed countries and the rate of death from breast cancer ranks 5^th among all cancer deaths.¹ Early detection of the tumor is important for better disease prognosis. Although started in different age groups in different countries, mammography and ultrasonography (US) are generally used for screening, and magnetic resonance imaging (MRI) for advanced examination or diagnostic purposes. However, while ultrasonography can be used for diagnostic purposes in young patients, MRI can also be used for screening in cases with a family history.

Although MRI as a screening tool has high sensitivity, it suffers from false positivity and high overdiagnosis rates. For this reason, unnecessary biopsies and follow-ups are more common if only MRI is acquired. To avoid this pitfall, MRI should be used as a problem-solving method to complement other methods as a part of multi-modality diagnosis.²

Some morphological findings (size, vascularity, number, invasion to neighboring structures) of the tumor are important in determining the type of treatment.³ MRI is superior to mammography and sonography in showing these parameters that affect the management options, operation decision and surgical method. This study aimed to determine the role of breast MRI in early stages of breast cancer (T1), to reveal the radiological findings of the tumor and to compare the results with mammography and US findings.

Patient selection

In this prospectively designed study, dynamic contrast enhanced breast MRI was acquired in 53 patients (52 women, 1 man) referred to our hospital in the last two years (2006-2008) with breast lesions evaluated as BI-RADS 4 or 5 based on mammography and ultrasonography.

The study group consisted of patients over 40 years of age who presented to our clinic for routine screening, and patients under 40 years of age who were referred to our clinic for family history of breast cancer, palpable lesions or stiffness in the breast, nipple discharge or axillary mass. US examination was performed in all patients. Mammography was not performed in patients under 40 years of age and without a family history (n=21).

Cases with definite and relative contraindications to MRI such as metallic prosthesis, stents or pacemakers were excluded.

Imaging indications and techniques

Mammography was used for screening purposes over 40 years of age, ultrasonography was used as a supplement to mammography in cases over 40 years old with dense breasts. US was used for diagnostic purposes in cases presenting with a specific complaint, and for screening purposes in patients under the age of 40 with a family history. MRI was performed for diagnostic purposes and as an additional imaging modality upon detection of suspicious lesions in mammography and ultrasonography. It was also used for screening in young patients with a family history. All these procedures were part of routine care in our clinic.

Mammographic examinations were performed with a Philips® MD4000™ mammography system (Philips Healthcare, Netherlands) using routine craniocaudal (CC) and mediolateral-oblique (MLO) projection and evaluated comparatively. Mammography was evaluated for mass lesions, microcalcification, parenchymal distortion, asymmetric density, skin thickening, and nipple retraction.

US examinations were performed with a Toshiba® Xario XG ultrasoundsystem (Toshiba, Japan) using 12 MHz linear transducer. Images were evaluated for size, echo and contour pattern, posterior acoustic shadowing or strengthening of the lesion and its relationship to surrounding structures.

MR imaging was performed with a 1.5 Tesla MR (Magnetom, Symphony Maestro Class, Siemens, Germany) using a double breast coil. T2 TIRM-sagittal (TR 10,000 msec-TE 70 msec), TSE-T2 axial (TR 6780 msec-TE 97 msec), dynamic T1 FLASH 3D-coronal (TR 10 msec-4.76 msec, section thickness 2.5 mm) and fat-suppressed late contrast enhancement T1 axial (TR 29 msec-TE 4.76 msec) images were obtained in all examinations.

Following pre-contrast images, 0.1 mmol/kg IV contrast agent (gadolinium) was injected with an automatic syringe at a rate of 3 mL/sec, and then the same region was imaged dynamically 5 times in succession. After the dynamic series, late-contrast T1-weighted images were obtained with fat suppression in the axial plane. Subtraction images were obtained at the end of the examination to reveal the contrast enhancement more clearly. Time-signal intensity curves were plotted using the post-processing feature of the device to demonstrate the contrast retention rate and intensity of lesions during the dynamic examination.

Analysis of MRI images

Images obtained from breast MRI were evaluated under the headings described below.

Background parenchymal enhancement

Breast imaging as part of screening was performed in 2-3 weeks of the menstrual cycle in premenopausal women to prevent intense ground enhancement. If the case was referred for a suspicious lesion, MRI was performed irrespective of the menstrual cycle as soon as possible. Ground enhancement was evaluated in 4 groups according to ACR BI-RADS for MRI, 4^th version (2013) criteria (minimal, mild, moderate, marked).

Evaluation of morphological characteristics

The borders, shape and internal structure of the mass determined in the morphological examination were evaluated. At the same time, lesions were evaluated as focus, mass enhancement, and non-mass enhancement according to the dynamic enhancement pattern. Contrast enhancements smaller than 5 mm were defined as a focal enhancement. A contrast enhancement consisting of more than one focus was defined as point enhancement. Contrast enhancement greater than 5 mm that pushes adjacent tissue was considered as a mass enhancement. The shape and boundary characteristics of the mass lesions were recorded on both non-contrast and dynamic contrast images. The masses were also evaluated for homogeneity, heterogeneity, rim, non-enhancing internal septa, enhancing internal septa and central contrast enhancement patterns relative to intra-mass enhancement characteristics. If the enhancement was not in the form of focus or mass, it was evaluated as a non-mass enhancement. Non-mass enhancement was defined as focal, linear, ductal, segmental, regional and diffuse according to the distribution of enhancement.

Analysis of kinetic properties

Time-signal intensity curves were drawn by selecting the sections demonstrating the lesion best on dynamic contrast imaging. According to the curves, enhancement pattern was grouped as type I, persistent; type II, plateau; and type III, wash out pattern. ROI was kept small because it may be affected by the necrotic or desmoplastic component of the lesion and result in a false-negative diagnosis. Dynamic curves were also evaluated in terms of the early arterial peak.

ACR-BIRADS classification system (4th version, 2013) was used for evaluation of all lesions on mammography, ultrasonography and MRI.⁴All images were evaluated separately by two radiologists with 15 years and 5 years of experience in breast imaging. In suspicious cases, images were re-evaluated in consensus

Histopathological findings

Lesions detected on breast MRI were evaluated according to the BI-RADS category. Lesions considered as BI-RADS 3 after MR examination were followed-up with sonography every 6-months. Biopsy was performed in patients with a final decision of BIRADS 4-5 according to mammography-sonography and MRI imaging features. All biopsies were performed under mammography or sonography guidance. No case underwent biopsy under MRI guidance. For microcalcifications visible only on mammography, markers were placed under mammography guidance, and the final results were obtained by examining the surgical material. Histopathological results were accepted as the gold standard for evaluation of the lesions and definitive diagnosis.

Informed consent was obtained from each patient prior to MRI and biopsy. Informed consent included future use of their data for research purposes. Institutional review board approval was received from Atatürk University.

A total of 53 patients (52 females, and 1 male) were examined with breast MRI. The age range of the study population was 25-77 years with a mean age of 45 years. The mean age of patients with final benign diagnosis was 37.3 years, and the mean age of patients with a final malignant diagnosis was 50.5 years.

Histology and types of lesions

Of the 53 cases, 22 (41.5%) had benign and 31 (58.4%) had malignant lesions on histopathology. While 22 (70.9%) of all malignant cases were invasive ductal carcinoma (IDC), granulomatous mastitis (22.7%) was the most common histological diagnosis in the benign group. Fibroadenoma were detected in 4 cases and fibrocystic changes (FCD) in 4 cases on biopsy (18%) (Table 1).

Table 1. Histopathological findings

Eight (25.8%) of 31 patients with malignant lesions had multicentric lesions and 4 (12.9%) had multifocal lesions. The smallest lesion was IDC, with a diameter of 6 mm, and the largest lesion was an angiomyolipoma, 16 cm in diameter.

Morphological features of lesions

All benign and malignant lesions were isointense or hypointense on T1-weighted images. A number of 28 of 31 malignant lesions were isointense or hypointense on T2-weighted images; 25 malignant lesions were hyperintense or heterogeneously hyperintense on STIR sequences.

Mass enhancement was detected in 21 (68%) of 31 malignant cases and 13 (59%) of 22 benign cases. For mass lesions, the positive predictive value (PPV) of spliced edge was 85% for malignancy, and PPV of smooth border for benignity was 87.5%. In malignant lesions, PPV of peripheral enhancement was 78.2% and hypointensity on T2-WI was 88% and heterogeneous enhancement was 83.3% on T2-weighted images (Figure 1a, 1b, 1c).

Non-mass enhancement was present in 10 (32%) of 31 malignant cases (Figure 2a, 2b, 2c) and 8 (36%) of 22 benign cases (Figure 3a, 3b, 3c). While non-mass enhancement was mostly detected in granulomatous mastitis cases among benign lesions, it was detected in 7 patients with IDC, 1 invasive lobular carcinoma (ILC) and 1 comedocarcinoma in malignant lesions. In these cases, parenchymal distortion with US, intense distal acoustic shadowing or microcalcifications were detected in mammography.

Overall, 59% of benign lesions showed type 1 kinetic curve and 27% type 2 kinetic curve, while 45% of malignant lesions showed type 2, and 48% showed type 3 kinetic curves after contrast administration. Type 3 kinetic pattern was observed in 14% of benign lesions and type 1 in 7% of malignant lesions (Table 2).

The pathology result was accepted as the gold standard, ROC analyzes were performed to test the consistency of mammography, US and MRI results with pathology results. While the area obtained for the US + MRI association was 0.870, the sensitivity was 96.8%, the specificity was 77.3%, the PPV was 85.7%, and the negative predictive value (NPV) was 94.4% (p=0.004; p<0.01). Its accuracy was determined as 88.7%. The area obtained for mammography + MRI association was 0.825, the sensitivity was 96.8%, specificity 68.2%, PPV 81.1%, and NPV 93.7% (p=0.036; p<0.01). Its accuracy was determined as 84.9%. No statistically significant difference was found between the areas obtained by MR in determining the lesions in the breast, between the areas obtained for US + MRI and Mam + MRI

(DeLong test p: 0.317; p: 0.055) (Table 3 and Figure 4).

Mammography and breast US are the main radiological methods used in breast imaging in our institution and throughout the world.⁵ Mammography can be applied in large patient groups as a scanning method.⁶ Although MRI is a valuable method in breast imaging, it cannot be used as a scanning method because of its expense and low specificity.⁷

On mammography, features with highest PPV for malignancy are linear and segmental distribution of microcalcifications, spiculated boundaries in mammographic mass lesions and spread within the breast tissue.⁸ On US, spicular-irregular edge, taller than wide shape, pronounced hypoechogenicity, and posterior acoustic shadowing are important criteria suggestive of malignancy.⁹

When assessing a lesion on MRI, morphological features, as well as contrast enhancement patterns and kinetic features, are evaluated. The shape and edges of the lesion are the most important factors that determine the morphology of the lesion. Spherical or oval shape has 97-100% PPV for benignity in some series, while in some series 20% of carcinomas were smooth, round and oval.^10,11 Similar rates were seen in our study.

On MRI, hyperintensity on T2-weighted images suggests benignity, while hypointensity suggests malignancy.¹ In our study, hyperintensity was mostly seen in cysts. Similarly, hyperintensity was detected in myxoid fibroadenomas and mastitis.¹² However, carcinomas may also have a high signal on T2-weighted images, particularly in mucinous subtypes or due to cystic or necrotic components and edema.¹³ In our study, the PPV of hypointensity on T2-weighted images for malignancy was 88%, compatible with the literature.¹¹

Non-mass lesions, whether benign or malignant, are generally isointense on T2-weighted images relative to the surrounding fibroglandular tissues.¹⁴ Similarly, all benign or malign non-mass lesions are isointense compared to fibroglandular tissues on T1-weighted images.¹⁵In the literature, non-mass lesions were mostly detected in fibrocystic diseases, while in our study non-mass enhancement was detected in 17 cases – 9 malignant and 8 benign.¹⁶ Non-mass enhancement was most commonly seen in invasive ductal carcinoma in malignant cases and granulomatous mastitis in benign cases. Non-mass enhancement was also detected in 2 of 4 cases with fibrocystic disease diagnosed histopathologically.

Angiogenesis is a pathological process observed in breast cancers as well as in other malignancies.¹⁷Contrast enhancement of the lesion on MRI depends on vascular permeability,

diffusion rate of contrast agent, content of interstitial tumor matrix, base T1 relaxation values, and T1 relaxation values after contrast agent.¹⁸When evaluating the contrast pattern of the lesion on MRI, initially mass versus non-mass enhancements are defined.¹¹ Lack of contrast enhancement is the most important descriptive character in distinguishing benign lesion from a malignant lesion and has a 100% PPV for benignity.¹⁹

The form of enhancement that was frequently observed in benign cases in this series was homogeneous enhancement and dark internal septa. The PPV value was 100% for both contrast patterns and is compatible with the literature findings.⁴ Homogenously enhancing lobular masses and non-enhancing internal septa have 93-97% PPV for benignity and these lesions are generally fibroadenomas in accordance with our study.²⁰ Circular (peripheral) enhancement is a common finding in invasive cancers.²¹ However, this enhancement pattern can lead to false-positive results as it can be seen in benign inflamed cysts and fat necrosis.²⁰

Heterogeneous enhancement is a contrast enhancement feature that varies according to cellular activity in different parts of tumor tissue and is observed due to necrosis.^22,23 It was the most common enhancement feature in malignant cases in our study.

In the literature, regional enhancement has been mostly described in ductal carcinoma in situ (DCIS), infiltrative tumor spread, fibrocystic diseases, and non-massive invasive lobular carcinoma. Linear, ductal and segmental enhancement is usually a form of enhancement detected in patients with DCIS.²⁴

Kinetic analysis is less reliable than morphological analysis. Generally, type 1 progressive curve is observed in benign cases, type 2 plateau curve is observed in both benign and malignant cases, and type 3 curve-wash-out pattern is mostly observed in malignant cases.²² Similar results were found in our study and type 2 curve was the most common kinetic curve pattern with 20 patients in all benign and malignant cases.

Whether or not there is contrast enhancement, the form and kinetics of contrast enhancement, intensities on T2-weighted images provide important information but they do not have significant diagnostic value alone.²⁴ Although the morphological features of the lesion enhancement are important in lesion characterization, the same morphological features can be seen in benign and malignant cases as shown in our study and in literature.¹¹ Lack of contrast enhancement in dynamic series is a reliable criteria for benignity. In malignancies, mass with irregular and spicular extensions and T2 hypointensity, peripheral and heterogeneous enhancement are the most frequently identified findings.²⁴

It is important to know the real extent of the disease in determining the appropriate treatment method. While multifocal cancer requires a greater excision than normally performed, multicentricity makes breast-conserving surgery impossible and mastectomy is required.²⁵ Skin, pectoral muscle, chest wall, and adjacent breast tissue involvement should also be known in planning the appropriate surgical method. Knowing whether there is involvement in the opposite breast is also important in surgical planning. In cases where there is an inflammatory spread, intense inflammation in breast tissue can mask mass lesions in the US and mammography.²³ Dynamic breast MRI more sensitively detects multifocality, multicentricity and distribution of the lesion, which cannot be detected by mammography and the US. However, especially in non-mass lesions of the breast such as DCIS, MRI should be used with other diagnostic modalities like mammography and sonography due to high false positive rates of MRI.

In our single-center study, the limited number of patients along with absence of mammography for some cases can be considered as limitations of the study.

Although, MRI examination reveals more valuable information about breast lesions such as morphology, kinetic pattern and real extent than mammography and ultrasonography, multi-level diagnostics and multi-parametric analysis should be used to avoid false positivities and negativities, unnecessary medical interventions, patient anxieties.

1. Azamjah N, Soltan-Zadeh Y, Zayeri F. Global trend of breast cancer mortality rate: a 25-year study. Asian Pac J Cancer Prev. 2019;20:2015-20.

https://doi.org/10.31557/APJCP.2019.20.7.2015

2. Baltzer PA, Benndorf M, Dietzel M, Gajda M, Runnebaum IB, Kaiser WA. False-positive findings at contrast-enhanced breast MRI: a BI-RADS descriptor study. AJR Am J Roentgenol. 2010;194:1658-63.

https://doi.org/10.2214/AJR.09.3486

3. Saxena S, Rekhi B, Bansal A, Bagga A, Murthy NS. Clinico-morphological patterns of breast cancer including family history in a New Delhi hospital, India – a cross-sectional study. World J Surg Oncol. 2005;3:67.

https://doi.org/10.1186/1477-7819-3-67

4. de Almeida JR, Gomes AB, Barros TP, Fahel PE, Rocha Mde S. Predictive performance of BI-RADS magnetic resonance imaging descriptors in the context of suspicious (category 4) findings. Radiol Bras. 2016;49:137-43.

https://doi.org/10.1590/0100-3984.2015.0021

5. Özmen V, Gürdal SÖ, Cabioğlu N, et al. Cost-effectiveness of breast cancer screening in Turkey, a developing country: results from Bahçeşehir Mammography Screening Project. Eur J Breast Health. 2017;13:117-22.

https://doi.org/10.5152/ejbh.2017.3528

6. Nelson HD, Weerasinghe R, Wang L, Grunkemeier G. Mammography screening in a large health system following the US Preventive Services Task Force recommendations and the Affordable Care Act. PLoS One. 2015;10:e0131903.

https://doi.org/10.1371/journal.pone.0131903

7. Mann RM, Kuhl CK, Moy L. Contrast-enhanced MRI for breast cancer screening. J Magn Reson Imaging. 2019;50:377-90.

https://doi.org/10.1002/jmri.26654

8. do Nascimento JHR, da Silva VD, Maciel AC. Accuracy of mammographic findings in breast cancer: correlation between BI-RADS classification and histological findings. Radiol Bras. 2010;43:91-6.

https://doi.org/10.1590/S0100-39842010000200008

9. Kim YR, Kim HS, Kim HW. Are irregular hypoechoic breast masses on ultrasound always malignancies?: a pictorial essay. Korean J Radiol. 2015;16:1266-75.

https://doi.org/10.3348/kjr.2015.16.6.1266

10. Morris E, Liberman L. Breast MRI: diagnosis and intervention. New York: Springer Science & Business Media; 2005.

11. Agrawal G, Su MY, Nalcioglu O, Feig SA, Chen JH. Significance of breast lesion descriptors in the ACR BI-RADS MRI lexicon. Cancer. 2009;115:1363-80.

https://doi.org/10.1002/cncr.24156

12. Hassan HHM, Zahran MHM, Hassan HE, Abdel-Hamid AM, Fadaly GAS. Diffusion magnetic resonance imaging of breast lesions: initial experience at Alexandria University. Alexandria J Med. 2013;49:265-72.

https://doi.org/10.1016/j.ajme.2012.11.004

13. Limaiem F, Ahmad F. Cancer, Mucinous Breast Carcinoma. In: StatPearls. StatPearls Publishing; 2019.

14. Chan WP. Magnetic resonance imaging of soft-tissue tumors of the extremities: a practical approach. World J Radiol. 2013;5:455-9.

https://doi.org/10.4329/wjr.v5.i12.455

15. Dağıstan E, Kızıldağ B, Gürel S, Barut Y, Paşaoğlu E. Radiologic and histopathologic review of rare benign and malignant breast diseases. Diagnostic Interv Radiol. 2017;23:286-92.

https://doi.org/10.5152/dir.2017.16085

16. Chen JH, Liu H, Baek HM, Nalcioglu O, Su MY. Magnetic resonance imaging features of fibrocystic change of the breast. Magn Reson Imaging. 2008;26:1207-14.

https://doi.org/10.1016/j.mri.2008.02.004

17. Longatto Filho A, Lopes JM, Schmitt FC. Angiogenesis and breast cancer. J Oncol. 2010;2010:576384.

https://doi.org/10.1155/2010/576384

18. Toprak N, Unal O, Avcu S. The efficacy of one molar contrast material in the evaluation of breast lesions with MR imaging. Pol J Radiol. 2015;80:401-10.

https://doi.org/10.12659/PJR.894863

19. Gunduru M, Grigorian C. Breast Magnetic Resonance Imaging (MRI). In: StatPearls. StatPearls Publishing; 2019.

20. Berg WA, Sechtin AG, Marques H, Zhang Z. Cystic breast masses and the ACRIN 6666 experience. Radiol Clin North Am. 2010;48:931-87.

https://doi.org/10.1016/j.rcl.2010.06.007

21. Santamaría G, Velasco M, Bargalló X, Caparrós X, Farrús B, Luis Fernández P. Radiologic and pathologic findings in breast tumors with high signal intensity on T2-weighted MR images. Radiographics. 2010;30:533-48.

https://doi.org/10.1148/rg.302095044

22. El Khouli RH, Macura KJ, Jacobs MA, et al. Dynamic contrast-enhanced MRI of the breast: quantitative method for kinetic curve type assessment. AJR Am J Roentgenol. 2009;193:W295-300.

https://doi.org/10.2214/AJR.09.2483

23. An YY, Kim SH, Cha ES, et al. Diffuse infiltrative lesion of the breast: clinical and radiologic features. Korean J Radiol. 2011;12:113-21.

https://doi.org/10.3348/kjr.2011.12.1.113

24. Macura KJ, Ouwerkerk R, Jacobs MA, Bluemke DA. Patterns of enhancement on breast MR images: interpretation and imaging pitfalls. Radiographics. 2006;26:1719-34.

https://doi.org/10.1148/rg.266065025

25. Houvenaeghel G, Tallet A, Jalaguier-Coudray A, et al. Is breast conservative surgery a reasonable option in multifocal or multicentric tumors? World J Clin Oncol. 2016;7:234-42.

https://doi.org/10.5306/wjco.v7.i2.234