6 Diagnostic MRI Interpretation

Gillian M Newstead and Michael S. Middleton

Abstract

This chapter outlines a standard method for interpretation and reporting of diagnostic MRI examinations and discusses the management of patients with challenging clinical or imaging diagnoses. Included is a structured reading method, providing clarity and uniformity of interpretation across breast MRI practices. Standard reporting includes not only lesion characterization and a final category assessment, but also specifics of the MRI acquisition, type and amount of GBCA administered, volume of fibro-glandular tissue (FGT) and amount of background parenchymal enhancement (BPE). Critical to accurate interpretation is the selection of reading protocols at the workstation which include the T2w Sequence, the MIP Image (First post-contrast T1w Image, both Source and Subtraction), the T1w Dynamic Sequence (Source and Subtracted Images) and the Kinetic Analysis. Sample protocols are shown. A discussion of challenging diagnoses including evaluation of patients with nipple discharge and papillary lesions, are supplemented with 23 case examples. A section on breast MRI Reporting reviews the Breast Imaging Reporting and Data System (BI-RADS®) developed by the American College of Radiology.

An excellent review of breast implants and the MRI findings of normal breast implants and their associated various abnormalities are further discussed in detail by DR Middleton This section includes the terminology and types of Implants, the MR appearance of various implant abnormalities, including types of rupture, standard reporting of implant cases and methods to improve the challenge of difficult diagnoses

diagnostic DCE-MRI interpretation – reading protocol – kinetic classification – interpretation challenges – nipple discharge – papillary lesions – reporting system (BI-RADS) – breast implants – implant imaging protocol – implant rupture – implant reporting

6 Diagnostic MRI Interpretation

6.1 Introduction

In this chapter, we will outline a standard method for interpretation and reporting of diagnostic magnetic resonance (MR) examinations and discuss management of patients with challenging clinical or imaging diagnoses, including evaluation of women with current or prior implants.

6.2 Interpretation Protocol

A standard reading method provides structure and uniformity of interpretation across breast MR practices. The radiology report should include not only lesion characterization and a final category assessment, but also specifics of the MR acquisition, the type and amount of gadolinium-based contrast agent (GBCA) administered, the volume of fibroglandular tissue (FGT), and the amount of background parenchymal enhancement (BPE). A standard interpretation method for diagnostic breast magnetic resonance imaging (MRI) examinations is outlined below, and a standard hanging protocol is illustrated (Fig. 6‑1).

**Fig. 6.1** Hanging protocols. This viewing protocol is shown as three separate displays; however, depending on available software, displays 2 and 3 could be condensed into one display for the dynamic sequence. A unifocal IDC is shown in the right breast. Hanging protocol #1 allows a technical assessment of image quality by first viewing of the MIP image **(a)**, providing a series overview with comparison to the T2w images **(b–d)**. Hanging protocol #2 shows the T1w precontrast image **(e)** and the source T1w postcontrast source images shown in MPR format **(f–h)**. Hanging protocol #3 exhibits T1w postcontrast subtraction images shown in MPR format **(i, l, m)**, with associated angiomap and TIC **(j, k)**. Histology yielded triple negative IDC.

6.3 MIP Image (First Postcontrast T1w Image, Source, and Subtraction)

The maximum intensity projection (MIP) images provide a useful overview of breast enhancement. Successful dynamic contrast-enhanced MRI (DCE-MRI) requires a rapid bolus injection of contrast, and this can be evaluated by identifying cardiac and major vascular enhancement on the first postcontrast series. The technologist should always check for the presence of contrast enhancement in the heart and great vessels during scanning. If contrast enhancement is not evident on the first postcontrast images, the technologist should check the injection site. Contrast injection into the subcutaneous tissue of the arm or a break in the tubing delivery mechanism could account for an unsuccessful DCE-MRI study; when this occurs, repeat examination is usually necessary. The radiologist should be certain that the dynamic sequence is acceptable by looking for adequate contrast bolus enhancement in the heart and nearby vasculature. Be aware that low delivery of contrast agent may occur in patients with cardiac failure, resulting in delayed contrast delivery. BPE can be assessed on the first postcontrast MIP image. Enhancing lesions and other findings may also be visible on MIP images (Fig. 6‑2) and, if seen, should be further evaluated on source and subtracted images within the complete dynamic series.

**Fig. 6.2** Maximum intensity projection (MIP) image. MIP image (subtracted), obtained at the first postcontrast time point (70 s), reveals a round enhancing mass with associated increased vascularity.

6.3.1 T2w Sequence

The T2-weighted (T2w) sequence can be acquired with fat saturation (typically at lower spatial resolution) or without fat saturation (typically at higher spatial resolution). The amount of FGT can be assessed on this series and should be included in the report. This sequence is useful for identification of high-signal fluid, as found in cysts, subareolar ducts, intratumoral necrosis, and subcutaneous/peritumoral edema. Other findings include visualization of normal or abnormal axillary, intramammary and internal mammary lymph nodes, breast masses, skin thickening, and postsurgical changes. The T2w series can also be used for assessment of postsurgical or postbiopsy marker clips and changes resulting from breast reconstruction surgery, and, as will be discussed later in this chapter, can also assist in the assessment of soft-tissue silicone resulting from implant rupture and from breast silicone fluid injections.

The high-spatial-resolution T2w series allows matching of images slice-by-slice with the T1w DCE series. This sequence accomplishes the need for both a T2w series and a non–fat-suppressed T1w precontrast series and incorporates both acquisition requirements into a single acquisition (Fig. 6‑3). This high-resolution T2w series provides improved assessment of mass morphology and identification of peritumoral and prepectoral edema, findings that may not be visualized on the thicker slices that accompany a fat-suppressed T2w acquisition. Although the T2w sequence may not be helpful for evaluation of noninvasive cancer (because NME exhibits very few precontrast findings) it is very useful for characterization of breast masses.

**Fig. 6.3** T2w series. T2w series (non–fat suppressed) shows a correlative hypointense mass shown in MPR format **(a–c)**.

6.3.2 T1w Dynamic Sequence (Source and Subtracted Images)

The pre- and postcontrast source and subtracted series are next reviewed. It is important to check the precontrast images to be sure that the fat saturation is uniform and to check the subtraction images for evidence of motion artifact. Next we search for abnormal enhancement, which is usually seen best at the first and second time points when the enhancement is most intense and is distinct from BPE. Analysis of any enhancing lesion should include a morphologic assessment of shape, margin, internal enhancement, and distribution characteristics; multiplanar reformatting (MPR) and slab images can be useful in this regard (Fig. 6‑4, Fig. 6‑5). Lesion size, location, laterality (right or left breast), and breast quadrant (including the appropriate use of “central”) and retroareolar and axillary tail should be reported. Lesion distance from the nipple, skin, or chest wall should be measured when appropriate. Although enhancing masses may be visible on the MIP images, careful evaluation of the first and second postcontrast source and subtraction images is needed for complete diagnosis. If an enhancing lesion is found on the T1w postcontrast series, review of the precontrast T2w series may identify a correlate lesion. Viewing of the later postcontrast sequences is also necessary to ensure identification of certain slowly enhancing cancers and for assessment of treatment response for patients undergoing serial MRI examinations during neoadjuvant chemotherapy; delayed enhancement is the only indication of residual disease in some cases. The ability to compare lesion characteristics on both T2w and T1w sequences with identical spatial resolution can result in improved diagnostic specificity.

**Fig. 6.4** T1w postcontrast source series. T1w postcontrast source images obtained at the first-time point (70 s) are shown in MPR format **(a–c)**.

**Fig. 6.5** T1w postcontrast subtraction images. Subtraction images from the dynamic sequence are shown in MPR format **(a–c)**. Angiomap and TIC demonstrate washout kinetics **(d, e)**. Histology yielded IDC grade 2, ER/PR (+), HER2/neu (−), Ki67: 10%.

6.3.3 Kinetic Classification

DCE-MRI kinetic techniques derived from imaging acquired over a standard time interval following contrast injection (5–7 minutes) include measures of the uptake and washout of contrast in tissues and contain diagnostically useful information. The shape of the signal-intensity-versus-time curve (TIC, or signal time-course, or kinetic curve), which plots signal intensity over time, has been found to be useful in the classification of enhancing lesions. Signal intensity is analyzed on a pixel-by-pixel basis within an enhancing lesion. TIC data can be obtained by using a manual technique, placing a region-of-interest (ROI) of at least 3 pixels on the most suspicious region of enhancement within an enhancing lesion. Changes in signal intensity are then monitored over time. Accurate analysis of TIC data strongly depends on predictable delivery of a GBCA using a bolus technique and a dose administered according to patient body weight. Most practices in the United States now utilize computer-aided analysis systems, allowing depiction of kinetic parameters on a pixel-by-pixel basis in parametric images describing intralesional variations in blood flow. These analytic tools can not only automatically display TICs, but also generate color maps of lesions that enhance above a set threshold. Thresholds are usually set between 50% (slow initial rise), 50 to 100% (medium initial rise), and >100% (fast initial rise). The parametric images reflect all lesions or tissues enhancing above a predetermined threshold below which no enhancement is measured, and exhibit “persistent” (continued increasing enhancement >10% above threshold), “plateau” (relatively constant signal intensity), and “washout” (decreasing signal intensity following peak enhancement >10% below threshold). Lesion color-coding of the delayed phase of enhancement can facilitate interpretation of the kinetic data (Fig. 6‑6).

**Fig. 6.6** Enhancement kinetics. A display of enhancement kinetics is shown.

Given that MRI is usually the last study in the diagnostic chain to be acquired, knowledge of the patient’s medical history and review of all prior imaging studies are essential for optimal interpretation of the breast MR examination.

6.4 Interpretation Challenges

Sources of error in breast MR interpretation may be attributed to technical incidents such as equipment malfunction and artifacts, host-related issues such as marked BPE or motion, and human errors of perception or misinterpretation. As discussed in ▒Chapter 5░, breast MRI is technically demanding and requires excellent fat saturation and high spatial and temporal resolution with rapid acquisition of postcontrast sequences. Technical errors in clinical practice that commonly affect interpretation include poor positioning, inadequate contrast injection, and patient motion. Careful assessment of image quality by the technologist and radiologist in routine practice is necessary to avoid these errors. Perception failures at screening account for missed cancers and may be exacerbated by marked BPE, which may mask small malignancies (Fig. 6‑7; Fig. 6‑8; Fig. 6‑9; Fig. 6‑10). Appropriate scheduling of the MR examination according to the timing of the patient’s menstrual cycle can often alleviate this problem. Small cancers are often best identified on the first postcontrast subtracted series in a standard acquisition, or on an ultrafast sequence where BPE is minimal (Fig. 6‑11; Fig. 6‑12). Small or even large in situ cancers may be difficult to detect even in the presence of mild BPE; high spatial resolution at 3 T facilitates detection of such cancers (Fig. 6‑13; Fig. 6‑14). Benign findings that may cause difficulty in interpretation and affect specificity negatively include certain lymph nodes, papillomata, fat necrosis (Fig. 6‑15; Fig. 6‑16; Fig. 6‑17), and fibroadenomata (Fig. 6‑18; Fig. 6‑19, Fig. 6‑20). These lesions may exhibit rapid enhancement, often with washout kinetics, careful morphologic analysis being necessary for accurate diagnosis; isotropic or near isotropic, MPR can be useful in these cases.

**Fig. 6.7** Asymmetric BPE. Example of asymmetric BPE: this is a normal finding.

**Fig. 6.8** Marked BPE: nodular pattern.

**Fig. 6.9** Marked BPE: nodular pattern—multiple bilateral benign enhancing masses representing known fibroadenomata are shown.

**Fig. 6.10** Marked BPE. nodular pattern.

**Fig. 6.11** Small IDC with DCIS. High-risk screening (BRCA 2), 4th module: MIP image reveals an irregular enhancing 8-mm mass in the lateral left breast (a, *arrow*). T1w postcontrast source and subtracted images show the irregular mass with associated posterior NME **(b, c)**. Angiomap shows heterogeneous, washout kinetics **(d)** and sagittal and coronal reformatted images are shown **(e, f)**. Histology yielded left breast IDC grade 2, size 8 mm, with DCIS within and adjacent to the mass. ER/PR (+), HER2/neu (–), Ki-67: 15%. Sentinel nodes were negative for malignancy (0/2).

**Fig. 6.12** Small ILC with pleomorphic LCIS. MIP image shows an irregular enhancing mass in the posterior aspect of the right breast (a, *arrow*), seen as an isointense mass on T2w image (b, *arrow*). T1w postcontrast source and subtracted images show mass enhancement with NME extending anterolaterally from the mass (**c, d**, *arrows*). Angiomap shows heterogeneous, washout kinetics **(e)** and sagittal and coronal reformatted images are shown **(f, g)**. Histology yielded ILC, 0.9 mm with associated pleomorphic LCIS, ER (+), PR (–), HER2/neu (–), Ki-67: 5–10%. Sentinel nodes were negative for malignancy (0/5).

**Fig. 6.13** Multifocal IDC and DCIS. A 43-year-old patient, MRI screening, normal mammogram BRCA mutation carrier 5th module: MIP image reveals extensive medial and central NME in the left breast **(a)**. Two irregular masses are seen on T2w and T1w postcontrast source images (**b, c** *arrows*). Subtracted image **(d)** exhibits two irregular enhancing masses with surrounding NME low (sub-threshold) enhancement was present. Sagittal and coronal reformatted slab images reveal the extent of NME **(e, f)**. Final histology (following MR-guided biopsy indicating IDC) yielded two IDC lesions 6 and 3 mm in size, grade 3, with associated extensive grade 3 DCIS, ER/PR (+), HER2/neu (–), Ki-67: 50–60%. Sentinel nodes were negative for malignancy (0/5).

**Fig. 6.14** Invasive cancer right breast and in situ cancer left breast. A 47-year-old patient, MRI screening, normal mammogram: MIP image of the **right** breast shows nodular marked BPE with a distinct small irregular mass noted (a, *arrow*). T2w image partially reveals an isointense mass (b, *arrow*), seen with robust enhancement on T1w source postcontrast image **(c)**. Spiculated mass margins are seen on subtracted image **(d)** and washout kinetics are noted on angiomap **(e)**. Sagittal and coronal reformatted subtraction images are shown **(f,g)**. Sagittal and coronal reformatted subtraction images are shown **(f,g)**. Histology yielded IDC grade 2, total extent 1.3 cm, with associated DCIS cribriform type and intermediate grade, ER/PR (+), HER2/neu (–). Evaluation of the **left** breast revealed normal MIP and T2w series. Careful review of the dynamic sequence revealed linear branching (ductal) enhancement, visible only on three slices. Precontrast T1w image exhibits segmental ductal fluid **(h)** and correlative postcontrast subtracted image reveals branching linear enhancement (i, *arrows*). Similar findings are shown on adjacent postcontrast source image **(j)** and subtracted image (k *arrows*). This finding is difficult to perceive, and high spatial resolution is necessary for diagnosis. Histology at simple mastectomy (following MRI-guided biopsy indicating DCIS) yielded low and intermediate DCIS, cribriform and solid type, ER/PR (+).

**Fig. 6.15** Intramammary lymph node (IMLN). MRI screening, normal mammogram: MIP image reveals mild BPE and a distinct 3-mm circumscribed enhancing mass in the right breast (a, *arrow*), seen as isointense on T2w image (b, *arrow*) and precontrast T1w image (c, *arrow*). Robust enhancement is seen on source postcontrast image **(d)** and subtraction image **(e)** with a minimal hilar notch noted (*arrow*). Washout kinetics are seen on angiomap **(f)** and reformatted sagittal and coronal images are shown **(g, h)**. IMLNs usually present with washout kinetics and identification of benign morphology is essential for diagnosis.

**Fig. 6.16** Papilloma. A 71-year-old patient, BRCA 2 carrier MRI screening 10th module: MIP image **(a)** reveals a new distinct 6-mm circumscribed enhancing mass in the right breast, 10 o’clock position anterior depth (a, *arrow*), seen as isointense on T2w image (b *arrow*). Robust enhancement is seen on source postcontrast image **(c)** and subtraction image **(d)** with persistent kinetics identified on angiomap **(e)**. Reformatted sagittal and coronal images are shown **(f, g)**. Despite benign morphology, MR-guided biopsy was performed with results yielding sclerosing papilloma without atypia.

**Fig. 6.17** Fat necrosis. Age 65, S/P RT lumpectomy 3 years ago for treatment of IDC grade 3: routine follow-up staging PET examination revealed uptake in the posterior right breast **(a)** at the site of the lumpectomy scar. T2W axial image with sagittal and coronal reformatting reveals an irregular mass at the scar site with central fat signal (**b–d**, *arrows*). Mass with central fat signal is seen on precontrast image **(e)** and rim enhancement and adjacent anterior NME is noted on postcontrast, source image (f, *arrow*) and rim enhancement on adjacent subtraction image **(g)**. Heterogeneous NME with washout kinetics is noted on angiomap **(h)**. Sagittal and coronal images are shown **(i, j)**. MR-guided biopsy yielded fat necrosis. This case emphasizes the importance of multiplanar reformatting for accurate lesion assessment.

**Fig. 6.18** Fibroadenoma. A 37-year-old, personal history of grade 3 DCIS treated with mastectomy 5 years ago. 5th MRI screening module: MIP image reveals autologous reconstruction of the right breast and a new 3-mm mass in the anterior left breast at 6 o’clock position (a, *arrow*). T2w image **(b)** shows the reconstructed right breast but no abnormal findings. Source and subtraction postcontrast T1w images identify the 3-mm enhancing mass (**c, d**, *arrows*), also seen with persistent enhancement on angiomap **(e)** and on sagittal and coronal reformatted image (**f, g**, *arrows*). MR-directed ultrasound identified a correlative oval mass, biopsy yielding fibroadenoma **(h)**.

**Fig. 6.19** Fibroadenoma. A 55-year-old patient. MIP image **(a)** reveals a circumscribed 13-mm enhancing mass with internal unenhancing septations, seen as isointense on T2w image (b, *arrow*). Precontrast T1w image reveals high signal fluid within a dilated duct which is compressed by the mass (c, *arrow*). Source postcontrast T1w image **(d)** shows both the high duct signal and the enhancing mass with unenhancing septations. The mass is also shown on subtraction image **(e)**, and persistent mass enhancement is seen on angiomap **(f)**. MR-directed ultrasound shows the benign mass partially obstructing the dilated duct **(g)**. Ultrasound-guided biopsy yielded fibroadenoma.

**Fig. 6.20** Subareolar papilloma and myxoid fibroadenoma. Patient age 54 years, routine screening. MIP image reveals a multilobulated mass in the right breast subareolar region **(a)**. Two lesions are present in the right breast. T2w image **(b)** reveals hyperintensity in the subareolar region (*arrow*) and a high signal oval mass is seen at 2 o’clock position **(c)**. T1w postcontrast source images show robust subareolar enhancing masses **(d)** and minimal enhancement in the mass at 2 o’clock position (e, *arrow*), also seen on subtraction images (f, g, *arrow*). Heterogeneous mass enhancement is noted with washout kinetics on angiomap **(h)** and persistent mass enhancement is on angiomap **(i)**. The subareolar location of the multilobulated circumscribed mass with washout kinetics is consistent with the subsequent ultrasound-guided biopsy yielding papilloma. The circumscribed mass at 2 o’clock position with minimal enhancement was also biopsied, yielding myxoid fibroadenoma.

6.4.1 Skin Enhancement

Skin enhancement may be seen at the site of recent percutaneous biopsy but may also reflect spread of malignancy to the dermis. Malignant skin involvement may be evident remote from the site of a newly diagnosed cancer or may directly involve the dermis. Cowden’s syndrome is a rare autosomal dominant inherited disorder characterized by diagnosis of multiple hamartomas and is associated with a predisposition for breast carcinoma. An example of a patient with Cowden’s syndrome and an aggressive inflammatory breast cancer (IBC) with direct skin involvement is shown in Fig. 6‑21.

**Fig. 6.21** IBC with skin involvement. Patient, age 37, with known Cowden’s syndrome presented with a painful anterior right breast mass with skin ulceration: marked BPE and a large anterior enhancing breast mass extending medially to the skin, with two additional small enhancing masses noted posterior to the index lesion, are seen on MIP image (a, *arrows*). T2w axial image **(b)** shows a hyperintense subareolar mass with a biopsy marker clip visible just below the skin (*arrow*). Sagittal reformatted T2w image **(c)** demonstrates the mass, also seen on the coronal T2w image, with an extruded marker clip and skin thickening (d, *arrow*). Precontrast T1w image reveals heterogeneous isointensity in the mass with an associated medial fluid collection and skin thickening **(e)**. Postcontrast image **(f)** shows the mass to enhance with two posterior satellites also identified; similar findings are noted on subtraction image **(g)** where prominent skin enhancement is noted overlying the mass and adjacent fluid collection is also present. Reformatted sagittal image **(h)** reveals rim enhancement in a satellite lesion (*arrow*). Coronal image **(i)** is shown. Final histology yielded IBC grade 3 triple negative, with squamous differentiation and necrotic inflammatory debris, Ki-67: 60%.

6.4.2 Inflammatory Changes

Inflammation due to mastitis may be focal or diffuse. Detection of segmental NME is usually associated with ductal carcinoma in situ (DCIS); benign diagnoses with this finding and distribution are uncommon. An example of inflammatory changes presenting as segmental NME, mimicking DCIS, is shown in Fig. 6‑22.

**Fig. 6.22** Benign segmental NME (1.5 T). Patient, age 54, presented with pain and a palpable “thickening” in the left breast; mammography and ultrasound were normal. MIP image reveals extensive NME in the central and lateral left breast with increased vascularity **(a)**. T2w image **(b)** is normal; however, diffuse NME is identified on postcontrast T1w source and subtracted images **(c, d)** and on reformatted sagittal and coronal images **(e, f)**. Percutaneous biopsy revealed periductal inflammation with histiocytes and poorly formed granulomas in association with ruptured ductules. Clinical findings subsided and follow-up MRI revealed no abnormal findings, as shown on MIP image **(g)**.

6.5 Postsurgical Complications

The T2w series is very helpful for evaluation of the posterior breast and chest wall, particularly when complications from implant augmentation or reconstructive surgery result. An example of postsurgical complications following expander placement for breast reconstruction is shown in Fig. 6‑23.

**Fig. 6.23** Expander dehiscence; patient aged 50 was diagnosed with right breast DCIS, and was treated with mastectomy and reconstructive surgery. A tissue expander was placed and then removed because of complication with infection. MIP image shows the mastectomy site **(a)**. T2w axial image **(b)** and postcontrast source image **(c)** reveal two fluid collections, also identified on sagittal reformatted images **(d, e)**. The fluid collections were drained and resolved with antibiotic treatment.

6.6 The Male Breast

Breast cancer in men is rare, accounting for less than 1% of breast cancers. Physiological gynecomastia occurs in neonates and puberty, and with obesity and ageing. Gynecomastia can be caused by an increased estrogen to testosterone ratio in men treated with estrogen therapy for prostate cancer and by a variety of other medications. Mammography is usually the recommended test when men are referred for imaging. An example of gynecomastia on MRI is shown in Fig. 6‑24.

**Fig. 6.24** Gynecomastia. Patient aged 62 presented clinically with painful bilateral gynecomastia, left breast greater than the right breast. Precontrast T2w and T1w images **(a, b)** show normal-appearing breast tissue behind the nipple, with a greater amount of tissue noted in the left breast. Postcontrast source and subtraction images reveal minimal enhancement of breast tissue **(c, d)**. No signs of malignancy are present.

6.7 Nipple Enhancement

Normal nipple enhancement can mislead the reader because the nipple often enhances with varying intensity due to the rich blood supply of the nipple–areolar complex, and the enhancement may not necessarily be symmetric. The radiologist should be careful when the nipple is inverted because normal nipple enhancement may simulate a subareolar enhancing mass. Abnormal nipple enhancement is found in patients with Paget’s disease, IBC, lymphatic obstruction, and inflammation. An example of an epidermal inclusion cyst within the nipple is shown in Fig. 6‑25.

**Fig. 6.25** Epidermal inclusion cyst in the nipple. Patient age 41, routine MR screening. MIP image reveals moderate BPE and linear enhancement in the left nipple **(a)**, also seen on T1w subtraction image **(b)**, with persistent kinetics shown on angiomap **(c)**. The finding is also noted on reformatted sagittal image **(d)**.

6.8 Breast Carcinoma in Augmented Breasts

Breast implant surgery has been performed routinely for augmentation and reconstruction purposes for over 50 years, employing a large variety of devices including saline, silicone double-lumen types using both saline and silicone, and polyacrylamide gel. Contrast-enhanced MRI is necessary for cancer screening or diagnosis of suspected tumor. A causal relationship between malignancy and implant placement has not been found. ¹ Breast carcinomas often contact the implant surface and tumors may spread along the contour of the implant. The use of MR MPR in such cases is particularly helpful for surgical planning. Examples of two patients with saline implants and associated malignancy are shown in Fig. 6‑26 and Fig. 6‑27.

**Fig. 6.26** ILC with a retropectoral saline implant. Patient age 52 presented with a palpable right breast mass. Partial MIP image **(a)** reveals an enhancing spiculated mass in the lateral posterior right breast with a second smaller anterior mass (*arrow*) and surrounding NME extending anteriorly. T2w axial image **(b)** exhibits a hypointense mass (*arrow*) and bilateral retropectoral saline implants. Source images **(c, e)** and subtracted images **(d, f)** show to advantage the satellite lesion (*arrows*) and the spiculated index lesion. Angiomap **(g)** demonstrates persistent kinetic mass enhancement, and sagittal source reformatted image **(h)** shows the index mass (*arrow*). Subtracted coronal reformatted image **(i)** reveals enhancing tumor spreading along the implant surface (*arrows*). Final histology yielded ILC grade 2 with associated LCIS, classic type. ER/PR (+), HER2/neu (–), Ki-67: 10%. One sentinel node was positive for malignancy (1/20).

**Fig. 6.27** IDC with a retropectoral silicone implant. Patient age 41 with a retropectoral silicone implant presented with a palpable **left** breast mass at 2 o’clock position. She underwent prior **right** breast cancer treatment (IDC and DCIS) with skin-sparing mastectomy, and reconstruction with a silicone implant. T2w axial image **(a)** exhibits bilateral silicone implants and a round hyperintense mass in the left breast corresponding with the palpable mass (*arrow*), also seen on reformatted sagittal and coronal planes (b, c, *arrows*). Source and subtracted axial images **(d, e)** show rapid heterogeneous enhancement within the mass (*arrows*), and angiomap (f) demonstrates heterogeneous washout mass enhancement (*arrow*), also seen on reformatted sagittal and coronal planes **(g, h)**. Final histology yielded IDC grade 3 with necrosis. ER (+) PR (–), HER2/neu (–). All axillary lymph nodes were negative for malignancy (0/13).

Breast implant–associated anaplastic large cell lymphoma (BIA-ALCL) is a rare, distinct type of T cell lymphoma which develops around implants, causing pain and breast swelling and, less commonly, a palpable breast mass. ² The underlying mechanism of this disease is thought to be due to chronic inflammatory change resulting from indolent infections, leading to malignant transformation of T cells that are anaplastic lymphoma kinase (ALK) negative and CD30 positive. Mean time to presentation is about 10 years following augmentation surgery and fluid is shown to develop around the implant. Immunohistochemistry confirms the diagnosis BIA-ALCL with CD30+ and ALK– expression. In most cases, surgical treatment is curative and includes capsulectomy and removal of the implant. When disease is more advanced, chemotherapy, radiotherapy, and lymph node dissection may be necessary.

6.9 Benign Papillary Lesions and Nipple Discharge

Papillary lesions of the breast represent a diverse group of lesions that share a common frond-like growth pattern with epithelium supported by a fibrovascular stroma. These breast lesions include benign papilloma, papillary DCIS, and invasive papillary carcinoma. Myoepithelial cells line the basement membrane in benign papillomata and papillary DCIS lesions, but are absent in invasive papillary carcinoma.

6.9.1 Solitary Intraductal Papillomata

These lesions are usually identified centrally in the large lactiferous subareolar ducts and cause bloody or serous nipple discharge when symptomatic. They are often nonpalpable, ranging from 3 to 5 mm in size, and are frequently mammographically occult. ³ The dominant feature of large duct papilloma is that of a circumscribed mass with or without calcifications on mammography, and a circumscribed hypoechoic mass arising within an ectatic duct or a complex mass with increased blood flow on ultrasound. Papilloma may be associated with a dilated duct on both mammography and ultrasound. ⁴ MRI features similarly identify a circumscribed subareolar mass smaller than 1 cm, with rapid homogeneous enhancement and varied kinetics in the delayed phase. ⁵ ^, ⁶ Dilated ducts exhibiting high T2w and T1w signal caused by hemorrhage may be associated with a papilloma. ⁷ Although the morphologic characteristics of most papillomata suggest benignity, the kinetic finding of washout in the delayed phase is not uncommon and differentiation from malignancy may require tissue sampling in some cases. Management of large duct papilloma generally depends on whether atypia is associated with the histologic diagnosis. ⁷ ^, ⁸ ^, ⁹ A diagnosis of papilloma with atypia confers a much higher patient risk (7.5 times) than that of a papilloma without atypia. ¹⁰ In clinical practice, most papillomata diagnosed on core biopsy with a finding of associated atypia are excised. There are no consensus recommendations for the management of papilloma without atypia, and in many cases follow-up with clinical examination and imaging is recommended.

6.9.2 Multiple Intraductal Papillomata (Multiple Peripheral Papillomatosis)

These lesions are much less common, arise from the terminal ductal lobular unit, and are usually found in the periphery of the breast. Despite their different site of origin and multiplicity, the imaging findings on MRI are similar to those of solitary large duct papilloma. A hereditary condition known as juvenile papillomatosis, first described by Rosen in adolescents and young women, confers a slightly higher breast cancer risk, the imaging findings being similar to those of multiple intraductal papillomata. ¹¹

6.9.3 Clinical Nipple Discharge

The standard-of-care imaging protocol for most women with spontaneous bloody or serous nipple discharge consists of imaging with mammography, ultrasound, and galactography. These studies identify the causative lesion in many, but not all cases, and the gold standard for treatment of patients with nipple discharge and negative imaging findings is surgery with large duct excision. ¹² ^, ¹³ Most causes of nipple discharge are benign; a report on 586 patients who underwent surgery for significant nipple discharge stratified the breast pathology as follows: papilloma 48%, fibrocystic change 33%, cancer 14%, and high-risk lesions 7%. ¹⁴ Studies have shown that the performance of breast MRI is superior to that of galactography for detection and characterization of papillary lesions, ¹⁵ ^, ¹⁶ and radiologists increasingly use MRI for definitive assessment of women with nipple discharge (Fig. 6‑28; Fig. 6‑29). The increased use of MRI for this indication has come into play for two main reasons, both resulting in change of management. Firstly, the addition of MRI to the presurgical imaging protocol results in higher sensitivity than other standard imaging tests and may detect unsuspected malignant disease not associated with the known papilloma, thus affecting patient management (Fig. 6‑30). Secondly, absence of significant breast enhancement would likely preclude diagnosis of cancer given the high negative predictive value of MRI, and perhaps obviate the need for surgical duct excision, unless the discharge were to be too troublesome for the patient. Further clinical studies are needed; however, thought should be given to the management of women with nipple discharge now that we have a test with exquisite sensitivity and a very high negative predictive value. A reasonable consideration could be to allow a patient the choice of surgery or follow-up when MRI is negative.