Breast Cancer

Invasive breast tumors are characterized by an ingrowth of neovascularity at their periphery. Tumor angiogenesis is associated with increased perfusion and abnormal leaky endothelium, leading to preferential enhancement of tumors versus normal breast tissue (Box 7-1). With bolus administration of an intravenous (IV) contrast agent, increased vascular flow and the rapid exchange rate of contrast between blood and the extracellular compartment cause invasive breast tumors to enhance more rapidly and more avidly than normal fibroglandular tissue, even in patients with dense breasts. Thus, invasive breast cancers have high signal intensity and are brighter than the surrounding normal tissue on the first postcontrast scan, which ideally should be obtained about 90 seconds after injection. As a result, MRI exquisitely reveals invasive tumors that are occult on mammography (Fig. 7-1). The sensitivity of MRI for invasive breast cancer is extremely high—over 90%. However, as discussed in detail in this chapter, contrast enhancement on MRI is seen in many benign conditions as well; the specificity of MRI varies between 39% and 95%. As detailed in this chapter, morphology, T1 and T2 characteristics, and the time course of contrast enhancement help differentiate benign from malignant lesions (Table 7-1).

Figure 7-1 Mammographically occult breast cancer. A, Mediolateral oblique mammogram in a woman with a palpable mass in the upper portion of the breast indicated by a metallic skin marker (arrow) revealed only dense tissue. B, Contrast-enhanced water-specific 3-D magnetic resonance image demonstrated a 1-cm rim-enhancing lesion (arrow). Lumpectomy revealed a 1-cm invasive ductal carcinoma.

Patient Preparation

Benign hormone-related enhancement of normal breast tissue, called background enhancement, occurs before the onset of menses and can lead to false-positive studies. When possible, patients should be imaged 7 to 10 days after the onset of their menstrual cycle, when spurious contrast enhancement of normal breast tissue is at its nadir (Box 7-2).

Before MRI scanning, the patient fills out an MRI safety form to exclude contraindications of entering the strong magnetic field, such as ferromagnetic vascular clips, metallic ocular fragments, pacemakers, and implanted electromechanical devices. A qualified person reviews the standardized MRI safety form before scanning (Fig. 7-2).

Figure 7-2 A and B, Magnetic resonance imaging (MRI) safety form. As part of routine safety interview procedures before MRI, patients fill out a history form to screen for implanted devices or other conditions that might affect the safety of MRI.

(Courtesy of Dr. Frank Shellock, MRI-Safety.com. Reprinted by permission.)

As with mammography, an MRI-specific breast history form is helpful to detail patient breast risk factors, family history, breast lumps, scars, or other areas of complaint. The technologist places MRI-compatible markers on the patient’s breast to indicate lumps or areas of concern and annotates them on the history form. The patient details the location, date, and results of previous breast biopsies because recent healing breast biopsies may normally show enhancement and are a cause of false-positive results. The patient also documents any use of exogenous hormone therapy and the phase of the menstrual cycle or menopause, because those factors may cause spurious background enhancement of normal tissue, which can produce false-positive results.

Equipment

An IV catheter is placed before scanning and is continuously flushed by using the keep vein open (KVO) setting of an MRI-compatible remote power injector. Placement of the catheter in the antecubital fossa contralateral to any known, previous, or suspected malignancy is preferred. The patient is placed prone on a dedicated breast coil (Fig. 7-3A). Prone positioning minimizes respiratory motion in the breast. Phased-array breast coils maximize the signal-to-noise ratio of the image. Patient discomfort is the primary cause of motion; the majority of patients remain most comfortable for the duration of the entire scan with both arms at their sides and wearing hearing protection (see Fig. 7-3B). The technologist spends considerable time discussing the importance of “holding still” with the patient to obtain the best scan. The patient then works with the technologist to obtain a comfortable position within the breast coil. Optional mild breast stabilization, or “compression,” may be used to reduce breast motion and decrease the volume of tissue to be scanned so that the whole breast is included. However, firm compression (as used routinely for mammography) should be avoided because it may negatively affect contrast enhancement. Scanners with a magnetic field strength of 1.5 Tesla (T) or 3.0 T provide the best signal-to-noise ratio. Magnets with high-performance gradients enable the fastest, highest-resolution scans (Box 7-3).

Figure 7-3 Dedicated breast coil and 1.5-Tesla (T) scanner. Prone positioning with the breasts in the apertures of a dedicated, phased-array breast coil (A) improves image quality by maximizing the signal-to-noise ratio of the image and minimizing respiratory motion. Imaging at 1.5 T enables more robust fat suppression and produces a higher signal-to-noise ratio than imaging at lower field strengths does. B, Conventional closed-bore scanners have stronger, faster gradient systems than open magnetic resonance imaging systems do; these systems enable faster scans with more slices and higher spatial resolution. Use of a remote power injector (arrow) enables administration of contrast during the dynamic scan protocol while the patient is within the magnet.

(A, Courtesy of Tom Tynes, MRI-Devices Inc, Waukesha, WI. Used by permission.)

MRI Protocols

Conventional breast MRI begins with T1-weighted images to define the position and anatomy of the breast. T1-weighted images using the signal from the “body coil” rather than the breast coil enable basic evaluation of the axillae, anterior mediastinum, chest wall, and supraclavicular fossa for enlarged regional lymph nodes. Thereafter, a dedicated breast coil signal should be used to perform all subsequent sequences. T2-weighted fast spin-echo (FSE) images are then obtained to characterize the breast and any lesions. T2-weighted scans using an FSE, turbo spin-echo (TSE), or rapid acquisition with relaxation inhibition (RARE) technique, produce high-quality images within reasonable scan times of 5 to 6 minutes. High fat signal on T2-weighted FSE images can be prevented with fat suppression and is most successful if unilateral scanning is performed (Table 7-2).

Table 7-2 Basic Bilateral Protocol for Breast Cancer MRI

Sagittal imaging for all scans except series 1 allows smallest field of view (∼20 cm for most patients) and thus highest resolution.

Frequency encoding in the anteroposterior direction minimizes artifacts from cardiac and respiratory motion in the breast.

Series in brackets are optional, and not yet standard of care.

1H, proton (hydrogen nucleus); ADC, apparent diffusion coefficient; EPI, echo-planar imaging; FOV, field of view; FSE, fast spin echo; Gd-DTPA, gadolinium diethylenetriamine pentaacetic acid; IPAT, integrated parallel acquisition techniques; NEX, number of excitations or signal averages; RARE, rapid acquisition with refocused echoes; SENSE, sensitivity encoding; STIR, short tau inversion recovery; TE, echo time; TR, repetition time; TSE, turbo spin echo.

* T2: Fast spin-echo (FSE, RARE, TSE, etc.) with effective TE 80–100 ms and TR at least 3000 ms provides good T2 weighting. Use 3- to 4-mm-thick slices and 256×192 matrix or higher for small FOV sagittal images. Fat saturation improves conspicuity of bright lesions on T2, although nonfat-suppressed T2 imaging allows fat signal intensity to be used as a reference signal intensity. Volume shimming improves fat suppression.

^† B-values have not been standardized yet, but most investigators use 500–1000 with good results. Parallel imaging may reduce distortions in echo planar imaging.

^‡ Both fat suppression and high spatial resolution (<2 mm in all directions) are essential to assess lesion morphology. Rapid imaging (60–90 sec per scan or less) is necessary to assess contrast uptake kinetics. Use “fast” spoiled 3-D gradient echo (TR ≤ 6 ms; FA ∼ 15° for T1 weighting). Intermittent “special” fat saturation pulses speed imaging substantially over conventional fat saturation. Fractional k-space (“1/2 NEX,” etc.), and parallel imaging (SENSE/IPAT, etc.) maximize resolution obtained during the limited scan time.

^§ Repeat rapidly for dynamic scans totaling approximately 7 minutes or more. Inject 0.1 mmol/kg standard low molecular weight gadolinium contrast agent (e.g., Gd-DTPA, gadoteridol, etc.) at 2 mL/sec followed by 20 mL flush (normal saline) at the start of the acquisition. Avoid negatively charged gadolinium agents if protocol includes spectroscopy, because they may reduce choline signal. (Lenkinski B, Wang X, Elian M, Goldberg SN: Interaction of gadolinium-based MR contrast agents with choline: implications for MR spectroscopy (MRS) of the breast, Magn Reson Med 61(6):1286–1292, 2009.

^¶ Single-voxel choline spectroscopy. A minimum voxel size of 1 cm × 1 cm × 1 cm is recommended for adequate signal-to-noise. Localized shimming and high-quality spatial saturation pulses, fat suppression, and partial water suppression improve quality of spectra.

Considerable variation exists worldwide in methods used for the contrast-enhanced portion of the examination. Most investigators agree that both the time course of enhancement provided by dynamic scanning and the morphology of lesions revealed by high-spatial resolution scanning provide distinct and useful information about the risk of malignancy in enhancing lesions. However, commercially available MRI pulse sequences necessitate a compromise between the dynamic and high-spatial resolution approaches. Currently, most scans are done with repeated T1-weighted, fat-saturated, 3-D spoiled gradient-echo (SPGR) scans. Slice thickness and resolution are selected to give maximum resolution and bilateral whole-breast coverage within a 60- to 90-second scan duration. Parallel imaging and intermittent/partial fat saturation substantially speed up imaging, allowing much higher resolution within the same scan time.

Volumetric T1-weighted 3-D SPGR imaging is repeated as rapidly as possible before, during, and for approximately 5 to 7 minutes after administration of a rapid IV bolus of 0.1 mmol/kg gadolinium contrast agent. Most investigators use axial or coronal images with a rectangular field of view to maximize efficiency. Images may be processed by subtracting the precontrast baseline images from subsequent dynamic images to reveal areas of enhancement. Region-of-interest analysis is used to assess the time course of contrast enhancement. Subtraction processing suppresses signal from bright fat because adipose tissue does not enhance significantly. Spatial resolution is limited by the need for rapid scan times and the large field of view required for bilateral scanning.

Proper shimming of the magnet and choice of center frequency are essential to ensure adequate fat suppression. Bilateral high-spatial resolution imaging is now possible on most scanners because of the development of hardware and software required for bilateral shimming and because coverage is possible with reasonable scan times.

In bilateral combined dynamic and high-spatial resolution imaging, sophisticated protocols and scanning techniques capture both rapid dynamic and high-spatial resolution images of both breasts during and after IV contrast injection. Approaches include time-resolved imaging of contrast kinetics (TRICKS) and interleaved protocols, in which dynamic scanning is interrupted for high-spatial resolution imaging. These protocols provide very high-quality images but they require rapid switching between different pulse sequences.

The American College of Radiology (ACR) now has guidelines for the performance of breast MRI and in 2010 launched a breast MRI accreditation program. The ACR MRI program recommends that facilities be able to obtain bilateral studies and be able to perform MRI-guided biopsies.

Normal Breast MRI Findings

Typical images from a normal patient who underwent both bilateral dynamic and high-spatial resolution imaging are provided in Figure 7-4. On T1-weighted noncontrast-enhanced images, aqueous tissues (including skin, fibroglandular tissue, muscle, and lymph nodes) have moderately low signal intensity when compared with the higher signal intensity of fat, which has a short T1 relaxation time. In the absence of previous surgery or pathology, a layer of subcutaneous and retromammary fat completely surrounds the mammary gland tissue except where it enters the nipple–areola complex. The mammary gland itself is composed of a mix of low-signal fibroglandular tissue and high-signal fat lobules. The mix and distribution of fat and fibroglandular tissue vary greatly between patients—from dense, uniformly glandular tissue with almost no visible fat, to heterogeneous, to predominantly fatty tissue separated by thin strands or septa of fibroglandular tissue. In the ACR Breast Imaging Reporting and Data System (BI-RADS®) MRI lexicon the amount of dense glandular tissue by volume is described in the same terms as used in the mammography lexicon. These include almost all fat (0% to 25% dense), scattered fibroglandular tissue (25% to 50% dense), heterogeneously dense (50% to 75% dense), or dense (>75% dense) (Box 7-4 and Fig. 7-5).

Figure 7-4 Normal breast magnetic resonance imaging (MRI). A, Axial T1-weighted spin-echo images reveal high-signal fat within the breast and axilla. Soft tissues, including skin, fibroglandular tissue, lymph nodes, and muscles, are dark. B, Sagittal, fat-saturated, T2-weighted fast spin-echo images reveal dark signal within fat and moderately low signal within the pectoralis muscle. Glandular tissue has mixed T2-weighted signal intensity. C, Precontrast water-specific 3-D spectral-spatial excitation magnetization transfer (3DSSMT) reveals dark fat and moderately low fibroglandular and muscle tissue signal. D, High-spatial resolution 3DSSMT after the intravenous administration of a 0.1-mmol/kg bolus of contrast reveals enhancing peripheral vessels (arrows) and mild nipple enhancement. E, Rapid dynamic 3-D spiral imaging reveals mild (<50%) gradual enhancement in the nipple and fibroglandular tissue and mild (<50%) rapid enhancement in the pectoralis muscle (F). Subtraction processing of early postcontrast multiplanar two-dimensional spoiled gradient echo images (G) minus precontrast baseline images (H) reveals diffuse low-level fibroglandular and muscle enhancement.

Figure 7-5 Examples of normal breast magnetic resonance imaging (MRI) density by volume on noncontrast sagittal 3-D spectral-spatial excitation magnetization transfer (3DSSMT) scans. A, Fatty sagittal postcontrast MRI shows mostly fatty tissue (0% to 25% density). B, Scattered fibroglandular tissue (25% to 50%). C, Heterogeneously dense breast tissue (50% to 75%). D, Dense glandular tissue (>75%).

T2-weighted noncontrast-enhanced images reveal heterogeneous fibroglandular tissue that is usually higher in signal intensity than adjacent muscle but still not as bright as the small subcutaneous blood vessels commonly seen at the periphery of the breast or as pure fluid (i.e., cysts, ducts).

After contrast injection, normal glandular tissue enhances to variable degrees. Normal fibroglandular breast tissue enhancement is called background enhancement in the BI-RADS® lexicon. An understanding of normal background enhancement is important because normal background enhancement can obscure cancers and make the MRI harder to read. Background enhancement describes slowly enhancing breast tissue within the breast. Specifically, normal fibroglandular tissue enhances with nonmasslike patterns, including stippled enhancement (tiny <5 mm foci of enhancement separated by normal tissue); scattered, regional, or multiple regions; or diffused stippled enhancement throughout both breasts. The areas of enhancement are usually separated by nonenhancing normal breast tissue between the normal background enhancing foci. In the BI-RADS® lexicon, the amount of background enhancement is described as a percentage of enhancing breast tissue with respect to the volume of the entire breast. Background enhancement is categorized in quartiles. The descriptors include none (0%), minimal (1% to 25%), mild (25% to 50%), moderate (50% to 75%), and marked (>75%) (Box 7-5 and Fig. 7-6).

Figure 7-6 Background enhancement on contrast-enhanced breast magnetic resonance imaging (MRI). A, No abnormal background enhancement. Contrast-enhanced breast MRI shows that the scant glandular tissue does not enhance. There is enhancement of an axillary lymph node. B, Minimal enhancement. Scattered fibroglandular tissue with minimal background enhancement. C, Mild enhancement. Scattered fibroglandular breast tissue with mild background enhancement. D, Heterogeneously dense tissue with moderate background enhancement. E, Marked enhancement. Dense glandular tissue with marked background enhancement.

The amount of background enhancement depends on the patient’s hormonal status, stage of the menstrual cycle, and whether the patient is premenopausal or postmenopausal. For example, in premenopausal women, normal breast tissue enhances the most right before the onset of menses. The normal breast tissue enhances the least 7 to 10 days after the onset of menses. Postmenopausal women on exogenous hormone therapy may enhance greatly. However, the amount of background enhancement depends not only on how much the tissue enhances, but also on the actual volume of dense breast tissue present. For example, a woman with extremely dense breast tissue, in which most of the breast is composed of fibroglandular tissue, may enhance from 0% to 100%. This woman might have minimal to marked background enhancement, because more than 75% of her breast is capable of enhancing. However, if the breast has only scattered fibroglandular tissue (25% to 50% dense), the greatest amount of background enhancement possible is mild background enhancement, or 25% to 50% enhancement, because only 25% to 50% of the breast is dense enough to enhance within and the remaining 50% is fat. Of note, cancers can occur in both dense and fatty parts of any breast.

Reporting normal background enhancement gives a clinician and other radiologists an idea of how likely it would be for the normal enhancing structures to hide a breast cancer and how confident the radiologist is in detecting breast cancer on the MRI.

After contrast injection, peripheral small subcutaneous vessels, the nipple, and adjacent retroareolar tissue enhances to variable degrees. The normal nipple enhances along its edge in a thin line; the rest of the nipple is dark, with an occasional dot of enhancement centrally within the nipple itself. The areolar complex is dark on the normal MRI. The normal skin is 2 to 3 mm in thickness and enhances slightly.

The rate and degree to which findings enhance gives a clue to whether the finding is normal or malignant. Dynamic imaging shows that fibroglandular tissue enhances mildly and slowly. Nipple enhancement curves are more avid, but still gradual. Muscle enhances rapidly at first, but never enhances very avidly. Subtraction imaging of a normal breast should show mild glandular and muscle enhancement. Usually cancer is much brighter than the background enhancement and will look much brighter on the first postcontrast scan compared to normal tissue. Normal breast tissue kinetic curves show a slow initial rise and a late persistent plateau. This normal slow initial and late persistent curve helps to distinguish normal breast tissue from cancers. Cancers will usually enhance rapidly initially, with a washout or a plateau of signal intensity in the delayed phase.

Common Breast MRI Artifacts

Ghosting from cardiac or respiratory motion occurs in the phase-encoding direction (Table 7-3). It can be prevented from obscuring breast tissue by careful selection of phase- and frequency-encoding directions. Poor fat suppression is usually due to poor shimming or incorrect choice of the excitation center frequency, especially in patients with silicone implants (Fig. 7-7) or non–MRI-compatible objects, such as BB markers, magnetic tissue expanders, scar markers, and metal infusion ports, in or near the breast. Patient motion may cause blurring of the image, so it is especially important that the patient hold still and breathe quietly during scanning. On subtraction imaging, patient motion causes alternating bright and dark bands at fat–glandular tissue interfaces (see Fig. 7-7).

Table 7-3 Common Artifacts

Figure 7-7 Common imaging artifacts. A, Inconsistent fat suppression on water-selective images is due to poor shimming or incorrect selection of the center frequency (arrow), or both. B, Repeat shimming and tuning improve fat suppression. C, Poor fat suppression on subtraction images is due to patient motion, causing misregistration artifacts at fat–skin or fat–glandular interfaces (arrows).

Poor breast tissue enhancement may be due to failed contrast injection, which can be confirmed by abnormal dynamic enhancement curves from the heart. The heart usually shows normal rapid, avid initial enhancement and rapid washout. In fact, one of the most common errors in contrast-enhanced breast MRI is the result of a poor bolus of IV contrast. Sometimes, the IV line is injected slowly or may even become detached from the vein; as a result, the contrast never enters the patient. To ensure rapid uptake of contrast and washout of signal intensity within the heart, one checks a region of interest over the heart or a large artery for a rapid intake bolus and late washout on the kinetic curve. This is important because only a good bolus of contrast will translate into breast cancer enhancing rapidly on the MRI scans. An abnormal or poor cardiac kinetic curve tells the radiologist that something is wrong with the contrast injection; further investigation is needed to determine why contrast in the heart did not rise in signal intensity rapidly and washout as expected. Scans that show poor contrast enhancement in the heart indicate a problem with the contrast injection and cannot be trusted to show cancer in the breast (Fig. 7-8).

Figure 7-8 Invasive ductal cancer (IDC) on ultrasound and signal void on postcontrast sagittal 3-D spectral-spatial excitation magnetization transfer (3DSSMT) magnetic resonance imaging (MRI) scans, missed because of late imaging. A, In the outer left breast, a focal asymmetry with possible spiculations (arrow) within dense fibroglandular tissue is seen only on the craniocaudal view. B, Left breast ultrasound shows a very hypoechoic irregular suspicious breast mass corresponding to the density seen in part A. C, Postcontrast sagittal 3DSSMT shows heterogeneously dense breast tissue and marked background enhancement in the left breast. There is signal void from the marker placed in the IDC in the upper breast, but the cancer itself is not seen, possibly obscured by the marker or the surrounding background enhancement. However, it was noted that the scan was delayed and the imaging was done in a later phase approximately 10 minutes after injection. A repeat MRI was done. D, Postcontrast sagittal repeat 3DSSMT shows marked background enhancement in the left breast, but because this scan was done at minutes, it shows an enhancing tumor at the signal void (arrow). E, In addition, a 6-mm mass is noted inferior to the main cancer; this was not seen previously and was undetectable in the background enhancement (arrow).

Breast Lesions: Approach and Lexicon

Initial studies of contrast-enhanced MRI reported a sensitivity of more than 90% for invasive breast cancer. The sensitivity of contrast enhancement has remained high for invasive breast cancer. However, achieving high specificity remains difficult because some benign breast conditions enhance more avidly than normal breast tissue and may resemble breast cancer. Specifically, benign fibroadenoma, papilloma, and proliferative fibrocystic change also enhance to a greater degree than normal surrounding breast tissue.

To interpret studies of lesions on MRI, radiologists use a combination of high-spatial resolution scanning (which produces sharp images for analysis of abnormally enhancing findings or produces analysis by lesion morphology) and dynamic imaging (which produces the kinetic curves of an abnormally enhancing finding). In addition, T2-weighted imaging plays a secondary role in distinguishing some benign and malignant lesions (Table 7-4).

Table 7-4 T2-Weighted Imaging of Breast Lesions

* The exceptions are rare mucinous carcinomas and some invasive ductal cancers, which may enhance and have high T2 signal; irregular, rim-enhancing morphology and dynamic enhancement curves may help with diagnosis.

Morphology

In the high-spatial resolution approach, lesion morphology is evaluated on fat-nulled, 3-D images to look for characteristic shapes, borders, or internal enhancement patterns characteristic of cancer. With this approach, Nunes and colleagues reported a sensitivity of 96% and a specificity of 80% for cancer. Leong and colleagues reported similar results. The morphologic characteristics of benign and malignant lesions are summarized in Table 7-1 and Box 7-6, and are shown in Figure 7-9. As discussed elsewhere in this chapter, the radiologist first decides if the finding is a mass or a nonmasslike enhancement. Consistent with mammography, masses with spiculated or very irregular borders are suspicious. Bright enhancement, particularly rim enhancement and enhancing septations, is usually suspicious for tumor angiogenesis. A ductal, linear, or segmental pattern of clumped enhancement is suspicious for ductal carcinoma in situ (DCIS), but it can also be seen in benign duct ectasia or fibrocystic change. As with mammography, entirely smooth, oval, or lobulated masses oriented parallel to Cooper ligaments suggest benign lesions, whereas lesions traversing Cooper ligaments are abnormal and suggest invasive ductal cancer. Nonenhancing internal septations in smooth, oval, or lobulated masses are highly specific for a benign fibroadenoma. Nonenhancing lesions are also benign. However, it is important to evaluate the dynamic curves of benign-appearing enhancing masses because round or oval homogeneous cancers mimic benign fibroadenomas. Sometimes the suspicious kinetic curves may be the only clue that the morphologically benign mass is a cancer.

Figure 7-9 Morphologic features suggesting malignancy include spiculation (A), rim enhancement (B, detail from Fig. 7-1B), and enhancing internal septations (C, arrow). Morphologic features suggesting benignancy include smooth borders (D) and nonenhancing internal septations (E, arrow).

Dynamic Contrast Enhancement

In the dynamic MRI approach, one evaluates a lesion’s signal intensity as a function of time during the bolus IV administration of contrast material (Box 7-7). The dynamic curves are evaluated according to initial and late enhancement. Initial enhancement describes the curve in the first 2 minutes during the bolus or when the curve begins to change. The late phase of enhancement occurs after the first 2 minutes or after the curve starts to change. The late-phase curve is described as persistent, plateau, or washout, in keeping with the ACR BI-RADS® MRI lexicon. The entire spectrum of the time course of enhancement may be categorized from most benign to most suspicious, according to the following scheme of Daniel and colleagues (Fig. 7-10): nonenhancing (type I), gradually enhancing (type II), or rapidly enhancing with a sustained gradual enhancement, plateau, or early washout (types III, IV, and V, respectively). In reference to the curve shapes depicted in Figure 7-7, types I and II typically indicate benignancy and types IV and V indicate a high likelihood of malignancy. Type III curves are indeterminate. Using a similar approach, Kuhl and colleagues reported a sensitivity of 91% and a specificity of 83%. Kuhl type I curves are gradually enhancing with a late persistent plateau. Kuhl type II curves are rapidly enhancing with a late plateau. Kuhl type III curves are rapidly enhancing with a late rapid washout. There are a few exceptions to these general principles. DCIS may exhibit any of the curve types, including nonenhancing or gradually enhancing curves, shown as types I and II in Figure 7-7. Benign papillomas may exhibit type I, II, III, or even type IV curves shown in Figure 7-7. The geographic distribution of dynamic enhancement also appears to be predictive, with tumors usually enhancing most rapidly at their periphery and benign lesions enhancing most rapidly at the center.

Figure 7-10 Classification of the time course of dynamic contrast enhancement from the most likely benign (type I) through the most likely malignant (type V). No enhancement (type I) or gradual enhancement (type II) suggests a benign lesion. Rapid initial enhancement followed by gradual late enhancement (type III) is indeterminate. Rapid initial enhancement followed by a plateau signal intensity (type IV) or early washout of signal intensity (type V) is suspicious for invasive malignancy.

(From Daniel BL, Yen YF, Glover GH, et al: Breast disease: dynamic spiral MR imaging, Radiology 209:499–509, 1998.)

A variety of image-processing techniques have been developed to automate analysis of dynamic images throughout the breast on a pixel-by-pixel basis, including the saturation model, the two-compartment pharmacokinetic model, the “three time point method,” and simpler enhancement ratios, wash-in slopes, and washout rates. Stand-alone workstations are available to perform these calculations and produce “functional images” that display abnormal areas of dynamic enhancement on corresponding anatomic images. Most current software determines whether to colorize a lesion based on whether it exceeds a percentage enhancement threshold at approximately 1 minute. The color of the lesion is set to reflect the late enhancement dynamics (i.e., persistent, plateau, and washout).

Not surprisingly, the highest sensitivity and specificity arise when both morphologic information and dynamic enhancement curves are taken into account, with a sensitivity of 95% and a specificity of 86% reported by Daniel and colleagues and a sensitivity of 91% and a specificity of 83% reported by Kinkel and colleagues. Pharmacokinetic scans or physiologic scans are scans that superimpose physiologic information, such as enhancement information, on morphologic images, thereby combining both types of information into one format. This type of scan usually shows the morphologic appearance of a lesion with the physiologic image superimposed in color.

T2-Weighted Imaging

T2-weighted imaging also plays an important role in discriminating which enhancing lesions are likely to be benign or malignant (Fig. 7-11; see also Tables 7-1 and 7-4). Lesions with very high signal, in which the lesion is much brighter than glandular tissue and even higher than fat on nonfat-suppressed T2-weighted FSE images, suggest benign lesions such as cysts, fluid-filled ducts, lymph nodes, or fibroadenomas. Invasive tumor, on the other hand, usually has a T2 signal similar to that of glandular tissue, that is, higher than muscle but not as high as fluid. Low-signal septations within very high-signal smooth oval or lobulated lesions on T2-weighted imaging also suggest benign fibroadenomas. Exceptions to this “rule” include mucinous cancer, which can be very bright on a T2-weighted image. Some invasive ductal cancers also may be bright on T2-weighted images. These T2-bright cancers may show irregular margins and inhomogeneity; however, it is important to check kinetic curves on all masses because occasionally cancer may be round or oval and have smooth borders.

Figure 7-11 T2-weighted imaging features of benign and malignant disease. Very high signal on T2-weighted fast spin-echo images that is brighter than fat (on nonfat-suppressed sequences) and substantially brighter than glandular tissue suggests a benign lesion such as a cyst (A, arrow), intramammary lymph node (B, arrows), or fibroadenoma (C). Low-signal septa are particularly specific for fibroadenoma (see Fig. 7-18A). D, Most malignancies, unless frankly necrotic, have a signal intensity that is similar to that of fibroglandular tissue (arrow).