Introduction

MR Enterography (MRE) is a noninvasive technique for imaging the entire gastrointestinal tract and extraintestinal organs in a single examination without ionizing radiation. MRE is particularly valuable for assessing patients with inflammatory bowel disease (IBD), allowing for repeated disease evaluation over time while sparing the risk of cumulative radiation exposure. In this setting, MRE can diagnose IBD and its subtypes, assess the bowel lumen, bowel wall, and surrounding structures simultaneously, evaluate disease activity over time, and assess response to therapy [1••]. Quantitative MRE indices have also been developed and validated for assessing treatment response [2•]. While MRE is particularly sensitive for detecting inflammation, accurately quantifying the presence and degree of fibrosis remains an area of active interest and investigation [3•]. Beyond IBD, the soft-tissue contrast resolution, lack of ionizing radiation, and the ability to perform dynamic and functional sequences has made MRE an option for evaluating the small bowel for neoplasms and other etiologies.

Technique

Optimizing MRE technique requires a complement of fluid-sensitive and postcontrast sequences, as well as fast imaging sequences to minimize bowel motion and ensure sufficient luminal distension. A set of expert panel consensus guidelines regarding MRE technique for IBD have recently been published by the Society of Abdominal Radiology, a summary of which, along with particular components and their utility, is outlined in Table 1 [1••].

Table 1 MR enterography—components of a typical exam
Full size table

Coronal and axial single-shot fast/turbo spin-echo (e.g., SSFSE, SSTSE, HASTE) sequences of the abdomen and pelvis are obtained for evaluating bowel distension, bowel wall thickness, and wall edema, but are relatively susceptible to motion artifact [4•]. Fat-suppressed T2-weighted sequences may also be used for increasing the conspicuity of inflammatory changes or edema in the bowel wall and mesenteric fat. To complement the T2-weighted sequences, balanced steady-state free precession sequences (BSSFP) are rapidly acquired that virtually eliminates motion artifact and depicts bowel wall thickening and mesenteric changes. However, these sequences are less specific in the soft-tissue contrast information due to mixed T1 and T2 contrast. These sequences can also be used to generate cine loops to identify areas of fixed narrowing, as well as to evaluate bowel motility. Fat-suppressed three-dimensional gradient-echo T1-weighted images are acquired pre- and dynamically postcontrast in the coronal plane and serve as the primary means for assessing the bowel wall enhancement pattern, mesenteric vasculature, interloop fistulas, and abscesses. However, contrast-enhanced sequences are also the most susceptible to motion, and thus, in order to ensure optimal image quality, postcontrast images are usually acquired after administration of pharmacologic bowel paralytics [1••]. Axial postcontrast images are also usually acquired, which can help colocalize findings, delineate anatomy, and identify extraintestinal findings.

Optional sequences include diffusion-weighted imaging (DWI), cine sequences, and magnetization transfer (MT) imaging. DWI is particularly useful for identifying inflammation in patients who cannot receive IV contrast. DWI at multiple b values also allows for calculation of apparent diffusion coefficient (ADC), which has been shown to have a linear correlation with markers of active inflammation [1••]. As such, DWI is a promising tool for quantifying inflammation in order to objectively measure response to therapy for patients with IBD. The added value of DWI is currently being investigated for severity indexing and treatment response evaluation, but there remains uncertainty due to difficulty in reproducing diffusion sequences and ADC values across different MR scanners and vendors. Multislice multiphase cine MRE sequences assess bowel peristalsis and can be generated using BSSFP sequences. Although currently considered optional, there is an increasing evidence suggesting that cine sequences may provide unique clinically useful information, improve sensitivity (especially in motion degraded studies), increase diagnostic confidence in equivocal cases, and have the potential to alter clinical management [5]. Cine MRE can also be used to assess functional motility disorders and to monitor postoperative bowel motility [6•]. MT imaging consists of acquisitions with and without an MT prepulse saturation in order to calculate the MT ratio, which can be used to identify tissue collagen [1••].

Optimal evaluation of the small bowel (particularly important in identifying early inflammatory changes) relies heavily on adequate distension with oral contrast material. For patients with known or suspected IBD, a hyperosmolar, nonabsorbable agent is required to ensure adequate distension of the terminal ileum [1••]. Biphasic oral contrast agents (e.g., VoLumen®) are low signal on T1-weighted sequences and high signal on T2-weighted sequences, and are preferred since they allow clear visualization of the bowel wall enhancement while also ensuring adequate contrast between the lumen and the bowel wall [4•]. Ingestion of 1350 mL of oral contrast over 45–60 min has been shown to be preferable over 900 or 1800 mL, although the actual amount ingested is largely patient dependent [4•].

To mitigate bowel motion and optimize image quality on postcontrast images, pharmacologic inhibition of bowel peristalsis is typically utilized. In the United States, glucagon (0.5–1.0 mg) is most commonly used via intravenous (IV) or intramuscular (IM) injection. Advantages of IM administration include longer duration and better side effect profile (e.g., less nausea, cramping, and vomiting). IV administration is convenient due to its rapid onset and ability to utilize the same IV access needed for IV contrast. However, IV administration is more frequently associated with nausea and vomiting, especially when given too rapidly. In Europe, butylscopolamine is the preferred agent (20–40 mg IV), although not available in the United States. Image quality is better with improved bowel paralysis, although care must be taken to minimize side effects from improper administration.

Proper coverage for MRE should include the entire small and large bowel, plus the perianal region for patients with known or suspected IBD. While concurrent dedicated high-resolution T2-weighted images of the perianal region are not always warranted or feasible, adequate coverage of the perineum on standard sequences generally allows the most severe complications to be detected. Patients may be imaged in the supine or prone position, depending on institution preferences. Advocates of prone imaging note that it decreases the anterior-posterior coverage to be scanned, and may help displace small bowel from the pelvis [7]. However, detractors note patient discomfort and resultant motion while in the prone position, particularly for obese individuals, those with orthopedic comorbidities or those with stomas.

Normal MRE Appearance of Bowel

Normal small bowel diameter is less than 3 cm with mean diameter decreasing from proximal to distal. Wall thickness is normally 0.1–0.2 cm, while fold thickness is normally 0.2 cm in the duodenum and slightly less than 0.2 cm in the terminal ileum [4•]. Normal jejunum has more surface area due to increased fold density, which results in more relative enhancement compared to normal ileum. Collapsed or underdistended loops of small bowel are a common pitfall since they will appear thickened and hyperenhancing. As MRE sequences are performed dynamically and acquired at multiple time points, queried segments should be correlated with other sequences to confirm whether findings persist or are transient (Fig. 1). Adequate bowel distention is the best way to differentiate decompressed bowel loops from true areas of bowel wall thickening and luminal narrowing.

Fig. 1
figure 1

Normal MRE with peristalsis. a Coronal T2W MRE image demonstrates decompressed loops of jejunum (white arrow) with apparent bowel wall thickening. b BSSFP sequence obtained later in the exam shows these same loops of bowel to have normal thickness (black arrow), confirming that the apparent wall thickening was transient and secondary to underdistention

Full size image

Indications

Inflammatory Bowel Disease

The two major subtypes of IBD are ulcerative colitis (UC) and Crohn disease (CD), which are characterized by a dysregulated immune-mediated inflammatory response that primarily affects the GI tract [8]. CD is a chronic transmural inflammatory disease that can affect any part of the GI tract and is thought to be related to altered gut mucosal immunity, which leads to overproduction of cytokines and leukocyte infiltration of the bowel wall [9]. UC involvement is restricted to the colon, so the diagnosis is frequently made endoscopically [10]. However, MRE may be useful to distinguish between CD and UC when the clinical, endoscopic, and histologic features are indeterminate.

The distribution of bowel involvement on MRE is frequently most useful for distinguishing CD from UC (Fig. 2). Classically, UC is characterized by contiguous colonic inflammation, which begins in the rectum and can extend to the cecum [8]. The small bowel is generally not involved and the concept of “backwash ileitis” is debated. By contrast, CD is characterized by discontinuous inflammation, most commonly involving the terminal ileum, but with skip lesions of disease involvement elsewhere in the small bowel and colon. Colonic involvement in CD is typically noncontiguous and is more frequently associated with fistulizing/penetrating and stricturing disease. Thus, any significant small bowel inflammation or perianal fistulizing disease should prompt suspicion for CD, even if pancolitis is present. When typical patterns are present, distinguishing CD from UC is relatively straightforward. However, up to 1/3 of CD patients will have isolated colonic involvement and 2/3 will have only luminal findings. This leaves a significant subset of CD patients with an imaging presentation closely resembling UC. Radiologists must also recognize that the initial diagnosis is not absolute: 5–15 % of new cases of IBD do not meet criteria for UC or CD and are termed “inflammatory bowel disease unclassified” (IBDU), while up to 14 % of patients initially classified as UC or CD will have the diagnosis altered over time [11]. For these patients, MRE findings may ultimately help lead to the definitive diagnosis. Differentiating between UC and CD is also important for patients considering ileal pouch-anal-anastomosis (IPAA) since there is a much higher complication rate (up to 90 %) in CD patients who receive this procedure, including perianal abscess, fistula, pouchitis, or anal stricture [11].

Fig. 2
figure 2

Comparison of Ulcerative Colitis versus Cohn disease. a Coronal BSSFP image shows diffuse pancolitis with loss of colonic haustration, colonic wall thickening (top arrow), and increased vascularity (bottom arrow) in a 57-year-old male with long-standing UC. b Coronal T1W postcontrast image in the same patient shows bowel wall thickening, mural enhancement, and increased mesenteric vascularity (arrow) involving the ascending colon (distribution was contiguous from rectum to cecum). c Coronal T1W postcontrast image in a 35-year-old male with CD shows typical discontinuous skip areas of colonic involvement with mild narrowing, increased enhancement, and bowel wall thickening affecting short segments of the transverse and sigmoid colon (arrows). Intervening colon segments were normal. d Coronal T1W postcontrast image in the same patient demonstrates typical active inflammatory changes in the terminal ileum with stratified mural enhancement (arrow)

Full size image

CD phenotypes are clinically classified as inflammatory (nonstricturing and nonpenetrating), stricturing, and penetrating based on presentation, management, and prognosis (Figs. 3, 4, 5). MRE findings help clinical classification and management by characterizing disease extent and distribution, as well as denoting active inflammatory, fibrostenotic, or penetrating/fistulizing disease (intra-abdominal or perianal). In practice, MRE findings exist on a spectrum with features of active inflammation coexisting with fibrostenosing and penetrating disease (Fig. 6). As such, the imaging appearance is best characterized based on the dominant features (i.e., predominantly active inflammation or predominantly fibrostenotic). Moreover, interpretation of MRE warrants more than simply noting the presence or absence of inflammation. When MRE is used to evaluate patients with CD, radiologists should comprehensively convey the relative contribution and predominant features of active inflammation, penetrating disease, and fibrostenosing disease, length and location of segments involved, coexistent dilated bowel and whether it is proximal to active inflammation or fibrosis, and extraintestinal complications [4•].

Fig. 3
figure 3

MRE from a 32-year-old male with active nonpenetrating nonstricturing Crohn disease. a Axial T1W postcontrast image shows the terminal ileum with bowel wall thickening, mural stratification, and mesenteric hypervascularity (arrow). b Axial T2W image demonstrates that the same loop of terminal ileum has high signal within the wall—a sign of active inflammation (arrow). c Coronal T2W image with fat saturation shows nonorganized mesenteric edema adjacent to the affected bowel loop (arrow)

Full size image
Fig. 4
figure 4

A 41-year-old male with stricturing Crohn disease prior to surgical resection. a Coronal T1W postcontrast image shows luminal narrowing of the terminal ileum with pseudosacculations (arrowhead). b Coronal BSSFP image shows the segmental luminal narrowing to be fixed throughout the exam. This stricturing disease resulted in frequent hospital admissions for recurrent small bowel obstructions and eventually required surgical resection

Full size image
Fig. 5
figure 5

MRE images from a 35-year-old male with penetrating Crohn disease. a Axial BSSFP image shows a clump of thickened small bowel loops tethered together in the central mesentery (arrow). b Axial T1W postcontrast image shows thickening and enhancement of the inflamed bowel loops with a complex interloop fistula (arrow)

Full size image
Fig. 6
figure 6

A 26-year-old male with recurrent small bowel obstructions who underwent surgical ileocecectomy that showed admixed active inflammation and chronic fibrosis/stenosis. a Axial T1W postcontrast image from preoperative MRE demonstrates luminal narrowing of the terminal ileum with stratified mural enhancement (arrow). b Axial T2W image of the same segment demonstrates wall thickening with high signal in the bowel wall. Although the enhancement pattern and T2 hyperintensity were consistent with active inflammation, the fixed luminal narrowing (which had been seen repeatedly over several prior studies) suggested coexistent fibrostenotic stricture, making the predominant process (active inflammation versus fibrostenotic disease) difficult to determine

Full size image

Findings of active inflammation include stratified mural enhancement, bowel wall thickening (>3 mm), increased signal intensity on T2-weighted images (relative to muscle), edema in the adjacent mesenteric fat, mesenteric hypervascularity or engorgement (“comb sign”), and mucosal ulcerations [12]. Stratified mural enhancement has been shown to be specific for active inflammation and refers to hyperenhancement along the mucosa and serosa with intervening hypoenhancing submucosal edema (Fig. 7) [1••]. Mesentery adjacent to actively inflamed bowel may enhance and have high signal on T2 W images, indicating reactive edema and inflammation. Actively inflamed bowel may also have luminal narrowing with or without proximal dilation (Fig. 8).

Fig. 7
figure 7

A short segment of actively inflamed small bowel in a 31-year-old male with Crohn disease. Typical stratified mural enhancement on this axial T1W postcontrast image is indicated by hyperenhancement along the mucosa and serosa (arrows) with intervening hypoenhancing submucosa (asterisk)

Full size image
Fig. 8
figure 8

Coronal T1W postcontrast (a) and T2W (b) images from an MRE in a 35-year-old patient with Crohn disease presenting with acute obstructive symptoms. In the central abdomen, there are multiple tethered segments of actively inflamed small bowel (arrows), with significant proximal small bowel dilation (asterisks). The patient underwent emergency surgery and an inflammatory closed loop obstruction was discovered

Full size image

In fibrostenosing disease, long-standing inflammation results in luminal narrowing secondary to collagen deposition and smooth muscle proliferation, which can lead to coexistent obstruction [9]. On imaging, fibrostenosing disease is characterized by luminal narrowing and proximal dilation with a relative lack of the imaging findings seen with active inflammatory disease (Fig. 9). Although wall thickening may be present, there is typically transmural enhancement rather than mural stratification (Fig. 10). Fibrosis is generally inferred when there is low signal intensity on T1- and T2-weighted images and luminal narrowing with proximal dilation, although the lack of inflammatory findings does not reliably predict the histologic presence of fibrosis [9]. Areas of narrowing can be confirmed on cine sequences as fixed nondistensible segments, distinguishing them from transient peristalsis and strictures [5].

Fig. 9
figure 9

A 57-year-old female with Crohn disease and fibrostenotic stricture. a Coronal T2W image with fat suppression demonstrates mildly dilated small bowel (asterisks) proximal to an area of wall thickening and luminal narrowing (arrow). Coronal T1W postcontrast (b) and axial T2W (c) images demonstrate that the narrowed segment has transmural enhancement (arrow) and no T2 hyperintensity (arrowhead) or mesenteric edema, suggesting fibrostenotic stricture

Full size image
Fig. 10
figure 10

Chronic fibrostenotic stricture in a 39-year-old female with long-standing Crohn disease. a Coronal T2W image of the terminal ileum shows luminal narrowing with bowel wall thickening and low signal in the bowel wall (arrow). b Coronal T1W postcontrast image through the stenotic segment demonstrates minimal transmural enhancement with lack of stratified mural enhancement (arrow). c Coronal T2W image with fat saturation shows no high signal edema in the adjacent mesentery

Full size image

Transmural inflammation and deep ulcers can ultimately lead to penetrating disease, including strictures, sinus tracts, fistulas, and abscesses. Fistulas frequently develop at or proximal to strictures and can form between actively inflamed bowel and any adjacent structure, including small and large bowel, bladder, vagina, scrotum, muscle, and skin surface (Fig. 11). Although a discrete fluid-filled enhancing tract is definitive evidence of a fistula, interloop fistula can also be suggested or inferred by clustered, tethered, or stellate arrangement of bowel loops (Figs. 5, 11). Abscesses may form in the adjacent mesentery, sometimes with communication to an inflamed loop of bowel, and are crucial to identify, since their presence typically precludes immunosuppressive therapy (Fig. 12). MRE is highly sensitive for detecting these complications, including obstruction (87–92 %) and abscess (86–100 %), although reported sensitivity for fistula detection is more variable (40–100 %) [13••]. Moreover, MRE has been shown to be highly accurate in guiding surgical management for patients with complicated CD (stenoses, fistulas, abscesses). In a prospective study, Spinelli et al. reported that MRE predicted the correct surgical approach and strategy in 90.7 % of patients with high per-patient accuracy (>85 %) for detecting stenosis, abscess, and fistula [14•].

Fig. 11
figure 11

A 32-year-old male with newly diagnosed Crohn disease and feculent drainage from a spontaneously opened abdominal wall wound. a Axial T1W postcontrast image shows an enterocutaneous fistula corresponding to the site of skin drainage (arrow). b Coronal T1W postcontrast image also demonstrates tethered small bowel loops with a complex interloop enteroenteric fistula (arrow)

Full size image
Fig. 12
figure 12

A 31-year-old male with unmanaged Crohn disease presents with abdominal pain, fever, and sepsis. Coronal T2W (a) and coronal T1W postcontrast (b) images show an extraluminal fluid collection in the abdominal mesentery (arrow in a) with thick rim enhancement (arrow in b). The mesenteric abscess was subsequently drained percutaneously with image guidance

Full size image

Active Inflammation Versus Stenosis

CD strictures may reflect acute transmural inflammation and/or chronic fibrosis and commonly result in acute symptoms, including abdominal pain, vomiting, difficulty passing flatus or stool, and bowel obstruction [3•]. Distinguishing between luminal narrowing secondary to active inflammation and fibrosis is challenging on MRE, but carries significant implications for management. Narrowing in the setting of active inflammation may resolve with medical management, whereas fibrostenosis may require stricturoplasty, resection, and possible additional medical management with biologic therapy [3•]. Stratified mural enhancement, engorged vasa recti, and T2 hyperintensity are suggestive of active inflammation, particularly in the absence of proximal dilation, although the imaging findings of active inflammation and fibrosis frequently coexist (Fig. 13) [13••]. Similarly, the absence of inflammatory findings does not reliably correlate with underlying fibrosis [9]. Accurately quantifying the relative contribution of active inflammation and fibrosis is an area of active ongoing research due to its considerable clinical implications.

Fig. 13
figure 13

A 38-year-old female with Crohn disease and acute-on-chronic obstructive symptoms. a Axial T2W image shows a long segment of bowel wall thickening with luminal narrowing, intermediate T2W signal (arrow), and upstream bowel dilation (asterisk). b Axial T1W postcontrast image shows mixed bowel wall enhancement pattern, including both transmural and stratified enhancement (arrows). The patient’s bowel obstruction was suspected to be related to a combination of acute inflammation and chronic stricture, and thus was treated with immunomodulating drugs with symptom improvement

Full size image

Colitis

MRE is primarily utilized for small bowel evaluation, although it is also valuable in concurrently demonstrating colonic inflammation. For patients with CD, there is a paucity of data regarding colonic findings, and reported sensitivities vary from 27 to 88 % depending on the severity of colonic inflammation [4•]. Active colonic inflammation manifests similar to small bowel inflammation with stratified mural enhancement, bowel wall thickening, increased signal intensity on T2-weighted images, and mesenteric hypervascularity or engorgement (Fig. 14) [4•]. Long-standing severe colonic inflammation and subsequent healing may result in pseudopolyps (Fig. 15) [9, 15]. For patients with UC, pseudopolyps are particularly important to note since they are a risk factor for developing colorectal cancer [8].

Fig. 14
figure 14

A 20-year-old patient presenting with abdominal pain and colitis, subsequently found to represent Crohn disease. Coronal T1 W postcontrast (a) and coronal T2W (b) MRE images demonstrate stratified mural enhancement (arrow in a), wall thickening, and bowel wall edema (arrow in b) involving a contiguous segment of sigmoid colon. Initial differentiation of Crohn disease versus ulcerative colitis was unclear based on MRE and colonoscopic findings, but follow-up biopsy favored CD, which was further supported when the patient developed perianal fistulizing disease

Full size image
Fig. 15
figure 15

A 27-year-old female with severe ileocolonic Crohn disease and pseudopolyps. Coronal (a) and axial (b) T1W postcontrast images demonstrate multiple enhancing polypoid lesions along the mucosa throughout the colon (arrows). Also note coexistent signs of active inflammation, including stratified mural enhancement (asterisk) and engorgement of the vasa recta (arrowheads). Subsequent colonoscopy also showed pseudopolyps with biopsies demonstrating signs of chronic inflammation, but no dysplasia or neoplasia

Full size image

Perianal Disease

Particular attention should be paid to the perianal region, even on standard MRE, since perianal fistulas or abscesses complicate CD in up to 36 % of patients and can be a key feature to help differentiate CD from UC (Fig. 16) [4•], [10]. For detailed characterization of perianal disease, particularly the relationship to the sphincter complex, dedicated high-resolution small field-of-view imaging of the perianal region should be performed, typically as a separate examination. Although dedicated perianal MRI can be combined with standard MRE to provide a comprehensive disease assessment in a single examination, tolerating a continuous 45–60-min examination can be quite challenging for patients.

Fig. 16
figure 16

A 51-year-old male with Crohn disease and perianal abscess. Coronal T2W (a) and coronal T1W postcontrast (b) images demonstrate a peripherally enhancing perianal abscess in the left ischioanal fossa (arrows). Internal opening and relationship to the sphincter complex are difficult to determine without dedicated imaging

Full size image

Ileoanal Pouch

Restorative proctocolectomy with IPAA is the procedure of choice for UC refractory to medical therapy and UC-associated colonic neoplasia. IPAA generally improves quality of life by restoring GI tract continuity and maintaining fecal continence, but is associated with early and late complications [16]. Early complications include postprocedural sepsis, anastomotic leaks, abscesses, small bowel obstruction, and portal vein thrombosis. Late complications include pouchitis, strictures, fistulas, and malignancy. Pouchitis is the most common late complication, with a reported prevalence of 27 % [16]. Since clinical symptoms correlate poorly with endoscopic and histologic findings of pouchitis, imaging may be useful for noninvasive diagnosis. A recent retrospective review by Sahi et al. used the MRE appearance of the ileoanal pouch to generate a composite score, which was then correlated with endoscopic and histologic findings of inflammation [16]. The MRE score included findings of active inflammation (pouch wall thickening, mural stratification, mucosal hyperenhancement, peripouch stranding, peripouch hyperemia, presacral widening, peripouch abscess, fistula, or sinus tract), subacute and chronic inflammation (peripouch fatty proliferation, presacral space widening, enlarged lymph nodes, and submucosal fat deposition), and inflammatory complications (inlet hyperenhancement and inlet stricture) (Fig. 17). The composite MRE had a strong correlation with endoscopic findings (r = 0.61; p = 0.0005) with four or more MRE findings of inflammation correlating best with endoscopic findings (sensitivity 86 %, specificity 79 %, accuracy 82 %). This suggests an increasing role for MRE in this patient population, offering the potential for noninvasive diagnosis without ionizing radiation.

Fig. 17
figure 17

A 44-year-old female with ulcerative colitis status post total colectomy with ileal pouch formation. Coronal T1W postcontrast image (a) shows pouch wall thickening with mild stratified mural enhancement (arrow). Coronal T2W image (b) shows pouch wall edema and peripouch free fluid, consistent with pouchitis (arrow)

Full size image

Extraintestinal Findings

Compared to ileocolonoscopy, MRE adds value by identifying extramural and extraintestinal findings, which are difficult or impossible to fully assess endoscopically. Although PSC complicates both CD and UC (PSC-IBD), it is more common in UC (2–7.5 % in UC versus 1.4–3.4 % in CD) and can be detected on MRE (Fig. 18) [10, 11, 17]. This is an important subset to identify since UC patients with PSC are at a much higher risk for developing colorectal cancer, typically presenting at a younger age with tumors in the right colon [18]. It has also been suggested that these patients have different imaging patterns with higher prevalence of rectal sparing and backwash ileitis, although the reported frequency is variable [17, 18]. Other extraintestinal manifestations of IBD that may be identified on MRE include sacroilitis, cholelithiasis, nephrolithiasis, and portal vein thrombosis [17].

Fig. 18
figure 18

A 49-year-old male with ulcerative colitis complicated by primary sclerosing cholangitis. Coronal T2W image demonstrates a beaded appearance of the biliary tree with alternating areas of stricturing and normal or slightly dilated intrahepatic ducts

Full size image

Follow-Up

Monitoring response to therapy and disease activity over time in CD has increasingly incorporated imaging findings in addition to traditional clinical indices. Imaging findings may be more useful to guide therapy since patient symptoms and other clinical parameters correlate poorly with disease activity, while MRE frequently identifies disease that would be underestimated (Fig. 19) [13••], [19]. Specifically, demonstrating complete mucosal healing is associated with clinical remission, decreased hospitalization rates, and decreased need for surgery [13••]. MRE findings have already been used to generate activity and severity scores, which correlate with endoscopic findings and can predict response [20, 21]. Reproducible quantitative MRE scores ultimately may serve as noninvasive, radiation-free biomarkers in clinical trials and routine clinical practice.

Fig. 19
figure 19

A 62-year-old male with Crohn disease at diagnosis and 1-year follow-up. Coronal T2W (a) and coronal T1W postcontrast (b) images at presentation show a short segment of active inflammation involving the terminal ileum (arrows). One year later, while asymptomatic and off therapy, repeat coronal T1W postcontrast (c) and coronal T2W (d) images show continued active inflammation, but with even longer segment involved (arrows)

Full size image

Other Indications

Although MRE is most well established for evaluating patients with IBD, its superior soft-tissue contrast resolution, lack of ionizing radiation, and the ability to perform dynamic and functional sequences has made it an option for evaluating the small bowel in patients who do not have IBD. Small bowel neoplasms are relatively uncommon, accounting for only 3–6 % of GI tract neoplasms, and are frequently difficult to diagnose owing to nonspecific symptoms, such as anemia, GI bleeding, and abdominal pain [22]. There is a relative paucity of data regarding use of MRE for identifying small bowel neoplasms when screening and evaluating symptomatic patients, but they may be detected incidentally during a work-up for suspected IBD. At least one series has shown that adding DWI to noncontrast MRE can increase the sensitivity and improve interobserver agreement for detection of small bowel neoplasms [23]. Screening patients with Peutz-Jeghers syndrome is an area of interest since MRE can detect and localize large polyps, which is useful for guiding surgical or enteroscopic resection [7]. MRE also offers better visualization of the mural and extramural extent of disease compared to optical or fluoroscopic techniques. Capsule endoscopy is still better for detecting smaller polyps (<10 mm), although smaller polyps are less relevant clinically [22].

Small bowel neoplasms that may be encountered on MRE include adenomas, lipomas, hemangiomas, polyps, GI stromal tumors, leiomyomas, carcinoid, adenocarcinoma, lymphoma, and metastases (Fig. 20). For small neoplasms, differentiation between benign and malignant lesions is frequently not possible. However, for larger nonpedunculated lesions with associated transmural extension, mesenteric fat infiltration, and lymph node enlargement, malignancy (e.g., adenocarcinoma, lymphoma) should be the primary consideration [24]. GI stromal tumors most commonly occur in the stomach and small bowel and should be considered for exophytic masses with internal hemorrhage or necrosis [22]. Carcinoid tumors most commonly arise from the distal ileum and appendix and can be suggested when there is a hypervascular lesion with characteristic spiculation and tethering of the adjacent mesentery [22]. Adenocarcinoma is the most common primary small bowel malignancy, is associated with CD, Peutz-Jeghers syndrome, and Lynch syndrome, and can be suggested when there is an enhancing annular lesion with eccentric or circumferential wall thickening [7, 22]. Lymphoma affecting the small bowel most commonly arises in the ileum and can have a variable appearance with solitary or multifocal presentations, including infiltrative pattern, exoenteric mass, polypoid intraluminal filling defect(s), segmental circumferential wall thickening, or aneurysmal dilation, frequently with associated enlarged lymph nodes [7, 22]. Secondary small bowel obstruction or intussusception may complicate benign or malignant small bowel neoplasms.

Fig. 20
figure 20

Various small bowel tumors seen on MRE. a Axial T1W postcontrast image in an 87-year-old patient with recurrent partial small bowel obstructions demonstrates an enhancing jejunal mass (arrow), subsequently found to be adenocarcinoma after surgical resection. b Axial T2W image in a 51-year-old female with early satiety shows a heterogeneous T2 hyperintense mass in the second part of the duodenum (arrow) that was proven to be a gastrointestinal stromal tumor upon surgical removal. c Axial T1W postcontrast image in a 65-year-old patient with advanced metastatic melanoma shows an aneurysmal small bowel metastasis (arrowhead), as well as multiple liver metastases (arrows). d Coronal T1W postcontrast image in a 62-year-old male who presented with metastatic neuroendocrine tumor in his liver. MRE showed a 1-cm enhancing nodule in the distal ileum (arrow), which was positive on an OctreoScan (not shown) and presumed to be the primary neoplasm

Full size image

Within the GI tract, the small bowel is the most radiosensitive and can develop acute or chronic radiation enteritis [25]. MRE findings may be indistinguishable from IBD with wall thickening, stratified enhancement, and mesenteric hypervascularity in the acute phase and transmural enhancement with fixed luminal narrowing in the chronic phase (Fig. 21). Other inflammatory processes that may be encountered on MRE include systemic vasculitides, ischemia, and chemotherapy-induced enteritis, all of which have similarly nonspecific imaging findings [7].

Fig. 21
figure 21

A 51-year-old female with abdominal pain and diarrhea status post pelvic radiation for anal cancer. a Coronal T2W fat-suppressed image demonstrates wall thickening with T2 hyperintensity involving a long segment of mid- and distal ileum (arrows). Coronal (b) and axial (c) T1W postcontrast images show stratified mural enhancement (arrows) and mesenteric hypervascularity (circle) in the affected segment. Based on the temporal relationship to radiation therapy, this was presumed to represent acute radiation enteritis and resolved with conservative management, although the imaging findings are identical to acitve inflammation in a patient with active Crohn disease

Full size image

Recently, MRE has been used to preoperatively evaluate patients with bowel endometriosis, demonstrating high accuracy and excellent interobserver agreement for localizing lesions above the rectosigmoid junction [26]. Such applications suggest a future role for MRE evaluating other noninflammatory etiologies.

Advantages and Disadvantages Versus CT Enterography (CTE)

For patients with suspected or known CD, the American College of Radiology (ACR) Appropriateness Criteria recommend MRE and CTE as the best imaging tests for diagnosis, evaluating acute flares, and long-term surveillance [13]. MRE has high sensitivity (77–82 %) and specificity (80–100 %) with similar accuracy to CTE for diagnosing IBD, but with superior soft-tissue contrast resolution, no ionizing radiation, and the ability to perform dynamic and functional sequences [4•], [13••], [27], [28•]. A recent prospective study by Masselli et al. comparing MRE and CTE for patients with suspected small bowel disease demonstrated MRE to be significantly more sensitive for overall detection of small bowel diseases and small bowel neoplasms [28•]. The lack of ionizing radiation is particularly important for patients with CD who are frequently diagnosed as children or young adults and expected to undergo repeated imaging during the course of their life. Radiation risk for these patients is significantly higher than the general population and a significant proportion of patients with IBD receive potentially harmful levels of lifetime radiation (>50 mSv) [1••], [29]. Both CTE and MRE can be used to correlate with response to therapy and mucosal healing, with MRE potentially offering superior detection of fibrosis [1••]. Lastly, unlike noncontrast CT, a noncontrast MRE can still use T2-weighted and diffusion-weighted sequences to provide diagnostic information even when IV contrast is contraindicated (e.g., pregnancy, renal failure, severe allergy) [1••].

The major limitation of MRE compared to CTE is its variability and susceptibility to respiratory and bowel motion artifacts, although such motion artifacts do not always affect the diagnostic accuracy [13••], [28]. As such, the ACR still recommends CTE over MRE for the initial evaluation in patients with suspected CD, although institutional preferences may vary [13••]. MRE may also be suboptimal for patients with CD presenting acutely or with severe symptoms who are unable to remain still for long acquisitions and cooperate with breathing instructions. MRE also requires time for protocol optimization and expertise, and has a steeper learning curve for the novice interpreter.

Advantages and Disadvantages Versus MR Enteroclysis

MR enteroclysis, which involves nasoenteric administration of luminal contrast, achieves better luminal distension and identification of mucosal lesions than MRE [12]. However, it requires ionizing radiation for placement of the nasoenteric tube, is not available at all institutions, and is generally less well tolerated by patients [7, 22].

Clinical Impact for IBD

The role of MRE in IBD practice goes beyond initial diagnosis in characterization of disease phenotype and distribution of disease involvement. MRE has already been widely shown to be accurate for identifying active inflammation, but is now being used to guide treatment decisions by quantifying disease severity and correlating with existing clinical indices. Using MRE to guide clinical management of patients after diagnosis is made, specifically in CD, is of significant interest.

As already described, defining the extent of disease is particularly important for patients with distal ileal CD who can also have more proximal small bowel disease not accessible by endoscopic visualization. Furthermore, distinguishing inflammatory from fibrostenotic disease impacts clinical decision-making regarding medical therapy, need for endoscopic dilation, and/or surgical resection.

The role of MRE continues to evolve in evaluating response to therapy in patients with CD. This is primarily due to current challenges that exist in clinical assessment of CD patients, specifically the lack of correlation between clinical indices, such as the Crohn’s Disease Activity Index (CDAI), and laboratory markers of inflammation, such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Endoscopic evaluation remains the gold standard in evaluation of treatment response and recent literature favors mucosal healing as a therapeutic endpoint [30]. However, the invasiveness of the procedure precludes serial endoscopy as a means for evaluating treatment response, and currently is reserved for patients who are symptomatic while on treatment. As a noninvasive technique that can image the entire gastrointestinal tract in a single examination without ionizing radiation, MRE is an appealing option to repeatedly assess therapeutic response and treatment efficacy. By using imaging findings to optimize medical treatment, MRE may help minimize overall costs and the potential adverse side effects or toxicities of therapy. Moreover, quantitative MRE indices can be used to objectively assess treatment response in an effort to alter the long-term disease course [31].

Rimola et al. used multivariate analysis to develop a Magnetic Resonance Index of Activity (MaRIA) based upon MR findings that best identified endoscopically active and severe disease, including bowel wall thickness, mural edema (T2 hyperintensity), relative contrast enhancement, and mucosal ulceration, demonstrating it to have good correlation with severity of endoscopic lesions and with the Crohn’s Disease Endoscopic Index of Severity (CDEIS) [32]. It was subsequently externally validated in a prospective study, showing high accuracy for detecting active and severe disease, demonstrating its utility as a reproducible quantitative metric [20]. In a recent prospective multicenter study, the MaRIA score was shown to be reliable for assessing treatment response and mucosal healing. Compared to ileocolonoscopy score using the CDEIS, the MaRIA score was shown to have 90 % accuracy in determining ulcer healing and 83 % accuracy for the evaluation of endoscopic remission [2•].

Another MRE index developed for evaluation of disease activity is the Crohn’s Disease Activity Score (CDAS). This was developed using a histopathologic grading score or the acute inflammation score (AIS) as a reference, and found significant association between MR features of active disease with acute inflammation [33]. More recently, a study modified CDAS into a global assessment score called the MRE global score (MEGS) and found that the MEGS model had a sensitivity of 65 % and specificity of 78 % for the detection of active disease [34].

Lastly, the Lémann score or the Crohn’s Disease Digestive Score is an instrument used to measure the cumulative structural damage to the bowel and takes into account disease damage location, severity, extent, progression, and reversibility, as measured by diagnostic imaging modalities (CTE or MRE) and history of surgical resection [35]. Rather than assessing the bowel at a single time point, this score accumulates over time and provides a graphic representation of structural bowel damage and risk for progression of disease. In order to facilitate widespread use and reproducibility, reliable imaging is needed, ideally a single modality and technique. In this role, MRE seems best suited for the development of a comprehensive scoring system to assess progression of CD and effect of treatment on long-term outcomes [36].

Future Directions

Emerging MRE sequences and fusion modalities continue to attempt to enhance accuracy, provide functional information, and serve as meaningful biomarkers.

Hybrid positron emission tomography (PET)/MRE serves to combine the metabolic information of 2-[fluorine 18]-fluoro-2-deoxy-d-glucose (FDG) PET with the anatomic detail and soft-tissue contrast of MRE, while also decreasing the radiation exposure compared to PET/CT [3•]. A recent retrospective analysis found three quantitative biomarkers (SUVmax, SI on T2-weighted images × SUVmax, and ADC × SUVmax), which were significantly lower in fibrotic strictures compared to active inflammation alone and mixed fibrosis and active inflammation [3•]. Of these, ADC × SUVmax < 3000 performed the best (accuracy 71 %, sensitivity 67 %, and specificity 73 %) and offers a promising tool for guiding management if it can be more widely adopted and validated.

MT imaging generates contrast determined by the fraction of large macromolecules or immobilized phospholipid cell membranes in tissue and can be used to directly quantify fibrosis and differentiate from inflammation [37]. High MT ratios correlate with the presence of fibrosis and collagen in rats and have subsequently been shown to be feasible in humans [38, 39]. However, further research is needed to demonstrate reproducibility and added value in patient management.

Conclusion

MRE is a valuable noninvasive technique for imaging the entire gastrointestinal tract and extraintestinal organs in a single examination without ionizing radiation. The value of MRE is best established in patients with IBD, allowing for accurate initial diagnosis and subtyping by assessing the bowel lumen, bowel wall, and surrounding structures simultaneously. MRE is superior to CTE in its ability to evaluate disease activity over time, thereby assessing response to therapy while sparing the risk of cumulative radiation exposure. MRE can also identify extraintestinal findings, such as PSC, which affect long-term prognosis, but are difficult or impossible to fully assess endoscopically. Reliably predicting and quantifying fibrosis remains a shortcoming, but is an area of active research and novel techniques, including fusion PET/MRE and MT imaging, offer promise. Quantitative MRE indices have also been developed and validated for assessing treatment response in patients with IBD, which can help guide treatment toward mucosal healing. Little data exist for screening and characterizing small bowel neoplasms on MRE, though its soft-tissue contrast resolution, lack of ionizing radiation, and the ability to perform dynamic and functional sequences suggest the potential for an emerging role.