Toward Molecular Imaging of Intestinal Pathology

Abstract

Inflammatory bowel disease (IBD) is defined by a chronic relapsing and remitting inflammation of the gastrointestinal tract, with intestinal fibrosis being a major complication. The etiology of IBD remains unknown, but it is thought to arise from a dysregulated and excessive immune response to gut luminal microbes triggered by genetic and environmental factors. To date, IBD has no cure, and treatments are currently directed at relieving symptoms and treating inflammation. The current diagnostic of IBD relies on endoscopy, which is invasive and does not provide information on the presence of extraluminal complications and molecular aspect of the disease. Cross-sectional imaging modalities such as computed tomography enterography (CTE), magnetic resonance enterography (MRE), positron emission tomography (PET), single photon emission computed tomography (SPECT), and hybrid modalities have demonstrated high accuracy for the diagnosis of IBD and can provide both functional and morphological information when combined with the use of molecular imaging probes. This review presents the state-of-the-art imaging techniques and molecular imaging approaches in the field of IBD and points out future directions that could help improve our understanding of IBD pathological processes, along with the development of efficient treatments.

This review presents the state-of-the-art imaging techniques and molecular imaging approaches in the field of IBD and points out future directions that could help improve our understanding of IBD pathological processes, along with the development of efficient treatments.

INTRODUCTION

Inflammatory bowel disease (IBD) is defined by a chronic inflammation of the gastrointestinal tract (GI) and 2 major disorders: ulcerative colitis (UC) and Crohn’s disease (CD).¹ Although over 3.1 million Americans have already been diagnosed with IBD, its incidence is increasing worldwide, particularly in newly developed countries.² The remitting and relapsing nature of this pathology has direct consequences on the quality of life of those affected: diarrhea, bowel obstruction, rectal bleeding, abdominal cramps, fever, weight loss, and fatigue are among the many different and sometimes debilitating symptoms patients intermittently suffer from.³ Although the etiology of IBD remains elusive, it is thought to arise from a dysregulated and excessive immune response to gut luminal microbes triggered by genetic and environmental factors.³ To date, IBD has no cure, and treatments are primarily directed at relieving symptoms and treating inflammation.⁴ Moreover, IBD is not only life-altering but also looms over patients with life-threatening complications such as perforation, bowel obstruction, abscess formation, or colorectal cancer, leading to hospitalizations and surgical procedures, which severely impact their lives and pose an economic burden to health care systems around the world.⁵

Though CD and UC share common symptoms, they affect different parts of the GI tract. Crohn’s disease is characterized by a transmural inflammation that can discontinuously involve any segment of the GI tract, from the mouth to the anus, with the last ileal loop being most commonly affected. Ulcerative colitis is limited to the large intestine and the rectum, which are affected in a continuous fashion, with the inflammation affecting only the mucosal and submucosal layers of the bowel wall.³ Fibrosis, the excessive accumulation of extracellular matrix (ECM), is a major complication of IBD, leading to the formation of fibrotic strictures and bowel obstruction. Strictures are generally classified as inflammatory, fibrotic, or mixed and are observed in 11% of CD patients at the time of initial diagnosis, which tends to increase over time, leading to significant morbidity. Differentiation between these categories has important clinical ramifications, as medical therapy is the preferred initial treatment for pure or predominant inflammatory strictures; endoscopic dilatation or surgical resection is required for pure or predominant fibrotic strictures. But these approaches provide only intermittent relief because 70% of patients will have endoscopic recurrence within a year and 40% will require additional intervention within 4 years.⁶ Despite the emergence of effective biological therapies directed at treating inflammation, no drugs are currently available to specifically target intestinal fibrosis or prevent the formation of fibrotic strictures and bowel narrowing.

Imaging plays a critical role in the diagnosis and management of IBD.⁷ In the absence of optimal noninvasive strategies, endoscopic evaluation remains the gold standard for the diagnosis of IBD in combination with biopsy, histopathology, blood analysis, and stool analysis.⁸ Although endoscopy helps differentiate CD from UC by disease location, it cannot inform on extraluminal complications or assess the nature of the strictures due to its inability to go beyond the mucosa. Moreover, endoscopy requires bowel preparation and sometimes anesthesia, which are not well tolerated—especially in pediatric patient population—and can adversely affect adenoma detection if inadequate. Once a patient has been diagnosed, serial endoscopic surveillance is required throughout the patient’s lifetime to follow the disease’s location and severity, monitor its activity, complications, and treatment response. Cross-sectional imaging modalities, including ultrasound (US), computed tomography (CT), and magnetic resonance (MR) enterography, are clinically used to complement endoscopy and have demonstrated a good diagnostic accuracy.^9–11 Hybrid techniques employing positron emission tomography (PET) paired with CT or MR imaging also enable the detection of inflammation sites using nonspecific radiotracers, mainly 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG).^{12, 13} Preclinical studies have primarily focused on the development of MR, PET, single photon emission computed tomography (SPECT), and near infrared (NIR) probes to interrogate the pathophysiology of IBD by targeting cellular receptors or labeling therapeutic agents.^{14, 15}

Characterizing morphological and functional changes in IBD is key for improving disease diagnostics, surveillance, and developing new treatments. By complementing conventional anatomical imaging, molecular imaging measures the underlying biological and molecular processes of a disease and has proved its value in clinical practice for many diseases such as cancer or cardiovascular disorders. The relative lack of validated imaging biomarkers in IBD is a major hurdle to overcome in IBD research, clinical trials, and clinical practice. Given multiple drug candidates, heterogeneous disease course, variable patient response to treatment, invasive assessment modalities, and severe disease complications, there is a significant unmet need in IBD to identify, validate, and qualify markers to advance drug development and improve patient care (Table 1). The goal of this article is to present the current state-of-the-art imaging techniques and molecular imaging approaches in the field of IBD and point out future directions that could help improve our understanding of IBD pathological processes, along with the development of efficient treatments.

TABLE 1.

Different Roles for Molecular Imaging in IBD

Diagnosis

Disease staging and monitoring

Personalized medicine

Prediction of treatment response

Evaluation of treatment response

Differentiation between the different types of strictures: predominantly fibrotic, predominantly inflammatory and mixed strictures.

Biomarker expression

Early detection of recurrence

Measure treatment response in clinical trial

PATHOPHYSIOLOGY OF IBD

Crohn’s disease and ulcerative colitis are characterized by repeated cycles of flare and remission. Inflammation is a complex and highly coordinated protective response to an insult such as microbial infection or tissue injury followed by wound healing to restore tissue integrity and function.¹⁶ Luminal bacteria activate tissue-resident immune cells (tissue macrophage, dendritic cells, lymphocytes, endothelial cells, fibroblasts, and mast cells) expressing Toll-like receptors (TLRs), which then produce soluble mediators such as cytokines and chemokines. These mediators regulate the expression of cell adhesion molecules on circulating leukocytes and endothelial cells, allowing the recruitment of leukocytes to the site of injury. Leukocytes, mainly neutrophils, become activated and eliminate tissue debris. Fibrosis is thought to be a consequence of local chronic inflammation and is defined by an excessive production of extracellular matrix (ECM).¹⁷ After the initial injury, mesenchymal cells (myofibroblasts, fibroblasts, and smooth muscle cells) become activated and produce ECM components such as collagen and fibronectin. Activation and differentiation of mesenchymal cells are mainly driven by luminal bacteria, cytokine, and growth factors secreted by immune and nonimmune cells.¹⁸ The relationship between inflammation and fibrosis in IBD is intricate: inflammation precedes fibrogenesis followed by fibrosis that self-propagates.¹⁹ Studies have demonstrated that treating the inflammation in IBD patients does not alter the natural course of fibrosis and that matrix stiffness is substantially involved in the autopropagation of fibrosis,²⁰ suggesting that other mechanisms not related to inflammation might promote fibrosis.¹⁸

In addition to inflammation and fibrosis, muscular alteration is another important contributor to the stricturing phenotype. Indeed, histopathological characterization of Crohn’s fibrostenotic strictures indicates that smooth muscle hyperplasia of the submucosa, hypertrophy of the muscularis propria, and inflammation were the most prevalent histological changes.²¹ It was shown that the degree of muscular hyperplasia/ hypertrophy is positively associated with the severity of chronic inflammation of the bowel wall and negatively associated with the degree of fibrosis, thus underlying the importance of the “inflammation-smooth muscle proliferation” axis in the development of CD’s strictures and fibrosis. Smooth muscle cell hyperplasia is regulated by increased endogenous insulin-like growth factor-I (IGF-I) and α _vβ ₃ integrin ligands, resulting in increased proliferation and decreased apoptosis.²² Moreover, α _vβ ₃ integrin present on smooth muscle cells is responsible for the activation of TGF-β1, a fibrogenic cytokine overexpressed in CD strictures. Overexpression of transforming growth factor (TGF)-β1 leads to increase type 1 collagen production and subsequent fibrosis. It was shown that the blockage of the α _vβ ₃ arginine-glycine-aspartic acid domain using Cilengitide induced a decrease in collagen production and fibrosis in animal models of CD. Thus, Cilengitide seems to be a promising treatment to prevent fibrosis in patients with stricturing disease.²³ Detailed underlying mechanisms of intestinal fibrosis and inflammation have been extensively reviewed elsewhere.^{16, 18, 24–26}

Current medical therapies for IBD have focused on treating the inflammation and include biological therapies, 5-aminosalicylates, thiopurines, methotrexate, and calcineurin inhibitors.²⁷ Biologic therapies aim to decrease or extinguish the local immune cell infiltration by targeting pro-inflammatory mediators, such as the pro-inflammatory cytokine tumor necrosis factor (TNF), interleukin (IL-12, IL-23), and adhesion molecules (eg, integrins) to reduce leukocyte migration. However, 20%–30% of patients do not respond to therapy, whereas 30% become nonresponders with time after showing signs of remission.²⁷ New drugs are currently being tested in humans and include janus kinase inhibitors (JAKs), anti-IL-6/23, anti-integrin, sphingosine-1-phosphate modulator, and phosphodiesterase type 4 inhibitor.²⁷ Despite medical treatment of inflammation, IBD often progresses to fibrosis followed by the formation of fibrotic strictures that require surgical intervention. Intestine resection is performed in medically refractory cases and can be curative in patients with UC, despite associated complications, but disease recurrence is usually observed in patients with CD. No drugs are currently available for treatment of fibrosis and stricture, but preclinical studies have shown promise with the development of antibody to IL-36 receptors, inhibitors of TNF-like cytokine 1A, and antifibrotic medication used to treat fibrosis in other organs.²⁸

In that context, molecular imaging in IBD serves an essential role in improving the understanding of pathophysiology, distinguishing between inflammation and fibrosis, identifying new biomarkers, monitoring treatment response, and developing personalized medicine (Fig. 1).

FIGURE 1.

Illustration of potential processes and biomarkers of inflammation and fibrosis for molecular imaging of IBD. (1) Recruitment of leukocytes at the site of injury: overexpression of integrins, cell adhesion molecules, cytokines, and growth factors. (2) Intestinal permeability is a characteristic of diseased mucosa with a loss of tight junction between epithelial cells. (3) Fibroblast recruitment, activation, and myofibroblast differentiation followed by abnormal buildup of extracellular matrix (ECM) components such as collagen, elastin, and fibronectin. (4) Overexpression of lysyl oxidase enzymes (LOXs) during fibrogenesis. These enzymes catalyze the oxidation of the collagen ε-amino groups to allysine aldehydes before collagen cross-linking. Allysine aldehydes are biomarkers of fibrogenesis. (5) ECM components such as type-I collagen are potential imaging targets for fibrosis detection in IBD.

IMAGING TECHNIQUES IN IBD: CURRENT STATUS

Endoscopic Molecular Imaging

Endoscopy is clinically used to establish the final diagnosis of IBD based on disease location and extent of inflammation.¹ Mucosal biopsy is commonly performed during endoscopic examination to obtain microscopic evaluation of disease. Endoscopy is widely available and well tolerated among IBD patients despite its invasiveness. Major limitations of endoscopy comprise the use of sedation, interobserver variability, the risk of bowel perforation, the technical challenge associated with reaching remote segments of small intestine or intestine beyond a stricture segment, the lack of information on extraluminal inflammation, and the inability to assess intramural inflammation. Recent technological developments have overcome some of those limitations while improving inflammation detection and neoplastic lesions by combining endoscopy, microscopy, and molecular imaging.

For instance, wireless capsule endoscopy has been recently developed to image the small intestine.²⁹ After bowel preparation, the patient swallows a small video capsule endoscope that moves through the bowel by peristalsis. Images are simultaneously streamed to a portable recorder. However, patients may require surgery to remove the capsule if trapped by a small bowel stricture.

Chromoendoscopy requires the use of a dye during endoscopy and is recommended for the detection of intraepithelial neoplasia, which is associated with a high risk of colorectal cancer in IBD patients.³⁰ The dye, usually methylene blue or indigo carmine, is sprayed over the mucosa during the examination to enhance the visibility of mucosal surface irregularities, thus guiding targeted tissue sampling for histopathology.

Endoscope-based confocal laser endomicroscopy (eCLE) has the potential to improve the diagnosis of IBD compared with standard endoscopy by detecting intestinal mucosal barrier defects and increased intestinal permeability in IBD patients.³¹ Notably, CLE plays an important role in assessing inflammation, mucosal healing, disease activity and relapse, response to therapy, and detecting colitis-associated dysplasia. That endoscopic modality integrates a confocal laser microscope within the tip of a conventional endoscope, enabling real-time in vivo histological assessments of the mucosal barrier at a 1000-fold magnification with a subcellular resolution.³¹ A fluorescent dye, usually nontoxic fluorescein, is administrated intravenously and is quickly distributed throughout the body. The leakage of the dye into the bowel allows the detection of intestinal permeability, characteristic of diseased GI mucosa with a loss of tight junction between epithelial cells.

Confocal laser endomicroscopy can detect microscopic inflammatory changes in IBD, which are nonapparent on standard endoscopy or histology and differentiate between remission and active disease.^{32, 33} Confocal laser endomicroscopy has also demonstrated its ability to evaluate mucosal healing. Macé et al showed that the colonic mucosa of UC patients in remission remains abnormal despite a normal macroscopic appearance. Microscopic findings included impaired crypt regeneration, persistent inflammation, distinct abnormalities in angioarchitecture, and increased vascular permeability.³⁴ In CD patients, the disease severity can be assessed in real-time by CLE using the Crohn’s Disease Endomicroscopic Activity Score (CDEAS), which evaluates the crypt number, crypt distortion, microerosions, cellular infiltrate, vascularity, and number of goblet cells. This classification showed a sensitivity, specificity, and accuracy for predicting histological inflammation in macroscopically uninflamed mucosa of 94%, 81%, and 87%, respectively.³⁵

In addition, CLE may be a useful tool for the prediction and prevention of flare in IBD. Kiesslich et al demonstrated that cell shedding and barrier loss detected by CLE could also predict disease relapse.³⁶ Using their Watson scoring system, they demonstrated that increased cell shedding and fluorescein leakage in patients in remission were associated with subsequent relapse within 12 months after endomicroscopic examination (P < 0.001).

Moreover, the value of combined chromoendoscopy and CLE for the diagnosis of intraepithelial neoplasia was assessed by Kiesslich et al in 161 UC patients in clinical remission.³⁷ Confocal laser endomicroscopy could help determine if UC lesions identified by chromoendoscopy should undergo biopsy examination: the detection of neoplasia increased 4.75-fold compared with conventional colonoscopy, with 50% fewer biopsies required. Combining these 2 techniques enabled the detection of neoplastic changes with high sensitivity (94.7%), specificity (98.3%), and accuracy (97.8%).

Confocal laser endomicroscopy has been used to predict the therapeutic effect of biological therapies in IBD patients.^{4, 38} Anti-TNF therapy is clinically approved in patients with moderate to severe CD. Antibodies to TNF such as adalimumab, certolizumab pegol, infliximab, and golimumab induce immunosuppression by binding to membrane-bound TNF (mTNF), which leads to mucosal healing. However, 30%–50% of those patients do not respond to therapy with little to no improvement of clinical symptoms while experiencing side effects.⁴ A fluorescent antibody (adalimumab) with a high affinity toward mTNF was used to detect mTNF-expressing cells in 25 patients with CD using CLE imaging (binding of adalimumab to mTNF on transfected cells K_D = 483 pM).³⁹ They found that molecular imaging of the gut mucosa using the fluorescent antibody has the potential to predict therapeutic responses to biological treatment: 92% of patients with high amount of mTNF⁺ showed high short-term response rates to anti-TNF therapy, whereas 15% of patients with low number of mTNF⁺ cells did.⁴⁰

Endoscopic photoacoustic (PA) imaging combines optical spectroscopy and ultrasonography; light absorbed by molecules present in tissue induces thermoplastic expansion, which can be detected as US waves. Using different wavelengths, the relative contribution of specific molecules can be determined, such as hemoglobin and collagen. Recently, Lei et al demonstrated the ability of endoscopic PA imaging to follow the development of intestinal fibrosis over time in animal models of IBD.⁴¹ The authors reported an increase in collagen deposition detected by endoscopic PA imaging, which was validated by histopathology.

Finally, the development of multiphoton imaging endoscopy is gaining interest and has already proven its potential in the context of IBD by allowing the detection of pathological changes in human biopsy samples.^{8, 42} This technique employs a near infrared excitation wavelength to overcome the scattering of biological tissues and water absorption. Notably, it offers a superior resolution and an increased penetration depth over confocal microscopy. Molecular components of the mucosa, such as type 1 collagen, can be imaged without requiring fluorophores based on specific autofluorescence or second-harmonic generation.⁴² Recently, multiphoton imaging endoscopy has been used to image the changes in tissue morphology in a mouse model of colitis, showing collagen deposition over time, but no macroscopic change could be observed by conventional endoscopy (Fig. 2).⁴³ Multiphoton microscopy systems undergoing miniaturization will most likely become a valuable asset for clinical diagnosis of IBD.^43–46

FIGURE 2.

Example of label‐free multiphoton endomicroscopy of murine colon mucosa in vivo from Dilipkumar et al.⁴³ A, Two‐photon excited autofluorescence from endogenous molecules such as NADH or FAD (green) and SHG from collagen‐I (blue) show the epithelial crypt structure and underlying lamina of the colon mucosa. B, Focal plane shifting by means of a fast ETL for 3D visualization of the tissue morphology. C, Repeated minimal invasive multiphoton endomicroscopy of healthy murine colon as a control for the colitis model. D, Changes in the tissue morphology during the time course of intestinal inflammation in a mouse model of ulcerative colitis ranging from deformations of the crypt pattern at day 3 to massive epithelial erosion and collagen matrix formation at day 9. Images were recorded in the same animal with minimal invasive multiphoton endomicroscopy. E, Conventional wide‐field endoscopy as reference showing clear symptoms of acute murine DSS colitis from day 6 to 9 and macroscopically unchanged mucosa at day 3. Scale bars: 50 µm.

CROSS-SECTIONAL IMAGING MODALITIES

Computed Tomography

The invasiveness, limitations, and inherent risk of complications of endoscopy spurred the employment of cross-sectional imaging modalities in IBD, which are able to provide information on extraluminal complications in a noninvasive way.⁴⁷ Computed tomography (CT) is widely available and offers rapid image acquisition while providing high spatial resolution, 3-dimensional images. Computed tomography enterography (CTE) requires the use of oral contrast agents, such as barium sulfate suspension, to achieve bowel distension to visualize the interface between the bowel wall and the lumen and detect abnormalities. Computed tomography enterography is routinely used in patients with CD to assess bowel wall thickening and enhancement, mesenteric inflammation, lymph node enlargement, and extraluminal and extraintestinal complications. The main disadvantage of CT is exposure to ionizing radiation. Indeed, repetitive CT examinations expose IBD patients to high cumulative levels of radiation, increasing the risk of developing malignancies.⁴⁸ This is of special concern, as the incidence of IBD is increasing in the pediatric population. Moreover, concerns about repeated exposures have limited the use of CTE for assessing the response to therapy and monitoring the disease. To overcome this drawback, low-dose CTE techniques are now clinically used to substantially reduce radiation exposure in IBD patients while retaining sufficient diagnostic accuracy.⁴⁹

Magnetic Resonance Imaging

The absence of ionizing radiation makes magnetic resonance imaging (MRI) an appropriate alternative to CT for IBD evaluation. Moreover, MRI offers a superior soft tissue contrast than CT.¹⁰ Similarly to CTE, MR enterography (MRE) is performed after administration of oral contrast agents to visualize bowel wall edema, enhancement, and thickening.¹⁰ Enteric contrast agents are categorized as positive, negative, or biphasic depending on their effects on signal intensity on T₁- and T₂-weighted images. Biphasic contrast agents, such as water, polyethylene glycol, and barium solution, are predominantly used mixed with high osmolarity additives to reduce the intestinal absorption of water. In some institutions, antiperistaltic agents such as butylscopolamine (Europe) or glucagon (United States) are administered intravenously before CTE or MRE to minimize bowel peristalsis during scanning and prevent motion artifacts.¹⁰ In addition, gadolinium-based contrast agents are used during MR examination to visualize and stage inflammation. To date, no specific MR contrast agent has been used in humans to probe the underlying pathophysiology of IBD and obtain functional data. Preclinical examples will be discussed later.

Multiparametric MR imaging has emerged as a promising tool for detecting and monitoring disease activity and extent in IBD patients using complementary MR techniques.

Diffusion weighted MRI (DWI) explores the random motion of water molecules in the body.⁵⁰ The motion of water molecules is dependent on the cellular density of the tissue; their diffusion is restricted in tissues with normal cellularity and intact cell membranes, and the diffusion is less restricted in damaged tissues with low cellularity. In 2010, Oussalah et al evaluated the potential of DWI to detect colonic inflammation in patients with CD and UC without requiring oral and IV contrast agents and bowel preparation.⁵¹ Diffusion weighted MRI hyperintensity could accurately detect colonic inflammation in patients with UC, whereas the sensitivity was lower in CD patients. A recent prospective study demonstrated that the diagnostic accuracy of DWI for active bowel inflammation is in agreement with contrast-enhanced MRE.⁵² Notably, DWI could be beneficial in patients for whom the use of MR contrast agents is contraindicated.

Cinematic MRI, or motility MRI, is used to evaluate the bowel peristalsis and therefore detect diseased bowel segments that are characterized by altered motility.⁵³ This technique does not require the use of antiperistaltic agents. In CD, diseased bowel segments show signs of hypomotility or even paralysis. Combining cinematic MRI and MRE has demonstrated better diagnostic accuracy than MRE alone.⁵⁴

Magnetization transfer (MT)-MRI is a technique that generates image contrast between protons in free water molecules and those within water molecules associated with large molecules. Tissues containing high concentration of macromolecules, such as collagen, exhibit a higher MT ratio, making this technique of interest for detecting and quantifying intestinal fibrosis. The potential of MT-MRI to detect fibrosis was first demonstrated in rats with peptidoglycan polysaccharide–induced fibrosis.⁵⁵ The mean MT ratio showed correlation with tissue collagen levels and is sensitive to changes in fibrosis over time. Recently, Li et al evaluated the diagnostic performance of MT-MRI for characterizing intestinal fibrosis and its ability to differentiate inflammatory from fibrotic strictures in 31 CD patients using histopathology as a reference standard.⁵⁶ Normalized MT ratios could accurately differentiate fibrotic from nonfibrotic bowel walls (area under the curve [AUC], 0.981) even in the presence of inflammation.

Hybrid Modalities

Hybrid modalities such as PET and SPECT imaging used in combination with MRI or CT provide both functional and morphological information.⁵⁷ This is of importance to determine the degree of inflammation and fibrosis in the strictures, as it has important therapeutic implications. Although PET and SPECT modalities have been used in humans, PET imaging predominates in clinical practice. PET and SPECT modalities involve the use of a radiotracer that can ideally target the diseased site with high specificity and sensitivity. To date, most of the radiopharmaceuticals used in IBD inform on inflammation location. PET and SPECT tracers have been developed for the diagnosis of IBD, with technetium (^99mTc), indium (¹¹¹In), and fluorine (¹⁸F) being the most commonly used radioisotopes.

SPECT imaging has predominantly involved the use of radiolabeled autologous white blood cells (WBCs) for the detection of inflammation. Most studies were based on the use of either ^99mTc-HMPAO (hexamethylpropylene amine oxime) or ¹¹¹In-oxine.⁵⁸ Due to their lipophilicity, those agents can cross the cell membrane of WBC and are subsequently retained in the cell. The main limitation of WBC labeling is the extensive labeling protocol.⁵⁹ Briefly, ~40 mL of blood is withdrawn from the patient, and the plasma is separated. White blood cells are extracted by centrifugation while keeping the supernatant for the next step. Radioactive material (^99mTc-HMPAO or ¹¹¹In-oxine) is added to the WBC, rinsed, and diluted with the WBC supernatant before administration into the patient. This procedure is tedious and presents a risk of cross-contamination.⁶⁰ Several articles have reported the diagnostic performance of SPECT with ^99mTc-HMPAO-WBC.^61–63 According to Stathaki et al, ^99mTc-HMPAO-WBC scintigraphy has been successfully used to detect bowel lesions and to assess disease activity with a sensitivity of 95%–100%, a specificity of 85%–100%, and an accuracy of 92%–100%.⁵⁸ Other radiopharmaceuticals have been tested to overcome the complex labeling method of WBC and avoid any blood manipulation. Kerry et al investigated the potential of ^99mTc-Leukoscan, a radiolabeled monoclonal Fab antibody fragment, to detect inflammation in 22 IBD patients by targeting glycoproteins on granulocyte surface.⁶⁴ However, the sensitivity and specificity of ^99mTc-Leukoscan were lower than ^99mTc-HMPAO-WBC. Moreover, an increased ^99mTc-Leukoscan signal could be observed at later time points in normal bowel due to radiotracer excretion. More recently, Aarntzen et al examined the accuracy of ^99mTc-labeled CXCL8 (also known as IL-8) to detect and localize inflammation sites in 30 patients with CD (n = 15) or UC (n = 15) by SPECT using endoscopy as a reference standard.⁶⁵ The chemokine CXCL8 targets CXCL8 receptors (CXCR1 and CXCR2) and is involved in the recruitment of immune cells to the inflammation sites. The uptake of ^99mTc-CXCL8 in intestinal lesions proved to be higher in patients with IBD during flare-ups. SPECT showed higher accuracy and sensitivity than endoscopy, but its specificity was lower. The use of ¹²³I-IL2 and ^99mTc-IL2 has also been tested in patients with CD.^{66, 67} Interleukin-2 is a cytokine involved in T-cell differentiation and proliferation, and high affinity receptors are expressed mainly in activated T-lymphocytes and monocytes. Higher intestinal uptake of labeled IL-2 radiopharmaceuticals could be a good predictor of clinical relapse in patients with inactive CD.

PET imaging is usually preferred over SPECT because of its increased spatial resolution.⁵⁷ To date, 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG), a radiolabeled glucose analog, is the only PET tracer used for IBD diagnosis. The ¹⁸F-FDG is taken up by cells undergoing increased glycolytic metabolism and can help visualize areas of increased metabolic activity such as inflammation. In 2007, a prospective study in 17 patients with newly diagnosed IBD showed that the detection rate of PET/CTE was significantly higher than colonoscopy and barium imaging, with the detection of 50 bowel segments vs 17 and 16, respectively.⁶⁸ In a retrospective study, Ahmadi et al evaluated the role of combining FDG-PET and CTE modalities for lesion detection in 41 CD patients. A total of 79% of diseased CTE segments showed a high to mild ¹⁸F-FDG uptake. All of the 38 diseased segments identified by ¹⁸F-FDG-PET could be visualized by CTE, whereas no radiotracer uptake could be observed in some abnormal bowel segments reported by CT due to the exclusive presence of fibrosis.⁶⁹ The ¹⁸F-FDG PET/CT has also been reported as a tool for monitoring disease activity and therapeutic response.^70–73 Spier et al performed FDG-PET/CT in 5 patients with CD (n = 3) or UC (n = 2) before and after medical therapy. Some of the pretreatment active segments displayed low to no radiotracer uptake after treatment, which correlated with symptom improvement, thus supporting the potential of PET imaging for disease monitoring.⁷² Lapp et al investigated the clinical utility of ¹⁸F-FDG PET/CT in 7 patients with known or suspected IBD. The ¹⁸F-FDG PET/CT proved to be effective for clinical decision-making and adapting medical treatment in this study.⁷¹

However, PET/MRI imaging presents several advantages over PET/CT such as an exquisite anatomic details and soft tissue contrast, simultaneous acquisition of PET and MR images that minimizes spatial mismatch due to bowel peristalsis, and a substantially reduced (20%–73%) exposure to ionizing radiation.⁷⁴ Representative ¹⁸F-FDG PET and MR images are shown in Figure 3. Pellino et al compared the accuracy and clinical impact of ¹⁸F-FDG PET/CTE and ¹⁸F-FDG PET/MRI in 35 CD patients scheduled for surgery. The authors demonstrated that PET/MRE exhibited a higher accuracy for the detection of extraluminal disease and fibrosis compared with PET/CTE.¹³ Moreover, Catalano et al showed that the diagnostic performance of PET/MRI is higher than each modality alone in detecting bowel inflammation using intraoperative findings as a reference standard. The sensitivity, specificity, and diagnostic accuracy were respectively 91.5%, 74%, and 84% for PET; 80%, 87%, and 83% for MR; and 88%, 93%, and 91% for PET/MR.⁷⁴ Moreover, PET/MRE allows the discrimination of fibrosis from inflammation in patients with CD with a high accuracy.⁷⁵

FIGURE 3.

(A) ¹⁸F-FDG PET, (B) contrast-enhanced, T₁-weighted, fat-suppressed image, and (C) fused PET/MR in a Crohn’s disease patient. PET/MR shows multiple and discontinuous areas of bowel wall thickening, with associated mucosal enhancement, regions of increased ¹⁸F-FDG uptake (SUVmax 5.6–9.2) vs background bowel (SUVmax 1.5–2.8) and mild prominence of the adjacent mesenteric vessels (white arrows), in addition to pseudosacculations (red arrows). Note fibrofatty mesenteric proliferation (*). This constellation of findings is consistent with coexistence of acute inflammatory changes (white arrows) over sequelae of prior inflammation/fibrosis (red arrows and *).

Hybrid modalities using ¹⁸F-FDG PET with CT or MRI have undoubtedly shown promise for the clinical management of IBD. However, ¹⁸F-FDG is a nonspecific radiotracer and is therefore unable to differentiate among infection, inflammation, and neoplastic disease. Additionally, ¹⁸F-FDG can result in false positives due to normal physiologic activity such as peristaltic activity of the muscular layer, bacteria uptake, luminal content, or lymphoid tissue inside the mucosa. Radiotracer uptake can also seem more pronounced in the pelvis where intestinal loops collapse and in patients receiving oral hypoglycemic agents.⁷⁶

Ultrasonography

Intestinal ultrasonography has emerged as an important tool in the management of IBD from the initial diagnosis to the assessment of complication and treatment response by measuring parameters such as bowel wall thickness and vascularization. Ultrasonography presents multiple advantages over endoscopy and other imaging modalities including availability, relatively low cost, real-time capabilities, and lack of ionizing radiation. It has been shown to be as sensitive and specific as MR, CT, and endoscopy for detecting IBD.¹¹

Contrast-enhanced ultrasonography (CEUS) can further increase the capabilities of intestinal US in detecting disease location, monitoring disease activity, and treatment response. Contrast-enhanced ultrasonography involves the use of microbubbles (MBs) that consist in an inert gas core such as perfluorobutane surrounded by a shell composed of varying lipids or polymers. Microbubbles usually range in size between 1 and 4 µm and are administered intravenously. Microbubbles resonate in an ultrasound beam, rapidly contracting and expanding in response to the pressure changes of the sound wave. Contrast-enhanced ultrasonography has shown higher sensitivity in detecting postoperative recurrence in CD patients with a positive predictive value of 98% as opposed to 89.7% without MBs.⁷⁷

Finally, elastography is another technique that might be used to detect small bowel lesions. Baumgart et al performed real-time elastography pre-, intra- and postoperatively on unaffected and affected bowel segments and demonstrated the potential of this technique to differentiate fibrotic from inflammatory stenosis in small and large bowel stenosis.⁷⁸ Additional studies are required to further validate these results.

PRECLINICAL IMAGING OF IBD

Animal models of human diseases are critical tools for improving our understanding of a pathology and developing new therapeutic agents. Inflammatory bowel disease is a complex disorder with an heterogenous disease course and is therefore difficult to translate into animal models while fully reproducing the clinical manifestations, pathophysiology, and treatment response observed in humans.⁷⁹ However, animal models contribute enormously to the identification of biomarkers of IBD both for targeted molecular imaging and the development of therapeutics (Table 2). In the case of molecular imaging, the target should ideally be overexpressed in the diseased area at a concentration compatible with the imaging modality. Target concentration should also vary with disease progression or be present at specific stages of the disease to enable quantification of disease activity and treatment response.

TABLE 2.

Probes Developed or Applied for Clinical and Preclinical Molecular Imaging of IBD

Imaging modality	Probea	Molecular process / Target	Information gained / Limitations	Species	Model	Ref
Confocal endo-microscopy	Fluorescein	Loss of tight junction between epithelial cells	Intestinal permeability	Human		32
Confocal endo-microscopy	Fluorescent adalumimabb	mTNF+ cells	Prediction of treatment response	Human		38,40
Chromo-endoscopy	Methylene blue Indigo carmineb	Difference in uptake between healthy and diseased mucosa	Intraepithelial neoplasia	Human		30
PET	18F-FDG	Glycolytic metabolism	Inflammation - False positive due to normal bowel activity Radiation exposure	Human		57
SPECT	99mTc-HMPAO-labeled leukocytes	Leukocytes	Inflammation – Tedious radiolabeling and high risk of cross contamination	Human		61–63
SPECT	111In-oxine labeled leukocytes	Leukocytes	Inflammation – High radiation dose, poor image quality	Human		58
SPECT	99mTc-CXCL8	Interleukin-8 receptors (CXCR1 and CXCR2)	Inflammation - Higher accuracy and sensitivity than endoscopy but lower specificity	Human		65
SPECT	99mTc/123I-IL2	Interleukin-2 receptors	Inflammation – Higher uptake could be a predictor of clinical relapse in patients with CD	Human		66,67
SPECT	99mTc-LeukoScan	Glycoproteins on granulocyte surface	Inflammation - Not suitable – Lower sensitivity and specificity than 99mTc-HMPAO-WBC	Human		64
Multiphoton micro-endoscopy	Second harmonic generationc	Collagen I	Epithelial crypt structure and underlying lamina of the colon mucosa. Changes in tissue morphology over time	Mice	DSS	43
Multiphoton micro-endoscopy	Autofluorescencec	NAHD FAD	Epithelial crypt structure and underlying lamina of the colon mucosa. Changes in tissue morphology over time	Mice	DSS	43
MRI	Gadofluorine-M	Nonspecific ECM protein binding	Discriminate between noninflamed, mildly inflamed, and severely inflamed colon wall	Rat	DNBS	85
MRI	Ultrasmall Iron Oxide SHU 555 C	Reticuloendothelial system (macrophages and dendritic cells)	Inflammation detection – Negative contrast agent	Rat	DNBS	84
MRI	Gd-DTPA	Nonspecific agent – Extracellular distribution	Similar and weak contrast enhancement in mild or severe colitis	Rat	DNBS	85
PET	64Cu-FIB504.64	⍺ 4β 7 and ⍺ Eβ 7 integrins	Inflammation - Long blood half-life and clearance	Mice	DSS	97
PET	64Cu-DAKT32	⍺ 4β 7 integrin	Not suitable for IBD imaging – weak signal in the bowel	Mice	DSS	96
PET	64Cu-FIB504.64-F(ab’)2	β 7 integrin	Inflammation - Fast clearance from nontarget tissues	Mice	DSS	96
PET	64Cu-FIB504.64-Fab	β 7 integrin	Inflammation - Fast clearance from nontarget tissues – Lower uptake in the inflamed tissue than 64Cu-FIB504.64-F(ab’)2	Mice	DSS	96
PET	89Zr-malDFO-GK1.5 cys-diabody	CD4+ T-cells	Inflammation – Probe uptake in the colon, ceca and mesenteric lymph nodes	Mice	DSS	98
PET	89Zr-lα-IL-1β	IL-1β cytokine	Strong correlation between disease severity and %ID/cc	Mice	DSS	99
PET	89Zr-α-CD11b	CD11b integrin	Not suitable for IBD – large uptake in many organs	Mice	DSS	99
PET	18F-DPA-714	Translocator Protein TSPO	Inflammation	Mice	DSS TNBS	102
SPECT	99mTc-scFv-VCAM-1	Vascular cell adhesion molecule-1 (VCAM-1)	High signal in the colon compared with controls and blocked	Rabbits	TNBS DSS	101
Optical imaging	MMP-activatable probe	MMP	Inflammation and colon cancer	Mice	Azoxymethane	88,92
Optical imaging	Cathepsin activatable probe	Cathepsin	Inflammation - Dysplasia Colitis-associated cancer	Mice	DSS IL-10 -/-	88–90
Optical imaging	NB200	Cathepsin S	Inflammation – Specificity tested ex vivo only	Mice	DSS	91
US	MB Selectin	E- and P-selectin	Inflammation	Mice	TNBS	93
US	MB P-selectin	P-selectin	Inflammation	Mice	TNBS	94

^aProbes are administered intravenously otherwise noted. _^bTopical administration. _^cNo dye administered. Abbreviation: DNBS, dinitrobenzene sulfonic acid.

Animal Models of IBD

Numerous animal models have been developed to study pathophysiology of human IBD and screen diagnostic or therapeutic agents.⁸⁰ Chemically induced murine models of colitis are generated by ingestion through drinking water or colonic instillation of a chemical agent (eg, dextran sulfate sodium [DSS] or trinitrobenzene sulfonic acid [TNBS]), resulting in epithelium damage, loss of barrier integrity, inflammation, and subsequent fibrosis. Several models of knockout and transgenic mice strains have also been developed and exhibit a change in the expression of mediators of inflammation or fibrosis (IL-10 deficiency, TGF-β1, and MCP-1 overexpression). Moreover, certain mouse strains can spontaneously develop intestinal fibrosis. For instance, SAMP1/YitFc mice will develop inflammation and fibrosis at 10 to 20 weeks of age and are an excellent model of CD but are limited by their low breeding rate. Finally, radiation and bacteria-induced models of IBD have also been studied. The ability of those animal models to translate human intestinal inflammation and fibrosis has been comprehensively reviewed elsewhere, along with the methods of induction and underlying mechanisms.^79–82 Although those models might not relate to the pathogenesis of IBD or fully recapitulate the disease, they are powerful tools and have been extensively involved in the development of new molecular imaging probes and preclinical testing.

Magnetic Resonance Contrast Agents

Often, MRI examinations rely on paramagnetic or superparamagnetic contrast agents to improve the sensitivity and specificity of the technique by inducing signal changes or contrast in the MR images. Gadolinium-based contrast agents, commonly used in clinical settings, are small molecular weight compounds that distribute in the extracellular space and are characterized by a short blood half-life.⁸³ Other MR contrast agents include ultrasmall superparamagnetic iron oxide nanoparticles (USPIOs) or superparamagnetic iron oxide nanoparticles (SPIOs). Frericks et al examined the role of USPIO SHU 555 C to quantitatively and qualitatively characterize the extent of disease in a chemically induced rat model of IBD.⁸⁴ Those nanoparticles are negative contrast agents that are taken up into the reticuloendothelial system. Results showed a significant signal to noise ratio reduction in T₁-, T₂-, and T₂*-weighted images 90 minutes after injection, which correlated with histology findings. In a similar study, the same group compared the MR findings after administration of gadolinium-based contrast agents.⁸⁵ Rats received either Gd-DTPA, a nonspecific linear contrast agent, or Gadofluorine-M, an amphiphilic experimental contrast agent known to bind albumin and extracellular matrix proteins.⁸⁶ Compared with Gd-DTPA, Gadofluorine-M could discriminate between noninflamed, mildly inflamed, and severely inflamed colon wall and showed a high enhancement ratio up to 24 hours post injection.

Optical Imaging Probes

Near infrared optical imaging has been employed in IBD for the detection of inflammation using in vivo imaging systems or endomicroscopy. Compared with visible-range light, in vivo imaging in the NIR spectrum (650 to 900 nm) presents the advantage of deeper tissue penetration due to the reduced light absorption and scattering from biological tissue.⁸⁷ So far, research has mainly focused on imaging cells expressing enzymes specific to inflammation such as cathepsins or matrix metalloproteinases (MMPs). Cathepsin and MMP activity can be detected in vivo using substrate-based probes, also known as “smart probes,” that produce a strong fluorescence signal after onsite enzymatic cleavage. Notably, this strategy aims at providing a high target to background ratio as the probe remains inactive in absence of substrate. A series of commercially available probes, activatable by cathepsins or MMPs, has been tested in chemically induced (DSS) and chronic (IL-10 ^-/^-) mice models of colitis and showed their ability to detect and quantify colonic inflammation ex vivo.⁸⁸ Finnberg et al explored the role of a cathepsin activatable probe (ProSense 680) to detect inflammation in DSS mice model of colitis.⁸⁹ The fluorescence signal correlated with increased infiltration of inflammatory cells, epithelial atrophy, and sterilization of the mucosa. Using the same NIR probe, Gounaris et al demonstrated that cathepsin activity could help distinguish between mild inflammation, dysplasia, and colitis-associated cancer in a IL10^-/^- mice model of colitis by performing near infrared fluorescence (NIRF) endoscopy.⁹⁰ Barlow et al designed an activatable NIRF probe (NB200) that could discriminate between cathepsin S and other inflammatory proteases in homogenized DSS mouse colons.⁹¹ Finally, Yoon et al developed an MMP-activatable NIR probe for the detection of colon cancer.⁹² The probe is based on a polymeric nanoparticle platform functionalized by a NIR dye (Cy5.5) and a NIR dark quencher (BHQ-3). The MMP-activatable probe was able to discriminate between normal, adenoma, and adenocarcinoma in this study. In clinical settings, specific NIRF probes could complement endoscopy by providing higher specificity and diagnostic accuracy. However, MMP and cathepsin probes can be activated in presence of inflammation or cancerous cells and therefore lack specificity.

Targeted Microbubbles for US Imaging

Monitoring inflammation in IBD at the molecular level can be performed by CEUS imaging using functionalized microbubbles. The cell adhesion molecules E- and P-selectin, markers of inflammation in IBD, have been targeted using MBs functionalized by P-selectin glycoprotein ligand 1, which is the natural ligand of both E- and P-selectin expressed on the surface of leukocytes. By functionalizing the MBs with an analog of the naturally occurring binding ligand P-selectin glycoprotein ligand 1 fused to a human Fc domain, Wang et al demonstrated that US with MB_Selectin enables detection and quantification of inflammation in a TNBS model of colitis. A strong correlation between in vivo US imaging with MB_Selectin and ¹⁸F-FDG uptake on PET/CT images was found (ρ = 0.89; P < 0.001).⁹³ Similarly, Deshpande et al have demonstrated that P-selectin-targeted MBs can be used to monitor the expression of P-selectin in TNBS-treated mice. In vivo US signal with use of MB_Pselectin quantitatively correlates (ρ = 0.69; P = 0.04) with P-selectin expression levels on inflamed intestinal vascular endothelial cells as assessed with ex vivo quantitative immunofluorescence.⁹⁴ Bachmann et al performed CEUS imaging in the SAMP1/YitFc mouse model of CD using MBs specific to the mucosal addressin cellular adhesion molecule-1 (MAdCAM-1) overexpressed during inflammation. Their MBs specifically accumulated in areas of inflammation and produced stronger acoustic echoes, measured by average video intensity, in SAMP mice as opposed to control mice (P < 0.001) or SAMP mice compared with nonspecific MBs (P < 0.001). A strong positive correlation was obtained between the video intensity after administration of MAdCAM-1 specific MBs with disease severity evaluated as total inflammatory scores (R² = 0.92).⁹⁵

PET and SPECT Probe Development

Recent advances in nuclear medicine have been focusing on the development of targeted PET and SPECT probes to image biomarkers specific to IBD. Radiolabeled antibodies, antibody fragments, or small molecular weight probes have been tested in animal models of colitis to image immune mediators or immune cells.

Notably, β ₇ integrin has been targeted using 64Cu-labeled antibodies and antibody fragments. The β ₇ integrin subunit expressed on lymphocytes activated in Peyer patches and mesenteric lymph nodes, when paired with the α ₄ subunit, is implicated in the recruitment of lymphocytes to the gastrointestinal tract through binding to MAdCAM-1.⁴ When β ₇ is paired to α _E subunit, the α _Eβ ₇ is involved in lymphocyte retention in the gut. As such, the anti-α ₄β ₇ integrin antibody vedolizumab is currently used for clinical treatment of IBD and decreases the inflammatory response by impeding the recruitment of lymphocytes through the blockage of the α ₄β ₇ integrin–MAdCAM-1 interaction.⁴ Dearling et al reported an increased uptake of the ⁶⁴Cu-labeled FIB504.64 antibody specific to the integrin β ₇ in the large intestine of mice treated with DSS compared with controls and after administration of the ⁶⁴Cu-labeled anti-α ₄β ₇ antibody DATK32.^{96, 97} The difference in uptake between the 2 antibodies is explained by a smaller α ₄β ₇ population in the colon of DSS injured mice. The long blood half-life and clearance of antibodies led to the use of antibody fragments including ⁶⁴Cu-labeled F(ab’)₂ that showed a large intestinal uptake and a fast clearance from nontarget tissues.⁹⁶

More recently, Freise et al studied the infiltration of CD4⁺ T-cells in a DSS mouse model using ⁸⁹Zr-GK1.5 cys-diabody, a ⁸⁹Zr labeled antimouse CD4 fragment.⁹⁸ The authors reported a higher uptake of the probe in the colon, ceca, and mesenteric lymph nodes of mice treated with DSS compared with healthy controls. Dmochowska et al compared PET imaging of antibodies with the cytokine IL-1β and the integrin CD11b, key mediators of inflammatory response.⁹⁹ PET imaging was performed in DSS mice using either ⁸⁹Zr-α-IL-1β and ⁸⁹Zr-α-CD11b in comparison with unenhanced MRI and ¹⁸F-FDG. The authors reported a statistically significant correlation between disease severity (percentage of body weight loss) and ¹⁸F-FDG uptake (P < 0.05), in addition to ⁸⁹Zr-α-IL-1β uptake (P = 0.09). However, no correlation could be observed for ⁸⁹Zr-α-CD11b, which exhibited a large uptake in many organs and is therefore not suitable for IBD imaging.

Radiolabeled antibodies and antibody fragments have also been employed for SPECT imaging of IBD.^{100, 101} For example, Liu et al targeted the vascular cell adhesion molecule-1 (VCAM-1) using a ^99mTc-labeled single-chain antibody fragment (^99mTc-scFv-VCAM-1) and showed a high uptake in the colon compared with controls.¹⁰¹ Other examples of radiolabeled immune cells and mediators for SPECT have been discussed previously.

Finally, Bernards et al used ¹⁸F-DPA-714, a translocator protein (TSPO) radioligand, to quantify the inflammation in DSS and TNBS treated rats.¹⁰² Translocator protein, also known as peripheral benzodiazepine receptor (PBR), is located on the outer mitochondrial membrane and is overexpressed in activated macrophages and microglia and in the colon of the DSS rat model. Results indicated that the PET signal originating from ¹⁸F-DPA-714 correlated with increased TSPO expression at cellular level and that the difference in signal uptake in healthy vs DSS- or TNBS-treated rats was significant. In this study, ¹⁸F-FDG showed a significant difference only in TNBS-treated rats.

IMAGING FIBROSIS: A MAJOR UNMET NEED

As demonstrated throughout this review, molecular imaging in the context of IBD has been primarily directed at detecting inflammatory pathways. In chronic inflammatory disorders such as IBD, repeated inflammation episodes drive the deposition of ECM components, such that 30% of patients with CD and 5% of patients with UC develop a fibrostenosing phenotype.¹⁰³ Despite the suppression of inflammation through therapeutic medications or surgical removal of stricture, strictures tend to recur, leading to repeated surgeries.¹⁷ Identifying markers of intestinal fibrosis would contribute to the development of new diagnostic and therapeutic approaches to help evaluate the risk of patients developing fibrosis and detect early stages of fibrosis to minimize symptom occurrence. Moreover, because fibrosis and inflammation often coexist, evaluating the degree of fibrosis using conventional imaging seems limited. Although some imaging modalities such as MT-MRI, PA imaging, or US elastography have shown promise in detecting fibrosis, there is an urgent need to develop new diagnostic tools able to detect and quantify fibrosis in presence of inflammation.

Fibrosis is the result of an imbalance between matrix metalloproteinases and tissue inhibitors of matrix metalloproteinases, leading to abnormal ECM buildup and tissue remodeling.¹⁷ Fibrosis is a major complication of many diseases such as atrial fibrillation, hypertrophic cardiomyopathy, nonalcoholic steatohepatitis (NASH), idiopathic pulmonary fibrosis (IPF), scleroderma, and chronic kidney diseases. It occurs in all organs and is associated with the loss of organ function. Notably, pathogenic pathways involved in intestinal fibrosis seem to display similarities with scleroderma and hepatic fibrosis.²⁸

Multiple PET, SPECT, and MR probes have been successfully designed to detect unique biochemical and cellular markers of fibrosis and fibrosis-associated processes in extraintestinal fibrotic diseases.^{104, 105} Type 1 collagen is a major component of fibrotic tissue present in the extracellular space at a concentration above 10 µM, thus making it accessible and compatible with both MR and PET imaging modalities. Collagen synthesis starts with the formation of procollagen trimers in fibroblasts followed by secretion of collagen monomers in the extracellular matrix; their self-assembly and cross-linking form mature collagen fibrils.¹⁰⁵ During fibrogenesis, lysyl oxidase (LOX) and LOX-like enzymes are upregulated and catalyze the oxidation of the collagen lysine ε-amino groups to allysine aldehydes, a precursor of cross-linked collagen. The abundance of allysine aldehyde makes it a suitable target for detection of active fibrosis.¹⁰⁵

Although not yet applied to the study of fibrosis in IBD, type 1 collagen-targeted MR and PET probes have been reported that can detect and stage fibrosis in animal models of lung, cardiac, and liver fibrosis.^106–108 For instance, EP-3533 is a peptide functionalized with 3 Gd-DTPA chelates that binds type 1 collagen reversibly and is suitable for MRI.^{109, 110} In mouse models of myocardial infarction, EP-3533 showed prolonged enhancement of the infarcted myocardium arising from specific collagen binding. The same probe was used in animal models of liver and lung fibrosis and showed a high correlation between MR signal enhancement and collagen content and the ability to use molecular imaging to stage fibrosis accurately. Other PET and SPECT probes have also been developed to specifically target collagen such as ⁶⁸Ga-CBP8, ^99mTc-collagelin, and ^99mTc-CBP1495.^111–113 In animal models of lung fibrosis, ⁶⁸Ga-CBP8 could detect and monitor disease activity and was successfully used to evaluate the effect of an antifibrotic treatment (α _vβ ₆-targeted antibody).¹¹¹ Additionally, ⁶⁸Ga-CPB8 was recently used in IPF patients. Results of the first in-human studies demonstrated the ability of the tracer to noninvasively measure increased lung collagen in subjects with IPF.¹¹⁴ Fibrogenesis has been imaged using allysine-targeted MR probes such as Gd-Hyd, Gyd-OA, and Gd-CHyd.^115–117 Those probes showed specific allysine targeting in a pulmonary fibrosis mouse model and a carbon tetrachloride liver fibrosis model and could also be used to monitor treatment response. Finally, the probe Gd-ESMA has been developed to image elastin, an ECM protein that accumulates late in the fibrosis process. The Gd-ESMA probe could noninvasively detect a plaque burden in a mouse model of atherosclerosis.¹¹⁸ Molecular probes for imaging fibrosis and fibrogenesis were recently included in a comprehensive review.¹⁰⁵ Exploring the potential of those probes in animal models of IBD and ultimately in patients could definitely advance our understanding of intestinal fibrosis and facilitate the development of new therapeutics.

CONCLUSION

Considerable progress has been made in the development of noninvasive imaging modalities to provide both anatomical and functional information for the diagnosis of IBD. But there is still an urgent need for detection of specific pathophysiological processes and identification of new biomarkers to advance drug development and improve patient care. Molecular imaging in IBD could enable disease staging and monitoring, early detection of disease recurrence, and prediction and evaluation of treatment response. So far, the development of molecular imaging probes has been mainly focused on targeting inflammation processes, immune cells, and mediators. However, intestinal fibrosis is a major complication of IBD for which surgical resection is currently the only treatment option, but there are antifibrotic therapies under development. As such, the study of intestinal fibrosis remains a major unmet need: understanding the pathogenesis of intestinal fibrosis could potentially lead to the development of more efficient diagnostic tools and novel therapeutics.

Glossary

Abbreviations:

CD

Crohn’s disease

CEUS

contrast enhanced ultrasonography

CLE

confocal laser endoscopy

CT

computed tomography

DNBS

dinitrobenzene sulfonic acid

DSS

dextran sulfate sodium

DWI

diffusion weighed imaging

ECM

extracellular matrix

IBD

inflammatory bowel disease

IPF

idiopathic pulmonary fibrosis

MAdCAM-1

mucosal addressin cellular adhesion molecule-1

MMP

matrix metalloproteinase

MRI

magnetic resonance imaging

MRE

magnetic resonance enterography

MT

magnetization transfer

NASH

nonalcoholic steatohepatitis

LOX

lysyl oxidase

PA

photoacoustic

PET

positron emission tomography

SHG

second harmonic generation

SPECT

single photon emission computed tomography

TNBS

trinitrobenzene sulfonic acid

UC

ulcerative colitis

US

ultrasonography

VCAM 1

vascular cell adhesion molecule 1.

Supported by: Study was supported by the National Institute of Diabetes and Digestive and Kidney Diseases with grants DK104956, DK104302, and DK121789.

Conflicts of Interest: PC has equity in and is a consultant to Collagen Medical LLC, which owns the patent rights to EP-3533 and CM-101; PC has equity in Reveal Pharmaceuticals Inc. and has research support from Pliant Therapeutics, Celgene, Takeda, and Indalo Therapeutics.