The role of molecular imaging in detecting fibrosis in Crohn’s disease

Abstract

Fibrosis is a pathological process that occurs due to chronic inflammation, leading to the proliferation of fibroblasts and the excessive deposition of extracellular matrix (ECM). The process of long-term fibrosis initiates with tissue hypofunction and progressively culminates in the ultimate manifestation of organ failure. Intestinal fibrosis is a significant complication of Crohn’s disease (CD) that can result in persistent luminal narrowing and strictures, which are difficult to reverse. In recent years, there have been significant advances in our understanding of the cellular and molecular mechanisms underlying intestinal fibrosis in inflammatory bowel disease (IBD). Significant progress has been achieved in the fields of pathogenesis, diagnosis, and management of intestinal fibrosis in the last few years. A significant amount of research has also been conducted in the field of biomarkers for the prediction or detection of intestinal fibrosis, including novel cross-sectional imaging modalities such as positron emission tomography (PET) and single photon emission computed tomography (SPECT). Molecular imaging represents a promising biomedical approach that enables the non-invasive visualization of cellular and subcellular processes. Molecular imaging has the potential to be employed for early detection, disease staging, and prognostication in addition to assessing disease activity and treatment response in IBD. Molecular imaging methods also have a potential role to enabling minimally invasive assessment of intestinal fibrosis. This review discusses the role of molecular imaging in combination of AI in detecting CD fibrosis.

Keywords:

• Crohn’s disease
• fibrosis
• molecular imaging
• PET
• SPECT

Introduction

Chronic inflammation causes disruption of the tissue destruction and epithelial barrier in inflammatory bowel diseases (IBD) (Crohn’s disease and ulcerative colitis). Fibrosis, which is a natural healing response, can become problematic and harmful in the context of chronic IBD where persistent tissue damage and subsequent healing processes lead to the formation of scar tissue [1]. At the tissue and cellular levels, fibrosis represents an amplified and dysregulated response characterized by the excessive deposition of collagen-rich extracellular matrix (ECM). This aberrant ECM accumulation arises from an increased presence of mesenchymal cells, including smooth muscle cells (SMCs), fibroblasts and myofibroblasts [2,3]. Fibrosis results in the irreversible stiffening of tissue along with functional impairment, serving as a prevalent pathway that contributes to morbidity and mortality in chronic diseases. The hallmark features of intestinal strictures in IBD are marked by the proliferation of fibroblastic cells and the concomitant accumulation of ECM. Fibromuscular and inflammation changes are transmural, leading to stricture development and progressive thickening of the bowel wall, even in the absence of inflammation CD. From a pathological perspective, intestinal fibrosis in Crohn’s disease (CD) is distinguished by the buildup of ECM and the expansion of mesenchymal cells, impacting all layers of the intestinal wall throughout the entire length of the intestine [4]. The process of intestinal fibrosis development is intricate, and its exact mechanism remains unclear. However, it is widely accepted among researchers that intestinal fibrosis progression involves the following stages: attraction of inflammatory cells, generation of transforming growth factor (TGF-β1), cellular injury and stimulation of myofibroblasts and cells responsible for collagen production [5]. Until recently, intestinal fibrosis was commonly viewed as an inevitable complication of IBD in patients who remained unresponsive to anti-inflammatory treatments. Consequently, surgical intervention was often necessary in such cases. While there is an increasing understanding of the pathophysiology of intestinal fibrogenesis, the clinical evaluation of the extent of fibrosis remains challenging. Therefore, an accurate and dependable diagnostic approach for identifying fibrotic segments of the bowel is crucial to facilitate optimal therapeutic decisions. The development of new anti-fibrotic drugs requires a non-invasive imaging-based method for accurately quantifying the extent of fibrosis. This approach is essential for selecting patients who are likely to benefit from the treatment and for determining the potential therapeutic benefits and efficacy of the drug.

Medical imaging modalities play a crucial role in the diagnosis and treatment of IBD. It also provides a dependable means of evaluating active transmural inflammation and identifying intestinal strictures in individuals with CD [6]. In the absence of ideal non-invasive approaches, endoscopic assessment, in conjunction with biopsy, histopathology, stool analysis and blood analysis, remains the gold standard for the diagnosis of CD [7]. However, endoscopic evaluation is limited by its inability to assess the nature of strictures beyond the mucosal layer and to provide information on extraluminal complications. Additionally, endoscopy necessitates bowel preparation and occasionally anaesthesia, which may not be well-tolerated, particularly in the pediatric patient population. These factors can have a negative impact on adenoma detection if the bowel preparation is inadequate.

Cross-sectional imaging techniques, such as magnetic resonance enterography (MRE), ultrasound (US) and computed tomography (CT), are utilized in conjunction with endoscopy to provide a comprehensive diagnostic evaluation and have demonstrated a favourable diagnostic accuracy of diagnosing CD [8].

Molecular imaging of CD holds substantial promise in advancing our understanding of its pathophysiology by effectively targeting and visualizing molecular and cellular events involved in the disease process. The fact that the molecular pathways involved in fibrogenesis span many fibrotic disease processes is especially interesting in the application of molecular imaging to fibrosis. All these modalities exhibit a high level of accuracy in detecting strictures, irrespective of the specific definition or criteria used to define and classify strictures. Although imaging techniques have shown promising results in detecting and characterizing fibrosis in CD, however, little is known about the role of molecular imaging in diagnosing CD fibrosis. This literature review aims to summarize the current knowledge on the role of molecular imaging in detecting fibrosis in CD as well as well as a future outlook into what is yet to come in this field of CD, such as machine learning (ML) and radiomics. This overview may be useful to a wider audience concerned with the understanding of diagnosing or detecting CD fibrosis. This overview will further enable physicians to better understand the added value of molecular imaging techniques that use molecular probes or tracers to target specific molecular pathways or cellular processes involved in fibrosis. An extensive review of the literature published in English was carried out using the Google Scholar and PubMed databases. The search terms included. The search terms included Chron’s disease (OR CD OR CD fibrosis), molecular imaging (OR SPECT OR PET) (OR CT-PET OR CT-SPECT OR MRI PET OR MRI-SPECT).

Imaging as non-invasive techniques to access fibrosis

Assessing fibrosis in IBD is a complex task and finding non-invasive approaches for accurately evaluating fibrosis remains a difficult endeavor. Despite the advancements in imaging and molecular technologies, the characterization, the precise identification and quantification of intestinal fibrosis in CD continue to rely on the histopathological examination of surgical specimens [9]. However, it is important to note that there is currently no universally validated or widely accepted scoring system based on histopathological evaluation specifically designed for this purpose [10, 11].

Histopathological analysis remains the established gold standard for identifying and assessing both inflammation and fibrosis in CD with the unavoidable difficulties of being invasive and limiting this study of intestinal damage to “end-of-stage” disease. In recent years, efforts have been made to migrate toward noninvasive techniques for assessing CD fibrosis.

Recently, several advance medical imaging techniques have been used to evaluate CD fibrosis mainly specific MRI sequences such as intravoxel incoherent motion imaging (IVIM), diffusion-weighted imaging (DWI) [12,13], diffusion kurtosis imaging (DKI) [14] and magnetization transfer (MT) [15,16], US elastography [17,18] and CTE [19]. However, none of these imaging techniques have been conclusively established as reliable for achieving this specific objective [20,21]. For example, MT is a type of MRI sequence that uses a contrast mechanism sensitive to intestinal collagen, and native T1 mapping, a quantitative technique capable of distinguishing fibrosis characteristics, have been identified as potential advances in the field of MR for bowel fibrosis detection and differentiation [15]. MT-MRI has demonstrated promising results in accurately identifying and measuring intestinal fibrosis in animal (rat) studies [15,22] as well as small cohort human studies [23,24]. In a study conducted by Li et al. in 2018, it was demonstrated that MT showed a significant positive correlation with fibrosis scores (r = 0.769, p = 0.000), indicating a strong association between MT ratio (MTR) and the extent of fibrosis. Furthermore, MTR was found to effectively differentiate between moderately-severely fibrotic bowel walls and non-fibrotic or mildly fibrotic bowel walls [23].

The use of DWI is another example of using advanced sequence in CD fibrosis. DWI is an advanced sequence that facilitates the visualization and mapping of water molecules’ diffusion in biological tissues. In patients with CD, a restricted diffusion pattern detected by DWI in the bowel wall is linked to the existence of active inflammation [25]. It is highly probable that the existence of fibrotic tissue results in a decrease in the extracellular space, leading to restricted diffusion of water molecules in the affected bowel wall and a subsequent reduction in the apparent diffusion coefficient (ADC) value. Tielbeek et al. (2014) conducted a study that revealed a correlation between the ADC values and the presence of fibrosis in histopathological specimens obtained from patients with CD [26]. According to two studies conducted on pediatric populations, bowel segments with a fibrostenotic imaging phenotype were observed to have lower ADC values in comparison to those that demonstrated an inflammatory imaging phenotype [27,28]. Overall, relatively new technologies, along with multiple developments and improvements based on existing technologies, have the potential to revolutionize the field of imaging fibrosis, hence enhancing diagnostic outcomes for innumerable patients. Despite the effectiveness these modalities, such as MRI and ultrasound, in detecting active transmural inflammation and intestinal strictures in CD patients, accurately estimating tissue composition, particularly fibrosis, and distinguishing between inflammation and fibrosis, remains a considerable challenge [29].

Basic principle of molecular imaging

Nuclear medicine molecular imaging operates on the principle of introducing tiny molecular probes into the body, which selectively bind to specific sites based on their distinct properties. Molecular imaging such as PET, SPECT and certain MRI sequences relies on either employ exogenous tracers (ligand or passive targeted) or direct detection of endogenous molecules that are tailored to a particular modality of choice such as radiolabeled tracers for PET/SPECT imaging or fluorescent probes for near-infrared (NIR) optical imaging to identify inflamed tissue [30]. Traditionally, exogenous tracers have typically been molecules with a small size of less than 2 kilodaltons (kDa) or less than 4 nanometres (e.g. chelated gadolinium for MRI and 18 F-FDG for PET and iodinated compounds for CT. Although small molecule tracers have made significant contributions to disease diagnosis, especially in oncology, they do have notable limitations when it comes to molecular imaging. These limitations include it has a rapid clearance from the body, relatively poor signal strength and potential toxicity and lack of adaptation to allow detection via imaging.

The role of molecular imaging in diagnosing CD in general

PET is a promising molecular imaging technique for the visualization and quantification of molecular processes underlying fibrosis. It is a highly sensitive, quantitative, and non-observer-dependent medical imaging modality that employs radiolabeled molecules to track biological processes and receptors. This makes PET an attractive option for the detection and monitoring of fibrosis, as it can provide accurate and reliable measurements of molecular changes associated with fibrosis in vivo [31].

Hybrid modalities such as SPECT and PET imaging used in combination with CT and MRI to provide combination of anatomic morphologic characteristics with functional metabolic information [32]. This is significant for determining the degree of inflammation and fibrosis in the strictures since it has important therapeutic implications. In clinical practice, the most utilized PET imaging agent is Fluorine-18-fluorodeoxyglucose (18 F-FDG). 18 F-FDG is capable of being absorbed by inflammatory cells, allowing for the precise localization of sites of inflammation in CD [33–36]. The role of combining CTE and FDG-PET modalities was evaluated to detect lesion in 41 CD patients in a retrospective study. The study found that a significant proportion of the diseased segments identified through CTE exhibited a notable level of 18 F-FDG uptake, ranging from high to mild, in approximately 79% of cases. In the study, all 38 diseased segments that were detected using 18 F-FDG-PET were also successfully visualized using CT. However, it was noted that in certain cases where there was exclusive presence of fibrosis, CT did not show any radiotracer uptake in the abnormal bowel segments [37]. In another prospective study involving 17 patients with newly diagnosed IBD showed that the detection rate of PET/CTE was significantly higher than barium imaging and colonoscopy. The study found that PET/CTE successfully detected 50 bowel segments, whereas barium imaging and colonoscopy detected only 16 and 17 segments, respectively [38].

In addition, the utilization of 18 F-FDG PET/CT has been recognized as a valuable approach for monitoring both disease activity and assessing therapeutic response in IBD [39–41]. For example, one study was aimed to investigate the clinical utility of 18 F-FDG PET/CT in 7 patients with known or suspected IBD. The 18 F-FDG PET/CT proved to be effective for clinical decision-making and adapting medical treatment in this study [41]. Spier et al. conducted a study using FDG-PET/CT in five IBD patients before and after medical therapy. They observed that active segments showing high radiotracer uptake prior to treatment displayed decreased or no uptake after therapy. This reduction in uptake correlated with symptom improvement, indicating the potential of PET imaging for monitoring disease activity and therapeutic response [41].

PET-MR and PET-CT imaging techniques can integrate both anatomical morphology and functional metabolic information to enable comprehensive diagnostic imaging. These techniques display comparable diagnostic accuracy in detecting pathological segments (with approximately 85% efficacy) [34,42]. In study included 35 CD patients scheduled for surgery aimed to compare the accuracy and clinical impact of 18 F-FDG PET/MRI and 18 F-FDG PET/CTE in these patients. The study found that as compared to PET/CTE, PET/MRE had a greater accuracy for detecting extraluminal disease and fibrosis [34]. Furthermore, using intraoperative findings as a reference standard, Catalano et al. demonstrated that the diagnostic performance of PET/MRI is superior to each modality alone in detecting intestinal inflammation [42]. The authors reported that MRI had sensitivity, specificity, and diagnostic accuracy of 80%, 87%, and 83%, respectively; PET had 91.5%, 74%, and 84%, and PET/MR had 88%, 93%, and 91% [42]. However, PET-MR is superior to PET-CT in its ability to identify intestinal fibrosis [34,43].

Several previous studies have shown that PET-CT provides benefits over PET scan in CD [44]. This imaging modality, known for its high sensitivity and specificity [45,46], has demonstrated superior capabilities and effectiveness compared to PET or CT alone [46]. Lapp et al, conducted a study using PET/CT scan in patients with known or suspected IBD. Pelvic and abdominal images were obtained using a multi-slice PET/CT scanner, where intravenous administration of FDG was performed at a dose of 0.14 mCi/kg. Subsequently, a low-dose, non-contrast CT scan was conducted with a current of 120 mA. In this specific CT component, the five observed bowel regions encompassed the small bowel, ascending colon, transverse colon, descending colon, and rectosigmoid colon. The study involved a cohort of participants consisting of four patients diagnosed with Crohn’s disease (CD), one patient with suspected IBD, and two patients who underwent ileal pouch-anal anastomosis surgery for UC. The CT/PET scan demonstrated significant value in guiding clinical decision-making regarding treatment selection and therapy, surpassing the limitations of standard tests that underestimated the activity of CD and yielded inaccurate result [40].

PET/MRE as hybrid imaging modality that combines metabolic information obtained from FDG PET with detailed anatomical imaging and enhanced soft-tissue contrast provided by MRI. The integration of PET/MRI imaging in the evaluation of Crohn’s disease (CD) patients provides valuable additional information regarding the metabolic activity of bowel segments, complementing the morphological and physiological data obtained. This integration, along with the utilization of hybrid biomarkers, offers exceptional prospects for enhancing the management and treatment strategies for CD patients [47]. The incorporation of combined PET/MRI metrics, including the assessment of ADC and metabolic inflammatory volume, contributes supplementary diagnostic value beyond the conventional magnetic resonance index of activity (MaRIA) score. Combining the aforementioned metric with the combined faecal levels of calprotectin and C-reactive protein can effectively differentiate between active and quiescent CD with high sensitivity and specificity, 83% and 100%, respectively [48]. Moreover, PET/MR index has been observed to exhibit an equal correlation with the Simple Endoscopic Score for CD [49]. PET-compared to PET-CT, PET-MRI at detecting extraluminal disease and can be used to determine whether patients are more likely to require faecal diversion during surgery. PET-MRI is superior to both PET-CT and MRI alone due to the quality of images is significantly higher for PET-MRI than for PET-CT and provides functional images that are not available with conventional MRI [34]. Furthermore, PET/MR demonstrates superior performance compared to each sub-modality alone, with a diagnostic accuracy of to 83% for MRI compared to 91% for PET/MR and 84% for PET. Furthermore, PET/MR imaging demonstrates enhanced specificity compared to Pet al.one, with PET/MR achieving a specificity of 93% compared to 74% for Pet al.one. Moreover, PET/MR imaging demonstrates enhanced specificity compared to Pet al.one, with PET/MR achieving a specificity of 93% compared to 74% for Pet al.one. Additionally, PET/MR exhibits comparable specificity to MR imaging, which is reported at 87%, in the detection of active inflammation in CD patients [42].

In general, decreased scan time, quantifiable functional data, unique tissue characterization and enhanced attenuation correction have led to new clinical applications for molecular imaging modality such as PET/MRI [50]. The co-acquisitions of PET-CT or PET- MRI allows for the combination of structural and functional information, resulting in a more comprehensive understanding of the data. By integrating these two techniques, the strengths of each modality complement and enhance one another [50].

In terms of detecting fibrosis, PET-MRI has been reported more accurately than both MRI alone and PET-CT. Using PET-MRI to identify patients for a trial with rescue medical treatment prior to surgery revealed that more than 70% of them did not require surgery [34]. Molecular imaging integrates anatomic and molecular data by utilizing ultrasound, MRI, PET, SPECT, and optical imaging in conjunction with imaging probes (Table1) [51].

The role of molecular imaging in diagnosing fibrosis and fibrogenesis

Molecular imaging is the in vivo characterization and measurement of molecular and cellular processes. It entails tissue identification and validation with high-affinity probes. Probes can be high molecular weight affinity ligands (recombinant proteins, monoclonal antibodies, peptides) or smaller molecules (enzyme substrate receptor ligands) [52].

Molecular probes can be used in different imaging modalities, such nuclear imaging techniques, such as SPECT and PET and there is no ideal modality. Both PET and SPECT techniques have the capability to detect extremely low concentrations in the picomolar range, enabling the detection of a wide range of biological targets. PET imaging, in particular, offers the advantage of providing absolute quantification of the detected signals. Notwithstanding their advantages in sensitivity and absolute quantification, these modalities exhibit lower spatial resolution compared to other imaging techniques. Additionally, the requirement for daily probe production and the involvement of ionizing radiation render these methods less suitable for longitudinal monitoring of patients over extended periods [53].

MRI has many advantages such as no requirement for ionizing radiation, imaging deep into tissue, the use of shelf-stable molecular probes, the advantages of relatively high spatial resolution (down to 0.01 mm), and additional image contrasts that yield functional and anatomical information in addition to the molecular image but is more limited in the range of molecular targets that can be detected. CT is not restricted by tissue depth and provides even higher resolution but, soft tissue contrast is poor and extremely high concentrations are necessary for detection in molecular imaging. US techniques are widely available, portable, and low cost, but operator dependent. It demonstrates high spatial resolution but has limited depth penetration. Nanomolar concentrations can be easily identified using nanoparticles, while tiny molecules normally require micromolar concentrations to be detected. Nuclear modalities, such as PET and SPECT, on the other hand, can image deep into tissue, and have exquisite sensitivity for detection (pm) and but compared with CT or MRI, it has much lower spatial resolution (0.5–10 mm). All mentioned modalities are expensive to be infrastructure [52].

In CD patients fibro-stenosing lesion displays specific histological characteristics such as muscularis propria and thickening of the muscularis mucosae that causes luminal narrowing and hardening of the intestinal wall [54]. In fibro-stenosing lesions, smooth muscle hyperplasia of the submucosa and hypertrophy of the muscularis propria are relatively prevalent, although fibrosis itself, defined as a deposit of collagenous fibres, contributes to a lower extent [55]. Despite being only partially understood, fibrogenesis in IBD is dependent on two simultaneous and parallel processes: the growth of the ECM in all layers and the expansion of the smooth muscle layers [56].

Direct molecular imaging with probes that target the components involved in fibrogenesis and fibrosis provides an innovative approach, with potential high specificity, to expand the diagnostic capabilities of current imaging methods. To date, several targets have been explored for fibrogenesis and fibrosis imaging, such as activated macrophages; activated fibroblasts; and other markers of integrins inflammation and ECM proteins, such as elastin, oxidized collagen overexpression, type I collagen synthesis, fibronectin, and fibrin. Recently, a type I collagen-targeted MR imaging probe was utilized in a mouse model to stage intestinal fibrosis CD; its efficiency was compared to that of the MR ageing contrast medium gadopentetatedimeglumine (Gd-DTPA) [57]. In comparison to Gd-DTPA, the probe exhibited a better improved effect and a greater impact. As a result, the imaging probe has potential for assessing the progression of intestinal fibrosis and following the therapy response in CD patients.

The role of molecular imaging in personalized medicine

The term of personalized medicine refers to the tailoring of medical treatment to each patient’s specific and unique characteristics. This aims to improve timing of health care and improve stratification by using biological biomarkers and information at the level of molecular disease pathways, metabolomics proteomics and genetics—essentially, embracing the concept of the application of system bioinformatics and biology and in health care. Systems biology enables the examination of how all the components within a specific biological context interact with each other [58]. In the context of IBD, personalized medicine has the potential to reduce uncertainty, to increase disease control, and improve quality of life for patients. As it entails examining the impact of environmental factors (so-called exposome), the gene expression at the mucosal levels including coding and non-coding RNAs (proteome and transcriptome) and the possible non-genetic modification of the genome (epigenome), predisposing genetic factors (genome), and the differences in gut microbiota and their produced metabolites (metabolome and microbiome). In patients with IBD personalized medicine might sound unrealistic, especially due to the complex pathophysiology and heterogeneity of the disease.

The role of molecular imaging in diagnosing intestinal fibrosis

As mentioned earlier, IBD is characterized by the ECM proteins in the intestine, resulting in fibrosis and structuring. Myofibroblasts have been identified as the primary cellular mediators of collagen deposition in IBD, with collagen types I and III being the predominant ECM proteins associated with intestinal fibrosis. These findings highlight the important role of myofibroblasts and ECM proteins in the pathogenesis of intestinal fibrosis in IBD [5]. Evidence suggests that intestinal fibrosis can persist in a self-perpetuating manner even in the absence of inflammation. The stiffness of the ECM has been identified as a key factor in perpetuating collagen deposition and promoting fibrosis. Specifically, increased ECM stiffness can trigger a positive feedback loop that leads to further ECM production and deposition, contributing to the ongoing progression of fibrosis [59,60]. Targeting signaling pathways that lead to the upregulation of intestinal fibrosis in individuals with IBD represents an important focus for both diagnostic and therapeutic purposes. By identifying and targeting these pathways, it may be possible to prevent or slow the progression of fibrosis in IBD, potentially improving patient outcomes and quality of life [61,62].

For optimal treatment planning in patients diagnosed with CD, it is imperative to achieve precise discrimination between fibrotic and inflammatory strictures. This differentiation holds significant importance as medical therapy is generally favoured for managing inflammatory strictures, whereas surgical intervention becomes frequently unavoidable in the case of fibrotic strictures [63,64]. In CD, this differentiation quite difficult as some bowel strictures, can be a mixed fibrotic and inflammatory nature, not solely fibrotic or not solely inflammatory [65]. Precise differentiation of these three types of strictures plays a crucial role in minimizing unwarranted surgical interventions [66].

Nuclear medicine and MR molecular imaging comprise the majority of molecular imaging for CD. Nuclear medicine imaging has been utilized for a long time, and PET is currently the most advanced display device. Molecular imaging is utilized in humans and other living systems to visualize, characterize, and measure biological processes at the molecular and cellular levels [67]. There are several molecular imaging tracers that can be used to evaluate fibrosis in CD such as ¹⁸F- ¹⁸FFDG and gallium 68 fibroblast activation protein inhibitor (FAPI) (gallium 68 (68Ga)). The most used PET imaging agent in clinical practice is ¹⁸FFDG. It can be taken up by inflammatory cells and used to precisely pinpoint inflammatory areas in CD [68].

Molecular imaging has been primarily directed at detecting inflammatory pathways in the context of IBD. Repeated inflammation episodes in chronic inflammatory illnesses such as IBD cause the deposition of ECM components, such that 5% of UC patients and 30% of CD patients develop a fibrostenosing phenotype [69,70]. Despite the reduction of inflammation with therapeutic drugs or surgical excision of the stricture, strictures tend to reoccur, requiring to repeated surgery [71]. By identifying markers of intestinal fibrosis would help in the development of new diagnostic and therapeutic methods to detect early stages of fibrosis to limit symptom manifestation and help evaluate the risk of patients acquiring fibrosis. Furthermore, it is limited to utilize conventional imaging to assess the degree of fibrosis because inflammation and fibrosis frequently coexist. The development of novel diagnostic methods that can identify and measure fibrosis in the context of inflammation is urgently needed, even though several imaging modalities, such as MT-MRI, PA imaging, or US elastography, have demonstrated promise in detecting fibrosis.

Hybrid imaging modalities, including PET and SPECT, when combined with MRI or CT, offer the advantage of providing both functional and morphological information. This can be of particular significance in determining the extent of inflammation and fibrosis present within strictures, which in turn can have important therapeutic implications. There are several studies that used molecular imaging in detecting CD fibrosis seen in Table 2. In terms of accuracy in detecting diseased segments (about 85%), they are comparable; however, PET-MR is more accurate than PET-CT in detecting intestinal fibrosis [34]. In one study, PET-MR was assessed retrospectively in 19 CD patients with strictures who underwent surgical resection and pathological confirmation. The authors reported that three of the PET-MR biomarkers namely signal intensity on T2WI, maximum standardized uptake value (SUVmax) and ADC × SUVmax are significantly reduced in the fibrosis group [43]. Catalano et al. demonstrated that PET/MRI has superior diagnostic performance in detecting bowel inflammation when compared to each individual modality alone, with intraoperative findings as a reference standard. PET/MRI exhibited sensitivity, specificity, and diagnostic accuracy values of 88%, 93%, and 91%, respectively, while Pet al.one showed values of 91.5%, 74%, and 84%, MR alone showed values of 80%, 87%, and 83% [32].

Table 2. Overview of the selected studies on the role of molecular imaging in diagnosing fibrosis CD.

Ref.	N of pt. Type of study	Indication	Imaging technique	Gold standard	Conclusion
Pellino et al. [34]	35 Prospective	To compare the accuracy and clinical impact of hybrid [PET_MRE and PET CTE in CD patients	PET/CTE PET/MRE	Histopathology (resection)	The PET-MRE approach was able to effectively discriminate between fibrotic and non-fibrotic strictures, exhibiting a sensitivity of 66.7%, specificity of 88%, and an AUC of 0.77.
Lenze et al. [73]	30 Prospective	To identify strictures in CD patients and to distinguish between those caused by inflammatory processes versus fibrosis	18F-FDG-PET/CT and MR enteroclysis, Trans-abdominal ultrasound	Histology and endoscopy	All three techniques investigated in this study demonstrated high rates of stricture detection when compared to the gold standard. However, none of the techniques were able to accurately differentiate between the nature of the strictures. Nevertheless, the use of a combination of these methods enabled the identification of all strictures that required surgical intervention.
Catalano et al. [43]	19	To distinguish fibrotic from inflammatory structures in CD patients	18F-FDG-PET/MRE	Post-surgical histology	The utilization of 18 F-FDG-PET/MRE provides credible biomarkers for the assessment of strictures. In CD patients, PET/MRE differentiated purely fibrotic strictures from inflammatory or mixed strictures.
Scharitzer et al. [72]	14 prospective single-center study	To assess the potential usefulness 68 (68Ga)- (FAPI) PET/MRE as a predictive tool for evaluating bowel wall fibrosis in CD.	68Ga-FAPI PET/MRE	Histopathology (Surgical resection)	A statistically significant positive correlation was found between higher SUVmax values and the presence of fibrosis in pathological specimens. There was a positive correlation observed between increasing SUVmax values and the severity of fibrosis identified in resected specimens.

Pellino et al. study indicates that PET/MR is a more precise method in detecting extra-luminal disease in patients with CD compared to PET/CT [34]. They conducted a study involving 35 CD patients with symptomatic small-bowel strictures who underwent 18 F-FDG PET/MRI and 18 F-FDG PET/CTE. Of these patients, 29 required surgical intervention. The results of the study showed that both modalities correctly identified 85% of pathologic segments. PET-MRE was found to be more accurate than PET-CTE in detecting fibrosis, with an area under the curve (AUC) of 0.77 compared to 0.51 for PET-CTE (p < .007). In detecting inflammatory segments, however, there was no significant difference between the two procedures, with PET-CTE and PET-MRE correctly identifying 75% and 87.5% of inflammatory segments, respectively (p < .99). PET-MRE also demonstrated a higher accuracy in detecting segments with fibrosis or mixed features compared to PET-CTE, with 67% and 28% detection, respectively (p = .043). A maximum standardized uptake value (SUVmax) cut-off value of 1.46 had an AUC of 0.944, 100% sensitivity, and 83% specificity in identifying segments with fibrosis and absent or mild inflammation (p < .0001). These findings suggest that PET-MRE may be a more reliable and accurate imaging technique for the detection of fibrosis in CD patients with small-bowel strictures [34].

Another study has demonstrated the possibility of using imaging biomarkers obtained from PET/MR to distinguish between fibrotic strictures and inflammatory strictures [43]. Catalano et al. conducted a retrospective study using 18 F-FDG PET-MRE in 19 patients with structuring CD who underwent bowel resection. Histological findings were available for 33 segments, with 7 exhibiting active inflammation without fibrosis, 11 displaying only fibrosis, and 15 having mixed composition. The study found that SUVmax, signal intensity on T2-weighted images × SUVmax, and ADC × SUVmax were significantly lower in fibrosis segments compared to inflammatory and mixed segments (p < .05). A combined PET/MRE biomarker of ADC × SUVmax with a cut-off value of less than 3000 was found to be the best discriminator between fibrosis and active inflammation, with accuracy, sensitivity, and specificity values of 71%, 67%, and 73%, respectively. These results suggest that 18 F-FDG PET-MRE imaging may provide a useful biomarker for the detection and differentiation of fibrosis and active inflammation in CD patients with structuring disease [43].

A newly developed radiotracer, called a gallium 68 fibroblast activation protein inhibitor (FAPI) (gallium 68 (68Ga)), has been designed using quinoline-based inhibitors that specifically target fibroblast activation proteins (FAPs). This radiotracer has been conjugated with 68Ga to enable selective targeting of FAPs. Scharitzer et al. [72]. conducted a prospective single centre study to assess the correlation between imaging features obtained from 68Ga-FAPI PET/MR enterography and the severity of fibrosis in the bowel wall of 14 participants with CD. The study utilized surgical pathological findings as the reference standard for evaluating the accuracy of the imaging technique. The degree of FAPI uptake was measured and expressed in numerical terms through the employment of the maximum standardized uptake value (SUVmax). A statistically significant positive association between the progressive elevation of SUVmax and the presence of fibrosis in pathological specimens was determined with a high degree of confidence, where the obtained P-value of less than 0.001 indicates a highly significant relationship between the two variables. When considering an SUVmax threshold of 3.5, the analysis yielded an area AUC of 0.94, with a sensitivity value of 93% and a specificity value of 83%. The study found that bowel segments with fibrosis exhibit higher SUVmax of the radiotracer compared to segments without fibrosis. Furthermore, the implementation of 68Ga-FAPI MRE imaging technique not only demonstrated its proficiency in precise detection and staging of fibrotic strictures, but also proved successful in distinguishing, albeit in a small subset of participants, between isolated active inflammatory strictures and mixed active inflammatory and fibrotic segments, owing to discernible differences in SUVmax values between the two groups. This distinction was statistically significant (p=.005), highlighting the potential utility of 68Ga-FAPI MRE imaging in differentiating between these distinct pathological processes [72].

The role of artificial intelligence (AI) alongside molecular imaging in diagnosing CD

AI can be employed to extract biometric information and imaging features from medical images that may not be easily discernible by human physicians. In recent times, there has been a growing utilization of AI-based radiomics to aid in the detection of lesions, pathological diagnosis, radiotherapy target delineation, and curative effect prediction for the treatment of cancer patients. This approach has been employed to enhance effective decision-making pertaining to treatment options. It been also applied to detect fibrosis CD [74]. Radiomic models were developed which provided significantly greater accuracy in characterizing enteric fibrosis than human radiologist capability in CD patients [75]. In a commentary published in 2020, Lin and colleagues explored the concept of computer-assisted image analysis within the context of IBD. They proposed the use of radiomics as a method to convert subjective assessments of fibrosis into quantitative data [76]. This discussion was prompted by a study investigating the feasibility of semi-automated analysis for quantifying structural damage in the bowel. The study demonstrated a high level of agreement between the measurements obtained through semi-automated analysis and those conducted by experienced radiologists [77]. Consistent with findings in other diseases, a limited number of studies have indicated that radiomics of MRE and CTE, which involves extracting high-dimensional data from cross-sectional imaging, hold promise as a valuable source of information for evaluating fibrosis in IBD [74,75].

One example of using AI in molecular imaging for CD is the use of convolutional neural networks (CNNs) to analyze PET images with FDG. Mori et al. developed an 18 F-FDG-PET/CT radiomic model to predict the recurrence and survival rates of patients with locally advanced pancreatic cancer (LAPC) following radiotherapy. The implementation of this model has demonstrated significant potential to enhance treatment outcomes by enabling the avoidance of over-treatment in patients with poorer prognostic indicators. This approach represents a valuable contribution to the field of personalized medicine, as it enables a more tailored treatment approach for patients with LAPC based on their individual predicted outcomes [78].

Discussion and conclusion

There is a dearth of literature reviews on the value of molecular imaging in diagnosing CD fibrosis. This study is the first to review the role of molecular imaging in combination of AI in the diagnosis of CD. In the current review, Table 2 summarize most of the recent previous studies demonstrating the use of molecular imaging in detecting CD fibrosis. Almost all studies employed a cross-sectional design, and the primary analyses were significant in distinguishing between fibrotic and non-fibrotic diseases. Several of the molecular targets mentioned earlier are not exclusive to fibrosis. Most of the enrolled studies used 18 F-FDG-PET.

In conclusion, molecular imaging techniques such as PET-MRI may have great promise in the detection and monitoring of fibrosis in Crohn’s disease, however, further studies are needed to explore they clinical utility. These studies should concentrate on the development of different techniques of cross-sectional molecular imaging, such as 68Ga-FAPI PET/CT or hyperpolarized 13 C spectroscopic MRI, to examine the pathophysiology of CD by labelling therapeutic agents or targeting cellular receptors. While there are limitations to their use, the emerging field of AI represents a promising avenue for improving the accuracy and efficiency of molecular imaging in the diagnosis and management of fibrosis. Further research is needed to fully realize the potential of these technologies in improving patient outcomes and quality of life.

Ethical approval

This review did not require ethical approval as it was based on previously published studies.

Authors contributions

Conceptualization: Ali S. Alyami, Wael A. Ageeli, Yahia Madkhali

Data curation: Ali S. Alyami, Yahia Madkhali

Investigation: Ali S. Alyami, Naif A. Majrashi

Writing – original draft: Ali S. Alyami, Naif A Majrashi, Yahia Madkhali,

Writing – review & editing: Naif A Majrashi, Wael A. Ageeli, Yahia Madkhali, Turkey Refaee, Bandar Alwadani, Meaad Elbashir, Sarra Ali, Heham El-Bahkiry and Abdullah Althobity.

Acknowledgments

The authors extend their appreciation to the Deputyship for Research& Innovation, Ministry of Education in Saudi Arabia for funding this research work through the project number ISP23-100.The authors would also like to thank Jazan university, Saudi Arabia, for using the online facilities to complete this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, [Dr Alyami], upon reasonable request.

Table 1. Shows the current applications for nuclear medicine imaging in different diseases.

Modality	Description	Radiotracers	Indications/Examples	Advantages
Gamma camera	Planar whole body and spot imaging, flow/cine.	Gamma-emitting radionuclides (^99mTc, ^81Tl, ^111In, ^131I, ^123I, etc.).	Main traditional nuclear medicine modality with broad spectrum of indications.	Functional imaging as opposite to traditional anatomic data provided by CT and radiography.
SPECT/CT	Tomographic reconstruction algorithm produces 3-D data. Multiplanar reformats can be created.	Gamma-emitting radionuclides can be simply added to planar imaging as needed.	Myocardial perfusion imaging, VQ scan, parathyroid localization	Improved accuracy and lesion localization compared with planar images.
PET/CT	Currently almost always in conjunction with CT for attenuation correction and lesion localization	^18FDG—nonspecific Multiple newer tracers	Oncologic imaging. Infection/inflammation (FUO, sarcoidosis including cardiac). Detection of active sites of CD within the small bowel and colon	Metabolism evaluation (FDG). Targeted molecular imaging (newer tracers). Higher resolution than SPECT.
MRI (PET/MR)	Magnetic Resonance Imaging used in conjunction with PET	PET tracers Gadolinium-based contrast. Ultrasmall superparamagnetic iron oxide particles (USPIOs).	Oncologic imaging, most commonly pelvic malignancies (cervical cancer, prostate cancer), hepatocellular carcinoma.	Superb soft tissue characterization, complimentary anatomic and functional information.
PET-MRE (and PET-CTE)		Assessment of the maximum standard uptake value of the tracer (SUVs)		Successful differentiation of fibrotic from nonfibrotic stricture
Optical imaging	Fluorescence	Fluorophores and quantum dots.	Preclinical	Limited depth (< 1–2 cm
	Bioluminescence	Luciferase enzyme.

Additional information

Funding

The authors extend their appreciation to the Deputyship for Research& Innovation, Ministry of Education in Saudi Arabia for funding this research work through the project number ISP23-100.