Abstract
Introduction: Inflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are associated to poor maternal and foetal outcomes during pregnancy, requiring a strict monitoring of the disease activity, preferably with non-invasive modalities. There are numerous data confirming intestinal ultrasonography (IUS) accuracy and efficacy for the detection of IBD disease activity and complications, but data in pregnant IBD patients are scarce. We aimed to evaluate the diagnostic performance and feasibility of IUS to monitor IBD activity throughout pregnancy. Methods: A systematic literature review was performed to identify studies on the use of ultrasound modality in pregnant inflammatory bowel disease women from the date of inception until April 2024 using MEDLINE, Cochrane Library, EMBASE, and ISI Web of Science databases, with keywords including (1) ultrasound/ultrasonography, (2) pregnancy, and (3) IBD (CD and UC). Additional relevant studies were identified from cross-referencing and hand-searches of references of the retrieved articles. We included fully published observational studies and abstracts. Results: Overall, five studies have been selected from 264 citations. All studies were highly heterogeneous in the definition of disease activity as reference standard, IUS protocols, and outcomes. Two of them used a cut-off value of faecal calprotectin (FCP) >100 μg/g. In one of them, clinical scores were used when the FCP value was ≥100–249 μg/g and FCP ≥250 μg/g was considered as an active disease independently of clinical scores. Only one study used a single reference standard with a Harvey-Bradshaw Index (HBI) >4. Across these 3 studies, results suggest a relatively good specificity (range 83–98%) but low sensitivity (range 33–84%) to detect disease’s activity. Only 1 study analysed IUS performance in detecting remission with a sensitivity of 80% and a specificity of 92% compared to the reference standard. The size of the uterus limits the visualization of the terminal ileum and the sigmoid from the second trimester and the third trimester, respectively. The evaluation of the rectum remains limited. Conclusion: IUS appears to offer a practical, readily available and non-invasive means of monitoring disease activity in IBD pregnant women. However, existing literature on IUS accuracy is limited and further studies with standardized reference comparator are needed.
What is already known?
- -
Many data support the accuracy and efficacy of intestinal ultrasonography (IUS) for detecting disease activity and complications in patients with inflammatory bowel disease (IBD). However, data regarding the use of IUS in pregnant patients with IBD are scarce, and its possible technical limitations due to the gravid uterus must be carefully evaluated.
- -
What is new here?
- -
Our study provides new insights into the evaluation of pregnant patients with IBD by synthesizing the available evidence regarding the diagnostic accuracy of IUS throughout all phases of pregnancy.
- -
How can this study help patient care?
- -
IUS appears to offer a practical, readily available and non-invasive means to monitor disease activity in IBD pregnant women which permits to avoid more invasive and poorly tolerated tests such as colonoscopy.
- -
Introduction
Inflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are characterized by chronic, relapsing, and destructive inflammation of the digestive tract that can lead to permanent damage of organ, impairing the patient’s quality of life even at young age [1]. IBD frequently affect women during their childbearing age [2] and adverse maternal and foetal outcomes during pregnancy are correlated with disease’s activity [3]. Several studies have shown an increasing risk of prematurity and low birthweight for the baby as well as an augmentation of caesarean section [2‒5]. Hence, the follow-up and the management of the disease become a priority to maintain a biological and clinical remission [6]. Treat-to-target strategies, allowing to achieve these goals, involve a tight control of the inflammation to confirm remission or guide a treatment optimization [7, 8]. Unfortunately, clinical scores use in non-pregnant patient to monitor the disease activity become unreliable given their similarities with pregnancy-related symptoms [9]. Moreover, biochemical parameters such as C-reactive protein, albumin, and haemoglobin evolve according to the stage of pregnancy [9, 10]. The use of faecal calprotectin (FCP) is more accurate to monitor the activity during pregnancy but hasn’t been consistently reported for pregnant women and can’t provide information on the extension and the complication of the disease [9, 10].
For non-pregnant IBD patient, more invasive modalities are used to monitor the disease such as colonoscopy and MRI imaging, which both present potential drawbacks: colonoscopy is a poorly tolerated and invasive procedure with a risk of bowel perforation and gives no information on small bowel disease activity in CD [11, 12]. Even if MRI became the gold standard to assess this part of the bowel, it remains an invasive, expensive, and access-limited technique [11]. The invasiveness and poor tolerability make these both techniques unfeasible for the frequent and repetitive monitoring of disease activity and have potential side effects for the foetus [3].
For the past decades, there has been a growing interest on intestinal ultrasonography (IUS) with several studies that have proven its feasibility and efficacy in non-pregnant IBD populations [13‒15]. IUS may provide a safe and non-invasive evaluation of the small and large bowel without risk for the mother and foetus during pregnancy [3, 16‒20]. However, data on IUS in pregnant patients are scarce and its possible technical limitation due to the gravid uterus must be evaluated. The aim of this systematic review is to assess diagnostic performance and feasibility of IUS to objectify IBD activity throughout pregnancy.
Methods
Search Strategy
Systematic search has been performed of the following databases from date of inception to April 2024 (MEDLINE, Cochrane Library, EMBASE, ISI Web of Science) for studies assessing ultrasound in pregnant patient with IBD. Medical subject heading and keywords will include (1) ultrasound/ultrasonography, (2) pregnancy, and (3) IBD or CD or UC (online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000541017).
Abstracts presented at major gastroenterology conferences (ACG, CDDW, DDW, UEGW, ECCO, JFHOD) in the past 5 years were also hand-searched. Additional relevant studies were identified from cross-referencing and hand-searches of references of the retrieved articles.
Study Selection and Patient Population
We included all human studies published in French and English that compared ultrasound modality to any comparator in the follow-up of pregnant patients with IBD (including CD or UC). All stages of pregnancy have been included. Publication types such as letters, notes, case reports, or comments were excluded.
Validity Assessment and Data Abstraction
Citations were handled in Endnotes X9 (Clarivate Analytics, Philadelphia, USA). Following removal of duplicates, two reviewers (J.P. and J.V.P.) independently evaluated the eligibility of citations based on the title and abstract. Selected full-text articles were reviewed for final inclusion/exclusion assessment. A third reviewer (O.K.) resolved conflicts. Risk of bias was assessed using the Quadas-2 for diagnostic testing (Fig. 1).
Choice of Outcomes
The primary outcome was diagnostic performance of IUS in comparison to any comparators to detect disease’s activity throughout pregnancy. Secondary outcomes were feasibility of IUS according to the pregnancy stage (separated in three trimesters). Other outcomes were reported such as the resultant changes to clinical management after the IUS and the safety indicators (potential adverse events including miscarriage, bleeding, and pain).
Data Items
Extracted variables from the articles included publication year, study type, country, and the number of institutions involved. Additionally, factors such as study enrolment duration, inclusion/exclusion criteria, and definition of active disease were considered. Material-related variables encompassed types of ultrasounds used, qualifications of sonographers, and diagnostic tests utilized. Factors such as illness duration, average age, IBD subtype, pregnancy trimester, BMI, comorbidities, past examinations, treatment regimens, and their duration were analysed. Outcome information, such as the total number of patients included, the availability of analysable data, and the specificity and sensitivity of each study, was assessed. Additionally, secondary and tertiary outcome measures were considered.
Results
Study Selection and Interventions
The structured literature search initially identified 328 citations. After review, a total of 54 studies were excluded. The PRISMA diagram is shown in Figure 2. We finally selected a total of 5 studies; all studies were fully published [16‒20]. All studies were observational and only 3 studies had a comparative test to detect disease activity (Table 1).
Study (first author), year of publication . | Study country; type of study; number of institutions . | Patient number; type of disease . | Criteria of inclusion; trimester (median weeks pregnant) . | Comparator (definition of active disease) . | US anormal definition . | Sonographer qualification . |
---|---|---|---|---|---|---|
De Voogd et al. [17] (2021) | Netherlands; Prospective cohort; 1 | 38; 22CD, 16UC | ≥18 years with diagnosis of IBD; 11 | FCP ≥250 μg/g (and absence of infection), FCP ≥100–249 μg/g + HBI ≥4 or/and SCCAI ≥3 | BWT (>2 mm ileum, >3 mm colon, >5 mm rectum) colour Doppler intensity (long stretches within the wall or extending into the mesentery), loss of wall layer stratification, loss of haustration, presence of fatty wrapping | International curriculum for IUS |
Flanagan et al. [20] (2020) | Australia; Prospective cohort; 3 | 90; 57CD, 29UC, 4 unclassified | Confirmed diagnosis of ileal and/or colonic IBD with at least one GIUS during pregnancy; unknown | FCP>100 μg/g or endoscopic data or cross-sectional imaging data | BWT >3 mm, Limberg score on colour Doppler | Gastroenterologist with adequate training and expertise in the procedure |
Leung et al. [19] (2019) | Canada; Prospective cohort; 1 | 82; 82CD | ≥18 years with diagnosis of CD defined by endoscopic/radiologic/histologic parameters; 20.1 | HBI >4 | BWT, inflammatory fat, blood flow on colour Doppler, complications (fistula, stricture, perforation, mass) | Sonographers with a high level of acquired skill and scans interpreted by radiologists |
Study (first author), year of publication . | Study country; type of study; number of institutions . | Patient number; type of disease . | Criteria of inclusion; trimester (median weeks pregnant) . | Comparator (definition of active disease) . | US anormal definition . | Sonographer qualification . |
---|---|---|---|---|---|---|
De Voogd et al. [17] (2021) | Netherlands; Prospective cohort; 1 | 38; 22CD, 16UC | ≥18 years with diagnosis of IBD; 11 | FCP ≥250 μg/g (and absence of infection), FCP ≥100–249 μg/g + HBI ≥4 or/and SCCAI ≥3 | BWT (>2 mm ileum, >3 mm colon, >5 mm rectum) colour Doppler intensity (long stretches within the wall or extending into the mesentery), loss of wall layer stratification, loss of haustration, presence of fatty wrapping | International curriculum for IUS |
Flanagan et al. [20] (2020) | Australia; Prospective cohort; 3 | 90; 57CD, 29UC, 4 unclassified | Confirmed diagnosis of ileal and/or colonic IBD with at least one GIUS during pregnancy; unknown | FCP>100 μg/g or endoscopic data or cross-sectional imaging data | BWT >3 mm, Limberg score on colour Doppler | Gastroenterologist with adequate training and expertise in the procedure |
Leung et al. [19] (2019) | Canada; Prospective cohort; 1 | 82; 82CD | ≥18 years with diagnosis of CD defined by endoscopic/radiologic/histologic parameters; 20.1 | HBI >4 | BWT, inflammatory fat, blood flow on colour Doppler, complications (fistula, stricture, perforation, mass) | Sonographers with a high level of acquired skill and scans interpreted by radiologists |
CD, Crohn’s disease; FCP, faecal calprotectin; HBI, Harvey-Bradshaw index; IBD, inflammatory bowel disease; IUS, intestinal ultrasound; PGA, Physician Global Assessment; SCCAI, Simple Clinical Colitis Activity Index; UC, ulcerative colitis; US, ultrasound.
Study Quality Assessment
Study quality assessment is summarized in Figure 1 using the Quadas-2 scoring system.
Definition of Comparator (Disease Activity)
Definitions of comparator are summarized in Table 1. All studies selected had a different definition of disease activity as reference standard. Two of them used a cut-off value for FCP >100 μg/g. De Voogd et al. [17] used additional clinical scores when the FCP value was ≥100–249 μg/g, but FCP ≥250 μg/g was considered as an active disease independently of clinical scores. Leung et al. [19] used only one reference standard with the HBI clinical score. The reference standard wasn’t always the same in the Flanagan et al. [20] study and depended on the data available for each patient.
Patient and Study Characteristics
First study published in 2017 by Leung et al. [19] is a prospective monocentric observational study in Canada, including 91 pregnancies in 82 pregnant adult women with CD (median 20.1 weeks of pregnancy, median body mass index [BMI] at the time of pregnancy 24.3 kg/m2) [19]. Most of the IUSs were performed before the end of the second trimester and only 16.5% in the third trimester. CD was diagnosed according to endoscopic, radiologic, and histologic parameters. All patients underwent IUS between July 2012 and December 2016. IUSs were performed by sonographers with a high level of acquired skill and CT scans were interpreted by radiologists. Active disease was defined by clinical indices like the Harvey-Bradshaw Index (HBI) >4 as the reference standard. Sonographic findings were graded as inactive or active disease according to the following criteria: bowel wall thickness (BWT), inflammatory fat, blood flow on colour Doppler, and complications (fistula, stricture, perforation, mass). The main objective was to report the ability of IUS to detect subclinical inflammation in pregnant women with asymptomatic CD.
In 2020, Flanagan et al. [20] published a multicentred observational study in Australia including 90 pregnant (median weeks of pregnancy unknown) IBD women (57 CD, 29 UC, and 4 unclassified). Most (55.9%) of IUSs were performed in trimester 2 (median gestation age 19 weeks). Active disease was assessed by the PGA, the HBI, or SCCAI score and defined by an FCP greater than 100 μg/g as the reference standard in the absence of endoscopic or cross-sectional imaging data. IUS was performed by gastroenterologist with adequate training and expertise in the procedure of IUS. Disease activity on IUS was defined by BWT >3 mm and the Limberg score on colour Doppler. The aim of this study was to assess the accuracy and feasibility of IUS in monitoring IBD during pregnancy.
De Voogd et al. [17] analysed, in a monocentric prospective cohort study published in 2021 in Netherlands, 38 adult pregnant women (11 median weeks of pregnancy) with 76 IUS examinations (22 CD and 16 UC). Based initially on the FCP value, the authors formed a composite standard reference. The clinical active disease was defined by FCP >250 μg/g and absence of infection as standard reference. When FCP was between 100 and 249 μg/g, a clinical disease activity score of HBI in CD ≥4 or SCAAI in UC ≥3 was used to consider an active disease. Three trained ultrasonographers performed the IUS and identified active disease based on BWT in ileum >2 mm), colon >3 mm, rectum >5 mm, colour Doppler intensity, loss of wall layer stratification, and haustration and presence of fatty wrapping. The aim of this study was to assess the feasibility and accuracy of IUS in the follow-up of IBD patients throughout pregnancy.
The studies conducted by Masseli et al. [16] and Jeanty et al. [18] did not incorporate a comparative test to identify disease activity. Without this crucial information, it's difficult to draw meaningful conclusions regarding the diagnostic accuracy of the tested methods. As a result, these studies will not be further elaborated on.
Primary Outcome
Leung et al. [19] revealed a sensitivity and specificity of IUS in detecting activity regarding the reference standard of 33% (95% CI: 10–65%) and 91% (95% CI: 83–96%), respectively (Table 2). Discordance between clinical HBI score and IUS findings occurred in 16.5%. Only 13% of women were clinically active with an HBI score >4. Most of IUS assessments (87.9%) validated quiescent disease. Although 8.9% patients were considered to be in clinical remission, they had active disease on IUS. In clinically symptomatic patients with inactive disease on IUS, women had a significantly lower median HBI score than women with symptoms and active disease at IUS (6.5 vs. 10.5, p = 0.05).
Study . | Primary outcome, detect activity . | Secondary outcome: feasibility . | Other outcomes . | Limitation(s) . | |||
---|---|---|---|---|---|---|---|
sensitivity . | specificity . | number of GIUS to see all bowel segments assessed . | median gestational age . | Visualization of TI per trimester . | |||
Leung et al. [19] (2019) | 33% (95% CI: 0.10–0.65) | SP = 91% (95% CI: 0.83–0.96) | 92.3% | Unknown | Unknown |
|
|
Flanagan et al. [20] (2020) | 74% (95% CI: 0.49–0.91) | 83% (95% CI: 0.70–0.92) | 69% | 16.5 weeks (IQR 10–21) | 1st = 95%, early 2nd = 89%, late 2nd = 59%, 3rd = 0% |
|
|
De Voogd et al. [17] (2022) | 84% (95% CI: 0.64–0.95) | 98% (95% CI: 0.88–1.00) | Unknown | Unknown | 1st = 91.3%, 2nd = 43.3%, 3rd = 21.7% |
|
|
Study . | Primary outcome, detect activity . | Secondary outcome: feasibility . | Other outcomes . | Limitation(s) . | |||
---|---|---|---|---|---|---|---|
sensitivity . | specificity . | number of GIUS to see all bowel segments assessed . | median gestational age . | Visualization of TI per trimester . | |||
Leung et al. [19] (2019) | 33% (95% CI: 0.10–0.65) | SP = 91% (95% CI: 0.83–0.96) | 92.3% | Unknown | Unknown |
|
|
Flanagan et al. [20] (2020) | 74% (95% CI: 0.49–0.91) | 83% (95% CI: 0.70–0.92) | 69% | 16.5 weeks (IQR 10–21) | 1st = 95%, early 2nd = 89%, late 2nd = 59%, 3rd = 0% |
|
|
De Voogd et al. [17] (2022) | 84% (95% CI: 0.64–0.95) | 98% (95% CI: 0.88–1.00) | Unknown | Unknown | 1st = 91.3%, 2nd = 43.3%, 3rd = 21.7% |
|
|
GIUS, gastrointestinal ultrasound; TI, terminal ileum; 1st, first trimester; 2nd, second trimester; 3rd, third trimester.
Flanagan et al. [20] showed a sensitivity and specificity of IUS to detect disease activity in comparison to FCP >100 µg/g as a reference standard of 74% (95% CI: 49–91%) and 83% (95% CI: 70–92%), respectively (Table 2). When FCP cut-off was changed for 150 μg/g, results were similar, with a sensitivity of 76% (95% CI: 50–93%) and a specificity of 81% (95% CI: 69–91%). This study reveals a significant correlation between FCP and maximal BWT in IUS (p = 0.03). In normal IUS, the median FCP was lower compared to the median FCP in patients with active IUS findings, respectively, of 20.9 µg/g and 186 µg/g. Authors also found a significant correlation between BWT and an increasing PGA score (p < 0.001) but no significant difference in the SCCAI score according to the activity in IUS (p = 0.72). All patients with hypervascularity on IUS (Limberg score ≥2) and an increase in BWT had a high FCP (median 217 µg/g).
De Voogd et al. [17] showed a sensitivity of 84% and a specificity of 98% of IUS in comparison to reference standards (Table 2). The evaluation per trimester showed a decrease in sensitivity in the third trimester (first trimester 83.3%, second trimester 92.3%, third trimester 66.6%), while the specificity remained high throughout pregnancy (first trimester 93.8%, second trimester 100%, third trimester 100%). The authors reported a moderate correlation between clinical scores and IUS activity (p = 0.60, p < 0.01) and a strong correlation with FCP (p = 0.73, p < 0.00).
Secondary Outcomes
In the Leung et al. [19] study, 92.3% of the IUS evaluations were able to assess all bowel segments (no median gestational age described). There was no correlation between the IUS quality and the term of the pregnancy or the location of the disease (ileocolonic, colonic, and ileal location).
Flanagan et al. [20] showed that IUSs were able to see the five bowel segments, with adequate views in 69% of cases, at a median gestational age of 16.5 weeks (IQR 10–21). 31% of IUS had at least one incorrectly visualized segment with a median gestational age of 22 weeks (IQR 20–26). All patients in the first trimester who didn’t have an acceptable view had a BMI over 28 kg/m2 and over 31 kg/m2 in the early second trimester. Regarding the views per segment across pregnancy, the terminal ileum (TI) was the segment most impacted by the progression of the pregnancy (visualization by trimester: 95% in the first, 89% in the early second, 59% in the late second, and 0% in the third). The visualization of the rectum hadn’t been analysed in this study.
De Voogd et al. [17] revealed a decrease in the visualization of sigmoid and TI was reported according to the growing size of uterus but not for the other segments. During the first trimester, 13% of patients with inadequate views on ultrasound had a BMI ≥29.5 kg/m2. The feasibility of IUS decreased during the second trimester for the TI (first trimester 91.3%, second trimester 43.3%, third trimester 21.7%) and during the third trimester for the sigmoid (first trimester 95.6%, second trimester 76.7%, third trimester 69.5%, p = 0.02). The rectum has been seen in less than 50% of the cases.
Other Outcomes
Changes to Clinical Management after the IUS
In the Leung et al. [19] study, the IUS findings of inactive disease among clinically symptomatic patients (8.8%) permitted to reassure them without further investigations. Conversely, clinical management was modified in asymptomatic women with active disease on IUS (1 started a biological therapy, 1 had antibiotics, 1 had a prolongation of treatment, and 1 started corticosteroids treatment).
In the Flanagan et al. [20] study, 61% of active disease on IUS resulted in a change of the patient’s therapy compared to 2% on IUS with quiescent disease. The study reported a non-significant difference between women with active disease on FCP or IUS regarding the gestational age at delivery and birthweight.
In the De Voogd et al. [17] study, treatment has been changed according to disease activity on IUS for 17.1% of patients but 11.8%, who had disease activity on IUS, continued their current treatment because they hadn’t an elevation of FCP and clinical score, two of them had fibrosis on IUS. Conversely, 4 cases with no activity on IUS (4/54 total cases, 2 CD and 2 UC) had an active disease according to the reference standard (FCP and clinical score were increasing) and the 2 UC had a change of their treatment for proctitis.
Safety Data
No safety data were reported in these studies.
Discussion
There are numerous data confirming IUS accuracy and efficacy for the detection of IBD disease activity and complications, and its utility is now well accepted in non-pregnant IBD patients particularly in CD [11, 13‒15, 21‒24]. Even if few studies have shown the value of IUS in assessing UC activity [22, 23], the last consensus of the European Crohn’s and Colitis Organization (ECCO) recommended the use of IUS for diagnosis and suggested it for monitoring IBD [11, 21].
Subsequently, there has been growing enthusiasm about the use of IUS in pregnancy given the need for safe, non-invasive, and regular monitoring of disease activity during this time to improve outcomes for both the mother and the foetus. The aim of this systematic review was to summarize existing evidence on IUS diagnostic performance, in pregnant IBD patients throughout all stages of pregnancy.
Surprisingly, despite accumulating data in non-pregnant IBD patients, only five observational studies including 162 CD and 46 UC specifically addressed IUS diagnostic performance during pregnancy, precluding any firm conclusions. Furthermore, only 3 studies used a comparative test to assess disease activity and all studies were highly heterogeneous in the definition of disease activity as reference standard, IUS protocols, and outcomes.
Overall, main results suggest a relatively good specificity (range 83–98%) but low sensitivity (range 33–84%) to detect disease’s activity. De Voogd et al. [17] indicate that the specificity remains higher than sensitivity throughout the pregnancy. The results on specificity mirror the results of a meta-analysis and the METRIC study for the diagnosis of non-pregnant CD, which reported a sensitivity of 84% and a specificity of 92% [15, 24]. The low sensitivity observed in the study by Leung et al. [19] (33%) may be attributed to the utilization of HBI score as a reference for inflammation. Pregnant women may experience gastrointestinal symptoms that are unrelated to intestinal inflammation but rather attributable to pregnancy itself. As a result, individuals classified as active by the HBI may not have been truly active, contributing to the low sensitivity of IUS in this study.
Despite a lower sensitivity, studies mention that IUS could be used to ensure quiescent disease and to detect subclinical inflammation [19, 20]. This would allow to confirm quiescent disease in patients who might have symptoms mimicking flare during pregnancy without active disease, avoiding an increase of their treatment or invasive investigations and providing reassurance particularly because irritable bowel disease is very frequent in this population [25]. The limited sensitivity of IUS is consistent with other studies in non-IBD population, due to its technical limitation in evaluating upper gastrointestinal and rectal disease [26, 27]. Still, the sensitivity in pregnant IBD patients tends to increase from 2017 to 2021 between Leung et al. [19] (33%), Flanagan et al. [20] (74%), and De Voogd et al. [17] (84%). This difference can be explained by patient-related factors and/or factors associated to the examiners themselves. Flanagan et al. [20] excluded IUS with inadequate ileal views in patient with ileal disease and inadequate colonic views in colonic disease that may have impacted the accuracy of IUS. The cut-off value used in all study for BWT (3 mm is the normal upper limit) and FCP (>100 µg/g) (Table 1) to define disease activity is likely to underestimate the specificity of IUS especially in fibrotic diseases and in small bowel disease due to strict lower cut-offs [20, 28]. Regarding the importance of a rigorous disease control, especially in pregnant women, BWT required to be interpreted with other parameters such as colour Doppler or ultrasound elasticity imaging that could differentiate chronic fibrosis and acute inflammation [29].
The Doppler velocity varies across studies, and the observed discrepancies are likely attributable to fundamental variations in signal processing depending on the machine used [30]. Furthermore, it is noteworthy that the velocity index used is frequently omitted, further complicating comparisons between studies. Therefore, the appropriate velocity needs evaluation, considering adjustments made to the energy gain to ensure full coverage of the sampling frame by colour, followed by gradual reduction until only the intravascular signal becomes visible [31].
One of the advantages of IUS is the ability to determine disease extent, severity, and complications of the disease such as fistulae, abscesses, and strictures while clinical and biochemical parameters merely inform on disease presence. Leung et al. [19] showed that clinical activity is not well correlated to IUS as two-thirds of patients with HBI >4 had no significant corresponding inflammation on IUS. Conversely, Flanagan et al. [20] and Voogd et al. [17] compared IUS essentially to FCP with good correlation. However, FCP doesn’t give any information on the extension of the disease. As the extension guides the treatment’s decision, IUS has all its importance. To improve ultrasonographic disease activity, several ultrasonographic activity scores have been proposed, such as the Limberg index or, for CD, the IBUS-SAS (International Bowel Ultrasound Segmental Activity Score) based on four elements with a good inter-observer correlation: wall thickness, wall stratification, Doppler signal, and inflammatory mesenteric fat anomalies [32‒34]. For UC, Milan ultrasound criteria, based on wall thickness and Doppler analysis of the colonic wall, can detect inflammation with a sensitivity of 71% and a specificity of 100% [23]. However, these scores still require external validation and simplification before they can be implemented in practice.
Concerns about the feasibility of IUS in pregnancy have been raised given the possibility of bowel views becoming obscured by the gravid uterus. There are no standardized protocols for abdominal ultrasound in pregnant women. However, it is possible to adapt known techniques for non-pregnant patients to perform the examination. In non-pregnant patient, ultrasound is performed in the supine position using a high-frequency (4–8 MHz) linear (or microconvex) probe, use for detailed bowel evaluation, or a low-frequency (1–5 MHz) convex probe for deep evaluation, for example, in case of overweight [13, 35, 36]. In studies included in our review, the common probe used is a linear (5–12 MHz) probe. De Voogd et al. [17]describe the different probe using for each part of the intestine: convex (5–1 MHz) for the rectum and linear (5–12 MHz) for the sigmoid. Overall, our results showed that IUS is feasible throughout pregnancy. However, visualization of the rectum is limited. In fact, ultrasound has a poor sensitivity to visualize the rectum, limiting its use to diagnose proctitis [37]. However, promising results have been obtained by Alloca et al. [23] comparing colonoscopy with ultrasound with sensitivity and specificity of 89% and 87% according to a BWT >3 mm in non-pregnant UC patients. Other modalities, such as perineal ultrasound, should be used in addition to IUS to detect proctitis [11, 38]. De Voogd et al. [17] and Flanagan et al. [20] concluded that the visualization of TI seems to be significatively impacted by the growing size of the uterus in the second trimester and in the third trimester for the sigmoid. However, the other segments seem to be less impacted. These two studies reported also some limitations concerning the feasibility of IUS by identifying the TI according to BMI (Table 2). Those elements limit the use of IUS in rectum, sigmoid, and TI in the second and mostly in the third trimester. However, the limited number of IUS performed during the third trimester in these studies precludes firm conclusion.
IUS may lead to changes in the management of pregnant women with IBD by detecting subclinical inflammation, resulting in the modification of therapy. In general, high specificity makes it likely that active disease on IUS truly reflects active IBD which, in a “point-of-care” setting, may have a key role in ongoing non-invasive monitoring of disease activity in patient with symptoms but middle-range inflammation parameters. De Voogd et al. [17] show a sensitivity of 80% and a specificity of 92% of IUS in detecting improvement to remission compared to the reference standard [20]. Moreover, they affirm that IUS can detect a modification in the activity (remission or relapse) with 78% accuracy and, most of the time, treatment has been changed according to disease activity on IUS. IUS could be used to indicate an early response to treatment and allow a tight control of the inflammation and an improvement of the “treat-to-target” strategies.
Limitation
The limited number of published studies represents a critical issue precluding meta-analysis and firm conclusions from this systematic review. The lack of research in IUS in IBD during pregnancy is surprising, as IUS plays a key role in the management of IBD non-pregnant patients. It can be explained by the following arguments: first, most practicians don’t have appropriate training on IUS that is not yet integrated in their curriculum. Access and expertise in IUS are still not universally accessible. Second, the common perception of the gravid uterus blocking the visualization of the intestine is still wildly present.
Moreover, no study compared endoscopy to IUS in pregnant women with IBD, which is understandable from an ethical standpoint. Besides, new scoring systems, such as the IBUS-SAS score, are currently in use in practice but were not incorporated into the studies included in our review. Future research should consider integrating these new scoring assessments to enhance the comprehensiveness and relevance of their findings. Additionally, technical norms of IUS, such as Doppler velocity, are not consistently reported in the studies of our review, and there is a lack of systematic description regarding the use of ultrasound examinations. A comprehensive description of standardized norms in IUS should be defined to better explore the performance of IUS in future research.
Another major limitation of this review is the lack of agreement between definitions of active disease in pregnant women across studies. The literature itself is unclear on this issue, with the definition of active disease changing regularly even in the overall IBD population. Further studies using standardized reference should be encouraged for pregnant women with IBD. Finally, standardized protocols for the ultrasound acquisition of GI during pregnancy, as it has been created for non-pregnant patient, should be developed and validated [39].
Conclusion
Notwithstanding the limited published data, IUS seems a feasible, accessible, and non-invasive method to monitor disease activity in IBD pregnant women. This is a promising method that allows the evaluation of disease activity in real time and could find its place in the close follow-up of patients. Interpretation of the data should be done with caution though, as existing literature on IUS accuracy is limited. Further studies with standardized reference comparators and protocols for IUS acquisition in pregnant patients with IBD are necessary to strengthen the evidence base.
Statement of Ethics
An ethics statement is not applicable because this study is based exclusively on the published literature.
Conflict of Interest Statement
The authors have no conflicts of interest for this article.
Funding Sources
No funding was received for this study.
Author Contributions
Study concept and design, drafting of the manuscript, and critical revision of the manuscript for important intellectual content: J. Pillet, J. Voirol-Perrin, M. Martel, O. Kherad, and S. Restellini. Acquisition of data: M. Martel, J. Pillet, and J. Voirol-Perrin. Analysis and interpretation of data: M. Martel and O. Kherad. Statistical analysis and technical or material support: M. Martel. Study supervision: O. Kherad and S. Restellini.
Additional Information
Julie Pillet and Julia Voirol-Perrin contributed equally to this work.
Data Availability Statement
All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.