Magnetic resonance enterography for predicting poor prognosis in Crohn’s disease: a systematic review with meta-analysis

Abstract

Background:

Magnetic resonance enterography (MRE) is a noninvasive and radiation-free option for managing Crohn’s disease (CD), for which various predictors have been reported to show significant prognostic value.

Objectives:

The aim of this review was to summarize MRE predictors that indicated poor prognosis, including surgeries, complications, relapses, therapeutic escalations, and persistent pathological changes in patients with CD.

Design:

Systematic review with meta-analysis.

Data sources and methods:

Web of Science and PubMed were systematically searched for articles exploring the predictive performance of MRE in patients with luminal CD, with the latest search performed on July 21, 2025. Pooled analyses of odds ratios (OR) and hazard ratios for predicting poor prognosis were conducted using a random-effects model when available.

Results:

Thirty articles were included in this study. Meta-analysis identified five significant MRE predictors for predicting CD poor prognosis over 12 months, including inflammation (OR 3.40, 95% confidence interval (CI 1.96–5.89)), bowel wall thickening (OR 2.88, 95% CI (1.95–4.25)), strictures (OR 4.95, 95% CI (1.99–12.34)), edema (OR 2.67, 95% CI (1.25–5.69)), and fistula (OR 5.15, 95% CI (2.72–9.76)). Furthermore, scoring systems, such as the Magnetic Resonance Index of Activity (MaRIA) and Clermont scores, showed significant predictive values.

Conclusion:

MRE is effective in predicting the prognosis of patients with CD, with significant predictors including inflammation, bowel wall thickening, strictures, edema, fistula, and MaRIA score.

Trial registration:

The study protocol was registered in the International Prospective Register of Systematic Reviews (registration number: CRD42022365965).

Plain language summary

MRE in CD prognosis

This systematic review demonstrated that MRE is effective in predicting the prognosis of CD, with significant predictors including inflammation, bowel wall thickening, strictures, edema, fistula, MaRIA, and Clermont scores.

Keywords

Crohn’s disease magnetic resonance enterography predictor prognosis

Introduction

Crohn’s disease (CD) is a subtype of chronic inflammatory bowel disease thought to arise from a complex interaction among genetic susceptibility, environmental factors, gut microbiota, and immune responses.¹ CD is characterized by abdominal pain, diarrhea, and weight loss, with skipping lesions under endoscopy.^1,2 CD usually follows a prolonged course, requiring lifelong management, which significantly influences patients’ quality of life and imposes a heavy burden on families, society, and the healthcare system.³ Notably, significant heterogeneity was observed in the prognosis of patients with CD, with over 30% developing stricturing or penetrating complications and 28% requiring surgery within 5 years.^4–6 With the advancement of personalized treatment and precision medicine, identifying predictors for disease prognosis has become increasingly critical for both doctors and patients.⁷

Predicting prognosis plays a crucial role in CD management, particularly in risk stratification, therapy selection, and therapeutic efficacy monitoring. Currently, various methods are available for predicting disease prognosis, including clinical symptoms, biomarkers, endoscopic performances, radiological examinations, and so on.¹ Among them, magnetic resonance enterography (MRE) is a radiation-free and noninvasive method that is gaining popularity, especially for those with small intestine involvement or with intestinal strictures where colonoscopies cannot proceed.⁸ In this way, MRE is well received by patients, with enhanced compliance and tolerance.^9,10

Recent studies have identified various MRE predictors of CD prognosis. Jauregui-Amezaga et al.¹¹ evaluated the prognostic value of MRE in predicting intestinal surgery, in which strictures and fistulas were observed as predictors of surgery within 2 years. Additionally, a retrospective cohort of 136 patients showed that the number and length of strictures were predictive of abdominal surgeries.¹² Furthermore, MRE-based scoring systems, such as the Magnetic Resonance Index of Activity (MaRIA) score, are associated with a decreased risk of surgery.¹³ These MRE findings and scoring systems have been utilized in CD management in recent years; however, a comprehensive review summarizing these predictors is still needed.

Therefore, the aim of this review was to summarize MRE predictors associated with poor prognosis in patients with CD.

Materials and methods

Search strategy and registration

This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.¹⁴ A comprehensive literature search was performed using Web of Science and PubMed databases to identify all studies published before July 21, 2025. Keywords focusing on “Crohn’s disease,” “magnetic resonance,” and their synonyms were used to formulate search queries in both databases. Studies categorized as meta-analyses, reviews, or case reports were excluded. The detailed search strategies are shown in Table S1.

The study protocol was registered in the International Prospective Register of Systematic Reviews (registration number: CRD42022365965).

Selection criteria

Studies meeting the following inclusion criteria were included: (1) patients were diagnosed with luminal CD; (2) MRE prognostic factors were used as predictors; (3) outcomes were related to any poor prognosis in CD, including surgical or endoscopic interventions, complications, hospitalizations, relapses, therapeutic escalations, and persistent pathological changes; (4) the follow-up time reached at least 12 months; (5) the studies were original articles, such as cohort or case-control studies.

The exclusion criteria were as follows: (1) studies that focused on pediatric CD; (2) studies that focused on perianal CD; (3) editorials, letters, case reports, and review articles; and (4) studies in languages other than English or without sufficient data to calculate.

Two reviewers (D.W. and R.C.) independently screened the titles and abstracts of all the articles. Full-text screening was then conducted to determine which studies would be included in the final analysis. Discrepancies were resolved through discussion, and if necessary, a third reviewer (Y.T.) was consulted.

Data extraction and quality assessment

The following data were collected from the eligible studies: authors, country, publication year, study design, sample size, patient age, follow-up time, outcomes, MRE predictors (e.g., MRE findings or scoring systems), and magnetic field strength. Odds ratios (ORs) and hazard ratios (HRs) for each prognostic factor were also collected for pooling analysis.

Two reviewers (D.W. and Z.F.) independently assessed the quality of the included studies using the Newcastle-Ottawa Scale. Any discrepancies were discussed to reach a consensus. The Newcastle-Ottawa Scale score ranged from 0 to 9, with scores of 7–9 and 4–6 indicating high and moderate quality, respectively.¹⁵

Prognostic factors and disease outcomes

All available MRE prognostic predictors for poor prognosis were included and divided into MRE findings and MRE scoring systems. Prognostic factors reported more than once were recorded for pooled analysis. Predictors of surgery were discussed separately. Other outcomes of interest included complications, hospitalizations, endoscopic interventions, relapses, therapy escalations, and persistent pathological changes. Prognostic factors reported only once are summarized in Table S4.

Disease outcomes included surgeries, complications, relapses, therapeutic escalations, and persistent pathological changes. Surgery was defined as an abdominal resection related to CD. Complications were defined as newly developed fistulas, pseudotumors, strictures, or abscesses. According to the definitions in eligible articles, relapse was defined as the need for therapy escalation, hospitalization, or surgery due to abdominal symptoms worsening (CD activity index >150 or 175^16,17 or Harvey–Bradshaw index >4 with an increase of at least 3 points compared to baseline¹⁸). Therapeutic escalation was defined as a change in medication dosage or type, and persistent pathological changes were defined as persistent creeping fat or bowel wall thickening.

Data analysis

The ORs/HRs and their 95% confidence intervals (CIs) were collected to evaluate the prognostic value of each MRE finding when available. When direct acquisition was not feasible, these values were calculated based on the data from the original texts. ORs were calculated by constructing a fourfold table based on exposures and outcomes, and the corresponding 95% CIs were obtained using the Woolf approximation.

When two or more studies were available, the ORs and corresponding 95% CIs were pooled and analyzed using a random-effects model. Heterogeneity was assessed by I²-test, where I² > 50% indicating high heterogeneity.¹⁹ To evaluate publication bias, funnel plots and Egger’s tests were conducted for each predictor. All analyses were performed using Stata MP 16.0 (StataCorp LLC., College Station, TX, USA). Statistical significance was set at p < 0.05.

Results

Literature search

The screening and selection process is illustrated in Figure 1. A total of 5841 studies were identified through PubMed (3099 studies) and Web of Science (2742 studies). After removing duplicates, 4758 studies were reviewed through title and abstract screening, of which 30 articles were included and 18 were eligible for meta-analysis.

Figure 1.

Flowchart for the study selection.

Characteristics of included studies

The characteristics of the 30 studies are shown in Table 1. Ten studies had a prospective design, and the others were retrospective. Among them, three studies had prospective recruitment, but the MRE observations were derived from a retrospective analysis of data.^13,20,21 There were no overlapping study groups or subjects among the included studies. Sample sizes ranged from 21 to 223 among different studies. Among these studies, 96.7% had a follow-up period of at least 12 months, except for one that reported an unclear follow-up duration. Twenty-three studies reported the prognostic values of MRE in predicting surgery, while six predicted disease relapses. All included studies had overall good quality (Tables S2 and S3).

Table 1.

Study characteristics.

Author	Publication year	Country	Study design	Enrollment period	Sample	Follow-up time, mo (median, IQR)	Age, y (median, IQR)	Magnetic strength	Predictor	Outcome
Jauregui-Amezaga et al.¹¹	2015	Spain	Prospective cohort	2007.6–2011.5	112	49 (22–65)	34 (27–46)	3T	MRE findings	Surgery
Gibson et al.²²	2015	Ireland	Retrospective cohort	2007.6–2013.6	75	16.7 (9.0–30.1)	24	1.5T	MRE findings	Surgery
Meunier et al.¹⁸	2015	Belgium	Prospective cohort	2012.3–2012.12	27	25.1 (22.1–27.1)	31 (26–40)	1.5T	MRE findings, MRE score	Relapse
Takenaka et al.²³	2017	Japan	Prospective cohort	2009.12–2014.12	200	21 (12–65)^†	16–71^†	1.5T	MRE findings	Surgery
Fernandes et al.²⁴	2017	Portugal	Retrospective cohort	NA	214	42 (12–94.8)	36.8 (16–77)	1.5T	MRE findings	Surgery, therapy escalation, hospitalization
Naganuma et al.¹⁷	2017	Japan	Prospective cohort	2013.1–2014.12	28	18.2 (12–36)^†	34.5 (16–56)^†	1.5T	MRE findings, MRE score	Relapse
Fiorino et al.²⁵	2017	Italy, France	Prospective cohort	2006.1.1–2014.12.31	142	58.8	28.8 (20.5–41.5)^†	1.5T or 3T	MRE score	Surgery, hospitalization
Amitai et al.²⁶	2018	Israel	Retrospective case-control	2012.1–2017.4	21	12	24 (20–25)	1.5T or 3T	MRE findings	Surgery, endoscopy intervention
Lee et al.¹⁶	2018	Korea	Retrospective cohort	2012.12–2016.3	173	15.9 (11.8–22.3)	29 (24–37)	3T	MRE findings	Relapse, hospitalization, therapy escalation
Rimola et al.²⁰	2018	Spain	Post hoc analysis of two prospective studies	2008.11–2016.11	28	31.65 (1–108)^†	34.9 (17–63.5)^†	1.5T or 3T	MRE findings; MRE score	Persistent pathological change
Takenaka et al.²⁷	2018	Japan	Prospective cohort	2012.7–2015.12	139	27 (20–37)	33 (12–68)^†	1.5T	MRE score	Surgery, relapse, hospitalization
Thierry et al.²⁸	2018	France	Retrospective cohort	2010.1–2016.12	96	26.4 (19.2–44.4)	33.1 (24.5–43.3)	1.5T	MRE score	Surgery
Althoff et al.²⁹	2019	Germany	Retrospective cohort	2003.1–2015.3	90	93 (52–94)	37 (26–47)	1.5T	MRE findings	Surgery, complication
Jones et al.³⁰	2019	UK	Retrospective cohort	2008–2012	104	69.5 (59.5–78.4)	47.9 (35.6–59.9)	1.5T	MRE findings	Surgery
Buisson et al.¹³	2019	France	Post hoc analysis of two prospective studies	2011.7–2014.5	63	28.8 ± 18.6^‡	34.8 ± 14.5^‡	1.5T	MRE score	Surgery
Sagami et al.³¹	2019	Japan	Retrospective cohort	2013.1–2017.12	50	14.97 (6.10–28.46)	34.5 (26.0–42.0)	1.5T	MRE score	Therapy escalation
Hallé et al.³²	2020	France	Retrospective cohort	2011.6–2017.6	115	17 (11.6–28.3)	32 (25–45.5)	1.5T	MRE findings	Surgery, endoscopy intervention
Schulberg et al.¹²	2020	Australia	Retrospective cohort	2011.1–2016.3	136	41 (26–56)	40 (29–48)	1.5T	MRE findings, MRE score	Surgery
Messadeg et al.³³	2020	France	Prospective cohort	2014.9–2015.12	46	46 (21.8–69.8)	34.9 ± 14.1^‡	1.5T	MRE score	Surgery, complication
Mainenti et al.³⁴	2021	Italy	Retrospective case control	2017.1–2018.6	70	12	35 (16–72)^†	3T	MRE findings, MRE score	Surgery
Kitazume et al.³⁵	2021	Japan	Retrospective cohort	2014.1–2015.2	89	60	33 (17–70)^†	1.5T	MRE score	Surgery, hospitalization
Lafeuille et al.³⁶	2021	France	Retrospective cohort	2012–2018	154	29.1(12.9–52.3)	⩾18	1.5T	MRE findings	Surgery, hospitalization
Pozios et al.³⁷	2022	Germany	Retrospective cohort	2010.6–2020.5	223	48.8 (20.0–82.9)	35 (26–48)	NA	MRE findings	Relapse, complication
El Ouali et al.³⁸	2022	USA	Retrospective cohort	2013.9–2016.4	86	63.4 (57–68.9)^§	36.7 (26.5–56.1)^§	NA	MRE findings	Surgery
Cankurtaran et al.³⁹	2023	Turkey	Retrospective cohort	2015.1–2021.8	116	12(12–79)^†	38.6 ± 12.9^‡	NA	MRE findings	Surgery
Chirra et al.⁴⁰	2023	USA	Retrospective case-control	2009–2015	73	60	36.3 ± 13.7^‡	1.5T or 3T	MRE findings	Surgery
Balestrieri et al.⁴¹	2023	Italy	Prospective cohort	2018.1–2020.12	103	24	36(16–67)^†	1.5T	MRE findings	Surgery, therapy escalation
Fernandez-Clotet et al.²¹	2024	Spain	Post hoc analysis of a prospective study	2012.2–2019.12	89	24	–	1.5T or 3T	MRE findings, MRE score	Surgery, relapse, hospitalization, therapy escalation
Dillman et al.⁴²	2024	USA	Prospective cohort	2019.12–2022.8	133	NA	23.9 ± 23.1	3T	MRE findings, MRE score	Surgery
Taylor et al.⁴³	2025	UK	Prospective cohort	2013–2022	194	⩾48	29(22–44)	1.5T or 3T	MRE score	Disabling disease

Continuous variables were described as median (interquartile range), median (range)^†, mean ± SD^‡, or median (95% confidence interval)^§.

IQR, interquartile range; MRE, magnetic resonance enterography; NA, not available.

Association between MRE findings and poor prognosis

We analyzed studies focusing on surgery and poor prognosis. The pooled analysis results are shown in Figures 2 and 3. Studies providing predictors not suitable for meta-analysis are described separately in Table 2.

Figure 2.

Forest plot showing the predictive value of magnetic resonance enterography predictors for poor prognosis using odds ratio as the effect size.

Figure 3.

Forest plot showing the predictive value of magnetic resonance enterography predictors for poor prognosis using hazard ratio as the effect size.

Table 2.

Summary of magnetic resonance enterography predictors for poor prognosis in Crohn’s disease.

Predictor	Study	Predictor type	Outcome	Effect size
Inflammation	Lee et al.¹⁶	Active inflammation	Relapse	aHR(95% CI): 8.603(1.026–72.138)
			Therapy escalation	aHR(95% CI): 5.113(1.187–22.029)
			Hospitalization	aHR(95% CI): 3.171(1.092–9.204)
	Althoff et al.²⁹	Proximal small bowel inflammation	Complication	OR(95% CI): 5(1.7–14.6)
		Small bowel inflammatory stenosis	Complication	OR(95% CI): 3.7(1.5–9.2)
			Surgery	OR(95% CI): 3.27(1.2–9.3)
Bowel wall thickening	Gibson et al.²²	Wall thickness	Surgery	aHR(95% CI): 3.292(0.31–35.52)
	Schulberg et al.¹²	Maximum wall thickness (mm)	Surgery	OR(95% CI): 1.31(1.12–1.55)
Stricture	Naganuma et al.¹⁷	Severe stricture	Relapse	OR(95% CI): 4(0.47–33.3)
	Takenaka et al.²³	Stricture	Surgery in 6 months	aHR(95% CI): 27.16(3.47–212.75)
	Schulberg et al.¹²	More than one stricture	Surgery	OR(95% CI): 1.98(1.00–3.96)
		Luminal diameter of narrowed segment ⩽2 mm	Surgery	OR(95% CI): 0.42(0.21–0.84)
		Stricture length >5 cm	Surgery	OR(95% CI): 3.26(1.61–6.62)
		Stricture length (mm)	Surgery	OR(95% CI): 1.00(1.001–1.008)
		Severe stricture	Surgery	OR(95% CI): 1.67(0.40–7.01)
	El Ouali et al.³⁸	Stricture length (cm)	Surgery	HR(95% CI): 1.04(1.01–1.07)
Ulcer	Meunier et al.¹⁸	Ulcer	Relapse	HR(95% CI): 12.5(1.13–138)
	El Ouali et al.³⁸	Ulcer	Surgery	HR(95% CI): 1.05(0.47–2.34)
Fistula	Meunier et al.¹⁸	Fistula	Relapse	HR(95% CI): 14.1(1.95–102)
	Fernandez-Clotet et al.²¹	Fistula or abscesses	Surgery	OR(95% CI): 2.6(0.9–6.9)

aHR, adjusted hazard ratio; CI, confidence interval; HR, hazard ratio; OR, odds ratio.

Inflammation

MRE can detect inflammation not visible via endoscopy.^8,44 The inflammation was often defined by a group of heterogeneous indicative signs in radiology, including bowel wall thickening >3 mm, increased contrast uptake, extra-mural hypervascularity, or ulcer.¹⁷ However, another study added bowel wall edema, the combs sign, lymphadenopathy, mesenteric fibrofatty proliferation, and creeping fat as indicative signs.²⁹ A pooled analysis of three studies showed MRE had stable prognostic performances in predicting surgery or relapse, with a pooled OR of 3.40 (95% CI (1.96–5.89)). No significant heterogeneity was observed (I² = 0.0%, p = 0.857).^12,17,29 A cohort of patients also reported the prognostic value of inflammation for therapeutic escalation (HR (95% CI): 5.133 (1.187–22.029)) and hospitalization (HR (95% CI): 3.171 (1.092–9.204)).¹⁶

Bowel wall thickening

Five studies observed that wall thickening increased the risk of poor prognosis, including surgeries, relapses, and persistent wall thickening with a pooled OR (95% CI) of 2.86 (2.03–4.03) and no significant heterogeneity (I² = 0.0%, p = 0.849).^{12,17,20,30,40} Notably, wall thickening was defined as wall thickness >3 mm in four studies, while one study defined it by 10 mm.¹² Considering this heterogeneity, further analysis of the four studies with consistent definitions yielded a stable conclusion (OR (95% CI): 2.88 (1.95–4.25), I² = 0.0%, p = 0.71). As for other related variables, maximum wall thickness was capable of predicting surgery as a continuous variable, with an OR (95% CI) of 1.31 (1.12–1.55) in a retrospective cohort of 136 patients.¹² In 2023, one study revealed that the ratio between the wall thickness at baseline and at 1 year could predict therapeutic responses, indicating the potential prognostic value of wall thickness change.⁴¹

Stricture

Although specific cutoff values varied between studies, stricture was defined as narrowing of the lumen with or without upstream dilatation.^{11,17,20,24,32,34} The pooled analysis of six studies showed that patients with stricture had nearly fivefold risk of poor outcomes in 1 year (OR (95% CI): 4.95 (1.99–12.34)), with significant heterogeneity (I² = 74.7%, p = 0.001).^{11,17,20,24,32,34} For 1-year surgery, a pooled OR (95% CI) of 6.64 (2.90–15.23) with low heterogeneity (I² = 31.5%, p = 0.23)^11,32,34 and a pooled HR (95% CI) of 8.95 (3.05–26.24) with high heterogeneity (I² = 60.8%, p = 0.078) were observed.^11,22,23

Ulcer

Although ulcers were one of the typical manifestations in CD, the pooled analysis of five studies showed no significance in predicting poor prognosis in 1 year (OR (95% CI): 1.41 (0.91–2.19)), with no heterogeneity(I² = 0.0%, p = 0.717).^{11,12,20,34,40} Of the two studies providing HRs, one reported that ulcers were associated with relapse (HR (95% CI): 12.5 (1.13–138)),¹⁸ while the other found no statistical significance³⁸ (Table 2).

Edema

Patients with edema in the mesentery or bowel wall had a higher risk of poor prognosis, with a pooled OR (95% CI) of 2.67 (1.25–5.69) from five studies.^{12,30,26,34,40}

Fistula

Three studies showed fistula detected by MRE exerted as a prognostic factor for surgery or additional endoscopic dilation in 1 year (OR (95% CI): 5.15 (2.72–9.76)) with no significant heterogeneity (I² = 0.0%, p = 0.498).^11,12,32

Others

Some other MRE findings were reported but not eligible for pooled analysis. These included fat-related findings, myopenia, multisegmental bowel involvement, and extramural hypervascularity (Table S4).

Association between MRE scoring system and poor prognosis

The MaRIA score demonstrated high and stable accuracy in detecting CD activity.^45,46 Due to various cutoff values across studies, a pooled analysis was conducted by treating the predictor as a dichotomous variable, categorizing patients into high or low MaRIA score group. Five studies using different thresholds were included (three studies used 7, 11, and 24 for the segmental MaRIA index, and two used 50 and 62 for the global MaRIA score), with no significant heterogeneity observed (Figure 2). The pooled analysis indicated a higher MaRIA score correlated with a higher risk of poor prognosis in 1 year (OR (95% CI): 3.90 (2.24–6.81); HR (95% CI): 3.62 (2.17–6.03)).^{12,13,21,27,31,35}

Other studies not suitable for meta-analysis are summarized in Table 3, where the MaRIA score was used in both continuous and dichotomous forms. One study used the logarithmic MaRIA score as a continuous variable; however, no significance was observed in predicting disease relapse.¹⁸ Recently, a multicenter prospective cohort also reported the simplified MaRIA score showed no predictive value of disabling disease in 5 years with an HR (95% CI) of 1.00 (0.98–1.01).⁴³ Decreases in MaRIA score were associated with corticosteroid-free remission in a cohort of 46 patients (OR (95% CI): 4.2 (1.3–13.3)).³³ Elements of the MaRIA score, such as relative contrast enhancement (RCE) independently showed statistical significance in predicting relapse (OR (95% CI): 2.56 (1.02–6.45), Table S4).¹⁸ However, in three other studies, RCE failed to predict surgery or relapse.^12,17,22

Table 3.

Summary of magnetic resonance enterography-related scoring systems for poor prognosis in Crohn’s disease.

Predictor	Study	Predictor type	Outcome	Effect size
MaRIA score	Meunier et al.¹⁸	MaRIA (logarithm)	Relapse	HR(95% CI): 0.85(0.07–11.2)
		Segmental maximal MaRIA (logarithm)	Relapse	HR(95% CI): 3.04(0.81–11.45)
	Fiorino et al.²⁵	MaRIA score	Surgery	HR(95% CI): 1.00(0.99–1.01)
	Takenaka et al.²⁷	MaRIA <11	Relapse	HR(95% CI): 0.24(0.11–0.54)
	Rimola et al.²⁰	Baseline MaRIA score	Persistent wall thickening	OR(95% CI): 0.469(0.25–1.2)
	Buisson et al.¹³	MaRIA <7	Surgery	aHR(95% CI): 0.16(0.043–0.63)
	Sagami et al.³¹	MaRIA score of ⩾50	Therapy escalation	HR(95% CI): 2.65(1.23–5.70)
			Shorter time to therapy escalation (in patients with endoscopic remission)	HR(95% CI): 4.14(1.09–15.6)
	Schulberg et al.¹²	MaRIA score for dominant stricture	Surgery	OR(95% CI): 1.08(1.04–1.13)
		MaRIA score for dominant stricture >24	Surgery	OR(95% CI): 3.59(1.73–7.47)
		MaRIA score for dominant stricture >11	Surgery	OR(95% CI): 3.57(0.39–32.85)
	Messadeg et al.³³	25% decrease of MaRIA at week 12	Complication	HR(95% CI): 0.36(0.1–1.2)
	Kitazume et al.³⁵	Global MaRIA ⩾62	Hospitalization	OR(95% CI): 3.57(1.39–9.18)
			Surgery	OR(95% CI): 2.48(0.77–7.98)
	Chirra et al.⁴⁰	sMaRIA	Surgery	HR(95% CI): 3.68(1.42–9.55)
	Fernandez-Clotet et al.²¹	MaRIA ⩾11	Relapse	OR(95% CI): 1.5(1.0–2.1)
	Dillman et al.⁴²	sMaRIA	Surgery	OR(95% CI): 1.76(1.32–2.34)
	Taylor et al.⁴³	Normalized global sMaRIA (%)	Disabling disease	HR(95% CI): 1.00(0.98–1.01)
Other scoring systems	Rimola et al.²⁰	Baseline London index	Persistent wall thickening	OR(95% CI): 1.81(1.22–2.68)
	Buisson et al.¹³	MaRIA <11or or Clermont score <8.4	Surgery	aHR(95% CI): 0.24(0.07–0.77)
	Messadeg et al.³³	25% decrease of the Clermont score at week 12	Complication	HR(95% CI): 0.21(0.0–0.9)
		Transmural response score ⩾2	Complication	HR(95% CI): 0.51(0.2–1.5)
	Kitazume et al.³⁵	Global MREC score ⩾11	Hospitalization	OR(95% CI): 5.84(2.14–15.9)
			Surgery	OR(95% CI): 3.61(1.05–12.39)
	Taylor et al.⁴³	Normalized global MEGS (%)	Disabling disease	HR(95% CI): 1.01(0.99–1.02)
		Normalized global sMaRIA (%)	Disabling disease	HR(95% CI): 1.00(0.98–1.01)

aHR, adjusted hazard ratio; CI, confidence interval; HR, hazard ratio; MaRIA, Magnetic Resonance Index of Activity; MREC, Magnetic Resonance Enterocolonography score; OR, odds ratio; sMaRIA, simplified Magnetic Resonance Index of Activity.

The Clermont score assesses ileal inflammation in patients with CD. It includes the components of the MaRIA score (except RCE, which requires gadolinium injection) and the apparent diffusion coefficient (ADC) derived from diffusion-weighted MRE.⁴⁷ Two studies reported the prognostic value of the Clermont score. One reported a decreased risk of surgery when the baseline Clermont score ⩽8.4 for each segment (HR (95% CI): 0.24 (0.07–0.78)).¹³ Another study showed that a 25% decrease in the Clermont score at week 12 indicated a lower risk of bowel damage at week 52 (HR (95% CI): 0.21 (0.0–0.9)).³³ Some studies focused on ADC values, indicating a negative correlation between ADC values and surgery risk. One study reported that an ADC value of <1 × 10⁻³ mm² s⁻¹ could predict higher risks of treatment failure (therapy escalation, need for surgery, or hospitalization), with an OR of 30.1.²⁶ In a case-control study of 70 patients, the mean ADC was associated with conservative management (OR (95% CI): 0.0003 (0.00–0.13)), with an area under the curve of 0.717 (0.607–0.810).³⁴

Others

Other reported scoring systems are listed in Table 3, including the London score, magnetic resonance enterocolonography score, and newly developed scoring systems.

Risk of publications bias

The funnel plot of strictures showed asymmetry (p < 0.05) in the Egger’s test, indicating publication bias (Figure S1). Using “trim-and-fill” analysis for correction, the pooled OR (95% CI) was 2.007 (1.388–2.903) and 2.205 (0.894–5.437) from the fixed and random models, respectively. The funnel plots of other predictors were symmetrical, indicating a low risk of publication bias.

Discussion

MRE has emerged as a valuable tool in CD diagnosis and management, and MRE predictors have been evaluated in multiple recent studies on CD prognosis. This review identified and summarized significant MRE predictors for CD prognosis, including inflammation, bowel wall thickening, strictures, edema, fistulas, and scoring systems. These predictors might provide valuable information for risk stratification and treatment planning. Besides, our results highlight the potential for MRE-derived parameters to noninvasively monitor treatment response, reducing reliance on repeated endoscopic evaluations in CD management. For the radiologists reporting MREs, these predictors may help optimizing MRE reporting protocols and we suggest for establishing standardizing measurement thresholds.

Through meta-analysis, we identified five MRE findings that effectively predicted poor prognosis in 1 year, including the presence of inflammation, bowel wall thickening, stricture, edema, and fistula. Mural inflammation is a typical symptom of CD, however, its definition varies across studies and may overlap with other MRE findings. Despite the pooled analysis indicating statistical significance, the lack of a unified definition may complicate its practical application. In contrast, the other four characteristics are more clearly defined. For example, a wall thickness cutoff value of 3 mm is the most frequently used, aligning with MRE reporting recommendations.⁴⁸ Stricture is mostly defined as luminal diameter reduction with prestricture dilation,⁴⁸ though some studies have used a distinct bowel wall thickness to define it.^20,34 The presence of stricture is typical in luminal CD and has been extensively studied for its predictive value. We did not explore the heterogeneity between studies; however, we suspect that our eligibility criteria did not restrict patient medication. As summarized in Table 2, the number and length of strictures also serve as predictors, showing considerable prognostic value in predicting relapse and surgery within 1 year.^12,38 Edema, often observed as a hyperintense signal on fat-saturated T2-weighted MRE, significantly correlated with inflammation severity and showed a strong correlation with poor prognosis in our meta-analysis.⁴⁹ Ulcers were not predictive of disease prognosis in our review, and we guess that this may be due to the limited ability of MRE to recognize ulcer, especially shallow ulcers, which may be somewhat restricted its predictive value.

This study also summarized some less frequently reported MRE findings. Although some of these findings remained controversial, they may inspire future research. For example, fat-related findings could reflect the inflammation level or absorption function of the digestive tract, indicating considerable potential for predicting disease prognosis. Two studies reported creeping fat as a predictor for various poor outcomes such as surgery and disabling disease.^20,29 Myosteatosis has a poor predictive value for hospitalization according to 2 studies involving 463 patients.^37,39 Findings such as fat stranding and myopenia require further investigation due to controversial results among studies.^20,37,39,40 Other predictors, like multi-segmental bowel involvement and extra-mural hypervascularity, also warrant further research. Upstream dilation of the intestinal tract on MRE might be a potential predictor, although computed tomography (CT) is used more often. This may partly result from CT reports of prestenotic dilation occurring more often than MRE reports when evaluating strictures.⁵⁰

The scoring systems summarized in this review were mostly based on typical MRE findings and parameters such as the ADC value. These scoring systems may reveal different predictive capacities by using different cutoff values or transformations.^18,27 The MaRIA score, for instance, was first used to assess ileocolonic CD activity and was based on wall thickness, RCE, edema, and ulcers to calculate, ranging from 3.4 to 35.0. Generally, cutoff values of 7 or 11, which indicated active or severe disease according to the derivation cohort, are most often used.⁴⁶ Some studies have divided the gastrointestinal tract to calculate a global MaRIA score and have reported encouraging results.^31,35 This may widen the application of the score.

This review had several strengths. We used broad search terms and did not restrict by country or race, ensuring the inclusion of as many MRE prognostic factors as possible. We also included studies that did not report effect estimates but provided sufficient data to calculate them. For predictors where OR/HR could not be calculated, we tabulated other parameters to reflect their predictive power (Table S5, which shows the results of studies ineligible for ORs or HRs). However, this review had some limitations. Firstly, heterogeneity occurred in our analysis due to the small sample size, various interventions, or distinct definitions of predictors. Besides, our review did not systematically assess the inter-rater reliability or the number of readers in the included studies. Variations in these methodological aspects (e.g., single vs multiple readers, inconsistent scoring criteria) may also introduce heterogeneity, potentially affecting the generalizability of our conclusions. In this way, our findings require further validation in cohorts with larger sample sizes. Furthermore, our review focused on the patients with essential disease activity on the MRE, which might neglect some manifestations of mild disease observed under endoscopy, such as aphthous ulcers or erosions. Notably, a previous systematic review has reported that MRE demonstrates moderate-to-excellent accuracy in evaluating disease activities in IBD, which showed superior performance than ultrasonography and CT, especially in proximal small bowel disease.⁵¹ Besides, though MRE might neglect some mild manifestations, mild disease may not need near-term treatment escalation or surgery.^51,52

Conclusion

In summary, this review summarized the MRE predictors that are effective for predicting poor prognosis in patients with luminal CD, including inflammation, bowel wall thickening, strictures, edema, fistula, and scoring systems like MaRIA score. This highlights the potential for MRE-derived parameters on monitoring treatment response noninvasively and reducing reliance on repeated endoscopic evaluations in CD management. However, further prospective studies with larger sample sizes and unified definitions are needed to confirm these results and enhance the clinical utility of MRE in predicting CD prognosis.

Supplemental Material

sj-docx-1-tag-10.1177_17562848251386317 – Supplemental material for Magnetic resonance enterography for predicting poor prognosis in Crohn’s disease: a systematic review with meta-analysis

Supplemental material, sj-docx-1-tag-10.1177_17562848251386317 for Magnetic resonance enterography for predicting poor prognosis in Crohn’s disease: a systematic review with meta-analysis by Yizhe Tie, Daiyue Wang, Zinan Fan, Li Li, Minhu Chen, Rirong Chen and Shenghong Zhang in Therapeutic Advances in Gastroenterology

Supplemental Material

sj-docx-2-tag-10.1177_17562848251386317 – Supplemental material for Magnetic resonance enterography for predicting poor prognosis in Crohn’s disease: a systematic review with meta-analysis

Supplemental material, sj-docx-2-tag-10.1177_17562848251386317 for Magnetic resonance enterography for predicting poor prognosis in Crohn’s disease: a systematic review with meta-analysis by Yizhe Tie, Daiyue Wang, Zinan Fan, Li Li, Minhu Chen, Rirong Chen and Shenghong Zhang in Therapeutic Advances in Gastroenterology

Footnotes

Appendix

Acknowledgements

We acknowledge Editage () for English editing.

Declarations

ORCID iDs

Yizhe Tie

Li Li

Minhu Chen

Rirong Chen

Shenghong Zhang

Supplemental material

Supplemental material for this article is available online.

References

Roda

Chien

Ng S

Kotze

, et al. Crohn’s disease. Nat Rev Dis Primers 2020; 6: 22.

Dolinger

Torres

Vermeire

Crohn’s disease. Lancet (London, England) 2024; 403: 1177–1191.

Coward

Clement

Benchimol

, et al. Past and future burden of inflammatory bowel diseases based on modeling of population-based data. Gastroenterology 2019; 156: 1345–1353.e4.

Tsai

Dulai

, et al. Contemporary risk of surgery in patients with ulcerative colitis and Crohn’s disease: a meta-analysis of population-based cohorts. Clin Gastroenterol Hepatol 2021; 19: 2031–2045.e11.

Fiorino

Bonifacio

Peyrin-Biroulet

, et al. Preventing collateral damage in Crohn’s disease: the Lémann index. J Crohns Colitis 2016; 10: 495–500.

Kayar

Baran

Ormeci

, et al. Risk factors associated with progression to intestinal complications of Crohn disease. Chin Med J 2019; 132: 2423–2429.

Plevris

Lees

CW.

Disease monitoring in inflammatory bowel disease: evolving principles and possibilities. Gastroenterology 2022; 162: 1456–1475.e1.

Bruining

Zimmermann

Loftus EV

, et al. Consensus recommendations for evaluation, interpretation, and utilization of computed tomography and magnetic resonance enterography in patients with small bowel Crohn’s disease. Gastroenterology 2018; 154: 1172–1194.

Buisson

Gonzalez

Poullenot

, et al. Comparative acceptability and perceived clinical utility of monitoring tools: a nationwide survey of patients with inflammatory bowel disease. Inflamm Bowel Dis 2017; 23: 1425–1433.

10.

Lee

Kim

Yang

S-K

, et al. Crohn disease of the small bowel: comparison of CT enterography, MR enterography, and small-bowel follow-through as diagnostic techniques. Radiology 2009; 251: 751–761.

11.

Jauregui-Amezaga

Rimola

Ordás

, et al. Value of endoscopy and MRI for predicting intestinal surgery in patients with Crohn’s disease in the era of biologics. Gut 2015; 64: 1397–1402.

12.

Schulberg

Wright

Holt

, et al. Magnetic resonance enterography for predicting the clinical course of Crohn’s disease strictures. J Gastroenterol Hepatol 2020; 35: 980–987.

13.

Buisson

Hordonneau

Goutorbe

, et al. Bowel wall healing assessed using magnetic resonance imaging predicts sustained clinical remission and decreased risk of surgery in Crohn’s disease. J Gastroenterol 2019; 54: 312–320.

14.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021; 372: n71.

15.

Wells

Shea

O’Connell

, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp (2009, accessed 3 May 2025).

16.

Lee

Park

Seo

, et al. Magnetic resonance enterography predicts the prognosis of Crohn’s disease. Intest Res 2018; 16: 445–457.

17.

Naganuma

Okuda

Hisamatsu

, et al. Findings of ulceration and severe stricture on MRE can predict prognosis of Crohn’s disease in patients treated with anti-TNF treatment. Abdom Radiol (NY) 2017; 42: 141–151.

18.

Meunier

Cousin

Van Kemseke

, et al. Persisting signs of disease activity at magnetic resonance enterocolonography predict clinical relapse and disease progression in quiescent Crohn’s disease. Acta Gastroenterol Belg 2015; 78: 274–281.

19.

Higgins

JPT

Thompson

Deeks

, et al. Measuring inconsistency in meta-analyses. BMJ (Clin Res Ed) 2003; 327: 557–560.

20.

Rimola

Alfaro

Fernández-Clotet

, et al. Persistent damage on magnetic resonance enterography in patients with Crohn’s disease in endoscopic remission. Aliment Pharmacol Ther 2018; 48: 1232–1241.

21.

Fernandez-Clotet

Ordas

Masamunt

, et al. Magnetic resonance enterography findings 46 weeks after initiation of biological therapy predict long-term adverse outcomes in Crohn’s disease. Aliment Pharmacol Ther 2024; 59: 1435–1445.

22.

Gibson

Murphy

Smyth

, et al. Magnetic resonance enterography findings as predictors of clinical outcome following antitumor necrosis factor treatment in small bowel Crohn’s disease. Eur J Gastroenterol Hepatol 2015; 27: 956–962.

23.

Takenaka

Ohtsuka

Kitazume

, et al. Magnetic resonance evaluation for small bowel strictures in Crohn’s disease: comparison with balloon enteroscopy. J Gastroenterol 2017; 52: 879–888.

24.

Fernandes

Rodrigues

Bernardo

, et al. Transmural healing is associated with improved long-term outcomes of patients with Crohn’s disease. Inflamm Bowel Dis 2017; 23: 1403–1409.

25.

Fiorino

Morin

Bonovas

, et al. Prevalence of bowel damage assessed by cross-sectional imaging in early Crohn’s disease and its impact on disease outcome. J Crohns Colitis 2017; 11: 274–280.

26.

Amitai

Klang

Levartovsky

, et al. Diffusion-weighted magnetic resonance enterography for prediction of response to tumor necrosis factor inhibitors in stricturing Crohn’s disease. Abdom Radiol (NY) 2018; 43: 3207–3212.

27.

Takenaka

Ohtsuka

Kitazume

, et al. Utility of magnetic resonance enterography for small bowel endoscopic healing in patients with Crohn’s disease. Am J Gastroenterol 2018; 113: 283–294.

28.

Thierry

Rousseau

Pouillon

, et al. Accuracy of diffusion-weighted magnetic resonance imaging in detecting mucosal healing and treatment response, and in predicting surgery, in Crohn’s disease. J Crohns Colitis 2018; 12: 1180–1190.

29.

Althoff

Schmiegel

Lang

, et al. Creeping fat assessed by small bowel MRI is linked to bowel damage and abdominal surgery in Crohn’s disease. Dig Dis Sci 2019; 64: 204–212.

30.

Jones

Fascì-Spurio

Kennedy

, et al. Faecal calprotectin and magnetic resonance enterography in ileal Crohn’s disease: correlations between disease activity and long-term follow-up. J Crohns Colitis 2019; 13: 442–450.

31.

Sagami

Kobayashi

Kikkawa

, et al. Combination of colonoscopy and magnetic resonance enterography is more useful for clinical decision making than colonoscopy alone in patients with complicated Crohn’s disease. PLoS One 2019; 14: e0212404.

32.

Hallé

Azahaf

Duveau

, et al. Radiological response is associated with better outcomes and should be considered a therapeutic target in Crohn’s disease. Dig Dis Sci 2020; 65: 2664–2674.

33.

Messadeg

Hordonneau

Bouguen

, et al. Early transmural response assessed using magnetic resonance imaging could predict sustained clinical remission and prevent bowel damage in patients with Crohn’s disease treated with anti-tumour necrosis factor therapy. J Crohns Colitis 2020; 14: 1524–1534.

34.

Mainenti

Castiglione

Rispo

, et al. MR-enterography in Crohn’s disease: what MRE mural parameters are associated to one-year therapeutic management outcome? Br J Radiol 2021; 94: 20200844.

35.

Kitazume

Takenaka

Ohtsuka

, et al. Crohn disease: magnetic resonance enterocolonography features of endoscopic ulcer stages reclassified with the healing process and the relationships to prognoses. Jpn J Radiol 2021; 39: 459–476.

36.

Lafeuille

Hordonneau

Vignette

, et al. Transmural healing and MRI healing are associated with lower risk of bowel damage progression than endoscopic mucosal healing in Crohn’s disease. Aliment Pharmacol Ther 2021; 53: 577–586.

37.

Pozios

Kaufmann

Boubaris

, et al. Impact of myopenia and myosteatosis on postoperative outcome and recurrence in Crohn’s disease. Int J Colorectal Dis 2022; 37: 791–804.

38.

El Ouali

Baker

Lyu

, et al. Validation of stricture length, duration and obstructive symptoms as predictors for intervention in ileal stricturing Crohn’s disease. United European Gastroenterol J 2022; 10: 958–972.

39.

Cankurtaran

Güneş

Dirican

, et al. Sarcopenia and myosteatosis assessed by magnetic resonance enterography may predict negative outcomes in patients with Crohn’s disease. Turk J Gastroenterol 2023; 34: 839–849.

40.

Chirra

Sharma

Bera

, et al. Integrating radiomics with clinicoradiological scoring can predict high-risk patients who need surgery in Crohn’s disease: a pilot study. Inflamm Bowel Dis 2023; 29: 349–358.

41.

Balestrieri

Ribolsi

Cimini

, et al. Wall thickness ratio—a new magnetic resonance parameter—is associated with the outcome of biological therapy in patients with ileal and ileocolonic Crohn’s disease. J Clin Gastroenterol 2024; 58: 64–70.

42.

Dillman

Tkach

Fletcher

, et al. MRI and blood-based biomarkers are associated with surgery in children and adults with ileal Crohn’s disease. Inflamm Bowel Dis 2024; 30: 2181–2190.

43.

Taylor

Kumar

Parry

, et al. Magnetic resonance enterography to predict subsequent disabling Crohn’s disease in newly diagnosed patients (METRIC-EF)-multivariable prediction model, multicentre diagnostic inception cohort. Eur Radiol. Epub ahead of print May 2025. DOI: 10.1007/s00330-025-11636-8.

44.

Coimbra

Rimola

O’Byrne

, et al. Magnetic resonance enterography is feasible and reliable in multicenter clinical trials in patients with Crohn’s disease, and may help select subjects with active inflammation. Aliment Pharmacol Ther 2016; 43: 61–72.

45.

Rimola

Rodriguez

García-Bosch

, et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn’s disease. Gut 2009; 58: 1113–1120.

46.

Rimola

Ordás

Rodriguez

, et al. Magnetic resonance imaging for evaluation of Crohn’s disease: validation of parameters of severity and quantitative index of activity. Inflamm Bowel Dis 2011; 17: 1759–1768.

47.

Buisson

Joubert

Montoriol

, et al. Diffusion-weighted magnetic resonance imaging for detecting and assessing ileal inflammation in Crohn’s disease. Aliment Pharmacol Ther 2013; 37: 537–545.

48.

Bruining

Zimmermann

Loftus EV

49.

Choi

Kim

Lee

, et al. Diffusion-weighted magnetic resonance enterography for evaluating bowel inflammation in Crohn’s disease: a systematic review and meta-analysis. Inflamm Bowel Dis 2016; 22: 669–679.

50.

Bettenworth

Bokemeyer

Baker

, et al. Assessment of Crohn’s disease-associated small bowel strictures and fibrosis on cross-sectional imaging: a systematic review. Gut 2019; 68: 1115–1126.

51.

Dal Buono

Faita

Armuzzi

, et al. Assessment of activity and severity of inflammatory bowel disease in cross-sectional imaging techniques: a systematic review. J Crohns Colitis 2025; 19: jjaf023.

52.

Koilakou

Sailer

Peloschek

, et al. Endoscopy and MR enteroclysis: equivalent tools in predicting clinical recurrence in patients with Crohn’s disease after ileocolic resection. Inflamm Bowel Dis 2010; 16: 198–203.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.24 MB