A Simple Endoscopic Score for Crohn’s disease (SES-CD) ≥ 7 predicts disease progression

Summary

Background:

Up to 30% of patients with Crohn’s disease (CD) will experience a mild disease course. However, there is no consensus definition for mild CD.

Aim:

To examine the Simple Endoscopic Score for Crohn’s disease (SES-CD) thresholds best associated with low likelihood of long-term disease progression

Methods:

We conducted a multicentre retrospective cohort study at three tertiary care centres in United States and Europe. We analysed data from 177 surgery-naïve patients with CD who had endoscopic assessment while not on immunosuppressive therapy. The primary outcome was disease progression (systemic steroids, biologic or immunomodulator therapy initiation; new stricturing or penetrating complications; or CD-related hospitalization or surgery). Univariable and multivariable Cox proportional hazards modelling identified predictors of the primary outcome at two- and five-years following endoscopy.

Results:

Disease progression occurred in 23% and 35% of patients at years two and five, respectively. Endoscopic severity at enrolment independently predicted disease progression. Compared to those with an SES-CD of 0, an SES-CD ≥ 7 had a greater risk of progression at two years (HR 2.50, 95% CI 1.09 – 5.72) and five years (HR 2.89, 95% CI 1.41 – 5.91). SES-CD ≥ 7 remained independently predictive of disease progression among the 129 immunosuppression-naïve patients (HR 5.65, 95% CI 1.49 – 21.52) and after excluding patients with prior penetrating disease (HR 2.32, 95% CI 1.00 – 5.45).

Conclusions:

SES-CD ≥ 7 predicts disease progression in mild CD. A score ≤ 6 may help identify patients less likely to progress and be part of the definition of mild CD.

Graphical Abstract

There is currently no consensus on the definition of mild Crohn’s disease (CD). A SES-CD of ≥7 is linked to long-term disease progression. A score ranging from 1 to 6 could serve as part of the criteria for defining mild CD.

INTRODUCTION

Crohn’s disease (CD) is a chronic inflammatory disease of the gut affecting nearly one million individuals in the United States.¹ In many, it can lead to irreversible bowel damage with complications including strictures, fistulae, and need for bowel resection. Epidemiological studies suggest that about 50% of patients with a non-stricturing, non-penetrating phenotype of CD will develop intestinal complications within 20 years of diagnosis.² However, approximately 20–30% of individuals with CD may experience a relatively mild disease course without progression to complications or need for advanced therapies.³ Current therapeutic recommendations advise early initiation of immunosuppressive therapy in most patients with CD to prevent disease progression. However, this recommendation is extrapolated from clinical trials that examined treatment efficacy and benefit only in moderate-to-severe CD. There is a dearth of both interventional trials as well as informed management recommendations in mild CD. An important contributor to the paucity of data on the management strategy for mild CD is the lack of formally validated consensus definitions for mild CD as well as incomplete understanding of determinants of this disease state.^4,5 Various societies have proposed working definitions of mild CD that use a combination of clinical features.^3,6,7 For example, a Crohn’s Disease Activity Index (CDAI) between 150 and 220 is considered consistent with mild disease.⁷ Other clinical factors associated with mild activity include the ability to ambulate and tolerate oral nutrition, absence of signs of systemic toxicity, abdominal pain, painful abdominal mass, intestinal obstruction, or significant weight loss.⁴ Limitations of these classifications include reliance primarily on symptom burden which has been shown to correlate poorly with objective inflammation;⁸ a sizeable fraction of patients with moderate-to-severe disease burden may have only mildly symptomatic disease. Thus, developing objective criteria to classify disease severity in CD is important.

Several endoscopic scoring systems developed to evaluate CD activity in clinical trials are used in practice to objectively quantify disease activity. ⁹ For example, the Simple Endoscopic Score for Crohn’s Disease (SES-CD) is commonly used in patients who have not undergone CD-related intestinal surgery and quantifies severity based on extent and severity of ulcers and other inflammatory features.¹⁰ Society guidelines and the majority of clinical trials in this population have considered an SES-CD of 3 to 6 as a criterion for mild CD.^3,11,12 However, this cut-off was informed primarily by expert opinion and has not been formally evaluated in studies with longitudinal follow-up.^9,13 In defining mild CD, it is important to examine the association of endoscopic score with risk of longer-term clinical outcomes and define optimal cut-offs for mild versus moderate-to-severe disease. Few prior studies have identified endoscopic prognostic factors indicative of a low risk of progression.^5,7,14

The aims of our study were to examine the predictive value of SES-CD in determining two- and five-year outcomes in surgery-naïve patients with CD not on immunosuppressive treatment and to validate existing SES-CD cut-offs for mild CD that indicate a lower risk of disease progression.

METHODS

Study Population

The study comprised surgery-naïve patients with CD who had prospectively scored endoscopic assessment while not on immunosuppressive therapy and subsequent longitudinal follow-up seen at one of three referral inflammatory bowel disease (IBD) centres, Massachusetts General Hospital (MGH) (Boston, USA), the Mount Sinai Hospital (New York, USA) and the Hospital Beatriz Ângelo (Lisbon, Portugal).

At MGH, participants included in this study were enrolled from a prospective patient registry (PRISM). After obtaining informed consent, participants completed a detailed intake questionnaire that assessed disease and medication history as well as any complications. Patients also provide blood and stool samples as well as biopsies during standard of care endoscopic procedures. For eligibility for this study, patients needed to have an established diagnosis of CD, undergo a prospectively scored endoscopic assessment while not being on immunomodulator or biologic therapy, and have no prior history of CD-related bowel surgery.

The Mount Sinai Crohn’s and Colitis Registry (MSCCR) is a longitudinal cohort of 464 patients with CD recruited between 2013–2016 at the time of a clinically indicated colonoscopy, previously described.¹⁵ Cohort data is prospectively linked with electronic medical records for longitudinal follow-up. Mean follow up in MSCCR is currently 5.8 years. For inclusion in this study, patients met similar criteria as outlined for MGH cohort, namely surgery-naïve patients who were not on immunosuppression at the time of endoscopic assessment.

Patients from Hospital Beatriz Ângelo were actively followed at the IBD consultation service at this public district hospital located in the South of Portugal. Eligible patients had a new diagnosis of CD and were identified through the hospital’s IBD database. Inclusion criteria (similar to the other two cohorts) and pertinent variables were obtained via electronic health records. Colonoscopies were performed by gastroenterologists with expertise in IBD who reported the SES-CD score at the time of examination.

Covariates

The primary variable of interest for this study was the endoscopic activity of CD defined by SES-CD score. At all study sites, endoscopies were prospectively scored at the time of the procedure by a gastroenterologist with expertise in IBD and endoscopic scoring. Both total SES-CD as well as segmental scores for the ileum and colon were calculated. Other variables of interest included demographics including age, sex, and race as well as disease location, phenotype, and perianal involvement according to the Montreal classification.¹⁶ We recorded details of whether patients had been previously on immunomodulator or biologic for the management of CD. However, no patients were on immunomodulators or biologics at the time of the endoscopy.

Outcomes

The primary study outcome was disease progression defined as need for systemic steroids, biologic or immunomodulator therapy initiation, development of new stricturing or penetrating (including perianal) complications, or IBD-related hospitalization or surgery. This was assessed through detailed review of the medical record at both study sites. Study outcomes were assessed at two years and five years after endoscopic assessment.

Statistical Analysis

Continuous variables were compared using the t-test and summarized using means and standard deviations while categorical variables were compared using the chi-square test and summarized using proportions. Person-time was calculated from the time of endoscopic assessment to either the composite adverse outcome, end of follow-up or two or five years respectively, whichever was the earliest event. Univariate and multivariable Cox proportional hazards models were used to identify independent predictors of experiencing the composite adverse study outcome at two years and five years following the endoscopic assessment. The total SES-CD was our primary variable of interest and was modelled both as a continuous variable as well as categorical variable at different cut-offs with an SES-CD of 0 as the reference. We also specifically examined the performance of the SES-CD cut-offs previously used in clinical trials to define mild disease activity. Pre-planned subgroup analysis included stratification by prior immunomodulator or biologic use and by disease location. A two-sided p-value < 0.05 indicated independent statistical significance. The study was approved by the institutional review boards at each hospital.

RESULTS

Study Population

The final study cohort included 177 CD patients, with no prior CD-related surgery, and with a prospectively scored colonoscopy while on no immunomodulator, biologic, or small molecule treatment. The mean age of the study population was 51 years (range 23 – 84 years) with a median disease duration at the time of endoscopy of 17 years (range 0 – 55 years) (Table 1). 42% were women, and 90% were White. None of the patients were on conventional immunomodulators, biologics, or small molecules at the time of the endoscopic assessment. The mean total SES-CD for the cohort was 3.6 (range 0 – 34) with a mean ileum and colon SES-CD of 1.13 and 2.6 respectively. One hundred and twenty patients (68%) had a SES-CD ≤ 3, consistent with current definitions of endoscopic remission,³ while 57 (32%) had SES-CD consistent with endoscopic activity. Only one-quarter of the cohort had previously been on an immunomodulator (26%) while 11% had previously been on a biologic agent. The most common disease locations were colonic (38%) or ileocolonic (31%). Only 5% of patients were noted to have penetrating (B3) disease phenotype and 10% had prior history of perianal disease. The indication for colonoscopy varied and included both the presence of symptoms as well as surveillance for ongoing disease activity in asymptomatic individuals.

Table 1:

Characteristics of patients with Crohn’s disease included in the study

Baseline Characteristics	N=177 %
Age (years), mean (range)	51 (23–84)
Female gender, N (%)	42%
White race, N (%)	90%
Mean disease duration (years), (range)	17 (0–55 years)
Disease location (%)
Ileal	30%
Ileo-colonic	31%
Colonic	27%
Upper GI	1%
Disease phenotype (%)
B1 - inflammatory	81%
B2 - stricturing	14%
B3 - penetrating	4%
Perianal disease, N (%)	10%
Prior therapies, N (%)
Immunomodulators	26%
Biologics	11%
SES-CD, N (%)
0	34%
3-Jan	33%
6-Apr	16%
≥ 7	17%
SES-CD, mean (range)	3.8 (0–34)
SES-CD ileum, mean†	1.13
SES-CD colon, mean†	2.6

^†Data only available for Mass General Hospital and Mount Sinai Hospital.

Predictors of Disease Progression

Over two- and five-years of follow-up, 23% and 35% of patients, respectively, met our primary composite outcome of requiring steroids, initiation of immunomodulator or biologic therapy, penetrating or stricturing complications or IBD-related hospitalization or surgery (Figure 1). Among those meeting definition of disease progression, the most frequent outcomes were initiation of systemic steroids (16%), biologic therapy (10%), or immunomodulators (6%). New stricturing (1%), penetrating (0.5%), perianal complications (0.5%) or IBD-related surgery (0.5%) were infrequent. The mean time to use of steroids, immunomodulators or biologics was 313, 317 and 238 days, respectively. At enrolment, 61 patients (34%) had a SES-CD of 0, 59 had an SES-CD of 1–3 (33%), while 28 (16%) had a score of 4–6, and 29 (16%) had a score ≥ 7. Over the two-year follow-up, fewer than one-fifth of patients with SES-CD of 0 had disease progression (18%) compared to 45% of those with SES-CD ≥ 7 (p=0.002). By five years, 62% of patients with SES-CD ≥ 7 had progressed compared to 25% of patients with SES-CD score of 0.

Figure 1:

Overall likelihood of disease progression over 2 years following endoscopic assessment in patients with Crohn’s disease

Compared to those with an SES-CD score of 0, only patients with an SES-CD ≥ 7 had a greater rate of progression at two years (HR 2.50, 95% CI 1.09 – 5.72) while those with scores of 1–3 (HR 1.17, 95% CI 0.49 – 2.76) and 4–6 (HR 1.64, 95% CI 0.58 – 4.59) were not significantly more likely to progress (Figure 2). Similarly, only a SES-CD ≥ 7 was predictive of disease progression at five years (HR 2.89, 95% CI 1.41 – 5.91) while SES-CD of 1–3 (HR 1.42, 95% CI 0.70 – 2.88) and 4–6 (HR 1.57, 95% CI 0.65 – 3.81) were not associated with disease progression. Endoscopic severity by disease location was available for two study sites (MGH and Mount Sinai Hospital). The total SES-CD was numerically more predictive of disease progression in those with ileal or ileocolonic disease (n=86) (HR for SES-CD ≥ 7: 7.70, 95% CI 2.52 – 23.53) than in those with colon only CD (n=60) (HR for SES-CD ≥ 7: 2.16, 95% CI 0.48 – 9.72). Given potential inter-observer variability in SES-CD, we examined the robustness of our findings of the prognostic value of endoscopic severity by varying the lower cut-offs of the highest strata to SES-CD scores of 6 or 8 (α 1 of the primary cut-off). These demonstrated that the highest strata of endoscopic severity remained associated with a higher rate of disease progression (SES-CD ≥ 6 vs. 0: HR 2.65, 95% CI 1.23 – 5.71; SES-CD ≥ 8 vs. 0: 3.70, 95% CI 1.63 – 8.40) while the two intermediate strata continued to show no association with disease progression.

Figure 2:

Likelihood of disease progression over 2 years following endoscopic assessment in patients with Crohn’s disease, stratified by endoscopic activity Log-rank p-value= 0.012

We next examined other clinical features, in addition to SES-CD, that were associated with disease progression. On univariate analysis, prior immunomodulator (HR 3.96, 95% CI 2.15 – 7.28), prior biologic (HR 2.95, 95% CI 1.49 – 5.89), and total SES-CD (for each 1-point increase in score: HR 1.06, 95% CI 1.02 – 1.10) were predictive of disease progression by two years (Table 2). In addition, ileal (p=0.012) but not colonic (p=0.13) segmental sub-scores were predictive of progression. Apart from endoscopic severity, only prior immunomodulator use was independently predictive of disease progression at two years in a multivariable model (HR 3.94, 95% CI 1.90 – 8.21) (Table 3). For progression at five years, similarly, prior immunomodulator use was the only significant independent predictor of progression (HR 3.88, 95%CI 2.10 – 7.19).

Table 2:

Univariate analysis of predictors of disease progression at two years among patients with Crohn’s disease

Characteristic	Hazard ratio (HR)	95% confidence interval (CI)
Sex – Female	1.72	0.93 – 3.16
Disease duration (in years)	1.00	0.98 – 1.02
Age at diagnosis (in years)	0.97	0.95 – 1.00
Prior immunomodulator use	3.96	2.15 – 7.28
Prior biologic use	2.95	1.49 – 5.89
Perianal disease	1.96	0.82 – 4.65
CD location
Ileum (L1)	Ref
Colon only (L2)	1.48	0.66 – 3.31
Ileocolonic (L3)	2.07	0.92 – 4.70
CD – Behavior
Non-penetrating, non-stricturing (B1)	Ref
Stricturing (B2)	1.28	0.54 – 3.06
Penetrating (B3)	2.00	0.61 – 6.54
Endoscopic Severity (Total SES-CD score)
SES-CD 0	Ref
SES-CD 1–3	1.10	0.48 – 2.48
SES-CD 4–6	1.36	0.50 – 3.67
SES-CD ≥ 7	3.12	1.39 – 6.97
Total SES-CD score	1.06	1.02 – 1.10

Table 3:

Multivariable analysis of predictors of disease progression at two years among patients with Crohn’s disease

Characteristic	Hazard ratio (HR)	95% confidence interval (CI)
Disease duration (in years)	0.98	0.95 – 1.01
Prior immunomodulator use	3.94	1.90 – 8.21
Prior biologic use	0.98	0.43 – 2.26
Total SES-CD score
SES-CD 0
SES-CD 1–3	1.21	0.50 – 2.91
SES-CD 4–6	1.94	0.70 – 5.39
SES-CD ≥ 7	2.49	1.10 – 5.63

Restricting the analysis to patients who were immunosuppression naïve (never been on immunomodulator or biologics prior) (n=129) revealed similarly higher rates of progression only in those with SES-CD ≥ 7 (HR 5.65, 95% CI 1.49 – 21.52) (Supplemental Figure 1). Only 9% of immunosuppression naïve patients who had a SES-CD of 0 experienced the composite outcome by two years compared to 33% of those with score of 7 or higher. A sensitivity analysis excluding 8 patients with prior penetrating disease revealed similar association between SES-CD ≥ 7 and disease progression (HR 2.32, 95% CI 1.00 – 5.45). In a sensitivity analysis stratifying by duration of disease, an SES-CD score ≥ 7 remained similarly predictive of disease progression in those with > 5 years (HR 3.06, 95% CI 1.34 – 7.00) or > 15 years of disease course (HR 4.18, 95% CI 1.48–11.78).

DISCUSSION

Existing treatment guidelines for CD are influenced in large part by data derived from patients with moderate-to-severe disease activity and all recent randomized trials of treatment (and treatment strategies) have focused on this population. It is important to derive evidence-based criteria for mild CD to facilitate studies of interventions and management strategies in this cohort. From a multi-centre international cohort of patients who underwent prospectively scored endoscopic assessment, we observed that an SES-CD ≥ 7 is predictive of disease progression compared to those with SES-CD score of 0. This finding was similar in patients with and without a history of prior immunomodulator or biologic exposure and was still significant after adjusting for potential confounders and other factors associated with progression of CD. Our findings suggest that an SES-CD ≤ 6 may be used to identify mild CD patients who are less likely to progress over time and may inform clinical decisions on appropriateness of escalation to advanced therapies.

In our cohort of patients with mild CD, over a quarter experienced disease progression within two years, increasing to 35% after five years. Thus, 65% of patients remained mild at five years after endoscopic assessment. These findings are comparable to those from prior observational studies. In two European cohorts, 56% and 46% of mild CD patients maintained their mild status after five and 15 years from diagnosis, respectively.^5,17 Notably, these cohorts based their disease severity assessment mainly on clinical symptoms, presence of complications and need for surgery, without accounting for endoscopic scoring. Additionally, our study population differed as it included patients with previous use of immunosuppressants. Another observational study involving 145 newly diagnosed CD patients found that 57% exhibited mild CD features (no complicated phenotypes, perianal disease, or need for biologic therapy or surgery) after a mean of 96 months of follow-up. Interestingly, in this study, patients with previous and current immunomodulator use were not viewed as having an unfavourable outcome, thus they were still categorized as having mild CD.¹⁸

International societies have typically relied on clinical symptoms to define mild CD. For instance, the European Crohn’s and Colitis Organization classifies mild CD as having a low risk for developing disabling disease within five years based on no perianal disease, ileocolonic location, young age at diagnosis, or use of systemic steroids.⁶ The American Gastroenterological Association uses a CDAI score between 150 and 220 to define mild CD.⁷ The American College of Gastroenterology considers mild clinical symptoms along with a SES-CD of 3 to 6 as indicative of mild disease.³ Although this cutoff is widely used in clinical trials, there is inconsistency in defining severity cutoffs and limited data on their long-term prognostic significance.¹⁹ Various SES-CD cutoffs have been suggested to define disease severity, with differing recommendations in the literature.¹¹ Most studies classify CD severity based on SES-CD scores: inactive (0–2), mild (3–6), moderate (7–15/16), and severe (>16).²⁰ Some variations exist, with studies classifying mild CD as scores of 4–9,²¹ 4–10,²² or mild to moderate as 4–14.²³

Previous research examining the correlation between endoscopic scoring and the long-term progression of the disease in patients with recently diagnosed moderate to severe CD revealed that CDEIS < 4 and SES-CD < 4 were associated with a low likelihood of disease progression following treatment with biologics.^24,25 Within our cohort of patients with mild CD, a SES-CD ≥ 7 was an independent predictor of disease progression over time. Notably, our cohort’s median disease duration was 17 years. So, our findings mostly apply to patients with long standing CD.

One advantage of SES-CD over CDEIS is its ability to assess ileal and colonic activity separately. Our data indicated that a SES-CD ≥7 predicted disease progression in ileal and ileocolonic disease but was less predictive in isolated colonic disease. These results align with population-based data demonstrating a higher risk of complications, such as stricturing or penetrating disease, in patients with terminal ileum or ileocolonic involvement compared to those with colonic disease alone.² Moreover, our findings are also consistent with the use of scoring systems that confers a higher weight to ileal involvement, such as Modified Multiplier Simple Endoscopic Score for Crohn’s Disease (MM-SES-CD).²⁵ Nonetheless, the prognostic value of this novel scoring system has only been assessed on post-hoc analysis²⁵ and retrospectives cohorts,²⁶ and further studies are needed to assess MM-SES-CD longitudinal predictive value. In addition to endoscopic scores, other attempts at defining disease severity in CD have used a composite of symptoms, biomarkers, and endoscopic features. ^27,28 These have demonstrated moderate longitudinal prognostic value though their performance has typically been studied in a heterogeneous cohort of patients of varying severity including a large fraction of current users of advanced therapies or prior surgery.

Our study benefits from several strengths, including its multicentre design with longitudinal follow-up of up to five years, allowing for a comprehensive assessment of disease progression over time. Colonoscopies were systematically and prospectively scored by gastroenterologists with expertise in IBD, ensuring consistency and accuracy in the evaluation of disease activity. This scoring was also independent of outcome ascertainment. While inter-observer correlation could not be examined and there was no central read, there was no difference in the mean SES-CD scores across the three sites, suggesting no systematic bias in scoring. Importantly, our findings provide an evidence-base for the current guidelines that classify mild disease activity as corresponding to a SES-CD between 3 and 6.³ Furthermore, our cohort comprised patients with well-established CD, with a mean disease duration close to 20 years, with one of the cohorts including patients since CD diagnosis. Our study suggests that in patients with established disease who have only mild endoscopic activity while not actively being treated with immunosuppression, the likelihood of disease progression is low. This suggests a need to examine alternate management strategies in this population, comparing the benefit of systemic immunosuppression to either active monitoring or safer non-immunosuppressive interventions such as diet, microbiome-based or gut-barrier directed treatments. In addition to endoscopic severity, a more comprehensive definition of mild CD may also need to include prognostic factors, clinical parameters, biomarkers such as faecal calprotectin and cross-sectional assessment including with magnetic resonance imaging, computed tomography scan, or intestinal ultrasound. It is important to note that our findings should not be inferred to suggest that partial endoscopic improvement is sufficient in patients with CD who are receiving immunosuppressive treatment. The treatment goal for patients with moderate-to-severe CD initiating advanced therapy remains an SES-CD score < 3 as suggested by professional societies and consensus guidelines.^3,6,7

Our study also has several limitations that warrant consideration. Firstly, its retrospective nature introduces inherent biases and limitations associated with data collection and interpretation. Secondly, there was some heterogeneity in the study population, particularly the small proportion of patients exposed to prior immunosuppressive therapies and relatively long disease duration. However, it is worth highlighting that none of them were receiving immunosuppressants at the time of endoscopic assessment and our findings remained significant after excluding those with prior immunosuppression use. Lastly, with two of our three sites being based at referral centers, our cohort likely represents a more complex patient population overall, potentially underestimating the prevalence of mild CD. However, it is worth mentioning that the observed rate of disease progression was consistent with previous research, indicating the generalizability of our findings.² There were also fewer patients with high endoscopic scores in our cohort; larger validation cohorts are an important next step.

Our findings contribute to the existing literature supporting the use of a SES-CD score cut-off of 1–6 for defining mild CD. Implementing this standardized definition in clinical practice and clinical trial protocols will improve the management of CD and enhance the development of novel therapeutic strategies for patients with milder disease.

Supplementary Material

Supinfo

Supplemental Figure 1: Likelihood of disease progression over 2 years following endoscopic assessment in immunosuppressant naïve patients with Crohn’s disease, stratified by endoscopic activity Log-rank p-value = 0.029