Introduction

Endoscopy in IBD is used to diagnose ulcerative colitis (UC) and Crohn’s disease (CD), prognosticate disease severity, obtain biopsies of intestinal mucosa for histological examination, monitor for dysplasia and risk of colorectal cancer and more recently to evaluate mucosal or endoscopic response to therapy. Historically, the aim of treatment for patients with inflammatory bowel disease (IBD) has been to induce and maintain clinical (symptomatic) remission. However, increasingly, there has been a paradigmatic shift in therapeutic approach with a push toward aiming for endoscopic remission or mucosal healing as a primary treatment goal. This is in recognition of the fact that treating to induce clinical remission is unreliable as IBD symptoms are subjective, with patients in clinical remission having significant endoscopic disease activity [1–5] and patients who feel unwell having no endoscopic findings of disease activity [3]. This puts a significant proportion of patients at risk of either disease progression due to inadequate treatment or at risk of over treatment with unnecessary medications if we rely on clinical symptoms alone to determine our anti-inflammatory treatment strategy. Furthermore, patients who do achieve endoscopic remission have improved outcomes compared to those who do not, with patients who achieve mucosal healing being found to have a decreased risk of clinical relapse, hospitalizations, surgery and colorectal neoplasia [1, 3, 6••, 7–14].

Therefore, in order to accurately assess disease activity and determine and quantify response to therapy, endoscopic assessment is required and is increasingly becoming the standard of care. This article will discuss the most common endoscopic scores used to measure endoscopic disease activity in IBD and their role in predicting the course of these diseases and their impact on decisions regarding medical strategies. We also provide a brief review of emerging non-invasive markers of mucosal healing that are increasingly being incorporated into the real world setting and discuss how the assessment of endoscopic disease activity should be incorporated into routine clinical practice.

Endoscopic Disease Activity Indices for Ulcerative Colitis

Endoscopic findings of mucosal inflammation in UC typically include erythema, edema with vascular congestion or loss of fine vascular pattern, and granularity [15]. As the disease progresses in severity, friability, spontaneous bleeding and macroscopic ulceration can also occur [15]. Truelove and Witts developed the initial mucosal scoring system for UC in 1955 [16]. They reported endoscopic lesions as normal or near normal, improved or no change/worse on sigmoidoscopy during a placebo-controlled trial of cortisone for the treatment of active disease. Since then, many endoscopic scoring systems for UC have been developed to measure endoscopic disease activity (Table 1).

Table 1 Endoscopic disease activity scoring systems in ulcerative colitis
Full size table

As there are many scoring systems available, this review will focus on the most common scoring system used in clinical trials of UC, the Mayo Clinic endoscopy sub-score, and the newest scoring system that is currently undergoing validation and is likely to be increasingly adopted in the future, the ulcerative colitis endoscopic index of severity (UCEIS).

The most common endoscopic score used in clinical trials to measure endoscopic disease activity in UC is the Mayo Clinic endoscopy sub-score [26]. This score has four components: erythema, friability, vascular pattern and erosions/ulceration with a maximum total score of 3 (Fig. 1). Although not formally validated, mucosal healing has generally been defined as a Mayo score of either 0 or 1 [29]. This definition has since been found to be fitting with a post hoc analysis of the active ulcerative colitis trials (ACT)-1 establishing that patients who achieve a post-treatment Mayo endoscopic score of 0 or 1 have equivalent rates of colectomy on follow-up and are significantly less likely to undergo colectomy over the subsequent year than those with higher Mayo endoscopic sub-scores [30]. However, of note, patients who achieved a score of 0 were found to have higher rates of steroid-free remission at 1 year compared to those who only achieved a score of 1 [30]. The strengths of the Mayo endoscopic sub-score lie in the frequency of its use in clinical trials and its ease of use. Its weakness lies in its lack of validation, the fact that it does not distinguish between deep and superficial ulceration [31•] and that the score only reflects the most severely affected segment of the bowel visualized without giving any indication of the extent or distribution of mucosal inflammation and setting no minimal insertion length. In addition, the original score includes variable degrees of friability in the score of 1 and 2, which results in high inter-observer discrepancy and inconsistent results [32]. In fact, because of this concern, some studies have used a modified Mayo scoring system that classifies the presence of any degree of friability as an automatic Mayo endoscopic sub-score of 2 [33–35]. This modified Mayo Clinic Endoscopic Score (MMCS) has been found on initial review to have excellent intra-observer and inter-observer reliability (intra and inter-class correlation coefficient and 95 % confident interval [95 % CI] 0.89 [0.85–0.92] and 0.79 [0.72–0.95], respectively) and is responsive to change [36].

Fig. 1
figure 1

Mayo endoscopic sub-score

Full size image

Due to the need for a prospectively validated endoscopic assessment tool that can assess mucosal healing in UC and be applied to clinical practice, the ulcerative colitis endoscopic index of severity (UCEIS) [37] and the ulcerative colitis colonoscopic index of severity (UCCIS) [27] have recently been developed and undergone initial validation. The UCEIS (Table 2) is the most cited of these tools and was prospectively developed using multiple validated steps with the final tool evaluating vascular pattern, bleeding and erosions and ulcers with the worst segment of the colon scored for each variable on a 0–2 or 0–3 scale giving a total score of 0–8 [37]. The final scoring system is easy to use and has a high intra and inter-observer agreement with an intra-observer kappa value of 0.82, 0.72 and 0.78 and inter-observer kappa values of 0.83, 0.56 and 0.77, respectively, for three main descriptor domains of vascular pattern, bleeding and erosion/ulcers [28•]. The correlation coefficient (r 2) between the UCEIS and overall severity evaluation was 0.94 (p < 0.0001), meaning it accounts for 88 % of the variance in overall assessment of severity between observers [28•]. The main limitation of this score currently is that there is still no threshold set for remission, mild, moderate and severe disease although these are anticipated in the near future. A preliminary study has shown that in patients admitted with acute severe colitis, a score of 7 or 8 out of 8 at the time of admission predicts inadequate response to intravenous steroids and need for rescue therapy with cyclosporine or infliximab [38]. This scoring system is currently being adopted in clinical trials and will likely be adapted for clinical practice in the future.

Table 2 The ulcerative colitis endoscopic index of severity (UCEIS) [28•]
Full size table

Endoscopic Disease Activity Indices for Crohn’s Disease

Endoscopic findings in CD consist of edema, erythema, apthoid ulceration, cobblestone appearance and strictures [15]. There are currently three major endoscopic indices for evaluating CD disease activity (Table 3). The two validated endoscopic activity scores for CD are the Crohn’s disease endoscopic index of severity (CDEIS) [39] and the simple endoscopic score for Crohn’s disease (SES-CD) [40]. Both tools have been prospectively validated and shown to be reproducible and have good inter-observer agreement [42–44].

Table 3 Crohn’s disease endoscopic disease activity scoring systems
Full size table

The CDEIS is often considered the gold standard for classifying endoscopic disease activity in CD. The endoscopic parameters of (1) presence or absence of ulcers, distinguished as superficial or deep, (2) percentage of surface ulcerated and/or affected, and (3) presence of stenosis, classified as ulcerated or non-ulcerated stenosis in the five bowel segments (terminal ileum, right colon, transverse colon and sigmoid, and rectum) are evaluated to give a total score of 0–44 [39]. It has good correlation with the Crohn’s disease activity index (CDAI), is highly reproducible and is sensitive to changes in endoscopic mucosal appearance and healing [30]. The CDEIS is the most commonly used endoscopic tool to assess disease activity in clinical trials although there is no agreement or formal validation regarding cut-off values for defining endoscopic response to treatment, endoscopic remission or mucosal healing and no data available on long-term clinical outcomes. In the available trials, endoscopic response has previously been defined as a decrease from the baseline score of at least 3 or 5 points [43, 45] although more recently, a post hoc analysis of the SONIC trial by Ferrante et al. [46], showed that defining endoscopic response as a decrease from baseline of the CDEIS score of at least 50 %, was most predictive of corticosteroid-free clinical remission by week 50, including that of a decrease in score of 3 or 5 points. In trials utilizing the CDEIS, endoscopic remission has been defined as “partial” using a cut-off of <6 [40, 45, 47], and “complete” using a cut-off of <3 [40, 45], <4 [48], ≤4 [49] or 0 [50]. The main limitation of the CDEIS is the fact that it is a complex tool that requires training and experience to utilize, resulting in a 2002 expert consensus statement that the CDEIS should be reserved for use in clinical trials only due to its complexity [51].

To overcome these limitations, a simplified index, the simple endoscopic score for CD (SES-CD) was developed. The SES-CD is reliable and correlates well with the CDEIS (correlation coefficient r 2 = 0.920) [40]. The endoscopic parameters of (1) ulcer size, (2) ulcerated and affected surfaces, and (3) stenosis are scored from 0 to 3 in each of the five bowel segments (terminal ileum, right colon, transverse colon and sigmoid, and rectum) to give a total score of 0–60 [40]. However, despite it being much simpler than the CDEIS, the SES-CD is still a complex index with limited use in clinical practice. In addition, as with the CDEIS, there is a lack of consensus on the definition of endoscopic response and remission. In previous clinical trials, a SES-CD score of <3 [49, 52–54] or equal to 0 [13, 50, 55–57] has been used to define endoscopic remission or minimal endoscopic activity and, more recently, Moskovitz et al. [58] validated the cut-off values for the SES-CD as 0–2 for endoscopic remission, 3–6 for mild endoscopic disease, 7–15 for moderate endoscopic disease activity and ≥16 for severe endoscopic disease activity. In regard to defining endoscopic response to treatment, as with the CDEIS, Ferrante et al. [46] demonstrated that a decrease from baseline of the SES-CD score of at least 50 % was most predictive of improved outcomes. With this evidence in mind, the International Organization for the Study of Inflammatory Bowel Disease is preparing an expert opinion publication stating that endoscopic response to therapy should be defined as a >50 % decrease in the SES-CD and that remission should be defined as an SES-CD of 0–2 [59].

The final endoscopic activity scoring system commonly used in CD is the Rutgeert’s score [41] (Fig. 2). The Rutgeert’s score assesses and quantifies endoscopic disease recurrence in the neo-terminal ileum after ileal or ileocolonic resection [41, 60] and is the most commonly used tool used to assess recurrence in postoperative CD trials. The numerical score ranges from 0 to 4; (0) normal mucosa; (1) <5 apthous lesions; (2) >5 apthous ulcers with normal intervening tissue; (3) diffuse inflammation with diffuse ulcers; (4) nodules and/or narrowing. Although it has not been fully prospectively validated, the severity of the Rutgeert’s score on endoscopy in an asymptomatic patient within 12 months of the ileocolonic resection has been shown to predict the risk of clinical recurrence with Rutgeert’s score of grade 0 or 1 being associated with a very low risk of clinical recurrence (80–85 % asymptomatic at 3 years follow-up) compared to those who have a score of 3 or 4 (<10 % asymptomatic at 3 years follow-up) [41]. Therefore, ileocolonoscopy is recommended within 1 year following surgical resection to determine if postoperative treatment is effective or if additional treatment is required.

Fig. 2
figure 2

Rutgeerts’ score for postoperative endoscopic recurrence

Full size image

Endoscopic Assessment Can Predict Disease Severity

Endoscopic severity may predict the future clinical course of IBD. In both UC and CD, severe endoscopic lesions predict an increased risk of colectomy. In CD, severe endoscopic ulceration increases the risk of colectomy to 31 % from a baseline of 6 % at 12 months in those without severe endoscopic lesions [61] and in UC, the odds ratio of colectomy when a patient is admitted for a severe attack is 41 in those with severe lesions on endoscopy compared to those without severe lesions [62]. In addition, it has been shown that only 34 % of patients who respond to medical therapy in severe colitis have severe endoscopic lesions compared to 91 % in those who do not respond to medical therapy (OR >20) [40].

Importance of Achieving Mucosal Healing in IBD

In recent years, mucosal healing has increasingly emerged as a major aim of therapeutic interventions in IBD. This is secondary to the growing evidence that demonstrates improved clinical outcomes in those achieving mucosal healing compared to those who do not.

Improved clinical outcomes in patients who achieve mucosal healing compared to clinical remission alone was first reported back in 1966 by Wright et al. [63] who found that UC patients not achieving mucosal healing when treated with steroids relapsed more frequently during a follow-up period of 1 year compared to patients who did (40 vs. 18 %, respectively). Since then, a plethora of studies have confirmed this finding and demonstrated that in both UC and CD, mucosal healing is associated with prolonged remission, fewer hospitalizations and surgical procedures, less bowel damage (fistulas) in CD, less immunosuppression therapy, a lower risk of colorectal cancer, and improved quality of life [1, 3, 6••, 7–14, 30, 62, 64–66].

Recently, it has also been demonstrated that the severity and chronicity of inflammation in the colon is associated with the risk of colorectal neoplasia [67–70]. The degree of endoscopic and histologic inflammation has been found to correlate with the risk of developing colorectal neoplasia on univariate analysis with more severe disease being associated with higher cancer risk [67, 68]. Despite the fact that on multivariate analysis only histological inflammation was an independent predictor of risk, a follow-up study of colorectal surveillance did find that UC patients who have mucosal healing or a macroscopically normal colon have a colorectal cancer risk similar to that of the general population on 5-year follow-up [68].

Endoscopic Assessment in Clinical Practice

Clinical disease activity is subjective and not a reliable indicator of endoscopic disease activity. It has been found that up to half of patients who are in clinical remission will still have endoscopic evidence of active disease [71]. In addition, a high prevalence of clinical symptoms has been noted in patients who actually have achieved mucosal healing [3, 72]. This leads to a situation in which patients may be either under- or over-treated in relation to their symptoms and disease activity if endoscopic assessment does not occur, and is the reason that assessment of endoscopic disease activity is increasingly being applied to treatment algorithms.

When to Look

In regard to timing of endoscopic assessment, due to the prognostic value of endoscopy in regard to long-term outcomes, patients who have a significant increase in clinical symptoms or are first presenting with symptoms should undergo a baseline endoscopy. This allows an appropriate treatment plan to be initiated that is titrated to the patient’s disease severity. Once therapy has commenced, it is now increasingly accepted that a follow-up colonoscopy should occur to assess for mucosal healing or endoscopic response to therapy. The timing of this is still controversial but should likely occur between 3 and 6 months (earlier if the faster acting anti-TNF therapies are utilized and later if the slower acting anti-metabolite or anti-integrin medications are used).

How to Document Endoscopic Activity and Mucosal Healing

The routine use of endoscopic scoring systems is currently limited to trial settings. The reason for this is secondary to the fact that currently there is no one accepted tool that has been standardized for this setting in either CD or UC, often the scoring systems are too complex and time-consuming to be used in clinical practice and many suffer from high inter-observer variability. In addition, the existing scoring systems do not have well-defined and validated thresholds for mucosal response or healing and there is no consensus on degree of mucosal healing that is required to limit future disability or change the natural history of the disease.

However, despite their limitations, the use of an endoscopic scoring system can aid in the reporting of endoscopic findings and allow easy comparison between a patient’s current and previous colonoscopy result. If it is feasible, we recommend the use of the Mayo sub-score for UC and the SES-CD score for CD. However, in clinical practice, generally documentation of endoscopic disease activity remains subjective. If the endoscopic scoring systems are not used, it is important to report in each segment of the bowel on the following:

  • The extent and location of inflammation

  • If the bowel involvement is continuous or involves skip areas

  • The presence of erythema, loss of vascular pattern, bleeding (contact or spontaneous), presence of erosions or ulceration (superficial or deep) and the presence of strictures or fistulas.

We also recommend specific language in the impressions to distinguish “clinical remission” from “endoscopic remission” from other end points (like histologic remission). Such examples include the following:

  • IMPRESSION: Endoscopically moderately active left-sided ulcerative colitis.

  • IMPRESSION: Endoscopically quiescent panulcerative colitis and clinical remission (deep remission).

  • IMPRESSION: Endoscopically mildly active patchy Crohn’s ileitis and proximal colitis.

In addition, on follow-up, ileocolonoscopy is important to note if the endoscopic disease activity has improved, worsened or is stable.

Although not yet suitable for adoption in the clinical setting, currently newer endoscopic scoring systems are being developed and future studies are likely to validate these scoring systems in the clinical setting and demonstrate their role in the day-to-day management of IBD patients. This will help with the comparison between drug efficacies and optimize a treat-to-target treatment algorithm in our patients.

How to Achieve Mucosal Healing

If a patient is symptomatic and has not achieved mucosal healing, then escalation of medical therapies should occur. If a patient who is in clinical remission is found to have unexpected mucosal inflammation, an open dialogue should occur about the goals of treatment. Symptom control and the side effects of therapy should be acknowledged but a discussion about the risks of uncontrolled inflammation and resulting progressive disease should also occur and short and long-term goals recognized. It is now thought that achieving mucosal healing will improve the long-term outcomes of inducing sustained clinical remission and reducing hospitalizations and surgery in patients with IBD and reduce or prevent progressive disease and disability. Therefore, adopting a “treat-to-target” approach is increasingly being accepted with the target being that of mucosal healing. After discussing the pros and cons of escalation of therapy with the patient, techniques to treat to mucosal healing include confirming adherence to medication and overcoming barriers to adherence, optimization of current medical therapies including assessment of medication metabolites or therapeutic monitoring of anti-TNF therapy and adjusting therapy as needed and if required consideration given to switching therapy to another drug within the same class or outside the class depending on the clinical context [73].

Preliminary retrospective data suggest that repeated assessment of endoscopic disease activity with adjustment of medical therapy to the target of mucosal healing is feasible in clinical practice and seems to be of benefit [74••, 75••]. However, although it is thought that mucosal healing will improve long-term outcomes, there are still many unresolved challenges in regard to incorporating endoscopic assessment and the target of mucosal healing into routine clinical practice (Table 4). It is still unclear just how much healing is required and it is yet to be demonstrated prospectively that mucosal healing can prevent disease progression or change the natural history of IBD [73]. Therefore, before any medication adjustment takes place, the risks of medical escalation must be weighed against the benefits of achieving mucosal healing as this escalation of therapy is likely to increase the associated risks of the medication [73, 76••]. For targets of healing, a recent expert statement from the International Organization for the study of Inflammatory Bowel Diseases (IOIBD) has recommended selecting a Mayo endoscopic sub-score of 0–1 to define endoscopic remission in UC and the resolution of ulceration at ileocolonoscopy in CD [77••].

Table 4 Unresolved challenges to the incorporation of routine endoscopic assessment and mucosal healing in IBD management
Full size table

Conclusion

Ileocolonoscopy is now considered the gold standard to assess disease severity, prognosticate a patient’s future disease course and quantify mucosal response and healing following treatment in inflammatory bowel disease and is more reliable in determining disease activity than relying on clinical symptoms alone. Numerous endoscopic scoring systems exist however most are limited due to their complexity and a lack of formal validation. In addition, there is currently limited consensus on the value or percentage improvement in these scores that should be used to define mucosal improvement and healing and there is limited data on how these scores can be utilized to predict long term improved clinical outcomes and therapeutic management strategies in regard to continuing or stopping therapy or changing the type of therapy completely. Despite these limitations, the assessment of disease activity and mucosal healing by endoscopy is increasingly becoming standard of care and should now be routinely implemented into clinical practice as part of a treat-to-target strategy.