Managing Ulcerative Colitis and Crohn’s Disease: Should the Target Be Endoscopy, Histology, or Both?

Abstract In inflammatory bowel disease (IBD), mucosal healing is the primary long-term treatment goal, encompassing both endoscopic and histological outcomes. This paper aims to overview the ability of new treatment options to promote endoscopic and histological healing and to discuss the prognostic significance of endoscopic and histological outcomes. The analysis included randomized-controlled trials (published since 2020) focused on the impact of pharmacological interventions on endoscopic and histological remission in IBD. Even though the Mayo endoscopic subscore is routinely used, the application of validated scoring systems for ulcerative colitis is uncommon. In Crohn’s disease (CD), the application of endoscopic scores remains limited to clinical studies. The standardized evaluation of histological features has been performed in several recent ulcerative colitis trials, resorting mostly to the Geboes score and the Nancy histological index. Still, the use of histological scores for CD remains elusive. Current evidence underscores that histological remission conveys the best long-term prognosis, supporting the inclusion of histology as a treatment guide in ulcerative colitis. In CD, data are promising but originated from a few retrospective studies. Further efforts are warranted to: (1) use validated histological indexes for ulcerative colitis, aiming their adoption as treatment targets; (2) promote the validation and utilization of histological scores for CD, at least in clinical studies; (3) confirm the prognostic impact of histological remission in CD; (4) integrate artificial intelligence assets to support grading, particularly in the setting of histology; (5) prospectively define the monitoring frequency of IBD patients who achieved histological remission.


Introduction
Inflammatory bowel diseases (IBD), comprising Crohn's disease (CD) and ulcerative colitis (UC), are chronic conditions characterized by inflammation of the gastrointestinal tract. 1,2reatment targets for IBD have shifted over time from symptom control to achieving objective evidence of mucosal healing (MH), which encompasses both endoscopic and histological healing. 3,4In the setting of UC, MH has been associated with favourable long-term outcomes, including prolonged periods of corticosteroid-free remission and reduced risk of complications such as hospitalization, surgery, and colorectal cancer. 2n 2013, the International Organization for the Study of IBD established the STRIDE program 5 -Selecting Therapeutic Targets in IBD-to assist clinicians in setting appropriate treatment goals for different stages of IBD; the original STRIDE statements were revised in 2021 to include treatment targets in both adult and pediatric IBD.6 The STRIDE program recommends a treat-to-target (T2T) approach based on evidence and expert consensus, including clinical indexes, biochemical biomarkers, endoscopic healing, and histological healing in the case of UC.All these goals are viewed as complementary and are integrated into the concept of disease clearance.4 In IBD, achieving MH is the primary long-term treatment goal.In this setting, endoscopic remission assessed through endoscopic indexes has proved to be a key target associated with improved long-term outcomes. 2However, a significant proportion of patients who achieve endoscopic healing may still present microscopic disease.7,8 In UC patients, histological activity, even in the presence of endoscopic remission, is associated with higher relapse rates, prolonged corticosteroid use, and long-term complications.9 Therefore, histological assessment using histological indexes has been proposed as a more accurate measure of disease activity.Recent international consensus has recognized that histological remission is an appropriate endpoint in both UC and CD clinical trials.10,11 However, histological healing has not yet been formally endorsed as a treatment goal in clinical practice, partly due to the existence of multiple unvalidated histological scoring systems.6 Furthermore, we are witnessing a growing interest in the use of molecular targets to evaluate IBD, including markers of inflammation, immune activation, and mucosal damage, which have shown utility in predicting disease activity and treatment response.6 The aim of this review was to discuss the evaluation of endoscopic and histological activity in UC and CD and to summarize recent evidence regarding the effect of new treatment options on those targets.It was also intended to overview the relationship between endoscopic and histological endpoints and IBD prognosis and to explore the unmet needs in this field.

Search strategy
A narrative review was conducted to collect the endoscopic and histologic scores used for grading IBD.To assess the effectiveness of recent medication in achieving endoscopic response and remission, we conducted a search on CENTRAL and PubMed for all randomized-controlled trials (RCTs) published since 2020, focusing on the impact of pharmacological interventions on endoscopic response and remission in patients with UC and CD; the following terms were applied: ((endoscopic score) or (endoscopic response) or (endoscopic remission)) and ((UC) or (CD)) and (RCT).

Are endoscopic response and remission achievable in UC?
Several studies have demonstrated that current ulcerative colitis (UC) medical therapies have the capability to induce favourable endoscopic outcomes. 12,13][16][17][18][19][20][21][22][23][24][25][26][27] Endoscopic remission after induction was achieved between 2.6 percent (after 12 weeks of mirikizumab 18 ) and 22.0 percent (after 12 weeks of olamkicept 25 ) of the patients.On the other hand, from 14.6 percent (after 58 weeks of filgotinib 16 ) to 30.6 percent (etrolizumab for 62 weeks 23 ) were in endoscopic remission during maintenance.Endoscopic response was achieved in 13.1 percent (for the oral α 4 β 7 antagonist PTG-100 20 ) to 60.7 percent (for apremilast 27 ) of patients after induction treatment.Regarding maintenance, improvement of endoscopic lesions has been reported to occur between 27.1 percent (with etrolizumab 26 ) and 55.3 percent (with upadacitinib 14 ) of the patients.All studied drugs demonstrated a positive effect on these outcomes when compared with the placebo, even though the differences were modest in some studies.The Mayo Endoscopic subscore (MES) was the most applied index, usually as a secondary outcome.MES is an easy-to-use tool, which assesses inflammation based on a 4-point scale (0-3) 28,29 ; MH is defined as an MES of 0 or 1, depending on the studies.However, MES presents some limitations: it lacks formal validation, has a low sensitivity to change over time, and does not provide an accurate distinction between moderate and mild friability, which leaves room for subjectivity and interobserver variability. 28

Do endoscopic outcomes impact UC prognosis?
In a recent meta-analysis of seventeen studies (five of whom prospective) that included over 2,500 UC patients, MES 0 patients had a 52 percent lower risk of clinical relapse compared with those with MES 1 (29 percent versus 14 percent). 30The predictive ability of endoscopic outcomes has been also demonstrated for the ulcerative colitis endoscopic index of severity (UCEIS), a score that has already been prospectively validated. 31,32Indeed, UCEIS has demonstrated a capability of predicting the need for colectomy 33 patients with acute severe UC with a UCEIS of 3 had a colectomy rate of 0 percent, while those with a UCEIS of at least 7 had a rate of 80 percent, after a median follow-up of eighteen months.UCEIS has also been reported to predict the response to biological treatment. 34More recently, a new index, the Paddington international virtual chromoendoscopy score in ulcerative colitis (PICaSSO), was prospectively validated. 35This index was able to accurately predict histological healing and long-term remission.Indeed, patients with a PICaSSO score ≤ 3 were less prone to UC clinical relapse than those with PICaSSO > 3 (hazard ratio ~0.20 at 6 and 12 months after colonoscopy). 36

Are endoscopic response and remission achievable in Crohn's disease?
The regression or disappearance of lesions in ileocolonoscopy of Crohn's disease (CD) patients has been regarded as endoscopic remission or MH and is, now, considered a therapeutic target.Over time, several scoring systems have been developed to increase the objectivity of endoscopic evaluation.The simple endoscopic score for CD (SES-CD) was introduced in 2004.A relevant drawback of SES-CD is the lack of clarity regarding its validation, responsiveness, and reliability in predicting outcomes in CD.Indeed, its thresholds are mostly arbitrary, corresponding endoscopic response, usually, to a reduction of at least 50 percent in baseline SES-CD, and remission to a score of up to 2 or 3 points.Since 2020, nine RCTs [37][38][39][40][41][42][43][44][45] on the effect of CD therapies on endoscopic outcomes was published; all the studies have used the SES-CD index (Table 1).Even though the interventions were diverse (including drugs targeting interleukins, [38][39][40][41] Janus-kinase inhibitors, 37 anti-tumor necrosis factor, 43 or integrins 45 ), it seems clear that currently approved drugs and those on the pipeline allow reaching endoscopic outcomes in a significant proportion of patients with moderate to severe CD; the rates of endoscopic response and remission were between 30 percent and 50 percent and 15 percent and 20 percent, respectively, after induction therapy.The achievement of endoscopic response and remission during maintenance was recently evaluated in four studies. 39,40,43,45The percentages of response were between 24.1 percent (after 52 weeks of etrolizumab 45 ) and 47.0 percent (52 weeks of risankizumab 39 ), while those if endoscopic remission ranged between 14.5 percent (for ustecinumab 40 ) and 34.2 percent (for risankizumab 39 ).Over the last decade, differences between ileal and colonic Crohn's disease in terms of therapeutic responses have been highlighted.A recent post-hoc analysis of an RCT of patients who received induction and maintenance therapy with anti-TNF agents found that endoscopic healing (evaluated using modified SES-CD) of the small bowel occurred in 36 percent, while 79 percent of the patients demonstrated endoscopic healing in the colon. 46Interestingly, all patients with small bowel healing had concurrent colonic endoscopic healing.This information was in line with the described by the posthoc analyses of the SONIC and TAILORIX trials, where it was also highlighted that the degree of endoscopic inflammation at baseline did not affect the likelihood of achieving endoscopic remission. 47,48Of the RCTs published since 2020, only one (comparing standard-versus high-dose adalimumab 43 ) provided data on the achievement of endoscopic outcomes depending on the location.In this study, the attainment of  43 Also, the effect of the high-versus standard-dose therapy was larger for patients with ileal versus colonic disease (delta of 19.0 vs. 6.1). 43

Do endoscopic outcomes impact CD prognosis?
A meta-analysis of 12 studies, in prospective cohorts with active CD, has demonstrated that patients who achieved MH, at first assessment (at least 1 month after study onset, but 6 months prior to the last follow-up), were more likely to achieve long-term clinical remission (at week 50), had lower rates of CD-related surgery, and were more prone to maintain long-term MH. 49In this meta-analysis, the SES-CD score was used in eight studies and the cutoffs applied to grade endoscopic changes varied (MH defined as SES-CD = 0 only in half).In 2017, another meta-analysis 50 of five studies revealed that small-bowel MH, assessed through capsule endoscopy and graded using the Niv or the Lewis score, was able to predict long-term clinical remission (up to 24 months).In addition, in a post-hoc analysis of the SONIC clinical trial, participants with an endoscopic response (decrease in SES-CD or CD endoscopic index of severity [CDEIS] of at least 50 percent, at week 26) were significantly more likely to achieve a 1-year corticosteroid-free clinical remission. 51In line with this, a follow-up analysis for the CALM study showed a correlation between the achievement of endoscopic remission (CDEIS of <4 with no deep ulcers), in individuals with early CD and a reduced risk of disease progression over a median duration of 3 years. 52However, in this study, deep remission (endoscopic and clinical remission) predicted prognosis with a higher effect size.More recently, the prognostic value of endoscopy was ascertained using another score, the modified multiplier simple endoscopic score for CD (MM-SES-CD).
Compared with SES-CD, this scoring system has the advantage of detecting disease heterogeneity, 53 by attributing varying weights to each individual parameter of SES-CD, according to their prognostic values for achieving endoscopic remission. 53A reduction ≥40 percent in MM-SES-CD, after biologic initiation, was found to be significantly more specific at predicting endoscopic remission after one year compared with SES-CD (95 percent vs. 63 percent, respectively). 53

Are histological response and remission achievable in UC?
While the value of histology in UC has been recognized since early clinical trials, 54,55 attention has been directed toward formal histological scoring only in recent years.Even though the concept of MH was initially defined by endoscopic evaluation alone, since 2016, guidance from the Food and Drug Administration recommends that it should encompass also histological healing. 56However, the histological scoring of UC is a complex process comprising twenty six distinct histopathological scoring systems. 57 inflammation, and surface epithelial injury, ranging from 0 to 33 58 ; histological remission is defined as RHI ≤ 3 and histological response as RHI ≤ 9. On the other hand, NHI assesses lamina propria chronic inflammation, neutrophilic inflammation, and surface damage (erosion/ulcers) on a scale that ranges from 0 to 4. 59 Histological remission is defined as an NHI of 0 and histological response as an NHI ≤ 1.This index has a strong correlation with RHI, good interobserver variability, and adequate response to change after the treatment. 3,60nother system that is deemed suitable to evaluate the histologic activity of UC is the Geboes score (GS).GS ranges from 0 to 5.4; being histological remission defined as GS ≤ 2.0 and histological response as GS ≤ 3.0. 3This score has not been fully validated 8,57 and its complexity limits its use in daily clinical practice.It is also important to acknowledge that it was not explicitly intended for quantifying changes in histologic activity because of its hierarchical scoring system. 57eal-world evidence has shown that drugs currently indicated for UC, including infliximab, 12 adalimumab, 13 vedolizumab, 61 and ustekinumab, 62 are able to induce histological remission in up to 45 percent of the patients.To overview the ability of most recent therapies to achieve histological outcomes, we have conducted a search for RCTs, published since 2020, that examined the impact of drugs on histological endpoints in UC; the search included the terms "((histologic score) or (histologic response) or (histologic remission)) and (ulcerative colitis) and (RCT)".Overall, ten studies were identified [14][15][16][19][20][21][22][23]27,63 (Table 2). As reportedin a previous meta-analysis, the most used score was the GS. 8 On the other hand, NHI was applied in three studies, [21][22][23] while two studies resorted to RHI. 20,63 In general, all studied drugs allowed achieving histological outcomes more frequently than placebo, yet with variable magnitudes of effect.The percentages of histologic response ranged from 24.4 percent to 58.6 percent on induction (for adalimumab 63 and upadacitinib, 14 respectively) and from 25.6 percent to 45.7 percent on maintenance (for adalimumab 63 and vedolizumab 63 ).

Do histological outcomes impact UC prognosis?
The presence of neutrophilic inflammation within the lamina propria and epithelium defines histological activity, 3 while its absence corresponds to histological remission. 7In a metaanalysis of twenty studies (twelve prospective), patients in endoscopic healing or remission (MES 0) and histological remission (evaluated using different scores) had a 61 percent lower risk of clinical relapse compared with those with persistent histological activity (14 percent vs. 5 percent). 30nother meta-analysis of seventeen studies has shown that patients in endoscopic remission (MES of 0 or 1) but with persistent histological activity had a threefold increased risk of presenting UC relapse. 64In these meta-analyses, a more stringent target was associated with a better prognosis, as also shown by Magro et al. 65 Indeed, these authors showed that GS is an independent risk factor in the 36 months after biopsy, with disease progression occurring earlier and more frequently in patients with higher GS.In detail, 26 percent of patients with GS ≤2B had progressive disease compared with 50 percent of those with GS > 4.0.More recently, it has been shown that the PICaSSO histologic remission index (PHRI), a new score based on the presence or absence of neutrophils (and recently externally validated), presented the strongest correlation to endoscopic activity and excellent prediction of clinical outcomes, among several histological grading systems.In fact, patients with PHRI above 0 had significantly more events associated with UC exacerbations than those with PHRI = 0 (49 percent vs. 14 percent, at month 12). 66The prognostic role of histologic remission was again demonstrated by Rath et al. 57 .Indeed, patients with RHI up to three points or NHI up to 1 had a significantly higher likelihood of not presenting UC-related adverse events (disease flares, IBD-related hospitalization or surgery, initiation or dose escalation of systemic steroids, immunosuppressants, or targeted therapy) during follow-up.Despite this evidence, it remains to be understood whether less rigorous monitoring may be proposed for UC patients in histologic remission. 57

Are histological response and remission achievable in CD?
The presence of neutrophilic epithelial infiltrate is, as in UC, a straightforward manner to, dichotomously, evaluate histologic activity in CD.Still, several other changes (like basal plasmacytosis, chronic inflammatory infiltrate in the lamina propria, and presence of erosions and ulcers) 3 have been reported to have prognostic relevance.Over the decades, several scoring systems have been developed, with variable ranges and remission/activity cutoffs defined only for some (Geboes, 67 Naini and Cortina, 68 Global Histologic Disease Activity Score [GHAS], 69 Regueiro, 70 and Saverymuttu 71 scores); a few systems were developed for specific sites (ileal 72 [Drews Score], colonic 73 [Gomes Score], or rectal 74  [Ward and  Webb] samples).One notable advantage of GHAS lies in the distinct evaluation of neutrophil infiltration in the lamina propria and of a persistent chronic inflammatory response, which remains after treatment.While extensive prospective studies have not yet validated this scoring system, it remains the most employed measure in clinical trials focused on assessing microscopic activity in CD.Likewise, none of the other scores has been validated nor is commonly used in routine practice.
As for UC, we searched for RCTs evaluating histological response and remission, published since 2020.However, the search on two databases (CENTRAL and PubMed) retrieved no results.A search among the abstracts presented at ECCO meetings since 2020 identified a trial (SERENTIY; NCT02891226) that evaluated the efficacy of mirikizumab 75 on histologic outcomes, ascertained through GHAS.Almost half of the patients showed histological response, after 12 and 52 weeks of therapy, while around 20 percent achieved remission.These results suggest that histological outcomes may be a realistic goal also in CD, as highlighted in a metanalysis of older studies. 76Further RCTs evaluating the impact of therapeutics on CD histopathology are awaited.

Do histological outcomes impact CD prognosis?
Data on the evaluation of the feasibility and impact of achieving histological outcomes in CD patients is limited and restricted to retrospective studies.This is mostly related to the patchy and transmural nature of CD.
In a study by Christensen et al. with hundred patients with ileal CD, the presence of histological remission (defined as GHAS = 0), but not endoscopic remission (no erosions nor ulceration), was associated with a reduction in clinical relapse, medication escalation, and need for corticosteroids. 77onversely, Hu et al. 78 did not observe a notable difference in the risk of relapse or quality of life in CD patients in endoscopic (SES-CD of 0), regardless of whether they achieved histological remission (defined as an absence of active or quiescent disease) or not.However, in this cohort, the duration of endoscopic remission, prior disease activity, and therapeutics were not assessed.In 2021, a study comprising 215 CD patients with histological assessment showed that, over a 2-year follow-up period, individuals who were in clinical, endoscopic, and histological remission (no evidence of disease or quiescent disease, per the Nancy index) had a 43 percent reduced risk of treatment failure compared with those who had ongoing histologic activity. 79However, no significant differences were observed in the magnitude of

Unmet needs for pursuing endoscopic outcomes
Even though the evaluation of endoscopic endpoints is of utmost relevance drug development, their systematic evaluation and reporting in everyday clinical practice are still suboptimal, particularly for CD.As such, the systematic use of endoscopic scores, such as SES-CD, outside clinical research settings, should be promoted, to improve the estimation of disease activity and prognosis.This will impact clinical decisions, including treatment escalation and de-escalation.
In this setting, research efforts are being directed at developing tools to increase clinician adherence to endoscopic scores.For instance, the novel simplified endoscopic mucosal assessment for CD (SEMA-CD) score was recently derived from the SES-CD with this objective. 80Even though simplified approaches have their appeal, some knowledge gaps may be filled by understanding patterns of endoscopic healing and focusing on challenging and harder-to-heal bowel segments.In addition, updated endoscopic scoring systems to predict postoperative recurrence following ileocecal resection are warranted.Indeed, the Rutgeerts score was developed during the end-to-end anastomosis era and, as such, may not be appropriate for modern surgical techniques.For example, patients who undergo side-to-side anastomosis may have a higher number of ileal lesions but fewer anastomotic lesions while an ileal body may be present near the inlet of the neoterminal ileum.
Regarding UC, even though endoscopic activity is often evaluated using MES, in real-world practice and in clinical trials, the use of more robust and validated indexes (such as the UCEIS 31 and the ulcerative colitis colonoscopic index of severity 81 ) remains suboptimal.
Several artificial intelligence (AI) tools are being developed to score IBD endoscopic lesions, particularly in the setting of UC.Indeed, several deep learning algorithms have been described to be able to predict UC severity (graded using the MES, UCEIS or, more recently, PICaSSO 82,83 ) from full-length colonoscopy videos.Concerning CD, most studies focused on capsule endoscopy readings for automated endoscopic grading. 84It is anticipated that the application of AI tolls will increase objectivity, reducing interindividual subjectivity and increasing adherence to scores.

Unmet needs for using histology as a target
Regarding CD, it is mandatory to validate the existing histopathological scores (particularly GHAS) and promote their incorporation as secondary endpoints in clinical trials.Indeed, since 2020, only one RCT did so. 75Furthermore, investigating the prognostic significance of histological remission through prospective studies is also warranted.Such prognostic information would be particularly valuable within the context of the treatment cycling concept, 85 in which the discontinuation of biological treatment may be considered for specific patients.In addition, it is not clear if histology is the key tool to monitor relapsing episodes, following treatment discontinuation.Also, even though data on the correlation between imaging scores (particularly regarding intestinal ultrasound) and endoscopic ones are starting to emerge, 86 the correlation with histological scores remains uncovered.
Regarding UC, although evidence supports that histological remission has prognostic relevance, several aspects hamper its utilization as a formal target.First, doubts persist regarding the added value of histology in patients who present clinical and/or endoscopic activity and/or present high values of fecal calprotectin.Second, the costs of this evaluation are high.Third, a standardized and uniform definition for histological remission is urgent but is still lacking.The use of simpler scores (based solely on the presence of neutrophils, as the PHRI), the implementation of formal training, and centralized reading may improve adherence, reporting quality, and patient monitoring. 57The impact of histological remission on monitoring frequency remains also underexplained.
Finally, a quick note about the application of AI algorithms for the histopathological evaluation in IBD.Recently, a convolutional neural network-based model was developed to classify the histologic features of patients with UC. 66 This algorithm categorizes patients as being on histological remission or nonremission, depending on the presence of neutrophils in whole slide images. 66The overall performance of this model was good, with values of sensitivity and specificity of 81 percent and 93 percent in the validation set. 66Other research groups have also developed similar models, achieving comparable diagnostic performances. 87Conversely, the application of automation for CD histological assessment is less studied.The single study found used a deep-learning model for the analysis of surgical specimens and to predict CD postoperative recurrence; its performance was outstanding (area under the curve of 0.995). 88These technical innovations will eventually lead to a standardized assessment of disease activity, which, up to now, is highly subjective and burdened by interobserver variability.
To finalize, future studies may provide further clarification on the feasibility and usefulness of molecular healing as a treatment target for IBD.Indeed, a recently published prospective study 2 involving CD and UC patients in clinical remission found that barrier healing, evaluated through confocal laser endomicroscopy, was more predictive of significant adverse outcomes than histologic remission.

Table 1 .
Endoscopic response and remission achieved in randomized-controlled trials of Crohn's disease and ulcerative colitis published since 2020.

Table 1 .
Continuedendoscopic response at week 12 was superior for patients with colonic activity only (40.3 percent vs. 32.7 percent).

Table 1 .
Continued Journal of the Canadian Association of 2024, Vol. 7, No. 1

Table 2 .
Histological response and remission in randomized-controlled trials published since 2020.