The Role of Mucosal Healing in the Treatment of Patients with Inflammatory Bowel Disease

Introduction

Classically, the goals of care when treating a patient with inflammatory bowel disease (IBD) were inducing and maintaining clinical remission and improving the patient’s quality of life [1, 2]. Unfortunately, many of the therapies for IBD have rare but potentially serious side effects [3–7]. Thus, when managing a patient with IBD, physicians must balance the benefits and risks of the various medications with the risk of untreated, or undertreated, disease [8]. A new treatment paradigm is emerging that focuses on “mucosal healing” (MH) in addition to the alleviation of symptoms [9, 10]. Some physicians are now advocating “deep remission,” or clinical plus endoscopic remission, as the goal for the treatment of IBD in the hope that attaining deep remission will improve the natural history of IBD. However, while this concept is intuitively appealing, not all patients will achieve MH, despite aggressive therapy, and in some cases it may expose patients to unnecessary risk from medications without an increased benefit. Changing the treatment paradigm for IBD to include deep remission requires quality data that MH healing is both attainable and improves clinical outcomes for patients with IBD.

Definition of mucosal healing

There is currently no singular validated definition for MH. For some, mucosal healing means a normal-appearing mucosa devoid of any inflammatory changes, although this can be quite difficult to attain. Practically, MH may be considered the absence of ulceration, which is potentially a more realistic goal. MH may also be defined as significant improvement in the appearance of the mucosa, but such a definition is notably more subjective. For the purposes of this review, we specify the definition of MH as noted in individual studies when discussing specific results.

Crohn’s disease

There are numerous endoscopic scoring systems for Crohn’s disease (CD), most of which are typically utilized only in the research setting. The Crohn’s Disease Endoscopic Activity Index (CDEIS) and the Simple Endoscopic Score for Crohn’s disease (SES-CD) are the most commonly used. The Rutgeerts score for CD is utilized in the postoperative setting, both clinically and in research trials. The CDEIS defines MH as a score of <6, with complete endoscopic remission being a score of <3 (which precludes any ulcerations). However, while often considered the gold standard, this system is not perfect. The CDEIS requires a complicated computation involving multiple measurements in various segments, thus limiting its practicality in the non-research setting [11]. Although the CDEIS can be laborious and time-consuming, it demonstrates good inter-observer correlation (r = 0.72) [11, 12]. Additionally, the CDEIS does not correlate with symptoms in the Crohn’s Disease Activity Index (CDAI) [13, 14]. The SES-CD was devised as an alternative to the CDEIS, and while reliable and reproducible, it is still cumbersome for routine use [15]. The Rutgeerts score is a validated tool to assess CD recurrence after ileocecal resection that is simple to use and an excellent predictor of symptomatic postoperative recurrence[16]. However, its use is limited to the postoperative setting. It is also important to emphasize that CD is a transmural process that may involve any segment of the GI tract, thus limiting the ability of endoscopy to accurately assess MH in every clinical situation. New scoring systems such as the Lemann score for CD are designed to incorporate transmural inflammation and allow assessment of the entire bowel [17].

Ulcerative colitis

There are more than 10 indices used to assess ulcerative colitis, none of which have been validated [11, 16, 18, 19]. The Mayo score is the most commonly utilized. For UC, it has been suggested that a Mayo endoscopic score of 0–1 be used as a cutoff for MH, with 0 being normal or inactive disease and 1 being mild erythema, decreased vascular pattern, or mild friability [20]. Other endoscopic scores in UC include the Baron endoscopic score or the endoscopic component of the UC Disease Activity Index (UCDAI) [21, 22]. In a consensus statement, D’Haens et al. propose that MH be defined as absence of all “friability, blood, erosions, and ulcers in all visualized segments” [23]. Additionally, they argue that due to a lag time in healing, histologic remission should not be a primary endpoint in a research protocol, although it is of significant interest and should be viewed as a secondary outcome. These scores are quite helpful, especially in the research setting, although the definition of MH can differ between trials and, historically, none of the tools have been validated. A new endoscopic scoring system, the Ulcerative Colitis Endoscopy Index of Severity (UCEIS) was recently developed[24] as a tool to measure endoscopic severity of UC based on flexible sigmoidoscopy. It consists of 3 descriptors (vascular pattern, bleeding, erosions/ulcers), each with a score ranging from 0–8. While there is no fully validated test for endoscopic severity of UC, the UCEIS has undergone an initial validation and demonstrated satisfactory reliability. This test promises to be advantageous as the first validated tool to measure endoscopic activity in UC and will likely become used in clinical trials.

How often is MH achieved with various medications?

In order to be able to recommend MH as a goal of therapy for patients with IBD, it is critical to understand how often that goal is attainable and to select a therapy that has the highest probability of attaining this objective. The following section will address the data regarding the ability of various medications to achieve MH. Table 1 shows the rates of MH seen with specific medications for CD, and Table 2 shows the same for UC.

Table 1.

Rates of mucosal healing with various drug therapies in Crohn’s disease

Therapy	Definition of mucosal healing	Rate of mucosal healing (%)	Time Endpoint
Steroids	Endoscopists’ assessment	29%	3–7 weeks [13]
Azathioprine	CDEIS	36–58	12–42 months [30,31]
Methotrexate	Disappearance of mucosal ulceration	11	24 months (median) [35]
Infliximab	Disappearance of mucosal ulcerations	45–50	6.7–12.5 months [36,37]
Infliximab+azathioprine	Disappearance of mucosal ulcerations, SES-CD	44–73	6–24 months [32,38]
Adalimumab	CDEIS	24	12 months [39]
Certolizumab pegol	Disappearance of mucosal ulceration	8	12.5 months [40]
Natalizumab	Decrease in SES-CD >70%	42	14.1 months (mean) [41]

Table 2.

Rates of mucosal healing with various classes of medications in Ulcerative Colitis

Therapy	Definition of mucosal healing	Rate of mucosal healing (%)	Time Endpoint
Steroids	Baron score 0/1	12 – 38%	2–12 weeks[51–53]
5-ASA	Mayo Endoscopy subscore 0/1, Modified UC-DAI	19–88	6–12 weeks[43–49,54]
Azathioprine	Baron Index of 0/1	53	6 months[54]
Infliximab	Mayo Endoscopy subscore 0/1	46–62	2–12.5 months[20]
Adalimumab	Mayo Endoscopy subscore 0/1	25–47	2–12 months[55,57]
Golimumab	Mayo Endoscopy subscore 0/1	42–45	1.5–12.5 months[56,58]
Vedolizumab	Mayo Endoscopic subscore 0/1	52–56	12 months[59]

Crohn’s Disease

Steroids

Steroids have long been used to induce clinical remission with CD, although the data demonstrating the effect of steroids on MH are limited. In a study of 8 patients with prior ileocecal or ileocolonic resections, oral prednisolone was unable to achieve endoscopic remission after 6–9 weeks of treatment despite clinical improvement [25]. In another cohort of patients who achieved clinical remission on steroids, only 29% had subsequent endoscopic remission (defined as absence of mucosal lesions or scarred lesions only) at 3–7 weeks [13]. A recent systematic review suggested that despite the effectiveness of second-generation corticosteroids for induction of remission, the data for mucosal healing are limited [26].

Thiopurines

Until recently, data for endoscopic healing with thiopurines in Crohn’s disease have been limited. In the 1990’s, only a few small trials included MH as an outcome [27–29]. More recent studies have used MH as an endpoint, and the results have been variable. In 2005, the GETAID group demonstrated that in patients in clinical remission on azathioprine for at least 42 months, only 36% had a CDEIS of 0 (complete MH) [30]. A subsequent randomized trial of MH, where MH was defined as a CDEIS score of less than 6, compared azathioprine to budesonide. Azathioprine demonstrated MH rates as high as 83% at 1 year, compared to 24% with budesonide alone [31]. However, when considering only complete MH (CDEIS of 0) and on intention-to-treat analysis, the rate of MH dropped to 58%. In the SONIC trial, treatment-naïve patients randomized to the azathioprine arm had a 6-month MH rate of 16.5% [32]. The SONIC trial defined MH as no ulcers among those who had ulceration at baseline. In addition to the follow-up time differences, this definition of MH is more stringent than a CDEIS score of <6 and thus may account for some of the variations seen in MH rates with azathioprine.

Methotrexate

Data regarding MH rates with methotrexate are also limited and quite variable. In a small study from 1989 that evaluated the efficacy of methotrexate for refractory IBD [33], 5 of 14 patients with Crohn’s colitis demonstrated mucosal and histologic healing after 12 weeks of methotrexate. Notably, none of the patients with UC had MH. Another small but more recent case series noted that up to 50% of CD patients with a clinical response will achieve MH [34]. Neither of these studies included a control arm, making the results difficult to interpret. A prospective observational study of patients in clinical remission defined MH as absence of mucosal ulcerations in all segments, and found MH rates of 11% with methotrexate versus 50% with azathioprine and 60% with infliximab [35]. However, the overall number of patients was small and each group had variable follow-up.

Anti-TNF

Unlike the therapies discussed thus far, many more studies have been done within the last decade to directly evaluate the effect of anti-TNF inhibitors on mucosal healing. In an ACCENT I substudy, MH (defined as the absence of mucosal ulcerations at follow-up in patients who had ulcerations at baseline) was achieved in 31% of patients at 10 weeks who received scheduled infliximab (IFX) and in 50% of patients at 54 weeks. Placebo rates of healing were 0% at week 10 and 7% at week 54 [36]. A second study of MH with IFX, where MH was similarly defined as disappearance of mucosal ulcerations, found that 45% of patients had complete healing after a median of 6.7 months (interquartile range 1.4–24.6 months) with sustained infliximab treatment [37].

The SONIC trial assessed MH in treatment-naïve patients who received azathioprine, infliximab, or the combination of azathioprine and infliximab. MH was defined as disappearance of mucosal ulcerations after presence of ulcerations at baseline [32]. At 6 months, the rates of MH were 43.9% on combination therapy, 30.1% on infliximab alone, and 16.5% on azathioprine alone. In the “Step Up versus Top Down” trial, patients were randomized to induction therapy with 3 infusions of infliximab and maintenance therapy with azathioprine versus traditional “step up” therapy with multiple courses of corticosteroids before institution of a thiopurine [38]. This longer-term study found a MH rate, defined as no ulceration, of 73.1% in the early combination treatment arm versus 30.4% in the traditional therapy arm at 2 years’ time.

Adalimumab has also been shown to induce MH. In the EXTEND trial, adalimumab had a MH rate of 27% versus 13% at week 12 (p = 0.56) and a MH rate of 24% versus 0% in placebo at 1 year (p<0.001) [39]. MH was defined as no ulcerations present in any of the 5 examined segments used in the CDEIS. There is also some data for certolizumab pegol and MH, although the numbers are less impressive [39,40]. In the MUSIC trial, the primary outcome was a mean change in CDEIS score at week 10, while secondary outcomes included MH at weeks 10 and 54. The mean CDEIS was 14.5 (standard deviation of 5.3) at baseline, with mean decrease of 5.7 at week (95% CI 4.6 to 6.8, p<0.0001) [40]. MH, defined as the absence of ulcerations, occurred in 4% of patients at week 10 and 8% of patients at week 54. The rate of complete endoscopic remission (or CDEIS score of <3) was only 10% at week 10 and 14% at week 54. The authors argue that the rates of complete MH were low given the severity of baseline ulcerations.

Anti-integrin inhibitors

A small retrospective study of MH with natalizumab in patients who had previously failed immunomodulators and anti-TNF agents demonstrated a MH rate of 42.3% after treatment for a mean duration of 14.1 months [41]. MH was defined as a decrease of the SES-CD of >70%, which are less stringent criteria than studies defining MH as complete absence of ulceration. MH was not an endpoint in the recent trial of vedolizumab for CD [42].

Ulcerative colitis

Given that the inflamed mucosa is easier to assess in UC, MH has been used as an endpoint in multiple landmark studies of the treatment of UC. However, interpretation of these studies is complicated by heterogeneity in the definition of MH, severity of disease in patients studied, extent of disease, and time points assessed. The ranges of rates of MH for various therapies in UC are presented in Table 2.

Mesalamine

The ASCEND I and II trials assessed the use of delayed-release mesalamine (Asacol) for the treatment of mild–moderate UC [43, 44]. In these trials, MH was defined as a Mayo endoscopy score of 0 or 1. A post hoc analysis determined that 80% of patients with moderately active UC achieved MH on mesalamine 4.8 g/day and 68% achieved MH on mesalamine 2.4 g/day (p=0.012). Higher rates of MH were achieved in those with mild UC, with 84% achieving MH on mesalamine 4.8 g/day and 88% on mesalamine 2.4 g/day (p=0.765) [45]. It is important to note that the definition of MH did include a Mayo score of 1, which includes mild erythema and friability [20].

Two large studies evaluating the efficacy of MMX mesalamine (Lialda) for the treatment of mild–moderate UC used a stringent definition of MH that only allowed minimal erythema, granularity, and decreased vascular patterns [46,47]. Lichtenstein et al. found that the combined endpoint of clinical and endoscopic remission by week 8 was achieved in 34.1% of patients on MMX mesalamine 2.4 g BID, 29.2% of patients on 4.8 g/day, and 12.9% of patients on placebo, with both therapy arms achieving significance versus placebo [46]. Improvement in endoscopic score by week 8 was seen in 61.4% of patients on MMX mesalamine 2.4 g BID, 69.7% of patients on 4.8 g/day, and 35.3% of patients on placebo [46]. Kamm et al. found clinical and endoscopic remission by week 8 in 40.5% of patients on MMX mesalamine 2.4 g/day, 41.2% of patients on 4.8 g/day, 32.6% of patients on Asacol 2.4 g/day, and 22.1% of patients on placebo [47]. Improvement in endoscopic score by week 8 was seen in 69% of patients on MMX mesalamine 2.4 g/day, 76.5% of patients on 4.8 g/day, 58.1% of patients on Asacol 2.4 g/day, and 41.9% of patients given placebo. A post hoc analysis of these studies determined that the combined rate of MH in these studies was 32% on MMX mesalamine 2.4 g/day, 32.2% on 4.8 g/day, and 15.8% on placebo [48]. The addition of topical mesalamine to oral mesalamine may improve the proportion of patients achieving MH. This was assessed in a prospective trial of 6 weeks of combined oral (4 g/day) and rectal (2 g/day) mesalamine[49]. By week 6, 69.1% of patients on this regimen achieved MH (defined as Mayo endoscopic subscore of 0– or 1).

A recent meta-analysis of patients with mild–moderate disease found MH rates in a range of 35–49% on oral mesalamine agents and 46–62% on topical mesalamine[50]. The higher rates of MH with oral mesalamine in the ASCEND I and II studies may be due to the definition of MH used. In the combined ASCEND I and II trials, if MH healing was defined as an endoscopic score of 0, it was achieved in only 32% of moderately active patients on mesalamine 4.8 g/day and 24% of patients on 2.4 g/day (p=0.125). Using the more stringent definition of MH in those with mild UC, the rate of MH decreased to 44% in patients on mesalamine 4.8 g/day and 42% in patients on mesalamine 2.4 g/day at 6 weeks (p=0.640)[45].

Steroids

In 1955, Truelove et al. described the effects of cortisone in ulcerative colitis. In this study, 30% of patients had “normal” or “near-normal” endoscopic appearance on sigmoidoscopy following cortisone 100 mg/day for 2 to 6 weeks (versus 11% in the control group) [51]. In a more recent cohort of UC patients receiving steroids, 38% were able to achieve MH at three months, where MH was defined as a Baron score of 0 (i.e., normal mucosa or mild erythema) [52]. Others, however, noted less impressive results. In a study comparing budesonide to prednisolone, only 11.8% of patients who received budesonide and 15.8% of patients who received prednisolone achieved endoscopic remission (defined as normal or non-inflamed mucosa at 4 weeks) [53].

Thiopurines

There are few studies in UC assessing the effect of thiopurines on MH. A study by Ardizzone et al. measured rates of MH after 6 months on azathioprine 2 mg/kg/day versus 5-ASA 3.2 g/day [54]. This trial randomized corticosteroid-dependent patients to azathioprine 2 mg/kg/day or mesalamine 3.2 g/day and utilized the Baron endoscopic score of 0 or 1 (range 0–3) as the definition endoscopic remission. A Baron score of 0 is normal mucosa, while a Baron score of 1 represents granular or edematous mucosa with loss of vascular pattern [21, 54]. The authors noted a MH rate of 53% on azathioprine and 19% on mesalamine at 6 months[54]. Thus, it is reasonable to conclude that thiopurines for steroid-dependent UC do achieve MH, and likely at a higher rate than mesalamine.

Anti-TNF

Anti-TNF medications including infliximab, adalimumab, and golimumab are approved for the treatment of moderate–severe UC [20, 55–57]. In the ACT 1 and 2 studies, MH was defined as Mayo endoscopy subscore of 0 or 1. At week 8, MH occurred in 59–62% of patients on infliximab and 30.9–33.9% of patients on placebo. At week 54, MH occurred in 45.5–46.7% of patients on infliximab and 18.2% of patients on placebo [20]. The ULTRA 1 and 2 studies evaluated adalimumab for moderate–severe UC and similarly defined MH as a Mayo endoscopic subscore of 0 or 1 [55, 57]. In ULTRA 1, 37.7–46.9% of patients who underwent induction followed by maintenance achieved MH at 8 weeks [55]. However, this was not significant compared to the placebo rate of MH of 41.5%. In ULTRA 2, the MH on adalimumab was 41.1% versus the placebo MH rate of 31.7% by week eight (p<0.05) [57]. By week 52, 25% had MH on adalimumab versus 15.4% on placebo (p<0.05). This effect appeared to be driven by patients who were anti-TNF-naïve. The PURSUIT-SC trial was a large phase 2/3 trial of golimumab for moderate–severe UC. MH was a major secondary endpoint defined as Mayo endoscopic subscore of 0 or 1. MH was achieved in 43.2–45.3% of patients on golimumab compared to 28.5% of patients on placebo (p<0.05) [56]. MH was achieved in 41.7–42.4% of patients at 30 and 54 weeks (versus 26.6% on placebo) [58].

Anti-integrin inhibitors

Unlike the trial for CD, the phase 3 trial of vedolizumab in UC did have MH as a study endpoint [59]. MH was defined as Mayo endoscopic subscore of 0 or 1. The vedolizumab arm had a MH rate at 52 weeks of 51.6 –56% versus 19.8% in the placebo arm (p<0.001) [59].

Why is mucosal healing important?

Using MH as clinical endpoint is helpful for multiple reasons. First, it allows the clinician to distinguish symptoms due to underlying inflammation from other etiologies such as irritable bowel syndrome and bile salt malabsorption, as symptom scores and endoscopic scores do not correlate[13]. One study documented that patients with IBS had higher CDAI scores than patients with CD [60]. This was driven by the subjective subcomponents of the CDAI. There is also a considerable overlap of IBS symptoms in patients with IBD [61]. As our medical therapies have rare but significant adverse events, it is important to confirm objective evidence of active IBD before stepping up therapy. This can help guide therapy and potentially prevent inappropriate “step-up” therapy when MH is already achieved. Additionally, MH has been associated with a number of improved outcomes in both CD and UC. Although the current data (discussed below) are limited, there appears to be a consistent association between MH and fewer disease-related complications such as surgery and hospitalizations. Additionally, the severity of endoscopic lesions can also help to predict the course of IBD. Studies have shown that patients with severe ulcerations have a worse clinical course, and other studies have demonstrated that MH tends to predict a sustained clinical response. Finally, for research purposes, MH is a more objective endpoint than clinical indices.

Crohn’s disease

Several studies have noted improved outcomes in patients with MH. Studies suggest that patients with MH are at lower risk of flaring, of developing complications of CD, and of requiring hospitalization or surgery. A single-center prospective cohort of patients with CD receiving IFX noted that patients who achieved MH had significantly fewer hospitalizations and abdominal surgeries than patients who did not achieve MH [62]. In the same study, patients who achieved partial MH (clear endoscopic improvement although ulcerations were still present) had similar surgical rates to those who achieved complete MH (absence of ulceration). Similar results were noted in a substudy of ACCENT I patients. Patients achieving MH had fewer hospitalizations, although the results were not statistically significant [36].

MH also appears to be associated with fewer flares of CD. Patients who achieved MH as part of the “Step Up versus Top Down Study” had fewer flares over a 2 year follow-up period compared to the group that did not achieve MH [38, 63]. In fact, complete MH (SES-CD of 0) was the only predictor of steroid-free remission in the 2-year follow-up period. Similarly, in a prospective study comparing the MH rates of methotrexate to thiopurines and infliximab, patients who achieved MH (defined as CDEIS <4), regardless of the specific therapy, had fewer flares over the 2-year follow-up time[35]. Thus, while data are still limited to subsets of patients in prior studies and observational cohorts, achieving MH appears to confer a long-term benefit [64].

Achieving MH also predicts a better clinical course. For example, healing early on in the course of therapy with IFX predicts MH at 1 year. Although the number of patients in the trial was small, a prospective observational study noted that patients who achieved MH at 3 months predicted MH at 1 year [65]. The rate of MH at 3 months was only 24%, but significantly more of these patients with early MH still had MH at 12 months (70% versus 17% of patients without early MH, p = 0.01). Thus the patients who achieve MH may represent a lower-risk group of patients.

Additionally, MH may also predict longer-term outcomes. In the IBSEN cohort, achieving MH at 1 year was found to be predictive of less inflammation seen on endoscopy at a 5-year follow-up (p = 0.02) [66]. Among patients who were on steroids at 1 year, those who achieved MH at 1 year were less likely to still be requiring steroids at 5 years (p = 0.02). Total steroid use, however, was the same at 5 years regardless of MH status at 1 year (p = 0.4). While there was a trend towards fewer surgeries (p = 0.1), there was no association with MH at 1 year. And while the cohort was large, the number of patients with 5 year MH data was smaller (124 patients). The study also had several limitations. First, it was spread out over many sites, increasing the potential variability in treatment decisions such as use of steroids or validation of a flare, and the definition of MH was for this cohort did not use a validated endoscopic scoring system. Although less inflammation was seen at 5 years for those who achieved early MH, improvement in other clinical outcomes requires more evaluation. To summarize, there is ample evidence that MH healing in CD is associated with better outcomes – specifically, fewer hospitalizations and surgeries and potentially fewer flares.

In addition, MH may have a role in predicting the course of disease. Patients with severe endoscopic disease are at higher risk for flaring and developing complications, such as surgery. Retrospective analysis suggests that patients with CD who have severe ulceration on endoscopy have an increased risk of colectomy compared to those who do not. Endoscopy can thus be used to help risk-stratify patients and identify those who may be more appropriate for early aggressive treatment with biologics.

Ulcerative colitis

Similar to CD, MH in UC is associated with improved outcomes and can help predict disease course. In a post hoc analysis of ACT 1 and ACT 2, patients who had a lower endoscopic subscore at week 8 were more likely to be colectomy-free at week 54 [67]. Interestingly however, there was no difference in outcomes in patients who achieved a subscore of 0 vs. 1. While it may be that complete MH healing (Mayo endoscopic score of 0) does not provide benefit beyond that of near-complete MH (Mayo endoscopic score of 1), it is possible that the study was underpowered to detect a difference between those groups. Another retrospective analysis found similar results, where patients achieving an endoscopic score of 0 or 1 had significantly fewer colectomies at 48 months [68]. The ISBEN cohort demonstrated that those who achieved MH at 1 year had significantly fewer colectomies reported at 5 years (p = 0.02) [66]. Given the small overall number of colectomies, only a limited analysis could be performed, but MH was consistently protective against colectomy at 5 years, with a RR of 0.22 (95% CI: 0.06– 0.79). However, MH was not able to predict patient-reported disease course at 5 years. Unlike the CD arm, MH in UC at 1 year was not associated with less inflammation at 5 years. Notably, no patients in this cohort were given biologics, and drug maintenance data is not reported. Given other data suggesting that maintenance medication is required for maintenance of MH, it is possible that variations in treatment among these patients over 5 years accounted for the lack of additional long-term benefit [36].

In patients with less severe UC, a prospective trial of combined oral and topical mesalamine for mild–moderately active UC found that patients who achieved MH had a significantly lower relapse rate, at 23%, as opposed to 80% in patients who only achieved clinical remission (p<0.0001) [49]. However, the majority of patients in this study who achieved clinical remission also achieved endoscopic remission. In Milan, another cohort of patients with UC who were followed for 5 years demonstrated that those with MH required less immunosuppressive therapy and experienced fewer hospitalizations and colectomies than those who did not achieve MH 3 months after steroids [52]. While only 60 patients achieved MH, on a multivariate analysis, the lack of MH after corticosteroid use was the only predictor of more aggressive disease (defined as need for immunosuppressive medications, hospitalizations, or surgery) at 5 years.

MH in UC is not only associated with improved outcomes but is also predictive of disease course. Using endoscopic lesions to predict the course in UC is not a novel concept. In 1996, for example, in developing a prediction tool for outcomes of severe UC, Travis et al. noted that the presence of severe rectal inflammation was more common in the group of patients that went to colectomy [69]. Additionally, in patients with severe UC who require rescue therapy with cyclosporine, those without severe lesions seen on endoscopy are more likely to avoid colectomy than those with severe lesions (p<0.001) [70]. This finding has also been demonstrated in patients with refractory UC receiving IFX. In a multivariate analysis, Laharie et al. found the absence of MH to be the only predictor of colectomy, with an OR of 18.01 (95% CI: 1.58–204.92; p = 0.02) [68]

To summarize, while the data for UC are still somewhat limited, patients who achieve MH appear to have improved outcomes, including fewer flares and fewer colectomies. However, based on the available data, it is not clear that MH in UC is associated with fewer hospitalizations.

Prevention of cancer

In IBD, MH may provide an additional benefit by reducing the incidence of colorectal neoplasia. Severe inflammation on histology is a risk factor for colorectal cancer in UC [71,72]. Similarly, it has also been suggested that endoscopic evidence of inflammation increases the risk of progression to neoplasia and that an endoscopically normal colon in a patient with UC has a similar risk of cancer as the general population [73]. Unfortunately, demonstrating this effect in a prospective controlled trial would require following a large number of patients over a long period of time. Thus far, evaluation of large data repositories has been unable to show a decrease of colorectal cancer with mesalamine over a short period of time [74]. However, a recently published case-control study demonstrated a protective effect of mesalamine and thiopurines against colorectal neoplasia in UC [75]. This study was large enough to stratify for degree of histologic inflammation and while mesalamine agents were not significantly associated with a decreased risk of colorectal neoplasia after adjustment (p = 0.56), an association was present with thiopurines (p = 0.011). If MH does have a chemoprotective effect, it is likely more relevant in UC than CD.

Limitations of mucosal healing

MH, despite its many advantages, is not a perfect endpoint. First, there is no single universally agreed-upon definition of MH. This makes it difficult to translate MH into practice as a clinically relevant endpoint. The majority of recent studies for UC use the Mayo endoscopic subscore of 0 or 1, which should be relatively easy to use clinically. There is no simple correlate in CD, however, and more comprehensive assessments such as the Lemann score may be required. A second limitation of MH as an endpoint is that it is not always achievable and may expose patients to unnecessary risk, particularly when it leads to escalation of therapy for a patient already in clinical remission. A more reasonable approach may involve a trial of escalated therapy for 6–12 months, with a plan to scale back treatment if endoscopic improvement is not achieved. A third limitation, more specific to Crohn’s disease, is that MH may be difficult to assess and may not be accessible by routine ileocolonoscopy, as in more proximal small-bowel Crohn’s disease. As such, other surrogate markers such as fecal calprotectin or cross-sectional imaging are likely to be required in this setting to assess for healing. Additionally, the majority of endoscopic scores are difficult to calculate in practice, thus limiting their use. Simply assessing for the presence of any ulceration is usually feasible and is potentially the best outcome of MH. Finally, the current evidence that MH is associated with better outcomes is largely of an indirect nature, and more data is necessary regarding long-term outcomes of MH in order to appropriately advise patients. To truly demonstrate that achieving MH is superior to clinical remission requires a large comparative-effectiveness trial randomizing patients to treatment to clinical remission or increasing therapies until MH is achieved.

Conclusion

In conclusion, MH is an appealing endpoint for IBD for a variety of reasons. Current data suggest that MH, when achieved, in both CD and UC, is associated with fewer disease flares, hospitalizations, and surgeries. Thus it is reasonable to educate patients on the potential benefits of MH and to discuss this option of as a goal of care. However, MH may not be an achievable endpoint for all patients. If therapy is escalated for the sole purpose of achieving MH, then reassessment within a year is recommended to ensure benefit. In the research setting, MH should continue to be utilized as a major variable, as it provides a more objective assessment tool than current clinical indices. The authors view MH as an important clinical endpoint. We routinely discuss MH with our patients and often strive to achieve it as a goal of care.

Opinion Statement.

Mucosal healing is a promising therapeutic endpoint in the treatment of inflammatory bowel disease. Currently, the treatment of IBD focuses predominantly on clinical remission, without requiring a demonstration of mucosal healing. However, recent studies suggest that achieving mucosal healing engenders better outcomes. In this review, we will examine the concept of mucosal healing, including its definitions and assessment techniques, and will evaluate the data regarding the potential benefits of achieving mucosal healing in the treatment of IBD. Furthermore, we will discuss how often mucosal healing can be expected with each of the therapies for IBD. While achieving mucosal healing is an ideal outcome, it may not be practical or feasible in any given patient. Although we suggest that mucosal healing be discussed with all patients, the practicing physician should realize that this goal may not be possible and should manage the patient accordingly.