In the management of Crohn's disease, earlier aggressive treatment is becoming accepted as a strategy to prevent or retard progression to irreversible bowel damage. It is not yet clear, however, if this same concept should be applied to ulcerative colitis. Hence, we review herein the long-term structural and functional consequences of this latter disease. Disease progression in ulcerative colitis takes six principal forms: proximal extension, stricturing, pseudopolyposis, dysmotility, anorectal dysfunction, and impaired permeability. The precise incidence of these complications and the ability of earlier, more aggressive treatment to prevent them have yet to be determined. (Inflamm Bowel Dis 2011;)
The treatment paradigm for Crohn's disease (CD) is recently shifting toward earlier, more aggressive treatment,1 in the hope of achieving “deep remission”2 and inducing “mucosal healing,”3 thus preventing or retarding progression to irreversible bowel damage.4 It is unknown, however, whether ulcerative colitis (UC) has the same inexorable tendency to progress to irreversible bowel damage, much less whether earlier, more aggressive therapy could alter this natural history.
In this review, therefore, we survey the available evidence concerning six major consequences of ongoing inflammation in ulcerative colitis, and we make the case for earlier intervention to prevent or retard this progression.
In contrast with CD, where disease location is fairly stable, UC is a dynamic disease with colorectal inflammation extent changing over time, which has important implications. The extent of colonic involvement in UC is clinically relevant. It has an impact on the need and type of medication, on the severity and activity of the disease, on the rates of colectomy and hospitalization, and on cancer and mortality risk.5–9 Several longitudinal referral-center based and population-based cohorts have shown that between 27%–54% of patients with proctitis or left colitis will show some proximal disease extension during the course of their disease.10–16
There is evidence that the late proximal spread of colitis, following a period of stable proctitis or left-sided disease, carries a particularly poor prognosis. For example, in a retrospective study performed by Ayres et al,13 disease extension was preceded by a clinical flare-up in 68% (36/53) of the patients that extended, and in 32% of those cases culminated in colectomy after a median period of 4.6 years. In a Norwegian cohort, there was likewise a trend toward a higher 10-year cumulative colectomy rate in patients who had progressed from proctitis or left-sided colitis to extensive colitis than in those who had extensive colitis initially (crude rate: 28% vs. 19%; P = 0.07).12
It would clearly be important to identify those clinical risk factors that could help to predict which patients are at a higher risk of disease proximal extension and who could therefore benefit from an early and intensive intervention. Young age at diagnosis,10, 17 extraintestinal manifestations,10 and nonsmoking14, 15 have been some of the clinical risk factors associated with disease extension in some, though not all, studies. In a series from Italy,14 proximal extension of disease occurred in 74 of 273 (27.1%) ulcerative proctitis patients followed for a mean period of 52 months. In univariate analysis the risk of proximal extension was higher in nonsmokers, in patients with >3 relapses/year, and in patients needing systemic steroid or immunosuppressive treatment. Interestingly, in the multivariate analysis, refractory disease (defined as >3 relapses per year, chronic disease activity despite continuous medical therapy, or need for systemic steroids or immunosuppressors) was the only independent prognostic factor, raising the question if earlier treatment with more potent therapy would be a way of preventing disease extension.
Although classically associated with CD, strictures can also complicate UC (Fig. 1). Edwards and Truelove18 described benign strictures in 6.3% of UC cases, and de Dombal et al19 reported a figure of 11.2%. In the largest series to date, Gumaste et al20 found benign strictures in 37 of 1156 (3.2%) UC patients at their referral center. Strictures in that study were defined as areas of constant narrowing, noted either by surgery, radiology, or endoscopy. The mean duration of colitis at the time of the diagnosis of benign strictures was 14.5 years; they were often asymptomatic and tended to be located in the left colon (68% in the rectum).
For a long time it was believed that UC-associated strictures developed due to muscular hypertrophy and thickening of the muscularis mucosae and muscularis propria rather than from fibrosis.21 Goulston and McGovern21 measured the various layers of the gut wall in 19 cases of benign stricture in patients with chronic UC and compared them to similar measurements in resected specimens from patients with UC without stricture as well as in normal sigmoid colons. The most striking finding was that in UC the muscularis mucosae was hypertrophied and in a state of contraction; indeed, in the zones of stricture the contraction was so strong that the inner circular coat of the muscularis mucosae was dragged away from the outer longitudinal layer, and thus produced narrowing of the lumen.
Recently, Yamagata et al22 reviewed a series of 1115 cases of UC and found strictures only in 17 (1.5%). The prevalence of strictures was smaller than in other series presumably because of these investigator's stricter definition (diameter less than 50% compared to the neighboring colon and a difficulty in passing the colonoscope) as compared to the looser criterion of any “constant narrowing of the colon.”20, 22 The authors analyzed the differences between 9 stenotic and 17 nonstenotic cases.22 They found that the stenotic cases had significantly longer disease duration, significantly higher longitudinal ulcer scar lengths both at the stenotic and nonstenotic sites (implying more pronounced fibrosis), and a significantly higher muscularis propria thickness as compared to the nonstenotic cases.22 Moreover, a significant increase in basic-fibroblast growth factor (b-FGF)-positive inflammatory cells and myofibroblasts was observed in stenotic portions; most b-FGF cells were also myeloperoxidase-positive, suggesting that the cause of colonic stenosis in longstanding UC is fibrosis, possibly induced by inflammatory neutrophils producing b-FGF.22
PSEUDOPOLYPOSIS AND BRIDGING FIBROSIS
Pseudopolyposis or postinflammatory polyps have been well described in both UC and CD.23, 24 These lesions have been supposed to result from ulceration which penetrates through and undermines the muscularis mucosae, thus allowing large mucosal tags drawn out by the propulsive action of the gut.25 In one clinical study, pseudopolyposis was found in 58 of 465 UC patients (12.5%).19 A higher prevalence of 48% was not surprisingly observed in a series of colonic resection specimens.25 Ulceration of these pseudopolyps can cause bleeding and anemia.26–28 More rarely, giant pseudopolyposis develops, having been found in 5% of one series of 86 consecutive colectomies for CD or UC.26 Such giant forms were observed mainly in quiescent disease, related more to disease extent than to activity. Complications of giant pseudopolyposis include colonic obstruction, caused either directly by bulk occlusion or indirectly by intussusception,29, 30 or even occasional protein losing enteropathy.28, 31
Besides pseudopolyposis, another type of postinflammatory lesion is bridging fibrosis.32 In one reported case, bridging fibrosis seemed to be associated with fecal urgency and incontinence even in the absence of endoscopic or histological evidence of disease activity.32
It has been long recognized that there is a colonic motility dysfunction in patients with UC, which may play an important role in the genesis of symptoms.33 The morphology and the histological changes found in “lead-pipe colon” (Fig. 2) (see Anorectal Dysfunction, below) suggest altered colonic elastic properties, compliance, and tone.34 Hence, patients who present with a narrow and stiff colon, after years of disease, will probably have altered motility, resulting in looser stools or even diarrhea, in the absence of active mucosal inflammation.
Indeed, there have been a number of studies of colonic motility in patients with UC.34–41 Disparate results (Table 1) may be attributable to small numbers of patients, different patient populations (active vs. quiescent, long-standing vs. early disease, left-sided disease vs. pancolitis), different measurement techniques, and different endpoints, but most studies agree that UC is characterized by a decreased contractility, with a reduction in the pressure or amplitude of segmental contractions.39, 40 A greater variability in the behavior of propulsive pressure waves and in the colonic response to meals has also been described.34, 36, 37, 41
Table 1. Summary of Colonic Motility Studies in UC
Colonic compliance (barostat, manometry) and transit
- Significantly increased fasting phasic activity in the upper descending colon in UC pts
- Colonic response to a meal
- 30 healthy controls
- Significantly reduced colonic contractile response to a meal in UC pts
- Effects of ulcerative proctosigmoiditis on the motor functions of noninflamed colon segments)
- No difference in colonic compliance and postprandial phasic activity as compared to controls in UC pts
- No difference after administration of iv nicotine in compliance, colonic tone or phasic motor activity in UC patients.
- Motor activity of the rectum and sigmoid
- 11 pts
- During fasting and a meal: frequency of contractions, amplitude of contractions, percentage activity, and motility index did not differ between quiescent and controls.
- Colonic response to a meal
(5 quiescent colitis; 6 active colitis)
- Amplitude of contractions, percentage activity and motility index during fasting significantly reduced in active vs controls.
- 13 healthy subjects
- Mean amplitude of contractions decreased in active disease compared to quiescent during eating
- After a meal: no difference in postprandial motility in 3 groups.
Besides in vivo studies, there have been a few in vitro experiments with colonic muscle. For example, Snape et al42 compared in vitro, strips of colonic circular muscle from patients with UC, diverticular disease, and cancer. The muscle strips, containing an intact myenteric plexus, were stimulated using electrical field stimulation (EFS). In presumably normal distal circular muscle from patients with cancer and diverticular disease, EFS initiated a small contraction during the electrical pulse (on-contraction) and a larger contraction after the electrical pulse was stopped (off-contraction).42 In strips from patients with UC, by contrast, the off-contraction was decreased, as was the force of contraction after bethanechol or potassium stimulation.42 In similar experiments, investigators from the Mayo Clinic studied spontaneous contractions and the contractions following EFS.43 Muscle strips from patients with UC presented a wider variability of spontaneous summation contractions than those from cancer or polyps controls.43 A decreased frequency of summation contractions was weakly associated with longer duration of symptoms, but not with severity of inflammation or age, suggesting that chronic disease could lead to altered colonic motility in patients with UC.
If there were to be any rationale for believing that earlier, more aggressive therapy could forestall the development of these motor abnormalities, it would have to reside in an understanding of how chronic inflammation disrupts neuromuscular integrity at the tissue level. In fact, histopathological studies have previously demonstrated that the enteric nervous system can be targeted during colonic inflammation, resulting in damage to the nerves, glial cells, and interstitial cells of Cajal (ICC), the main regulators of coordinated smooth muscle colonic activity.44, 45 Recently, Bernardini et al45 collected full-thickness archival samples from 10 patients with UC and from 10 cancer controls and assessed the distribution and number of neurons, glial cells, and ICC in the neuromuscular compartment. They observed that patients with UC displayed considerable morphologic alterations in the whole neuromuscular compartment, with a reduced density of myenteric neurons glial cells and ICC.45 Other studies have found different results, with some showing an increased number or unchanged number of the enteroglial cells and ICC of the muscular layer and myenteric plexus.46, 47 In any event, it is not yet clear whether alterations in the enteric are a primary event or the result of long-standing inflammation. Nonetheless, these alterations in the nerve and muscle compartments show that in UC other tissue components beyond the mucosal layer can be targeted during inflammation and contribute to symptoms.
Likewise, another recently published study demonstrated that oxidative stress during active inflammation permanently impaired the integrity of DNA in colonic smooth muscle cells and caused persistent damage in the expression of calcium channels.48 This phenomenon occurred even after inflammation had subsided, and could explain the smooth muscle dysfunction and persistence of symptoms even in patients with quiescent inflammatory bowel disease (IBD) and healed mucosae.49
A particularly distressing form of dysmotility in UC is impaired anorectal function, leading to urgency, tenesmus, and frank incontinence.50 If is true that these symptoms are more serious during active and severe inflammation, it is also true that patients with longstanding disease may have similar complaints even when their proctitis is mild or quiescent.
Some of these complaints can be attributed directly to changes in the length and caliber of the colon and rectum. In the past, when barium enemas were frequently performed for disease diagnosis and evaluation of activity, it was not uncommon to find patients who presented with the so-called “lead-pipe colon.” This expression was used to describe a colon with a decreased length and caliber, decreased ability to distend when filled with barium, and absent haustra.51 These features were classically associated with long-standing, “endstage” disease and were not necessarily related to current disease activity.51 In other words, permanent damage to the bowel wall, beneath the mucosa, can evidently occur late in the disease. Additional features of chronic disease are rectal narrowing and the widening of the presacral (retro-rectal) space. Alp et al52 measured the width of the presacral space between in 66 patients with IBD (49 patients with UC and 17 patients with Crohn's colitis) and in 100 healthy controls. While the width of the presacral space at the S3–S4 disc level, in a lateral projection, varied between 2–16 mm (mean 7 mm) in controls, in IBD patients it varied between 2 and 42 mm. The authors found that there was a significant direct association of the excessive width of the space with the duration of the disease, but not with the severity of sigmoidoscopic inflammation. Specifically, UC patients with a presacral width >16 mm had a mean disease duration of 10.5 years, compared with patents with a presacral space width <16 mm, who had mean disease duration of 4.5 years (P < 0.001). These dramatic morphologic changes are believed to result from proliferation of perirectal fat and fatty infiltration of the submucosa layer52; they can be appreciated on computed tomography (CT) as a “target” appearance of the rectum and by an increase in the number of nodular and streaky soft-tissue densities with an abnormal attenuation value, 10–20 HU higher than the normal.53 It is especially remarkable that some of the largest spaces are seen in patients with chronic disease and shortening of the colon and rectum.54
However, the precise contribution of the narrowed rectum and widened presacral space to the incontinence complaints of patients with long-standing proctitis has not been established. Besides decreases in the length and caliber of the colon and rectum, functional motor abnormalities presumably also play an important role in producing these symptoms.
Anorectal manometric studies performed in patients with active ulcerative proctitis have shown that rectal distension, with small volumes of air, elicits rectal pressures significantly greater than in controls. In other words, patients with active proctitis have hypersensitive and less compliant rectums, accounting for symptoms such as tenesmus and urgency.55–58 In quiescent disease, though, results are less predictable. Rao et al56 studied anorectal sensitivity and responses to rectal balloon distension in 12 patients with active disease proctitis, 11 patients with quiescent disease, six patients in both phases, and 12 controls. Both peak and steady-state rectal pressures were significantly higher in active disease, and rectal sensitivity and compliance normalized during inactive disease. However, when a saline solution was instilled into the rectum the infused volume retained until leakage first occurred, and the total volume retained were both significantly lower in patients with active colitis (P < 0.001) and in those with quiescent disease (P < 0.001) when compared with normal controls.57 Disease duration was similar between groups. Loening-Baucke et al57 performed anorectal manometry in 11 patients with active UC, seven with quiescent disease, and 18 healthy controls. Although the rectums of patients with quiescent colitis showed normal sensitivity and contractility, a significant decrease in the rectal wall compliance was noted, which, according to the authors, could have resulted from chronic changes in the rectal wall producing rectal stiffness.57
ALTERED COLONIC PERMEABILITY
There is accumulating evidence that colonic physiology is impaired in UC patients, as demonstrated indirectly by in vitro and in vivo permeability studies.59–61 The reduction in the efficiency of the epithelial barrier can, by itself, cause chronic mucosal inflammation and then could be considered as a primary involvement in the pathogenesis of UC, and indeed most studies support the idea that impaired intestinal permeability relates to colonic inflammation, since it correlates with disease activity.60–67 However, there is also evidence for an impaired permeability in UC patients in remission or with mucosal healing, thus indicating impaired gastrointestinal (GI) physiology independently from disease activity in these patients.60, 68–70 As an example, impaired transepithelial transport pathways have been identified irrespective of presence of inflammation and in quiescent disease.70, 71 If persistent inflammation can further impair colonic permeability or if more intensive treatment would prevent it is unknown.
There is evidence that in chronic UC the disease may extend proximally and may also be complicated by structural and functional damage beyond the mucosal layer, leading to strictures, giant pseudopolyposis, and bridging fibrosis, dysmotility, anorectal incontinence, and possibly impaired permeability. The true prevalence and relevance of these later complications in clinical practice is unknown; however, they provide the basis to accept that UC, as CD, should be regarded as a progressive disease. Earlier and more aggressive disease control might be a way of preventing or forestalling these consequences.
Until now, major endpoints of natural history studies of UC have been colectomy rates, cancer, disease activity, and disease extension. The deleterious consequences of ongoing inflammation in the colonic physiology have not been adequately studied. Being only a mucosal disease, UC is not typically accompanied by the stricturing and fistulizing complications that can be seen in CD; hence, there is a tendency among physicians to consider it a less progressive disease.72 This attitude results in a reluctance to introduce more potent treatments earlier in the course of disease, even if this approach has been shown to have better outcomes.73
In the modern era of endoscopy, evaluation of UC severity and activity is based on clinical symptoms and on endoscopic indices of inflammation. In most clinical trials and in clinical practice, therapeutic goals for UC have translated into achieving mucosal healing (MH). However, what is considered a healed mucosa by the current endoscopic indexes does not necessary translate into a healthy mucosa or maintenance of GI physiology. MH may not be a satisfactory therapeutic endpoint if it is only achieved after damage to the colon physiology has already occurred, as a consequence of undertreated disease.
In the future, new studies using technological advances in imaging and in motility should be conducted in order to assess the progression of bowel damage in UC and the impact of disease-modifying agents in preventing it.