Diagnosis and Monitoring of Ulcerative Colitis

Abstract

Ulcerative colitis is one of the two main subtypes of inflammatory bowel disease, along with Crohn's disease. Understanding the clinical and endoscopic features of ulcerative colitis is critical in achieving a timely diagnosis. An initial evaluation includes assessing clinical symptoms, inflammatory markers, endoscopic findings, and determination of the presence or absence of extraintestinal manifestations. Initial disease management should consider disease severity at the time of diagnosis as well as prognostication, or the determination of risk factors present with a high likelihood of severe disease in the future. Once appropriate therapy has been initiated, ongoing monitoring is crucial, which may include repeated clinical assessments over time, measuring noninvasive markers of inflammation, and endoscopic and histologic reevaluation. An important aspect of disease monitoring in ulcerative colitis is dysplasia surveillance; there are many patient-specific risk factors which influence surveillance strategies. Utilizing appropriate surveillance techniques is necessary for early detection of dysplasia and colorectal neoplasia.

Diagnosis of Ulcerative Colitis

Clinical, Laboratory, and Radiographic Diagnosis

Clinical, laboratory, endoscopic, and histopathologic information are used to establish a diagnosis of ulcerative colitis (UC), one of the two main subtypes of inflammatory bowel disease (IBD), along with Crohn's disease (CD). ¹ UC is a chronic, relapsing, and remitting inflammatory condition of the gastrointestinal tract that presents most commonly in early adulthood with peak incidence between ages 15 and 30 years old. ^{2 3} Symptoms such as bloody diarrhea, mucous in the stool, urgency, tenesmus, and abdominal cramping may be consistent with UC. ^{4 5} On initial presentation, health care providers should assess for the frequency of bowel movements, severity of bleeding, presence of nocturnal bowel movements, and degree of weight loss to help determine disease severity. ⁶ Extraintestinal manifestations (EIMs) of UC should also be evaluated at time of index clinical presentation such as fever, joint swelling, cutaneous manifestations, ocular symptoms, or anemia. ^{7 8} A history should be obtained to evaluate potential contributors to UC such as smoking cessation, nonsteroidal anti-inflammatory use, and enteric infections such as Clostridioides difficile infection (CDI). ^{9 10 11}

No singular laboratory test is diagnostic for UC. However, checking inflammatory markers can be helpful for prognostication and disease monitoring, although they are nonspecific. C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) may correlate with endoscopic severity and be predictive of need for colectomy; however, up to 25% of patients will have normal CRP during a flare. ^{12 13 14} Fecal leukocytes are neither sensitive nor specific for monitoring UC. Fecal calprotectin (FC), a marker of neutrophilic inflammation in the gut, is helpful in distinguishing inflammatory etiologies, such as UC, from noninflammatory causes like irritable bowel syndrome. FC is also a noninvasive marker utilized to assess response to therapy in established UC. ^{13 15 16 17 18} Checking hemoglobin and albumin as surrogate markers of inflammation are also useful for assessing severity of disease, prognosis, and response to medical therapy. ¹⁹

Serologic markers including antineutrophil cytoplasmic antibodies (ANCA), perinuclear ANCA (pANCA), and anti- Saccharomyces cerevisiae antibodies (ASCA) have been explored as aiding in the diagnosis of UC. These tests are commonly found in laboratory panels marketed as IBD panels. A positive ANCA or pANCA with negative ASCA can be found in up to two-thirds of patients with UC and have been associated with treatment refractory disease. However, sensitivity is poor therefore there is currently no indication for testing ANCA, pANCA, or ASCA to establish diagnosis. ^{20 21}

Excluding infectious mimickers of UC may also be indicated in the correct clinical scenario. Infections associated with bloody diarrhea such as Escherichia coli O157:H7, Salmonella , Shigella , Yersinia , Campylobacter , and parasitic infections can be tested using stool studies such as polymerase chain reaction (PCR)-based assays. Certain enteric infectious may exist concomitantly along with newly diagnosed or relapsing UC including CDI; this carries an increased risk of hospitalization, longer length of stay, higher mortality, and risk of colectomy. ^{22 23 24} Thus, testing for CDI using two-step PCR-based enzyme-linked immunosorbent assay or antigen testing modalities should be done at initial presentation and any subsequent flares. ^{6 25}

Radiographic modalities provide little diagnostic yield for luminal manifestations of UC as compared to endoscopic evaluation. Imaging the small bowel with computed tomography (CT) or magnetic resonance imaging (MRI) is not routinely recommended unless there is a clinical suspicion for CD or complex intra-abdominal complication. If there is clinical concern for toxic megacolon, a potential surgical emergency abdominal X-ray imaging may be performed quickly in order to assess colon diameter. ²⁶ Imaging can be helpful for diagnosing other disease complications in emergent settings as well as certain EIMs such as primary sclerosing cholangitis (PSC). ²⁷

Endoscopic and Histopathologic Diagnosis

Endoscopy is an important diagnostic tool used to distinguish UC from CD, and other non-IBD-related colitides, evaluate extent of disease, and obtain biopsies. Colonoscopy with intubation of the terminal ileum is standard for diagnosis with histologic confirmation. Routine upper endoscopy is not required for diagnosis of UC. ²⁸ Colonoscopy is considered safe with a low risk of adverse events. Relative contraindications include toxic megacolon and severe colitis; in these scenarios, diagnostic sigmoidoscopy may be considered. ²⁸

UC has a distinct endoscopic appearance with continuous mucosal inflammation extending proximally from the rectum and may feature granularity, loss of vascular markings, friability, deep ulcerations, and spontaneous bleeding. ²⁸ The described extent of disease is most frequently based off on the Montreal Classification, which delineates proctitis, left-sided disease, and extensive colitis (defined as extension beyond the proximal splenic flexure). ^{29 30} Establishing pretreatment extent of disease is imperative for subsequent management decisions as well as evaluating response to medical therapies.

Other endoscopic features of UC may include a cecal patch or isolated area of inflammation around the appendiceal orifice in the setting of otherwise endoscopically normal right colon. This should not be confused with the skip segments of CD. The presence of a cecal patch does not appear to have a more severe or refractory course and may occur in up to 20% of patients on index diagnostic colonoscopy. ³¹ Backwash ileitis can also be seen in up to 25% of patients with UC pancolitis and should not be misinterpreted as ileal involvement of CD. Backwash ileitis is seen with extensive UC with cecal involvement of inflammation and endoscopically appears as a short, contiguous segment of mildly inflamed ileum without stricture, stenosis, or significant ulcerations. The inflammation appears to be a continuation of the colitis in the cecum. Lastly, less than 5% of patients with UC have rectal or relative rectal sparing, meaning disease is more severe proximally than in the rectum. This is most frequently noted in pediatric patients and data is lacking regarding the subsequent clinical course. ^{32 33 34 35}

Endoscopic evaluation can underestimate the extent of disease compared to the histology. On index pretreatment colonoscopy, biopsies are typically obtained from each segment of the colon as well as the terminal ileum. Biopsies should include abnormal and normal tissue for pathology to review and assess for microscopic signs of active and chronic inflammation. ²⁸

As UC is a chronic inflammatory condition limited to the mucosa, pathologic signs of activity and chronicity aid in its diagnosis. Active inflammation is identified as the presence of epithelial injury with neutrophil infiltrating crypt epithelium (cryptitis), crypt abscesses, or neutrophilic infiltration of surface epithelial with or without ulceration. Histologic features of chronicity include crypt architectural distortion, basal plasmacytosis, increased cellularity of the lamina propria, and Paneth cell metaplasia. Evidence of chronicity can help distinguish UC from acute self-limiting infectious colitis, which tends to have preserved architecture. ^{36 37} Ischemic colitis can histologically appear similar to IBD with edema, distorted crypts, mucosal and submucosal hemorrhage, inflammatory infiltration in the lamina propria, granulation tissue, and necrosis. However, ischemic colitis is endoscopically segmental and rarely involves the rectum due to dual blood supply. ³⁸

Disease Activity Measures for Diagnosis

Several scoring systems and classifications have been used to risk stratify UC activity and severity. Classically, the Truelove and Witts' criteria have been used to assess mild versus severe disease activity. Using this criteria, mild colitis was defined at < 4 bowel movements daily, normal temperature and heart rate, hemoglobin > 11 g/dL, and ESR < 20 mm/h. Severe disease was defined as > 6 bowel movements per day with fever, tachycardia, anemia, or ESR < 20 mm/h. ³⁹ However, this scoring system did not include endoscopic data or other pertinent clinical patient-reported outcomes (PROs). Other scoring systems have been created including the Mayo Score for endoscopic severity and UC activity, ⁴⁰ UC Endoscopic Index of Severity (UCEIS), and the Simple Clinical Colitis Activity Index (SCCAI), ⁴¹ Seo Index, ⁴² and Proposed American College of Gastroenterology UC Activity Index. ⁶ Although most scoring systems are not validated and there is often interobserver variance endoscopically, they generally correlate to endoscopic disease activity. ⁴³ The Seo Index uses laboratory and clinical data while the SCCAI uses exclusively patient symptoms. The UCEIS and Mayo endoscopic subscore (MES) use endoscopic data. The Proposed American College of Gastroenterology UC Activity Index utilizes UCEIS with the MES in addition to clinical and PROs. Clinically, the total Mayo Score and MES are most commonly used.

Distinguishing CD from UC

While CD and UC are both IBDs with distinct features, they may be mistaken for each other. Distinguishing CD from UC prior to initiation of therapy can be helpful in determining best treatment options and prognostication, specifically if surgery is warranted, as well as aid in proper disease monitoring given their differing potential complications and natural history. Clinical, endoscopic, and histologic diagnostic information may aid in making this distinction. In one prospective study, index colonoscopy was accurate in distinguishing CD from UC in up to 89% of cases. ⁴⁴ While UC generally features diffuse and contiguous inflammation from the rectum extending proximally and limited to the colon, CD has more variable endoscopic phenotypes which may include small bowel or upper gastrointestinal involvement. However, when CD is limited to the colon and granulomas are not present on biopsy, these two diseases can be more difficult to distinguish. Endoscopic features on colonoscopy more consistent with CD include discontinuous inflammation with areas of normal bowel, rectal sparing, involvement of the terminal ileum, identification of a fistulous tract, and anal or perianal disease. ⁴⁵ Other endoscopic features suggestive of CD include aphthous ulcers, deep ulcers, serpiginous ulcers, and cobblestoning. Conversely, UC features include granularity, loss of vascularity, and usually demonstrates continuous inflammation without skip lesions proximal to the anal canal. ²⁸

Histologically, both CD and UC have evidence of chronicity underlying active inflammation. Granulomas can be seen with CD, however, is not pathognomonic. Granulomas can be seen in tuberculosis, fungal infections, sarcoidosis, and even in severe UC with pericrypt granulomas. Granulomas in the lamina propria not associated with crypt injury are associated with CD and may be seen in 13 to 55% of biopsies. ^{46 47} Active ileal involvement is also associated with CD, although this should be distinguished from backwash ileitis of UC. Crohn's ileitis would have the absence of contiguous pancolitis and discrete ulcers or strictures can be seen in the terminal ileum and ileocecal valve. ⁴⁵ Features that favor a diagnosis of CD rather than UC with backwash ileitis include length of involved small bowel, greater inflammatory activity, and mucosal injury in the ileum than the cecum, and transmural ileal inflammation with granulomas. ⁴⁸

Prognosis of Ulcerative Colitis

Risk of Colectomy and Prognostication of Disease Course

UC is not associated with increased mortality. However, up to 10% of patient with UC require colectomy within the first 10 years of diagnosis, though that number is decreasing, especially for nonemergent colectomies. ^{49 50} Risk factors for colectomy should be assessed at time of diagnosis. ⁷ Extensive colitis, age younger than 40 at diagnosis, need for systemic steroids, elevated CRP/ESR, UC-related hospitalization, and prior CDI or cytomegalovirus infection are associated with increased risk of colectomy and poor prognosis. ^{49 51} The greater the number of high-risk features, the increased likelihood of colectomy. ²

Other Disease-Related Complications and Extraintestinal Manifestations

Anemia is one of the most common systemic complications of UC. Approximately 20 to 25% of UC patients are anemic at the time of diagnosis. The cause is often multifactorial including iron deficiency, vitamin B12 deficiency, and anemia of chronic inflammation. Anemia has been found to be independently linked to elevated CRP levels and disease extent. ^{52 53} Iron deficiency is likely caused by a combination of decreased intestinal absorption and chronic gastrointestinal bleeding from active UC. Identifying and treating anemia with iron supplementation may be helpful to improve patient quality of life. ⁵⁴

Rates of venothrombolic events (VTEs), including deep vein thrombosis and pulmonary embolism, are increased in patients with UC, and to a greater degree than in CD. Studies have shown a 1.5 to 3.5 times increased risk of VTE compared to non-IBD patients. ⁵⁵ Disease activity, extensive disease, fulminant colitis, use of corticosteroids, hospitalization, and smoking all increase the risk of VTE. ^{2 56} Low molecular weight heparin is used as a pharmacologic prophylaxis for hospitalized patients and is safe; it does not worsen gastrointestinal hemorrhage even when rectal bleeding is present.

EIMs include disease involvement outside of the gastrointestinal tract. Due to variance in defining EIMs, the range of prevalence is anywhere from 6 to 47%. ⁵⁷ EIMs tend to affect patients with early-onset IBD and younger patients. They can present before or after diagnosis of IBD. Musculoskeletal EIMs are the most common. ^{57 58} They can be subdivided into peripheral arthritis and axial arthritis/spondyloarthritis. Peripheral arthritis occurs in 5 to 14% of UC patients and can be broken down into pauciarticular and polyarticular arthritis. Pauciarticular arthritis affecting, fewer than five joints, involving preferentially large joints (ankles, knees, hips, wrists, elbows, and shoulders), and typically occurs with active IBD. Polyarticular arthritis occurs independent of IBD inflammation and occurs in five or more joints, is symmetric mainly affecting small joints of both hands with pain, swelling, or effusion, and tends to be more chronic. Axial arthritis includes sacroiliitis and ankylosing spondylitis which can precede diagnosis of IBD. ^{58 59}

Cutaneous manifestations occur in 5 to 15% of patients with IBD; erythema nodosum (EN) and pyoderma gangrenosum (PG) are the most common. ⁶⁰ EN is described as a raised, red or violet, tender lesions typically found on the extensor surfaces of lower extremities. In patients with UC, EN was found in 2% of patients with inactive disease and 4.7% of patients with active disease. ⁵⁸ PG is described as an erythematous pustule or nodule which may rapidly develop into deep ulcers with irregular violaceous edges and purulent material in the ulcer ground, which is sterile on culture. In patients with UC, PG was reported in 1.5% patients with inactive disease and 3% of patients with active disease. ⁵⁸ The diagnosis should be made clinically as skin biopsy can worsen pyoderma.

Ocular EIMs are the third most common EIM after musculoskeletal and cutaneous involvement. Scleritis, episcleritis, and anterior uveitis are the most common of these. Ocular involvement is more common in CD than in UC; however, it still can occur especially in conjunction with other joint or skin EIMs. ⁶¹ Episcleritis and scleritis tend to follow intestinal inflammation. Scleritis can lead to vision loss. Anterior uveitis generally occurs independent of IBD activity.

PSC is the primary hepatobiliary-related EIM. In patients with PSC, 60 to 80% also have UC and up to 5% of UC patients have PSC. ⁶² Elevated alkaline phosphatase and gamma-glutamyl transferase may raise suspicion for PSC and it can be radiographically seen on MRI/magnetic resonance cholangiopancreatography. Identifying PSC in UC patients is critical as it is progressive and has significantly greater morbidity and mortality than UC alone. The median survival without liver transplant is 10 to 12 years after diagnosis. ⁶² Concomitant PSC and UC increases the risk of colorectal cancer and impacts screening protocols for colorectal cancer, cholangiocarcinoma, and gallbladder cancer.

Monitoring Disease Activity in Ulcerative Colitis

Appropriate and timely disease monitoring is critical in the successful management of patients with IBD. According to the most recent Selecting Therapeutic Targets in Inflammatory Bowel Disease-II (STRIDE-II) expert consensus guidelines, the most important short-term target of therapy is clinical response, which they define as at least 50% improvement in rectal bleeding and stool frequency. ¹⁶ Achievement of clinical and endoscopic remission remain among the top long-term goals for treatment of patients with IBD. ¹⁶ Clinical remission is defined as absence of rectal bleeding and normal stool frequency; definitions of endoscopic remission are detailed below. Other long-term targets include normalization of serologic inflammatory markers such as CRP and ESR, normalization of stool inflammatory markers such as FC, transmural healing, and histologic healing. ¹⁶ However, these targets are generally considered secondary outcomes that are not as routinely targeted as clinical remission and endoscopic remission, although achievement of these targets are associated with improved long-term clinical outcomes. ^{63 64}

Treat-to-target, or the intentional fulfillment of objective clinical goals such as clinical response, clinical remission, and endoscopic remission, is thought to improve clinical outcomes in IBD. In UC specifically, systemic reviews of natural histories of the disease have estimated the 5- and 10-year incidence of colectomy at 10 to 15%; however, studies have reported lower rates of colectomy if mucosal healing was achieved. ^{2 65 66 67 68} Other clinical outcomes associated with achievement of clinical and endoscopic remission include decreased rates of hospitalization and sustained clinical remission. ^{65 66 69} Thus, if the predetermined targets have not been achieved, consideration of change in medical (or surgical) management is advised. ¹⁶

Regardless of the specific clinical scenario, the key to disease monitoring is close and ongoing clinical follow-up which may include a combination of symptom monitoring, physical exam, laboratory monitoring, endoscopic evaluation, and radiographic assessment. Symptom monitoring by itself is insufficient for clinical monitoring given the possibility of ongoing mucosal disease in the setting of clinical response or remission. ⁷⁰ In general, clinical monitoring should be more frequent if the patient has not yet achieved their targets, but may reasonably be performed less frequently for a stable and well-controlled patient who has achieved their clinical targets. Biomarkers such as FC have been proven to correlate with endoscopic mucosal lesions as well as histologic activity, and therefore may have a role in noninvasive monitoring of disease as they may be performed more frequently than radiography or invasive disease monitoring methods such as endoscopy. ^{71 72 73}

Multiple disease activity indices are widely used for the purposes of clinical monitoring. The partial Mayo Score, which does not include the endoscopic component, is also widely utilized for this purpose. ^{72 73} The MES, as well as the UCEIS may be used to document and track endoscopic evolution over time in patients with UC. Endoscopic remission as defined in STRIDE-II is either MES score 0 or UCEIS ≤ 1. ^{16 74 75} Timing of repeat endoscopic evaluation after a new therapy or change in therapy is generally suggested to be performed at around 6 months; however, may be reasonably performed earlier or later depending on patient-specific clinical circumstances. ⁷⁶

Radiographic assessment, including intestinal ultrasound and cross-sectional imaging such as CT or MRI, are adjunctive tools for monitoring disease activity in UC; however, access, radiation exposure, and cost may be determinative factors against their routine use for some. ⁷⁷ Intestinal ultrasound is noninvasive, portable, lacks radiation exposure, and has been demonstrated to correlate with endoscopic activity, however, is not widely available nor therefore performed routinely in the office, and also may be operator-dependent. ^{77 78} Certain features on CT enterography have been demonstrated to correlate with endoscopic activity but radiation exposure limits the use of this as a form of disease activity monitoring. ^{77 79 80} MR enterography has been demonstrated to correlate well with endoscopic activity, however, access and cost limit its use. ^{77 81}

Monitoring for Colorectal Neoplasia in Ulcerative Colitis

Monitoring for the development of colorectal neoplasia (CRN) is a critical component of the long-term management of UC given that UC confers an increased risk of CRN due to ongoing mucosal damage from inflammation. ⁸² By some accounts, the rates of CRN have been decreasing over time, however, other studies report stable rates. ^{82 83} There exist many risk factors for the development of CRN in UC including male sex, smoking, duration of disease (longer duration correlating to higher risk), anatomical disease extent (more extensive disease correlating to higher risk), concomitant PSC, personal history of dysplasia, endoscopic features such as strictures, possibly pseudopolyps (most likely as a surrogate marker of past inflammation), tubular colon, and family history of CRN. ^{84 85 86 87 88 89 90 91} Overall disease severity and cumulative inflammatory burden are emerging concepts in the risk factors for CRN in UC as well. ^{92 93}

Per current clinical guidelines by the American Gastroenterological Association (AGA), screening should start 8 to 10 years after the diagnosis of UC or immediately in patients with UC who have been diagnosed with PSC. ⁹⁴ In those patients with UC and PSC, annual screening is generally recommended. In those patients with UC alone, repeated screening is recommended to be performed every 1 to 3 years except in the case of isolated proctitis. ⁹⁵ Many patient-specific factors may influence screening intervals including the presence or absence of risk factors for dysplasia as discussed above.

The gold standard for CRN screening in UC is colonoscopy; other tests offered for colorectal cancer screening in the general population such as stool tests (e.g., fecal immunohistochemical tests) or CT colonography are not appropriate in this patient population that have a significantly increased risk of CRN compared to the general population. ⁸² Appropriate screening techniques include high-definition endoscopy, dye spray chromoendoscopy with indigo carmine or methylene blue (particularly if only standard-definition endoscopes are available), and virtual chromoendoscopy, according to the AGA. ⁹⁴ Biopsies of any abnormal mucosa or lesions suspicious for dysplasia are recommended, in addition to random biopsies of colonic mucosa every 10 cm in a four-quadrant fashion. The goal of random biopsies is to detect invisible dysplasia although advancements in image quality during endoscopy has thrown this entity into question. ⁹⁵