Interval colorectal cancers (CRCs) may account for approximately one half of all CRCs identified during IBD surveillance. The etiology of interval CRCs is multifactorial, with procedural factors likely to play a major role. Molecular events promoted by inflamed mucosa may augment the cancer risk and perhaps explain some interval CRCs. This article reviews key studies relating to CRC risk in the patient with IBD, paying particular attention to the occurrence of interval CRCs. The most common factors implicated in the etiology of interval CRCs, in particular missed, incompletely resected lesions, the adherence to recommended surveillance intervals and biologic pathways associated with a faster progression to cancer are examined. Basic concepts for quality and effectiveness of colonoscopic surveillance in IBD are summarized.
Key points
- •
Interval colorectal cancers (CRCs) may account for approximately half of all CRCs identified during IBD surveillance, which highlights the need for improvements.
- •
The cause of interval CRCs is multifactorial, with procedural factors likely to play an important role.
- •
Molecular events promoted by inflamed mucosa may augment the cancer risk and perhaps explain some interval CRCs.
The past decade has witnessed considerable progress in the management of inflammatory bowel disease (IBD), including improvements in the quality and effectiveness of colonoscopic surveillance. Patients with ulcerative colitis (UC) or Crohn’s colitis have a greater risk of colorectal cancers (CRC), which may develop earlier and progress more rapidly than sporadic CRCs. Although most societies now endorse intensive colonoscopic surveillance to reduce the CRC risk, the efficacy of this strategy remains controversial. Several recent studies have cast doubt about the limited effectiveness of colonoscopy at reducing the incidence of sporadic CRC in the general population, especially in the proximal part of the colon, resulting in the occurrence of interval CRCs. Little is known, however, about the magnitude of this problem in patients with IBD and the most common explanations. Similar to the sporadic interval CRCs, two factors contribute to interval CRCs in IBD: clinician-dependent factors, such as missed, incompletely resected lesions or suboptimal surveillance; and molecular features of the inflamed mucosa underlying the development of cancer. The endoscopic knowledge, equipment, and techniques have evolved in recent years, contributing to a paradigm shift in the diagnosis and endoscopic resection of CRC precursors. The nonpolypoid (flat or depressed) colorectal neoplasms (NP-CRNs) play a significant role in the genesis of interval CRCs. Such subtle-appearing lesions are indeed more likely missed or incompletely resected endoscopically than their polypoid counterparts, and a subgroup of them harbor an aggressive biologic behavior.
This article provides insight into the magnitude and most common factors underlying the cause of interval CRCs during surveillance for IBD. Milestones of the literature regarding CRC risk in patients with IBD are reviewed. Specifically examined to the occurrence of interval CRCs are the contribution of missed, incompletely resected lesions; the adherence to surveillance; and distinct biologic features of the inflamed mucosa. Key principles are presented for ensuring the quality of IBD surveillance practice.
Incidence of CRC and interval CRC
A casual glance at the overall incidence of CRC in patients with IBD reveals discrepant outcomes, with a few studies showing similar CRC rates in patients with IBD versus the general population, whereas others show greater rates. In a nationwide cohort of close to 50,000 Danish patients with IBD who were followed over three decades (1979–2008), CRC was identified in 338 (0.71%) cases (268 in patients with UC and 70 in patients with Crohn’s disease). The overall CRC risk among patients with UC in this study was similar to that of the general population (relative risk, 1.07; 95% confidence interval, 0.95–1.21). In contrast, a North American study conducted from 1998 through 2010 found that the incidence of CRC in patients with Crohn’s disease or UC was 60% higher than in the general population.
The Danish study found a marked decline in the overall relative risk of CRC among patients with UC over the past decades, from 1.34 (95% confidence interval, 1.13–1.58) in 1979 to 1988 to 0.57 (95% confidence interval, 0.41–0.80) in 1999 to 2008, possibly reflecting refinements in the anti-inflammatory arsenal (ie, immunosuppressive therapy, biologicals), but perhaps also caused by a gradual adoption of CRC screening and surveillance. Conversely, the North American study found a fairly stable CRC rate in patients with IBD over time. Controversies surrounding the time-trends in CRC risk are not surprising, and likely reflect the cumulative effect of several factors, such as advancements in endoscope technology, a greater awareness, and improvements in the quality of colonoscopic performance.
As a common denominator, such epidemiologic studies lack relevant information about the disease duration, degree and extent of inflammation, presence of risk factors (ie, primary sclerosing cholangitis, personal or family history of CRC), and patients’ compliance with the recommended follow-up. Although clinical studies provide such details, most have focused on the optimal frequency of surveillance, paying less attention to the quality of examination. A systematic characterization of the lesions phenotype, in particular the location, size, shape, and histology, is often lacking.
Very few data are available about the occurrence of interval cancers during surveillance for IBD. The first paper dates back to 1982. In this surgical review of 676 patients with UC undergoing long-term follow-up, a total of 35 CRCs were identified. Twelve of these were diagnosed because of symptoms, 10 as incidental findings at proctocolectomy, and 13 CRCs were diagnosed during the follow-up at least 1 year after the initial UC diagnosis. This latter subgroup was referred to as “interval CRCs.” In a St Mark’s study reviewing the UC surveillance program over approximately three decades, a total of 74 patients (12.3% of the total population) developed neoplasms, including 30 CRCs. The authors defined interval CRCs as “cancers presenting after a negative index-colonoscopy or advanced (Dukes’ C/disseminated) cancers detected at surveillance.” During a median follow-up of 1.5 years, nine patients were identified with Dukes’ C cancers and four patients with disseminated cancers (4 of these 13 cases were diagnosed within 12 months). In three cases, CRC was diagnosed at colonoscopy because of symptoms; one of these was attributable to noncompliance. Of note, more than half (16 out of the 30) of the CRCs identified with this program were interval cancers, raising concerns about the effectiveness of colonoscopic cancer prevention. A statistically significant reduction in CRC rates over time was observed in this study ( r = −0.40; P = .04), especially in the proximal colon.
From these data, we can conclude that there is sparse understanding of the magnitude and clinical significance of interval CRCs in patients with IBD. Indeed, a wide variation exists with regard to the terminology used in endoscopy and pathology diagnostic protocols across countries, IBD centers, and studies.
Standardization of the nomenclature and clinical protocols, and uniformity in reporting on interval CRCs during IBD surveillance, would help to define quality targets. As a first step, a universal terminology is required for dysplasia and interval cancers. Previously used terms, such as flat dysplasia or dysplasia associated lesion or mass, need to be revisited. A rigorous description of the endoscopic shape and histologic features of the detected lesions is required, using international classifications (ie, Paris-Japanese endoscopic classifications and the World Health Organization histopathologic classifications ). Interval cancers should be considered those invasive cancers diagnosed after a negative screening examination, but before the next recommended follow-up colonoscopy, as endorsed by the current international IBD surveillance guidelines.
Incidence of CRC and interval CRC
A casual glance at the overall incidence of CRC in patients with IBD reveals discrepant outcomes, with a few studies showing similar CRC rates in patients with IBD versus the general population, whereas others show greater rates. In a nationwide cohort of close to 50,000 Danish patients with IBD who were followed over three decades (1979–2008), CRC was identified in 338 (0.71%) cases (268 in patients with UC and 70 in patients with Crohn’s disease). The overall CRC risk among patients with UC in this study was similar to that of the general population (relative risk, 1.07; 95% confidence interval, 0.95–1.21). In contrast, a North American study conducted from 1998 through 2010 found that the incidence of CRC in patients with Crohn’s disease or UC was 60% higher than in the general population.
The Danish study found a marked decline in the overall relative risk of CRC among patients with UC over the past decades, from 1.34 (95% confidence interval, 1.13–1.58) in 1979 to 1988 to 0.57 (95% confidence interval, 0.41–0.80) in 1999 to 2008, possibly reflecting refinements in the anti-inflammatory arsenal (ie, immunosuppressive therapy, biologicals), but perhaps also caused by a gradual adoption of CRC screening and surveillance. Conversely, the North American study found a fairly stable CRC rate in patients with IBD over time. Controversies surrounding the time-trends in CRC risk are not surprising, and likely reflect the cumulative effect of several factors, such as advancements in endoscope technology, a greater awareness, and improvements in the quality of colonoscopic performance.
As a common denominator, such epidemiologic studies lack relevant information about the disease duration, degree and extent of inflammation, presence of risk factors (ie, primary sclerosing cholangitis, personal or family history of CRC), and patients’ compliance with the recommended follow-up. Although clinical studies provide such details, most have focused on the optimal frequency of surveillance, paying less attention to the quality of examination. A systematic characterization of the lesions phenotype, in particular the location, size, shape, and histology, is often lacking.
Very few data are available about the occurrence of interval cancers during surveillance for IBD. The first paper dates back to 1982. In this surgical review of 676 patients with UC undergoing long-term follow-up, a total of 35 CRCs were identified. Twelve of these were diagnosed because of symptoms, 10 as incidental findings at proctocolectomy, and 13 CRCs were diagnosed during the follow-up at least 1 year after the initial UC diagnosis. This latter subgroup was referred to as “interval CRCs.” In a St Mark’s study reviewing the UC surveillance program over approximately three decades, a total of 74 patients (12.3% of the total population) developed neoplasms, including 30 CRCs. The authors defined interval CRCs as “cancers presenting after a negative index-colonoscopy or advanced (Dukes’ C/disseminated) cancers detected at surveillance.” During a median follow-up of 1.5 years, nine patients were identified with Dukes’ C cancers and four patients with disseminated cancers (4 of these 13 cases were diagnosed within 12 months). In three cases, CRC was diagnosed at colonoscopy because of symptoms; one of these was attributable to noncompliance. Of note, more than half (16 out of the 30) of the CRCs identified with this program were interval cancers, raising concerns about the effectiveness of colonoscopic cancer prevention. A statistically significant reduction in CRC rates over time was observed in this study ( r = −0.40; P = .04), especially in the proximal colon.
From these data, we can conclude that there is sparse understanding of the magnitude and clinical significance of interval CRCs in patients with IBD. Indeed, a wide variation exists with regard to the terminology used in endoscopy and pathology diagnostic protocols across countries, IBD centers, and studies.
Standardization of the nomenclature and clinical protocols, and uniformity in reporting on interval CRCs during IBD surveillance, would help to define quality targets. As a first step, a universal terminology is required for dysplasia and interval cancers. Previously used terms, such as flat dysplasia or dysplasia associated lesion or mass, need to be revisited. A rigorous description of the endoscopic shape and histologic features of the detected lesions is required, using international classifications (ie, Paris-Japanese endoscopic classifications and the World Health Organization histopathologic classifications ). Interval cancers should be considered those invasive cancers diagnosed after a negative screening examination, but before the next recommended follow-up colonoscopy, as endorsed by the current international IBD surveillance guidelines.
Potential etiologic factors of interval CRCs
Similarly to sporadic CRCs, most interval CRCs in IBD probably can be explained by clinician-dependent factors, such as missed, incompletely resected lesions or deviation from surveillance protocols. The understanding of the underpinnings of such interval CRCs is of importance because it may permit identification of modifiable factors, for example gaps in knowledge and training on the recognition of nonpolypoid neoplasms and their endoscopic resection. In this case, tailored educational programs would improve the awareness and help to shape practical skills, to ultimately safeguard the quality of colonoscopy. Furthermore, it is important to understand whether certain molecular features of the inflamed mucosa could augment the risk of cancer progression. Such information may help to develop personalized (ie, molecular-based) surveillance strategies.
Missed Lesions
Two recent studies exploring the cause of sporadic interval CRCs in the general population found missed lesions represent by far the most important contributor (>50% of all interval CRCs). Undoubtedly, missed lesions are likely to account for a significant proportion of interval CRCs in IBD, although a thorough analysis using structured algorithms has not yet been performed. A recent population-based analysis by Wang and colleagues, using SEER cancer registry data from 55,008 older patients with CRC, found rates of early/missed CRCs were three-fold greater in IBD than in patients without IBD (15.1% for Crohn’s disease, 15.8% for UC vs 5.8% for patients without IBD; P <.001). Early/missed CRCs were defined as CRCs identified within 6 to 36 months after a colonoscopic examination that did not detect cancer. This study was based on administrative data, and therefore lacked detail about the completeness of colonoscopy, bowel preparation, extent of colitis, characteristics of mucosal lesions identified at the baseline examination, and resection outcomes. Such observations underscore the importance of meticulous inspection of the entire colonic mucosa, which should be ideally clean and free of inflammation, and the need for formal training of the endoscopist in the recognition of IBD neoplasms. Presence of active or chronic background inflammation and the diversity in endoscopic appearance of dysplasia by IBD may, however, increase the complexity of diagnosis. Fig. 1 illustrates a lateral spreading tumor of granular subtype, which could have been missed at a previous examination.
A substantial number of studies demonstrated that indigo carmine– or methylene blue–guided chromoendoscopy (CE) improves the diagnostic yield of dysplasia and invasive CRC during IBD surveillance. This is not surprising, because a significant proportion of dysplastic lesions in patients with IBD appear to have a flat appearance, as illustrated in Table 1 . Pancolonic CE delineates the borders and permits a detailed analysis of the epithelial surface, thus facilitating the diagnosis of subtle lesions and their endoscopic resection. A few meta-analyses now demonstrate CE with targeted biopsies is superior to white-light colonoscopy with random biopsies in the detection of dysplasia and invasive CRCs. CE yielded a 7% increase in the detection of any dysplasia. Compared with white-light colonoscopy with random biopsies, the likelihood to detect any dysplasia with CE and targeted biopsies was 8.9-fold greater, and 5.2-fold greater for detecting nonpolypoid dysplasia. In a Mainz study of 165 patients with long-standing UC who were randomized to undergo standard colonoscopy using white light versus CE (0.1% methylene blue), significantly more intraepithelial neoplasms were detected in the CE group (32 vs 10; P = .003). CE detected more intraepithelial neoplasms in “flat mucosa” than white-light endoscopy (24 vs 4; P = .0007), and more invasive cancers (3 vs 1).
Related posts:
Learning to Detect Flat Neoplastic Lesions in Inflammatory Bowel Disease
Quality Bowel Preparation for Surveillance Colonoscopy in Patients with Inflammatory Bowel Disease Is a Must
Image-Enhanced Endoscopy Is Critical in the Surveillance of Patients with Colonic IBD
The Unique Pathology of Nonpolypoid Colorectal Neoplasia in IBD
Toward a Consensus on Endoscopic Surveillance of Patients with Colonic Inflammatory Bowel Disease
Surgical Management of Nonpolypoid Colorectal Lesions and Strictures in Colonic Inflammatory Bowel Disease
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
