The best magazine
Abstract and Introduction
Abstract
Objectives: To evaluate the yield and clinical impact of random biopsies taken during colonoscopic surveillance of patients with longstanding ulcerative colitis (UC).
Methods: Retrospective analysis of 1,010 colonoscopies performed from 1998 to 2008. Colonoscopy and pathology reports were reviewed to assess the yield and clinical impact of random biopsies. In total, 475 patients with UC who underwent colonoscopy at the Academic Medical Centre Amsterdam were included in this study. The main outcome measures are neoplasia yield per-colonoscopy and clinical impact per-patient of random biopsies.
Results: Of all colonoscopies, 466 were performed for surveillance (in 167 patients) during which 11,772 random biopsies were taken (median 29). Overall, neoplasia was detected in 88 colonoscopies (53 patients): in 75 colonoscopies (85%) by targeted biopsies only and in 8 (9.1%) by both targeted and random biopsies. Neoplasia was detected in random biopsies only in five (5.7%) colonoscopies in four (7.5%) patients. Two of these four patients with neoplasia detected only by random biopsies had visible neoplasia in previous colonoscopies. One patient had unifocal low-grade intraepithelial neoplasia (LGIN) that could not be confirmed in three subsequent colonoscopies. The last patient had multifocal LGIN and suspicious appearing ulcerations. Proctocolectomy confirmed the presence of neoplasia.
Conclusions: The yield of random biopsies is low whereas UC-associated neoplasia is macroscopically visible in 94% of colonoscopies. During 10-year surveillance, neoplasia was detected in only random biopsies in four patients of whom only one had clinical consequences. The low yield and lack of clinical consequences from random biopsies in this high-risk population raise questions about the necessity and cost-effectiveness of their routine use during UC surveillance.
Introduction
As the risk of colorectal cancer is increased in patients with longstanding ulcerative colitis (UC), guidelines recommend colonoscopic surveillance in these patients in order to detect neoplasia at an early and potentially curable stage. Surveillance implies performing biannual colonoscopies, even annually in case of concomitant primary sclerosing cholangitis (PSC). During these colonoscopies targeted biopsies of suspicious lesions and random sampling of inconspicuous appearing mucosa for pathological evaluation should be performed. The rationale for taking random biopsies is that UC-associated neoplasia may be difficult to visualize. To obtain a high sensitivity for the detection of neoplasia, four random biopsies every 10 cm of colon are presumed to be necessary.
Recent reports, however, suggest that most UC-associated neoplasia is in fact macroscopically visible. Advanced endoscopic techniques have also demonstrated improved neoplasia detection compared with conventional colonoscopy. Pancolonic dye spraying has shown to be superior to conventional colonoscopy in several reports, and recently autofluorescence imaging has shown promising results as well. Remarkably, when using advanced techniques for surveillance the yield of random biopsies diminishes as a result of which their need is more and more questioned.
In addition, Rutter et al. suggested that a number of endoscopic features could help to predict an increased risk of neoplasia in individual patients. Certain endoscopic appearances that were a result of severe inflammation in the past (e.g., post-inflammatory polyps) were associated with neoplasia. The authors even postulated to reduce the surveillance interval in case no endoscopic features were present. This raises the question whether random biopsies may be reserved only for patients with endoscopic features of previous severe inflammation.
The aims of this retrospective study were to evaluate (i) the yield of neoplasia from random biopsies during a 10-year surveillance program of patients with UC in a tertiary referral center in the Netherlands, (ii) to determine whether random biopsies had clinical consequences, and (iii) to validate whether endoscopic features were associated with neoplasia in our UC population.
Source: ...