Too much of a good thing? ▸

Patients with multiple adenomas or those ⩾1 cm are at increased risk of subsequent colorectal cancer, and surveillance colonoscopy every three years can reduce this risk. There are increasing pressures being placed on colonoscopy services with the introduction of population screening and it is important to use this investigation appropriately. The survey reported by Mysliwiec et al is therefore worrying as it suggests that a considerable proportion of colonoscopy surveillance being performed is unnecessary. There was an 83% response rate, with 317 gastroenterologists and 125 surgeons undertaking surveillance colonoscopy; 24% of gastroenterologists and 54% of surgeons stated they performed surveillance colonoscopy for patients with hyperplastic polyps. Furthermore, 43% of surgeons recommended surveillance at an interval of less than three years for this group of patients. Hyperplastic polyps occur in 10% of patients colonoscoped and future colorectal cancer risk is similar to the general population. This survey was taken from a geographically representative sample of US clinicians but the results may be biased in either direction. For example, respondents may overestimate the colonoscopy rates to exaggerate workload or underestimate investigations that do not comply with guidelines. The “fee for service” type funding of US health care might encourage inappropriate surveillance colonoscopy so it would be interesting to conduct a similar survey among UK colonoscopists.

Barrett’s: the cutting edge? ▸

Oesophagectomy for high grade intraepithelial neoplasia (HGIN) or superficial adenocarcinoma in Barrett’s carries risks, especially in the elderly or unfit, yet data on endoscopic mucosal resection (EMR) in Barrett’s are limited, with little long term follow up and the problem of managing the remaining Barrett’s mucosa. Two studies address some of these issues. In the first, the results of EMR in 295 patients are described; 31% of lesions were >2 cm and most were removed in one piece using the “suck and ligate” method. Specimens were analysed for completeness of resection (basally and laterally) as well as depth of invasion; 80% of lesions were intramucosal carcinomas (IMC), 16% invaded the submucosa, and the remainder were dysplastic. Only 3.5% had lymphatic invasion and all bar one of these occurred in submucosal cancers. Similarly, positive basal resection margins were virtually only seen in this group in the submucosa. Only 27% of lesions were completely resected after one session, rising to 74.5% after two procedures, but better EMR equipment and techniques may further improve these impressive results. The procedure was safe and effective, and allowed definitive detection of those with submucosal involvement who require further therapy.

How to tackle the remaining Barrett’s mucosa after resection of focal lesions was addressed in the second study, describing circumferential EMR in 21 patients with HGIN or IMC. The lesion and surrounding hemicircumference of Barrett’s were resected piecemeal and one month later the other hemicircumference was resected, in procedures lasting an average of 45 minutes (under propofol). Complete resection was achieved in 83% (although two late recurrences were seen) and all Barrett’s epithelium was eradicated in 75%. Immediate bleeding occurred in four patients and settled with endoscopic therapy, no late bleeding or perforations occurred, and at 18 months no oesophageal strictures had been encountered. This promising technique offers yet another means of eradicating HGIN and superficial cancer in Barrett’s and in future more endoscopists will need to become competent in EMR techniques.

Colour me beautiful ▸

The bile pigment bilirubin has potent antioxidant properties. Biliverdin reductase catalysed redox cycling of unconjugated bilirubin may be important for cytoprotection against oxidative injury. The evolutionary development of bilirubin metabolism (that consumes significant amounts of energy) as well as high prevalence of polymorphism leading to Gilbert’s syndrome suggests that a modest elevation of bilirubin might be adaptive. Zucker et al analysed the 3rd National Health and Nutrition Examination Survey (NHANES III) to determine the demographics and correlates of serum bilirubin levels in the general population. In a weighted analysis representing 176 748 462 subjects, bilirubin levels were significantly higher in men and were reduced in active smokers. Baseline bilirubin concentrations were significantly lower in individuals reporting prior non-dermatological cancer. Lower bilirubin level was strongly associated with a history of gastrointestinal cancers. An increase in serum bilirubin level of 1.0 mg/dl was associated with a markedly lower prevalence of colorectal cancer with an odds ratio of 0.295 (95% confidence interval (CI) 2.91–2.99) in men and OR of 0.186 (95% CI 0.183–0.189) in women.

Based on these fascinating findings, the authors hypothesise that unconjugated bilirubin suppresses carcinogenesis. In case of colon cancer, this could be due to passive diffusion of unconjugated bilirubin from serum across the intestinal mucosa. But the NHANES III database did not stratify serum bilirubin into conjugated and unconjugated fractions. One should also not assume that such associations demonstrated in a cross sectional database are causal. However, chemopreventative function of bilirubin certainly warrants further investigation.