Background: Primary sclerosing cholangitis (PSC) is the strongest risk factor for colorectal neoplasia (CRN) in inflammatory bowel disease (IBD). While prior studies in this population have estimated the prevalence of advanced CRN (aCRN) (high-grade dysplasia or colorectal cancer), little is known about the incidence rate of aCRN after a diagnosis of low-grade dysplasia (LGD) to, and its potential risk factors.
Methods: PSC-IBD patients were identified from two existing large surveillance databases (the Mount Sinai Hospital Surveillance Database (2005–2015) and the Dutch multicentre surveillance database (2000–2013)) and compared to non-PSC IBD patients. All patients had undergone at least two surveillance colonoscopies. Clinical information, as well as endoscopic and histologic data were recorded. The prevalence of LGD and aCRN and the incidence of aCRN after an index LGD lesion (first LGD within study period) were compared between groups. Cox-regression was used to determine predictors of dysplasia progression.
Results: 301 patients with PSC-IBD were compared to 1100 non-PSC IBD patients. Patients with PSC-IBD were younger at IBD diagnosis [median 24 y (range 69) vs 28 y (72), p=0.003], more frequently male (69% vs 50%, p<0.001), and had a shorter IBD duration (median 10y  vs 14y , p<0.001). Median time of follow-up for the total cohort was 4.8 years. PSC-IBD patients had a statistically significant increase in the frequency of aCRN as compared to non-PSC IBD patients (Hazard ratio [HR] 3.1, 95% CI 1.7–5.8). The overall incidence rate of developing aCRN in PSC-IBD compared to non-PSC IBD patients was 1.3 versus 0.4 per 100 patient-years follow-up (pty). Despite similar frequencies of LGD between PSC-IBD and non-PSC IBD patients (20.3% versus 21%, p=0.8, Table 1), the rate of developing aCRN following detection of LGD was higher in PSC-IBD patients (HR 2.8, 95% CI 1.2–6.5). The incidence rate of aCRN after a diagnosis of LGD was 7.4/100 pty for PSC-IBD patients compared to 2.3/100p ty for non-PSC IBD patients. Using Cox-regression analysis, older age at study entry and a history of prior neoplasia were significant risk factors for development of CRN (LGD, or HGD, or CRC) among PSC-IBD patients; there was a non-significant trend for an association of multifocal dysplasia with higher risk of developing aCRN after an LGD diagnosis (HR 3.2, 95% CI 0.8–11.8, p=0.086).
Conclusions: PSC-IBD patients have a similar incidence of LGD compared to non-PSC IBD patients, but the risk of developing aCRN after a diagnosis of LGD is significantly higher in PSC-IBD patients. Our findings substantiate recommendations for annual surveillance in this very high-risk population.