While the traditional role of epidemiology was the study of communicable diseases, it has evolved to a modern approach geared towards the study of the increasing incidence of non-communicable chronic diseases. Epidemiology may be defined as the collection of large sample sizes and the measurement of numerous variables from stored samples to facilitate estimation of disease risk over time, and may involve the development and use of new techniques to acquire dependent and independent variables . Nevertheless, the traditional view of epidemiology remains embedded as the study of epidemics. In studying Inflammatory Bowel Disease (IBD) as an epidemic, several aspects were addressed at ECCO'19 Copenhagen Congress, including changes in the incidence/prevalence of IBD in different populations (countries), effects of therapy on surgical rates, the incidence of complications such as colorectal cancer, the impact of treatment patterns and findings in smaller sub-populations such as patients with elderly-onset and perianal Crohn’s Disease (CD). Although the modern definition carries a risk that studies will be conducted without a specific hypothesis thereby generating statistically significant observations of no clinical relevance nor benefit, the studies presented at ECCO 2019 addressed clinically important questions.
One of the many difficult questions to answer has been the impact of biologic therapy on disease course. Complete remission with mucosal healing is postulated to reduce surgical rates by halting disease progression, thereby avoiding complications. Results from large-scale studies have been conflicting. Three studies presented at ECCO 2019 addressed this topic. The first, from the United Kingdom, used a hospital dataset linked to a prescription dataset. Within the limits of the study, the authors reported an association, albeit a non-significant one, between reduction in Ulcerative Colitis (UC) surgery rates and the introduction of infliximab in 2008 . Similarly, while a decline in colectomy rates due to anti-TNF therapy was observed in a study from Denmark in 334 children and adolescents with UC, this reduction did not reach the level of significance. Instead the population showed a benefit in terms of reduced use of steroids following the start of anti-TNF therapy . In contrast, analysis of a large administrative dataset from Quebec in Canada concerning a newly diagnosed population suggested an increased probability of major surgery during the years after the introduction of biologic drugs despite increasing use of these drugs .
Inherent limitations of retrospective studies are the indirect analysis of treatment and surgery rates as well as the assumption that prescription of biologic drugs is associated with remission. Studies investigating comparative surgical rates between patients receiving drug treatment versus no drug treatment and assessing disease activity to determine the association between remission and surgery may explain some of these inconsistent findings. A post-hoc analysis of GEMINI trials addressed the issue of effects on surgical rates and showed a lower rate of surgery with vedolizumab (VDZ) therapy versus placebo at one year. In patients who continued VDZ therapy for five years, low rates of surgical intervention were observed . Another potential confounder is disease duration. Disease duration within the GEMINI studies was >7 years, raising the question of whether administration of biologic therapy early in the disease course might influence outcomes to a greater extent. In a paediatric population, surgical resection rates declined with increasing use of anti-TNF drugs in CD but not UC. It is noteworthy that in this study, anti-TNF therapy within three years of diagnosis had a greater impact on surgical rates . Further studies with a higher surgical event rate and collecting additional variables such as short disease duration before biologic therapy and correlation between response to biologic therapy and measures of disease activity would be invaluable in clarifying the impact on surgery rates.
The emergence of IBD as an epidemic in low-incidence countries has opened up the potential for studies of the risk factors underlying this trend. Abstracts presented from Brazil, Iran, Thailand, Singapore and French Polynesia all highlighted the global impact of IBD and the introduction of databases or registries to monitor disease patterns. The incidence in French Polynesia remains the lowest among Eastern countries [7–11].
Similarly, studies are required to explore the causes and risk factors underlying temporal trends in populations in high-incidence countries. In this context, observations stemming from well-developed databases were presented, addressing the alarming rise in incidence and prevalence in Scotland and the continued rise in Israel, the United States and Spain [12–15].
Reports on treatment outcomes across IBD population cohorts in Denmark and Sweden and across Europe offered a better understanding of treatment patterns and outcomes in response to different therapeutic interventions. A study in Denmark compared outcomes of combination therapy with thiopurine and allopurinol versus thiopurine monotherapy in 10,367 IBD patients . No significant differences were noted between the groups with regard to either benefits (surgery, hospitalisation) or risks (adverse events), even though previous studies have indicated improved clinical remission with the studied combination. Drug survival assessed through the Swedish database was greater for adalimumab than for infliximab, both when it was used as a first-line anti-TNF drug and when it was used as a second-line treatment. This result encourages use of adalimumab as a first-line biologic in CD . In the Epi-IBD study, elderly patients were found to have received less aggressive therapy; further studies are needed to explore the reason for this, and particularly whether the finding reflects a less severe disease course in late-onset IBD .
The nature of the relation between cancer and IBD is an important topic, and two abstracts presented findings on the increased risk of cancer in patients with paediatric-onset IBD [19, 20]. In another study, the risk of colorectal cancer following low-grade dysplasia (LGD) was explored in a Dutch cohort of 1215 IBD patients with colonic LGD. The significance of LGD recurrence at a subsequent surveillance colonoscopy was investigated in comparison with absence of such recurrence. It was found that 21.3% of patients developed a recurrence of LGD within three years of the index lesion. Within this group, 17.8% developed subsequent advanced neoplasia compared with 10% without a second LGD lesion, suggesting that LGD recurrence is a high-risk precursor of future cancer .
Epidemiology research continues to address questions of clinical importance. Progress is anticipated through the acquisition of large prospective datasets that collect specific standardised variables as well as the harmonisation of data across databases to enable collaborative studies across the world. These measures may help to close the gaps in knowledge, and one may look forward to presentation of further illuminating studies at future ECCO Congresses.