Fistula Presence As A Cancer Risk Factor In Crohn’s Disease
Overview
In this retrospective cohort study, the researchers aimed to delve into the cancer risk among patients with Crohn’s disease (CD) based on the presence or absence of perianal fistulas (PF). The study spanned from January 1, 2013, to December 31, 2014, utilizing the comprehensive German InGef research database. The cohort consisted of 10,208 identified CD patients, with 824 individuals (8.1%) having perianal fistulas (CPF).
The investigation focused on two key aspects: the prevalence and incidence of cancer in both the CPF and non-PF CD groups. The 6-year crude malignancy prevalence in patients with CPF was found to be 8.13% (95% confidence interval [CI] 6.36%–10.21%), which was notably lower than the prevalence in patients with non-PF CD, standing at 19.8% (95% CI 19%–20.6%). The incidence of cancer, measured per 100,000 person-years, was 1184 (95% CI 879–1561) for CPF and 2365 (95% CI 2219–2519) for non-PF CD.
Interestingly, when comparing the two groups, the adjusted incidence rate ratio (IRR) of cancer showed no statistically significant difference. The calculated IRR for CPF compared with non-PF CD was 0.83 (95% CI 0.62–1.10), with a p-value of 0.219. This lack of significance indicates that, despite a numerical difference in cancer prevalence, patients with CPF did not exhibit a significantly higher incidence of cancer compared to those with non-PF CD.
It’s noteworthy that, despite the absence of statistical significance, individuals with CPF demonstrated a higher numerical risk of cancer than the general German population. This observation suggests that while the study did not establish a conclusive disparity in cancer risk between the two CD groups, the potential for an increased risk in patients with perianal fistulas warrants continued attention and further investigation.
Introduction
This study addresses the heightened risk of cancer, particularly gastrointestinal cancer, in patients with Crohn’s disease (CD). The focus is on individuals with CD who experience perianal fistulas (PF), a complication associated with increased morbidity and healthcare costs compared to CD patients without PF (non-PF CD). Existing evidence links CD treatments, such as immunomodulators and biologics, to cancer risk, with reports suggesting an elevated risk of skin cancer and lymphoma in CD patients with perianal fistulas (CPF).
The study aims to fill a critical gap in population-based research by elucidating the prevalence and incidence of various cancers, including site-specific ones like perianal, lymphoma, and skin cancers, in patients with CPF. The investigation becomes especially relevant in the context of an ongoing Post Authorization Safety Study (PASS) examining the risk of malignancies among patients treated with darvadstrocel, a treatment involving stem cells.
To provide a comprehensive background, the researchers designed a real-world evidence (RWE) study focusing on the German population. The study seeks to estimate the prevalence and incidence of cancer in both CPF and non-PF CD groups. Additionally, it aims to calculate the incidence rate ratio (IRR) of cancer between these two groups and the standardized incidence ratio (SIR) of cancer diagnoses in patients with CPF compared to the general German population.
The research recognizes the severity of PF complications, their resistance to spontaneous improvement, and the necessity for medical or surgical intervention. Despite indications from non-clinical studies suggesting low tumorigenicity risk with certain stem cell treatments, concerns persist regarding their potential impact on cancer risk in humans. The study aims to contextualize the ongoing PASS results by establishing the baseline rates of cancer within the specific population of interest—individuals with CPF. This initiative contributes valuable real-world evidence to the understanding of cancer risks in CD patients with perianal fistulas, offering insights essential for both clinical practice and ongoing research endeavors.
Method
This retrospective cohort study utilized the InGef (Institute for Applied Health Research Berlin) database, a comprehensive anonymized German health claims database encompassing longitudinal data from around 4.8 million insured members across various federal states. The database, reflective of the German population in terms of age and gender, includes demographic information, inpatient and outpatient care details, drug dispensing records, information on remedies and devices, as well as sick leave and sickness allowance data. It also provides cost information from the statutory health insurance (SHI) perspective, with diagnoses coded according to the International Classification of Diseases 10th revision German modification (ICD‐10 GM), procedures classified using German Procedure Classification codes, and prescriptions recorded through the Anatomical Therapeutic Chemical classification.
The study focused on the period from January 1, 2013, to December 31, 2014, encompassing all patients with a recorded diagnosis of Crohn’s disease (CD) in the InGef database. The index date for study entry was set as January 1, 2015, with patients tracked from this point until December 31, 2020. The end of follow-up was determined by the earliest occurrence of outcomes such as a cancer record (excluding prevalence analysis), the termination of data collection (indicative of emigration, death, or a switch to another healthcare provider), or the conclusion of the study period. This robust methodology aimed to provide a comprehensive understanding of cancer prevalence and incidence in patients with CD, particularly those with perianal fistulas, offering valuable insights for clinical considerations and ongoing research initiatives.
The analysis included two patient groups: those with Crohn’s disease (CD) and perianal fistulas (CPF) and those with CD without perianal fistulas (non-PF CD). Patients in the non-PF CD group were considered as such unless they had a record of perianal fistulas during the selection period, at which point their follow-up was considered as contributing to the CPF group.
The study variables encompassed perianal fistula exposure and malignancy exposure. CPF was defined as having at least one medical code for CD and one medical code for perianal fistula or undergoing surgery for perianal fistula within a hospital or outpatient setting. The perianal fistula exposure was considered a time-varying variable, transitioning from unexposed to exposed if an individual developed perianal fistula during follow-up.
Malignancies were classified using ICD-10 GM codes in hospital or outpatient settings, with analyses conducted for any cancer, digestive tract cancer (including colorectal and anal/perianal cancer), and extra-intestinal cancers (such as lung cancer, lymphomas, and skin cancer).
Covariates considered as potential confounders included region, sex, and age, with age estimated at the index date. Additional baseline confounders, identified in the first two quarters after January 1, 2015, comprised major colorectal surgery, smoking (identified through proxy algorithms involving COPD and nicotine abuse/dependence ICD-10 GM codes), alcohol abuse, obesity, comorbidities (primary sclerosing cholangitis, diabetes, autoimmune diseases), and CD treatments associated with an increased risk of lymphoma or skin cancer (thiopurines, anti-TNF, other biologics). The study aimed to account for these variables to enhance the accuracy and validity of the findings.
Inclusion Criteria
Eligibility criteria for inclusion in this study encompassed individuals aged 18 years and above, ensuring a minimum of 2 years of continuous insurance coverage within the InGef database during the specified selection period. Furthermore, potential participants were required to have at least one diagnosis of Crohn’s disease (CD), as indicated by the relevant International Classification of Diseases 10th revision German modification (ICD‐10 GM) code (K50.x). This diagnosis could be either the primary or secondary hospital discharge diagnosis, or an outpatient diagnosis code recorded between January 1, 2013, and December 31, 2014. These stringent inclusion criteria aimed to establish a well-defined cohort of individuals with a confirmed history of CD for robust analysis within the study.
Exclusion Criteria
The analysis excluded patients with a prior diagnosis or related procedure code for perianal fistulas (PF) within a specified timeframe preceding the diagnosis of Crohn’s disease (CD). Specifically, individuals were excluded if PF diagnosis occurred 30 days prior to the initial CD diagnosis in hospital records or during the same quarter in outpatient records. Additionally, patients with a concurrent diagnosis of ulcerative colitis during the study period were excluded from the analysis.
To ensure the accuracy of incidence analyses, individuals diagnosed with cancer during the selection period were excluded. This precaution aimed to prevent the misclassification of pre-existing cancers (those predating 2015) as incident cancers during the study period. Furthermore, individuals who underwent major colorectal surgery were also excluded from the analysis, contributing to the refinement of the study cohort and the precision of subsequent findings.
Statistical Analysis
The study conducted prevalence and incidence calculations for cancer in patients with Crohn’s disease (CD) with perianal fistulas (CPF) and those without perianal fistulas (non-PF CD). Prevalence and incidence rates were analyzed by calendar year, covering the complete study period, and included various age subgroups within the CPF and non-PF CD patient populations. Due to data privacy regulations, categories with fewer than five patients were not reported.
The prevalence and incidence rates were calculated per 100,000 person-years, employing the direct standardization method based on the German standard population at the end of 2017. The prevalent study population included patients with a cancer record within the follow-up period, while the incident study population comprised patients with a record of cancer during the follow-up but not in the selection period.
Incidence rate ratios (IRR) of cancer between CPF and non-CPF groups were estimated using Poisson regression with Generalized Linear Models. Adjustments were made for various covariates such as gender, age, time from CD record and CPF record, region, and potential confounders (nicotine abuse, alcohol abuse, obesity, primary sclerosing cholangitis, diabetes, autoimmune disease, and CD treatment). Standardized incidence rates (SIR) for cancer in patients with CPF were compared to rates in the general German population, utilizing estimates from the Global Cancer Observatory for the year 2020.
A sensitivity analysis was conducted in a subgroup of newly diagnosed CD patients to account for varying disease duration. This subgroup included patients with a CD record in 2014 but not in 2013. The analysis encompassed crude and age-standardized prevalence and incidence of malignancies in CPF and non-PF CD, along with the IRR of cancer between CPF and non-PF CD in this specific subgroup.
Result
This retrospective study, utilizing the InGef database, aimed to assess the prevalence and incidence of cancer in Crohn’s disease (CD) patients with perianal fistulas (CPF) compared to those without perianal fistulas (non-PF CD). Among the 10,208 eligible patients, 8% were in the CPF group, and 11% had a record of cancer during the selection period, leading to exclusion from the incidence analysis.
Baseline demographics were largely comparable between CPF and non-PF CD groups in the incidence study population, except for a higher mean age in the non-PF CD group and increased use of thiopurines or anti-TNF treatments in the CPF group. Over the study period, 8.8% of patients had CPF, contributing to 45,316 person-years of follow-up.
Crude prevalence of any cancer was lower in the CPF group (8.13%) compared to the non-PF CD group (19.81%). The age-standardized cancer prevalence in patients with CPF was 8.65%, with a higher prevalence in males than females. Incidence rates for all cancers, including anal/perianal cancer, were lower in the CPF group than the non-PF CD group.
Crude incidence of any cancer was higher in the non-PF CD group (2365/100,000 person-years) compared to the CPF group (1184/100,000 person-years). The age-standardized cancer incidence in patients with CPF was 1378/100,000, with slightly higher rates in males. The incidence rate ratio (IRR) of cancer between CPF and non-PF CD groups was not statistically significant after adjusting for potential confounders.
Standardized incidence rates (SIR) for any cancer in patients with CPF were 1.58-fold greater than in the general German population, based on the 2020 International Agency for Research on Cancer database. However, SIR for digestive tract cancers in CPF patients aligned with the general population.
In a sensitivity analysis focusing on newly diagnosed CD patients in 2014, the adjusted IRR for any cancer in CPF versus non-PF CD was 0.82, suggesting a lower risk for cancer in the newly diagnosed CPF subgroup.
This comprehensive analysis provides valuable insights into cancer prevalence and incidence among CD patients with perianal fistulas, emphasizing the importance of considering disease-specific factors in assessing cancer risk.
Conclusion
This pioneering real-world evidence (RWE) study, conducted in Europe and utilizing the German InGef research database from 2015 to 2020, aimed to compare cancer prevalence and incidence in Crohn’s disease (CD) patients with and without perianal fistulas (PF). The adjusted incidence rate ratio (IRR) for any type of cancer between the two groups showed no statistically significant differences, emphasizing the robustness of the results in a sensitivity analysis focused on newly diagnosed patients.
While the study revealed a standardized incidence ratio (SIR) of any cancer in patients with PF more than 1.5 times higher than the general German population, the age-standardized prevalence and incidence rates of cancers were higher in non-PF CD patients. This intriguing observation might be attributed to the age difference between the groups, with CPF patients being, on average, seven years younger than non-PF CD patients. Age, a well-known cancer risk factor, demonstrated an increasing trend in cancer prevalence and incidence for both groups.
Contrary to expectations based on previous literature associating PF with increased inflammation and cancer risk, this study found no heightened risk of any cancer type in patients with CPF compared to those without. The authors suggested that any observed differences in cancer rates between CD patients with or without PF are likely influenced by the presence of CD itself.
The study acknowledged several strengths, including the use of the InGef research database, a large and representative real-world evidence source, and the robust identification of PF exposure through various codes. However, limitations, such as a limited follow-up period, small sample size in the PF group, and potential coding inaccuracies, were also acknowledged. Further research with extended follow-up and larger cohorts is recommended to validate these findings.
In conclusion, this study challenges previous assumptions and indicates that the presence of PFs in CD patients may not increase the risk of any type of cancer when compared to those without PFs. Despite limitations, this pioneering research provides valuable insights and sets the stage for future, more extensive observational studies.
Oncology Related Tools
Other
Latest Research
