The risk of disabling, surgery and reoperation in Crohn's disease - A decision tree-based approach to prognosis

Development and validation of diagnosis model for inflammatory bowel diseases based on a serologic biomarker panel: A decision tree model study

BACKGROUND AND STUDY AIMS

Currently, an increasing amount of experimental data is available on newly discovered biomarkers in inflammatory bowel diseases (IBD), but the role of these biomarkers is often questionable due to their limited sensitivity. Therefore, this study aimed to build a diagnostic tool incorporating a panel of serum biomarkers into a computational algorithm to identify patients with IBD and differentiate those with Crohn's disease (CD) from those with ulcerative colitis (UC).

PATIENTS AND METHODS

We studied sera from 192 CD patients, 118 UC patients, 60 non-IBD controls and 60 healthy controls. Indirect immunofluorescence (IIF) assays were utilized to determine several serum biomarkers previously associated with IBD, and the decision tree algorithm was used to construct the diagnosis model. Performances of models were evaluated by prediction accuracy, precision, AUC and Matthews's correlation coefficient (MCC). The "Inflammatory Bowel Disease Multi-omics Database (IBDMDB)" cohorts were used to validate the model as external validation set.

RESULTS

The prediction rates were determined and compared for decision tree models after each data was developed using C5.0, C&RT, QUEST and CHAID. The C5.0 and CHAID algorithms, which ranked top for the prediction rate in the IBD vs. non-IBD model and the CD vs. UC model, respectively, were utilized for final pattern analysis. The final decision tree model achieved higher classification accuracy than the approach based on conservative marker combinations (sensitivity 75.0% vs. 79.5%, specificity 93.8% vs. 78.3% for differentiating IBD from non-IBD; and sensitivity 84.3% vs. 73.4%, specificity 92.5% vs. 54.9% for differentiating CD from UC, respectively). The model prediction consistency was 93% (28/30) in the external validation set.

CONCLUSION

The decision-tree-based approach used in this study, based on serum biomarkers, has shown to be a valid and useful approach to identifying IBD and differentiating CD from UC.

The Risk of Developing Disabling Crohn's Disease: Validation of a Clinical Prediction Rule to Improve Treatment Decision Making

BACKGROUND

Crohn's disease (CD) is characterized by the development of complications over the course of the disease. It is crucial to identify predictive factors of disabling disease, in order to target patients for early intervention. We evaluated risk factors of disabling CD and developed a prognostic model.

METHODS

In total, 511 CD patients were retrospectively analyzed. Univariate and multivariate logistic regression analyses were used to identify demographic, clinical, and biological risk factors. A predictive nomogram model was developed in a subgroup of patients with noncomplicated CD (inflammatory pattern and no perianal disease).

RESULTS

The rate of disabling CD within 5 years after diagnosis was 74.6%. Disabling disease was associated with gender, location of disease, requirement of steroids for the first flare, and perianal lesions. In the subgroup of patients (310) with noncomplicated CD, the rate of disabling CD was 80%. In the multivariate analysis age at onset <40 years (OR = 3.46, 95% confidence interval [CI] = 1.52-7.90), extensive disease (L3/L4) (OR = 2.67, 95% CI = 1.18-6.06), smoking habit (OR = 2.09, 95% CI = 1.03-4.27), requirement of steroids at the first flare (OR = 2.20, 95% CI = 1.09-4.45), and albumin (OR = 0.59, 95% CI = 0.36-0.96) were associated with development of disabling disease. The developed predictive nomogram based on these factors presented good discrimination, with an area under the receiver operating characteristic curve of 0.723 (95% CI: 0.670-0.830).

CONCLUSION

We identified predictive factors of disabling CD and developed an easy-to-use prognostic model that may be used in clinical practice to help identify patients at high risk and address treatment effectively.

Secondary Indicators for an Evaluation and Guidance System for Quality of Care in Inflammatory Bowel Disease Centers: A Critical Review of the Inflammatory Bowel Disease Quality of Care Center

The number of patients with inflammatory bowel disease (IBD) has increased remarkably in recent years. However, the level of health care for IBD patients varies greatly among regions of China. Standardization of health care for IBD patients is essential to improve quality of care (QoC). The mission of the IBD Quality Care Evaluation Center (IBDQCC) is to establish indicators for QoC. Since 2017, the IBDQCC has developed structure, process, and outcome indicators with the steering committee of IBD specialists and methodologists; 28 core and 13 secondary IBD QoC indicators were selected using a Delphi method. Applications for certification of IBD quality care units were made voluntarily and preliminarily screened through the IBDQCC committee. Regional units had to meet all core indicators, and units of excellence were required to meet all core indicators together with an additional 50% of secondary indicators. As of 2019 and 2020, 69 IBD units (all from tertiary referral hospitals) have been certified as regional IBD units in China. The certification of excellence of the IBD units is currently undergoing auditing. The awareness of and appreciation for QoC in IBD is increasing in China, especially through the quality control evaluation program initiated by the IBDQCC, with a higher number of IBD units applying for the next round of certification. Although secondary indicators seem to play relatively minor roles in QoC, they suggest additional requirements for high-level centers.

Composite outcomes in observational studies of Crohn's disease: a systematic review and meta-analysis

BACKGROUND

This systematic review and meta-analysis aims to assess composite and aggregate outcomes of observational studies in Crohn's disease and to evaluate whether the number and type of variables included affect the frequency of the outcome.

METHODS

MEDLINE [via PubMed], Scopus and Web of Science were searched to identify observational studies that enrolled patients with Crohn's disease and evaluated a composite or aggregate outcome. The proportion of patients achieving the outcome was determined and a random-effects meta-analysis was performed to evaluate how the frequency of each outcome varies according to the reporting of predefined variables.

RESULTS

From 10,257 identified records, 46 were included in the qualitative analysis and 38 in the meta-analysis. The frequency for composite and aggregate outcomes was 0.445 [95% confidence interval (CI): 0.389-0.501] and 0.140 (95% CI: 0.000-0.211), respectively. When comparing composite outcomes by number of included variables, the frequency was 0.271 (95% CI: 0.000-0.405) and 0.698 (95% CI: 0.651-0.746), for one and six variables, respectively. The frequency of the composite outcome varied according to the identity of the variables being reported. Specific pairs of predefined variables had a significant effect in the frequency of composite outcomes.

CONCLUSION

Composite outcomes with increasing number of predefined variables show an increase in frequency. Outcomes including variables such as 'Surgery' and 'Steroids' had higher frequencies when compared with the ones that did not include these variables. These results show that the frequency of composite outcomes is dependent on the number and type of variables being reported.

Prognostic biomarkers to identify patients likely to develop severe Crohn's disease: a systematic review

BACKGROUND

Identification of biomarkers that predict severe Crohn's disease is an urgent unmet research need, but existing research is piecemeal and haphazard.

OBJECTIVE

To identify biomarkers that are potentially able to predict the development of subsequent severe Crohn's disease.

DESIGN

This was a prognostic systematic review with meta-analysis reserved for those potential predictors with sufficient existing research (defined as five or more primary studies).

DATA SOURCES

PubMed and EMBASE searched from inception to 1 January 2016, updated to 1 January 2018.

REVIEW METHODS

Eligible studies were studies that compared biomarkers in patients who did or did not subsequently develop severe Crohn's disease. We excluded biomarkers that had insufficient research evidence. A clinician and two statisticians independently extracted data relating to predictors, severe disease definitions, event numbers and outcomes, including odds/hazard ratios. We assessed risk of bias. We searched for associations with subsequent severe disease rather than precise estimates of strength. A random-effects meta-analysis was performed separately for odds ratios.

RESULTS

In total, 29,950 abstracts yielded just 71 individual studies, reporting 56 non-overlapping cohorts. Five clinical biomarkers (Montreal behaviour, age, disease duration, disease location and smoking), two serological biomarkers (anti-Saccharomyces cerevisiae antibodies and anti-flagellin antibodies) and one genetic biomarker (nucleotide-binding oligomerisation domain-containing protein 2) displayed statistically significant prognostic potential. Overall, the strongest association with subsequent severe disease was identified for Montreal B2 and B3 categories (odds ratio 4.09 and 6.25, respectively).

LIMITATIONS

Definitions of severe disease varied widely, and some studies confounded diagnosis and prognosis. Risk of bias was rated as 'high' in 92% of studies overall. Some biomarkers that are used regularly in daily practice, for example C-reactive protein, were studied too infrequently for meta-analysis.

CONCLUSIONS

Research for individual biomarkers to predict severe Crohn's disease is scant, heterogeneous and at a high risk of bias. Despite a large amount of potential research, we encountered relatively few biomarkers with data sufficient for meta-analysis, identifying only eight biomarkers with potential predictive capability.

FUTURE WORK

We will use existing data sets to develop and then validate a predictive model based on the potential predictors identified by this systematic review. Contingent on the outcome of that research, a prospective external validation may prove clinically desirable.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42016029363.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 45. See the NIHR Journals Library website for further project information.

Ileal Crohn's Disease Exhibits Similar Transmural Fibrosis Irrespective of Phenotype

INTRODUCTION

In Crohn's disease (CD), the assessment of transmural inflammation and fibrosis is of utmost importance. This study aimed to quantify these parameters in CD ileal specimens and correlate them with disease progression.

METHODS

This is a retrospective unicentric study based on the analysis of archived specimens (n = 103) of primary ileal resection. Data were retrieved from a prospective national inflammatory bowel disease registry. Two pathologists, blinded for CD phenotype and clinical indications for surgery, examined 3 sections per patient and graded inflammation and fibrosis, based on a histopathological score.

RESULTS

Penetrating (B3, n = 74) CD exhibited significantly higher inflammation in diseased areas, compared with stricturing (B2, n = 29) disease (score 3: 96% vs 76%, P = 0.005 in inflamed areas; 78% vs 55%, P = 0.019 in most affected areas). This was also observed for the comparison of B2 CD with B3 CD with (B3s, n = 54) and without associated stricture (B3o, n = 20): B3s vs B2: 81% vs 55%, P = 0.033 in most affected areas; B3o vs B2: 100% vs 76%, P = 0.006 in inflamed areas; 70% vs 55%, P = 0.039 in most affected areas. We could not show differences in fibrosis scores between the subphenotypes. Postoperative new penetrating events occurred only in B3s (n = 6, 11%, P = 0.043) patients. The changing of biologic therapy after surgery correlated with severe inflammation at the proximal ileal margin (55% changed vs 25% not changed, P = 0.035).

DISCUSSION

In our cohort, fibrosis scores and fibromuscular changes were comparable, irrespective of CD phenotype. Inflammation severity was the major differentiator between penetrating and stricturing disease.JOURNAL/cltg/04.03/01720094-202104000-00012/inline-graphic1/v/2021-04-13T161901Z/r/image-tiff.

Age at Diagnosis Is Determinant for the Outcome of Inflammatory Bowel Disease: Is It a Myth?

INTRODUCTION

Patients with elderly-onset inflammatory bowel disease were previously associated with a less aggressive course of the disease. However, there are conflicting data that need further validation. We aimed to determine the association between age at diagnosis and the development of progressive disease in patients with Crohn's disease (CD) and ulcerative colitis (UC).

METHODS

This cohort study included patients with CD and UC followed in 6 secondary and tertiary care centers in mainland Portugal. Patients were divided into a derivation (80%) cohort and a validation (20%) cohort. The primary outcome was progressive disease. Logistic regression analysis, receiver operating characteristic curves, and the areas under the curve (AUC) were performed. Odds ratios with 95% confidence intervals (CIs) were estimated.

RESULTS

The derivation cohorts included 1245 patients with CD (68% with progressive disease) and 1210 patients with UC (37% with progressive disease), whereas the validation cohorts included 302 patients with CD and 271 patients with UC, respectively, with similar outcome proportions. In our final model, age at diagnosis older than 60 years was significantly associated with a lower risk of developing progressive disease (odds ratio 0.390, 95% CI 0.164-0.923, P = 0.032), with a high discriminative power (AUC 0.724, 95% CI 0.693-754) in patients with CD. However, according to this model, no significant associations were found between age at diagnosis and the risk of developing progressive disease in patients with UC. No differences were observed in the AUC values between the validation and the derivation cohorts.

DISCUSSION

Patients with elderly-onset CD, but not patients with UC, were associated with a less progressive course of the disease.

Clinical Criteria Can Identify Children With Osteopenia in Newly Diagnosed Crohn Disease

OBJECTIVES

Chronic inflammation of Crohn disease (CD) is associated with reduced bone mineral density (BMD). As bone mass is almost exclusively accrued during childhood, early recognition of osteopenia is especially important in pediatric CD. We aimed to identify variables associated with osteopenia to guide dual-energy X-ray absorptiometry (DXA) scan screening to those who most need it.

METHODS

This was a retrospective inception cohort study of children newly diagnosed with CD, and routinely referred to DXA scans. Demographic and explicit clinical data were recorded along with whole-body less head BMD, adjusted for age, sex, and height by z-scores.

RESULTS

Of the 116 included children (mean age 13 ± 3.1 years, 67 [58%] boys, mean body mass index [BMI] 16.7 ± 2.6), 63 (54%) had normal BMD (z-score > -1) or borderline osteopenia (-1 ≥ z-score > -2) and 53 (46%) had osteopenia (z-score ≤ -2). Osteopenia was associated with lower BMI z-score (-0.8 ± 1.2 vs -1.8 ± 1.1, P < 0.001) and higher PCDAI (33.7 ± 15.2 vs 25.7 ± 16.5; P = 0.009) than those with BMD z-score >-2. In total, 59% of children with BMI z-score <-0.5 had moderate-severe osteopenia and only 18% of those with higher z-scores. Multivariate logistic regression identified BMI z-score as the sole risk factor (OR 1.28 [95% CI 1.08-1.52], P = 0.005). BMI z-score ≥-0.5 excludes osteopenia with a sensitivity 87%, specificity 49%, NPV 82%, and PPV 59%.

CONCLUSIONS

Osteopenia was found in nearly half of children with newly onset CD. BMI z-score <-0.5 should prompt referral to DXA screening.

Inflammatory Bowel Disease Reoperation Rate Has Decreased Over Time If Corrected by Prevalence

Despite the recent emergence of expensive biologic therapies, hospitalization and surgery remain important contributors for the overall costs of inflammatory bowel disease (IBD). In this study, we aimed to describe the burden of reoperations in patients with IBD by evaluating reoperation rates, charges, and risk factors over 16 years.

What is Machine Learning? A Primer for the Epidemiologist

Machine learning is a branch of computer science that has the potential to transform epidemiologic sciences. Amid a growing focus on "Big Data," it offers epidemiologists new tools to tackle problems for which classical methods are not well-suited. In order to critically evaluate the value of integrating machine learning algorithms and existing methods, however, it is essential to address language and technical barriers between the two fields that can make it difficult for epidemiologists to read and assess machine learning studies. Here, we provide an overview of the concepts and terminology used in machine learning literature, which encompasses a diverse set of tools with goals ranging from prediction to classification to clustering. We provide a brief introduction to 5 common machine learning algorithms and 4 ensemble-based approaches. We then summarize epidemiologic applications of machine learning techniques in the published literature. We recommend approaches to incorporate machine learning in epidemiologic research and discuss opportunities and challenges for integrating machine learning and existing epidemiologic research methods.

Development and Validation of Risk Matrices Concerning Ulcerative Colitis Outcomes-Bayesian Network Analysis

BACKGROUND

Ulcerative colitis [UC] is a chronic inflammatory disease often accompanied by severe and distressing symptoms that, in some patients, might require a surgical intervention [colectomy]. This study aimed at determining the risk of experiencing progressive disease or requiring colectomy.

MATERIAL AND METHODS

This was a multicentre study: patients' data [n = 1481] were retrieved from the Portuguese database of inflammatory bowel disease patients. Bayesian networks and logistic regression were used to build risk matrices concerning the outcomes of interest.

RESULTS

The derivation cohort included a total of 1210 patients, of whom 6% required a colectomy and 37% had progressive disease [over a median follow-up period of 12 syears]. The risk matrices show that previously hospitalised patients with extensive disease, who are not on immunomodulators and who are refractory to corticosteroid treatment, are the ones at the highest risk of undergoing a colectomy [88%]; whereas male patients, with extensive disease and less than 40 years old at diagnosis, are the ones at the highest risk of experiencing progressive disease [72%]. These results were internally and externally validated, and the AUC [area under the curve] of the ROC [receiver operating characteristic] analysis for the derivation cohort yielded a high discriminative power [92% for colectomy and 72% for progressive disease].

CONCLUSIONS

This study allowed the construction of risk matrices that can be used to accurately predict a UC patient's likelihood of requiring a colectomy or of facing progressive disease, and can be used to individualise therapeutic strategies.

The timing of early therapeutic strategies has a significant impact on Crohn's disease prognosis

BACKGROUND

Abdominal surgery and immunosuppressive pharmacological treatments are two different therapeutic options used to manage Crohn disease. This study aimed to determine whether the timing of these interventions had an impact on patients' prognosis.

METHOD

This manuscript entails a retrospective analysis of a multicentric cohort involving 498 CD patients that had bowel surgery after diagnosis and prior to immunosuppression treatments. Two endpoints were considered: the occurrence of disabling disease and the need to undergo further bowel surgeries.

RESULTS

Disabling disease affected 71% of all patients, whereas 39% needed reoperation. The odds ratios (OR) of being affected by disabling disease were higher when patients had upper tract involvement [3.412 [1.285-9.061]], perianal disease (2.270 [1.239-4.157]) and a longer time elapsed from diagnosis to first surgery (13-36 months: 2.576 [1.207-5.500]). On the other hand, the need to undergo further surgical interventions was significantly increased in smoking patients (2.294 [1.187-4.432]), but decreased in patients who started pharmacological therapy not later than six months after the first surgery (0.256 [0.093-0.704]).

CONCLUSIONS

Our results suggest that the timing of therapeutic strategies does affects the CD outcomes: whereas an early surgery had a preventive effect on the occurrence of disabling events, the introduction of medication in the first semester after surgery had a preventive effect on the need for reoperation.