Value of colonoscopy for prediction of prognosis in patients with ulcerative colitis

Predictive models for endoscopic disease activity in patients with ulcerative colitis: Practical machine learning-based modeling and interpretation

Background

Endoscopic disease activity monitoring is important for the long-term management of patients with ulcerative colitis (UC), there is currently no widely accepted non-invasive method that can effectively predict endoscopic disease activity. We aimed to develop and validate machine learning (ML) models for predicting it, which are desired to reduce the frequency of endoscopic examinations and related costs.

Methods

The patients with a diagnosis of UC in two hospitals from January 2016 to January 2021 were enrolled in this study. Thirty nine clinical and laboratory variables were collected. All patients were divided into four groups based on MES or UCEIS scores. Logistic regression (LR) and four ML algorithms were applied to construct the prediction models. The performance of models was evaluated in terms of accuracy, sensitivity, precision, F1 score, and area under the receiver-operating characteristic curve (AUC). Then Shapley additive explanations (SHAP) was applied to determine the importance of the selected variables and interpret the ML models.

Results

A total of 420 patients were entered into the study. Twenty four variables showed statistical differences among the groups. After synthetic minority oversampling technique (SMOTE) oversampling and RFE variables selection, the random forests (RF) model with 23 variables in MES and the extreme gradient boosting (XGBoost) model with 21 variables in USEIS, had the greatest discriminatory ability (AUC = 0.8192 in MES and 0.8006 in UCEIS in the test set). The results obtained from SHAP showed that albumin, rectal bleeding, and CRP/ALB contributed the most to the overall model. In addition, the above three variables had a more balanced contribution to each classification under the MES than the UCEIS according to the SHAP values.

Conclusion

This proof-of-concept study demonstrated that the ML model could serve as an effective non-invasive approach to predicting endoscopic disease activity for patients with UC. RF and XGBoost, which were first introduced into data-based endoscopic disease activity prediction, are suitable for the present prediction modeling.

External validation of the Paddington International Virtual Electronic ChromoendoScopy ScOre as a good endoscopic score to define mucosal healing and predict long-term clinical outcomes in ulcerative colitis

OBJECTIVES

To define endoscopic and histological remission in ulcerative colitis (UC) accurately, several scoring systems have been established. A novel virtual electronic chromoendoscopy score, the Paddington International Virtual ChromoendoScopy ScOre (PICaSSO), was developed, validated, and reproduced to assess inflammation grade and predict patient prognosis. We externally verified and validated the clinical value of PICaSSO in UC patients.

METHODS

This prospective study enrolled 63 UC patients. The Mayo endoscopic score (MES), UC Endoscopic Index of Severity (UCEIS), and PICaSSO were adopted for endoscopic evaluation. All biopsy samples were scored using the Robarts histological index (RHI), Nancy histological index (NHI), and "Extent, Chronicity, Activity, Plus additional findings" (ECAP) score. Patients with an endoscopic MES of 0-1 at baseline were followed up during a median time of 23.5 months.

RESULTS

PICaSSO was strongly correlated with other endoscopic and histological scoring systems. PICaSSO ≤3 had advantages in assessing histological remission (HR), with the highest accuracy of 88.9% for ECAP-HR. Relapse-free survival rates were significantly different between patients with MES 0 and MES 1 and patients with PICaSSO ≤3 and >3 (P = 0.010 and 0.018, respectively).

CONCLUSIONS

PICaSSO was externally validated with strong correlations with other endoscopic and histological scoring systems in UC, and PICaSSO-ER might potentially predict a better long-term clinical outcome in UC patients.

Advances in Imaging Specific Mediators of Inflammatory Bowel Disease

Inflammatory bowel disease (IBD) is characterized by chronic remitting and relapsing inflammation of the lower gastrointestinal tract. The etiology underlying IBD remains unknown, but it is thought to involve a hypersensitive immune response to environmental antigens, including the microbiota. Diagnosis and monitoring of IBD is heavily reliant on endoscopy, which is invasive and does not provide information regarding specific mediators. This review describes recent developments in imaging of IBD with a focus on positron emission tomography (PET) and single-photon emission computed tomography (SPECT) of inflammatory mediators, and how these developments may be applied to the microbiota.

How to predict clinical relapse in inflammatory bowel disease patients

Inflammatory bowel diseases have a natural course characterized by alternating periods of remission and relapse. Disease flares occur in a random way and are currently unpredictable for the most part. Predictors of benign or unfavourable clinical course are required to facilitate treatment decisions and to avoid overtreatment. The present article provides a literature review of the current evidence on the main clinical, genetic, endoscopic, histologic, serologic and fecal markers to predict aggressiveness of inflammatory bowel disease and discuss their prognostic role, both in Crohn’s disease and ulcerative colitis. No single marker seems to be reliable alone as a flare predictor, even in light of promising evidence regarding the role of fecal markers, in particular fecal calprotectin, which has reported good results recently. In order to improve our daily clinical practice, validated prognostic scores should be elaborated, integrating clinical and biological markers of prognosis. Finally, we propose an algorithm considering clinical history and biological markers to intercept patients with high risk of clinical relapse.

Factors associated with poor outcomes in adults with newly diagnosed ulcerative colitis

It is a challenge to accurately identify patients with early stage ulcerative colitis (UC) who are at highest risk for a poor outcome and therefore might require salvage therapy. Several epidemiologic and clinical studies have analyzed factors associated with poor prognosis and increased risk for colectomy. We review prognostic factors for adults with newly diagnosed UC and discuss which patients might benefit from rapid and progressive therapy. Patients with poor prognoses tend to be young nonsmokers with high levels of inflammatory biomarkers, low levels of hemoglobin, and extensive disease, based on colonoscopy. We examine these risk factors in 2 hypothetical patients who have been newly diagnosed with UC.

Ulcerative colitis six years after colon cancer: only a coincidence?

The association between inflammatory bowel disease and colorectal cancer is well known. Ulcerative colitis is a risk factor for the development of colorectal cancer, and this risk increases with the activity and duration of bowel inflammation. Here we describe the case of a 52-year-old man who developed ulcerative colitis 6 years after the diagnosis and treatment of colon cancer. Although this could be a coincidence, there could be additional possibilities, like pre-existence of quiescent colitis, late effect of therapy, or maybe the existence of common pathogenetic factors contributing to the development of ulcerative colitis and colorectal cancer.

Transparent cap colonoscopy versus standard colonoscopy to improve caecal intubation

BACKGROUND

Colonoscopy is considered the gold-standard investigation for screening and diagnosis of colorectal cancer. It is also becoming increasingly desirable for assessment, management, diagnosis and follow-up of other colorectal diseases, such as inflammatory bowel diseases and acute diverticulitis. Hence, due to the increasing demand for colonoscopy, devices to advance examination techniques are highly sought-after and the colonoscope with the transparent cap could be one of these.

OBJECTIVES

To identify and review all relevant data in order to determine whether colonoscopy with a transparent cap is a more effective diagnostic tool than colonoscopy.

SEARCH METHODS

We searched the MEDLINE, EMBASE and CINAHL databases, and the Cochrane Central Register of Controlled Trials for all randomised controlled trials (RCTs) comparing the use of colonoscopy with a transparent cap with standard colonoscopy.

SELECTION CRITERIA

Studies were included if they were randomised controlled trials which compared the use of colonoscopy with a transparent cap with standard colonoscopy.

DATA COLLECTION AND ANALYSIS

Data on study methods, participants, interventions used and outcomes measured was extracted from each study. Data was entered into the Cochrane Review Manager software (RevMan 5.0, 2008) and analysed using Cochrane MetaView.

MAIN RESULTS

In the present meta-analysis, we considered 14 randomised controlled trials so far published. The findings of our work indicate that colonoscopy with transparent cap has a faster caecal intubation time when compared with standard colonoscopy. Reviewing studies individually would also seem to favour colonoscopy with transparent cap for polyp detection rate and pain during procedure but due to lack of comparable data meta-analysis was not feasible.

AUTHORS' CONCLUSIONS

This review suggests that a transparent cap on the end of the colonoscope may give a marginally faster caecal intubation time compared with standard colonoscopy. It also suggests that there is a better polyp detection rate and less pain with the cap. However, the authors feel that further randomised controlled trials in this area would provide more clinically significant information on this adjunct to colonoscopy.

Transparent Cap Colonoscopy versus Standard Colonoscopy for Investigation of Gastrointestinal Tract Conditions

BACKGROUND

Colonoscopy is considered the gold-standard investigation for screening and diagnosis of colorectal cancer. It is also becoming increasingly desirable for assessment, management, diagnosis and follow-up of other colorectal diseases, such as inflammatory bowel diseases and acute diverticulitis. Hence, due to the increasing demand for colonoscopy, devices to advance examination techniques are highly sought-after and the colonoscope with the transparent cap could be one of these.

OBJECTIVES

To identify and review all relevant data in order to determine whether colonoscopy with a transparent cap is a more effective diagnostic tool than colonoscopy.

SEARCH STRATEGY

We searched the MEDLINE, EMBASE and CINAHL databases, and the Cochrane Central Register of Controlled Trials for all randomised controlled trials (RCTs) comparing the use of colonoscopy with a transparent cap with standard colonoscopy.

SELECTION CRITERIA

Studies were included if they were randomised controlled trials which compared the use of colonoscopy with a transparent cap with standard colonoscopy.

DATA COLLECTION AND ANALYSIS

Data on study methods, participants, interventions used and outcomes measured was extracted from each study. Data was entered into the Cochrane Review Manager software (RevMan 5.0, 2008) and analysed using Cochrane MetaView.

MAIN RESULTS

In the present meta-analysis, we considered for the first time all five randomised controlled trials so far performed. The findings of our work indicate that colonoscopy with transparent cap has a faster caecal intubation time when compared with standard colonoscopy. Reviewing studies individually would also seem to favour colonoscopy with transparent cap for polyp detection rate and pain during procedure but due to lack of comparable data meta-analysis was not feasible.

AUTHORS' CONCLUSIONS

This review suggests that a transparent cap on the end of the colonoscope may give a marginally faster caecal intubation time compared with standard colonoscopy. It also suggests that there is a better polyp detection rate and less pain with the cap. However, the authors feel that further randomised controlled trials in this area would provide more clinically significant information on this adjunct to colonoscopy.

[The usefulness of Doppler sonography in the assessment of disease activity of ulcerative colitis]

BACKGROUND/AIMS

disease activity in ulcerative colitis (UC) is generally assessed using symptoms, laboratory data, endoscopic findings, and histology of the biopsy specimens. In this study, we compared disease activity of UC as determined by clinical features and endoscopic findings, and aimed to assess the clinical usefulness of Doppler sonography.

METHODS

the duplex Doppler sonography of superior mesenteric artery (SMA) and inferior mesenteric artery (IMA) of 10 patients with clinically inactive UC and 20 patients with active UC were evaluated by one radiologist who was blinded to clinical information. Peak systolic velocity (PSV), end diastolic velocity (EDV), resistance index (RI), and pulsatility index (PI) of the SMA and IMA were evaluated. All patients underwent biochemical and endoscopic evaluations thereafter. Correlation between disease activity by the Truelove-Witts classification and the Mayo scoring system was measured, and we compared hemodynamic parameters between active and inactive UC.

RESULTS

correlation rate of disease activity between these two scoring systems was 93.3%. Flow velocities (PSV, p<0.001 and EDV, p=0.03) and PI (p=0.03) were significantly higher in patients with active UC than inactive UC. PSVs of the SMA and IMA were also significantly correlated with disease severity. The active UC could be accurately diagnosed using Doppler sonography (AUC=0.83; 95% confidence interval 0.68-0.99).

CONCLUSIONS

in patients with UC, clinical stage was well matched with endoscopic disease activity. Doppler sonography was a readily available method, and PSV of SMA would be clinically useful in predicting of disease activity and severity.

Novel endoscopic activity index is useful for choosing treatment in severe active ulcerative colitis patients

AIM

Clinical symptoms are the most important factors used by physicians to evaluate the severity and extent of ulcerative colitis (UC). In this context, colonoscopy is also a useful diagnostic tool. We have recently developed an endoscopic activity index (EAI) to assess the severity of UC. Here, we assess the correlations among the EAI, other endoscopic indices, and clinical scores. The usefulness of the EAI for choosing treatment options, such as intravenous corticosteroid or cyclosporine A (CsA), in severe UC patients was also evaluated.

METHODS

Clinical symptoms and endoscopic finding were evaluated in 396 patients with UC (454 colonoscopies). The EAI was scored using the following six items: ulcer size, ulcer depth, redness, bleeding, edema, and mucus exudates. The patients were also scored using Matts' grade, Rachmilewitz's endoscopic index, and the Lichtiger index.

RESULTS

Our results showed that (1) the EAI scores were closely correlated with those of the Lichtiger index, Matts' grade, and Rachmilewitz's endoscopic index; (2) the EAI scores significantly decreased in patients who responded to treatment, while Matts' grade did not change in some responders treated with intravenous CsA and steroid; (3) patients with a higher EAI (14-16) tended to be refractory to corticosteroid therapy (responders 19%) compared to CsA (77%), while steroid treatment was effective in 58% of patients with EAI scores of 11-13.

CONCLUSIONS

The EAI is equivalent to other endoscopic indices and relatively more useful in choosing a treatment for patients with severe UC.