Computed tomography findings of bowel wall thickening: its significance and relationship to endoscopic abnormalities

Radiomics-based machine learning in the differentiation of benign and malignant bowel wall thickening radiomics in bowel wall thickening

PURPOSE

To distinguish malignant and benign bowel wall thickening (BWT) by using computed tomography (CT) texture features based on machine learning (ML) models and to compare its success with the clinical model and combined model.

METHODS

One hundred twenty-two patients with BWT identified on contrast-enhanced abdominal CT and underwent colonoscopy were included in this retrospective study. Texture features were extracted from CT images using LifeX software. Feature selection and reduction were performed using the Least Absolute Shrinkage and Selection Operator (LASSO). Six radiomic features were selected with LASSO. In the clinical model, six features (age, gender, thickness, fat stranding, symmetry, and lymph node) were included. Six radiomic and six clinical features were used in the combined model. Classification was done using two machine learning algorithms: Support Vector Machine (SVM) and Logistic Regression (LR). The data sets were divided into 80% training set and 20% test set. Then, training took place with all three datasets. The model's success was tested with the test set consisting of features not used during training.

RESULTS

In the training set, the combined model had the best performance with the area under the curve (AUC) value of 0.99 for SVM and 0.95 for LR. In the radiomic-derived model, the AUC value is 0.87 in SVM and 0.79 in LR. In the clinical model, SVM made this distinction with 0.95 AUC and LR with 0.92 AUC value. In the test set, the classifier with the highest success distinguishing malignant wall thickening is SVM in the radiomic-derived model with an AUC value of 0.90. In other models, the AUC value is in the range of 0.75-0.86, and the accuracy values are in the range of 0.72-0.84.

CONCLUSION

In conclusion, radiomic-based machine learning has shown high success in distinguishing malignant and benign BWT and may improve diagnostic accuracy compared to clinical features only. The results of our study may help ensure early diagnosis and treatment of colorectal cancers by facilitating the recognition of malignant BWT.

A case of concurrent gastrointestinal, peritoneal and urinary bladder tuberculosis

Tuberculosis is a top 10 leading cause of death worldwide. Lungs are primarily involved organs in tuberculosis. The rest of cases are extrapulmonary tuberculosis (14% reported in 2017). Extrapulmonary tuberculosis always presents with non-specific symptoms, thus at risk of delay diagnosis and management. In genitourinary tuberculosis, kidney alone and kidney with urinary bladder or ureter is affected in more than 70% of cases. The ureter and urinary bladder infections are almost always secondary to tuberculous involvement of the kidney. Bacilli haematogenic spreading is a known transmission pathway to the kidney. In this case, we diagnosed isolated urinary bladder tuberculosis caused by direct gastrointestinal tuberculosis infiltration, a rare occurrence of extrapulmonary tuberculosis. We illustrate the multiorgan involvement in tuberculosis infection including pulmonary, gastrointestinal, peritoneal and urinary bladder.

Spectrum and Relevance of Incidental Bowel Findings on Computed Tomography

A wide spectrum of incidental bowel findings can be seen on CT, including but not limited to, pneumatosis intestinalis, diverticular disease, non-obstructive bowel dilatation, transient small bowel intussusception, and submucosal fat. Radiologists should be aware that such findings are almost always benign and of little clinical significance in the absence of associated symptoms. Conversely, vigilance must be maintained when evaluating the bowel, because malignant neoplasms occasionally come to clinical attention as incidental imaging findings. When suspicious incidental bowel wall thickening is detected, the radiologist can alert the clinical team to the finding prior to the patient becoming symptomatic, potentially leading to definitive management at an early, more curable stage.

Predictors of Neoplasia in Colonic Wall Thickening Detected via Computerized Tomography

Introduction

Colonic wall thickening (CWT) is frequently observed incidentally via abdominal computerized tomography (aCT). Although the general approach to evaluating incidental CWT is a colonoscopic examination, there is a lack of definitive recommendation guidelines. Thus, we aimed to determine neoplasia rates and identify the factors predictive of neoplasia via colonoscopic examinations of patients with CWT incidentally diagnosed via aCT.

Methods

We retrospectively reviewed 5,300 colonoscopy reports. A total of 122 patients who had CWT incidentally observed via aCT were included in the study. CWT was graded as mild (3-5 mm), moderate (6-12 mm), or severe (≥12 mm). A logistic regression model was used to determine the predictive factors for neoplasia.

Results

The mean age of the patients was 60 years, and abnormal findings were noted in 52% of the colonoscopies. Neoplastic lesions were detected in 24 patients (19.6%), while colon adenocarcinoma was detected in 8 patients (6.5%). Multivariate analysis showed that moderate-severe, focal, and asymmetric CWT were independent factors for predicting neoplasia (p=0.049, p=0.033, and p=0.018, respectively).

Conclusion

Pathological findings can be noted via colonoscopic examination in cases of incidental CWT; therefore, patients with moderate-severe, focal, or asymmetric CWT require colonoscopic examination for the purpose of detecting neoplasia.

Colonic Wall Thickening Reported in Abdominal CT: Does It Always Imply Malignancy?

Aim/Background

Early diagnosis of patients with colon cancer is one of the most important parameters affecting the survival of patients. In this study, we aimed to examine the effect of the age, hemoglobin (Hb), albumin, neutrophil lymphocyte ratio (NLR), thrombocyte lymphocyte ratio (PLR), and mean platelet values (MPV) on the separation of benign and malignant diseases in patients with suspected colon wall thickness (CWT) observed in abdominal computed tomography (CT) examination.

Method

The study included 116 patients between the ages of 18 and 95 who had CT examination where the colon wall could be evaluated and who also had colonoscopy. Patients suspected for CWT in CT with difficulties in differential diagnosis were divided into two groups according to colonoscopic-histopathological evaluations. Normal or benign pathological causes were included in the first group, while malignant causes constituted the second group. Whether the two groups differed in terms of CWT, Hb, age, albumin, NLR, PLR, and MPV values was investigated with descriptive statistics.

Results

One hundred and sixteen patients (74 males, 42 females) with CT examinations and colonoscopic sampling were included in the study. After colonoscopic and histopathological evaluations, there were 64 cases in the first group and 52 in the second group. According to the results of the univariate analysis and multivariate logistic regression, CWT, Hb, NLR, and MPV were identified to be independent variables for determining colon cancer.

Conclusion

A combined evaluation of Hb, NLR, and MPV values in patients with CWT in abdominal CT examination may contribute to the separation of benign and malignant pathologies.

Significance of ileal and/or cecal wall thickening on abdominal computed tomography in a tropical country

Background and Aim

Clinical significance of ileocecal thickening on computed tomography (CT) is uncertain. We conducted this prospective study to determine the clinical relevance of ileal and/or cecal thickening on CT.

Methods

All patients with ileocecal thickening on CT were prospectively evaluated with ileocolonoscopy, biopsy, and other relevant investigations.

Results

Fifty patients (29 males, mean age 36.8 ± 13.21 years) were studied. Thirty nine (78%) patients presented with abdominal pain. On CT, 46 (92%) had a thickened wall of terminal ileum, 25 (50%) cecum, and 21 (42%) of both cecum and ileum. The mean wall thickness of ileum and cecum on CT was 7.23 + 3.2 mm and 5.5 + 3.1 mm, respectively. Final diagnosis was tuberculosis in 24 (48%) patients, Crohn's disease (CD) in 10 (20%), and adenocarcinoma in 1 patient. Colonoscopy demonstrated abnormal findings in 41 patients (82% patients with mucosal ulcerations being most common (n = 20 (40%). Of 15 (30%) patients with ileocecal bowel wall thickening, 4 (8%) patients had normal colonoscopy and histopathology (incidental ileocecal wall thickening), and in the remaining 11 patients, histopathology showed non-specific findings and these patients were asymptomatic without any specific treatment on last follow up ranging from 3 to 24 months. Involvement of cecum with ileocecal junction, ascending colon involvement, peri-ileocecal stranding, and long-segment stricture was significantly more common in patients with underlying disease as compared to nondiseased patients (P < 0.05).

Conclusion

A majority of patients with ileocecal wall thickening on CT have an underlying disease and should be further investigated by ileocolonoscopy and biopsy.

Systematic review and meta-analysis: does colonic mural thickening on CT correlate with endoscopic findings at colonoscopy?

BACKGROUND

Colonic mural thickening (MT) is often reported on standard CT examinations of the abdomen and pelvis. It often presents a dilemma for the clinician on whether any further evaluation is needed, especially in the absence of any set guidelines.

OBJECTIVE

To evaluate the significance of colonic MT and to assess its correlation with colonoscopy.

METHODS

The search strategy was initially developed in Medline and adapted for use in Embase, Medline, NHS Evidence and TRIP. Studies were included if they had reported colonic MT and subsequent colonoscopy in adults.

RESULTS

A total of 9 cohort studies examining 1252 patients were selected having undergone both CT and colonoscopy. Of the 1252 patients with MT, 950 had an abnormal colonoscopy. In the presence of MT, the pooled positive predictive value (PPV) of having any abnormal findings at colonoscopy was 0.73 (95% CI 0.60 to 0.84). The pooled PPV for colorectal cancer, in the presence of MT reporting suspicion of cancer, was 0.63 (95% CI 0.49 to 0.75), and MT suggestive of inflammation confirmed at colonoscopy was 0.97.

CONCLUSION

The probability of having an abnormal colonoscopy in the presence of MT identified on CT is high, especially for inflammation. Asymptomatic cancers may also be detected; hence, further endoscopic confirmation is reasonable when a finding of MT is demonstrated on CT examinations. Small sample sizes of the available studies and lack of data on the description of MT detected are the main limiting factors in this review.

TRIAL REGISTRATION NUMBER

CRD42016039378.

Poor Agreement Between Imaging and Histologic and Colonoscopy Findings in Pediatric Patients

BACKGROUND

Computed tomography scans (CTs), more recently magnetic resonance imaging, are often used to assess the gastrointestinal tract in patients complaining of abdominal pain. We aim to determine the strength of agreement among abdominal imaging, endoscopic, and histologic findings.

METHODS

Retrospective chart review of pediatric patients who underwent colonoscopy between January 1, 2012, and December 31, 2014, at Women and Children's Hospital of Buffalo. Patients who had abdominal and pelvic CTs or magnetic resonance imaging within 30 days before or after a colonoscopy were included.

RESULTS

One hundred two patients were included: mean age 12.7 ± 3.8 years, 66% girls. A total of 109 imaging studies were performed. Overall 61% of imaging studies were abnormal. The most frequent intestinal radiological findings were colonic wall thickening (CWT) (55%) and colonic wall enhancement (CWH) (24%). Free fluid (20%) and fat stranding (18%) were the most common extra-intestinal findings. Imaging studies agreed with histology in 81% and with colonoscopy in 75% with a moderate strength of agreement (k: 0.59 and 0.466, respectively). CWT agreed with histology in 74% with a moderate strength of agreement (k: 0.47). History of weight loss (OR 5.35, P = 0.041), chronic diarrhea (OR 4.22, P = 0.014), a positive lactoferrin (OR 7.00, P = 0.011), and presence of CWT on imaging study (OR 5.20, P = 0.001) were predictive of having abnormal histology.

CONCLUSIONS

The strength of agreement among imaging, endoscopic, and histologic findings was suboptimal. Colonoscopy and imaging are both likely to be necessary in patients with suspected inflammatory bowel disease. Although colonoscopy may be superior in diagnosis of colitis, imaging may provide more information regarding small bowel disease.

Metastatic infiltrating ductal carcinoma of the breast to the colon: a case report and literature review

True metastatic involvement of the colon is rare. Colonic metastases occur most commonly secondary to peritoneal metastases from intra-abdominal malignancies. Breast cancer is the most common malignancy that metastasizes hematogenously to the colon. Colonic metastatic disease mimics primary colonic tumors in its presentation. Colonic metastatic involvement is a poor prognostic sign, and the pathologist should be informed about the history of the primary breast cancer when examining the pathologic specimens. In this paper, we report a case of an ileocecal mass found to be histologically consistent with metastatic ductal breast cancer, and then we review the literature about breast cancer metastases to the gastrointestinal tract in general and colon in particular.

The performance of routine computed tomography for the detection of colorectal cancer

INTRODUCTION

Although colonoscopy and computed tomography (CT) colonography in expert hands are the most sensitive investigations for colorectal cancer, some patients may not tolerate the necessary bowel preparation and insufflation of gas into the colon. We assessed the performance of unprepared contrast CT for the detection of colorectal cancer.

METHODS

A retrospective review was undertaken of all patients who had contrast CT of the abdomen and pelvis and then went on to have colonoscopy at our institutions between 2007 and 2010.

RESULTS

Overall, 96 patients were identified as having had CT prior to colonoscopy. The sensitivity of CT in detecting colorectal cancer was 100% (95% confidence interval [CI]: 19.8-100%) and the specificity was 95.7% (95% CI: 88.8-98.6%). The positive predictive value was 33.3% (95% CI: 6.0-75.9%) and the negative predictive value was 100% (95% CI: 94.8-100%).

CONCLUSIONS

Non-targeted CT that is negative for colorectal malignancy is usually reassuring but the decision for further investigations should be made on a case-by-case basis, taking into account of the likelihood of underlying colorectal malignancy and the underlying co-morbidities of the patient. However, video colonoscopy is usually necessary to assess positive CT findings.

Metastatic breast cancer to the gastrointestinal tract: report of five cases and review of the literature

Luminal gastrointestinal (GI) metastases from breast cancer are rare, reports are fragmentary and poor. The purposes of this study are to assess the gastrointestinal involvement from breast cancer in a retrospective study at a single institution and reviewing the related literature. Between January 2007 and December 2011 a total of 980 patients with breast cancer were treated at our institution, patients' records and report database were analysed. Institutional Review Board approval was obtained for this study. A search of the literature using PubMed, CancerLit, Embase, was performed. Selected for the present review were papers published in English before June 2012. Five of 980 patients (0.5%) showed gastrointestinal metastases from breast cancer, 3 patients had gastric involvement, 1 jejunum, and 1 rectum. Reviewing the literature, 206 patients affected by gastrointestinal metastasis from breast cancer were identified: the most frequent site of metastasis was the stomach (60%). The majority of the patients underwent chemotherapy and endocrine therapy, someone surgery and radiotherapy. GI metastases from breast cancer are rare, but possible, and a very late recurrence can also occur. Cyto-histological diagnosis is mandatory, to differentiate GI metastases from breast cancer to other diseases and to allow an adequate treatment.