Advancing Early Detection of Colorectal Adenomatous Polyps via Genetic Data Analysis: A Hybrid Machine Learning Approach

In this study, a hybrid machine learning (HML)-based approach, incorporating Genetic data analysis (GDA), is proposed to accurately identify the presence of adenomatous colorectal polyps (ACRP) which is a crucial early detector of colorectal cancer (CRC). The present study develops a classification ensemble model based on tuned hyperparameters. Surpassing accuracy percentages of early detection approaches used in previous studies, the current method exhibits exceptional performance in identifying ACRP and diagnosing CRC, overcoming limitations of CRC traditional methods that are based on error-prone manual examination. Particularly, the method demonstrates the following CRP identification accuracy data: 97.7 ± 1.1, precision: 94.3 ± 5, recall: 96.0 ± 3, F1-score: 95.7 ± 4, specificity: 97.3 ± 1.2, average AUC: 0.97.3 ± 0.02, and average p-value: 0.0425 ± 0.07. The findings underscore the potential of this method for early detection of ACRP as well as clinical use in the development of CRC treatment planning strategies. The advantages of this approach are highly expected to contribute to the prevention and reduction of CRC mortality.

BRAF, K-ras and BAT26 mutations in colorectal polyps and stool

AIM： To assess the feasibility of using BRAF, K-ras and BAT26 genes as stool-based molecular markers for detection of colorectal adenomas and hyperplastic polyps （HPs）. METHODS： We applied PCR-SSCP and direct sequencing to detect BRAF mutations of polyps and paired stool samples. Primer-mediated restriction fragment length polymorphism （RFLP） analysis and mutant-enriched PCR were used in detection of K-ras mutations of polyp tissues and paired stool samples respectively. BAT26, a microsatellite instability marker was examined by detection of small unstable alleles in a poly （A） repeat. RESULTS： No genetic alterations were detected in the 36 colonoscopically normal patients in either tissues or stools. BRAF, K-ras and BAT26 mutations were found in 4 （16%）, 10 （40%） and 3 （12%） of 25 adenoma tissues and among them, 75%, 80% and 100% of patients were observed to contain the same mutations in their corresponding stool samples. In HPs, mutations of BRAF and K-ras were detected in the tumor DNA of 2 （11.1%） and 8 （33.3%） of 18 patients respectively, all of whom had identical alterations in their stools. Taken together, the three genetic markers detected 15 （60%） of 25 adenomas and 8 （44.4%） of 18 HPs. The sensitivity of stool detection was 80% for adenomas and 100% for HPs with an overall specificity of 92% for adenomas and 100% for HPs. CONCLUSION： BRAF, K-ras and BAT26 genes have the potential to be molecular markers for colorectal adenomas and HPs, and can be used as non-invasive screening markers for colorectal polyps.

CLINICAL VALUE OF ALARM SYMPTOMS IN DIAGNOSIS OF LOWER GASTROINTESTINAL DISEASES

Objective To evaluate the clinical value of alarm symptoms in diagnosis of lower gastrointestinal diseases. Methods Data of consecutive autochthonous patients referred to the endoscopy center of Renji Hospital during the period of Oct. 2002 to Dec. 2003 were retrospectively analyzed. The prevalence of alarm symptoms in Shanghai patients with colorectal malignancies or other organic or functional lower gastrointestinal diseases was investigated. Results 83 （4. 9% ） cases of colorectal malignancies were found in 1681 patients referred to the center for colonoscopy because of lower gastrointestinal symptoms. All these malignancies were verified to be in the progressive stage. The prevalence of alarm symptoms was 81.9% （68/83）. Hematochezia （ OR 4. 1, 95% CI 3.3 -5.2, P 〈0. 001）, melena （0R6.4, 95%CI3. 7-11. 0, P 〈0. 001） and anemia （OR 9.6, 95%C13. 7 - 25. 0, P 〈 0. 001 ） were the most common and specific alarm symptoms. All the patients without alarm symptoms were above the age of 40 years. 264 （15. 7% ） cases of organic colorectal diseases other than malignancies and 1334 （ 79. 4% ） cases with no causal pathology identified were found in 1681 patients, and the prevalence of alarm symptoms in these two groups was 48. 5% （128/264） and 14. 8% （197/1334）, respectively. Conclusion Alarm symptoms including hemotochezia, melena, and anemia were useful in distinguishing organic from functional colorectal diseases in patients over 40 years old at the onset of symptoms. Furthermore, hematochezia, melena, anemia, severe weight loss, and abdominal mass were helpful in differentiating malignant from non-malignant colorectal diseases. Colonoscopy should be recommended for patients regardless of age with these alarm symptoms, and so do patients above the age of 40 years with no alarm symptoms before the diagnosis of functional diseases are made.