Development of a gastric cancer risk calculator for questionnaire-based surveillance of Iranian dyspeptic patients.

BACKGROUND: Gastric cancer (GC) is considered a silent killer, taking more than three quarters of a million lives annually. Therefore, prior to further costly and invasive diagnostic approaches, an initial GC risk screening is desperately in demand. METHODS: In order to develop a simple risk scoring system, the demographic and lifestyle indices from 858 GC and 1132 non-ulcer dyspeptic (NUD) patients were analysed. We applied a multivariate logistic regression approach to identify the association between our target predictors and GC versus NUD. The model performance in classification was assessed by receiver operating characteristic (ROC) analysis. Our questionnaire covering 64 predictors, included known risk factors, such as demographic features, dietary habits, self-reported medical status, narcotics use, and SES indicators. RESULTS: Our model segregated GC from NUD patients with the sensitivity, specificity, and accuracy rates of 85.89, 63.9, and 73.03%, respectively, which was confirmed in the development dataset (AUC equal to 86.37%, P < 0.0001). Predictors which contributed most to our GC risk calculator, based on risk scores (RS) and shared percentages (SP), included: 1) older age group [> 70 (RS:+ 241, SP:7.23), 60-70 (RS:+ 221, SP:6.60), 50-60 (RS:+ 134, SP:4.02), 2) history of gastrointestinal cancers (RS:+ 173, SP:5.19), 3) male gender (RS:+ 119, SP:3.55), 4) non-Fars ethnicity (RS:+ 89, SP:2.66), 5) illiteracy of both parents (RS:+ 78, SP:2.38), 6) rural residence (RS:+ 77, SP:2.3), and modifiable dietary behaviors (RS:+ 32 to + 53, SP:0.96 to 1.58). CONCLUSION: Our developed risk calculator provides a primary screening step, prior to the subsequent costly and invasive measures. Furthermore, public awareness regarding modifiable risk predictors may encourage and promote lifestyle adjustments and healthy behaviours.

A Bayesian latent class extension of naive Bayesian classifier and its application to the classification of gastric cancer patients.

BACKGROUND: The Naive Bayes (NB) classifier is a powerful supervised algorithm widely used in Machine Learning (ML). However, its effectiveness relies on a strict assumption of conditional independence, which is often violated in real-world scenarios. To address this limitation, various studies have explored extensions of NB that tackle the issue of non-conditional independence in the data. These approaches can be broadly categorized into two main categories: feature selection and structure expansion. In this particular study, we propose a novel approach to enhancing NB by introducing a latent variable as the parent of the attributes. We define this latent variable using a flexible technique called Bayesian Latent Class Analysis (BLCA). As a result, our final model combines the strengths of NB and BLCA, giving rise to what we refer to as NB-BLCA. By incorporating the latent variable, we aim to capture complex dependencies among the attributes and improve the overall performance of the classifier. METHODS: Both Expectation-Maximization (EM) algorithm and the Gibbs sampling approach were offered for parameter learning. A simulation study was conducted to evaluate the classification of the model in comparison with the ordinary NB model. In addition, real-world data related to 976 Gastric Cancer (GC) and 1189 Non-ulcer dyspepsia (NUD) patients was used to show the model's performance in an actual application. The validity of models was evaluated using the 10-fold cross-validation. RESULTS: The presented model was superior to ordinary NB in all the simulation scenarios according to higher classification sensitivity and specificity in test data. The NB-BLCA model using Gibbs sampling accuracy was 87.77 (95% CI: 84.87-90.29). This index was estimated at 77.22 (95% CI: 73.64-80.53) and 74.71 (95% CI: 71.02-78.15) for the NB-BLCA model using the EM algorithm and ordinary NB classifier, respectively. CONCLUSIONS: When considering the modification of the NB classifier, incorporating a latent component into the model offers numerous advantages, particularly within medical and health-related contexts. By doing so, the researchers can bypass the extensive search algorithm and structure learning required in the local learning and structure extension approach. The inclusion of latent class variables allows for the integration of all attributes during model construction. Consequently, the NB-BLCA model serves as a suitable alternative to conventional NB classifiers when the assumption of independence is violated, especially in domains pertaining to health and medicine.

Helicobacter pylori acquisition rates and the associated risk factors amongst newlywed couples; a prospective cohort study in Tehran, Iran.

BACKGROUND: The rates and routes of Helicobacter pylori transmission, in a high-prevalent country like Iran, with gastric cancer as the leading cause of male cancer mortality, are of great essence. Here, we have studied the H. pylori-associated risk factors and the likelihood of interspousal transmission. METHODS: In a cohort of 686 young prewed couples, questionnaires were self-administered and serum samples were collected, for assessment of risk factors and H. pylori serostatus, at baseline and follow-up. Of the 475 H. pylori single- or double-seronegative couples, 201 returned for follow-up. The average follow-up duration was 2.2 (SD 0.6) years, with a total of 560.1 person-years. Logistic regression and Cox regression models were used to estimate the odds ratios (ORs) and hazard ratios (HRs). RESULTS: The risk of infection was higher in men than women (OR: 1.3, 95% CI: 1.0-1.8) and among metropolitan than rural residents (OR = 1.4, 95% CI: 1.1-1.9). It was also significantly higher among those with three (OR = 1.6, 95% CI: 1.1-2.2), and four or more siblings (OR = 1.4, 95% CI: 1.0-1.9), in reference to those with one or no siblings. Adult H. pylori acquisition occurred in 10.9% (27/247) of the seronegative participants. The risk of the acquisition was significantly associated with age (P value for trend=0,000). It was also significantly lower among participants who had various degrees of education as compared to illiterate subjects (HR = 0.2, 95% CI: 0.1-0.9). Nevertheless, our analysis did not find any evidence for interspousal transmission (HR = 1.0, 95% CI: 0.4-2.2). CONCLUSION: Whilst H. pylori acquisition was detected in the young adult Iranian population, our findings did not support interspousal transmission, as a mode of acquisition.

dupA as a risk determinant in Helicobacter pylori infection.

The Helicobacter pylori duodenal ulcer promoting (dupA) gene has been previously described as a risk marker for duodenal ulcer (DU) development and a protective factor against gastric cancer (GC). Recent studies which have assessed the application of dupA in the prediction of clinical outcomes have been controversial. In the current study, the association of dupA with the clinical outcomes and histopathological changes following H. pylori infection was evaluated in Iranian patients. A total of 157 H. pylori-infected patients with DU (n=30), gastric ulcer (n=23), gastritis (n=68) or GC (n=36) were assessed. The presence of jhp0917 and jhp0918 genes was determined by gene specific PCR. Gastric histopathological changes were recorded according to the updated Sydney system. Seventy-eight (49.7 %) and 71 (45.2 %) of the 157 tested strains, respectively, were positive and negative for both genes. The remaining 8 (5.09 %) of the 157 strains were jhp0917-positive/jhp0918-negative. Univariate analysis showed inverse associations between dupA and histological features including dysplasia as the penultimate stage of GC and lymphoid follicles as a consequence of relatively long-standing H. pylori-associated gastritis. The degrees of nucleotide sequence identity of Iranian strains to Colombian, Brazilian and Indian strains ranged from 86.1 to 100 % for the aligned regions of jhp0917, from 88 to 98.8 % for jhp0918 and from 93.4 to 99.5 % for the partial sequences of the dupA gene. Despite the fact that possession of the dupA gene showed no association with any disease category in our population as reported in several other countries, association of dupA-negative strains of H. pylori with pre-malignant lesions calls for additional studies to evaluate the role of this gene as a protective marker against GC.