Background: The COVID-19 pandemic highlighted the vulnerability of immunocompromised individuals, including liver transplant recipients (LTRs), who often exhibit reduced vaccine immunogenicity. While initial vaccine doses and subsequent boosters improved immune response, LTRs were prioritized for vaccination. Studies have shown increased antibody response after each booster dose. Vaccine hesitancy, defined as delayed or refused vaccination despite availability, poses a public health challenge, often fueled by misinformation. This study aimed to evaluate anti-spike antibody responses in vaccinated LTRs after two initial doses and at least one booster, also assessing adherence to subsequent doses. Methods: We conducted a retrospective observational study at a transplant center in Milan, Italy, between January 2021 and December 2023. LTRs who had received four or more doses of mRNA vaccines (Pfizer or Moderna) were included. Anti-spike antibody levels were measured 60–80 days after each dose. Data on vaccination status were collected in January 2024. Statistical analysis was performed to compare antibody responses and identify predictive factors. Results: LTRs showed a significant increase in anti-spike antibody responses after the first booster compared to the second dose with a trend versus a further increase following the fourth dose in a subgroup of the patients receiving two booster doses. However, adherence to booster doses decreased over time. In LTRs, predictors of a weaker response after the second dose were chronic kidney disease and metabolic etiology at transplant. Conclusions: The study highlighted that in LTRs, multiple doses of the COVID-19 vaccine led to a continuous increase in anti-spike antibody responses. The progressive decline in adherence of LTRs “to further booster doses” should be related to the fact that after the spread of vaccination programs worldwide, COVID-19 is still a current infection, but it is much less severe than before.

Background/Objective: Lactose malabsorption (LM) is common, yet not all malabsorbers develop lactose intolerance (LI). We assessed whether symptoms recorded during the hydrogen breath test (HBT) predict LM and whether methane (CH₄) measurement adds diagnostic value. Methods: In this retrospective study, 199 adults underwent HBT (January 2022–February 2025). Pre-test and intra-test symptoms were graded on a 0–10 numerical scale; LI was classified as mild, moderate, or severe. LM was defined by a ≥ 20 ppm rise in H₂; CH₄ positivity by a ≥ 10 ppm rise. Statistical analyses were performed to evaluate the association between symptoms and test results. Results: LM was present in 56.8% of patients. Flatulence during testing predicted LM (p < 0.01) and correlated with peak H₂. Severe LI doubled the likelihood of LM, whereas pre-test symptoms had no predictive value. The addition of CH₄ measurement uncovered a further 3.5% of malabsorbers, that the HBT would have missed. Methane production was observed in 19.1% of patients, and higher baseline CH₄ anticipated a greater post-challenge increase. Conclusions: Flatulence recorded during the hydrogen breath test is a reliable clinical marker of lactose malabsorption in an adult population with unexplained gastrointestinal symptoms. Grading intolerance according to symptom severity effectively identifies lactose malabsorbers. Adding methane measurements further increases diagnostic accuracy. Taken together, these findings support an individualized diagnostic pathway that blends gas profiling with standardized symptom scoring—guiding targeted management while sparing patients from unnecessarily restrictive diets.

Background: Machine learning (ML) is increasingly applied in medicine, underscoring the need for transparent and clinically relevant models. In gastrointestinal oncology, most ML studies rely on raw imaging data, which limits clinical adoption due to poor interpretability and the difficulty of collecting high-quality, large-scale video and image datasets in routine practice. Endoscopic ultrasound (EUS) plays a central role in the evaluation of pancreatic cancer; however, structured EUS features remain underused in predictive modeling. Objective: To assess the performance and interpretability of ML models for diagnosing pancreatic ductal adenocarcinoma (PDAC) using routinely collected EUS variables. Methods: We conducted a retrospective multicenter study using data from two Italian hospitals (n = 641) for model training and internal validation and from a third hospital (n = 120) for external validation, collected from 2015 to 2023. Decision trees, random forests, naïve Bayes and other classifiers were developed and evaluated. Model performance was assessed in terms of discriminative ability, calibration, and selective prediction. Results: All models demonstrated high discriminative performance (AUC ≥ 0.90). Decision trees provided the most favorable balance between interpretability and accuracy (balanced accuracy = 0.87; sensitivity = 0.89). Calibration and selective prediction analyses confirmed the robustness of the models. Conclusions: These findings demonstrate the feasibility of implementing interpretable yet high-performing ML models for PDAC diagnosis in real-life endoscopic settings.