Chiara Castellani

Background: Heart failure remains a leading cause of morbidity and mortality, and current therapies rarely repair established myocardial damage. Nanoparticle-based interventions have been investigated across heterogeneous models of myocardial injury, remodeling, cardiomyopathy, and heart failure, but the distribution and translational maturity of this evidence remain unclear. Methods: A systematic search of PubMed, Embase, Scopus, and Web of Science was conducted from database inception to June 2024. Eligible reports were mapped according to disease model, experimental system, carrier-level nanoparticle platform, payload, route, comparator, outcomes, biodistribution, safety assessment, and translational characteristics. Reports of non-therapeutic nanoparticle exposure were retained in a separate contextual safety/toxicology stratum and were not included in the therapeutic evidence-density map. Risk of bias was evaluated using design-appropriate tools. Results: Of 2640 records screened, 157 independent studies met the criteria: 140 in the main therapeutic/platform evidence map and 17 in a separate contextual safety/toxicology stratum. Within the main corpus, polymeric systems were the largest platform class (n = 50), followed by inorganic/mineral (n = 35), biological/biomimetic (n = 24), lipid-based (n = 23), carbon-based (n = 5), and hybrid/multicomponent systems (n = 3). Evidence was concentrated in acute myocardial injury (n = 76), while direct same-agent comparisons, long-term safety assessment, repeated dosing, quantitative biodistribution, and clinically aligned heart-failure models remained limited. Conclusions: The field demonstrates substantial formulation diversity and biological activity, but translation is constrained by fragmented characterization, sparse comparative evidence, and incomplete assessment of biological fate and safety.

Background and Objective: The analysis of high-dimensional omics data is severely hampered when only a small number of labelled samples are available, a condition common in rare or resource-intensive clinical settings. Current approaches poorly accommodate this regime: methods achieving strong predictive performance require training-set sizes that such cohorts cannot provide, whereas strategies tailored to small samples seldom couple robust classification with the interpretable, reproducible feature selection needed to derive biologically verifiable biomarkers. Methods: We present a machine-learning pipeline for binary miRNA-based phenotyping under data-scarcity constraints. Class-conditional Gaussian data augmentation, applied within training folds only to prevent leakage, addresses limited sample size and class imbalance; L1-regularised logistic regression within a nested cross-validation performs stability-based selection of miRNAs; and a kernel support vector machine classifies on the selected features. Robustness under controlled data scarcity was first assessed on a public lung cancer microRNA dataset of over 2,000 samples, training on subsets as small as 10 samples per class and testing on the remaining cohort across three binary classification tasks. As a clinical proof-of-concept, it was then applied to endomyocardial biopsies from heart transplant recipients to discriminate acute cellular rejection from cytomegalovirus infection. Results: In the lung cancer experiment, the pipeline achieved 70–88% accuracy across the three tasks, with consistent performance between cross-validation and large independent hold-out sets at the smallest training size. In the transplant proof-of-concept, on a necessarily limited cohort, it achieved 91% accuracy, outperforming a baseline logistic regression (75%) and an augmented variant (82%). Pathway enrichment of the selected microRNAs revealed biologically meaningful associations: hsa-miR-514a-3p and hsa-miR-198 with the extracellular matrix receptor interaction pathway, and hsa-miR-490-5p with the glutamatergic synapse pathway. Conclusions: The pipeline provides an interpretable framework for high-dimensional omics classification under data-scarcity constraints, designed to be platform-agnostic. The transplant findings remain a proof-of-concept given the limited cohort size, but consistent performance from minimal training sets to large independent cohorts positions it as a promising tool for translational bioinformatics and biomarker discovery where sample availability is inherently limited. Prospective validation on larger cohorts and further platforms will be conducted to translate these findings towards clinical practice.

Background and Aims: Tissular gene expression profiling applicable to formalin-fixed, paraffin-embedded (FFPE) endomyocardial biopsies (EMBs) may refine the diagnosis of cardiac rejection while being easily implemented in clinical practice. This study aimed to develop and validate the first FFPE-based molecular diagnostic system dedicated to heart transplant rejection. Methods: An international study was conducted (NCT06436027), establishing a deeply phenotyped cohort of heart transplant recipients. EMBs were graded according to international classifications, and gene expression was analysed on FFPE-EMBs using the Banff Human Organ Transplant Panel. Molecular classifiers for antibody-mediated rejection (AMR) and acute cellular rejection (ACR) were developed, with discrimination and calibration assessed in internal and external validation cohorts. Results: A total of 671 biopsies were included, with 591 in the main cohort (AMR: n = 188, ACR: n = 289, matched non-rejection: n = 114); this was split into a derivation set (n = 475) and an internal validation set (n = 116). The external validation cohort comprised 80 biopsies (AMR: n = 20, ACR: n = 32, non-rejection: n = 28). AMR was associated with significant transcripts related to the interferon-gamma pathway, endothelial activation, and monocyte-macrophage recruitment, while ACR was characterized by transcripts related to T-cell receptor signalling, CD3 receptor activation, and CD28 signalling. Molecular ACR and AMR models accurately identified rejection in the validation cohorts (internal: ROC-AUC: AMR = 0.812, ACR = 0.849; external: ROC-AUC: AMR = 0.822, ACR = 0.815) and were strongly associated with pathologic severity. Calibration was adequate. An automated report was developed to enhance the clinical applicability of these classifiers. Conclusions: A novel FFPE-based molecular diagnostic system accurately identified cardiac allograft rejection. This tool is readily applicable in clinical practice as a companion to pathology and has the potential to refine the diagnosis of rejection.

Background: Heart transplant rejection, particularly acute cellular rejection (ACR), remains a critical post-operative concern, despite declining incidence rates. Current diagnostic standards rely on invasive endomyocardial biopsy, which presents limitations in sensitivity and reproducibility. There is an unmet need for noninvasive, accurate biomarkers that can detect and monitor rejection. This study aims to evaluate whether extracellular vesicle (EV) surface antigens, analyzed through flow cytometry and interpreted with artificial intelligence (AI), can serve as reliable biomarkers for ACR detection and monitoring in heart transplant recipients. Methods: We conducted a prospective longitudinal cohort study involving 24 heart transplant recipients over a median follow-up of 303 days. A total of 285 blood samples were analyzed for EV surface antigens exploiting two flow cytometry-based protocols. An adaptive AI model (random forest regressor) was developed to interpret EV antigen profiles, dynamically calibrating thresholds per patient. Results: Here we show that 14 EV surface antigens progressively increase with ACR severity. These changes are evident even before histological diagnosis. The AI model achieves an accuracy of 93.3% at leave-one-out testing (AUC 0.968), and 78.9% at validation in an independent cohort (AUC 0.832), with high specificity and negative predictive value. EV profiling outperforms conventional biochemical markers and provides anticipatory insight into rejection dynamics. Conclusions: EV profiling, enhanced by patient-specific AI modeling, offers a powerful noninvasive method for early detection and monitoring of ACR. This approach holds the potential to reduce reliance on biopsies and tailor immunosuppressive strategies more precisely.

Given the increasing complexity of omics datasets, a key challenge is not only improving classification performance but also enhancing the transparency and reliability of model decisions. Effective model performance and feature selection are fundamental for explainability and reliability. In many cases, high dimensional omics datasets suffer from limited number of samples due to clinical constraints, patient conditions, phenotypes rarity and others conditions. Current omics based classification models often suffer from narrow interpretability, making it difficult to discern meaningful insights where trust and reproducibility are critical. This study presents a machine learning based classification framework that integrates feature selection with data augmentation techniques to achieve high standard classification accuracy while ensuring better interpretability. Using the publicly available dataset (E MTAB 8026), we explore a bootstrap analysis in six binary classification scenarios to evaluate the proposed model's behaviour. We show that the proposed pipeline yields cross validated perfomance on small dataset that is conserved when the trained classifier is applied to a larger test set. Our findings emphasize the fundamental balance between accuracy and feature selection, highlighting the positive effect of introducing synthetic data for better generalization, even in scenarios with very limited samples availability.

Population pharmacokinetic (PopPK) modelling is a fundamental tool for understanding drug behaviour across diverse patient populations and enabling personalized dosing strategies to improve therapeutic outcomes. A key challenge in PopPK analysis lies in identifying and modelling covariates that influence drug absorption, as these relationships are often complex and nonlinear. Traditional methods may fail to capture hidden patterns within the data. In this study, we propose a data-driven, model-free framework that integrates Variational Autoencoders (VAEs) deep learning model and LASSO regression to uncover key covariates from simulated tacrolimus pharmacokinetic (PK) profiles. The VAE compresses high-dimensional PK signals into a structured latent space, achieving accurate reconstruction with a mean absolute percentage error (MAPE) of 2.26%. LASSO regression is then applied to map patient-specific covariates to the latent space, enabling sparse feature selection through L1 regularization. This approach consistently identifies clinically relevant covariates for tacrolimus including SNP, age, albumin, and hemoglobin which are retained across the tested regularization strength levels, while effectively discarding non-informative features. The proposed VAE-LASSO methodology offers a scalable, interpretable, and fully data-driven solution for covariate selection, with promising applications in drug development and precision pharmacotherapy.

BACKGROUND: Thrombosis is linked to neutrophil release of neutrophil extracellular traps (NETs). NETs are proposed as a mechanism of resistance to thrombolysis. This study intends to analyze the composition of thrombi retrieved after mechanical thrombectomy, estimate the age and organization of thrombi, and evaluate associations with the use of thrombolysis, antiplatelets, and heparin. METHODS: This retrospective observational study involved 72 samples (44 from cerebral and 28 coronary arteries), which were stained with hematoxylin and eosin, anti-NE (neutrophil elastase) antibody, and anti-histone H2B (histone H2B) antibody, representing different components in NET formation, all detectable during the later stages of NETosis, for histochemical and digital quantification of NET content. The histological and morphological evaluations of the specimens were correlated, through univariate and mediation analyses, with clinical information and therapy administered before intervention. RESULTS: The results demonstrated that the composition of cerebral and coronary thrombi differs, and there were significantly more lytic cerebral thrombi than coronary thrombi (66% versus 14%; P =0.005). There was a considerably higher expression of NETs in the cerebral thrombi as testified by the higher expression of H2B ( P =0.031). Thrombolysis was remarkably associated with higher NE positivity (average marginal effect, 6.461 [95% CI, 0.7901–12.13]; P =0.02555), regardless of the origin of thrombi. There was no notable association between the administration of antiaggregant therapy/heparin and H2B/NE amount when adjusted for the thrombus location. Importantly, the age of the thrombus was the only independent predictor of NET content without any mediation of the thrombolytic treatment ( P =0.014). CONCLUSIONS: The age of the thrombus is the driving force for NET content, which correlates with impaired clinical outcomes. The therapy that is currently administered does not modify NET content. This study supports the need to investigate new pharmacological approaches added to thrombolysis to prevent NET formation or enhance their disruption, such as recombinant human DNase I (deoxyribonuclease I).

Human papilloma virus (HPV) is known as the main cause of cervical cancer. Data also indicate its role in head–neck cancer, especially oropharyngeal cancer. The correlation between high-risk HPV and oral cancer is still controversial. HPV-related lesions of the oral cavity are frequent and, in most cases, benign. The primary aim of this study was to establish if there is a different follow-up necessity between HPV-positive compared to HPV-negative oral lesions. The secondary aim was to evaluate the recurrence of HPV-related lesions. All patients who underwent a surgical procedure of oral biopsy between 2018 and 2022, with ulterior histopathological examination and HPV typing, were examined. A total of 230 patients were included: 75 received traumatic fibroma as diagnosis, 131 HPV-related lesions, 9 proliferative verrucous leukoplakia, and 15 leukoplakia. The frequency and period of follow-up varied in relation to HPV positivity and diagnosis. This study confirms what has already been reported by other authors regarding the absence of recommendations of follow-up necessity in patients with oral mucosal lesions. However, the data demonstrate that there was a statistically significant difference in the sample analyzed regarding the follow-up of HPV-positive vs. HPV-negative patients. It also confirms the low recurrence frequency of HPV-related oral lesions.

Introduction: Vitiligo is an acquired chronic pigmentation disorder of the skin. Even if the role of the immune system seems to be well established, new pathogenetic hypothesis are rising in these years. It has been recently suggested by the development of an animal model that a protein called Melanoma Inhibitory Activity (MIA) is involved in the pathogenesis of vitiligo. This protein interacts with the adhesion molecules expressed on the melanocytes causing its detachment from extracellular matrix proteins and creating the depigmented macules. A topical preparation based on oligopeptides able to inhibit the actions of the MIA protein has been introduced to the market, claiming activity on vitiligo. Patient concerns and diagnosis: A patient affected by non-segmental vitiligo for 10 years, recalcitrant to any treatment (such as steroids, immunomodulators, kellin, UVB-NB and UVA) came to our observation. Interventions: We used this topical preparation containing the MIA inhibitors peptides in selected areas (face and sides of the trunk) leaving untreated other areas as control (legs and arms). The patient was required to be sun exposed or to have some UVA sessions during the treatment to stimulate the melanocytes replications. Outcomes: After 9 months of treatments, he recovered from 50% to 80% of repigmentation only in the treated areas, without any side effects locally or systemically. Conclusion: Even if other studies are required to better determine the efficacy of this approach, this first observation about the use of the MIA-inhibitors peptides for the treatment of non-segmental vitiligo indicates that this topical preparation containing the MIA inhibitors peptides could be a very promising option for the cure of this disease.

Background : The present study evaluates outcomes of LVAD patients, taking into account the device strategy and the INTERMACS profile. Methods: We included 192 LVAD-patients implanted between January 2012 and May 2021. The primary and secondary end-points were survival and major adverse events between Profiles 1–3 vs. Profile 4, depending on implantation strategies (Bridge-to-transplant-BTT; Bridge-to-candidacy-BTC; Destination-Therapy-DT). Results: The overall survival was 67% (61–75) at 12 months and 61% (54–70) at 24 months. Profile 4 patients showed significantly higher survival (p = 0.018). Incidences of acute right-ventricular-failure (RVF) (p = 0.046), right-ventricular-assist-device (RVAD) implantation (p = 0.015), and continuous-venovenous-hemofiltration (CVVH) (p = 0.006) were higher in Profile 1–3 patients, as well as a longer intensive care unit stays (p = 0.050) and in-hospital-mortality (p = 0.012). Twelve-month and 24-month survival rates were higher in the BTT rather than in BTC (log-rank = 0.410; log-rank = 0.120) and in DT groups (log-rank = 0.046). In the BTT group, Profile 1–3 patients had a higher need for RVAD support (p = 0.042). Conclusions: LVAD implantation in elective patients was associated with better survival and lower complications incidence. LVAD implantation in BTC patients has to be considered before their conditions deteriorate. DT should be addressed to elective patients in order to guarantee acceptable results.