Artificial Intelligence-Augmented Epigenetic Phenotyping in Sepsis- a Pathway to Precision Medicine. A Comprehensive Narrative Review

Sepsis and septic shock remain major causes of mortality worldwide, affecting approximately 49 million people annually and accounting for an estimated 11 million deaths, representing nearly 20% of all global deaths. Despite sustained efforts in early diagnosis and treatment, sepsis continues to impose a substantial global health burden. Sepsis is defined as a life-threatening organ dysfunction secondary to a dysregulation of the host response to infection. Despite guidelines-based approaches to its diagnosis and therapy, the mortality rates remain higher even in developed countries, reaching up to 30% in intensive care units. High mortality rates in sepsis are largely driven by clinical heterogeneity, highlighting the critical role of epigenetic modulation mechanisms in shaping diverse clinical phenotypes. The main epigenetic mechanisms that modulate the immune response and outcome in sepsis are: DNA methylation, microRNA synthesis, histone modulation, and RNA methylation. This narrative review examines the role of epigenetic mechanisms in sepsis heterogeneity, phenotyping, management, and outcomes, with particular emphasis on the potential contribution of artificial intelligence to the integration of molecular and clinical data for precision medicine. Consequently, the study of AI-integrated strategies for better tailoring of the diversity of epigenetic changes should be prioritized. Thus, machine learning algorithms, deep learning frameworks, and large language models offer helpful insights in redefining the role of AI in sepsis management from this epigenetic perspective. Despite the support of AI in integrating the epigenetic mechanisms into clinical practice, some disadvantages should be also considered like: black box configuration for deep learning frameworks, inequalities from the past data used for training the models, data privacy, ethical concerns, cohort size, number of parameters considered etc. In conclusion, AI tools represent important means in tailoring the personalized decision tree, especially considering the epigenetic influence on the disease course.