RESUMO
The development of machine learning (ML) tools in many different medical settings is largely increasing. However, the use of the resulting algorithms in daily medical practice is still an unsolved challenge. We propose an epistemological approach (i.e., based on logical principles) to the application of computational tools in clinical practice. We rely on the classification of scientific inference into deductive, inductive, and abductive comparing the characteristics of ML tools with those derived from evidence-based medicine [EBM] and experience-based medicine, as paradigms of well-known methods for generation of knowledge. While we illustrate our arguments using liver transplantation as an example, this approach can be applied to other aspects of the specialty. Regarding EBM, it generates general knowledge that clinicians apply deductively, but the certainty of its conclusions is not guaranteed. In contrast, automatic algorithms primarily rely on inductive reasoning. Their design enables the integration of vast datasets and mitigates the emotional biases inherent in human induction. However, its poor capacity for abductive inference (a logical mechanism inherent to human clinical experience) constrains its performance in clinical settings characterized by uncertainty, where data are heterogeneous, results are highly influenced by context, or where prognostic factors can change rapidly.
RESUMO
BACKGROUND & AIMS: Acute-on-chronic liver failure (ACLF) is characterized by severe systemic inflammation, multi-organ failure and high mortality rates. Its treatment is an urgent unmet need. DIALIVE is a novel liver dialysis device that aims to exchange dysfunctional albumin and remove damage- and pathogen-associated molecular patterns. This first-in-man randomized-controlled trial was performed with the primary aim of assessing the safety of DIALIVE in patients with ACLF, with secondary aims of evaluating its clinical effects, device performance and effect on pathophysiologically relevant biomarkers. METHODS: Thirty-two patients with alcohol-related ACLF were included. Patients were treated with DIALIVE for up to 5 days and end points were assessed at Day 10. Safety was assessed in all patients (n = 32). The secondary aims were assessed in a pre-specified subgroup that had at least three treatment sessions with DIALIVE (n = 30). RESULTS: There were no significant differences in 28-day mortality or occurrence of serious adverse events between the groups. Significant reduction in the severity of endotoxemia and improvement in albumin function was observed in the DIALIVE group, which translated into a significant reduction in the CLIF-C (Chronic Liver Failure consortium) organ failure (p = 0.018) and CLIF-C ACLF scores (p = 0.042) at Day 10. Time to resolution of ACLF was significantly faster in DIALIVE group (p = 0.036). Biomarkers of systemic inflammation such as IL-8 (p = 0.006), cell death [cytokeratin-18: M30 (p = 0.005) and M65 (p = 0.029)], endothelial function [asymmetric dimethylarginine (p = 0.002)] and, ligands for Toll-like receptor 4 (p = 0.030) and inflammasome (p = 0.002) improved significantly in the DIALIVE group. CONCLUSIONS: These data indicate that DIALIVE appears to be safe and impacts positively on prognostic scores and pathophysiologically relevant biomarkers in patients with ACLF. Larger, adequately powered studies are warranted to further confirm its safety and efficacy. IMPACT AND IMPLICATIONS: This is the first-in-man clinical trial which tested DIALIVE, a novel liver dialysis device for the treatment of cirrhosis and acute-on-chronic liver failure, a condition associated with severe inflammation, organ failures and a high risk of death. The study met the primary endpoint, confirming the safety of the DIALIVE system. Additionally, DIALIVE reduced inflammation and improved clinical parameters. However, it did not reduce mortality in this small study and further larger clinical trials are required to re-confirm its safety and to evaluate efficacy. CLINICAL TRIAL NUMBER: NCT03065699.