Valutazione delle interazioni farmacologiche con gli inibitori della tirosin-chinasi nella leucemia mieloide cronica: valutazione della precisione di DDI Predictor per la personalizzazione dei regimi terapeutici

Abstract Chronic Myeloid Leukaemia (CML) is a clonal, myeloproliferative disease; it was the first hematological disease associated with a particular genetic defect. Myelo-proliferative diseases, like CML, are characterized by the uncontrolled growth of myeloid cells.  CML was identified for the first time in 1845. The chromosomal aberration typical of CML was identified in 1960 and then called the Philadelphia chromosome. Since the introduction of tyrosine kinase inhibitor (TKI) drugs (at present there are six TKIs approved for therapy), survival rates have improved significantly. Imatinib (Glivec ® ) was the first TKI and it has now a history of two decades of safety and efficacy. Other TKIs are dasatinib, nilotinib, bosutinib, ponatinib and asciminib.  Since CML patients can also be polypathological patients, it is possible that two or more molecules in their therapy follow the same metabolic pathway, and, as a result, drug-drug interactions may occur, causing an increased or decreased plasma concentration of the drugs. As a part of this work, different interactions were divided in three categories, reflecting their strength: weak, moderate, and strong. The consequences of these interactions can be: reduced efficacy of the therapy, or toxicity. To limit these consequences, assessments of the potential drug- drug interactions are conducted in clinical contexts, quantifying their magnitude, with the aim of taking appropriate measures to mitigate associated risks. These predictions are generated using computational tools that enable the prediction of the magnitude or severity of these interactions. As a result, it is possible to tailor the dosages accordingly. The primary objective of this work was to establish the feasibility of utilizing prediction tools for drug-drug interactions involving TKIs, and the quality of these predictions. To achieve this goal, a comparative analysis was conducted, contrasting data derived from the prediction tool DDI Predictor with data extracted from human clinical studies documented in the existing literature. To facilitate a practical comparison of the data, a linear regression analysis was employed. After comparing the data using this method, we can conclude that the results of this study affirm the efficacy and reliability of this prediction tool in assessing drug-drug interactions, particularly within the context of TKIs. These findings lend strong support to the practical utility of this tool for clinicians in tailoring dosages effectively and managing patient treatments.