Developing precision medicine is a major trend in clinical oncology. The main adverse effects of ifosfamide, actinomycin D and vincristine (IVA) treatment for rhabdomyosarcoma are haematological toxicities such as neutropenia or thrombocytopenia. The severity of these effects vary among patients but their dynamic profiles are similar. A non-empirical adjustment of the chemotherapy dose to avoid severe toxicities could help secure the treatment administration. Twenty-four patients with rhabdomyosarcoma treated with IVA chemotherapy courses were selected. Before and during each cycle, routine multiple blood cell counts were performed allowing for a dynamic study of the haematological toxicities. We developed a machine learning analysis using a gradient boosting regression technique to forecast the ifosfamide induced haematological toxicities as a function of neutrophils and platelets initial levels and the initial ifosfamide dose. To validate models' accuracy, predicted and observed neutrophils and platelets levels were compared. The model was able to reproduce the dynamic profiles of the haematological toxicities. Among all cycles, the mean absolute errors between predicted and observed neutrophils and platelets levels were 1.0 and 72.8 G/L, respectively. Adjusting a patient's ifosfamide dose based upon the predicted haematological toxicity levels at the end of a treatment cycle could enable tailored treatment.
Keywords: artificial intelligence; children; haematological toxicity; machine learning; neutropenia; rhabdomyosarcoma; thrombocytopenia.