Title:
Principles of clinical reasoning in personalized oncology with limited evidence
Abstract:
As knowledge of the micro-physiological mechanisms underlying carcinogenesis advances, personalizing prognostic reasoning and decision-making becomes an increasingly important issue. Personalized oncology enables interventions sensitive to inter-individual variability, resisting an excessive standardization of medical care, and stressing each patient’s specific clinical history and personal preferences. This presentation introduces clinical logic as a normative and epistemic framework for reasoning under uncertainty in the context of personalized oncology. It outlines a systematic procedure to justify the aggregation and integration of evidence from population studies and multi-omic clinical records to motivate therapeutic decisions at the patient’s bedside. In doing so, this approach departs from evidence-based medicine, which tends to evaluate the effectiveness of medical interventions through fixed epistemic hierarchies, privileging randomized controlled trials and meta-analyses over non-statistical evidence of physiological mechanisms. Clinical logic implements tools from direct inference and probability theory to estimate the measure of support that the available evidence confers to a hypothesis of effectiveness related to a specific treatment, and often requires substantive consideration of the methodologies through which this evidence is generated. This framework is applied to assess hypotheses of effectiveness in the field of rare cancers, in which the lack of robust statistical evidence needs to be compensated by knowledge of the physio-pathological mechanisms influencing prognosis. The presentation then discusses some of the methodological challenges that the nature of rare cancers seems to raise for the application of clinical logic to the individual patient. These include the fact that therapeutic effectiveness in oncology is rarely apparent and that oncological prognostic profiles are subject to often dramatic changes over time.
The talk will be presented on Zoom.