Science and art of anticancer drug dosing: nine steps to personalised therapy

Stephen P. Ackland, Michael Michael, Paul de Souza, Jennifer H. Martin, Stephen J. Clarke, Kay Francis, Christos S. Karapetis, Howard Gurney

    Research output: Contribution to journalArticlepeer-review


    Cancer medicine is a challenging field with an increasing range of promising therapies and combinations. Increasingly, personalised medicine shows promise to improve cancer outcomes – response, symptom control, survival and cure. However, optimal dosing is an under-appreciated aspect of personalised anticancer therapy, with few clinical trials addressing this specific issue. This position paper aims to inform various health professionals about the principles that guide anticancer drug dose selection and modifications. We discuss the available evidence base for personalised dosing, as well as the professional judgement required by experienced oncology physicians to determine the most appropriate dose for each patient. We provide nine steps to guide clinicians and trainees in this process, based on: pharmacology of each agent (absorption, distribution, metabolism, elimination and mechanism of action); scientific evidence for recommended doses; professional knowledge of patients' unique phenotype (adiposity, comorbidities, etc.); previous drug tolerance; individual dose adjustment in combination therapy; communication and documentation, with the added need for ongoing monitoring and adjustment. We strongly propose professional education and future research towards optimal dosing.

    Original languageEnglish
    Pages (from-to)992-996
    Number of pages5
    JournalInternal Medicine Journal
    Issue number8
    Publication statusPublished - 1 Aug 2020


    • cancer chemotherapy
    • drug dosing
    • personalisation
    • personalised medicine
    • pharmacology


    Dive into the research topics of 'Science and art of anticancer drug dosing: nine steps to personalised therapy'. Together they form a unique fingerprint.

    Cite this