Introduction

Precision oncology approaches have enabled matching of the best available anticancer drug to each individual patient’s particular tumor type. Nevertheless, treatment even with such targeted agents is often short-lived or unsuccessful. One factor that may significantly contribute to failure of treatment, despite successful selection of the correct drug, is the failure to select the correct drug dose.
Oral small molecule anticancer drugs are typically prescribed at fixed drug doses regardless of patient weight, age, or gender. Differences in bioavailability, metabolism, and adherence further increase pharmacokinetic (PK) variability from patient to patient. This means that, despite being prescribed the same dose of the drug, individual patients may have significant differences in the resultant blood drug levels (i.e. drug exposure), with some being under-dosed and others over-dosed.
One efficient way to optimize drug dosing is through therapeutic drug monitoring (TDM) of patient blood drug levels. TDM is most effective for the subset of drugs for which the inter-patient variation in systemic drug exposure (e.g. blood drug levels) is high relative to the therapeutic index (concentration range of adequate efficacy and minimal toxicity), and for which the exposure is strongly correlated with clinical response. For such drugs, a flat-fixed dose will be optimal for only a fraction of patients. Numerous oral small molecule anticancer drugs that are already in clinical use have been shown to exhibit such high inter-patient PK variabilities and strong pharmacokinetics-pharmacodynamics (PK-PD) relationships, and seem to be strong candidates for benefiting from TDM as described in several excellent reviews on this subject.1–5
Candidate drugs for TDM can be actively identified throughout various phases of drug development, which includes the post-marketing phase. For newly approved drugs and those in clinical development, critical analysis of PK and PD data may provide clues to potential benefits of dose individualization. PK and PD data is typically available for every anticancer drug on the market or in late phases of clinical development, but the type of analysis of this data for the purpose of further drug development differs widely from company to company. Regulatory authorities usually only require non-compartmental analysis of PK data, and the relationships with therapeutic and adverse drug effects are often explored only with correlative analysis. Alternatively, PK and PKPD modeling and simulation can provide additional insights into the often complex relationship between dose and effects and, in fact, there are several efforts from the regulatory authorities to encourage the use of modeling and simulation in drug development.6,7Consequently, the results from such additional analyses can be quite useful for selection of optimal dose regimens for further clinical studies, identification of candidates for TDM and, ultimately, for patient care.
Drug dosing regimens are typically established during early phase trials involving a small number of participants, with dosing decisions based on population level data rather than individual level data. Late phase trials often do not even include a PK component, which in the context of establishing dosing for optimal drug exposure may be a missed opportunity. Compared to early phase trials, phase 3 studies are usually much larger and longer (involving hundreds to thousands of participants over months to years), and are therefore better suited for accurate characterization of inter- and intra-patient PK variability as well as for determination of the relationships between systemic exposure, relevant clinical outcome parameters, and side effects. The PK and PD results that could be obtained in such studies (e.g. PK variability and exposure-efficacy/toxicity relationships) would inform the need for optimization and perhaps individualization of drug dosing for subsequent studies and for patient care.
This article discusses some important considerations for evaluating the utility of TDM for new oral anticancer drugs during clinical trials as well as during patient care. Similar to other types of clinical interventions, studies evaluating TDM of oral small molecule oncology drugs are likely to progress through several stages of increasing complexity, with additional levels of evidence in support of the intervention being gathered at each stage. To help conceptualize this progression as it relates to TDM of oral small molecule oncology drugs, we have subjectively categorized it into four stages (Figure 1) that are detailed in the following sections. We first focus on measuring systemic drug exposure to assess inter- and intra-patient PK variability, followed by correlating drug exposure to efficacy and toxicity, then evaluating whether TDM can effectively optimize drug exposure, and finally testing whether the implementation of TDM-guided precision dosing, i.e. adjusting an individual’s drug dose based on measured blood drug concentrations, might improve clinical outcomes indeed.