Therapeutic drug monitoring—Does it really matter?

Abstract

Therapeutic drug monitoring (TDM) is a vital process in clinical pharmacology that enables healthcare providers to tailor medication regimens to individual patients, maximizing therapeutic effectiveness while minimizing potential side effects. Since its inception in the early 1970s, TDM has become an integral part of managing certain medications, often referred to as ‘narrow therapeutic index’ (NTI) drugs that exhibit significant variability in absorption, distribution, metabolism, excretion and patient response. The objective of this editorial is to examine the current role and value of TDM in drug treatment of human patients.

A key component of TDM includes the determination of specific drug (and/or metabolite) concentrations in a biological matrix, typically blood, plasma or serum using a selective bioanalytical assay. Modern high-resolution mass spectrometry-based bioanalytical methods have significantly advanced traditional TDM practices. However, assay inconsistencies may negatively affect clinical decisions, drug adjustments, and patient outcomes, thus highlighting the importance of traceability, standardized reference materials, and the establishment of appropriate reference procedures for TDM. Secondly, plasma/blood concentrations must be interpreted in the context of individual patient factors such as age, body weight, organ impairment, pharmacogenomics, comorbidities and concomitant drug therapy. This personalized medicine approach ensures drug concentrations remain within a therapeutic range that is both effective and safe. Thirdly, TDM is required for a limited subset of therapeutic drugs, particularly those with a NTI. For NTI drugs, TDM is a cornerstone in personalized medicine and should be interpreted alongside clinical guidelines, laboratory values and patient symptoms to guide therapy effectively. This collaborative approach is especially crucial in managing complex cases, such as in those involving the elderly patients and/or those with renal and/or hepatic impairment. The usefulness of TDM measurements has sometimes been questioned due to significant inter-occasion variability, including random variation between dosing occasions unrelated to inherent changes in drug exposure. Additionally, residual unexplained variability may arise from sample handling procedures assay conditions, model misspecification or other intra-individual factors.^{1, 2}

This special issue was stimulated by a recent themed issue in BJCP, where Holford et al. raised the concern that TDM focuses solely on using the determined drug concertation to asses whenever a patient is under- or overexposed, without proposing dose adjustment.³ In a recent opinion paper, Lemaitre and Hesselink agreed with Holford et al. regarding the need for TDM to evolve toward guiding subsequent drug doses using algorithm-based pharmacokinetic-pharmacodynamic (PK-PD) population models.⁴ This evolution of traditional TDM is referred to as model-informed precision dosing (MIPD), which represents a shift toward personalized drug dosing tailored to individual patient characteristics. MIPD moves beyond traditional TDM by integrating mathematical predictions of dosing and accounting for patient-specific factors (e.g., patient characteristics, drug measurements), alongside various sources of variability (e.g., characterized through population PK models). The research, development and validation of mechanism-based models across various disease areas represent a significant innovation, as there is a tremendous potential to integrate biomarkers into pharmacometric models together with complete pharmacogenetic profiles to fully establish personalized medicine. This approach considers not only drug exposure but also individual treatment responses.⁵ A recent study proposed a method for case-by-case assessment of the benefits of TDM versus pharmacogenetic testing.¹ In this special issue, Udomkarnjananun et al. discuss how the progression of personalized treatment depends on technical advancement, including the development of techniques for target-site concentration measurement and biomarker quantification. Patient convenience is another critical focus, with innovations emphasizing outpatient self-monitoring of drug concentrations (e.g., microsampling) as discussed by van Gelder et al. in this special issue.

One question that could be raised is: Why it is hard to demonstrate benefit of TDM? If the objective is to show benefit with TDM, a clinical trial in the target patient population would need to demonstrate that a TDM-guided approach achieves a high rate of safety and effectiveness compared to fixed dosing. This would likely require more study participants than the original phase 3 clinical trials, which were typically designed to compare a fixed dose against placebo rather than superiority to a treatment previously shown to be effective. A review of the literature suggests that most trials investigating the benefits of TDM include far fewer participants than the original registrational trials and are often not designed to account for the magnitude of the expected difference between TDM-guided dosing and fixed dosing.

TDM is currently costly due to the need for individualized sampling, high-quality and rapid bioassays, and the involvement of qualified personnel (e.g., clinical pharmacologists and clinical pharmacists) for interpretation and dose adjustment. These factors may limit its widespread use in many countries. However, it is evident that many physicians consult TDM experts in cases where patients are difficult to treat. For instance, tuberculosis (TB) patients with multidrug-resistant strains may require higher-than-recommended doses⁶

Another approach to improving drug treatment is development of new formulations that reduce pharmacokinetic variability. After switching to these new and improved formulations, several drugs have demonstrated reduced inter- and intra-individual variability in plasma exposure.^{7, 8}

In this special issue, 10 articles explore TDM from various perspectives. These include the site of concentration measurements (intracellular vs. whole blood), the potential and challenges illustrated with various drugs, patient-controlled home-based self-monitoring of drug concentrations, the role of the clinical pharmacists in drug management and finally, an economic evaluation of the cost-effectiveness of pharmacokinetically guided dosing of 5-fluorouracil (5-FU) in patients with metastatic colorectal cancer.

In one paper in this special issue, Udomkarnjananun et al. raise the interesting question of whether TDM of tacrolimus should be based on intracellular concentrations within T lymphocytes rather than whole blood. The determination of the intracellular tacrolimus concentration is still in its early stages of development and not yet ready for clinical use. However, the authors identify potential clinical advantages, and promising findings from initial studies suggest that further investigations are warranted to address the technical issues and clinical applications of this novel approach.⁹

In one review, Bergan and Vethe discuss the medical evidence available for biologicals used in solid organ transplantation (SOT) and the potential for improving treatment through TDM and MIPD. The various biologics currently used in the treatment of SOP are often dosed off-label, relying on the clinical experience from their use on labelled indications. The authors conclude that the benefits of TDM and MIPD have not yet been fully recognized, and the future challenge lies in designing and performing sufficiently large clinical trials in well-defined, real-world patient populations to establish their usefulness.¹⁰ Using an example of valganciclovir prophylaxis for cytomegalovirus (CMV) infection prevention after SOT, the authors propose that further potential improvements in this crucial prophylactic therapy, may be achievable through personalized medicine and TDM. These include reducing risks of myelotoxicity, late CMV infection and viral resistance. They stress that extensive well-designed clinical investigations are needed to explore these possibilities.¹¹

In a futuristic article, Hazenbroek et al. suggest a strategy for patient-controlled, home-based TDM of the immunosuppressant tacrolimus following SOT. While current TDM practice will remain standard of care, they are expensive and resource intensive, highlighting the need for further development. One potentially successful strategy involves integrating patient. Controlled, home-based dosing with dosing algorithms derived from population data and individual tacrolimus exposure measured using dried blood samples (DBS) and telemedicine. The authors believe this approach could enable precise tacrolimus dosing, increase patient adherence and participation, improving quality of life and reduced healthcare-associated costs.¹²

Ingelman-Sundberg and Molden predict the future use of TDM, liquid biopsies and pharmacogenomics for predicting human drug metabolism and response. They identify that liquid biopsies, together with pharmacogenomics and TDM, could become useful in clinics settings. However, they emphasize that larger clinical trials are necessary to establish this approach. These larger studies must include well-characterized patients and carefully control for confounding factors such as diet, co-medication and physical health.¹³

Nersesjan et al. present a case report in which paliperidone, in its long-acting parenteral formulation, lead to poisoning. They highlight how TDM might be applied as a tool in such cases, noting that treatment with St. John's wort, an inducer of elimination, may reduce overexposure of the antipsychotic. The authors underscore the importance of determining how such long-lasting and extensive exposures can be avoided.¹⁴

Brown et al. report on a study where a secure analytics platform for electrotonic patient records. OpenSAFELY was used to safely deliver health services in the United Kingdom during pandemic lockdowns. Disease-modifying antirheumatic drugs (DMARDs) may cause serious adverse effects, requiring patients to adhere to regular safety monitoring. As the COVID-19 pandemic disrupted general monitoring services in England, the use electrotonic patient records became essential.¹⁵

Van Gelder et al. analyse whether TDM for tacrolimus after kidney transplantation should be based on trough (or predose) concentration or area-under-curve (AUC) monitoring. Tacrolimus is a NTI drug with a high inter-patient variability in PK, and TDM is widely accepted to individualizing doses to prevent rejection and toxicity. Van Gelder also proposes that in the future, patients could take repeated samples themselves to determine AUC. The authors expect that this approach could lead optimized tacrolimus exposure, representing a valuable advancement in the clinical use of this cornerstone drug in SOT.¹⁶

Mawardi et al. contributed a short essay emphasizing the role of the clinical pharmacists in healthcare teams. One of their conclusions was that optimization of drug therapy will require increased interprofessional communication across different parts of the healthcare system.¹⁷

In the final paper of this special issue, Erku et al. argue that pharmacokinetic dose management of 5-FU for metastatic colorectal cancer (mCRC) patients appears to be a cost-saving strategy in Australia. This is partly due to enhanced efficacy and fewer adverse events, despite some uncertainties in model assumptions. Their analysis incorporates direct healthcare costs, quality-adjusted life years (QALYs) and incremental cost-effectiveness ratios (ICERs) and includes both one-way and probabilistic sensitivity analyses.¹⁸

In conclusion, based on the current status and the contributions in this special issue, it is clear that TDM undoubtedly plays a critical role in optimizing drug therapy for specific patient populations, particularly for drugs with narrow therapeutic indices or high inter-patient variability. However, its broader application and universal importance remain areas of active investigation. While TDM provides a framework for tailoring drug doses based on individual pharmacokinetics, the variability in clinical outcomes suggests that its effectiveness might be highly patient- and context-dependent.

Therefore, the real question is not whether TDM matters—it clearly does for some drugs and patients—but rather to what extent it impacts outcomes across diverse clinical scenarios and whether it justifies the associated costs and resource allocation. Further studies are needed to determine its value in routine practice and to identify patient subgroups most likely to benefit from this approach.

[Correction added on 4 April 2024, after first online publication: References 9-18 have been added in this version.]