Cancer poses a serious threat to human health, with high morbidity and mortality rates observed globally, and it also leads to significant economic and social problems (1). Drug therapy is one of the primary treatment modalities for cancer. Anticancer drugs, including chemotherapy agents, endocrine therapy drugs, molecular targeted drugs and immune checkpoint inhibitors (ICIs), exhibit complex pharmacokinetic properties and significant inter-individual variability (2–4). These factors lead to fluctuations in blood drug concentrations, making precise control essential to optimize efficacy and minimize safety risks (5–7).

Therapeutic drug monitoring (TDM) is a discipline dedicated to individualized dosing. It aims to optimize clinical outcomes by measuring blood drug concentrations and adjusting doses accordingly. Over the past years, TDM has gained increasing attention in the management of drugs that exhibit low therapeutic indices, narrow therapeutic windows, or a high risk of severe adverse reactions. It is also commonly used for drugs requiring long-term administration and those showing substantial variability between individuals. The research types encompass both clinical and basic studies. Currently, TDM is relatively more widely utilized in areas such as immunosuppressants (8, 9), antimicrobial agents (10), and antiepileptic drugs (11, 12).

In contrast, TDM in oncology was introduced at a later stage and remains limited in scope and clinical adoption (13, 14). Although used for certain agents since the 1970s, it is not universally applied, and therapeutic targets are not well established for most drugs (6).

Due to the typically prolonged courses of anticancer treatment, long-term medication may further exacerbate variations in drug exposure levels. Although there is a clear necessity for TDM of anticancer drugs, several challenges remain in practical implementation.

Common challenges apply broadly regardless of drug type. For instance, obtaining the area under the curve (AUC) for a single drug often requires multiple blood samples, which is highly inconvenient for patients (2). Meanwhile economic considerations also need to be taken into account. Additionally, some anticancer drugs are prodrugs, whose active metabolites are difficult to detect in the bloodstream (15). Furthermore, due to the unique blood supply of solid tumors, measuring the concentrations of the parent drug or its active metabolites in tumor tissues is extremely challenging (16–18).

Drug class-specific further complicate the implementation of TDM. In chemotherapy, regimens are typically based on combinations of drugs, making pharmacokinetic (PK)-pharmacodynamic (PD) relationships more difficult to model and target AUC values harder to define for individual agents (2). Tyrosine kinase inhibitors and anti-hormonal drugs typically require long-term daily administration, yet fixed dosing remains widely used. This is problematic because many of these agents exhibit considerable inter-individual variability in PK exposure and have a narrow therapeutic range. Moreover, well-defined exposure-response and exposure-toxicity relationships are lacking for some drugs at the approved dose, and suitable bioanalytical methods for monitoring are not always available (19). For ICIs, their PK and PD properties are fundamentally different from those of small-molecule drugs, which makes the development of TDM equally challenging (4). Nevertheless, researchers have conducted extensive exploration in the field of TDM for anticancer drugs. However, comprehensive analyses of the research hotspots and trends in this area are rarely reported.

Bibliometrics is a quantitative analysis tool that reveals hot topics and trends in scientific research (20). By systematically analyzing large volumes of literature, it identifies high-impact keywords, core authors, international collaboration networks, key research institutions and technologies. This study employs bibliometric methods to explore the advances in TDM for anticancer drugs globally since 1990, analyze research hotspots and trends, and provide references for researchers.

Since its introduction into clinical practice in the early 1960s, TDM has seen a rapid increase in research activity, primarily in non-oncological therapeutic areas. However, its application in oncology remained very limited until the 1980s (13, 14). Routine clinical use was confined to methotrexate, which was then predominantly employed in the treatment of diseases such as acute lymphoblastic leukemia (ALL) (14). For instance, Evans et al. (28). established the relationship between systemic methotrexate clearance and the probability of relapse in children with ALL. Additionally, approximately one-third of the studies focused on bioavailability assessment, primarily involving mercaptopurine and melphalan (29, 30). For other agents such as cytarabine (Ara-C), doxorubicin, and cyclophosphamide, research was still in the stage of pharmacokinetic investigation and preclinical or early clinical data accumulation (31–33).

Since then, the emergence of new technologies and therapeutics, along with a deeper understanding of cancer and anticancer drugs, has gradually expanded the role of TDM in oncology. This study used bibliometric methods to map and describe the research hotspots and trends in the field of TDM for anticancer drugs globally between 1990 and 2024. Over the past 35 years, the number of publications worldwide in this field has steadily risen. Despite rapid progress in oncology, the relative focus on TDM for anticancer drugs has steadily increased over time, reflecting growing interest in optimizing therapeutic outcomes through individualized monitoring.

Keyword analysis employed co-occurrence, clustering, and burst detection. The most frequent terms were “therapeutic drug monitoring” and “pharmacokinetics.” TDM is crucial for optimizing outcomes and minimizing toxicity in anticancer therapy due to interpatient variability in drug metabolism. Although pharmacokinetic studies are common in early drug development, their clinical relevance remains high, especially for personalized dosing. The expanding use of targeted therapies and immunotherapies further underscores the importance of PK research.

The research in this field has largely revolved around three main categories related to TDM: drugs, quantitative detection methods, and clinical efficacy and safety. In terms of drugs, the studies mainly cover four classes of anticancer therapies, namely: chemotherapeutic agents, endocrine therapy drugs, targeted therapy drugs and ICIs. Chemotherapeutic agents often exhibit high inter-individual PK variability, a well-established exposure-response relationship such as the correlation between area under the curve and pharmacodynamic endpoints, and sufficient time delay between sampling and clinical effect; these features collectively represent key criteria supporting TDM implementation (2). Busulfan remains the most studied agent in this class, particularly in hematopoietic stem cell transplantation settings (34–37), while methotrexate and 5-fluorouracil also demonstrate clear associations among drug exposure, efficacy, and toxicity (38–41). For TKIs, TDM is increasingly relevant because these drugs are metabolized by cytochrome P450 enzymes, are prone to drug–drug interactions, and exhibit plasma concentration variability driven by genetic polymorphisms that significantly affect clinical outcomes (42); imatinib, crizotinib, and osimertinib exemplify TKIs for which emerging evidence links specific PK parameters to improvements in survival outcomes (43–49). Monoclonal antibodies (mAbs) display more complex pharmacokinetics determined by target-mediated disposition, FcRn-mediated recycling, and inter-patient differences in target expression, yet trastuzumab and bevacizumab continue to appear frequently in TDM-related research (50, 51). TDM for endocrine agents such as tamoxifen is complicated by CYP2D6-dependent metabolic activation and inter-individual variation in metabolizer status (52–54). Similarly, for ICIs, inconsistent exposure-efficacy relationships, the absence of established therapeutic concentration ranges, and confounding effects of target binding currently limit the clinical utility of TDM (4, 51, 55).

Although preliminary pharmacology and pharmacokinetic studies on antibody-drug conjugates (ADCs) have emerged during our study period, and although ADCs are often termed “magic bullets” due to their precise tumor targeting and typically narrow therapeutic window, there remains no consensus on the optimal analyte(s) for clinically meaningful TDM (56). Accumulating evidence indicates that different analytes may differentially inform efficacy versus toxicity. The intact (conjugated) ADC, as the pharmacologically active species, shows stronger association with antitumor efficacy. In trastuzumab emtansine (T-DM1), higher exposure—particularly the trough concentration at cycle 1 day 21 (Cmin, C1D21)—was significantly correlated with longer progression-free survival (PFS) and overall survival (OS) in patients with HER2-positive metastatic breast cancer (57). Similarly, in trastuzumab deruxtecan (T-DXd), greater systemic exposure to the intact ADC was linked to higher objective response rates in early-phase trials (58). Conversely, released payload (e.g., DM1, DXd, MMAE) appears more predictive of toxicity. Elevated plasma concentrations of free MMAE were associated with increased risk of grade ≥2 peripheral neuropathy and neutropenia in patients receiving brentuximab vedotin (59). Likewise, higher systemic exposure to released deruxtecan (DXd) has been implicated in interstitial lung disease, a dose-limiting toxicity of T-DXd (58). In contrast, total antibody has not been clearly demonstrated to be consistently associated with efficacy or safety outcomes (60).

Given the increasing clinical use of bispecific antibodies (bsAbs), no TDM-related studies on bsAbs were identified in our bibliometric analysis, underscoring their limited representation in the current literature under our search criteria. Their unique mechanisms, such as dual-target engagement, target-mediated drug disposition, and high immunogenicity, pose distinct TDM challenges that differ from those associated with conventional monoclonal antibodies and ADCs (61–63). This underscores the need for future research to establish evidence-based TDM frameworks tailored to bsAbs.

In addition to drugs, we also identified research hotspots related to analytical methods, diseases and special populations. Detection methods such as liquid chromatography, mass spectrometry, LC-MS, and tandem mass spectrometry have been relatively maturely applied, while emerging technologies like electrochemical sensors (64) and biosensors (65) are beginning to emerge. A highly cited paper utilized paper spray mass spectrometry for the quantitative analysis of therapeutic drugs in dried blood spot samples, with the method capable of detecting sunitinib at extremely low concentrations (1 ng/mL) (66).Regarding diseases, the burst detection of disease-related keywords was largely associated with the use of specific drugs, reflecting that research in TDM is often driven by drug-specific therapeutic contexts. For example, bursts in “chronic myeloid leukemia” coincided with research on busulfan and imatinib, while “metastatic colorectal cancer” emerged alongside studies on 5-FU. Meanwhile, children/pediatric patients have consistently been a hotspot in TDM research for anticancer drugs, with the keyword remaining in a state of emergence over the long term. Since 2022, it has entered a strong emergence phase, reflecting a recent surge in studies focusing on pediatric populations. This growing attention highlights important opportunities for optimizing drug dosing in children, though challenges remain—such as limited PK data, ethical constraints in sampling, and variability in drug metabolism due to developmental changes.

The publishing and co-citation patterns of journals reflect the interdisciplinary nature of TDM research. Productive journals are primarily specialized in pharmaceutical and analytical sciences, whereas highly co-cited journals are rooted in clinical oncology and pharmacology. This divergence aligns with the dual-map overlay results and underscores that while TDM methodology is developed in technical disciplines, its impact is realized through high-impact clinical research.

From the publication volume by country and the international collaboration map, it can be observed that most of the countries with high publication volumes are major economies (67), a trend similar to that seen in the field of antimicrobial TDM research (10). Notably, despite the Netherlands not ranking among the top ten in overall GDP or per capita GDP (based on 2024 data from the International Monetary Fund) (67), it not only stands out in the research output and citation count within the field of TDM for anticancer drugs, but also has the strongest research intensity, indicating the country’s strong commitment and capability in this area. Indeed, according to a 2024 report by the Commonwealth Fund, an independent healthcare research organization, the Netherlands ranked second among high-income countries for healthcare system performance and topped the list for the best accessibility and availability of health services (68). While the goal of TDM is to achieve precision medicine, which aims for better clinical outcomes, it also signifies greater investment in health economics from both individual and national perspectives (69). Several European countries play significant roles in this field, with close inter-country collaborations and high RII. The United States has the highest publication and citation count for all its publications, which correlates positively with its high health spending (70) and reflects the global impact of their research. However, despite its high overall publication output, the country’s RII in research on anticancer drug TDM is not outstanding. Among the top ten institutions by publication volume, all are from France and the Netherlands, with each institution publishing at least 41 articles on average, 67.2 papers, and maintaining strong connections with other global institutions. Meanwhile, among the institutions or study groups, UNICANCER (the French national hospital network exclusively specialized in oncology) were found to have significantly higher research intensity than others, reflecting the strong focus of this French collaborative network. Notably, Slotervaart Hospital in the Netherlands was previously active in this field. However, the hospital ceased operations in 2018 due to financial difficulties. As a result, several leading researchers including Lankheet NAG and Huitema ADR moved to other institutions. Consequently, Slotervaart Hospital has had no recent research output in TDM.

Through bibliometric methods, we have identified and visualized several prominent keywords and research hotspots. However, some less explored areas can also be integrated with existing findings to collectively outline the future of TDM for anticancer drugs (Figure 12). Despite the current lack of widely accepted effective TDM strategies for pediatric cancer patients, further exploration in this field holds promise for significantly improving clinical outcomes and providing insights for personalized treatment in critically ill patients and the elderly (2). Additionally, deeper investigations into tumor heterogeneity are essential for better understanding drug mechanisms and individualized treatment needs (71, 72). On the pharmacological front, while some studies have examined the PK and PD of mAbs, many concentration-effect relationships have not been fully elucidated—particularly for ICIs, ADCs and bsAbs, whose multi-component exposure profiles, relevant biomarkers, and their links to efficacy and toxicity require systematic investigation (51, 56, 61, 73). In terms of analytical techniques, there is an urgent need to develop more precise, rapid, convenient, and cost-effective methods, such as biosensor-based platforms, alongside exploring alternative sample matrices like saliva, interstitial fluid, or dried blood spots to simplify collection processes (64, 74, 75). Future developments could include real-time visualization of test results via mobile applications, enhancing patient engagement and adherence, thus facilitating home-based TDM. Advanced modeling and simulation techniques should be introduced, leveraging high-quality clinical samples and incorporating machine learning algorithms to improve the predictive accuracy and personalization of TDM (76, 77). Furthermore, given the slowdown in global research progress over the past five years and insufficient investment by most countries, it is encouraged to strengthen international and institutional collaborations to share centralized laboratory resources and accelerate technological advancements. Simultaneously, fostering a supportive TDM medical culture is necessary, encouraging clinicians to adjust dosages based on monitoring results and urging pharmaceutical companies to address the complexities of personalized medicine (44). More high-quality clinical trials and prospective studies, especially Phase III trials, are critical for validating the effectiveness of new methods and technologies. Finally, health economic evaluations should be intensified, with national and governmental policy support needed to integrate TDM into healthcare systems, thereby enhancing its accessibility and prevalence (78). Achieving these goals requires the concerted efforts of clinicians, research institutions, pharmaceutical companies, regulatory bodies, and patients.

This study has several limitations. First, despite a relatively long search period, there is still a possibility of omissions, particularly as the analysis relied solely on the Web of Science database, which may introduce selection bias and limit comprehensiveness. Second, due to visualization requirements and threshold settings in bibliometric tools, not all keywords, institutions, or countries were fully represented, potentially leading to underrepresentation of weaker associations. Third, similar or misspelled terms—such as “tyrosine kinase inhibitor” and “tyrosine kinase inhibitors,” or “leukemia” and “leukeamia”—may have affected data accuracy. We attempted to merge these variants as much as possible, but some inconsistencies may remain due to the large volume of text. Additionally, bibliometric analysis cannot assess the qualitative depth of research or reflect actual clinical outcomes, so findings should be interpreted in conjunction with critical appraisal of study quality. Nevertheless, these limitations do not affect the overall conclusions of the article.

This study utilized bibliometric and visualization analysis methods to explore research trends and hotspots in TDM for anticancer drugs from 1990 to 2024, based on the WOSCC database. Over these 35 years, the number of papers on TDM for anticancer drugs has shown steady growth. Research primarily focused on drug exposure in special populations, hematologic diseases, TKIs, quantitative detection methods, chemotherapeutic drugs, and population pharmacokinetics. In the past three years, kinase inhibitors and pediatric patients have emerged as new research hotspots within the field. The thematic trend indicates a shift towards convergence from Molecular/Biology/Immunology and Health/Nursing/Medicine towards Medicine/Medical/Clinical. At the national level, the Netherlands and the United States have had significant impacts on this field, while at the institutional level, institutions from the Netherlands and France have contributed the most research outputs. Through this study, we hope to assist researchers in identifying new directions, hotspots, and frontiers in related areas.