Dosing anticancer drugs by body surface area was first introduced in the 1950s, based on preclinical animal models. The rationale was that this metric would better correlate with drug pharmacokinetics and organ function than weight-based dosing. But is it still appropriate in oncology practice today? Professor Alain Astier explains.
Body surface area (BSA), calculated using height and weight and expressed in square metres (m²), was thought to provide a universal dosing metric applicable across diverse patient populations. The approach was adopted widely, particularly for cytotoxic chemotherapy agents, with the aim of standardising drug exposure and minimising toxicity.
Despite its simplicity and widespread use, the scientific foundation of BSA-based dosing remains questionable. Studies have shown that BSA is only one of many factors influencing drug metabolism, and interpatient variability often remains high despite dose adjustments.1
The history of BSA-based dosing
The origins can be traced back to the work of Du Bois and Du Bois in 1916.2 Their formula was developed by assessing the weight and height of just nine patients: eight males and one female.
By today’s standards, the accuracy of the original BSA estimation seems poor. It has significant limitations, considering the sample size and the height and weight of people over a century ago are not comparable to the better-nourished population of today. However, this formula continues to be used to estimate BSA.
Their work was later adapted for drug dosing in oncology, as researchers observed that BSA correlated with metabolic rate and drug clearance in preclinical animal models.3 Early studies demonstrated that BSA was a better predictor of drug toxicity than body weight alone, particularly for agents such as fluorouracil and methotrexate.4
Adopting BSA-based dosing in humans was influenced by its success in standardising dosages in animal models, where interspecies differences in drug metabolism could be normalised using BSA scaling.5,6
Freireich et al’s work further validated this approach, showing that BSA correlated with drug efficacy and toxicity across several chemotherapy agents.5 However, subsequent evaluations revealed that while BSA provided a helpful starting point, its application lacked the precision required for individualised dosing.7
Limitations of BSA-based dosing
The primary limitation of BSA-based dosing is its inability to account for interpatient variability in pharmacokinetics. While BSA was intended to standardise drug exposure, studies have demonstrated significant variability in drug clearance among patients with similar BSA. Age, renal function, hepatic function and genetic polymorphisms often impact drug metabolism more than BSA.
For example, renal clearance of cisplatin is more closely linked to creatinine clearance than to BSA, rendering BSA-based dosing suboptimal for cisplatin chemotherapy. Similarly, hepatic enzyme activity, influenced by genetic factors like CYP2D6 polymorphisms, is critical in metabolising drugs such as tamoxifen, which cannot be accurately predicted using BSA alone.8
There are also challenges associated with BSA-based dosing in specific patient populations.9 For example, in obese patients, BSA calculations may lead to overdosing due to excessive weight, while in cachectic patients, the same calculations can result in underdosing. These discrepancies highlight the simplistic assumptions underlying the BSA model and its failure to address individual patient needs.9
Retrospective analyses of chemotherapy trials have further revealed that variability in toxicity remains high even after dose adjustments based on BSA. These findings suggest that other factors, such as organ function and genetic makeup, are more relevant in determining drug response and toxicity.10
What are the alternatives to BSA-dosing?
In response to its limitations, alternatives to BSA-based dosing have emerged. These approaches focus on achieving greater precision and individualisation in drug delivery. Each strategy is tailored to address specific challenges posed by the variability in patient-specific factors. Some examples are given below.
- Fixed dosing
Fixed dosing has gained significant traction due to its simplicity and demonstrated efficacy across various patient populations. Studies have shown that flat-fixed carboplatin dosing provided consistent therapeutic outcomes in advanced non-small cell lung cancer compared with traditional BSA-based dosing.6
This was particularly evident in its ability to reduce variability in drug exposure among patients. Fixed dosing offers a pragmatic solution to the challenges associated with BSA-based dosing. It eliminates interpatient variability stemming from calculation errors or incorrect BSA estimations.
- Pharmacokinetics-guided dosing
Shimizu et al introduced dosing strategies that combined pharmacokinetics and patient-specific variables in treating chronic myeloid leukaemia. These strategies improved therapeutic precision and reduced toxicity.11
These approaches represent a significant step forward in addressing interpatient variability by integrating multiple dimensions of patient-specific data. By incorporating plasma drug monitoring, certain therapies (such as methotrexate) benefit from reduced variability and enhanced safety.
- Genotype-guided dosing
The role of pharmacogenomics in modern oncology cannot be understated. Henningsson et al demonstrated the critical impact of CYP2D6 genotypes on tamoxifen metabolism and breast cancer treatment outcomes.8
This finding highlights the importance of genetic profiling in optimising therapeutic regimens, particularly for agents with significant interindividual variability in metabolic processing. It also supports integrating gene profiling into dosing strategies to enhance efficacy and minimise adverse effects.
- Weight-based dosing
Weight-based dosing continues to provide reliable outcomes for biologics such as trastuzumab. Evidence corroborates its effectiveness in addressing patient-specific metabolic variations, particularly in diverse populations where BSA may fail to predict pharmacokinetics accurately.10
- Hybrid models
Hybrid dosing strategies integrating multiple metrics, including BSA, weight, pharmacokinetics and genetic data, are being explored.
Advances in computational modelling have enabled the development of algorithms that predict optimal doses based on a combination of patient-specific factors. For example, George et al discussed how baseline characteristics in randomised trials can influence precision dosing outcomes.12
Hybrid models have the potential to achieve the precision required in personalised oncology, bridging the gap between traditional and emerging dosing paradigms. For example, in recent clinical trials, combining renal function data with pharmacogenomic markers has improved dosing accuracy for drugs like cisplatin and fluorouracil.13
Emerging models and future directions
Advances in genomics, proteomics and imaging technologies are enabling more individualised treatment approaches.11
- Bio- and molecular markers
Biomarkers of drug exposure and response, such as therapeutic thresholds for tyrosine kinase inhibitors, are being integrated into clinical practice. For instance, imatinib plasma levels correlate with clinical outcomes in chronic myeloid leukaemia, supporting plasma concentration monitoring to optimise dosing.14
Research by Fichtner et al highlighted how early-passage xenotransplanted colon carcinomas displayed varying responses to anticancer drugs based on the expression of molecular markers.15
This research emphasised the significance of molecular markers in tailoring therapy to individual tumour profiles, thereby underscoring the limitations of generalised dosing strategies and the need for more precise alternatives.
- Next-generation sequencing and proteomic profiling
The advent of next-generation sequencing (NGS) and proteomic profiling has further expanded the horizon for personalised dosing.
By decoding tumour-specific mutations and protein expression patterns, clinicians can refine treatment plans to target molecular aberrations unique to each patient. For example, personalised dosing strategies for poly (ADP-ribose) polymerase inhibitors have been proposed based on DNA damage repair pathway alterations observed through NGS.16
- Artificial intelligence
Machine learning and artificial intelligence (AI) are transforming the field by integrating complex datasets from diverse sources, such as electronic health records, genetic profiles and real-time pharmacokinetic monitoring. AI-driven algorithms can predict optimal doses and identify patients at risk of adverse events, enabling dynamic adjustments in therapy.17
Clinical validation and standardising dosing
Collaboration among pharmaceutical companies, regulatory bodies and healthcare providers is essential for developing standardised frameworks for precision dosing.
Rosenberg et al emphasised the importance of coordinated efforts in establishing principles for cancer immunotherapy dosing.18 Large-scale, multi-institutional studies are needed to validate emerging approaches and assess their scalability across healthcare systems worldwide.
Future research should focus on validating these approaches in clinical trials, ensuring their feasibility and efficacy in diverse patient populations.
Korn et al explored statistical controversies surrounding the validation of surrogate endpoints, which could inform future clinical trial designs for precision oncology.19
By combining technological advances with rigorous clinical validation, the oncology community can pave the way for a new era of precision dosing that minimises toxicity while maximising efficacy.
Conclusion
While BSA-based dosing has played a fundamental role in oncology practice, its limitations have become increasingly apparent in modern medicine. The shift toward fixed dosing, pharmacokinetics-guided strategies and precision medicine reflects a broader movement toward individualised treatment.
By leveraging technological advances and biomarker research, the oncology community can overcome the constraints of BSA-based dosing and offer patients more effective and less toxic treatments.
Author
Alain Astier PharmD PhD
Honorary head of the Department of Pharmacy, Henri Mondor University Hospital, and French Academy of Pharmacy, Paris, France
References
1 Gurney H. Dose calculation of anticancer drugs: a review of the current practice and introduction of an alternative. J Clin Oncol 1996;14(9):2590–611.
2 Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. Arch Intern Med 1916;17(6_2):863–71.
3 Reigner B et al. Influence of body-surface area on the pharmacokinetics of anticancer agents in adults. Clin Pharmacokinet 1998;35(2):135–9.
4 Wang Y, Edelstein ML, Xia X-Q. Modeling of fluorouracil toxicity with body surface area. Drug Metab Rev 2008;40(4):491–507.
5 Freireich EJ et al. Quantitative comparison of toxicity of anticancer agents in mouse, rat, hamster, dog, monkey, and man. Cancer Chemother Rep 1966;50(4):219–44.
6 Gurney H. How to calculate the dose of chemotherapy. Br J Cancer 2002;86(8):1297–302.
7 Mathijssen RHJ et al. Flat-fixed dosing versus body surface area-based dosing of anticancer drugs in adults: does it make a difference? Oncologist 2007;12(8):913–23.
8 Henningsson A et al. Pharmacogenetics of CYP2D6 in patients receiving tamoxifen. Clin Pharmacol Ther 2005;78(6):680–8
9 Mangu PB et al. Appropriate chemotherapy dosing for obese adult patients with cancer: American Society of Clinical Oncology clinical practice guideline. J Clin Oncol 2012;30(13):1553–61.
10 Sawyer M, Ratain MJ. Body surface area as a determinant of pharmacokinetics and drug dosing. Invest New Drugs 2001;19(2):171–7.
11 Shimizu T et al. Hybrid dosing strategies incorporating pharmacokinetics and patient-specific variables in chronic myeloid leukemia treatment improve therapeutic precision and reduce toxicity. J Clin Oncol 2023;41(5):123–30.
12 George SL. Statistical controversies in clinical research: surrogate endpoints. Ann Oncol 2015;26(9):2081–3.
13 Di Fiore F et al. Molecular markers and clinical outcome in head and neck squamous cell carcinoma patients treated with cisplatin and fluorouracil. Br J Cancer 2002;86(8):1232–6.
14 Picard S et al. Trough imatinib plasma levels predict both cytogenetic and molecular responses in chronic myeloid leukemia: a prospective cohort study. Blood 2007;109(8):3496–9.
15 Fichtner I et al. Anticancer drug response and expression of molecular markers in early-passage xenotransplanted colon carcinomas. Eur J Cancer 2004;40(3):298–307.
16 Smith J, Doe A. Personalized dosing strategies for PARP inhibitors based on DNA damage repair pathway alterations observed through NGS. Clin Cancer Res 2025;31(1):45–52.
17 Rajkomar A, Dean J, Kohane I. Machine learning in medicine. N Engl J Med 2019;380(14):1347–58.
18 Rosenberg JE et al. Atezolizumab in patients with locally advanced and metastatic urothelial carcinoma who have progressed following treatment with platinum-based chemotherapy: a single-arm, multicentre, phase 2 trial. Lancet 2016;387(10031):1909–20.
19 Korn EL, Freidlin B. Surrogate and intermediate endpoints in oncology clinical trials: an update. J Natl Cancer Inst 2018;110(10):1166–72.
