What is pharmacometrics?
Pharmacometrics, or quantitative pharmacology, uses mathematical models based on biology, pharmacology, physiology, and disease to quantify the interactions between drugs and patients. These models characterize the relationship between the dose of an administered drug, its exposure (pharmacokinetics) and its clinical efficacy and safety (pharmacodynamics) in individual patients and in whole populations. It aims to quantify the effect on this relationship of covariates such as age, sex, organ (dys)function, or concomitantly administered drugs. This methodology has been used successfully to identify appropriate doses of a medicine for certain patient populations, by describing variability between individuals that may be associated with lack of efficacy in certain patients or with toxicity in others. This model-based approach initially gained utility in populations where collection of large amounts of data in each patient was difficult or impossible - for example children, very old or severely sick patients - and its application is now used more widely.
Read more about our research areas below.
Tuberculosis
Tuberculosis (TB) is an infectious disease caused by the bacillus Mycobacterium tuberculosis and it is the leading cause of death from a single infectious agent (ranking above HIV), worldwide. In 2018, an estimated 10 million people fell ill with TB and 1.4 million lives were lost. Standard TB treatment requires a six months combination therapy with up to four drugs (rifampicin, isoniazid, pyrazinamide and ethambutol). Infections with resistant TB require treatment with additional drugs. We focus on the pharmacokinetics, efficacy and safety of these drugs using pharmacometrics model-based analysis. With our work we aim to improve treatment guidelines and support decision making for regulators, industry and physicians treating individual patients.
TB disease biomarkers are generally known to be noisy, and nonlinear mixed effect modelling, which can handle repeated sampling and jointly model more than one outcome, can improve the understanding of the PK/PD relationships. We have developed a semi-mechanistic model describing the progression of days to positivity of MGIT culture using a repeated time to event model approach. We have combined classical time-to-event modelling techniques, and time-to-event trees to identify the factors and their interactions that contribute to time-to-stable culture conversion. We also perform simulations to determine optimal doses of tuberculosis drugs associated with a probability of PD target attainment of >90%.
TB patients are often co-infected with other infectious diseases such as the human immunodeficiency virus (HIV) and malaria. To treat these infections co-administration with TB treatment is necessary which can cause unwanted interactions and as result a decreased efficacy and more side effects. Therefore, we also specially focus on drug-drug interactions.
HIV
The human immunodeficiency virus (HIV) epidemic remains one of the greatest health challenges of our generation. Sub Saharan Africa bears the highest burden of the disease, with approximately 20.7 million people living with HIV as of 2019. Although the availability of testing services and access to antiretroviral treatment has increased significantly over the years, the number of new HIV infections remains high. For example, in 2018, there were approximately 800,000 new HIV infections.
WHO recommends antiretroviral therapy (ART) for all people with HIV as soon as possible after diagnosis without any restrictions of CD4 counts. Standard ART consists of a combination of antiretroviral drugs to maximally suppress the HIV virus, stop disease progression and prevent onward transmission. Currently, dolutegravir-based regimens are recommended as the preferred first and second line regimens for people living with HIV initiating ART.
People living with HIV often present with other co-morbidities such as tuberculosis. Our research focuses on understanding drug-drug interactions between antiretroviral therapy and drugs often used for these co-morbidities with the aim of informing dosing decisions that best optimize favorable treatment outcomes.
Malaria
A combination of patient characteristics, malaria disease severity, and antimalarial drug exposure and activity drives treatment efficacy. We apply nonlinear mixed-effects modelling to determine drug exposure by estimating pharmacokinetic parameters at population and individual level simultaneously and quantify variability due to differences between individuals. The overall aim is to optimise antimalaria drug dose to levels that maximise treatment efficacy and minimise toxicity, factoring in the effects of known characteristics that contribute to drug exposure variability. Our work is mainly supported through collaboration with the WorldWide Antimalarial Resistance Network.
We have successfully developed population pharmacokinetic models for lumefantrine, amodiaquine and desethylamodiaquine, and a combination of sulfadoxine and pyrimethamine using individual patient data from multiple studies. Our model detected significant drug-drug interactions between lumefantrine and concomitant medication: lopinavir-ritonavir-based antiretroviral therapy (ART), efavirenz-based ART, and rifampicin-based antituberculosis treatment. Lower and higher exposures to salfadoxine and pyrimethamine, respectively, were detected during pregnancy compared to postpartum using a model-based analysis. For children, we constructed a model-based dosing regimen that would achieve salfadoxine and pyrimethamine exposures within the range observed in adult patients. Once-daily 10 mg/kg doses of amodiaquine result in reduced exposure amongst children with low weight and patients weighing more than 62 kg; we proposed daily doses of amodiaquine in the range of 9.8 to 19.9 mg/kg.
Drug-drug interactions
One of the research focusses of the pharmacometrics group at the University of Cape Town is to characterise and overcome drug-drug interaction (DDI) through the use of mathematical and statistical (nonlinear mixed-effects) models. DDI occurs when the drug's pharmacological activity is altered by the presence of another drug taken concomitantly.
DDIs are one the most common cause of medical error and are known to induce the development of adverse drug reactions or reduce clinical efficacy of a drug. DDI are also costly for the health system as they are associated with lengthening hospital stays and hospital costs. DDI can be classified as pharmacokinetics DDI; altering the exposure of the drug, or as pharmacodynamics DDI; which either increase, reduce or cancel out the drug effect. Even though the co-administration of multiple drugs often increases therapeutic effectiveness, certain combinations might be harmful. Therefore, research focused on DDI is vital to ensure the safety and efficacy of drugs when concomitantly administered
Pharmacogenomics
The pharmacometrics group at the University of Cape Town characterizes pharmacogenomic determinants of drug response using nonlinear mixed-effects models. Pharmacogenomics studies the role of genes on interindividual variation to drug disposition and response. It is a tool for individualized therapy through optimization of drug selection and dosing. This is critical in Africa where there is a great burden of infectious diseases, such as malaria, TB, and HIV/AIDS, along with a challenge of adverse drug reactions and lack of efficacy associated with their treatment. The situation is further complicated as the African population is characterized by a great level of genetic and phenotypic variation compared to the non-African population. Much of the variation in drug response is due to genetic polymorphisms resulting in differences in the activities of drug-metabolizing enzymes and transporters.
We have developed population pharmacokinetic models for isoniazid and efavirenz (in special populations such as pregnant women and children), and the exposures of these drugs are heavily influenced by genetic polymorphisms. The main determinant of Isoniazid exposure, i.e. N-acetyltransferase 2 (NAT2) status, has been reported to have the highest level of within-population diversity in Africans. And the main determinant of efavirenz exposure CYP2B6 genotype is reported to have a high proportion of slow metabolizers in the African population, increasing the risk of efavirenz-related toxicities in this population. Our population pharmacokinetic models for these two drugs in pregnant women and children have highlighted the importance of genetic testing and the need for individualized treatment to ensure similar exposures in a diverse population.
Special populations - paediatrics
In infectious disease pharmacotherapy, children are assumed to have similar disease progression and response to drugs as adults, therefore, paediatric dosing of anti-infectives aims to achieve same exposures as adults. However, optimizing doses in paediatric population remains a challenge, mostly due to changes in physiological development, including physical growth and the maturation of organs, transporters and enzymes. These changes contribute to large pharmacokinetic (PK) and pharmacodynamic (PD) variability in children. Therefore, the focus of pediatric dose-optimization is primarily to account for the effects of body size and maturation, thus achieving a similar pharmacokinetic profile to that in adults. Because of these physiological changes within the different sub-groups in the paediatric population the doses are calculated using weight-based dosing.
To help with dose optimization in the paediatric population, the Pharmacometrics group at the University of Cape Town in collaboration with the WHO Paediatric Anti-retroviral Working Group have recently created a generic paediatric dosing tool based on Microsoft Excel to ease the calculation of expected AUCs in children relative to adult targets adjusting for allometric scaling and maturation. Further details can be found in the link provided https://www.rosaandco.com/webinars/2020/what-drives-policy-change. For in-depth analysis the Pharmacometrics group utilizes mathematical and statistical (nonlinear mixed-effects) models to quantify and explore relationships of factors contributing to paediatric PKPD variability to support paediatric dosing recommendations.
Special populations - pregnancy
In pregnancy, maternal physiology changes to accommodate development and growth of the placenta and foetus, these physiological variations may alter the pharmacokinetics or pharmacodynamics of a drug. Most changes are observed during the first trimester and are most pronounced in the third trimester. These physiological variations are not fixed throughout pregnancy, but progressively change during pregnancy. Enzymatic activity exhibited by the placenta and foetus may also affect maternal pharmacokinetics. Dosing strategies in pregnant women are often based on data from healthy male volunteers and non-pregnant women, with little adjustment for the complex physiology of pregnancy. This is because most drug studies exclude pregnant women based on concerns regarding foetal risk. Therefore, understanding pregnancy physiology is crucial to achieve effective treatment and limit maternal and foetal risk.