Africa is still too often brought into drug and vaccine development late, most commonly during Phase III, post-marketing, or implementation studies. Scientifically, that is a weak model. Phase I and II studies are where development programmes establish core assumptions about dose, exposure, early safety, pharmacodynamic activity, and, for vaccines, immunogenicity. If those assumptions are built without sufficiently diverse populations, developers may carry avoidable uncertainty into later-stage trials. The case for earlier African inclusion is therefore not merely rhetorical; it is scientific.
Africa remains underrepresented in global clinical research, with less than 3% of worldwide clinical trials conducted on the continent (1). This creates a structural gap where early development decisions are frequently made without incorporating the full spectrum of human biological diversity. This is not primarily an issue of equity; it is an issue of scientific validity. For developers, the practical consequence is that key early assumptions about dose, exposure-response, and immune performance may be built on datasets that are less globally predictive than they appear.
Phase I and II: Where Development Assumptions Are Set
Phase I and II studies are designed to reduce uncertainty before later-stage investment. They establish pharmacokinetic profiles, pharmacodynamic activity, dose-response relationships, early safety signals, and, in vaccine development, initial immune responses. If those parameters are derived from biologically narrow populations, developers may overestimate how well the resulting assumptions will generalize across settings. Earlier inclusion of African populations can therefore strengthen the external validity of foundational datasets rather than simply broaden participation. Africa contains the greatest amount of known human genetic diversity, making it especially important in efforts to build development datasets that better capture biological variation (2).
Importantly, early inclusion is not only useful because it captures more diversity. It is useful because it does so at the stage when protocol decisions remain flexible. By Phase III, many key assumptions about dose, schedule, biomarkers, and endpoints have already hardened. If biologically important variability is discovered only late, the cost of adjustment is much higher.
Pharmacokinetics and Pharmacogenomics: Exposure Cannot Be Assumed
Drug exposure varies across populations, and one important contributor is genetic variation in drug-metabolizing enzymes and transporters. African populations show substantial diversity in pharmacogenes including CYP2D6, CYP2B6, CYP2C8, and CYP3A5, all of which are relevant to clinically important pathways in drug metabolism (3,4). CYP2D6 alone is involved in the metabolism of roughly a quarter of commonly prescribed medicines (4). Where African populations are underrepresented in pharmacogenomic datasets, early estimates of exposure variability may be incomplete.
Real-world evidence reinforces this risk. Studies of efavirenz (an antiretroviral drug) pharmacokinetics in East African populations have demonstrated that both genetic variation and regional differences influence drug exposure and outcomes, indicating that findings are not always directly transferable even across different African populations (5). This is a useful reminder that African populations should not be treated as a single pharmacogenomic category. Early inclusion improves science only if it is designed with enough granularity to capture meaningful variation.
This matters particularly in programmes where exposure is sensitive to enzyme activity, concomitant medications, nutritional status, or coinfections. In such cases, early pharmacokinetic work in diverse populations can improve formulation strategy, inform dose selection, and reduce the risk of late-stage bridging exercises that could have been anticipated earlier.
Pharmacodynamics: Similar Exposure, Different Effects
Pharmacokinetic similarity does not guarantee pharmacodynamic similarity. Even where exposure appears comparable, downstream biological effects may differ because of variation in receptor biology, immune activation, inflammatory state, coinfections, and other host or environmental factors. For developers, this matters because exposure-response assumptions that seem stable in one population may be less reliable when moved across settings without early testing.
Warfarin is a familiar illustration of this problem. Variants identified in European and Asian populations do not fully explain dose variability in African populations, which led researchers to identify additional variants that are more relevant in African settings (6). The broader lesson is not limited to anticoagulation. Where pharmacodynamic variability is not explored early, developers may end up with dose-optimization strategies that are less precise than expected.
The same logic applies to biomarker strategy: if predictive markers are calibrated too narrowly in early development, later interpretation may be less robust across heterogeneous populations.
Vaccines and Immunology: Population Context Shapes Response
The scientific case for earlier inclusion may be even stronger in vaccine development. Phase I and II vaccine studies establish reactogenicity, antibody kinetics, cellular immune responses, and early correlates that shape subsequent programme decisions. Yet vaccine performance is not uniform across populations. A 2024 review in Nature Reviews Immunology highlights substantial geographic variation in vaccine immunogenicity and efficacy, linked to differences in prior pathogen exposure, microbiome composition, nutrition, and immune conditioning (7).
In many African settings, immune systems are shaped by repeated pathogen exposure, chronic parasitic infection, distinct microbiome environments, and environmental antigen exposure. These factors can influence baseline immune tone, vaccine reactogenicity, and the durability of immune responses (8). Reviews focused on African immunology suggest that these dynamics remain insufficiently characterized in early-phase vaccine studies (7,9). Malaria vaccine research offers a concrete example, where in one study it was suggested that genetic variation in immune-response pathways may contribute to differences in vaccine efficacy across populations (10). If developers want to refine correlates of protection, optimize booster strategies, and design next-generation vaccines with stronger real-world performance, those questions need to be examined early in populations where both exposure history and host biology may differ materially from the populations that usually dominate first-in-human and early expansion studies.
This is especially relevant when developers are relying on immunobridging, surrogate endpoints, or early immune signatures to guide programme progression. If immune baselines and response patterns vary by geography and exposure history, then early immunology datasets become more informative when they are built across settings rather than inferred across them.
From Scientific Risk to Development Opportunity
Earlier inclusion of African populations is not an added burden on development programmes; it is a way to improve them. It can strengthen dose selection, identify exposure-response variability earlier, improve biomarker interpretation, and produce datasets that are more globally informative. In practical terms, that can reduce avoidable uncertainty before pivotal trials and lower the risk of late-stage surprises or post-approval variability. WHO’s 2024 guidance on best practices for clinical trials also supports more representative trial populations and stronger country-led evidence generation, reinforcing the broader shift toward context-relevant development science (11).
The argument for earlier inclusion is also more practical than it once was. African trial ecosystems have expanded substantially, and recent literature highlights ongoing growth in trial capacity, regulatory strengthening, and preparedness infrastructure across the continent (1). That does not eliminate operational variation across countries and sites, but it does weaken the assumption that African participation is only realistic late in development.
Conclusion
If drug and vaccine development is fundamentally about understanding how interventions behave in humans, then excluding the world’s most genetically diverse populations from early-phase studies introduces avoidable uncertainty. Africa should not enter development only once dose assumptions, immune-response models, and exposure-response expectations have already been set elsewhere.
The scientific case for earlier inclusion is straightforward. Better early-phase representation can improve the external validity of pharmacokinetic and pharmacodynamic datasets, strengthen interpretation of immunology signals, and make development programmes more predictive across populations. For industry scientists, this is not about symbolic inclusion. It is about building stronger evidence earlier, when programmes are still most able to learn.
References
- NdembiN, Mekonen TT, Folayan MO, Dereje N, Kruger A, Fokam J, et al. Strengthening and expanding capacities in clinical trials: advancing pandemic prevention,preparedness and response in Africa. Nat Commun. 2024 Oct 7;15(1):8662. doi:10.1038/s41467-024-53126-3
- TishkoffSA, Williams SM. Genetic analysis of African populations: human evolution and complex disease. Nat RevGenet. 2002 Aug;3(8):611–21. doi:10.1038/nrg865
- Matimba A, Del-Favero J, Van Broeckhoven C,MasimirembwaC. Novel variants of major drug-metabolising enzyme genes in diverse African populations and their predicted functional effects. Hum Genomics. 2009 Jan;3(2):169–90. doi:10.1186/1479-7364-3-2-169 PubMed PMID: 19164093; PubMed Central PMCID: PMC3525272.
- TwesigomweD,Drögemöller BI, Wright GEB, Adebamowo C, Agongo G, Boua PR, et al. Characterization of CYP2D6 Pharmacogenetic Variation in Sub-Saharan African Populations. Clin Pharmacol Ther. 2023 Mar;113(3):643–59. doi:10.1002/cpt.2749 PubMed PMID: 36111505; PubMed Central PMCID: PMC9957841.
- MugusiS,Habtewold A, Ngaimisi E, Amogne W, Yimer G, Minzi O, et al. Impact of Population and Pharmacogenetics Variations on Efavirenz Pharmacokinetics and Immunologic Outcomes During Anti-Tuberculosis Co-Therapy: A Parallel Prospective Cohort Study in Two Sub-Sahara African Populations. Front Pharmacol. 2020 Feb 7;11. doi:10.3389/fphar.2020.00026
- Fung E,PatsopoulosNA, Belknap SM, O’Rourke DJ, Robb JF, Anderson JL, et al. Effect of Genetic Variants, Especially CYP2C9 and VKORC1, on the Pharmacology of Warfarin. Semin Thromb Hemost. 2012 Nov;38(8):893–904. doi:10.1055/s-0032-1328891 PubMed PMID: 23041981; PubMed Central PMCID: PMC4134937.
- van Dorst MMAR,PyuzaJJ, Nkurunungi G, Kullaya VI, Smits HH, Hogendoorn PCW, et al. Immunological factors linked to geographical variation in vaccine responses. Nat Rev Immunol. 2024 Apr;24(4):250–63. doi:10.1038/s41577-023-00941-2
- Cose S, Bagaya B, Nerima B,JolobaM, Kambugu A, Tweyongyere R, et al. Immunology in Africa. Trop Med Int Health. 2015 Dec;20(12):1771–7. doi:10.1111/tmi.12599 PubMed PMID: 26391634; PubMed Central PMCID: PMC4737115.
- WoldayD, Ndungu FM, Gómez-Pérez GP, de Wit TFR. Chronic Immune Activation and CD4+ T Cell Lymphopenia in Healthy African Individuals: Perspectives for SARS-CoV-2 Vaccine Efficacy. Front Immunol. 2021 Jun17;12:693269. doi:10.3389/fimmu.2021.693269 PubMed PMID: 34220854; PubMed Central PMCID: PMC8249933.
- Tukwasibwe S,Mboowa G, Sserwadda I, Nankabirwa JI, Arinaitwe E, Ssewanyana I, et al. Impact of high human genetic diversity in Africa on immunogenicity and efficacy of RTS,S/AS01 vaccine. Immunogenetics. 2023;75(3):207–14. doi:10.1007/s00251-023-01306-8 PubMed PMID: 37084013; PubMed Central PMCID: PMC10119520.
- New global guidance puts forward recommendations for more effective andequitable clinical trials [Internet]. [cited 2026 Mar 18]. Available from: https://www.who.int/news/item/25-09-2024-new-global-guidance-puts-forward-recommendations-for-more-effective-and-equitable-clinical-trials