A cancer patient receives treatment.
Cancer is a democratic affliction; it does not only target the wealthy, yet its effect on South Africans is deeply unequal. Most of us know someone close who has struggled with, or succumbed to, the disease. It remains a silent killer, often ignored because we fear the “C-word” or overlook the telling signs of unusual fatigue and weight loss. As we approach World Cancer Day 2026, we must confront a harsh reality: we cannot simply “borrow” solutions from the Global North if we hope to save lives at home.
The history of cancer research is long, stretching back to 1713 when physician Bernadino Ramazzini observed that nuns faced an extremely high risk of breast disease. While his notes were not strictly scientific, they pointed to the lifestyle choices underlying what he called the “Plight of Nuns”. Today, we have the molecular clarity he lacked. We understand that breast cancer risk is tied to the repetitive cycles of cell division during menstruation, where hormone fluctuations cause breast cells to proliferate and contract, potentially introducing genetic mutations.
Modern science has even uncovered the protective power of the immune system in this context. Studies have shown that specialised immune cells, called CD8+ T cells, remain in breast tissue for decades after childbirth to guard against cancer formation. Women who breastfeed have much higher concentrations of these “guardian” cells, providing a biological explanation for why these life stages reduce long-term risk.
The complexity of the “hallmarks”
Eliminating a tumour is never a simple matter because tumours are complex organisms defined by distinct biological “hallmarks”. These eight capabilities acquired as a normal cell evolves into a neoplastic state allow cancer to sustain growth, resist cell death, and even reprogramme its energy metabolism to evade the body’s immune system. These hallmarks provide the organising principle we need to rationalise the complexities of the disease.
Historically, cancer classification relied on how a tumour looked under a microscope, but this morphological approach has serious limitations. Two tumours that look identical can follow vastly different clinical courses and respond differently to therapy.
The revolution began in 1999 when Todd Golub used computational biology to prove that cancer could be classified based on gene expression. His methods resulted in the discovery of subtypes of leukaemia, such as acute myeloid leukaemia (AML) and acute lymphoblastic leukaemia (ALL), which require entirely different treatments. This ushered in the era of molecular subtyping, the foundation of personalised cancer care.
A South African solution
At our laboratories at the University of Cape Town (UCT), we are applying these same principles to our own population on a much smaller scale with a fraction of the budgets of large American labs. We have developed machine learning models that analyse the differential expression of specific genes to distinguish between various cancer types. This research is forming the basis of clinical trials to lay the foundation for molecular biomarker discoveries specifically for South African patients.
Research laboratories in South Africa are looking towards the future of “liquid biopsies”. While tumour tissue is the gold standard, it is now possible to detect non-coding RNA and circulating DNA in the bloodstream. These blood-based biomarkers offer a significant opportunity for non-invasive early detection and monitoring, providing a more reliable alternative to older, less sensitive markers like CA15-3. However, for these biomarkers to be effective, they have to be specific to the cancer subtype that will point towards the prognosis and programme of care to be followed. For this, we need studies of cancer in South African populations.
Why we cannot wait
The global landscape of cancer is changing. We are seeing an alarming rise in cancers, including colon and breast cancer, in people younger than 50, likely due to shifts in our collective lifestyle.
While we often lean on solutions discovered in American or European laboratories, the trend towards personalised medicine makes this “borrowing” increasingly impossible. Present-day cancer genomic subtyping and classification rely on large cellular/molecular data banks and studies of homogenous ethnic groups in the Global North. Personalised care solutions rely on vast and variable data drawn from studies of different populations, ethnicities, and lifestyles. We must discover solutions for our own people.
Currently, South Africa invests a mere 0.62% of GDP into research, with public health research receiving only a tiny fraction of that. Within this limited pool, cancer research must compete with the historical giants of HIV/AIDS and tuberculosis. The result is visible in the long lines outside our public oncology wards, where we catch cancer in our most vulnerable citizens far too late.
To conquer this silent killer, we must invest in our own science. We cannot depend on global solutions alone to solve a uniquely South African crisis.
Professor Kevin Naidoo is a professor of physical chemistry in the Department of Chemistry at the University of Cape Town, director of the Scientific Computing Research Unit, and holds the position of SARChI Research Chair in Scientific Computing.