How South Africa can stop HIV drug resistance in its tracks

If patients contract drug resistant HIV, they may have to be prescribed more expensive drugs. Finbarr O'Reilly/Reuters

Health care is entering a new era of personalised medicine, where treatment is tailored to the individual patient. To usher in this era, researchers across the world are trying to create cheaper tests that can find DNA mutations in humans, bacteria and viruses.

In South Africa, researchers are using these new, high-throughput testing methods to crack the problem of HIV drug resistance. Drug resistance testing is considered a critical part of an effective HIV treatment programme, but until now testing has been very costly.

Now, novel research in South Africa has come up with a solution that cuts the cost up to tenfold, raising the possibility of an HIV drug resistance test that could be accessible to almost anyone.

More than 30 years after its discovery, HIV continues to present an exceptional medical and public health challenge. Although anti-retroviral drugs (ARVs) are effective at disrupting the virus’s life cycle, the way HIV replicates is very unstable. Mutations creep into the virus’s DNA as it replicates, which can make the virus resistant to ARVs.

Drug resistance on the rise

HIV drug resistance could become a ticking time bomb for public health if resistant strains continue to spread. Patients with resistance to cheaper first-line anti-retrovirals must take more expensive alternatives, which both limits their options and vastly increases the costs of treatment.

The World Health Organization estimates that in 2010 one in 20 HIV-positive patients worldwide was infected with a drug-resistant strain. Later studies suggest this figure is rising in low- and middle-income countries.

In South Africa, over 6.4 million people are infected with HIV. Of these, 2.5 million are taking anti-retrovirals. There are 18 different drugs available in the country and patients are usually prescribed three or more drugs concurrently to lessen the chances of resistance.

Around nine percent of patients are failing first-line anti-retroviral therapy year-on-year in South Africa. If a patient fails treatment because of resistance to their prescribed anti-retrovirals, the resistant virus is free to replicate unchecked. If this resistance is not picked up and treated, the virus will destroy the patient’s immune system and speed the progression to AIDS.

Patients who have developed drug resistance are also more infectious than those on a fully effective treatment regimen. There is more virus in their bodily fluids, which means that there is a higher chance they will infect someone else. To make matters worse, the newly-infected person may contract the resistant strain.

Preventing the time bomb

Most public health guidelines recommend a drug resistance test when an HIV-positive patient is diagnosed, and another drug resistance test when a patient’s treatment fails.

A drug resistance test works by ‘reading’ the DNA of the HIV in a sample of a patient’s blood (called sequencing), and checking it against a known list of drug-resistant mutations.

The South African HIV Clinicians Society currently recommends a test only when first-line anti-retroviral therapy has failed. However, fulfilling even this recommendation in practise is a significant challenge. Commercially available tests cost around R5000 per patient, which is an expensive burden on the South African public health sector.

New solutions

New technology, however, has the potential to lessen this burden. Our novel research at the university’s South African National Bioinformatics Institute reduces this cost up to tenfold, making an HIV drug resistance test accessible to almost anyone.

We have developed a web application that analyses HIV samples from many patients, which were sequenced concurrently using a high-throughput sequencing machine. Sequencing in parallel like this significantly reduces the per-patient cost.

High-throughput sequencing machines produce up to several million short, error-prone ‘reads’ of DNA sequences, which must be identified and corrected before they can be used to reconstruct a picture of the HIV in each patient’s blood.

Although this data is much more complex, our process can return results that are not only cheaper but also more accurate than the conventional, expensive approach. More copies of each patient’s HIV can be sequenced, which means it is more likely that drug-resistant mutations will be detected.

The software combines cloud computing with novel algorithms that correct errors in the data and accurately detect drug resistance mutations. The application then returns a simple traffic light report that indicates the drugs to which a patient is resistant (red), or susceptible (green). This report can be interpreted by a primary care physician and used to tailor a treatment regimen to a patient’s drug resistance profile.

This approach to drug resistance testing is inexpensive enough to be included as a routine part of HIV treatment in low-resource settings. It will not only improve individual patient outcomes, but also reduce the public health burden of new infections and transmitted drug resistance.