New HIV drugs should be classed as inferior if they carry a higher risk of resistance


New antiretroviral regimens should be considered inferior to older ones unless trials can demonstrate that their failure doesn’t lead to more drug resistance than standard treatment, Italian researchers argue in Lancet HIV.

They say that recently completed trials of injectable cabotegravir and rilpivirine raise concerns about the risk of developing resistance to two drug classes if the regimen fails.

New antiretroviral drugs and drug combinations are tested before approval in studies known as non-inferiority trials. The aim of these studies is to show that the new drug or combination is non-inferior, or no worse than, the existing standard of care in suppressing HIV and avoiding virological failure.


viral rebound

When a person on antiretroviral therapy (ART) has persistent, detectable levels of HIV in the blood after a period of undetectable levels. Causes of viral rebound can include drug resistance, poor adherence to an HIV treatment regimen or interrupting treatment.

non-inferiority trial

A clinical trial which aims to demonstrate that a new treatment is not worse than another. While the two drugs may have comparable results in terms of virological response, the new drug may have fewer side-effects, be cheaper or have other advantages. 

virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

integrase inhibitors (INI, INSTI)

A class of antiretroviral drugs. Integrase strand transfer inhibitors (INSTIs) block integrase, which is an HIV enzyme that the virus uses to insert its genetic material into a cell that it has infected. Blocking integrase prevents HIV from replicating.


Refers to the mouth, for example a medicine taken by mouth.

What counts as non-inferior? The US Food Administration sets standards for clinical trials, including the statistical analysis. Pharmaceutical companies are required to report not only the headline rate of viral suppression but also the confidence margins – the 95% probability that the true rate of viral suppression lay between two values.

So, for example, the headline difference between two regimens in a comparison of once-daily and twice-daily raltegravir was –5.7%. But the confidence interval of the estimated difference was –0.83% to –10.7%. In this study, the prespecified margin for non-inferiority was –10%, and because the lower margin exceeded –10%, once-daily raltegravir was judged to be inferior to twice-daily raltegravir.

The US Food and Drug Administration recommends that in antiretroviral drug trials in previously untreated people, the inferiority margin should be 12%.

But for studies where virologically suppressed people switch from one regimen to another, the standard is tighter. In 2016, the US Food and Drug Administration told pharmaceutical companies that it would only consider a drug to be non-inferior if the difference in virological failure rates between the new product and the standard of care was no more than 4%.

This tight margin is intended to ensure that people who switch to a regimen that may be more convenient or have fewer side effects will not run a substantially greater risk of viral rebound compared to people who stay on their existing treatment.

However, this way of judging if new regimens are non-inferior doesn’t take into account what happens to the people who experience viral rebound in switch studies.

In their Lancet HIV commentary, Italian physicians Diego Ripamonti and Mauricio Zazzi draw attention to the concerning resistance patterns that emerged in people who switched treatment and experienced viral rebound in several studies in which rilpivirine was combined with an integrase inhibitor.

In the SWORD studies of switching from three-drug treatment to dolutegravir and rilpivirine, 11 people experienced virological rebound. Six people who switched to dolutegravir and rilpivirine developed resistance to rilpivirine. No one who experienced viral rebound on three-drug treatment developed major resistance mutations.

In the LATTE study, participants received oral cabotegravir plus two nucleoside reverse transcriptase inhibitors (NRTIs) for 24 weeks before switching to oral cabotegravir and rilpivirine if they suppressed viral load. Eight people experienced viral rebound and six developed drug resistance, including three cases of integrase inhibitor resistance. Again, no one who experienced viral rebound on the three-drug regimen developed resistance.

The ATLAS-2M study evaluated monthly or two-monthly injections of cabotegravir and rilpivirine in virally suppressed people who had either been taking three-drug treatment or monthly injections of cabotegravir and rilpivirine. Although the numbers were low, more people in the two-monthly injections group experienced virological failure (approximately one in forty) than in the monthly injections group (one in two hundred). Ten participants developed resistance to cabotegravir and eight to rilpivirine.

Developing resistance to both rilpivirine and cabotegravir rules out the use of two drug classes in HIV treatment and limits the range of drug combinations that are available to suppress HIV in the future.

“Given the current availability of oral high genetic barrier regimens, we believe the rate of treatment-emergent resistance should be incorporated in an updated definition of HIV therapy success,” say the Italian physicians.

“Modern treatment options should be ranked according to the risk of failure with resistance [… and] patients should be informed about the potential (although minimal) risk of resistance to integrase inhibitors.”

They note that British HIV Association guidelines recommend that people with HIV who are considering a switch to injectable cabotegravir and rilpivirine should be informed about the risk of resistance to both drugs.


Ripamonti D, Zazzi M. Emergent HIV drug resistance in non-inferiority trials. Lancet HIV, 10: e632-34, 2023.