Five-year safety of Cabenuva

Most clinical trials of long-acting formulations of cabotegravir + rilpivirine (the combination is called Cabenuva) have lasted one to two years. However, data collected from 274 participants who were taking Cabenuva every four or eight weeks for up to five years is now available. This data showed that between 74% and 94% of participants maintained an undetectable viral load. Although injection site reactions were common, only 2% of participants stopped using Cabenuva because of them. Three people experienced serious adverse reactions and three people died (their deaths were not likely due to Cabenuva). In general, long-term use of Cabenuva was well tolerated and effective.

Study details

A clinical trial called Latte-2 began in 2014 and recruited participants from the following countries:

  • Canada
  • France
  • Germany
  • Spain
  • U.S.

Latte-2 was a randomized clinical trial with a complex design.  

All participants were not taking HIV treatment (ART) before they began taking the study medicines in Latte-2. Participants were given once-daily oral formulations of HIV treatment for a total of 20 weeks—a combination of pills containing cabotegravir, abacavir and 3TC. For the final four weeks, the drug rilpivirine was added. This oral lead-in period was meant for patients to achieve an undetectable viral load and to assess their tolerance for the medications. When Latte-2 started in 2014, not much was known about the safety and tolerability of cabotegravir in combination with rilpivirine.

Today, people initiating Cabenuva have a much shorter oral lead-in—usually four weeks. Furthermore, other clinical trials have found that an oral lead-in may become optional in the future, as cabotegravir is very well tolerated. Rilpivirine is an older medication, and so doctors are familiar with its safety.

Participants in Latte-2 who completed the first 20 weeks of oral therapy were randomly assigned to one of the following groups:

  • long-acting formulations of injectable cabotegravir + rilpivirine given every four weeks – 115 people
  • long-acting formulations of injectable cabotegravir + rilpivirine given every eight weeks – 115 people
  • continued oral cabotegravir + abacavir + 3TC for 96 weeks, followed by a switch to injectable therapy – 56 people

Note that doses and volumes of injectable medicines given every eight weeks are greater than those given every four weeks.

After 96 weeks, participants who were on oral ART were offered the opportunity to switch to long-acting formulations of injectable cabotegravir + rilpivirine, given every four or eight weeks. This study design ensured that all participants had the chance to receive long-acting therapy—a factor that likely motivated enrollment.

The first two groups—long-acting cabotegravir and rilpivirine given every four and eight weeks—were called the randomized groups by the researchers. The third group—prolonged use of oral meds—was called the extension group by the researchers.

About injectable cabotegravir and rilpivirine

Once the injection phase of a particular group commenced, nurses injected participants in one buttock with long-acting cabotegravir and the other buttock with long-acting rilpivirine—for a total of two injections. People who received long-acting drugs every eight weeks received a greater dose and volume of these drugs than people who were injected every four weeks.

About the participants

Upon entry to the study, participants, on average, were in their 30s, more than 90% were men, and they were mostly White (80%) or Black (10%). CD4+ cell counts were between 450 and 500 cells/mm3 and viral load was 30,000 copies/mL.


After five years, the proportions of participants with a suppressed viral load were as follows:

  • long-acting formulations of injectable cabotegravir + rilpivirine given every four weeks – 74%
  • long-acting formulations of injectable cabotegravir + rilpivirine given every eight weeks – 88%
  • extension group – 90% among those getting injections every four weeks; 94% among those getting injections every eight weeks

Note that the third group in this study consisted of people who were on oral ART for 96 weeks and then switched to injectable therapy and remained on injectable therapy for more than three years (though they were in the trial for a total of five years).

By the fifth year of the study, different proportions of people on each regimen left the study prematurely because of adverse events (these are explained later in this report). As a result, there was no information about these people’s viral loads by the fifth year of the study. To minimize statistical bias when interpreting the virological results at year five, the people who dropped out are considered to have a detectable viral load and still counted as being in the trial for the overall results. Cases of confirmed virological failure among people who remained in the study appear below.

Confirmed virological failure

Participants were considered to have virological failure with two successive viral load test results of 200 copies/mL or greater. Using this metric, virological failures were distributed as follows:

  • randomized to injections every four weeks – 3% had virological failure
  • randomized to injections every eight weeks – 0% had virological failure
  • extension group – 3% had virological failure; all were getting injections every eight weeks

Adverse events

The term adverse events in clinical trials describes any unfortunate events that occur. These may be caused by drug side effects, the underlying disease process or activities outside of the trial (such as accidents).

More people in the group randomized to receive injections every four weeks left the study prematurely for the following reasons:

  • acute kidney injury
  • attempted suicide
  • coronary artery disease
  • persistent fatigue
  • complications of addiction
  • muscle weakness
  • persistently swollen lymph nodes
  • hepatitis C virus infection
  • excessive blot clot formation

For the most part, it is very unlikely that these were related to the study medicines. Indeed, few adverse events related to treatment were sufficiently bothersome that participants left the study. Two of those adverse events are as follows:

  • injection site pain – two people receiving injections every eight weeks and one every four weeks
  • injection site nodule

Common side effects while receiving injectable formulations were mainly injection site reactions—swelling, redness, discomfort and pain. These side effects were generally mild or moderate and resolved without treatment after a few days. Over the course of the study, injection site reactions became less common.

Other side effects after injection included the following:

  • fever
  • back pain
  • fatigue

These were usually temporary.

Three people who were receiving injections every four weeks died during the study for the following reasons:

  • seizures – this was unrelated to the study medicines; death occurred at week 30
  • coronary artery disease – possibly caused by use of cocaine; death occurred at week 223
  • heart attack – potentially related to the study medicines but ViiV and Janssen (the study sponsors) dispute this; death occurred at week 139

Lab tests

Severely abnormal lab test results occurred with 13% of participants who received long-acting therapy having elevated levels of the enzyme creatine kinase in their blood samples. There are several subtypes of creatine kinase, and the subtypes in these cases was not revealed. However, elevated creatine kinase can occur with inflammation and injury of skeletal muscles, the heart or brain. This adverse effect was temporary.

Another severely abnormal laboratory test result—elevated levels of the enzyme lipase— occurred with 8% of participants who received injectable therapy. Lipase is an enzyme produced by the pancreas gland. Excessive levels of lipase in the blood may suggest inflammation of the pancreas gland. However, there were no accounts of abdominal pain associated with elevated lipase in the report on Latte-2 released by ViiV. This adverse event was temporary.

Less than 1% of participants had elevations in total cholesterol and LDL-cholesterol (so-called “bad” cholesterol).

Cardiograms done at the start and end of the study did not find any abnormalities.

—Sean R. Hosein


Smith GHR, Henry WK, Podzamczer D, et al. Efficacy, safety and durability of long-acting cabotegravir and rilpivirine in adults with human immunodeficiency virus type 1 infection: 5-year results from the Latte-2 study. Open Forum Infectious Disease. 2021. Aug 25;8(9):ofab439.