People with very high pre-treatment viral loads who took dolutegravir were significantly less likely to have experienced treatment failure nine months after starting treatment compared to people taking other drugs recommended for first-line treatment, a review of more than 2000 people in the United States shows.
The findings, published in the journal Open Forum Infectious Diseases, come from an analysis of a large US cohort, OPERA, which records data on around 90,000 people with HIV receiving treatment in 85 clinics across the United States.
Several large clinical trials have demonstrated differences in response to antiretroviral drugs according to baseline viral load, notably studies of rilpivirine and abacavir. Rilpivirine has been licensed for use in people with baseline viral load below 100,000 copies/ml as a result of these study findings.
The proportion of people starting treatment with high viral load – above 100,00 copies/ml – is unclear. In the United States, previous studies have reported that between 19% and 44% of people starting treatment have viral loads above 100,000 copies/ml and high viral load may be more common in people with low CD4 counts, presenting late for HIV care.
The lack of head-to-head studies of commonly prescribed antiretroviral agents, especially in people with high viral loads, means that it is unclear whether there are differences in the efficacy of these drugs at higher viral loads. Treatment failure in people with high viral load might increase the risk of drug resistance compared to failure of treatment at lower viral loads, so it is important to know if there are differences. Treatment failure in people with high viral load may also lead to disease progression.
To investigate differences between antiretroviral drugs, researchers associated with the OPERA cohort looked at virological failure in people who started treatment at participating clinics with one of four core drugs recommended for first-line treatment between 2013 and 2017: the integrase inhibitors dolutegravir (Tivicay, in Triumeq), elvitegravir boosted by cobicistat (in Stribild or Genvoya) and raltegravir (Isentress), and the protease inhibitor darunavir (Prezista, also in Rezolsta or Symtuza).
Cohort members were selected for this analysis if they had a baseline viral load of at least 100,000 copies/ml, and they started treatment for the first time with one of these drugs as part of a three-drug regimen.
The investigators looked for virologic failure, defined as either failure to achieve viral load below 200 copies 36 weeks after starting treatment, or viral rebound above 200 copies after a previous viral load below 50 copies, within 36 weeks of starting treatment.
They also looked at core agent discontinuation, for whatever reason.
During the study period, 6,223 people started treatment with a three-drug regimen and one-third of these (2,038) had a viral load of at least 100,000 copies/ml. A treatment regimen containing dolutegravir (36%) or elvitegravir (46%) was more common than a regimen containing darunavir (16%) or raltegravir (2%) in those starting treatment with high viral load.
More than two-thirds of people who took dolutegravir did so as part of the three-drug formulation Triumeq (with abacavir/lamivudine). Everyone took elvitegravir as part of a single-tablet regimen, either Stribild or Genvoya. A single-tablet regimen containing darunavir did not become available until after the study period ended, so everyone taking darunavir or raltegravir took these drugs with other tablets.
There were several significant differences between people who started treatment with dolutegravir compared to people who started treatment with darunavir or raltegravir. People receiving darunavir were older, more likely to be African-American and more likely to be covered by Medicare. They also had lower CD4 counts and were somewhat more likely to have viral loads above 500,000 copies/ml.
People receiving raltegravir – a much smaller group – were more likely to be female and to have lower CD4 counts.
Participants were followed for a median of 18 months and discontinuation rates for core agents were high. Almost 80% of those who started treatment with raltegravir discontinued during the follow-up period compared to 54% of those taking darunavir, 41% taking elvitegravir and 32% taking dolutegravir.
People taking dolutegravir were significantly more likely to achieve viral suppression than people taking other core drugs (86% for dolutegravir vs 82% for elvitegravir, 63% for raltegravir and 60% for darunavir, p<0.001). Loss of suppression was rare; in most cases, virological failure was the result of a failure to achieve viral suppression after starting treatment.
After adjusting for demographic factors, HIV-related factors including baseline viral load, year of treatment initiation and insurance coverage, people taking elvitegravir were 46% more likely to experience virologic failure (adjusted hazard ratio 1.46, 95% CI 1.05-2.03), those on darunavir more than twice as likely to experience virologic failure (aHR 2.24, 95% CI 1.50-3.34) and people taking raltegravir at least four times more likely to experience virologic failure (aHR 4.13, 95% CI 1.85-9.24). However, because of the small number of people taking raltegravir, this risk estimate is especially uncertain.
The study investigators say their results provide real-world confirmation of the findings of a meta-analysis of seven clinical trials, which showed that people taking dolutegravir had a greater likelihood of viral suppression than people taking other core drugs, especially if they had high viral load.
They add that more data are needed on the efficacy of new antiretroviral drugs such as bictegravir and new classes of antiretrovirals in people with high viral loads, so that doctors can judge which core drugs are likely to be more effective in this population.
Mills AM et al. Virologic outcomes among people living with human immunodeficiency virus with high pretherapy viral load burden initiating on common core agents. Open Forum Infectious Diseases, 8 (8), ofab363, August 2021 (open access).