Adherence to ART: is the gold standard still needed for viral suppression?

The levels of adherence to antiretroviral (ARV) regimens needed to achieve and maintain HIV viral suppression may be lower today than they previously were, conclude Dr Kathy Byrd of the Centers for Disease Control and Prevention (CDC) and colleagues in the Journal of Acquired Immune Deficiency Syndromes (JAIDS).

The researchers wanted to examine if over 95% adherence to antiretroviral therapy (ART) – typically considered as gold standard to reach viral suppression – is still necessary in the context of the more effective antiretrovirals available today. For this purpose, they used data from a US demonstration project integrating community-based HIV-specialty pharmacists with HIV clinicians that provided services to 765 adults living with HIV across the United States between 2014 and 2016.

Information collected – 12 to 48 months of data for each participant – included prescription fulfilment figures, HIV viral load results and participants’ demographics.

The proportion of days covered (PDC) was key to the calculation of adherence. This reflects the proportion of days for which a person has medication available during a given period of time. To obtain the ART PDC, one must divide the number of days of ART coverage – defined in this study as at least three antiretrovirals – during the chosen period, by the number of days in the same period.

For this analysis, the investigators chose a measurement period of 365 days preceding each viral load test result, and five PDC categories:

• < 50%.
• 50 to 80%.
• 80 to 85%.
• 85 to 90%.
• ≥ 90%.

As per the study objective, the outcome variable for the analysis was a viral load inferior to 200 copies. Although a more stringent criterion such as below 50 copies could have been chosen, this was based on the US Department of Health and Human Services’ definition of virologic treatment failure. A viral load test result could only be included in this analysis if it had a corresponding PDC value (requiring 365 days of ART data before the test).

The researchers calculated the adherence rate required to achieve viral suppression in 90% of viral load tests. Again, a more stringent criterion such as viral suppression in 95% of tests could have been chosen.

Of the 765 clients of the demonstration project, 570 were eligible for the analysis (at least one HIV viral load with a corresponding PDC value). Most of these 570 individuals were older than 50 years (53%), non-black or of unknown race/ethnicity (59%), male (78%), and did not have private insurance (85%).

Two-thirds of the 2427 viral load test results included in the analysis coincided with a PDC of ≥ 90% within the 365 days, and the majority of these results were lower than 200 copies.

Participants' ARV regimens were:

• Integrase inhibitor (INSTI)-based regimens (31%).
• Non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens (21%).
• Protease inhibitor (PI)-based regimens (18%).
• “All other types” regimens, that is, any regimen not falling into the previous categories (30%).

Each of the three first categories contained two nucleoside reverse transcriptase inhibitors (NRTIs) as backbone. The paper does not say which specific ARVs were taken by participants.

After adjustment for all co-variates, individuals who were over 50 years of age (adjusted odds ratio, or aOR 2.33), male (aOR 1.49) and privately insured (aOR 1.77) showed better chances of having a suppressed viral load, compared with persons who were younger, female and non-privately insured. Non-Hispanic black people had lower chances of reaching viral suppression (aOR 0.46) than people from all other races/ethnicities combined.

PI-based regimens required a higher adherence level to achieve viral suppression

No significant differences in the aOR of viral suppression were found between individuals with a PDC of 80 to 85%, or 85 to 90%, compared to that of people with a PDC of ≥ 90%.

Individuals from the “all other” regimens category had lower viral suppression chances than those taking a PI-based regimen. The differences in adjusted viral suppression odds between PI, INSTI and NNRTI-based regimens were not statistically significant.

Importantly, the study also revealed that overall, the adherence rate required to achieve viral suppression in 90% of all the viral load tests was 82%. However, this varied according to the regimen types:

• 75% for INSTI-based regimens.
• 78% for NNRTI-based regimens.
• 87% for PI-based regimens.
• 99% for “all other” regimen types.

Therefore, it may be that in our today’s era of antiretrovirals with better pharmacokinetics (PK) and safety profiles than their predecessors, the adherence levels necessary to the achievement of viral suppression is lower than the conventional values of over 95% (or for some, over 90%), and more in line with what it is for other chronic diseases such as hypertension (80%).

While it is difficult to compare studies on the adherence threshold required for viral suppression (different HIV RNA cut-off levels; different adherence measures, such as patients’ self-report, pill counts and proportion of days covered; different thresholds reflecting treatment failure), the finding of there being no significant difference between the upper adherence rates found is consistent with previous studies.

However, although no viral suppression odds differences were found between either INSTI-based or NNRTI-based regimens, when compared with PI-based ones, it should be highlighted that the PI-based regimens required a higher adherence level to achieve viral suppression. The even higher adherence observed threshold for “all other” regimens to reach viral suppression is not surprising: this category includes regimens not recommended by treatment guidelines and may represent non-standard or salvage treatments for people who have experienced previous treatment failure.

Find out more: Adherence to HIV treatment

Overall, the main findings from this small study are encouraging as they suggest that today’s antiretrovirals are more forgiving than yesterday’s. However, while acknowledging that their results reflect the improved potency and efficacy of newer PK regimens, the authors of the JAIDS paper clearly state that although they discovered an ART adherence rate to reach viral suppression in 90% of test results that is lower (82%) than the gold standard rate of over 90% or 95%, clinicians should continue to advise patients to take their antiretrovirals as strictly prescribed as possible, and offer them adherence support whenever they consider it necessary.

Larger studies on adherence levels to reach viral suppression may be needed to help clinicians fine-tune their choice of optimal ARV regimens for patients (taking into account their adherence factors, such as lifestyle, understanding of HIV and of the need to be treated, etc.) and provide better adherence advice.