A single-tablet combination of cobicistat (Tybost)-boosted darunavir (Prezista) plus emtricitabine (Emtriva) and tenofovir alafenamide (Vemlidy) (D/C/F/TAF) proved virologically noninferior to continuing a suppressive boosted protease inhibitor (PI) plus emtricitabine/tenofovir disoproxil fumarate (F/TDF, Truvada) at 48 weeks in the phase 3 randomized Emerald trial. Adverse event rates were similar in people who switched to D/C/F/TAF and in those who continued their baseline boosted PI.
D/C/F/TAF (800/150/200/10 mg once daily) is licensed in the European Union and under regulatory review in the United States. EMERALD is a 96-week open-label phase 3 international trial randomizing patients taking an effective boosted PI plus F/TDF to continue that combination or change to D/C/F/TAF. Participants had a viral load below 50 copies/mL for at least two months on their boosted PI. No one had a history of darunavir failure or darunavir-associated resistance mutations if they had a historical genotype. The primary U.S. Food and Drug Administration (FDA)-stipulated endpoint was the proportion of people with virologic rebound at 48 weeks, defined as (1) a confirmed viral load at or above 50 copies/mL, or (2) early discontinuation for any reason with a last viral load at or above 50 copies/mL, or (3) a single viral load above 50 copies/mL at week 48.
The 763 people randomized to D/C/F/TAF were similar to the 378 randomized to the control regimen in median age (46 years), proportion of women (18%), proportion of whites (74.9%), median years since HIV diagnosis (9.3), and proportion with prior virologic failure (14.8%). Most participants (70.4%) were taking darunavir in their baseline regimen.
Nineteen people randomized to D/C/F/TAF (2.5%) and eight randomized to continue their baseline PI (2.1%) had viral rebound by week 48. The 0.4% difference and 95% confidence interval (-1.5% to 2.2%) confirmed the noninferiority of D/C/F/TAF to the control regimen in an analysis with a noninferiority margin of 4%. No one stopped treatment for virologic nonresponse. Twelve of the 19 people (63%) with study-defined virologic failure of D/C/F/TAF and four of the eight with failure in the control arm regained a viral load below 50 copies/mL by week 48 without changing antiretrovirals. In an FDA snapshot analysis, 48-week virologic success rates were 94.9% with D/C/F/TAF and 93.7% with a continued PI. Viral genotyping of one rebounder in the D/C/F/TAF arm and three in the control arm found no mutations conferring resistance to darunavir, other PIs, tenofovir, or emtricitabine.
Proportions of participants with one or more grade 3 or 4 adverse events were 6.8% in the D/C/F/TAF arm and 8.2% in the control arm. Respective proportions with serious adverse events were 4.6% and 4.8%, and proportions with adverse events leading to treatment discontinuation were 1.4% and 1.3%. Nasopharyngitis and upper respiratory tract infection were the most common adverse events, affecting about 10% in each study arm.
Median gain in "good" high-density lipoprotein (HDL) cholesterol through 48 weeks favored the D/C/F/TAF group (+2.7 versus 0 mg/dL, P < .001), while total-to-HDL cholesterol ratio change slightly but significantly favored the control arm (+0.1 versus +0.2, P = .036). A nonsignificantly higher proportion of participants assigned to D/C/F/TAF started antilipid therapy during the trial (2.6% versus 1.9%, P = .54).
Changes in estimated glomerular filtration rate (eGFR) were modest through 48 weeks and favored either D/C/F/TAF or the control regimen depending on whether the investigators calculated eGFR by serum creatinine or cystatin C. Hip and spine bone mineral density changed little over 48 weeks in the control arm while rising about 1.5% in people who switched to D/C/F/TAF (P < .001 for D/C/F/TAF versus control at both hip and spine).
Emerald investigators proposed that single-tablet D/C/F/TAF combines TAF-related safety advantages with already-appreciated darunavir efficacy and a high genetic barrier to resistance.