A total of 360 patients were equally randomised into arms
that omitted or added NRTIs as part of their optimised therapy. As previously
reported, the most common optimised regimen was raltegravir, boosted darunavir
and etravirine (56%), followed by raltegravir, boosted darunavir and maraviroc
(14%). Almost all of those adding in NRTIS used tenofovir disoproxil, together
with lamivudine, emtricitabine and/or zidovudine.
Robust increases in CD4 cell count were observed in both arms
by week 96: mean CD4 cell count was 391 cells/mm3 in the omit
group and 428 cells/mm3 in the add NRTIs group.
Virological data clearly showed that treatment that omitted
NRTIs was non-inferior to therapy that added NRTIs. At week 96, rates of viral
suppression (below 200 copies/ml) were 70% in the omit arm compared to 65% in
the add NRTIs group, with 61% and 59%, respectively, having a viral load below
Analysis of the 316 patients who remained in follow-up at week
96 showed that rates of suppression below 200 copies/ml were 79% in the omit
group and 75% among the patients who added NRTIs (69% and 68%, respectively,
for below 50 copies/ml).
The cumulative probability of virological failure at week 96
was 28% for the omit arm and 30% for the NRTI arm. Factors associated with
higher odds of viral suppression included older age and starting a higher
number of active antiretroviral drugs. Only in those adding NRTIs, people with
a poorer quality of life at baseline were less likely to be virally suppressed
at 96 weeks.
Patients experiencing virological failure rarely developed
new resistance mutations (16% NNRTIs; 3% protease inhibitors; 11% integrase
Patients in the omit group had higher cholesterol values and
also higher Framingham cardiovascular risk-scores than patients in the NRTI
group. The investigators speculate that this is because their regimens did not
include tenofovir, which is known to lower lipids.
Conversely, renal outcomes were better among the omit group,
again probably due to the omission of tenofovir.
Cumulative 96-week morality rates in the omit and NRTI arms
were 0.6% and 5.7%, but the difference was not statistically significant. There
was no evidence suggesting common mechanisms or causes.