Switching from a Protease Inhibitor to Efavirenz, Nevirapine or Abacavir
Written by Jules Levin
In suppressed patients on a 2 NRTI- 1 PI regimen, switching from the PI to
nevirapine (NVP), efavirenz (EFV), or abacavir (ABC) is a common practice to
improve convenience and compliance and to potentially reverse or slow or
prevent toxicities/side effects/lipodystrophy.
Summary of study: This 12-month study compared patients with undetectable
viral load on a PI regimen who switched from a PI to NVP, EFV or Abacavir. In
this Study, they found that switching from a PI to an ABC regimen can improve
total cholesterol and triglycerides for some patients. Of 8 patients with
>400 fasting triglycerides before switching to ABC, 4 had <400 after 12
months following the switch to ABC. Median fasting total cholesterol
decreased after 3 months following the switch to abacavir from 200 to 180 and
remained at that level for the 12 months of the study. 26 patients had total
cholesterol >240 before switching to ABC and 9 of these 26 had <240 after 12
months on ABC. It appeared as though cholesterol could decline quickly, but
triglycerides took more time to improve. Patients with elevated total
cholesterol & triglycerides who switched to NVP or EFV did not see
improvement in total cholesterol. Gatell reported that bad cholesterol (LDL)
decreased and good cholesterol (HDL) increased in the efavirenz and
nevirapine arms. Median glucose was normal for patients at the start of the
study and remained normal. Patients were able to maintain viral suppression
equally in all 3 switch groups by an ITT analysis. As expected there were
more drug discontinuations in the EFV & NVP arms due to adverse events & side
effects and more discontinuations in the abacavir arm due to viral failure.
It appeared that most or all viral failures in the abacavir arm were due to
prior suboptimal nuke experience resulting in nuke resistance.
In the Late Breaker session of the final day of the Conference, the NEFA
study was presented by Jose Gatell and a group of Spanish researchers. He
said it was the first randomized comparative study of these 3 drugs.
Study patients were on HAART for a median of 29-31 months (6-64 months), no
differences between the arms. Patients had to have <200 copies/ml HIV RNA for
at least 6 months prior to this study. 46%-56% had prior NRTI mono or dual
therapy before the HAART regimen they were on at the start of this study: 50%
in NVP arm, 56% in EFV arm, 46% in ABC arm. The CD4 counts were 500-550
before the switch. About 35% previously had AIDS. 34-42% of study
participants reported IVDU as exposure risk for HIV. About 25% of patients
The study primary endpoint was the proportion of patients developing
treatment failure at 12 months (ITT switch=faulre). Secondary endpoints were
tolerability and CD4 response. Quality of life, metabolic/lipodystrophy and
resistance substudies were conducted. About 150 patients were randomized to
each of the 3 study arms. 9 patients in the ABC arm switched due to adverse
event, 6 due to viral failure; in the NVP arm, 26 patients switched due to
adverse event, 2 due to viral failure; in the EFV arm, 26 switched due to AE
and 1 due to viral failure.
40%-50% of the patients were on an indinavir regimen with AZT/3TC or d4T/3TC.
About 20% of patients were taking a nelfinavir regimen with d4T/3TC or
By the ITT analysis the proportion of patients with undetectable viral load
(<200 copies/ml) was about the same (no statistically significant
differences) between the 3 treatment groups at 12 months: 77% ABC, 72% EFV,
77% NVP. By the on-treatment analysis there was a statistically significant
difference (p=0.02) between the ABC (86%) and the EFV & NVP arms (94%).
Gatell said this was due to more viral failures in the ABC arm. There were
more discontinuations due to adverse events in the NVP & EFV arms.
The number of patients developing viral failure in the ABC arm was
significantly (p=0.019) higher (16) than in EFV (5) and NVP (8) arms. But
significantly (p=0.056) less patients on ABC (9) switched due to adverse
events compared to NVP (26) and EFV (26).
Gatell performed genotypic resistance testing and concluded that the most
likely reasons for the ABC viral failures were pre-existing NRTI resistance
due to prior suboptimal mono and dual NRTI therapy.
NRTI There were 29 viral failures (16 ABC, 5 EFV, 8 NVP), and 23 of the
failures were reported due to suboptimal NRTI therapy prior to HAART. Gatell
reported 14 of the 16 ABC viral failures were due to prior suboptimal nuke
therapy. And 5/8 of the NVP and 4/5 of the EFV viral failures were due to
prior suboptimal NRTI therapy.
12 of the 16 ABC viral failures had genotypic resistance testing and all 12
had NRTI mutations. Of the 8 NVP viral failures, 4 patients had genotypic
testing and 4 had NRTI mutations and 4 had NNRTI mutations. They genotyped 1
of the 5 EFV failures and that patient had NRTI & NNRTI mutations. The
implication from these findings is that ABC viral failures were due to
previous existing NRTI resistance from prior suboptimal mono and dual
therapy. This and several other studies have clearly shown that there is a
risk of viral failure when switching to ABC triple nuke regimen if patient
had prior nuke experience and resistance. These study findings suggest that
if the patient had no prior nuke resistance switching from a PI to an ABC may
be safe from viral failure.
Summary of Adverse Events Leading to Switch of Therapy
There were significantly (p=0.031) more adverse events experienced in the NVP
(55%) and EFV (58%) arms than in the ABC arm (44%). There was a 3% rate of
ABC hypersensitivity. 35% of patients receiving EFV were reported to
experience neuropsychiatric adverse events compared to 12% in the NVP arm and
15% in the ABC arm. 20% in the NVP arm experienced rash compared to 12% in
the EFV arm and 5% in the ABC arm. The ABC arm had slightly more GI toxicity
reported. Liver toxicity was reported by 8% in the NVP arm vs 3% in the ABC &
NVP discontinuations were mainly due to rash; EFV to neuropsychiatric side
effects; and ABC hypersensitivity
NVP: 13 rash, 4 neuropsychiatric, 4 liver toxicity, 3 rash+liver toxicity, 2
GI toxicity, 0 hypersensitivity
EFV: 20 neuropsychiatric, 3 rash, 2 GI toxicoty, 0 liver toxicity, 1
ABC: 1 neuropsychiatric, 4 GI toxicity, 4 hypersensitivity
About 60% of the patients across all 3 study groups had normal fasting
triglycerides before switching therapy, and about 70-80% had normal
cholesterol (<240). The number of patients on NVP or EFV with >400 mg/dL or
200-400 fasting triglycerides did not improve after 12 months on therapy.
Gatell reported that the number of patients with >400 mg/dL fasting
triglycerides significantly (p=0.042) improved from 8 to 4 following 12
months on ABC. The number of patients on ABC with 200-400 triglycerides did
not change over the 12-month study.
Gatell reported that median total fasting cholesterol significantly improved
(p=<0.001) from about 200 to 180 while fasting cholesterol remained about the
same for the EFV & NVP arms. The percentage of patients with mean total
cholesterol >240 mg/dL remained about the same in the EFV & NVP arms. Gatell
reported that good cholesterol (HDL) increased and bad cholesterol (LDL)
decreased in the EFV and NVP arms. There were 26 patients who received ABC
with >240 total cholesterol at the beginning of the study, and 8 after month
3, 15 after month 6, 11 after month 9, and 9 at the end of the 12 month
study. The changes at months 3, 9, and 12 were statistically significant
compared to the start of the study.
The median glucose levels remained stable and normal for patients throughout
the study, no changes.
The numbers of patients reporting clinical lipodystrophy did not change much
in any of the three arms throughout the study.