Several
recent clinical trials have indicated that structured antiretroviral treatment
interruption is a risky strategy, but the reasons for this remain unclear.
Two studies presented this week at the 15th Conference
on Retroviruses and Opportunistic Infections (CROI 2008) in Boston shed further
light on the biological mechanisms underlying the increased risks of intermittent
therapy.
SMART
The
large Strategies for Management of Anti-Retroviral Therapy (SMART) trial included
5472 mostly treatment-experienced participants with a CD4 cell count above 350
cells/mm3 at baseline. Most (73%) were men, 29% were black, the median age was
43 years, and the median CD4 count was 597 cells/mm3.
Subjects were randomly
assigned to 2 treatment strategy arms. Patients in the "drug conservation"
(DC) arm interrupted antiretroviral therapy when their CD4 count was above 350
cells/mm3 and resumed when it fell back to 250 cells/mm3. Participants in the
"viral suppression" (VS) arm remained on continuous therapy throughout
the study.
As
previously reported, SMART was halted in January 2006 after interim results
showed that participants in the treatment interruption arm had not only higher
rates of opportunistic illness and death due to any cause, but also unexpectedly
higher rates of serious non-opportunistic heart, liver, and kidney events.
At
the Retrovirus conference, members of the SMART team presented data from a study
that aimed to explain the increase in cardiovascular events. They analyzed a variety
of biomarkers associated with inflammation, blood coagulation, and endothelial
dysfunction:
•
high sensitivity-C reactive (hsCRP);
•
interleukin-6 (IL-6);
•
serum amyloid A;
•
serum amyloid P;
•
D-dimer;
•
prothrobmin fragment 1+2 (F1.2).
The
investigators conducted both a case-control study of SMART participants who died
and an analysis of blood samples from a random subset of study subjects.
In
the case-control part, the researchers paired 85 patients who died through January
2006 (55 in the DC arm, 30 in the VS arm) with 2 surviving participants matched
for sex, age, and country. In the blood sample analysis, they assessed stored
blood samples from a random subset of 500 individuals (half in each arm) collected
before and 1 month after study randomization.
Results
•
Patients who died had a lower baseline CD4
count (545 vs 614 cells/mm3), higher risk of diabetes (26% vs 15%), more treatment
of hypertension (39% vs 25%), more prior cardiovascular disease (12% vs 5%), and
were more likely to smoke (58% vs 32%) compared with survivors.
•
HIV RNA levels and HAART use did not differ
in the 2 groups.
•
Higher levels of IL-6 and D-dimer were strongly
associated with an increased risk of death (adjusted odds ratio [OR] 11.8 and
26.5, respectively).
•
Hs-CRP and amyloid A had a weaker association
with increased mortality (adjusted OR 2.8 and 2.6, respectively).
•
Amyloid P and F1.2 were not linked to an increased
risk of death (adjusted OR 1.1 and 1.2, respectively).
•
In the blood sample analysis, levels of IL-6
and D-dimer increased as viral load rose after interrupting therapy in the DC
arm, but remained stable in the VS arm.
•
Both baseline levels of IL-6 and D-dimer,
as well as increases in these markers after interrupting therapy, were associated
with elevated mortality.
•
Elevated IL-6 and D-dimer levels were associated
more strongly with all-cause mortality than with death specifically due to cardiovascular
disease.
Conclusion
"In
summary, elevated levels of D-dimer and IL-6 identify patients at very high risk
of death," the researchers concluded. "Increases in IL-6 and D-dimer
following art interruption may explain in part the increased risk of all-cause
mortality and cardiovascular disease in the DC compared to the VS group."
"The association
with all-cause mortality suggests that HIV infection results in activation of
coagulation and inflammatory pathways that may impact multiple organs," they
continued. "It is possible that HIV viremia directly [a]ffects vascular endothelium
resulting in increased tissue factor transcription, which in turn initiations
a coagulation cascade."
Presenting
investigator Lewis Kuller characterized the findings as "remarkable,"
noting that he had never seen such a strong association between biomarkers and
increased mortality in studies of the general population. STACCATO
Researchers
also presented data from the HIV Netherlands Australia Thailand Research Collaboration's
STACCATO trial. This study included 430 participants, mostly (80%) from Thailand.
Participants first took antiretroviral therapy for 6 months, then those with a
CD4 count above 350 cells/mm3 and viral load below 50 copies/mL were randomly
assigned to continue therapy or undertake a treatment interruption. Treatment
was restarted if CD4 count fell below 350 cells/mm3.
As
previously reported, participants in the interruption arm had more minor opportunistic
infections and those on continuous therapy experienced more drug side effects,
but none in either arm experienced any AIDS-defining events or deaths due to HIV/AIDS
or drug toxicity.
At the Retrovirus conference, STACCATO investigators
reported on a study of biomarkers of endothelial dysfunction (damage to the blood
vessel lining) in 145 Thai participants (97 in the interruption arm, 48 in the
continuous therapy arm).
STACCATO
subjects had fewer baseline risk factors for cardiovascular disease compared with
SMART; more were women (62%), blood lipid levels were lower, and they were less
likely to smoke. On the other hand, they had more advanced HIV disease and a lower
median CD4 cell count.
The
researchers analyzed blood samples collected pre-treatment, while on treatment
before study randomization, at 12 weeks (during treatment interruption), and (for
the interruption arm) 24 weeks after resuming therapy. Here, too, a variety of
biomarkers signaling endothelial dysfunction, inflammation, and metabolic disruption
were measured, including:
•
s-VCAM-1;
•
p-selectin;
•
CRP;
•
IL-6;
•
IL-10;
•
MCP-1;
•
MIP 1-alpha;
•
leptin;
•
adiponectin.
Results
•
12 weeks after randomization, different changes
in biomarkers were observed in the continuous therapy and treatment interruption
arms.
•
Differences were significant for MCP-1, MIP
1-alpha, and adiponectin.
•
MCP-1 and s-VCAM-1 increased after treatment
interruption, followed by a non-significant decrease after treatment resumption.
•
IL-10 and adiponectin decreased during treatment
interruption and did not increase after treatment resumption.
•
These biomarker changes were strongly correlated
with changes in viral load.
•
However, no significant association was observed
between viral load and IL-6, CRP, or the other markers.
Since
none of the study participants experienced cardiovascular events or died of cardiovascular
causes, the researchers could not look for associations between the biomarkers
and these outcomes.
The
STACCATO researchers concluded that there were significant differences in biomarkers
of endothelial dysfunction between the treatment interruption and continuous therapy
arms, and that changes in the interruption arm only partially reversed after restarting
therapy.
Taken
together, the SMART and STACCATO findings suggest that treatment interruption
is associated with physiological changes -- including increased inflammation and
blood coagulation -- that may increase the risk of cardiovascular events and death.
These changes seemed to be mediated by alterations in HIV RNA levels, not a direct
effect of the drugs themselves. Based on these results, it may be prudent to monitor
cardiovascular and other metabolic biomarkers in HIV infected patients.
02/08/08
References
L
Kuller and others (SMART Study Group). Elevated Levels of Interleukin-6 and D-dimer
Are Associated with an Increased Risk of Death in Patients with HIV. 15th Conference
on Retroviruses and Opportunistic Infections. Boston, MA. February 3-6, 2008.
Abstract 139.
A Calmy, A Nguyen, F Montecucco, and others (STACCATO study
team). HIV Activates Markers of Cardiovascular Risk in a Randomized Treatment
Interruption Trial: STACCATO. 15th Conference on Retroviruses and Opportunistic
Infections. Boston, MA. February 3-6, 2008. Abstract 140.
Mechanisms
for Elevated Cardiovascular Risk in SMART and STACCATO Treatment Interruption
Trials 
Several
recent clinical trials have indicated that structured antiretroviral 
