Children did just as well on antiretroviral therapy without
routine laboratory monitoring of CD4 cell counts or drug side-effects as those who received regular monitoring during
four years of follow-up, researchers
from Uganda and Zimbabwe reported on 5 March at the 20th Conference
on Retroviruses and Opportunistic Infections (CROI 2013) in Atlanta.
The researchers say that money spent on routine laboratory
monitoring would be better spent on expanding access to treatment for children,
since only 28% of children with HIV who needed treatment were receiving it in
2011, according to UNAIDS statistics.
The findings come from the ARROW study, which recruited over
1200 children with HIV in Uganda and Zimbabwe.
Laboratory monitoring of antiretroviral therapy in adults
and children should ideally consist of regular checks on the CD4 cell count,
viral load and blood and liver markers such as haemoglobin and liver enzymes,
in order to monitor potential side-effects and signs of treatment failure.
In settings where laboratory facilities and resources are
limited, laboratory monitoring may not include viral load or CD4 counts.
Laboratory monitoring strategies for adults were
investigated in the DART study, which found that patients who received regular
CD4 cell monitoring while taking antiretroviral drugs in sub-Saharan Africa had
only slightly better outcomes after five years when compared to those who
received clinical monitoring alone. CD4 cell counts or clinical symptoms were
used to determine when first-line treatment was failing, leading to a switch to
Researchers from Uganda, Zimbabwe and the United Kingdom
found that the difference in outcomes only became apparent after the second
year of treatment in the five-year study, leading them to conclude that
clinical monitoring alone is feasible during the first two years of treatment,
and that CD4 counts should be reserved for monitoring treatment beyond this
The UK’s Medical Research Council and the UK Department for
International Development funded a second study, to examine whether laboratory
monitoring is necessary in children, and whether starting treatment with a
four-drug regimen before stepping down to three-drug treatment results in
better long-term outcomes than three-drug treatment alone.
The study was conducted in Zimbabwe and Uganda, and
recruited 1206 children ranging in age from 4 months to 17 years (median age 6
years) with fairly advanced HIV disease (median CD4 cell percentage 12%, 14%
with WHO stage 4 HIV disease and 56% with stage 3 disease).
Participants were randomised to receive either:
- Laboratory monitoring of biochemistry, full
blood count and CD4 count every 12 weeks, with a switch to second-line treatment in cases
of a new or recurrent WHO stage 4 HIV-related illness, or a 30% decline in CD4
count from its on-treatment peak.
- Clinical monitoring every 12 weeks, with
laboratory monitoring only at specific request of the clinician, and switch to
second-line treatment in the event of a new or recurrent WHO stage 4 HIV-related illness.
Participants received abacavir/3TC plus an NNRTI, with
factorial randomisations to once- or twice-daily abacavir/3TC and to either an
induction/maintenance regimen of four drugs stepping down to three, or to a
three-drug regimen. Neither of these randomisations had any impact on the
overall study results, and there was no significant difference in outcome
between any of these arms, so these will not be discussed further in this
report. All participants received cotrimoxazole, and caregivers were also asked
to consent to randomisation to stopping or continuing cotrimoxazole
prophylaxis. Results of the cotrimoxazole sub-study will be described in a
The primary endpoints of the study were new WHO stage 4
events or death (clinical efficacy) and stage 3 or 4 adverse events (safety).
After a median of four years of follow-up, there was no
significant difference between arms in the proportion of children who remained on
first-line treatment (94% in the laboratory arm vs 95% in the clinical monitoring
arm) and no significant difference in clinical events or deaths (96% survival in the laboratory arm vs 95% survival in the clinical monitoring arm). There was a
small and non-significant elevated risk of WHO stage 4 event or death in the
clinical arm (hazard ratio 1.13, 95% confidence interval 0.73-1.53), and a
small but significant difference in survival after the first year of treatment
(an excess of deaths of 0.6/100 child-years in the clinical arm, p=0.009).
Twelve of the fourteen excess deaths in the clinical monitoring arm occurred in
children over the age of eight years, in whom adherence difficulties had been
No significant difference in CD4 cell count gains was
observed, and nor was there a significant difference in the proportion of
children with a viral load suppressed below 400 copies/ml (78 vs 77%) or 80 copies/ml after
four years on treatment. Critics of clinical monitoring argue that viral load testing is essential in order to ensure that high-level resistance does not develop due to protracted periods of viral replication. No data were presented on responses to second-line treatment because responses to second-line treatment were not an endpoint in this study.
There was no significant difference in grade 3 and 4
adverse events, and 70% of the grade 3 events were cases of asymptomatic
neutropenia. Indeed, only 1.4% of all laboratory toxicity tests were carried
out at the request of a doctor to assist in clinical management; the remainder
(just over 125,000 tests) were carried out routinely.
The study also found
that laboratory monitoring was not cost-effective, and in an accompanying
policy briefing, the trial investigators urge that laboratory resources should
be focused on carrying out tests that are clinically indicated, rather than
routinely monitoring all patients. They suggested that CD4 counts should be
used to check whether it is necessary to switch to second-line treatment in
children who appear to be clinically failing.
The researchers concluded that their findings demonstrate
that antiretroviral treatment for children is safe and highly effective in the
absence of laboratory monitoring.