Stopping treatment after early ART in infants

Polly Clayden, HIV i-Base

In an oral late breaker, Mark Cotton presented the final results from the Children with HIV Early Antiretrovirals (CHER) trial.

Interim results from CHER, announced in 2007, demonstrated the need for early ART in HIV-infected infants and influenced guidelines worldwide. The standard of care is now universal treatment for infants less than one year (and in WHO guidelines in children up to two years), initiated as early as possible.

In the study, infants <12 weeks of age with CD4 percent >25% were randomised to receive deferred ART (ART-Def), immediate ART stopping after 40 weeks (ART-40W) or immediate ART stopping after 96 weeks (ART-96W). The recommendation that enrolment to ART-Def cease and all children be assessed to start ART was made by the DSMB in June 2007, as starting ART immediately reduced deaths by 75%. All children received treatment with lopinavir/ritonavir (LPV/r) + 3TC + AZT.

Treatment initiation in the deferred arm and re-initiation in the other two were in accordance with previous WHO guidance at a threshold of CD4 percent <25% in infants and <20% after infancy or with clinical disease progression. The primary endpoint was death or progression to CDC severe B or CDC C disease. All analyses were intent-to-treat using time-to-event methods.

A total of 377 infants were enrolled in CHER between 2005 and 2007. They were a median age of 7.4 weeks with a CD4 percent of 35% at baseline and 341 (91%) completed the study. Median follow up was 249 weeks (4.8 years) and the maximum was 309 weeks (5.9 years). The overall proportion of follow up time on ART in the three study arms were 81%, 70% and 69% in ART-Def, ART-40W and ART-96W respectively.

Approximately 75% in the ART-40W and 65% in the ART-96W arms re-initiated treatment by 240 weeks. The median time to starting ART in ART-Def was 20 (IQR 16-25) weeks and to restarting ART after interruption in ART-40W and ART-96W was 33 (IQR 26-45) and 70 (IQR 35- 109) weeks, respectively. Difference between the two deferred arms was 37 (95% CI -11 to 85) weeks, p=0.13. By the end of the trial only 7 children had switched to second line antiretroviral therapy.

When the investigators looked at the total primary endpoints in the study they found 39 (25%) in ART-Def, 31 (25%) in ART-40W and 25 (20%) in ART 96-W. This was mainly due to at least double the number of deaths in ART-Def compared to the other two arms: 22 (18%), 11 (9%) and 9 (7%) in ART-Def, ART-40W and ART-96W respectively. Dr Cotton noted that there were no cases of regimen limiting toxicity.

Time to primary outcome compared to ART-Def, showed 23% fewer events in ART-40W, 42% fewer in ART-96W and 35% fewer in the two arms combined. Hazard ratios (HR) relative to ART-Def were ART-40W, 0.73 (95% CI 0.46-1.17), p=0.19; ART-96W, 0.58 (95% CI 0.35-0.96), p=0.05 and ART-40W/ ART-96W 0.65 (0.43 – 0.98), p=0.04.

Progression to CDC B or C or death was also reduced by 50% and 60% respectively. HR relative to ART-Def were ART-40W, 0.5 (95% CI 0.3-0.8), p=0.005 and ART-96W and 0.4 (95% CI 0.3-0.7), p=0.0003. There were 43, 27 and 18 events in the ART-Def, ART-40W and ART-96 arms respectively. For encephalopathy there were 9, 5 and 2 events.

When the investigators compared the ART-40W and ART-96 arms – including 34 additional children included after ART-Def stopped enrollment (n=143 in each remaining arm) – there was no difference between the two in time to primary outcome, HR 0.84 (95% CI 0.51 – 1.4), p=0.49. The majority of deaths in both arms occurred early, during the initial period of ART.

At the end of the trial 30 (25%) children in ART-40W and 46 (33%) in ART-96W never started continuous ART and CD4 percent was a median of approximately 30% in both arms.

Dr Cotton concluded that treatment during early infancy protects against HIV-related high mortality and morbidity and ART interruption after infancy appeared to be safe. But further analyses of virologic suppression and resistance and immunological response to restarting and interrupting treatment are needed.


Stopping treatment in infants who receive ART during acute infection appeared to be safe in CHER. Although Andy Prendegast questioned the value of the short duration off treatment before restarting (33 and 70 weeks after 48 and 96 weeks of early ART respectively), in his excellent overview exploring “controversies and consequences of early initiation” of ART in infants. [2]

Conflicting results to those from CHER were reported from The Optimising Paediatric HIV-1 Therapy 03 (OPH03) Study – a randomised trial of continued vs interrupted treatment in infants with CD4 >25%, following at least 24 months of ART and restarting if CD4 dropped to 25% – which was stopped by the DSMB due to the high proportion of children restarting by three months. [3] In this study, 42 children were randomised (21 in each arm) and 18/21 in the interrution arm (86%) restarted, the majority (14/21) before 3 months. The children in OPH03 differed from those in CHER in that they were initially treated with ART at a median of about 5 months of age with a lower pre- ART CD4 percentage. However although lower CD4 percentage at randomisation was predictive of starting treatment after <3 months (p=0.04 vs > 6 months) but neither age at ART or pre-ART CD4 were (both p=0.7).

Dr Prendegast questioned whether one or two years early treatment was enough and noted that there was no comparison to early continuous ART.

Emphasising the controversies, a speaker from the floor declared he was “shocked” that treatment interruptions were even being considered in children and suggested such an approached belonged in the “middle ages”.


  1. Cotton M et al. Early ART followed by interruption is safe and is associated with better outcomes than deferred ART in HIV+ infants: final results from the 6-Year randomised CHER trial, South Africa. 19th CROI, 5–8 March 2012, Seattle. Oral abstract 28LB.
  2. Prendergast A. ART for the HIV+ infant: controversies and consequences of early initiation. Symposium abstract. 113.
  3. Wamalwa D et al. Treatment interruption in infants following 24 months of empiric ART: Kenya. Oral abstract 27.

Links to other websites are current at date of posting but not maintained.