7.1.1 When to start ART in adults and adolescents
Consolidated ARV guidelines, June 2013
Rationale and supporting evidence: Initiating ART based on CD4 cell count
Since 2010, evidence and programmatic experience have continued to shift the riskbenefit ratio towards initiating ART earlier. Increasing evidence also indicates that untreated HIV may be associated with the development of several non-AIDS-defining conditions (including cardiovascular disease, kidney disease, liver disease, several types of cancer and neurocognitive disorders) (15–17) and that initiating ART earlier reduces such events and improves survival.
Recent evidence (18) also show that ART substantially reduces sexual transmission in HIV-serodiscordant couples, but not all studies have reported survival benefits. At the same time, more convenient and less toxic regimens have become more widely available, and ARV costs have continued to fall. How early ART should be started is still debated, and the Guidelines Development Group paid close attention to evaluating the potential benefits and harms to the individual and community in developing these new recommendations.
Initiating ART in individuals with symptomatic and asymptomatic HIV disease at a CD4 count ≤350 cells/mm3 as a priority
The benefits of initiating ART are greatest among individuals with symptomatic HIV disease or those with lower CD4 counts. The 2013 Guidelines Development Group did not change the strength and quality of evidence for this recommendation established in the 2010 ART guidelines (2). Moderate-quality evidence from two randomized controlled trials and several observational studies shows that initiating ART at CD4 ≤350 cells/mm3 significantly reduces mortality, disease progression and the incidence of opportunistic diseases, especially TB and non-AIDS-defining conditions (2).
Initiating ART at a CD4 count between 350 and 500 cells/mm3
The risk-benefit analysis of the rationale for ART initiation between 350 and 500 CD4 cells/mm3 in these guidelines was debated. The Guidelines Development Group agreed that impact on HIV transmission is strongly supported by the evidence. The quality of evidence for clinical benefit of earlier ART initiation was rated as moderate using the GRADE system, as it mostly relies on observational data mainly from high-income countries.
The Guidelines Development Group strongly recommended earlier ART as a public health approach. In settings where feasibility of implementation is a concern, the Guidelines Development Group suggested conducting operational research during implementation to assess context-specific factors such as feasibility, linkage to and retention in care, adherence and resource allocation.
The recommendation for initiating ART at CD4 counts between 350 and 500 cells/mm3 is based on a systematic review with GRADE evidence profiles (Web Annex) that assessed the quality and strength of the evidence from 21 observational studies (8,19–39) and three randomized controlled trials (3,18,40) reporting morbidity, mortality and immunological and virological outcomes. They showed that initiating ART at a CD4 count >350 cells/mm3 compared with treatment at a CD4 count ≤350 cells/mm3 reduced the risk of progression to AIDS and/or death, TB, development of a non-AIDS-defining illness and increased the likelihood of immune recovery. Although no studies suggest that earlier ART causes individual harm, these studies were of limited duration.
The pooled analysis of the observational studies found a consistent decreased risk of death with earlier initiation of ART in 13 studies (21–23,26,29–31,34–39) and a decreased risk of progression to AIDS or death in 9 studies (21,23,26,27,30,33,34,36,39) and 3 randomized controlled trials (3,18,40) , with a low level of heterogeneity, supporting moderate-quality evidence for earlier treatment.
A further subgroup analysis showed a reduced risk of mortality with a CD4 threshold for initiating ART of 500 cell/mm3. The impact on immune recovery was inconsistent and rated as low- to very-low-quality evidence (20,24,28). Two studies found no significant difference in the likelihood of virological suppression (<500 copies/ ml), risk of virological failure and viral rebound when treatment is initiated at higher or lower CD4 cell counts (20,36).
In the pooled analysis of two randomized controlled trials (3,18) there was low-quality evidence supporting ART initiation at higher CD4 thresholds for reducing mortality, disease progression or the combined outcome of death and/or progression and, in one trial, the risk of non-AIDS-defining illnesses.
The risk of severe adverse events did not differ significantly, but the risk of Grade 3 or 4 laboratory abnormalitiesii was increased in one randomized controlled trial (40). Since treatment in the delayed arm of the SMART trial (3) was initiated when the CD4 count fell below 250 cells/mm3 (rather than 350 cells/mm3), the quality of the evidence for clinical benefit was graded as low because of imprecision and indirectness.
A separate systematic review (41) identified one randomized clinical trial (18) and two observational studies (42,43) reporting a decreased risk of TB when individuals initiated ART with CD4 counts exceeding 350 cells/mm3. ART also reduces recurrent TB by about 50% (44). Dynamic models have suggested ART initiation above 350 cells/mm3 could lead to a more substantial reduction in population tuberculosis incidence (45).
Finally, there is high-quality evidence from one randomized controlled trial (18) indicating that earlier ART can markedly reduce the risk of sexual transmission to HIV-negative sexual partners. This is supported by the secondary outcomes of a trial that also found a 92% reduction in HIV sexual transmission from partners with HIV taking ART (46).
ii Grade 3 and 4 laboratory abnormalities are considered as severe drug adverse reactions and usually requires discontinuation of ARV drugs until the patient is stabilized and substitution for an alternative drug (See Web Annex)