Long-term use of antiretroviral therapy (ART) in HIV-positive persons may be challenged by the need for high-level adherence, development of drug resistance, toxicities, and cost. Treatment strategies conferring durable virological control, whilst minimising ART exposure are highly desirable. With this goal in mind, strategic interruption of ART was the focus of several studies –.
However, interruption of ART is no longer a recommended strategy  and the level of HIV plasma viral load (pVL) following ART stop has been shown to reach levels comparable to pre-treatment values –, increasing onward transmission risk .
Inaccessible reservoirs of latently-infected resting memory CD4 T-cells are hypothesised to be the major source contributing to viraemia rebound after stopping ART , .
Recent research has shown the dramatic effect of ART to prevent onward viral transmission , and mathematical models predict that it may potentially be possible to eliminate HIV infection at a population level with universal treatment coverage for all HIV-positive individuals, irrespective of CD4 count . However, although not recommended, consideration of the potential impact of individuals choosing to stop ART could be considerable, and data are needed on subsequent viral rebound to better inform future transmission models. Furthermore, final results from SPARTAC suggested that ART initiated in primary HIV infection (PHI) was associated with a change in pVL set-point out to 60 weeks after stopping therapy  whilst the SMART trial reported that interruption of ART in chronic infection (CHI) was associated with an increased risk of all-cause mortality
The level of viral rebound following interruption of ART commenced in at different stages of HIV infection is, therefore, highly relevant from both a clinical and public health perspective and warrants further investigation.
We, therefore, wanted to compare the pVL changes observed after cessation of ART initiated in chronic HIV infection with those in PHI by comparing viral rebound between individuals enrolled in two protocol-indicated ART interruption studies; SPARTAC and SMART.
The magnitude of HIV viral rebound following ART cessation has consequences for clinical outcome and onward transmission. We compared plasma viral load (pVL) rebound after stopping ART initiated in primary (PHI) and chronic HIV infection (CHI).
Two populations with protocol-indicated ART cessation from SPARTAC (PHI, n = 182) and SMART (CHI, n = 1450) trials.
Time for pVL to reach pre-ART levels after stopping ART was assessed in PHI using survival analysis. Differences in pVL between PHI and CHI populations 4 weeks after stopping ART were examined using linear and logistic regression. Differences in pVL slopes up to 48 weeks were examined using linear mixed models and viral burden was estimated through a time-averaged area-under-pVL curve. CHI participants were categorised by nadir CD4 at ART stop.
Of 171 PHI participants, 71 (41.5%) rebounded to pre-ART pVL levels, at a median of 50 (95% CI 48–51) weeks after stopping ART. Four weeks after stopping treatment, although the proportion with pVL≥400 copies/ml was similar (78% PHI versus 79% CHI), levels were 0.45 (95% CI 0.26–0.64) log10 copies/ml lower for PHI versus CHI, and remained lower up to 48 weeks. Lower CD4 nadir in CHI was associated with higher pVL after ART stop. Rebound for CHI participants with CD4 nadir >500 cells/mm3 was comparable to that experienced by PHI participants.
Stopping ART initiated in PHI and CHI was associated with viral rebound to levels conferring increased transmission risk, although the level of rebound was significantly lower and sustained in PHI compared to CHI.