HIV rebound linked to higher baseline level, slower suppression and low-level viraemia

Laura Waters of Mortimer Market Centre speaking at EACS 2013. Photo by Liz Highleyman,
This article is more than 11 years old. Click here for more recent articles on this topic

Higher baseline viral load, longer time to viral suppression and low-level viraemia below the level of detection of standard tests raise the risk of viral rebound amongst people with HIV on antiretroviral therapy (ART), according to a pair of studies presented at the 14th European AIDS Conference this month in Brussels.

Only about 20 to 25% of people with HIV in Europe, the US and sub-Saharan African are on ART with full viral suppression, whilst the remainder fall out at various stages of the 'treatment cascade'. Lack of access and poor adherence are known problem areas, but reasons for loss of virological control amongst people who achieve viral suppression are not fully understood.

Laura Waters of the Mortimer Market Centre in London reported findings from a study of the effects of baseline viral load and time to viral suppression on risk of viral rebound amongst participants receiving first-line combination ART in the UK Collaborative HIV Cohort. UK CHIC regularly collects data from HIV-positive adults attending 13 centres for HIV care, including demographic information, ART treatment history, clinical events and laboratory test results.


virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

viral rebound

When a person on antiretroviral therapy (ART) has persistent, detectable levels of HIV in the blood after a period of undetectable levels. Causes of viral rebound can include drug resistance, poor adherence to an HIV treatment regimen or interrupting treatment.


The presence of virus in the blood.


first-line therapy

The regimen used when starting treatment for the first time.


Lowest of a series of measurements. For example, an individual’s CD4 nadir is their lowest ever measured CD4 count.

Several clinical trials have shown lower rates of viral suppression amongst people with high baseline viral load, but the impact of high pre-treatment viral load on longer-term virological outcomes in patients who achieve viral suppression is unknown, the researchers noted as background.

Eligible participants were selected from amongst 14,477 treatment-naive UK CHIC participants who started treatment with combination ART including a protease inhibitor or non-nucleoside reverse transcriptase inhibitor (NNRTI) after the year 2000. People who started ART within three months of HIV diagnosis, experienced AIDS-related events within three months after starting ART or lacked follow-up data were excluded, leaving 8184 individuals.

Of this group, 7475 people (91%) achieved viral suppression on their initial ART regimen and were included in the main analysis. Rates of suppression were 92% for people with viral load <10,000 or 10,000 to 100,000 copies/ml at baseline, 91% for those with 100,000 to 500,000 copies/ml and 87% for those with >500,000 copies/ml prior to treatment.

The participants in the main analysis were mostly men (80%), two thirds were white, one quarter were black and the median age was 38 years; 65% were men who have sex with men and 29% were heterosexual. The median baseline CD4 cell count was 245 cells/mm3. All started three-drug combination ART, with 76% starting on a NNRTI-based regimen and 24% starting on a protease inhibitor.

Looking at baseline viral load, 23% had <10,000 copies/ml, 42% had 10,000 to 100,000 copies/ml, 30% had 100,000 to 500,000 copies/ml and only 5% had >500,000 copies/ml.

Most participants achieved viral suppression to <50 copies/ml either within three months (42%) or three to six months (40%) after starting first-line ART. However, 15% did so between six and twelve months after starting therapy and 5% took longer than a year to reach undetectable viral load.

A total of 1289 participants (17%) experienced viral rebound -- defined as two consecutive measures >50 copies/ml – during a median follow-up period of 3.5 years. Adjusting for other factors, the probability of viral rebound was significantly higher for people who started treatment with 100,000 to 500,000 copies/ml (hazard ratio [HR] 1.34, or 34% higher than those with 10,000 to 100,000 copies/ml) and highest for those with >500,000 copies/ml (HR 1.67).

In addition, viral rebound rates also rose significantly for people who took six to twelve months to achieve suppression (HR 1.47 compared to those who took three to six months) and even more so for those who took longer than a year (HR 2.49, or nearly 2.5-fold higher risk).

Patterns for both baseline viral load and time to viral suppression were similar when looking separately at people who started NNRTIs and those who started protease inhibitors. Again, people with the two highest baseline viral load levels and those who took either six to twelve months or more than a year to achieve viral suppression were at significantly greater risk for viral rebound.

The overall median CD4 cell rise between baseline and time of viral suppression was 115 cells/mm3, but gains rose with baseline viral load: 69 cells/mm3 for those with <10,000 copies/ml, 110 cells/mm3 for those with 10,000 to 100,000 copies/ml, 153 cells/mm3 for those with 100,000 to 500,000 copies/ml and 187 cells/mm3 for those with >500,000 copies/ml.

CD4 cell gains were likewise larger for those who took longer to achieve viral suppression: 75 cells/mm3 for those who took less than three months, 135 cells/mm3 for those taking three to six months, 180 cells/mm3 for those taking six to twelve months and 190 cells/mm3 for those taking more than a year.

"Baseline viral load correlated significantly with achievement of viral suppression on initial combination ART," the researchers concluded. "Amongst individuals who have achieved viral suppression on first-line combination ART, higher baseline viral load and slower time to suppression were also associated with a higher chance of subsequent virological rebound."

Although this study was limited by lack of adherence data, Waters noted that all participants included in the main analysis managed at least good enough adherence to achieve undetectable viral load.

"[T]his study adds to the trial data that suggests we may need to investigate new strategies, with modern drugs, for individuals with high pre-treatment viral load," the investigators added. "We recommend that patients with high baseline viral load should continue to be monitored closely, even after virological suppression, as they remain at higher risk of rebound."

Anecdotally, Waters said, her group has been trying Atripla (efavirenz/tenofovir/emtricitabine) plus raltegravir (Isentress) for people starting treatment with very high viral load.

Low-level viraemia

In a related study, Andrea Calcagno of the University of Torino in Italy and colleagues evaluated whether residual low-level virus while on ART and other clinical factors raise the risk of viral rebound over two years amongst people who have achieved 'undetectable' viral load according to standard tests.

Low-level viraemia can persist for years amongst people on suppressive ART, the researchers noted as background. Treatment intensification by adding more drugs does not eliminate low-level viraemia, but some data indicate levels may be lower in people using NNRTIs, especially nevirapine (Viramune).

This retrospective analysis included 1065 people with HIV on combination ART who had at least two consecutive HIV RNA measurements <50 copies/ml between the first half of 2010 and the second half of 2012. Most (72%) were men, the mean age was 47 years and nearly 30% were co-infected with hepatitis C.

Participants had been HIV-positive for 13 years on average and had had undetectable viral loads for three years. The mean current CD4 count was 587 cells/mm3, with a mean nadir (lowest-ever) CD4 cell count of 220 cells/mm3. Equal proportions were taking protease inhibitors and NNRTIs (about 41% for each) whilst 4% were taking raltegravir; 76% also used tenofovir (Viread, also in Truvada and single-tablet regimens).

The researchers measured viral load using a sensitive assay (CAP/CTM Roche TaqMan v. 2.0) that gave results of 'target not detected' (TND), <20 copies/ml or 20 to 50 copies/ml. Viral rebound was again defined as two consecutive viral loads >50 copies/ml.

At baseline, 78, 77 and 70% of people taking raltegravir, NNRTIs and protease inhibitors, respectively, had a measurement showing 'target not detected', with the difference reaching statistical significance; 10, 19 and 20%, respectively, had <20 copies/ml, whilst 12, 10 and 10% had 20 to 50 copies/ml.

In an adjusted multivariate analysis, the only three factors that independently predicted absence of low-level viraemia ('target not detected') were female sex (adjusted odds ratio [OR] 1.69), CD4 nadir above 200 cells/mm3 (adjusted OR 1.32) and use of either raltegravir or NNRTIs (adjusted OR 1.24 compared to protease inhibitors).

Overall, 8% of participants experienced viral rebound at two years, a proportion that remained the same when looking at the 867 people who did not change their ART regimen. Rebound occurred in 7% of people without low-level viraemia, 11% with <20 copies/ml and 16% with 20 to 50 copies/ml. The duration of viral suppression was about 30 months for people who experienced rebound compared to 50 months for sustained responders.

Independent predictors of viral rebound were female sex (adjusted OR 1.79), hepatitis C co-infection (adjusted OR 1.68), viral suppression <50 copies/ml for less than two years (adjusted OR 2.11), low-level viraemia (<20 or 20 to 50 copies/ml) at baseline (adjusted OR 1.75) and using protease inhibitors (adjusted OR 2.48, or nearly 2.5-fold higher risk).

A risk score comprised of these variables – low risk if 0 to 2 factors or high risk if >2 factors – accurately predicted viral rebound, which occurred in 5% of those with a low risk score, compared to 19% of those with a high risk score.

In thisstudy, low-level viraemia predicted virological rebound at two years of follow-up, the researchers concluded.

However, they noted, the use of a lower cut-off for defining complete viral suppression is called into question by the few available interventions in case of low-level viraemia, possibly including adherence counselling or switching to a NNRTI-based regimen.


Jose S et al (Waters L presenting). The impact of baseline viral load (VL) and time to viral suppression on treatment responses to first-line combination antiretroviral therapy (cART). 14th European AIDS Conference, Brussels, abstract PS7/3, 2013. View abstract on the conference website.

Calcagno A et al. Level of HIV Viremia below 50 copies/ml predicts virologic rebound in HAART-treated patients. 14th European AIDS Conference, Brussels, abstract PS7/4, 2013. View abstract on the conference website.