HIV detectable in CSF of 10% of patients taking successful antiretroviral treatment

Ten per cent of patients taking antiretroviral therapy who have an undetectable viral load in their blood have detectable levels of HIV in their cerebrospinal fluids, Swedish and US investigators report in the online edition of the Journal of Infectious Diseases.

Detectable viral load – or viral escape – in cerebrospinal fluid (CSF) was associated with evidence of immune activation in the brain, longer duration of antiretroviral therapy and treatment interruptions.

“Previous studies have shown that CSF HIV-1 RNA generally responds well to antiretroviral therapy,” comment the authors. However, they add, “our findings…suggest that viral escape in CSF, even in subjects with successful systemic treatment with contemporary regimens, is a more common occurrence than previously reported.”

The author of an editorial that accompanies the paper “congratulated” the authors on their research, and writes: “It is concerning that this ongoing viral presence represents a serious threat to brain injury over the many years that patients are expected to live.”

Thanks to antiretroviral therapy, the prognosis of many HIV-positive patients is excellent. This treatment suppresses HIV replication in the blood to very low levels allowing the immune system to recover and fight infections. Successful therapy also leads to reductions in localised and systemic inflammation.

However, not all anti-HIV drugs cross the blood-brain barrier or blood-CSF barrier, and few studies have looked at the presence of a detectable viral load in CSF in patients who are taking HIV therapy and have an undetectable viral load in their blood.

Therefore investigators in Gothenburg, Sweden, and San Francisco, California, designed a cross-sectional, or snapshot, study involving 67 patients. All the patients were taking modern anti-HIV drugs and had had an undetectable viral load in their blood for at least six months. None had any neurological symptoms. The patients were recruited between 2002 and 2010.

A total of seven (10%) of individuals had detectable HIV in their CSF. The median CSF viral load was 121 copies/ml (range 52 to 860 copies/ml).

Detectable CSF was associated with increased inflammation in the brain. Levels of CSF neopterin (a marker of inflammation and macrophage activation) were significantly higher in patients with detectable CSF viral load than those with undetectable CSF viral load (9.2 vs 5.1 nmol/ml, p = 0.03).

In addition, 71% of individuals with detectable levels of HIV in their CSF had neopterin levels above the ”normal” value (5.8 nmol/l) compared to 40% of patients with undetectable CSF viral load.

Blood viral load did not differ significantly between the patients with detectable and undetectable CSF viral load.

Longer duration of HIV treatment had a significant association with detectable viral load in CSF. Individuals with detectable HIV in their CSF had been taking treatment for a median of 77 months, compared to a median of 35 months (p = 0.002) for individuals with undetectable CSF load.

In addition, detectable CSF viral load was associated with a greater number of viral load “blips” – transient increases in blood viral load above 50 copies/ml. Those with detectable HIV in their CSF had a median of 2.5 blips, compared to a median of zero for those with fully suppressed CSF viral load (p < 0.01). The investigators suggest that this could be related to adherence.

Treatment interruptions were also more common among those with detectable CSF virus (p < 0.01). The investigators believe that such interruptions were leading to the “intermittent reseeding” of viral load in the brain and CSF.

There was no evidence that any anti-HIV drug was associated with detectable CSF load. Nor was penetration of drugs into CSF and the brain significant.

“We demonstrate that 10% of subjects had CSF HIV-1 RNA >50 copies/ml,” comment the investigators.

They add, “subjects with detectable CSF virus had significantly longer exposure to ART and higher levels of intrathecal immune activation; treatment interruptions were also more common in these subjects.” 

The investigators are uncertain of the clinical significance of their findings.

None of the patients with detectable CSF viral load had neurological symptoms, “suggesting that viral escape in CSF may, at least in the short term, be clinically benign or silent in treated individuals”.

Aware that their study had a cross-sectional design and small sample size, the researchers conclude by calling for larger, longitudinal studies to examine the issues raised by their research.

This call is echoed by the author of the accompanying editorial, who was especially concerned that ongoing HIV replication in the brain could have an impact on the neurocognitive performance of HIV-infected individuals as they age.