Background: A viral load between 50 and 400 copies/mL in patients (pts) previously undetectable (< 50 copies/mL) on HAART may be due to laboratory factors, a transient viraemia, or the portent of imminent virological failure. We have examined whether such low level viraemia (LLV) values are repeatable on the same sample, whether there is evidence of previous prior ongoing virologic replication below 50 copies in those who subsequently fail virologically (viral load > 400 copies/mL), and virologic outcome for those who have transient viraemia.

Methods: A database of 2,885 pts who were on stable HAART for whom viral loads were routinely collected was interrogated. A cohort of those who had at least two prior values of viral load < 50 copies/mL (Versant HIV-1 RNA Assay bDNA), before one of > 50 but < 400copies/mL (LLV) and subsequently became undetectable (< 50 copies/mL) had their viral loads remeasured on the same plasma sample at LLV. Those whose repeated value was still > 50 but < 400 copies/mL were called repeatable blips (RB) and those with a repeat < 50 copies/mL were called non-repeatable (NRB). These groups were compared with a control cohort who developed > 400 copies/mL immediately after the LLV sample, called the virologic failure (VF) group. Pre-LLV samples were not remeasured, but the optical density normalized for background was calculated in at least two samples of < 50 copies/mL prior to LLV as an indirect measure of ongoing viral replication. Kruskal-Wallis, Mann-Whitney U test and Log rank c² test were used for the appropriate analyses.

Results: Of 247 blip pts, 102 pts were RB (41%) and 145 were NRB (59%); 139 were in the VF group. At the time of LLV, there was a significant difference in viral load and optical density or between the VF group and both NRB and RB groups (p < 001). There was evidence of greater ongoing viral replication at < 50 copies/mL in the VF group from samples taken at median of 5.8 and 2.8 mos before LLV and in the NRB group from 2.8 mos. This was not seen in the RB group.

Conclusion: On the same sample, 59% of all blips are non-repeatable. Low level viraemia leads to subsequent risk of virological failure. We speculate that this is a function of the degree of viral replication present at the time of transient viraemia, which is in turn predicted by the level of virus in plasma 3 to 6 mos prior to that event at quantities below the level of detection of currently available commercial assays.