The intermittent detection of viral load in patients who normally have undetectable viral loads (<50 copies/mL) has been referred to as "viral load blipping." There has been considerable concern among both patients and providers that these blips are early warning signs of treatment failure and could be an early demonstration of the development of drug resistance.

Nettles and colleagues frequently monitored the viral load of a small group of patients to determine the clinical significance of viral load blipping. The goal of this study was to characterize the frequency, amplitude and duration of blips, any patient characteristics, clinical events (infections or immunizations) or antiretroviral drug blood concentrations that might contribute to blipping.

Viral load and resistance testing were performed on blood samples from 10 patients. The blood samples were obtained every Monday, Wednesday and Friday for 3 months. Baseline and end-of-study genotypes for antiretroviral drug resistance were also obtained. The viral loads were measured using the Roche Amplicor HIV-1 Monitor ultra sensitive assay and performed in 2 different laboratories. Patients were also questioned about any significant clinical events at every visit.

Patients enrolled in the study had been on their current antiretroviral regimen for 11 to 79 months. A few of the patients had documented blips from conventional (normal time interval) visits. The total number of historical blips was 6/125 (5%).

Ninety percent of the patients experienced viral load blips, with an average of 0.6 blips per month, a median increase in viral load of 79 copies/mL. The median duration of the blip was 60 hours. The total number of blips detected during the study was 26/723 (3.6%), or an average of 2 blips every 3 months per person.

The detection of blips correlated poorly between testing laboratories. Nine blips were detected by one lab and 8 blips by another lab. Only one blip was detected by both labs.

The presence of blips did not correlate with either patient variables (age, race or gender) or clinical variables (CD4 nadir, CD4 at study baseline, time of HIV infection, number of antiretroviral medications, number of prior blips, influenza vaccination or acute infections, such as sore throats or herpes outbreaks).

Nine out of 10 patients' virus was successfully sequenced; 951 protease (mean 106/patient) and 1,079 reverse transcriptase (mean 120/patient) gene clones were analyzed for the presence of drug resistance mutations. No new significant reverse transcriptase mutations, and only 1 protease gene mutation (46I), were detected when viral load blipping was noted.

The authors conclude from their small study that almost all patients have some degree of blipping and that most blips were the result of assay variation -- in other words, the blips did not indicate treatment failure or the development of resistance. This is suggested by the fact that no new resistance mutations were found during blipping, that no patient's viral load had a sustained increase and that 2 independent laboratories doing the viral load testing had very poor correlation when blips were found.

This short-term study confirms longer-term observations (which had less-frequent viral load measurements) that blipping does not predict or predate virologic failure. This study is important in that practitioners can be somewhat more assured that they need not change antiretroviral regimens at the first sign of a low, but detectable, viral load.