This Study Suggests Viral Load Blips Matter
     Reported by Jules Levin

Study Finds That The Risk of Viral Failure (>500 copies/ml) Increases with Repeated Viral Load Values Between 51-500 (after repeatedly being below 50 copies/ml), and Risk Increases When Last Value Was 300-500

(This is the full text of the article Study Suggests Potent HAART & Intensification May Prevent Viral Rebound)

At the Resistance Workshop in the Summer of 2000, Diane Havlir from UCSD reported data on the likelihood of viral load "blips" leading viral failure. In her study, treatment-naïve patients received indinavir/3TC/AZT and had well suppressed viral load (<50 copies/ml). The patients were followed for 84 weeks. Viral load blips were defined as >50 copies/ml and failure was defined as 2 tests >200 copies/ml. In essence, she found that having a couple of viral load tests of 75 copies/ml did not lead to viral load failure. It may be relevant to bear in mind that the PI used in this study was indinavir which has a relatively higher barrier for failure. In other words, you need a number of mutations to fail indinavir. There were 241 patients in this study.

Havlir concluded that in this study population, where they saw 343 intermittent blips, the blips did not lead to viral rebound within the time frame they looked at (84 weeks): 9/96 (9.3%) of patients with intermittent viremia had viral rebound; 20/145 (13.8%) with viral suppression (without blips) had viral rebound. Baseline viral load was the only predictor of virologic failure in the model they used. In other words, the higher a patient's viral load was at baseline the more likely they were to fail therapy. Since follow-up was only 84 weeks, it's possible that if followed longer viral failure may have occurred more often in individuals with blips >50 copies/ml. She also reported on a small group of patients followed for 4.5 years who were also on indinavir/3TC/AZT, and 0/6 patients with intermittent viremia had viral rebound, and 0/7 patients with viral suppression had viral rebound.

The New Study

At the recent Retrovirus Conference, Greub and colleagues reported on a large group of patients with viral load "blips" of up to 500 copies/ml, which is higher than the "blips" in the Havlir study. In this study patients patients had <50 copies/ml. In essence, Greub found that 2 or more consecutive blips of viral load of 100-500 could lead to viral load rebound over 500 copies/ml. Importantly, Greub found that patients who changed therapy after the third viral load blip were more likely (2.65 times) to stay <50 copies/ml than people who did not change therapy. Greub also found that people who had ART experience prior to their current HAART regimen were more likely to experience viral load rebound following 3 consecutive blips (100-500), but this was not statistically significant.

Grueb described the purpose of his study was to determine the frequency of low level rebound in clinical practice, and to evaluate the virologic outcome of single or repeated viral load tests of 51-500 according to whether therapy was modified.

Over 2,000 patients were included in this study (1,063 from the Frankfurt HIV Clinic Cohort, and 1,481 from the Swiss HIV Cohort. HIV viral load was measured with the Roche Amplicor assay with a lower limit if detection of 50 copies/ml. Patients had to have two consecutive VL's <50 copies/ml and at least 2 additional test results. Follow-up was 68 weeks in the Frankfurt group and 43 weeks in the Swiss group.

Patients With 1 Low-Level Viral Load Rebound

41% (n=1054) experienced at least 1 low level rebound (51-500). Most of these patients (about 70%-85%) went back down to <50 copies/ml. 20%-30% experienced viral load failure (>500 copies/ml), regardless of whether the low level rebound was 100-300 or 301-500. Of course, it's important to bear in mind that I don't think Greub talked about adherence, so you don't know if viral load rebound was due to non-adherence.

The More Viral Load Blips A Patient Experienced To 101-500 copies/ml, the More Likely They were To Experience Viral Load Failure To >500 copies/ml: patients with 2 viral load values 51-500 increased likelihood of viral failure

323 patients had 2 consecutive values of 51-500. In this group less patients went back down to <50 copies/ml. I'm not sure if both values were 100-300 or 301-500, or if just the 2nd value was 100-300 or 301-500. But when 2 consecutive tests were 51-500, only 40%-70% went back down to <50 copies/ml. The higher the viral load went to the more likely viral failure was to occur. For example, if viral load values were 51-100 (n=118), 70% went back down to <50 copies, and about 20% experienced viral load failure to >500 copies/ml. Among patients with viral load blips to 101-300 (n=118), about 45% went back down to <50 copies/ml, and about 45% experienced viral load failure to >500 copies/ml. For patients whose viral load blipped to 301-500 (n=17) about 53% went back down to <50 copies/ml, and about 30% experienced viral load failure.

127 Patients Experienced 3 Consecutive Values 51-500, and They Were More Likely To Experience Viral Load Failure Than Patients Experiencing 2 Values Between 51-500

The likelihood of going back down to <50 copies/ml was less in this group experiencing 3 low-level rebounds than in individuals with 2 consecutive values 51-500. When values were 301-500, 55% experienced viral load failure to >500 copies/ml. When viral load values were 101-300 40% experienced viral load failure. And, 5% to 20% of these patients had their viral load remain between 51-500, they did not go back down to <50 copies/ml.

Odds ratio of Reducing Viral Load to <50 copies/ml After 3 Consecutive Viral loads 51-500 (using a logistic regression model)

If the third value was 301-500, patients had an 80% less chance of getting back to <50 copies/ml (p=0.004), compared to if the third value was 51-100. If the third value was 101-300, patients had a 60% (p=0.04) reduced chance of going back down to <50 copies/ml.

Changing Therapy Helps Reduce VL to <50

Patients who modified therapy after the third value of 51-500 were 2.65 times more likely to go back down to <50 copies/ml than patients who did not change therapy (p=0.04). If a patient had ART experience prior to current HAART that reduced their chance by 50% of going back down to <50 copies/ml after experiencing 3 values of 51-500, but this was not statistically significant (p=0.11).

The Risk of Failure After Going Back Down To <50 Following 51-500 Viral Load

In patients who had a viral load of 51-500 but then went back down to <50 they were about 2 times more likely to experience failure (>500) compared to patients with consistent viral loads <50. But, patients who had sustained viral loads 51-500 who were not able to get back down to <50 copies were 5 times more likely to experience viral load failure. In other words, patients who had viral loads 51-500 after having viral load <50 were 5 times more likely to see viral load rebound to >500.

Greub concluded the majority of low level viral rebound resolve spontaneously. But, spontaneous resolution is associated with a modest increase in risk for subsequent failure. The risk of viral failure increases with repeated observations of values 51-500, and with the level of the most recent viral load within this range. Grueb did not reveal what the various regimens were that patients were taking, as opposed to the Havlir study in which everyone was receiving an indinavir regiman. One other qualification is that you don't know if blips were due to laboratory errors, but having 2-3 consecutive blips reduces the odds that it was due to lab errors. However, since blood samples were bundled it's less likely lab error was involved.