0001) compared with those who were virally suppressed >90% of the time, and those who had virally rebounded in the year prior to baseline had a 3.1-times higher rate of virological failure compared with patients who had never virally rebounded (95% CI 1.84–5.25; P<.0001). The analyses were also repeated using a lower limit of detection for viral load of 50 copies/mL;

901 patients were included Crizotinib solubility dmso in the analysis and 41% experienced virological failure (defined as a viral load >50 copies/mL), with an IR of 14.3 per 100 PYFU (95% CI 12.8–15.8). Those who had virally rebounded in the year prior to baseline had an 84% higher rate of virological failure compared with patients who had never virally selleck chemicals rebounded (95% CI 1.33–2.57; P=0.0003) and patients who were virally suppressed <50% of the time they were on cART had a 13% higher rate of virological failure (95% CI 0.79–1.64; P=0.50) compared with those who were virally suppressed >90% of the time, although this was not statistically significant after adjustment. Five hundred and forty-four patients (29%) had some resistance data available at baseline. Four hundred and five patients (75%)

had a GSS ≥3 for their baseline cART regimen; there was no significant difference in rate of virological failure in patients with a GSS<3 compared with those with a GSS ≥3 (IRR 1.41; 95% CI 0.89–2.23; P=0.14) after adjustment for all demographic variables, percentage of time suppressed and time since last rebound. A patient's history of viral suppression can provide important information about the risk of viral failure after a change in ARVs. The variables describing the history of viral suppression after cART initiation but before a change in regimen were highly predictive of future virological failure, in addition to the traditional baseline predictors. The most important factors were the percentage of time spent with suppressed viral load since starting cART prior to baseline and time since last viral rebound. After adjustment for these factors, none of the other Interleukin-2 receptor markers of previous patterns of suppression

was a significant predictor of virological failure after baseline. There was a clear inverse relationship between time suppressed and risk of future virological failure. Patients with viral suppression <50% of the time prior to baseline had almost double the rate of virological failure compared with those with viral suppression >90% of the time. A study in patients with CD4 counts >200 cells/μL found that time with undetectable viraemia was a significant predictor of clinical progression [30]. In addition, previous studies have found that patients with a history of persistent low-level viraemia (51–1000 copies/mL) were more likely to experience virological failure [31], as were those with intermittent viraemia above 400 copies/mL, compared with those who sustained an undetectable viral load [32].