 |
|
|
| |
"Historic" Resistance Profile Helps Predict PI and Overall Salvage Failure
|
| |
| |
6th European HIV Drug Resistance Workshop
March 26-28, 2008
Budapest, Hungary
Mark Mascolini
A resistance profile accounting for failure of earlier antiretroviral regimens proved a stronger predictor of salvage failure than a simple look at the latest genotype [1]. Italian investigators found class-wide nucleoside resistance in the latest genotype a strong predictor of failure. Class-wide resistance to protease inhibitors (PIs) in the historic genotype also strongly predicted failure.
A team from three Roman centers analyzed the predictive prowess of genotyping in 177 people who had at least three genotypic resistance tests from 2000 through 2006. The researchers defined class-wide resistance to nucleosides and PIs as detection of three or more major mutations conferring resistance to each class, while class-wide resistance to nonnucleosides required only one major nonnucleoside mutation. The historic genotype consisted of mutations in all three genotypes from each patient and thus would include mutations not detectable in the current genotype but still probably archived at low levels in cells. Such hidden mutations may easily reemerge during treatment with a drug that elicits that mutation. For this analysis treatment failure meant consecutive viral loads above 50 copies 6 months after the third genotype.
The study group had tried a median of four nucleosides, one nonnucleoside, and three protease inhibitors. People had taken antiretrovirals for an average 46 months (interquartile range [IQR] 35 to 56) and had a median viral load of 4.47 log (about 30,000 copies) at the time of their third genotype.
Class-wide resistance proved consistently more common in the historic genotype than in the current genotype:
· Class-wide resistance to one class: 20.3% historic versus 19.8% current
· Class-wide resistance to two classes: 37.9% versus 27.1%
· Class-wide resistance to three classes: 29.4% versus 13.0%
· Class-wide nucleoside resistance: 74.0% versus 49.7%
· Class-wide nonnucleoside resistance: 71.2% versus 39.0%
· Class-wide PI resistance: 39.0% versus 24.3%
After a median 13 months of follow-up (IQR 6 to 72), 143 people (80.8%) suffered a viral rebound. Overall failure probability at 1 year measured 69%.
Multivariate analysis adjusted for CD4 count, viral load, and AIDS diagnosis determined that class-wide resistance to all three antiretroviral classes on the latest genotype raised the risk of failure 55%, a correlation that fell just short of statistical significance (hazard ratio 1.55, 95% confidence interval [CI] 0.97 to 2.47, P = 0.06). Class-wide resistance to all three classes on the historic genotype raised the failure risk 87%, a highly significant correlation (HR 1.87, 95% CI 1.31 to 2.66, P = 0.001). But failure rates were high regardless of how many antiretroviral classes got knocked out by class-wide resistance on the historic genotype: 88% failure with triple class-wide resistance versus 63% with class-wide resistance to fewer classes.
Class-wide resistance to nucleosides in the current genotype raised the failure risk 80% (HR 1.8, 95% CI 1.2 to 2.9, P < 0.002), while class-wide resistance to PIs on the historic genotype upped the failure risk 60% (HR 1.6, 95% CI 1.1 to 2.5, P < 0.0015). CD4 count, viral load, or an AIDS diagnosis did not correlate with failure in any statistical analysis.
This study requires careful interpretation because some PIs may control virus with three major mutations, and the new nonnucleoside etravirine may stop replication of virus carrying a single classic nonnucleoside mutation. Also, newer PIs and etravirine can team up with antiretrovirals in new classes to suppress virus with wide resistance to the first three classes. But the results stress the importance of undetectable mutant populations in influencing response to a new regimen.
The findings confirm results of an earlier British Columbian study that analyzed current and historic mutations in 1734 treatment-experienced people with at least three genotypes and with a median of five genotypes [2]. The most recent genotype consistently underestimated historic resistance mutation prevalence, especially for nucleoside mutations. For example, current genotypes detected the M184V/I mutations in 25.5% of samples versus 58.8% of historic samples.
References
1. Zaccarelli M, Lorenzini P, Tozzi V, et al. Are current or archived resistance mutations predictive of treatment failure in heavily pre-treated HIV patients? 6th European HIV Drug Resistance Workshop. March 26-28, 2008. Budapest. Abstract 42.
2. Harrigan PR, Wynhoven B, Brumme ZL, et al. HIV-1 drug resistance: degree of underestimation by a cross-sectional versus a longitudinal testing approach. J Infect Dis. 2005;191:1325-1330.
|
| |
|
|
|
|
|
