The effects of transmitted drug resistance on viral load are specific to each of the major drug classes, according to data presented to the Fourteenth Conference on Retroviruses and Opportunistic Infections in Los Angeles on Monday. Compared to recent seroconverters with wild-type virus, those with PI- and NRTI-associated resistance had consistently lower viral loads over time, and those with NNRTI-associated resistance had significantly higher viral loads over time.
Prevalence of transmitted drug resistance
The prevalence of transmitted drug resistance (TDR) , also known as primary resistance, varies within countries and populations. The most recent data from the United Kingdom suggest that prevalence of transmitted drug resistance amongst antiretroviral-naive individuals declined from a peak of 16% in 2002 to 9% in 2004.
Unpublished studies presented in Monday’s plenary by Dr Susan Little of the University of California, San Diego suggest that in primary infection, prevalence of TDR ranges from 11% to 15% among newly infected treatment-naive individuals. On the other hand, studies of newly diagnosed treatment-naive individuals (with HIV infection of an unknown duration) suggest prevalence of TDR ranges from 7% to 11%. Taken together, these studies suggest that TDR becomes less readily detectable over time.
However, Dr Little pointed out that all of these studies are based on standard genotypic resistance testing which can only measure drug-resistant HIV that makes up 20% or more of the entire population. Consequently, she said that these studies “may underestimate the true prevalence of transmitted resistance” and that more sensitive assays may “double the estimate.”
Does TDR affect the natural history of HIV disease progression?
Dr Little and her colleagues then sought to ascertain whether there were any differences in the viral load set point between untreated recently-infected individuals with and without TDR.
She presented data from the DACSOO3 study, which included individuals enrolled in the Acute Infection and Early Disease Research Program (AIEDRP) within a year of HIV antibody seroconversion. A total of 1,555 individuals were included in the programme, the majority of whom were white gay men. At baseline, median CD4 counts were 548 cells/mm3 and viral load was 4.8 logs. Of the 913 individuals with genotype testing results at baseline, 93 (10.2%) had TDR.
The individuals with TDR did not appear to have a significantly different viral load at baseline compared with the group infected with wild-type virus. However, when the TDR group was stratified by resistance to a specific drug class, a different picture emerged.
Individuals with NNRTI-resistant virus had a baseline viral load that was 0.4 logs higher than those with wild-type virus (p = 0.003). Those with NRTI- resistant virus had a baseline viral load that was 0.7 logs lower than those with wild-type virus (p
Dr Little told the conference that some of the participants had been followed-up for more than three years, and that the differences seen at baseline continued in those with NNRTI- or NRTI-resistant viruses.
She also noted that these differences occurred independently, and that if someone had both NNRTI-resistant and NRTI-resistant virus, they could, in effect, “cancel each other out”.
Persistence and fitness of TDR
In addition, it has been suggested that the replication capacity of transmitted drug-resistant viruses may be lower than wild-type.
Dr Little presented data suggesting otherwise from a small study based on fourteen antiretroviral-naive, newly infected individuals who were identified an average of 66 days after their estimated date of infection, and followed for a median of two years. Nine had HIV that was resistant to a single drug class; three were dual-class resistant; and two had triple-class multidrug resistant (MDR) HIV.
Using standard population-based genotypic testing, Dr Little and her colleagues measured the average time to the first appearance of wild-type HIV (i.e. the time that it took for at least 20% of the TDR virus population to mutate back to wild-type): this was one week shy of two years – although this ranged from just under a year to over four years.
They then measured the average time to the last appearance of drug-resistant HIV (i.e. the time that it took until TDR became less than 20% of the viral population). Although TDR ‘disappeared’ in one of the fourteen, it persisted in the other thirteen for an average of 2.7 years. Dr Little noted, however, that 2.7 years was a “conservative estimate” of TDR persistence and that a “less conservative estimate would suggest that lifelong persistence of resistance is a possibility.”
The mean replication capacity of the transmitted drug-resistant HIV in these fourteen individuals was 87% (range 16-208%) compared with wild-type. When this was compared with the replication capacity of viruses from another 33 individuals with TDR, and with drug class-specific or wild-type viruses, no significant difference was seen.
However, a significant difference was seen in individuals with TDR virus that had protease inhibitor (PI) and nucleoside (NRTI) associated resistance, but notnon-nucleoside (NNRTI) resistance (50% versus 92%; P
Dr Little noted that, overall, reversion of resistance is gradual and usually incomplete, resulting in the persistence of a mixture of wild-type and drug-resistant strains for many years after initial infection. This, she said, supports the routine use of drug resistance testing in antiretroviral-naive patients.
Dr Little concluded by saying that the pathogenic potential of TDR appears to be “very antiretroviral class-specific,” with NNRTI-resistance associated with high viral load, and NRTI- and PI-resistance associated with lower viral load.
She added that slower treatment responses for individuals with transmitted PI-resistance and NRTI-resistance at codon 215 may be associated with treatment failure.
Her take home message, however, was that the prevalence and persistence of TDR strongly supports the routine use of HIV resistance genotyping for all newly diagnosed HIV-positive individuals, but that cross-resistance may limit the potency of available first-line regimens in those with MDR TDR even when guided by prospective primary resistance testing.
How does this translate to treatment response?
Given the results of the above study, Dr Little and her colleagues hypothesised that it should take longer for individuals with NNRTI-resistant virus to reach ‘undetectable’ levels than those with PI- or NRTI-resistant virus, once they started treatment.
Dr Little presented data from the DACS002 study which included 793 individuals from the AIEDRP, 84 of whom had TDR, and who began antiretroviral therapy a median of seven months after their estimated date of infection.
Median baseline viral load in the group as a whole was 5.17 logs. However, the individuals with TDR had significantly lower viral loads than those with wild type virus (4.97 vs. 5.19 logs; p=0.02).
Of the 84 with TDR, 67 had NNRTI-resistance (15% of whom also had PI-resistance and 21% of whom also had NRTI-resistance). All but one of the 25 individuals with NRTI-resistant virus also had resistance to other drug classes: 56% to NNRTIs and 40% to PIs. All of the 18 individuals with PI-resistance virus had resistance to another drug class: 56% to NNRTI and 56% to NRTIs (and, consequently, at least two individuals had triple class resistance).
However, in this study, most participants received more than three antiretrovirals. Consequently, the average number of active drugs across the whole cohort was 3.61.
Dr Little notes that this was probably the reason why this study saw no significant difference in the time to complete suppression between those with TDR and those with wild-type virus (p = 0.069). She added that of the 67 individuals with NNRTI-resistant virus, “only 15% received an NNRTI-based regimen, which may explain why no difference was seen.”
However, the study did suggest a possible delayed treatment response among participants with PI-resistance and with NRTI-resistance at codon 215, although this may have been due to the fact that the majority of individuals with these mutations had fewer active drugs than most people in the study.
During questions from the audience Dr Little noted that she observed no significant effect on CD4 counts, nor any difference in clinical outcome, between those with MDR HIV and those with wild type virus.
Little S. Transmission of HIV drug resistance and treatment response. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 60, 2007.