UK HIV Drug Resistance database: background and recent outputs.
The UK HIV drug resistance database was established in 2001 as a central repository of resistance tests carried out as part of routine clinical care in the United Kingdom. Sequence data from laboratories that perform the vast majority of genotypic resistance tests in the UK are submitted to the database on a yearly basis and linkage with a clinical dataset in the UK CHIC cohort allows extensive analyses.
Over time there has been an increase in the number of resistance tests performed each year in the UK and in 2006 there were approximately 40,000 sequences with more than half of them linked to very detailed treatment histories. Selected recent analyses from the database are described.
TIME TRENDS IN PI RESISTANCE IN TREATMENT-EXPERIENCED PATIENTS
This analysis addressed the prevalence of PI resistance among treated patients, where resistance was defined as the presence of one or more major mutations from the IAS-USA list (2006). Figure 1 shows the calendar time trends of drug resistance in treatment-experienced patients complete to the end of 2005. Resistance to any drug class peaked at approximately 80% in 1999-2000 and it has declined ever since. In 2005, only 50% of treatment-experienced patients had resistance mutations. This general downward trend was mirrored in the trend in NRTI resistance and to a lesser extent in NNRTI resistance. However, the biggest proportionate change was seen in the prevalence of PI resistance, which halved in prevalence during the study period. By linking the resistance data to the patients' drug histories, there was indirect evidence linking this decrease to changes in prescribing patterns.
[FIGURE 1 OMITTED]
Figure 2 shows regimens received at the time of the resistance test. As would be expected, there has been a decline in the proportion of patients receiving unboosted PIs and an increase in the proportion of those taking boosted PI regimens. When looking at the prevalence of PI resistance with respect to current therapy, the highest rates of resistance were found in those taking an unboosted PI. However, between 2001 and 2005, this fell from 60% to 30%. To differentiate between the possible reasons for this change, the analysis was taken further and patients were classified not just by current therapy but rather by lifetime exposure to PIs (Figure 3).
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Figure 4 shows that the highest rate of PI resistance is seen among patients who have received both boosted and unboosted PIs. This is unsurprising as these patients will have had the longest and most complex exposure to PIs. Also consistent with expectations is the much higher rate of resistance in those patients who have received unboosted PIs in the past compared with patients who have only received boosted PIs. Furthermore, the rate of resistance in patients who have received only boosted PIs also declined over time, possibly reflecting the recent introduction of more potent PIs.
In summary, PI resistance is becoming less common among treatment-experienced patients who undergo resistance testing. There is indirect evidence that the decline in PI resistance may be due to increased use of boosted PIs. There are however limitations to this type of surveillance approach and detailed analyses of patterns of resistance linked to drug histories are required, together with a prospective approach at monitoring the emergence of resistance in clinical practice among patients currently starting HAART.
RATE OF DEVELOPMENT OF RESISTANCE AMONG RECENTLY TREATED PATIENTS
The background to this analysis was that much of the currently detectable resistance was induced by mono- and dual-therapy regimens prior to 1997. One of the key questions today in terms of long-term prognosis is the rate of development of resistance for patients who are starting with a HAART-type regimen.
The study used a prospective approach. From the database, just under 4500 treatment-naive patients were identified who started first-line therapy with three or more drugs and for whom there was a median follow-up of 2.9 years. Virological failure was defined as two consecutive plasma viral load values above 1000 copies/ml or a single value above 1000 copies/ml followed by the initiation of at least one new drug. It was found that after 5 years, roughly 35% of patients had experienced virological failure and at least a quarter of these had one major resistance mutation.
It was also found that the lowest rate of resistance was observed in patients prescribed a boosted PI in combination with two nucleosides. Among these, the probability of developing a PI mutation was roughly one-third of the risk of developing a non-nucleoside resistance mutation among patients receiving NNRTI-based HAART. In addition, there was also a reduced risk of developing resistance to any of the nucleosides in the backbone. These results do give some insight into the robustness of regimens against the development of resistance.
Correspondence to: Dr David Dunn, MRC Clinical Trials Unit, London, UK. Email: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Dunn, David; Pillay, Deenan|
|Publication:||Journal of HIV Therapy|
|Date:||Dec 1, 2007|
|Previous Article:||Antiretroviral resistance in the developing world.|
|Next Article:||Antiretroviral drug resistance in HIV-infected children: implications for long-term treatment.|