YOU ARE HERE:
No Nukes?
| Last updated: 27.10.04 |
Hopes that triple nucleoside therapy could provide a simple and effective means of treating HIV have been severely quashed by the findings of several recent studies. In particular, it has been established that once-daily triple nucleoside regimens including tenofovir and abacavir are suboptimal, and switching to abacavir/AZT/3TC (Trizivir) is associated with an increased risk of viral rebound.
Triple nucleoside therapy involves the use of only nucleoside and nucleotide reverse transcriptase inhibitors (NRTIs). This single class approach has been promoted as a means of keeping protease inhibitors (PIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs) in reserve for later treatment.
However, current British and US antiretroviral guidelines do not endorse triple nucleoside regimens as first line treatment. According to guidelines published by the US Department of Health and Human Services (DHHS), Trizivir and abacavir/d4T/3TC “should ONLY be used when an NNRTI-based or a PI-based regimen cannot or should not be used as initial therapy (e.g. for important drug-drug interactions)”.1 The British HIV Association (BHIVA) guidelines make a similar recommendation.2
The guidelines take an even tougher stance on tenofovir/abacavir/3TC and tenofovir/ddI/3TC. Following very poor efficacy in several studies, drug regulatory authorities in Europe and the US and pharmaceutical companies Gilead Sciences and GlaxoSmithKline have warned against the use of these two triple nucleoside regimens in any circumstances.
Initial therapy: Trizivir
The recommendation against the use of Trizivir (AZT/3TC/abacavir) was made following the interim analysis of the ACTG 5095 study, which showed that Trizivir alone was less effective than the efavirenz-containing combinations it was compared with in the study.
ACTG 5095 was a randomised, double-blind study comparing Trizivir, efavirenz/AZT/3TC and efavirenz/Trizivir. After an average of 32 weeks treatment, 79% of the triple nucleoside recipients had a viral load below 200 copies/ml compared with 90% of people taking efavirenz-containing HAART. After 48 weeks, the efavirenz arms of the study were performing significantly better than the triple nucleoside (89% versus 74% with viral load below 200 copies/ml).3
A retrospective, non-randomised study of people who started efavirenz/AZT/3TC or abacavir/AZT/3TC confirmed this finding. The average time to treatment failure was 441 days in the triple nucleoside group and 1222 days in the efavirenz group.4
However, ACTG 5095 did not verify earlier findings that triple nucleoside therapy was less effective in people with baseline viral loads above 100,000 copies/ml.5 Instead, Trizivir was found to be less effective than recommended regimens regardless of baseline viral load.
In contrast to ACTG 5095, several earlier studies had suggested that Trizivir was roughly equivalent to indinavir/AZT/3TC and nelfinavir/AZT/3TC.5,6 However, these combinations are no longer the yardstick of antiviral efficacy. Response rates of 40-58% (viral load below 50 copies/ml at 48 weeks), as achieved in these studies, would now be regarded as below par. Recommended combinations with efavirenz or lopinavir/ritonavir achieve undetectable viral load in up to 90% of participants in clinical trials.
Although the triple nucleoside arm of ACTG 5095 was halted because it was clearly less effective than the efavirenz-based combinations, it still performed reasonably well with viral suppression in three quarters of patients at one year.7 However, given the evidence that first line Trizivir is inferior to other combinations, both BHIVA and the US DHHS recommend that Trizivir should not be considered for first line therapy unless specific circumstances mean that more potent regimens cannot be taken.
Initial therapy: tenofovir/abacavir
Tenofovir/abacavir/3TC has also proved to be an inferior first line combination. The ESS30009 study compared once-daily tenofovir/abacavir/3TC with efavirenz/abacavir/3TC in treatment-naïve patients. Data on 194 people who completed at least eight weeks of treatment showed that 49% of those randomised to tenofovir/abacavir/3TC were classified as virologic failures, compared to 5% in the efavirenz/abacavir/3TC (p<0.001). Virologic failure was defined as a viral load reduction of less than 2 logs or an increase of 1 log following an initial drop in viral load.7
ESS30009 corroborated the findings of a pilot study of 19 people who took tenofovir/abacavir/3TC. A high rate of non response or treatment failure was found, with only eight achieving a viral load reduction of 2 logs out to 16 weeks.8
Tenofovir/abacavir/ddI has also performed badly in a small pilot study of 24 treatment-naive people. Twenty-two patients (91%) failed to achieve a 2 log reduction in viral load after 12 weeks of treatment. Of 21 people who underwent resistance testing, 95% had the M184I/V mutation and 50% had the K65R mutation.9
As a result, the US DHHS have added tenofovir/abacavir/ddI and tenofovir/abacavir/3TC to the list of combinations which should not be offered at any time.
Not fully understood
The reasons these combinations have performed so poorly are not fully understood. Dr Martin Fisher of Brighton and Sussex University Hospitals told ATU that previously undescribed drug interactions, once-daily dosing, common resistance pathways, and inherent lack of potency may all have contributed to lower-than-expected efficacy.
“None of these individually stand up to scrutiny,” Dr Fisher said. “Initial data show no apparent unexpected interactions, the Zodiac study10
suggests that abacavir/3TC can be given once daily (with efavirenz), since not all individuals with virological failure demonstrate the K65R [mutation] common resistance pathways cannot explain all, and inherent lack of potency seems unlikely if one considers the individual drug effects on viral load when used as monotherapy. It may be that there are components of each that add up to explain the findings.”
Dr Joel Gallant, one of the authors of the ESS30009 study, has discussed the reasons for the poor performance of tenofovir/abacavir/3TC on the Clinical Care Options website. Although he admits that there is a possible role for dosing and drug interactions, his focus is on the issue of resistance. Genotypic data on 36 people in the ESS30009 study found all had the M184V mutation associated with resistance to 3TC and abacavir, and 23 (64%) had the K65R mutation associated with resistance to tenofovir and abacavir. The overlapping pathways to resistance of the three drugs in this combination may be a major factor in its poor clinical efficacy.11
However, some people have failed the tenofovir/abacavir combinations without detectable resistance, suggesting that other factors are at play. In the pilot study of tenofovir/abacavir/3TC, only one third of people with treatment failure had the K65R mutation but nine had a baseline viral load over 100,000 copies/ml. This points to a possible lack of potency with the once-daily tenofovir/abacavir combinations.
Initial therapy - other triple nuke regimens
Two randomised studies - CLASS and Atlantic - have compared d4T-containing triple nucleoside regimens with PI and NNRTI-based regimens. The CLASS study found that the efavirenz regimen was superior to ritonavir-boosted amprenavir and d4T regimens when combined with abacavir/3TC after a year.12 In the Atlantic study, the d4T-containing regimen was inferior to nevirapine and nelfinavir when combined with ddI/3TC after two years of therapy.13
A Danish team has compared three first line regimens in 180 people. Once again, the triple nucleoside regimen performed badly: 43% of people taking d4T/abacavir/ddI had viral load below 20 copies/ml after 48 weeks compared with 69% of people on nevirapine/nelfinavir/AZT/3TC and 62% of the ritonavir/saquinavir/AZT/3TC group.14
These studies point to the inferiority of d4T-containing triple nucleoside regimens compared to recommended highly active antiretroviral therapy (HAART), which may be due in part to the role of d4T-associated side effects in treatment failure.
In the UK, d4T is no longer recommended as first line therapy because of its association with long-term toxicities, including lipoatrophy. Furthermore, US guidelines now recommend against the d4T/ddI nucleoside ‘backbone’.
Triple nukes for switch-maintenance?
Triple nucleoside regimens are sometimes used to simplify dosing or to reduce toxicities in people who are on suppressive NNRTI- or PI-based HAART. This approach has been dubbed ‘switch-maintenance’ therapy.
Several studies have looked at the use of triple nucleoside regimens in this context.
An early randomised study reported that replacing a PI with abacavir was associated with sustained viral suppression and improved lipids.15 However subsequent studies have not been as encouraging, suggesting that abacavir is not the best switch option.
A Spanish study called NEFA was published in the New England Journal of Medicine in September 2003.16 Four hundred and sixty people who had a viral load suppressed to below 200 copies/ml for at least six months were randomised to replace their PI with either efavirenz, nevirapine, or abacavir. After twelve months follow-up abacavir-treated patients showed a trend towards greater likelihood of viral rebound. However, rebound was most common among people who had previously taken suboptimal dual or single NRTI therapy, as found in a previous Trizivir ‘switch-maintenance’ study.17 This indicates that people with underlying NRTI resistance may be most vulnerable to treatment failure when switching to a triple nucleoside regimen.
A meta-analysis of nine randomised studies confirmed that a switch to abacavir is associated with increased risk of viral rebound, although the extent to which prior suboptimal nucleoside treatment contributed to this higher failure rate is not clear.18
Another important switching study, known as Trizal, investigated a switch from PI- or NNRTI-based therapy to Trizivir. Sponsored by GlaxoSmithKline, the marketers of Trizivir, and conducted in nine European countries, 219 people with viral load below 400 copies/ml with an average CD4 count of approximately 500 were randomised to continue their PI or NNRTI, or to switch to Trizivir. There was no difference in the rate of treatment failure between the two groups (22% of each group stopped treatment due to viral rebound or adverse events).19
Some benefits
Despite the evidence of poorer virological outcomes with triple NRTI therapy, studies have found some benefits associated with the Trizivir switch. In the NEFA study, people taking abacavir were less likely to report side effects and were less likely to stop treatment due to toxicity. Another positive feature of the triple nucleoside regimen was the improvement in blood lipids after the switch to abacavir. Improvements in cholesterol (but not triglycerides) were also reported among switchers in the Trizal study.
Given that Trizivir involves simply taking one pill twice daily, it is not surprising that some Trizivir ‘switch-maintenance’ studies have reported boosted patient adherence to medication.
The benefits of switching to Trizivir, in terms of improved adherence and lower cholesterol, point to the potential usefulness of Trizivir for people struggling with complex dosing or high pill burden, and those who develop elevated lipids.
Despite inferior efficacy, there remains a role for Trizivir in treatment-experienced people in some contexts. For example, for a person with resistance to the NNRTIs who is experiencing high lipids on PI-based therapy may consider a switch to Trizivir. Alternatively, a person commencing treatment for tuberculosis may elect to take Trizivir for the duration of their TB therapy to minimise the risk of drug interactions.
However, the potential role of Trizivir should not be extrapolated to all triple nucleoside regimens. For instance, data shows that replacing a suppressive HAART regimen with tenofovir/abacavir/3TC is a risky strategy. A small pilot study from the Netherlands found that five of eight people who made such a switch experienced virological failure after an average of 19 weeks. None had previously experienced virological failure, yet four of the five had developed key NRTI resistance mutations at positions 194 and 65.20 As noted above, US guidelines recommend against the use of tenofovir/abacavir/3TC and tenofovir/abacavir/ddI under any circumstances.
A future for triple nukes?
Leading UK HIV clinician, Dr. Martin Fisher acknowledge an ongoing role for Trizivir in patients experiencing adherence difficulties, side effects, or drug interactions due to use of other medications. In particular, the use of Trizivir may become more attractive when doctors begin screening patients for abacavir hypersensitivity using a genetic test that is currently available but not widely used.
The poor results with tenofovir/abacavir/3TC highlight the importance of treatment decisions that are informed by clinical trial evidence. “I think this serves as a reminder that we still live in an era of uncertainty and practice should still be data driven,” comments Dr. Fisher. “These unorthodox triple nuke combinations (TDF/ABC/3TC and ddI/3TC/TDF) should be avoided completely.”
Although triple nucleoside therapy took a battering in 2003, there is still hope that four-drug nucleoside/nucleotide treatment may be effective. So far, ongoing UK and US studies involving tenofovir/Trizivir have not been discontinued, suggesting no early evidence of poor performance.
Summary
Tenofovir/ddI/3TC: should not be used at any time due to extremely poor efficacy.
Tenofovir/abacavir/3TC: should not be used at any time due to extremely poor efficacy.
Trizivir (AZT/3TC/abacavir): not recommended as first line therapy because less effective than recommended first line regimens. May be used where alternative regimens will cause more severe toxicities, drug interactions, or regimen complexity.
References
1.http://www.aidsinfo.nih.gov/guidelines/default_db2.asp?id=50
2.http://www.bhiva.org/guidelines/2003/hiv/index.html
3.Gulick RM et al. Second IAS Conference on HIV Pathogenesis and Treatment, Paris, abs 41, 2003.
4.Nassar N et al. Second IAS Conference on HIV Pathogenesis and Treatment, Paris, abs 554, 2003.
5.Staszewski S et al. JAMA 285(9):1155-63, 2001.
6.Matheron S et al. Antiviral Therapy 8(2):163-171, 2003.
7.Gallant JE et al. 43rd ICAAC, Chicago, abs H-1722a, 2003
8.Farthing C et al. Second IAS Conference on HIV Pathogenesis and Treatment, Paris, abs 43, 2003.
9.Winston A et al. XII Int HIV Drug Resistance Workshop, Los Cabos, Mexico, abs 137, 2003.
10.Gazzard BG et al. 43rd ICAAC, Chicago, abs H-1722b, 2003.
11.Gallant JE. http://clinicaloptions.com/HIV, 2003.
12.Bartlett JA et al. XIVth Int AIDS Conference, Barcelona, 2002.
13.van Leeuwen R et al. AIDS 17(7):987-99, 2003.
14.Gerstoft J et al. AIDS 17(14):2045-2052.
15.Clumeck N et al. AIDS 15(12):1517-1526.
16.Martinez E et al. NEJM 349: 1036-1046, 2003.
17.Opravil M et al. J Infect Dis. 185(9):1251-1260, 2002.
18.Bucher HC et al. AIDS 17(17):2451-2459, 2003.
19.Katlama C et al. HIV Medicine 4: 79-86, 2003.
20.Hoogenwert M et al. Lancet 362(9400):1979-1980, 2003.