OCTANE Trial DSMB finds ritonavir boosted lopinavir superior to nevirapine in HIV-positive women previously exposed to single dose nevirapine
Polly Clayden HIV i-Base
On 28 October 2008 the National Institute of Allergy and Infectious Disease (NIAID) in the US issued a press release in which they explain that an interim DSMB review of a large clinical trial has found LPV/r containing HAART to be significantly more effective than NVP containing HAART in women who had previously received a single dose of NVP to reduce mother to child transmission.
The A5208/Optimal Combination Therapy After Nevirapine Exposure (OCTANE) study is a Phase 3 trial of 745 women at 10 sites in seven African countries.
The trial objectives are to evaluate which of two antiretroviral regimens is more effective for women previously exposed to single dose NVP and whether single dose NVP compromises subsequent NVP containing HAART.
OCTANE is composed of Trial 1 and Trial 2. Women in Trial 1 (n=243) were NVP exposed prior to the study and women in Trial 2 (n=502) were unexposed. Women in each trial were randomised to receive either NVP + FTC/TDF or LPV/r + FTC/TDF. Women were eligible for treatment at CD4 </= 200 cells/mm3.
The primary endpoint for both trials is time from randomisation to virologic failure of death. Viorologic failure is defined as <1 log10 below baseline 12 weeks after starting treatment or as viral load >/=400 copies/mL at or after 24 weeks of treatment (whether or not randomised treatment is being taken at the time of failure).
On October 6, 2008, the DSMB evaluated interim data from OCTANE (median follow up 66 weeks) and found LPV/r to be significantly more effective than NVP in Trial 1, and also associated with fewer side effects.
At trial entry, the women in Trial 1 were a median of 31 years of age with a median CD4 count of 139 cells/mm3 and viral load of 5.15 log/10 copies/mL. At the time of randomisation the time from receiving single dose NVP was 6 to <12 months for 32% of women, 12 to <24 months for 40% and >/=24 months for 28%. The DSMB found that 39 women discontinued treatment (NVP or LPV/r). This included 34/123 (28%) in the NVP arm and 5/120 (4%) in the LPV/r arm (p<0.0001).
Significantly more, 29(24%) women in the NVP arm, failed to achieve undetectable viral load (n=25) or died (n=4), compared to 8 (7%) women in the LPV/r arm who failed to achieve undetectable viral load (n=7) or died (n=1) (p=0.0005). None of the 5 deaths were reported to be associated with antiretroviral treatment.
The DSMB also found that of the 25/200 women who began the study with resistance to NVP, those in the NVP arm where more likely to be unsuppressed or die than those in the LPV/r arm. 5/13 women (38%) in the NVP arm who began the study with NVP-resistance either had detectable viral load or died, while all of the 12 women in the LPV/r arm who began with NVP-resistance are alive with undetectable viral load.
Additionally they found a trend for the negative effects
of prior single dose NVP exposure to decrease the greater the interval between receiving single dose NVP and initiating NVP containing HAART (see Table 1).
Table 1: OCTANE Trial 1 – Women experiencing negative effects of prior single dose NVP exposure on NVP containing HAART by interval since exposure
|Time since s/d NVP||NVP arm||LVP/r arm|
|6 to <12 months||n=15 (37%)||n=1 (3%)|
|12 to <24 months||n=11(24%)||n=4 (8%)|
|>24 months||n=3 (8%)||n=3 (10%)|
n= number of women who reached an endpoint at time of DSMB review.
Following the review DSMB recommended that these findings about Trial 1 be unblinded. As there are no safety concerns with Trial 2, the DSMB advised that it continue as planned.
The study investigators are notifying all participants of the findings of the DSMB. Treatment with LPV/r is available to the women in Trial 1 who have been receiving NVP, should the participant and investigator choose to change their therapy. Trial 2 will continue as planned until the end of the study June 2009 to determine how the two treatment options compare for women who have not previously taken NVP.
That NVP exposed women receiving NVP-containing HAART did less well than those receiving LPV/r-containing HAART is unsurprising.
One thing that is notable about these findings though is that the women in the LPV/r arm had results that were better than might be expected from other studies: 93% were alive and below detection (<400 copies/mL; <50 copies/mL data was not shown) at a median of 66 weeks. This compares to results from the CASTLE study 76% of patients on LPV/r had <50 copies/mL at 48 weeks; ARTEMIS, 78% on LPV/r had <50 copies at 48 weeks; KLEAN, 71% on LPV/r had <400 copies at 48 weeks and in ACTG 5142, 86% on LPV/r had <200 copies/mL at 96 weeks. Although backbones and endpoints vary, the OCTANE LPV/r arm appears to have performed very well. So the results from Trial 2 of LPV/r vs NVP in unexposed women will also be useful and that part of the trial is still ongoing.
Most importantly, the median CD4 of the women at enrollment in OCTANE was 139 cells/mm3, and median interval since receiving single dose NVP was about 12-18 months. So many of these women may have met the eligibility criteria to start HAART at 200 cells/mm3 for their own health in pregnancy had it been available (and probably all if the threshold was 350 cells/mm3). These women with low CD4 during pregnancy receiving single dose NVP would be at higher risk of NVP resistance and therefore be more likely to fail NVP containing HAART if they started within a year or so of exposure.
That there was a trend for the negative effects of prior single dose NVP exposure to decrease as the interval between receiving single dose NVP and initiating NVP containing HAART became greater is consistent with other studies. MASHI-Plus, NEVEREST and the CDC studies have all reported that response to NVP containing HAART in women with sd NVP exposure <12 months or so prior to starting HAART was poor compared to those without exposure, but response in women with sd NVP exposure >12 months or so prior to starting
HAART was similar to those without exposure. In OCTANE, for women starting HAART >/=2 years after sd NVP exposure there was no significant difference between LPV/r and NVP containing HAART. One explanation being that these women had higher CD4 at the time of exposure and therefore less at risk for
acquisition of resistance.
Although programmes are increasingly using short course AZT with sd NVP for PMTCT this strategy appears to have only modest impact on acquisition of resistance.
So what needs to be done?
1. HIV-positive pregnant women should have CD4 counts in pregnancy, and these need to be available at all health service levels.
2. Women who meet the criteria for HAART initiation should receive treatment for their own HIV as a matter of urgency. Ideally the threshold should be <350 cells/mm3, but even following the minimum guideline recommendation of 200 cells/mm3 correctly will exclude women most at risk for acquisition of resistance, of mother to child transmission and (often neglected in PMTCT discussion) AIDS and death. This should continue post partum. This is uncontroversial.
3. More controversial is what is the best intervention for women not in need of treatment for their own health. Current WHO guidelines recommend a two-tiered approach in which healthy women receive short course AZT plus single dose nevirapine with AZT/3TC for 7 days post partum to reduce resistance risk. These data re-emphasise the need to include the AZT/3TC “tail” intervention with this strategy to reduce the risk of NNRTI resistance, which has not been adopted in all PMTCT programmes.
4. Should guidelines change to offer short course HAART to healthy women stopping after delivery or breastfeeding? If so what regimen? NVP is out of the question because of hepatoxicity risk at higher CD4 counts, PIs are very expensive and triple nucleosides have limited data and may be less effective. Additionally following the SMART results concerns have been raised about the safety of treatment interruptions.
5. So one thing is clear, we need more data. Trials are planned or underway, for example, in Thailand and Botswana, and a large trial is planned with sites in US/Brazil and Africa and Asia to address many of these questions. Additionally the Botswana national programme is planning to roll out a pilot of universal HAART in pregnancy. The next WHO guidelines revision will need to address this issue and WHO is covening an initial consultative meeting on this in November 2008.
i-Base will include a detailed review of the current data, ongoing and planned trials and outstanding research questions in the next issue of HTB.
Thanks to James McIntyre and Lynne Mofenson for discussion of these results.
Source: NIAID press release. “DSMB finds ritonavir-boosted lopinavir superior to nevirapine in HIV-positive women who previously took singledose nevirapine”. (28 October 2008).