Don't look now, but a two-drug lamivudine (3TC) + LPV/r strategy did just as well as a standard 3-drug regimen of 2 NRTIs + LPV/r. Better, actually, since virologic outcomes were the same and the two-drug regimen had fewer side effects.
Here are the key details about the GARDEL study, presented just this week by Pedro Cahn at the European AIDS Clinical Society meeting, or EACS:
- Study design: Open-label, randomized clinical trial in 426 treatment naive patients.
- Randomization was to 3TC 150 mg BID or 3TC plus an investigator-selected 2nd NRTI; all study participants received LPV/r twice-daily.
- At 48 weeks, 88% of the double-therapy and 84% of the triple-therapy arms had HIV RNA < 50 copies/mL
- 1% vs 5% (p = 0.03) in the double- and triple-therapy arms met failure criteria due to no data at week 48 due to stopping treatment or death (mostly stopping treatment; there was only 1 death).
- Rates of virologic failure and resistance were not significantly different between arms.
Ah, but we're all thinking, what about the high viral load stratum -- surely this group would need the extra potency of a three-drug regimen.
But surely we'd be wrong: 87% vs 78% were < 50 at week 48, so the difference favoring double-therapy was even greater.
In the "Timing is Everything" category, my recent review of the failed MODERN study of maravirc + DRV/r included this bit of prescience when discussing why the various two-drug regimens have failed:
What remains unclear is why these two-drug regimens have been so disappointing. Is two drugs not enough? Or maybe just the two drugs tested to date in these clinical studies? Is there something magic about the NRTIs? Or certain NRTIs? (One vote could be for 3TC or FTC, which have been part of every truly great HIV regimen since the late 1990s.)
I added the bolding, because it's always advisable to highlight when you're right to help balance out all those times that you're wrong.
(And believe me, there have been lots of the latter over the years. Just ask my kids. And in HIV treatment, too -- remember ddI/d4T/hydroxyurea? What were we thinking?)
So in the HIV world, this study is pretty big news, that much is clear. A two-drug regimen has never done better than a three-drug treatment in a fully powered study.
But will it influence clinical practice? Not right now, I don't think, which makes this more of a paradigm-shifter than a game-changer. (See, doctors can mobilize business-school cliches too.)
Here are at least four reasons why:
- It's a 3-pill, twice-daily regimen, and commonly used first line regimens are now all easier than that.
- Related, boosted-PI based regimens have lots (and increasing) competition in first-line therapy, especially from integrase-based strategies.
- Clinical practice and treatment guidelines have moved away from lopinavir/r due to study data and clinical experience showing that it has higher rates of adverse effects (GI, lipids) than once-daily atazanavir and darunavir.
- The generalizability of the study results might be limited given that the most common second NRTI chosen by the investigators was zidovudine (54%), followed by tenofovir (37%) and abacavir (9%).
These caveats notwithstanding, the GARDEL study raises several interesting questions:
- Would ATV/r or DRV/r have done as well with just 3TC? Or is LPV/r in this context better? I can't see why, but of the various two-drug PI plus raltegravir studies, this one with LPV/r seemed to do the best.
- Would once-daily 3TC have done as well? Remember, half-life of 3TC is shorter than FTC.
- What future fixed-dose regimens could we envision, especially as cobicistat-based PI combinations emerge and 3TC is now generic?
- How would a two-drug, 3TC or FTC plus boosted-PI regimen fare as a maintenance strategy? I suspect quite well, though for published data all we have is this small study with ATV/r.
So there you have it, the GARDEL study in all it's disruptive innovation glory. You can mail the MBA to my home address.
Paul Sax is Clinical Director of Infectious Diseases at Brigham and Women's Hospital. His blog HIV and ID Observations is part of Journal Watch, where he is Editor-in-Chief of Journal Watch AIDS Clinical Care.
Comment by: Josep M Llibre
Wed., Oct. 30, 2013 at 4:00 am EDT
However, it must be highlighted that LPV/r is not a preferred regimen nor a proper comparator. Even with the justification that it is a regimen popular in the countries where the study recruited.
It is BID, inferior to EFV, ATV/r and DRV/r, associated to high rates of diarrhoea, dyslipidemia, and associated with CVR and AMI in many cohorts. Therefore, it is hard to understand the design of the study, particularly allowing that more than half of the subjects in the triple arm received ZDV, which is again not preferred, more toxic, and inferior in response to both ABC or TDF.
People might think that the high response rate in the LPV/r + 3TC minimizes all this, and it is partly true. However, let's remind that even though the study had no CD4 count limit in the inclusion, only 19-21 % had a CD4 count lower than 200 cells. It happened the same in the Flamingo, QUAD 102 & 103, and all recent studies. Therefore, efficacy rates rise.
Reported efficacy rates of LPV/r + TDF/FTC were 78% in ARTEMIS and 76% in CASTLE, and 67% in HEAT. Therefore, the 83.7% found in GARDEL using ZDV in most than 50% of the subjects is really impressive, as it is obviously the 88.3% of LPV/r + 3TC as well.
However, the absence of a proper comparator makes me be very cautious with the results. The drop-out rates due to AEs were 0.9% in LPV/r + 3TC and 4.9% in the LPV/r + 2NRTIs arm (p=0.03). It would be helpful to see the difference in efficacy of LPV/r + 3TC vs LPV/r + TDF or ABC (w/o ZDV).