January 16th, 2013

More Evidence That Early HIV Treatment — REALLY Early — Is Beneficial

Management of recently acquired HIV infection — especially acute HIV, pre-seroconversion — has long been controversial, with the risks and benefits of treatment versus observation debated now for nearly two decades.

(Yes, it’s been that long since the publication of this controlled trial of zidovudine monotherapy. Amazing.)

On the risk side of the equation is the toxicity of lifelong treatment, especially since some patients may have normal or near-normal CD4 cell counts after recovery from acute HIV, and may remain asymptomatic for many years. Benefits include preservation of immunologic function, reduction in transmission risk to others, and even potentially a reduction in the HIV reservoir — the last making patients who are treated for recently acquired HIV of particular interest for HIV cure strategies.

Now, two studies have been published in the New England Journal of Medicine that strongly suggest that we manage early HIV infection the same way we do long-established disease — that is, with antiretroviral therapy (surprise!), and the sooner it’s started, the better.

  • Le and colleagues conducted a prospective observational study involving 468 patients (95% men) in San Diego with acute or early HIV infection — 213 who initiated ART within 4 months after HIV acquisition and 384 who did not receive it until later. Sixty-four percent of patients treated early achieved the primary endpoint (CD4 recovery to ≥900 cells/mm3 during 48 months of observation) versus 34% of those treated later (P<0.0001). For those treated later, the median time to CD4 count < 500 — a threshold at which the clinical evidence of treatment benefit becomes much stronger — was only 12 months, suggesting that even if treatment is not started early, it is likely to be needed relatively soon.
  • In the open-label SPARTAC Trial, conducted at sites in Europe, South America, Australia, and Africa, 371 patients (60% men) with primary HIV infection (variously defined) were randomized to receive 12 weeks of ART,  48 weeks of ART, or no ART (standard of care). The primary endpoint was either a CD4 count 3 or initiation of long-term ART. During a median follow-up of 4.2 years, only the 48-week treatment group demonstrated a delay to CD4 count 3, with 29% falling to this level versus 40% of the 12-week group and 39% of the no-ART group. The proportion who eventually went on long-term treatment initiation was similar among groups. In addition, 36 weeks after the end of short-course therapy, the mean change in HIV RNA level from baseline was significantly greater in the 48-week group than in the no-ART group (difference, –0.44 log10copies/mL). Among participants in the 48-week group, treatment started closer to the time of HIV acquisition appeared to confer greater benefit. (Quick aside — HIV-specific CD4 or CD8 immune responses did not correlate with the outcome, raising questions about whether these laboratory-based findings have any clinical relevance, or are mere associations.)

There are of course limitations to these studies, most notably the non-randomized design and skewed demographics in the first, and the fact that treatment interruption makes both of the treatment strategies in the second study of limited relevance today. Furthermore, the various definitions of early HIV infection — a problem which has plagued this field from the outset — imply a very heterogeneous patient population: In the second study, some participants had acute symptomatic HIV infection pre-seroconversion at one extreme, and others were diagnosed solely based on a newly positive HIV antibody. And neither study had sufficiently long follow-up to provide hard clinical endpoints of the benefits of early treatment.

These limitations notwithstanding, there is a consistent message from the aggregate results, which is that treatment of early HIV infection — and the earlier the better — is associated with significant improvements in the most important surrogate markers of HIV disease, the CD4 cell count, and the HIV viral load. For these outcomes alone, early treatment is warranted — especially since we’re generally recommending treatment for everyone else with HIV already!

2 Responses to “More Evidence That Early HIV Treatment — REALLY Early — Is Beneficial”

  1. David says:

    I did not understand why are you so encouraged by the results of these studied. Is it that surprising that people with (acute/early) HIV are dropping their CD4 counts later if they are treated with ARVs? The SPARTAC study conclusion was “A 48 week course of ART in patients with primary HIV delayed disease progression, although not significantly longer than the duration of the treatment”. What makes you to say so decisively “For these outcomes alone, early treatment is warranted – especially since we’re generally recommending treatment for everyone else with HIV already!”

    • Paul Sax says:


      Good point, the CD4 effect was not statistically significant in SPARTAC — but the the VL difference was. Moreover, the other paper really did find a difference in ultimate CD4 recovery in those who started early vs waited, which reinforces the theme the best way to retain CD4 cells as high as possible is never to let them fall too far to begin with. Finally, there’s a consistent finding here and in other studies that most people with symptomatic acute HIV actually progress to commonly used thresholds for starting therapy relatively soon — makes me think it’s worth starting sooner rather than later, for both the proven and theoretical benefits. Why wait only on average a year or so?

      Eric Rosenberg at MGH is doing a randomized study of early treatment vs deferred therapy for acute HIV, so perhaps we’ll find out some additional pros/cons there.


HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.