Long-Acting HIV Antiretrovirals May Be Revolutionary. But Will They Be Worth It?

Executive Editor

Once-a-month HIV antiretroviral dosing is no longer a fanciful dream. Long-acting antiretrovirals are currently in development, and the first glimpses of early safety/efficacy data are not terribly far away. So it's not outlandish for us to start thinking about where these drugs might fit into the HIV treatment armamentarium once they arrive.

It's within that context that Eric Ross, a research assistant at Massachusetts General Hospital's Medical Practice Evaluation Center, strode to the podium at IDWeek 2013 to offer the results of a "what if" simulation model.

Ross and his colleagues asked themselves: What if the U.S. Food and Drug Administration approved a once-a-month HIV antiretroviral -- possibly GSK744 or TMC278 LA, both of which are presently in early development and have not yet been studied for efficacy in HIV-infected people? Heck, what if enough of them were approved that a complete, long-acting regimen could be formed? How might that regimen be utilized? How might it increase the lifespan of people with HIV? How much would it cost -- and would that cost be worth the benefits it might provide?

To explore these questions, Ross et al fabricated an imaginary cohort of treatment-naive, HIV-infected patients and stratified it into four groups:

  • control group (takes currently available, once-daily oral HIV treatment)
  • multi-failure group (took two or more once-daily, oral HIV treatment regimens, and lost viral control on them all; then switched to a once-monthly regimen)
  • second-line group (failed their first once-daily, oral HIV treatment regimen; then switched to a once-monthly regimen)
  • first-line group (takes a once-monthly regimen as their initial antiretroviral therapy)

Ross et al also made a few assumptions about the baseline characteristics of their study population, based on recent research into the current demographics of the U.S. epidemic:

  • mean baseline CD4+ cell count of 320 cells/mm3
  • mean baseline age of 43
  • mean gender split of 84% men, 16% women
  • mean adherence to oral antiretroviral therapy of 89% (vs. 100% on long-acting antiretrovirals)

Estimates were also made regarding the effect of adherence on virologic suppression and loss to follow-up.

An estimated drug cost of $53,000 per patient per year for this as-yet-nonexistent, once-a-month drug was factored into this statistical model. The estimate was based on assessments of the costs of existing oral HIV medications, as well as the costs of novel, long-acting drugs developed for other diseases. (Ross noted that $53,000 was roughly 85% more expensive than oral regimens based on boosted protease inhibitors.)

When all of these data were plugged into their simulation, Ross et al found that -- assuming generally good adherence -- there was only a small benefit to life expectancy among patients receiving long-acting antiretroviral therapy (LA-ART) compared to patients receiving once-daily treatment: an additional .5 years for those with multiple prior treatment failures, an additional .6 years for those with one prior treatment failure and an additional .7 years for those who were taking LA-ART as their first regimen. The relative benefits increased significantly, to upwards of three years, among patients with the worst adherence to daily antiretroviral therapy, the simulation suggested.

Regardless, the exploratory analysis appeared to show a benefit to life expectancy. The remaining question was: Is that benefit high enough to make the drug cost-effective?

The answer was: maybe. Using generally approved measures for incremental cost-effectiveness, Ross et al determined that LA-ART would be a cost-effective option only for patients with multiple prior failures on standard, once-daily therapy. The cost of LA-ART would need to be $29,000 per patient per year -- roughly the same as a typical boosted protease inhibitor-based regimen in the U.S. -- for the regimen to be cost-effective in their simulation, they calculated.

Ross et al tinkered with some of their initial assumptions regarding baseline patient characteristics and other parameters (including adherence levels and loss to follow-up) in order to establish a better range of cost-effectiveness possibilities, but ultimately found similar results: LA-ART's greatest value, and highest likelihood of cost-effectiveness, would be for patients who had already failed multiple once-daily regimens. For it to be "worth it" even as a second-line treatment option, the cost would need to be dramatically lower than estimated, in the $27,000 to $34,000 range.

Ironically (though logically), Ross noted that good adherence made LA-ART less cost-effective. This is why its appeal was so much greater in multi-failure scenarios: Those situations tend to assume that poor adherence was the reason for the failures (as it often is), and LA-ART eliminates daily adherence problems as a concern. Among patients who would normally be adhering just fine to once-daily treatment, LA-ART is -- from a cold, calculating standpoint -- merely an overly expensive nicety, although from a more subjective standpoint such regimens could be reasonably expected to have a tremendous amount of appeal.

Obviously, there are a tremendous number of possibly mistaken assumptions that went into Ross et al's simulation. The exercise also completely ignores (by necessity) the extraordinary quality-of-life and psychological benefits that LA-ART would provide for people living with HIV. (Ross additionally noted that the analysis ignores any possible reduction in HIV transmission rates that may result from the ability of LA-ART to keep viral load more reliably suppressed in the patient population.) But the study provides valuable food for thought as we begin to ponder the future role of LA-ART alongside other treatment options. It also provides some insight into the types of calculations that may go into drug-pricing decisions.

Myles Helfand is the editorial director of TheBody.com and TheBodyPRO.com.

Follow Myles on Twitter: @MylesatTheBody.