Pre-exposure prophylaxis would maintain HIV prevalence among high-risk gay men at the current level and prevent a prevalence increase of up to 5%, a mathematical model has shown, even if it were taken only half the time and only prevented one in two infections.
However for PrEP to significantly reduce HIV prevalence in this population it would need to be 90% effective and be taken continuously.
If prescribed in either of these two ways it would be barely cost effective. Intermittent PrEP could be more cost effective, but to be so it would need to have higher efficacy than yet demonstrated – around 90%.
The model was presented to the European AIDS Conference by Jonathan Anderson of the Australian National Centre in HIV Epidemiology and Clinical Research. He noted that past analyses of whether PrEP could be cost-effective, based on the US epidemic, had produced contradictory results.
The current model is based upon HIV prevalence and dynamics in Australian gay men.
The base-case model assumed a current HIV prevalence of 9% in this population and that PrEP using tenofovir and FTC would cost $7536 a year (in US dollars). The annual cost of managing a case of HIV infection was put at $13,920 a year and the base-case scenario assumed a PrEP efficacy of at least 87%, similar to 100% attempted condom use. The prevalence of drug-resistant HIV was assumed to be low (under 3%) as was the incidence of serious adverse drug reactions (under 4%).
Other factors inputted included mortality, types of sex, and the effect of HIV treatment on infectiousness. The model was able to include changes in risk behaviour and HIV prevalence over time. It is able to predict changes over a 40-year timeframe.
It found that if continuous PrEP with more than 87% efficacy was used, the programme would be barely cost effective, costing US$47,745 per quality-adjusted life year (QALY) saved. The threshold for government willingness to pay for health interventions is assumed to be $50,000.
The model assumed that the programme would be very targeted to risk behaviour; Anderson said it certainly would not be cost-effective if it included heterosexuals, or gay men without high-risk behaviour.
Under these strictly-defined circumstances, the model shows that HIV prevalence among high-risk gay men would gradually decline to 4.36% over a 40-year period.
PrEP that was intermittent, only taken 50% of the time, would be more cost-effective if its efficacy was high. If it prevented nine in ten infections it would only cost $6816 per QALY saved.
Few animal studies of even continuous PrEP have demonstrated efficacy of this order and certainly not human studies.
However intermittent PrEP would continue to be cost affective down to a threshold of 46% efficacy, at which point it would reach the $50,000 dollar per QALY saved threshold.
According to the model, intermittent PrEP would not reduce HIV prevalence - it would stay at 9% over the 40-year period. However if PrEP was not introduced at all and all other influences on HIV transmission remained the same, HIV prevalence in gay men would increase gradually to 13.6%.
The model is very sensitive to even small increases in risk behaviour. A 2% annual increase in unprotected serodiscordant sex, for instance, would render a continuous-PrEP programme no longer effective.
The continuous-PrEP programme envisaged would cost $330 million per year, with a minimum cost per individual of $35,000 once factors like publicity, education and administration are taken into account.
Anderson urged more research into intermittent PrEP to find if costs could be brought down without sacrificing too much efficacy.
Anderson J and Cooper D. Cost-effectiveness of pre-exposure prophylaxis for HIV in an MSM population. 12th European AIDS Conference, Cologne. Abstract BPD1/6. 2009.