HIV: Predicting Treatment Response More Accurately

03/09/2013
Photo: Patient

Knowing the possible evolutionary paths allows for more accurate predictions if the AIDS pathogen is likely to develop a resistance to a drug; © panthermedia.net/Alexander Raths

A new statistical model calculates the genetic evolution of the HI virus in individual patients. Knowing the possible evolutionary paths allows for more accurate predictions if the AIDS pathogen is likely to develop a resistance to a drug and thus if the treatment is likely to become ineffective in a specific patient.

This is the conclusion of a research project funded by the Swiss National Science Foundation (SNSF).

The HI virus is feared, not least, because of its great adaptability. If the virus mutates at precisely the point targeted by a drug, it is able to neutralise the attack and the treatment fails. To minimise these viral defence mechanisms, doctors treat patients with modern combination therapies involving the simultaneous administration of several drugs. This approach forces the virus to run through a series of mutations before it becomes immune to the drugs.

"It is not easy to decide which of the over 30 combination therapies is best suited to a patient," says Huldrych Günthard from Zurich University Hospital. The decision is based on the prospects of success and therefore on the genetic make-up of a particular virus. The established prediction models already consider the genetics of the virus but they neglect that the virus continuously evolves through sequential mutations.

In cooperation with the Swiss HIV Cohort Study, Niko Beerenwinkel and his team from the Eidgenössische Technische Hochschule (ETH) Zurich have now developed a more accurate prediction model based on a probabilistic method. This model calculates the possible evolutionary paths of the virus and yields a new predictive measure for the development of resistances: the so-called individualised genetic barrier. When applied retrospectively to 2185 patients of the HIV Cohort, the new approach made it possible to predict treatment success more accurately compared to the existing models. "We are now introducing the individualised genetic barrier in a pilot project and hope that it will help us in the future to identify the best therapy for each patient," says Günthard.

MEDICA.de; Source: The Swiss National Science Foundation (SNSF)