On June 2, the Computerworld Honors Program selected IBM as one of its 2009 laureates for its contribution to EuResist, a free online computer program designed to assist doctors in determining optimal drug combination therapies for HIV patients.

EuResist recommends a combination of drugs to help prevent the development of drug resistance, which is the ability of the virus to mutate or change so that the drug is no longer effective. This is not only dangerous for the infected individual but also provides real challenges to drug developers, who must constantly develop new drugs to combat the evolving HIV virus.

EuResist recommends treatments that will work longest and help prevent the virus from replicating, which hinders the virus’s ability to mutate and develop resistance to the antiretroviral therapy.

HIV has a high mutation rate and short life cycle, two factors that make it likely to develop drug resistance. Its high mutation rate is due its poor ability to check for mistakes when replicating.  In humans, there are multiple mechanisms that check to make sure genetic information is copied accurately and that correct any mistakes. In HIV, however, many mistakes are not caught during replication.  This coupled with the fact that the virus may replicate up to 10 billion times in one day, makes it possible for new genetic strains to develop. These strains may require a different assortment of antiretrovirals to keep the virus in check.

EuResist is capable of providing an updated treatment plan specific to the strain of HIV according to the location and type of mutations present in its DNA.

IBM researchers in Haifa, Israel developed EuResist’s mathematical models and prediction engines to allow the program to analyze the world’s largest database on HIV resistance. The program has access to information composed from more than 33,000 patients and 98,000 therapies, as well as 370,000 viral load measurements. Viral load, or the number of HIV copies in one milliliter of blood, is used to determine the severity of the HIV infection and its subsequent reduction shows the relative success of drug therapies.

EuResist integrates clinical data and the HIV’s strain to predict the most effective combination of drugs in each patient’s particular case.

The system’s predictions are successful in achieving an undetectable virus load nearly 76 percent of the time, which is more accurate than current HIV treatment prediction models and even human experts. In the “Engine versus Experts” study, EuResist was able to outperform nine out of the 10 international experts in the field. This study was a simulated competition in which EuResist and the 10 humans were presented with 25 patient case studies and were asked questions about their treatment plans.

For over 20 years, members on the Chairman’s Committee of the Computerworld Honors Program have granted awards annually to recognize individuals and institutions who have used information technology to benefit society. This year, IBM was one of over 150 worldwide institutions recognized at the gala ceremony in Washington, D.C.

For more information, please see the IBM press release.