Abstract
A rule-based expert system, CTSHIV (Customized Treatment Strategies for HIV) that recommends an individualized treatment strategy for HIV patients is described. Since the HIV virus mutates rapidly, a patient can develop a resistance to particular antiretroviral agents. CTSHIV addresses this problem by recommending a treatment strategy that avoids antiretroviral agents for which a resistance has developed. By monitoring the HIV virus of the patient, the treatment strategy can be switched in response to mutations of the virus. CTSHIV contains a knowledge base that encodes information from the medical literature on drug-resistant mutations. It also contains additional rules that rank and weight combinations of antiretroviral agents based upon antiviral activities, redundant mechanisms of action, and overlapping toxicities. To recommend a treatment for an HIV-infected patient, plasma is obtained from the patient and the RNA extracted. Segments of the HIV pol gene encoding the entire protease, reverse transcriptase and integrase proteins are amplified by RT-PCR (using a total of 3 primer pairs) and cloned. The patient's HIV sequencing data is entered into the expert program and the information is downloaded directly into the CTSHIV program. The program produces the five most effective 2, 3, and 4 drug regimens coupled with an explanation for their choice. Thus, the CTSHIV system couples efficient genetic sequencing with an expert program that recommends regimens based upon information in the current medical literature. Clinical trials are underway to evaluate the effectiveness of this novel approach to the management of HIV-infected patients.