Application of an expert system in the management of HIV-infected patients

A rule-based expert system, Customized Treatment Strategies for HIV (CTSHIV), which encodes information from the literature on known drug-resistant mutations was developed. Additional rules include ranking and weighting based on antiviral activities, redundant mechanisms of action, overlapping toxicities, relative levels of drug-resistance, and proportion of drug-resistant clones in the HIV quasispecies. Plasma was obtained from HIV-infected patients and the RNA was extracted. Segments of the HIV pol gene encoding the entire protease, reverse transcriptase, and integrase proteins were amplified by reverse transcriptase-polymerase chain reaction (using a total of three primer pairs) and cloned. Sequencing was performed on five clones from each of two patients. When the patient's RNA sequencing data were entered into the expert program, and the information was downloaded directly into the CTSHIV program, the five most effective two, three, and four drug regimens coupled with an explanation for their choice were displayed for each patient. Thus, the CT-SHIV system couples efficient genetic sequencing with an expert program that recommends regimens based on information in the current medical literature. It may serve as a useful tool in the design of clinical trials and in the management of HIV-infected patients.