A pharmacokinetic-pharmacodynamic model of chemotherapy of human immunodeficiency virus infection that relates development of drug resistance to treatment intensity.

RNA viruses, including retroviruses, have mutation rates that are about 100 times higher than those of DNA viruses, bacteria, or eukaryotes, so that resistance to AIDS drugs emerges very rapidly. This has been shown to limit the effectiveness of the treatment of AIDS by reverse transcriptase inhibitors, such as zidovudine (AZT) and resistance to the new class of HIV aspartyl protease inhibitors has already been reported. The technique of pharmacokinetic-pharmacodynamic simulation has now been used to predict ways of delaying the development of resistance to these two classes of antiretroviral agents. A model is described that includes pharmacokinetic, pharmacodynamic, and cytokinetic equations, and expressions describing effects if the HIV on the immune system and destruction of virally infected cells by cellular immunity. The model predicted that the degree of viral drug resistance in relation to the sustainable blood level of drug would be the major determinant of response duration. Early treatment was consistently superior to late treatment, both with a drug that caused cumulative toxicity and with a drug that did not. Making reasonable assumptions about the likely degree of viral resistance, in conjunction with typical blood levels achievable for reverse transcriptase inhibitors or aspartyl protease inhibitors led to predicted response durations of several months to a few years, despite the rapid mutation rate of HIV. Preliminary studies of combination chemotherapy showed that predicted response durations were greater than for monotherapy, though less than the sum of responses to the individual drugs. Strategies for delaying the development of resistance include early treatment, combination chemotherapy, and developing novel agents with a high ratio of plasma level to antiviral efficacy.