Mo, Shirley2019-03-262016-052016-07-062016http://nrs.harvard.edu/urn-3:HUL.InstRepos:38811453While many HIV therapies are now available, greatly improving disease prognosis for patients, long-term adherence to treatment regimens remains a serious problem. Recent advances in drug delivery have made possible new extended-release drug formulations, such as injectable nanosuspensions, subdermal implants, and gastric release devices which may only need to be taken on a weekly or monthly basis. Questions remain in how these new extended-release drugs might affect disease outcomes and the risk of drug resistance. A major concern is that while the new formulations have extended half-lives and decrease the frequency of dosing, they may maintain drug concentrations at suboptimal levels near the end of a dosing interval for a longer period of time with unknown consequences. Here, we develop a hybrid stochastic-deterministic model of viral dynamics capable of tracking multiple viral strains with multi-mutations. In linking time-dependent drug dosing to viral dynamics, we are able to predict patient outcomes in response to regular and extended-release drug formulations. We seek to find an optimal extended-release dosing pattern and explore the tradeoff between adherence and Cmax, the highest drug concentration available in the blood for extended-release drugs. We found that among four drug candidates considered for development, ATV/r appears to be the most promising.application/pdfenThesis or Dissertation2019-03-26