| dc.description.abstract | HIV/AIDS is a global pandemic that has claimed the lives of 39 million people and currently afflicts 36.9 million more. These effects are disproportionately felt in sub-Saharan Africa which has almost 70% of those current cases and where women account for 59% of current infections. These numbers are even more disproportionate for young women and complications associated with HIV/AIDS are the leading cause of death in women aged 15-44. This creates a need for female-controlled tools to prevent infection. Microbicides are one such tool. Microbicides are vaginally-delivered topical products that combine a therapeutic agent with activity against HIV and a delivery vehicle. The delivery vehicle can consist of many different formulations, depending on the therapeutic agent used and the delivery mechanism desired. Polymer solutions (known as “gels” in the microbicide field) have received the most attention and have been incorporated into the most formulations for use in clinical trial evaluations. There have been several high profile microbicide clinical trial failures, including nonoxynol-9 and cellulose sulfate, as well as one success, CAPRISA 004. Nonoxynol-9 failed in the direction of harm by producing host cell toxicity. The reasons for the failure of cellulose sulfate are not fully understood, but one hypothesis suggests unmeasured benefits provided by the delivery vehicle placebo caused the treatment arm to fail in the direction of harm. Conversely, CAPRISA 004 displayed moderate protection against HIV, which was highly correlated with adherence. The VOICE and FACTS 001 follow-up studies failed to prove the same efficacy, specifically citing low adherence as the reason for failure. These previous studies illustrate the need for delivery vehicles with fully understood physical properties that are safe, allow for effective microbicide delivery, and are acceptable to patients. This study focused on rationally designing next generation microbicide delivery vehicles that are safe, effective, and acceptable. This rational design was accomplished using the methodology of computational molecular design (CMD). This process involves measuring relevant physical properties for a set of known compounds, creating structure-property correlations that relate these data to molecular structure, formulating an optimization problem that incorporates these correlations, and solving this problem to generate novel structures that possess target physical properties. Novel delivery vehicles were designed to have ideal safety, efficacy, and adherence by creating a predictive model that considered rheology, biocompatibility, and drug compatibility. Rheology describes the viscous and elastic character of a fluid. These properties influence the flow behavior of the fluid. The efficacy of liquid microbicide formulations depends upon the delivery vehicle flowing from the point of delivery to cover the entire epithelium to release the active therapeutic over this whole surface area as well as provide a lubricating barrier to pathogens. Additionally, patient acceptability and thus adherence is greatly influenced by how the fluid feels during insertion and use, what the consistency and color looks like, and whether the fluid leaks out prematurely. This study measured the rheology of a set of cellulose ethers, common liquid microbicide delivery vehicles, and related these data with molecular weight, molecular structure, and concentration to form the first part of the overall predictive model. Biocompatibility characterizes the safety and incidence of harm to host cells caused by compounds. This study defined biocompatibility as in vitro vaginal cell cytotoxicity, inflammation represented as the upregulation of two cytokines, and inhibition of C. trachomatis infection. These properties were measured for the same set of cellulose ethers and correlated with molecular structure and molecular weight to form the second part of the overall predictive model. Drug compatibility considers behavior of the active therapeutic in the delivery vehicle. This study defined drug compatibility as solubility of the drug in the delivery vehicle and release of the drug out of the delivery vehicle. Tenofovir, used in the successful CAPRISA 004 trial, was chosen as the model active therapeutic. The properties were again measured for the same set of compounds and correlated with molecular structure and molecular weight to form the final part of the predictive model. This predictive model was finally incorporated into an optimization problem formulation along with structural feasibility constraints. Target properties for all the measured physical properties were selected and the optimization problem was solved to minimize the difference between the properties predicted by the model and these targets. This optimization process resulted in novel structures for candidate delivery vehicles with improved properties compared to the measured set. These candidate delivery vehicles can be synthesized and incorporated into future microbicide formulations. | |
