Pharmacokinetic (PK) and pharmacodynamic (PD) modeling of host-pathogen interactions has enhanced our understanding of drug resistance. However, how combinations of drug resistance mutations affect dose-response curves remains underappreciated in PK-PD studies. The fitness seascape model addresses this by extending the fitness landscape model to map genotypes to dose-response functions, enabling the study of evolution under fluctuating drug concentrations. Here, we present an empirical fitness seascape in E. coli harboring all combinations of four drug resistance mutations. Incorporating these data into PK-PD simulations of antibiotic treatment, we find that higher mutation supply increases the probability of resistance, and early adherence to the drug regimen is critical. In vitro studies further support the finding that the second dose in a drug regimen is important for preventing resistance. This work bridges empirical fitness seascapes and computational PK-PD studies, revealing insights into drug resistance.