Predictive Pharmacodynamic Model Of Fluoroquinolones

Bacteria-specific PK/PD indices prediction model, Bacteria-general PK/PDindices prediction model and structure-pharmacodynamic model were built in thisstudy to predict fluoruoquinolone’s pharmacodynamic characteristics, which wereexpected to accelerate new drug development and provide information for decisionmaking process of new drug approval.Three different models were built in this study.（1） Bacteria-specific PK/PDindices prediction model.4bacteria were included in this study.15fluoroquinolones’PK/PD indices and91descriptors were used to build models.（2） Bacteria-generalPK/PD indices prediction model. Linear mixed effect model was used to estimate therelationship between descriptors and PK/PD indices, while inter-bacteria variabilitywas set as random effect.（3） Pharmacodynamics-Structure model. The data used inthis model was collected from Tianjin First Center Hospital. Body temperature ofantipyretic process was treated as pharmacodynamic indicator and nonlinear mixedeffect method was used to build the model. The model was combined withbacteria-general model in part2in the end.4bacteria-specific models were separately built for4bacteria. These models areS.aureusATCC291213: log10AUIC=-11.476+2.317*CHI_3CH+0.396*CHI_V0+3.928*IAC_Mean, E.faecaliATCC29212: log10AUIC=-4.061+0.0327*Zagreb+1.492*<2.590Num_Rings6>, E.coliATCC25922: AUIC=-2564.696+1367.216*CHI_3CH+1416.158*IAC_Mean+14.623*Num_Rings62+24.301*LogD, Ps.aerugi ATCC27853: log10AUIC=0.2131.4885*CHI_3P+0.0198*E_ADJ_mag+2.729*IAC_Mean0.411*Num_HDonors. Bacteria-generalmodel is log10fAUC24/MIC=-0.356*Num_HDonors–0.312*Num_Rings6+0.00347*E_ADJ_mag+1.331*CHI_3CH, and final pharmaocynamic-structuremodel is TEMP=36.739+（A0-36.739）*exp (-（0.02613*10（-0.356*Num_HDonors–0.312*Num_Rings6+0.00347*E_ADJ_mag+1.331*CHI_3CH）*Dose*DAY）.The three different models built in this study allow us to use descriptors, such asnumber of hydrogen bond donors and number of rings of size6, to predictbacteria-specific/general PK/PD indices and clinical pharmacodynamic information. These information can provide reference infromaion for screening and optimizationprocess in new drug development and decision making in new drug approvals.