(W1030-03-16) Design and Biopharmaceutical Characterisation of Ciprofloxacin Ester Salts: Assessing Solubility and Permeability Dynamics through Molecular Descriptors

Purpose: Ciprofloxacin (CPR), a widely used fluoroquinolone (FQ) antibiotic, has limitations in permeability due to poor solubility. Enhancing permeability through esterification offers a promising approach to overcome these limitations and improve therapeutic efficacy. Although the biopharmaceutics classification system (BCS) classifies CPR as a class IV drug (low solubility and permeability profile) (1). CPR’s oral bioavailability varies from 56% to 77% (2). This study investigates CPR esters and compares them to five other FQ esters to understand how differences in physicochemical properties impact drug performance. Key properties analysed include the partition coefficient (Log P), solubility (Log S), Critical volume, Gibbs free energy, Henry's Law constant, and topological polar surface area (tPSA).

Methods: This study compared ciprofloxacin with 37 FQs based on key physicochemical properties. Five were selected to ensure a broad property range. A library of 294 esters was built using 49 alcohols per drug, assessing how chain length and branching patterns influence the physiochemical activity of the drug.

Results: Results indicated that the most substantial relationship was that of Log S between LogP. It was then seen that this relation had a strong negative correlation with the correlation coefficient R (-0.7364), while the p-value was less than 0.001. This is the hallmark of a good negative correlation: increases in Log S go with decreases in Log P. Using the coefficient of determination, 54.2% of the variation in LogP was accounted for by Log S alone. More than half of the variation in Log P is explained just by Log S. The other predictors indicated weaker and less significant relationships with LogP. Additionally, critical volume moderately influenced LogP, indicating it may contribute to predictions. Critical Volume and Log S together modestly predict LogP. After adjusting for the number of predictors, the adjusted R-squared was 0.17, indicating that the predictors in the model can explain about 17% of the variability in LogP. The Y-intercept (1.2695), with a t-statistic value of 6.20 and a p-value less than 0.001, indicates that the intercept value is significant.  The 95% confidence interval for the intercept was from 0.8534 to 1.6857, indicating much confidence in the model. The coefficient for Log S is statistically negative, with (p-vale = 0.009), indicating a significant correlation.

Conclusion: Regression analyses determined that Log S was the best determinant as the main predictor for Log P, thereby defining a strong negative correlation where high solubility comes with low hydrophobicity. While other factors, such as critical volume, presented weaker correlations, the overall model of regression was statistically significant in indicating the crucial role of solubility. These findings give insight into optimising the oral permeability of prodrugs and enhancing therapeutic efficacy, especially for drugs like CPR.

References: 1. Wu, C.-Y. and Benet, L.Z. (2005) Predicting Drug Disposition via Application of BCS. Pharm Res. 22 (1), 11–23.
2. Junkert et al., (2023) Pharmacokinetics of oral ciprofloxacin in adult patients: A scoping review. British Journal of Clin. Pharm. 90 (2), 528–547.