Percutaneous Penetration Kinetics of Nitroglycerin and Its Dinitrate Metabolites Across Hairless Mouse Skin in Vitro

The percutaneous penetration kinetics of the antianginal, nitroglycerin (GTN), and its primary metabolites, 1,2- and 1,3-glyceryl dinitrate (1,2- and 1,3-GDN), were evaluated in vitro, using full-thickness hairless mouse skin. GTN and the 1,2- and 1,3-GDNs were applied (a) in aqueous solution as pH 7.4 phosphate-buffered saline (PBS) and (b) incorporated into lipophilic ointment formulations. The cutaneous transformation of GTN to its dinitrate metabolites was detected, but no interconversion between 1,2-GDN and 1,3-GDN was observed. Following application of the nitrates in PBS solution, all three compounds exhibited steady-state transport kinetics. The steady-state flux of GTN (8.9 ± 1.5 nmol cm^–2 hr^–1) was significantly greater (P < 0.05) than those of 1,2-GDN (0.81 ± 0.54 nmol cm^–2 hr^–1) and 1,3-GDN (0.72 ± 0.20 nmol cm^–2 hr^–1). The corresponding permeability coefficient () for GTN (20 ± 3 × 10^–3 cm hr^–1) was significantly larger than the corresponding values for 1,2-GDN (1.4 ± 0.9 × 10^–3 cm hr^–1) and 1,3-GDN (1.2 ± 0.4 × 10^–3 cm hr^–1), which were statistically indistinguishable (P > 0.05). Further analysis of the transport data showed that the differences between GTN and the GDNs could be explained by the relative stratum corneum/water partition coefficient (K _s) values of the compounds. The apparent partition parameters, defined as = K _s · h where h is the diffusion path length through stratum corneum (SC)] were 19.8 ± 2.5 × 10^–2 cm for GTN and 1.91 ± 1.07 × 10^–2 and 1.81 ± 0.91 × 10^–2 cm for 1,2- and 1,3-GDN, respectively. However, when the nitrates were administered in an ointment base, the apparent partition parameter (') and permeability coefficient (') of GTN markedly decreased, to 2.51 ± 0.75 × 10^–2 cm and 1.6 ± 0.3 × 10^–3 cm hr^–1, respectively. In contrast, the ' and ' results for 1,2- and 1,3-GDN were not significantly different (P > 0.05) from the corresponding and values, which were measured following dosing as aqueous solutions. As a result, the steady-state fluxes of all three nitrates from the ointment formulation were comparable (GTN, 154 ± 28 nmol cm^–2 hr^–1; 1,2-GDN, 162 ± 22 nmol cm^–2 hr^–1; 1,3-GDN, 162 ± 34 nmol cm^–2 hr^–1). It follows that the dinitrates can be as efficiently delivered across the skin as GTN when a suitable formulation is employed. This finding may support transdermal therapy using 1,2- or 1,3-GDN if, indeed, they are found to be pharmacologically effective.