Due to pathological changes in the ceruminous glands, the total amount of lipids (% by weight) of canine cerumen of OE ears has been described to differ significantly from healthy ears . In our analysis, however, the detailed analysis of various lipid classes in canine cerumen from healthy and inflamed ears demonstrated comparable lipid compositions except for sterol ester (Table 1). The lipid composition of the SCC used in the present study differed from that implemented by Sánchez-Leal et al. . However, these investigators used synthetic cerumen based on the average results of human and canine cerumen studies [9, 10, 14] with diminished concentrations of oleic acid to obtain adequate consistency and malleability. Thus, the main components in their synthetic cerumen were fatty acids (myristic acid, palmitic acid and oleic acid). In contrast, triglycerides dominated in our cerumen composition derived from the cerumen of OE-affected dogs. This difference may be due to variant methodologies of extraction and analysis or the time of the year when the sampling was done. Furthermore, it is known that storage before analysis can promote lipase activity in cerumen with the result that the triglycerides are metabolised to fatty acids. In our study, the lipids were immediately extracted, but information about storage times is not available in the literature data on which the synthetic cerumen (JSL) described by Sánchez-Leal et al.  was based. To overcome this lack of information, the diffusion studies were performed using both these lipid mixtures. The results were found to be similar: in the case of persistent cerumen films in the auditory canal, Aurizon® would exhibit the fastest diffusion through the cerumen lipids.
The lipid classes present in the synthetic cerumen resulted in a particular lipid architecture , which had to be softened and dissolved to provide adequate space for diffusion in our experiments. It seems that the content of sorbitan oleate in Aurizon® with a molecular structure similar to several skin penetration enhancers like sucrose oleate  plays a major role in this respect. Nevertheless, it cannot be ruled out that the other excipients used in Aurizon® such as the medium-chain triglycerides, silicon dioxide or propyl gallate interact with the cerumen lipids and promote diffusion activity. Since all the other otic preparations tested, except for the ear cleaners, contain liquid paraffin, the lower diffusion activity of these products may even be due to its supplementation. Liquid paraffin is a mixture of various alkanes, which are only marginally detectable in canine cerumen, which may be a reason why liquid paraffin-containing products exhibit low diffusion activities. In contrast, the composition of Aurizon® is more comparable to the composition of canine cerumen with regard to its triglyceride content and, therefore, it is not surprising that Aurizon® showed a better diffusion activity.
Penetration through lipid-rich domains in the skin is mainly determined by the lipophilicity of a drug. In addition, the melting point and the molecular weight can influence the diffusion process . Oil red O was chosen in order to visualise the diffusion of a lipophilic drug (LogP 7.6) with a moderate molecular weight of 409 g/mol and a melting point of 120°C. For comparison, the diffusion of a hydrophilic drug (marbofloxacin) with similar molecular weight (362 g/mol) and a melting point of 268°C was selected. Marbofloxacin has comparable diffusion behaviour to Oil red O, although a main physicochemical characteristic for diffusion through lipid layers, its lipophilicity, differs markedly. The HPLC analysis ruled out that only lipophilic compounds like Oil red O diffused through cerumen. From this result it can be assumed that active ingredients with physicochemical drug characteristics similar to the marker compounds used in the diffusion test would exhibit similar diffusion behaviour.
The ceruminolysis results showed that the high viscosity of some of the products led to a weight gain due to impregnation (Tests 1–3). Since impregnation can be caused either by incorporation of the product and disintegration of cerumen lipids or by adhesion to the cerumen surface, the products were removed from the tubes in Test 4 thereby allowing the evaluation of the ceruminolytic effects of the viscous products. Aurizon®, Otomax®, Panolog® and Posatex® exhibited a high potency to impregnate the cerumen, while Eas Otic® and Surolan® showed an intermediate potency. The ear cleaners Epi Otic® and Otifree® had very little impregnation potency, as expected. These differences may be due to the different amounts of viscous excipients like liquid paraffin (Eas Otic®, Otomax®, Posatex®, Surolan®, Panolog®) or medium-chain triglycerides (Aurizon®), all of which are not included in the ear cleaners.
Since the investigated otic preparations (with the exception of Epi Otic® and Otifree®) have to remain in the external ear canal to exhibit optimal antimicrobial, antifungal or antiparasitic activity, a high impregnation potency will contribute to a good efficacy. With regard to the results of the diffusion study, Aurizon® seems to provide ideal conditions for its active compound marbofloxacin to get into the cerumen. On the other hand, none of the products (except for Eas Otic® with a high standard deviation) performing a high or intermediate impregnation effect exhibited any ceruminolytic activity. A slight weight loss was only found with Otifree® and Eas Otic®. It can just be assumed that propylene glycol (excipient in Otifree®) caused disturbances in the lipid architecture of the cerumen and elimination, since propylene glycol has been described as being able to modify skin penetration of topically applied drugs such as tenoxicam or bupranolol [21–24].
The penetration of the active components of the tested products is based on the assumption that they have similar physicochemical properties as Oil red O or marbofloxacin. Both these compounds have comparable molecular weights (362–409 g/mol) and melting points (120-268 C). The penetration effect of the products was evaluated assuming that penetration is mainly influenced by the lipophilicity of the drug (similar to skin permeability through lipid-rich domains). Oil red O and marbofloxacin are extremely different concerning their lipophilicity. They were chosen to cover a wide range of different lipophilicities. Nevertheless, the eight otic preparations also contain active components with other physicochemical qualities, which may not be covered by the present study. Furthermore, missing components of natural cerumen like proteins or keratinocytes may affect the penetration of active components as well, due to their breaking up the lipid matrix and unknown interactions with the active components.
The ceruminolytic activity was evaluated on the basis of the lipolytic effect of the products. The addition of other components of natural cerumen like keratinocytes or proteins would provide results closer to the in-vivo situation, albeit possibly with increased test variability. Due to the lack of information on these components of natural cerumen, further studies are required.