3.3. Scanning Electron Microscopy (SEM)
Images obtained by SEM of the cross sections of the freeze-dried bigels are shown in Figure 2
. A porous structure can be seen in all cases as a result of the sublimation of the frozen water in the freeze-drying process [55
]. However, some differences can be established between the different batches. Since the batches differ in the concentration and/or nature of the polymer included in the hydrogel in the systems, these are the factors that will explain the differences.
Depending on the polymer concentration, smaller pores can be seen in bigels with a higher amount of polymer (fd3P << fd2P, fd3C < fd2C, fd2H < fd1H); this phenomenon was also reported by authors such as Shen et al. and Furst et al. [56
]. A higher concentration of polymer can be considered to produce greater viscosity and a denser polymeric framework in the corresponding hydrogel, resulting in smaller water droplets being trapped inside, whose elimination during the freeze-drying process gives rise to the corresponding pores [58
]. The arrangement of the polymer chains during the formation of the hydrogel can be said to vary with the amount of polymer; this is more obvious in the case of pectin batches.
When comparing systems that contain different polymers but in the same proportions, freeze-dried bigels containing pectin can be seen to have larger pores than bigels based on chitosan (fd3P > fd3C and fd2P > fd2C). Batches containing HPMC reveal a microstructure with less defined pores that appear to be connected, forming ducts.
3.4. Hardness and Deformability Test
The hardness and deformability results of the freeze-dried systems are shown in Figure 3
A,B, respectively. As can be seen in Figure 3
A, the proportion of pectin and HPMC in the freeze-dried bigels affects their hardness, and the greater the proportion of the polymer, the higher the value of this parameter. Student’s t
-test corroborated these differences. This direct correlation between hardness and polymer concentration was also found by Furst et al. for hydroxyethyl cellulose sponges [49
]. A porous structure resulting from the freeze-drying of a hydrogel could be expected to be harder due to the greater density of the three-dimensional network of the previous hydrogel. The previous SEM micrographs suggest that smaller pore size can be associated with the greater hardness of freeze-dried bigels containing pectin or HPMC. However, this does not occur in the case of chitosan as there is no difference between the hardness of fd2C and fd3C. Student’s t
-test showed that the nature of the polymer at the same concentration (2%) has no effect on hardness, since no significant differences were found between fd2P, fd2C, and fd2H. However, this test also confirmed that fd3P is significantly harder and fd1H significantly less hard than the other batches.
B shows the deformability curves obtained with the maximum force applied in each compression cycle. Force values were expressed as a percentage, considering the maximum force of the first cycle to be 100%. In all cases the force required to compress the systems by 1 mm decreases from one cycle to another, indicating that these freeze-dried bigels are increasingly deformable. This suggests that the deformability of batches based on pectin or HPMC becomes greater as the proportion of polymer in the hydrogel (fd2P >> fd3P and fd1H > fd2H) decreases, due to a more pronounced reduction in force. It is especially significant in the case of pectin. Furst et al. also obtained a reverse correlation between polymer proportion and the deformability of hydroxyethyl cellulose sponges [49
]. In view of the results of the previous tests, deformability increases with larger pore sizes and the hardness of the structure resulting from lyophilisation decreases. In batches based on chitosan, the higher the concentration of polymer (fd3C > fd2C), the more deformable the system, although with no significant differences between them. These results also reveal that the proportion of polymer affects the deformability in different ways depending on the nature of the polymer included in the hydrogel.
According to the results of these tests, batch fd3P is the hardest and least deformable freeze-dried bigel of all the formulated batches, making it the most suitable for vaginal administration.
3.5. Mucoadhesion Test
The phenomenon of mucoadhesion is based on establishing interfacial forces between a material—the dosage form in this case—and a mucous membrane, whose surface is upholstered by a mucus layer [59
]. The main components of this mucus layer are water and mucin glycoproteins, which possess sialic and sulphated residues. These groups are ionized at a pH of over 2.6 (pKa), giving a negative charge to the molecule [60
]. The mucoadhesion of pectin is usually attributed to hydrogen bonds between its carboxylic groups and mucin glycoproteins [60
]. However, the pKa of this polymer (3–4) is closer to vaginal pH, so some of its carboxylic groups could be ionized in this medium. This would create an electrostatic repulsion between the carboxylic groups and the similarly negatively charged mucin glycoprotein groups. Sriamornsak et al. suggested this repulsion could aid the formation of bonds by polymer coil expansion. In pectins with a high degree of esterification, methoxyl groups confer a hydrophobic character on the polymer, resulting in greater adsorption on the mucin surface [61
]. Chitosan amino groups are positively ionized at pH values of less than 6.2–7 (pKa) [62
], so they interact ionically with the negatively charged residues of mucin glycoproteins and produce the mucoadhesion of chitosan. Hydrogen bonds between the hydroxyl and amino groups of chitosan and mucus also contribute to the mucoadhesion of this polymer. Since HPMC is a non-ionic polymer, the pH of the medium does not affect its mucoadhesion, which is explained by the formation of bonds (including hydrogen bonds) between its hydroxyl groups and the functional groups of the mucus components [64
shows the average values of the work required to detach each freeze-dried system from the mucosa, a parameter that we will call “mucoadhesion work”. As can be observed, this parameter is modified in the same way by varying the proportion of the polymer, whatever its nature, and therefore increases with the proportion of pectin, chitosan, or HPMC in the hydrogel, although only slightly in the last case. This expected correlation between the polymer concentration and the mucoadhesion work, which was also indicated by Furst et al. [57
], may be because greater amounts of polymer allow more interactions with mucus per unit of surface area. However, Student’s t
-test failed to establish any significant differences between the mucoadhesion work data for the batches containing different proportions of the same polymer. The statistical analysis revealed no significant differences between fd2P and fd2H, but significantly higher values of these batches with respect to fd2C.
In view of the above, it can be confirmed that these variations in the proportion of polymer do not affect the mucoadhesion work of the systems, although the nature of the polymer (at the same concentration) does. Thus freeze-dried bigels containing pectin and HPMC have better mucoadhesive properties than those based on chitosan.
3.6. Swelling Test
shows the data resulting from the swelling test of the freeze-dried systems in both SVF and the SVF/SSF mixture. Each graph groups the swelling profiles of the batches containing the same polymer in both media. In most cases, an initial swelling increase can be observed until a maximum value is reached, as the dominant process in this first stage is water capture from the medium. The mucoadhesive polymer traps the water in the three-dimensional structure of each freeze-dried bigel, resulting in the formation of a gel [65
]. This process can therefore be considered to reconstitute the fresh bigels. It should be noted that these formulations do not swell excessively, unlike other vaginal dosage forms such as tablets [66
], which makes them more comfortable for patients. This reduced swelling of bigels can be attributed to the small amount of swellable polymer in the dosage form and to the hydrophobic character conferred by the organogel. After the aforementioned maximum swelling value, the weight of the systems progressively decreases as they become destructured by erosion and/or dissolution.
As can be seen in Figure 5
A, the batches containing pectin begin to gain weight in the same way in SVF. However, fd2P reaches its maximum swelling value faster than fd3P (0.5 h as opposed to 24 h). fd2P begins to lose its structure earlier, and has significantly lower swelling values than fd3P from 3 h to 48 h. After this point, both batches show similar profiles until the end of the assay. The swelling profiles are more similar in the SVF/SSF mixture, as fd2P presents its maximum weight at 0.5 h and fd3P at 2 h. Nevertheless, in this short period the weight of fd2P diminishes faster than fd3P, with significant differences from the beginning of the test to 24 h. From 48 h on, these differences disappear until the end of the test, as fd2P and fd3P have similar weight variations. The fact that the lower the proportion of pectin, the sooner the maximum swelling value is reached could be because a lower amount of polymer requires less time to rehydrate [57
]. Regarding the swelling of the same batch in both media, fd2P attains its maximum weight at 0.5 h in both SVF and the SVF/SSF mixture. Both profiles are similar, although a significant difference can be observed at 24 h, when this batch has a higher value in SVF than in the SVF/SSF mixture. More differences are observed for fd3P. This batch continues swelling for longer in SVF (until 24 h), so its weight loss begins later than in the presence of SSF; this freeze-dried bigel shows significantly higher swelling values in SVF than in the SVF/SSF mixture from almost the start of the test. This faster loss of structure in the SVF/SSF mixture than in SVF is due to the acidic character of pectin. This polymer is acid-stable [68
] while highly soluble at a pH equal to or greater than 7 [69
]. At the pH of the SVF/SSF mixture (around 7.5), the carboxylic groups of the polymer are ionized and generate a mutual repulsion that hinders the association between the polymer chains and prevents the formation of a gel. However, at vaginal pH these functional groups can be considered non-ionized, which allows the cross-linking of the polymer chains and results in a gel [70
B shows the swelling profiles of freeze-dried bigels containing chitosan. In SVF they have practically overlapping profiles until 3 h, when fd3C reaches its maximum swelling. Batch fd2C does so at 6 h, so its weight decreases later and less markedly, but with significantly higher values than fd3C at only 6h into the test. In the SVF/SSF mixture, the two stages of the swelling profiles mentioned above are not observed in the case of chitosan-based batches. Initially, fd2C and fd3C undergo a weight increase until they reach a value that remains almost constant until the end of the test. Although fd2C and fd3C exhibit the same behaviour in this medium, there are differences between their swelling degrees; fd2C has higher values than fd3C, which are significant from 1h to 96 h. Their maximum weights are reached at 72 h in the case of fd2C and at 120 h for fd3C, both in the peculiar swelling plateaus that characterize these batches. In this case, the profiles obtained in the SVF/SSF mixture also reveal a relationship between the lower polymer concentration and the shorter time taken to reach maximum swelling. Very different swelling profiles are obtained in both media for batches containing chitosan. For fd2C, the values recorded in SVF and the SVF/SSF mixture overlap until 6 h. From this point on, the weight decreases progressively after this batch reaches its maximum value in SVF, with values that are very significantly lower than in the SVF/SSF mixture. In the case of fd3C, lower swelling values were obtained in the SVF/SSF mixture at the start of the assay, and were significant until 3 h. However, from 6 h on, fd3C shows the same behaviour as fd2C, although significant differences can be seen between both media from 48 h. These differences in the swelling profiles of chitosan-based batches from one medium to another can be explained by the basic character of this polymer. At vaginal pH, its amino groups are protonated, causing the chitosan to dissolve in the medium. However, at a higher pH—as in the SVF/SSF mixture—these functional groups are non-ionized, so the polymer becomes insoluble and forms a precipitate [62
The weight evolution of freeze-dried batches containing HPMC in both media can be seen in Figure 5
C. fd1H and fd2H have very similar profiles in SVF and reach their maximum weights at 24 h. Although fd1H appears to maintain its structure longer and lose weight more slowly than fd2H, their values are not significantly different at any point in the test. The swelling profiles of these batches are also similar in the SVF/SSF mixture, although there are some differences between them. While fd1H reaches its maximum at 4 h, fd2H does so at 6 h, so fd1H begins to lose its structure earlier, with lower values than fd2H that are significant at 96h and 120h. Again it is worth noting the correlation between the greater proportion of polymer and the longer the time taken to reach the maximum swelling value. Both fd1H and fd2H swell more in SVF than in the SVF/SSF mixture, and this difference is significant for both batches during much of the test. However, it is more evident for fd1H, since fd2H has very similar swelling values in both media at 72 h and 96 h due to its very acute weight loss from 48 h to 72 h. This could be because the nature of the medium determines the arrangement of the polymer chains and the solid–liquid interaction in the reconstitution of the hydrogel. Tritt-Goc et al. reported that the diffusion mechanism of a solvent into HPMC matrices varies depending on its pH and related it to the chemical exchange between the medium and the polymer [43
3.7. Drug Release Test
shows the results obtained from the drug release tests of the freeze-dried bigels in both SVF and the SVF/SSF mixture. Each graph groups the release profiles of the batches containing the same polymer in both media.
According to Figure 6
A, batches fd2P and fd3P release the total amount of TFV in 72 h in SVF, although over 90% of TFV is released by 24 h. Both batches show almost the same drug release profile in this medium, maybe slightly more controlled for fd3P. According to the f2
similarity factor, they can be considered equivalent. Relatively similar drug release profiles were also obtained for fd2P and fd3P in the SVF/SSF mixture, in which both batches delivered TFV for 24 h. However, f2
did not reveal any similarity between these two profiles in this case, which could be due to the fact that the average values of TFV released from fd2P are higher—and therefore less sustained—than those of fd3P from 1 h to 4 h of the test. The faster loss of structure of fd2P observed in the swelling profiles would explain these minor differences between the release profiles of fd2P and fd3P. Notable differences can be observed when comparing the profiles in both media. The release of TFV is faster in the SVF/SSF mixture than in SVF, with significant differences from the first hour of the assay for both fd2P and fd3P, which were supported by the f2
factor. In the SVF/SSF mixture, around 90% of the drug is released in the first 6h of the test, whereas only about 60% of the dose is released from the same formulations in SVF. This agrees with the results obtained from the swelling test, which showed a faster loss of structure in the SVF/SSF mixture than in SVF, since pectin is more soluble at a higher pH.
B shows the drug-release profiles of batches containing chitosan. In SVF, fd2C and fd3C release the drug for 96 h and 72 h respectively, although over 90% of the dose is released at 24 h in both cases. Higher values of TFV released from fd3C than fd2C can be observed from 3 h to 6 h. Despite this, the f2
similarity factor did not indicate any significant differences between these profiles. In the SVF/SSF mixture, fd2C and fd3C allowed a controlled release of TFV for 24 h and 72h respectively, although around 90% of the drug is delivered in 24 h in the second case. Differences can be noted at 24 h, when fd2C reaches a higher percentage of released drug than fd3C; however, their profiles are equivalent based on the f2
value. Although some differences can be observed when comparing the release of TFV in SVF and in the SVF/SSF mixture, the f2
factor proved they are significant neither for fd2C or fd3C. Considering that freeze-dried bigels containing chitosan maintain their structure in the SVF/SSF mixture, a more sustained release of TFV could be expected from C batches in this medium than in SVF, and yet there are no significant differences between the release profiles of both media. This could be because chitosan precipitates at the pH of the SVF/SSF mixture, thus becoming a solid additive which is unable to control the release of the drug.
As can be seen in Figure 6
C, a controlled release of TFV for 96 h was obtained in SVF for batches containing HPMC, although the release from fd1H should only be considered until 72 h, since the very low increase in the percentage of drug released in the last 24 h would not be effective according to Karim et al. [72
]. Over 90% of the drug is released at 48 h for both batches. The f2
value indicates that the drug release profiles of fd1H and fd2H are similar, although fd2H shows significantly higher TFV release values than fd1H at 24 and 48 h. In the SVF/SSF mixture, fd1H releases the drug for 72 h while fd2H does so for 48 h, although both batches exceed 90% of the dose delivered at 24 h. No major differences can be established between these profiles, as confirmed by the f2
similarity factor. In terms of differences in TFV release according to the medium, a more controlled release of the drug is found in SVF than in the SVF/SSF mixture for both fd1H and fd2H until 48 h. This is corroborated by the f2
factor, which found no similarity between their profiles in one medium or the other. This could respond to the swelling results, such that the greater swelling of H batches in SVF could translate into greater control over the release of TFV than in the SVF/SSF mixture.
Based on these results, the batches containing pectin or HPMC are the best suited to the proposed objective, as they allow a controlled release of TFV in SVF that is accelerated in the presence of SSF. It should be highlighted that the fastest TFV delivery in the presence of SSF occurs in the case of fd2P and fd3P, whose profiles are not equivalent to the others according to the f2 data. This is supported by the high solubility of pectin at a pH equal to or higher than 7, as mentioned earlier. This demonstrates that the pH of the medium influences the behaviour of the pectin formulations, as we stated in the objective of this work. We can therefore confirm that these are pH-responsive systems that can be called “smart”, and would constitute a useful tool for preventing infection in women, since when intercourse takes place with the partner infected by HIV, the release of TFV from the freeze-dried bigel is accelerated in the presence of the seminal fluid—with a pH of around 7.5—thus increasing the ability to fight against the sexual transmission of the virus.
The main results obtained from fitting the drug release results to Korsmeyer-Peppas and Higuchi kinetics are shown in Table 2
. The TFV release profiles of all the batches can be said to have a good fit to both models.
The drug release profiles of batches containing pectin in SVF have the highest correlation coefficients for the Korsmeyer-Peppas kinetic. The value of the n exponent between 0.5 and 1.0 indicates an “anomalous transport” based on drug diffusion and structural modification of the dosage forms as the mechanism responsible for TFV release. The fact that n is closer to 0.5 than to 1.0, and the high correlation coefficient for the Higuchi model, point to the major role of the diffusion process in TFV delivery from fd2P and fd3P in this medium. In the SVF/SSF mixture, fd2P and fd3P best fit the Korsmeyer-Peppas kinetic model. n values close to and over 1.0 point to an “anomalous transport” and a “transport super-case II”; the main cause of TFV release from these formulations in this medium is their structural changes. This is supported by the values of KH from the Higuchi model, which are significantly higher in the SVF/SSF mixture than in SVF for these batches. The mechanisms responsible for the release of TFV from P batches are reflected in the swelling profiles, which in the SVF/SSF mixture are mostly weight loss by destructuration. This would also justify the slower drug release in SVF, where pectin forms a gel and diffusion and structural modification occur in the dosage form; and the faster delivery in the mixture, where pectin dissolves and there are mainly structural changes in the freeze-dried bigel.
In batches containing chitosan, fd2C has the highest correlation coefficient for the Higuchi kinetic in both SVF and the SVF/SSF mixture. The mechanism involved in the release of TFV from these freeze-dried bigels is diffusion. However, the profile in SVF also presents a good fit to Korsmeyer-Peppas. Its n value is close to but slightly higher than 0.5, showing that both diffusion and structural changes in the formulation induce the release of the drug in this medium. The same occurs for fd3C. This batch best fits the Higuchi model in the SVF/SSF mixture, and diffusion is again the process responsible for the drug release. In SVF, a high correlation coefficient was obtained for Higuchi but even higher for Korsmeyer-Peppas, with a value of n between 0.5 and 1.0. This indicates that both diffusion and structural changes cause the release of TFV from fd3C in this medium. The mechanisms explaining the release of TFV from batches containing chitosan can be deduced from the swelling profiles. Their weight decrease in SVF by dissolution of chitosan would explain the structural modification, and the absence of any weight decrease in the systems due to the insolubility of the polymer in the SVF/SSF mixture points to diffusion as the main mechanism of drug release.
In the case of H batches, the drug release profiles in SVF show a very good fit to the Korsmeyer-Peppas model. Since n is lower than 0.5, diffusion can be said to be the only causal mechanism of TFV release from fd1H and fd2H in this medium. The best fit for fd2H is observed for the Higuchi model, thus confirming the above. In the SVF/SSF mixture, diffusion is again the main cause of drug release, as the highest correlation coefficients were obtained for the Higuchi kinetic. This is supported by n values from Korsmeyer-Peppas of close to 0.5; fd1H and fd2H also had a good fit to this model in the SVF/SSF mixture. Nevertheless, these values of n are slightly higher than 0.5 and thus higher than those obtained for these batches in SVF. This indicates that the structural modification of fd1H and fd2H could play a role in the release of the drug in the SVF/SSF mixture, which could lead to less control over TFV delivery in this medium than in SVF. The release of the drug by diffusion is correlated with the swelling profiles of these batches, which show more sustained losses of structure after the maximum swelling value than in batches containing another polymer. The release profiles can also be explained by these mechanisms. The diffusion of the drug through the gel layer means the delivery of TFV from batches containing HPMC is more controlled in SVF than in those containing pectin or chitosan.
Regardless of the correlation coefficient values, the results of the statistical analysis by the AIC showed minimum values for Higuchi kinetics in all cases, indicating that diffusion is the mechanism that best explains the release of TFV from the freeze-dried bigels.