In vitro permeation of desmopressin across rabbit nasal mucosa from liquid nasal sprays: The enhancing effect of potassium sorbate

In vitro permeation of desmopressin across rabbit nasal mucosa from liquid nasal sprays: The enhancing effect of potassium sorbate

e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42 available at journal homep...

351KB Sizes 0 Downloads 18 Views

e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

available at

journal homepage:

In vitro permeation of desmopressin across rabbit nasal mucosa from liquid nasal sprays: The enhancing effect of potassium sorbate Bortolotti Fabrizio a , Balducci Anna Giulia a,b , Sonvico Fabio b , Russo Paola c , Colombo Gaia a,∗ a b c

Department of Pharmaceutical Sciences, University of Ferrara, I-Ferrara, Via Fossato di Mortara 17/19, 44100 Ferrara, Italy Department of Pharmacy, University of Parma, I-Parma, Italy Department of Pharmaceutical Sciences, University of Salerno, I-Fisciano, Italy

a r t i c l e

i n f o

a b s t r a c t

Article history:

Nasal spray products containing desmopressin acetate (DDAVP) were tested in vitro to evalu-

Received 23 October 2008

ate the effect of the contained preservatives on drug permeation across rabbit nasal mucosa.

Received in revised form

Experiments were performed using Franz-type diffusion cells with rabbit nasal mucosa as

20 December 2008

model barrier. Transport profiles obtained in comparison with a preservative-free solution

Accepted 22 December 2008

evidenced that in the presence of preservatives DDAVP permeation in vitro always increased

Published on line 31 December 2008

(p < 0.05), although at different extents (chlorobutanol < benzalkonium < sorbate).


rier properties, the effect of sorbate on drug transport was further investigated, being less


studied. After having found that sorbate permeated together with DDAVP, the hypothesis


that the two compounds formed an ion pair in solution with improved permeability was

Nasal spray

made. Additional experiments with aqueous test solutions reconstructed ad hoc contain-

Potassium sorbate

ing desmopressin and varying sorbate concentrations confirmed the enhancing effect of

While for benzalkonium structural damage of the mucosa could occur decreasing its bar-

sorbate, which however resulted to be independent of sorbate concentration. In conclusion, preservatives significantly enhanced desmopressin permeation in vitro across rabbit nasal mucosa with different mechanisms. If a correlation existed between these data and in vivo DDAVP bioavailability after nasal administration, this could strengthen the safety concerns related to the use of this medication in adults and children. © 2008 Elsevier B.V. All rights reserved.



Desmopressin (DDAVP), a synthetic nine-aminoacid cyclic peptide analogue of vasopressin, is used as antidiuretic drug both in adults and children for treating nocturnal enuresis and central diabetes insipidus (Cvetkovic and Plosker, 2005). Desmopressin is successfully administered via the nasal route, which guarantees higher bioavailability compared to

Corresponding author. Tel.: +39 0532 455909; fax: +39 0532 455953. E-mail address: [email protected] (C. Gaia). 0928-0987/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2008.12.015

the oral route, due to reduced enzymatic degradation. Worldwide, generic liquid formulations (nasal drops or sprays) are available for DDAVP administration to the nose, usually containing 0.05–0.1 mg/ml of desmopressin acetate together with preservatives, required to maintain product stability during storage and use. It has been shown that preservatives used in liquid nasal formulations cause adverse effects, such as nasal conges-

e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

tion, hypersensitivity reactions, itching and reduced ciliary beat frequency, in some cases affecting the integrity of the mucosa (Graf, 2001; Verse et al., 2003). The issue becomes even more relevant in case of prolonged or chronic use of these products. This explains the interest in developing nasal formulations without preservatives, such as solutions with alternative components or packaged in single-dose containers or powder formulations (Sacchetti et al., 2002; Bommer et al., 2004; Russo et al., 2006). More recently, safety issues have raised concerning intranasal desmopressin formulations, since especially in children an increased risk for developing severe hyponatremia and seizures was observed (Odeh and Oliven, 2001; Das et al., 2005). These side-effects would be related to uncontrolled drug bioavailability due either to the formulation itself or to its misuse/abuse. As such, for all desmopressin nasal products the regulatory agencies in Europe and the US have removed the indication for the treatment of primary nocturnal enuresis (PNE) and requested the manufacturers update the prescribing information ( Drug/infopage/desmopressin/default.htm). A few experimental studies focused on in vitro transport of desmopressin in order to assess the permeation parameters of the drug across a model mucosa (Law et al., 2001; Colombo et al., 2007). The aim of this research was to study the in vitro permeation of desmopressin from liquid nasal products across rabbit nasal mucosa focusing on the role of preservatives in affecting the peptide transport. Desmopressin in vitro transport from marketed nasal sprays was compared with that obtained from a preservative-free solution. In vitro drug permeation parameters across rabbit nasal mucosa were determined in order to anticipate possible differences in availability among commercially available products containing different preservatives. In fact, the tested nasal solutions all contained the same concentration of desmopressin acetate, but different preservatives, i.e., benzalkonium (BZK) chloride, chlorobutanol and potassium sorbate (PS). The effect of potassium sorbate on DDAVP transport was studied in detail.




Materials and methods

Pure desmopressin acetate (Arbeitsstandard, batch #L00022623) was kindly supplied by Gebro Pharma GmbH (A-Fieberbunn) and used as analytical standard and for test solution preparation. All other chemicals used were of analytical grade (Carlo Erba Reagenti SpA, I-Rodano). Three marketed nasal formulations (sprays) were used as received and stored in the conditions indicated by manufacturers. These products were: Product #1: desmopressin 0.1 mg/ml and benzalkonium chloride (pH 5.2), (Nocutil nasal spray, batch #351408), Product #2: desmopressin 0.1 mg/ml and chlorobutanol (pH 4.7) (Desmogalen, batch #3481), and Product #3: desmopressin 0.1 mg/ml and potassium sorbate (pH 4.8) (Desmopressin-acetaat Dumex, batch #359781).


Test nasal solutions in 50 mM sodium acetate buffer (pH 5.0) were also prepared, all containing desmopressin at the same concentration as in the marketed solutions and different potassium sorbate concentrations: Solution PS-A: DDAVP 0.1 mg/ml, without potassium sorbate, pH 5.0, Solution PS-B: DDAVP 0.1 mg/ml + 0.05% (w/v) potassium sorbate, pH 5.0 (molar ratio 1:40), Solution PS-C: DDVAP 0.1 mg/ml + 0.1% (w/v) potassium sorbate, pH 5.0 (molar ratio 1:80), Solution PS-D: DDVAP 0.1 mg/ml + 0.2% (w/v) potassium sorbate, pH 5.0 (molar ratio 1:160), and Solution PS-E: contained 0.1% (w/v) of potassium sorbate without desmopressin, pH 5.0. For these solutions, the pH value chosen was close to the values of the commercial solutions tested.


In vitro diffusion experiments

Rabbit nasal mucosa was selected as permeation tissue (Sacchetti et al., 2002; Russo et al., 2006). On the day of the experiment, rabbit heads were collected from a local slaughterhouse (Pola, I-Finale Emilia) and specimens of nasal mucosa were dissected within 2 h from the animal death. Briefly, upon incision of the nasal bone and exposition of the nasal cavity, the nasal septum was extracted and the mucosa layers were carefully detached from the septum cartilage by means of a pair of tweezers. The specimens removed (surface area 0.7–1 cm2 ; average thickness 100 ␮m) were rinsed with phosphate buffered saline (PBS) pH 7.4 and immediately inserted between the donor and receptor compartments of a vertical diffusion cell, with the mucosal side facing the donor. Franz type diffusion cells (0.58 cm2 permeation area) were used (VETROTECNICA S.r.l., I-Padova). In order to assess the proper cell assembly as well as the mucosa integrity, the donor compartment was filled with saline solution, checking that no liquid passed to the empty receptor due to inappropriate mounting or lack of tissue integrity. Afterwards, the receptor compartment was filled with PBS pH 7.4 and the assembled system was allowed to equilibrate at 37 ◦ C for half an hour. Then, after having removed the saline from the donor, 0.4 ml of each test formulation (corresponding to 40 ␮g of desmopressin acetate, i.e., the maximum daily dose in humans) was introduced in the donor compartment, which was closed with a screw-cap to prevent evaporation. The receptor solution was magnetically stirred to avoid any boundary layers effect. All experiments were carried out over a 4-h period of time. At predetermined time-points, samples were withdrawn from the receptor compartment and either immediately analyzed for desmopressin content or frozen at −20 ◦ C until the analysis. At the end of the experiment, the donor solution was quantitatively collected and assayed in order to measure the residual desmopressin amount and calculate the mass balance. Afterwards, the donor compartment was filled with a 1% (w/v) crystal violet aqueous solution in order to assess the integrity of the mucosa based on dye leakage into the receptor.


e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

All experiments were replicated at least three times; results are expressed as the means ± SEM. The content of desmopressin in the samples was determined by reverse phase-HPLC. Isocratic elution was carried out with a mixture (v/v) of 25% acetonitrile and 75% water containing 0.08% (v/v) trifluoroacetic acid. The detection wavelength was set at 220 nm. The column used was a Hypersil BDS C18 5U, 150 mm × 4.6 mm (Alltech, I-Milano) equipped with a Hypersil BDS C18 guardcartridge (Alltech, I-Milano). The flow rate was 1 ml/min and the temperature was ambient. The injection volume was 40 ␮l. In these conditions, the retention time of desmopressin acetate was between 3.5–4.5 min. The method was validated for linearity (R2 = 0.994), repeatability (RSD 0.01%, n = 6 injections) and limit of quantification (0.05 ␮g/ml). The transport parameters, i.e., steady-state flux and permeability of desmopressin across the membrane, were calculated (Xiang et al., 2002) according to the steady-state solution of Fick Equation (1): Jss =

dM 1 = PeC dt A


where Pe is the apparent permeability coefficient of diffusant (cm s−1 ), C is the initial donor concentration and JSS is the flux at steady-state (␮g s−1 cm−2 ); dM is the amount of drug (␮g) transported through the membrane during the infinitesimal time dt and A is the diffusion area (cm2 ). The permeability coefficient across the mucosa was calculated from the slope of the linear part of the line obtained by plotting mass transported per unit area against time. Pe is related to the diffusion coefficient D (cm2 s−1 ) and mucosa-medium partition coefficient K by Eq. (2): Pe =

DK h


the DDAVP concentration were detected, thus confirming that no peptide degradation occurred in this medium during the experiment time. All experiments were completed within 6 h from the animal death to limit mucosa alteration due to prolonged hypoxia. A preliminary experiment was performed to compare DDAVP permeation across freshly excised mucosa and specimens stored at 2–4 ◦ C for 24 h. The results confirmed that the amount of drug permeated per unit area was significantly lower in the case of the refrigerated mucosa (data not shown). The transport profiles obtained for the three commercial nasal sprays in comparison with the preservative-free desmopressin solution (Solution PS-A) are illustrated in Fig. 1. The preservative’s presence increased the amount of drug transported already after 30 min in comparison with the preservative-free solution. In addition, 1 h after the application differences among the various preservatives became evident. Regression analysis of experimental data was performed between 1 and 4 h in order to approximate the permeation parameters at steady-state, although such a long time is not realistic in vivo given the short residence time of aqueous liquids in the nasal cavity (Merkus et al., 1998). Product #1 contained 0.1% (w/v) benzalkonium chloride, a compound commonly used to prevent microbial growth in nasal and ophthalmic aqueous solutions. After 1 h, the permeation profile became rather linear (Eq. (1): y = 0.52821 + 0.97214x, R2 = 0.98481), indicating that quasi steady-state drug diffusion was attained. In 4 h the cumulative amount of DDAVP permeated per unit area from this product was 4.31 ± 0.54 ␮g cm−2 that corresponded to about 6% in weight of the desmopressin amount loaded into the donor at time zero (40 ␮g). The calculated apparent permeability coefficient was equal to (2.94 ± 0.72) × 10−6 cm s−1 . As reference, a preservative-free desmopressin solution 0.1 mg/ml in acetate buffer at pH 5 (Solution PS-A) was

where h is the mucosa thickness (cm). Statistical analysis of permeation parameters was performed by applying an unpaired two-tailed Student’s t-test. Significance was accepted at p < 0.05.


Results and discussion

The nose represents an attractive route for systemic drug delivery, particularly in the case of small peptides owing to limited enzymatic activity. Three marketed nasal spray products containing 0.1 mg/ml of desmopressin acetate were tested in vitro in order to know the extent of peptide permeation across the nasal mucosa. The experimental conditions for the permeation study, derived from the literature (Law et al., 2001), were optimized. In particular, receptor solution (pH, volume), peptide stability in such solution and freshness of the mucosa were considered as the relevant parameters. Thus, PBS pH 7.4 was selected as receptor solution to maintain the mucosa in physiological conditions, despite the fact that desmopressin appears to be more stable at pH values between 4.5 and 5.5 (Law et al., 2003). The stability of the peptide in the chosen medium was checked by incubating a DDAVP solution in PBS pH 7.4 for 4 h at 37 ◦ C. No significant changes in

Fig. 1 – DDAVP permeation profiles from Product #1 containing benzalkonium chloride (empty circle, n = 6), Product #2 containing chlorobutanol (empty triangles, n = 9) and Product #3 containing potassium sorbate (empty squares, n = 7) in comparison with Solution PS-A (full circle, n = 4). Data are expressed as the mean ± SEM. Experimental points and theoretical straight lines according to Eq. (1).


e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

tested in the same conditions. The cumulative amount of drug permeated per unit area from this solution was 1.04 ± 0.10 ␮g cm−2 , i.e., almost one fourth of that transported in the presence of benzalkonium. The apparent permeability coefficient was also lower and equal to (0.78 ± 0.09) × 10−6 cm s−1 . Given that the main difference between the two products was the presence of benzalkonium, it was deducted that the increased desmopressin diffusion across the nasal mucosa was due to the preservative. The amphiphilic structure of this molecule accounts for it acting as a surfactant on the mucosa. Several studies in literature (van de Donk et al., 1982; Cho et al., 2000; Merkus et al., 2001; Lebe et al., 2004) have shown how, in case of chronic use of BZK-containing solutions, this preservative caused structural and functional damage (namely ciliary beat impairment) of the nasal mucosa, decreasing its barrier properties toward drug permeation. More recently, an in vitro study by Ho et al. (2008) on human nasal epithelial cells showed that clinical nasal preparations containing BZK severely damaged the cells, eventually leading to the lysis of cell membranes. Hence, the enhanced desmopressin diffusion across the nasal mucosa observed in our in vitro experiments was likely due to an impairment of the barrier damaged by benzalkonium chloride. The second nasal spray examined (Product #2) contained 0.5% (w/v) chlorobutanol as preservative. The cumulative amount of drug permeated per unit area from this product after 4 h was 2.10 ± 0.33 ␮g cm−2 (Fig. 1). This amount was significantly lower than the one obtained with benzalkonium (p < 0.05). The apparent permeability coefficient was (1.21 ± 0.21) × 10−6 cm s−1 , i.e., about half the value calculated for the benzalkonium-containing product. Therefore, even if less than benzalkonium, chlorobutanol enhanced DDAVP transport across nasal mucosa in vitro. In fact, the DDAVP cumulative transport for the chlorobutanol-containing product was twice as higher as for the preservative-free solution, despite the permeability coefficients at steady-state were not significantly different (p > 0.05), indicating that the mucosa barrier was not impaired. Therefore, the observed higher cumulative amount permeated was due to more DDAVP transported during the first hour, before quasi steady-state was attained. Few data are available in the literature about alterations of the nasal mucosa induced by this compound. An ex vivo study by Merkus et al. (2001) on the effect of nasal medications on ciliary beat frequency showed that chlorobutanol contained in a nasal spray (Minirin® ) exhibited a ciliostatic effect. The third product (Product #3) evaluated in vitro for desmopressin transport contained 0.1% (w/v) potassium sorbate as preservative. As Fig. 1 shows, during the first 90 min of

experiment the permeation profile in presence of potassium sorbate was almost completely superposed to that with benzalkonium. However, after this initial similar behavior, the profile for the potassium sorbate-containing product became steeper, reaching a significantly higher cumulative amount of 6.06 ± 0.64 ␮g cm−2 permeated in 4 h. The apparent permeability coefficient in the presence of sorbate was also higher [(4.41 ± 0.52) × 10−6 cm s−1 ] than that obtained for the benzalkonium-containing product. In the same in vivo studies evaluating the toxicity of benzalkonium chloride to the nasal mucosa of rats, potassium sorbate was described as responsible of nasal irritation (sneezing, rubbing) and structural damage of the respiratory epithelium (Cho et al., 2000; Lebe et al., 2004). However, the more recent study by Ho et al. (2008) comparing the effects of benzalkonium and potassium sorbate in vitro on human nasal epithelial cells showed no significant cell damage in the sorbate group even at higher concentrations than clinically used (Ho et al., 2008). In summary, given that these three products are marketed as generic desmopressin nasal sprays in the same country, they are supposed to be bioequivalent and interchangeable. Obviously, bioequivalence cannot be demonstrated in vitro, but the experimental set-up adopted here showed non-similar diffusion profiles for the three products. These differences were definitely linked to the presence and type of preservative, even though in the case of benzalkonium vs sorbate they were not immediately evident in the early times of diffusion. At the end of all experiments, the mass balance was calculated knowing the initial amount of drug loaded into the donor (40 ␮g) and the drug concentration in both receptor and donor compartments at the end of the experiment. Table 1 shows the data for each commercial product expressed as percentage of the corresponding amounts. In all cases it was impossible to recover 100% of the drug introduced in the donor compartment at the beginning of the experiment. Even taking into account the drug remaining inside the mucosa (not measured), this would be a too small quantity to justify about 40–50% in weight of missing DDAVP. An explanation for this result is that degradation of the peptide structure of desmopressin occurred during the experiment. Degradation would be caused by proteolytic enzymes of the rabbit mucosa, as previously reported (Jonsson et al., 1992). Indeed, during HPLC analysis of permeation samples several secondary peaks were detected in the chromatographic patterns that could be attributed to degradation products. Based on the relative position of these peaks with respect to the DDAVP peak, the degradation products were expected to be more hydrophilic compounds than DDAVP.

Table 1 – DDAVP mass balance at the end of transport experiments with marketed nasal sprays and Solution PS-A (preservative-free) as reference. Data are expressed as mean ± SEM (n = 4–9). Product Solution PS-A (preservative-free) Spray #1 (benzalkonium) Spray #2 (chlorobutanol) Spray #3 (potassium sorbate)

DDAVP permeated in 4 h (%) 1.5 6.2 3.0 8.8

± ± ± ±

0.1 0.8 0.5 0.9

DDAVP recovered from donor (%) 55.9 43.4 45.2 54.2

± ± ± ±

6.0 0.5 3.7 6.0

Mass balance (%) 58.1 49.7 48.2 62.9

± ± ± ±

4.1 0.5 3.9 6.1


e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

It was not in the intent of this study to identify and quantify DDAVP degradates, nevertheless a simple experiment was carried out to demonstrate that degradation depended on contact between the DDAVP solution and the mucosa. A 0.1 mg/ml DDAVP solution in 50 mM acetate buffer pH 5 was loaded into both donor and receptor compartments of diffusion cells, using as barrier either rabbit nasal mucosa or a regenerated cellulose artificial membrane (MW cut-off 12,000–14,000 Da, tubing width 32–34 mm, Dexstar Visking, Medicell International Ltd., London, UK). The cells (n = 4) were incubated at 37 ◦ C for 4 h, sampling both compartments at the end of the experiment for DDAVP assay. We found that the concentration of the loaded DDAVP solution did not change in the compartments of the cells equipped with the artificial membrane. Moreover, the corresponding HPLC traces did not present peaks other than that from DDAVP. In contrast, in the cells equipped with the biological membrane, a decrease in DDAVP concentration was observed in both donor and receptor of about 10–30% with respect to the initial concentration. This range was lower than that observed in the actual permeation experiments, but could be explained considering that here no diffusion was expected as DDAVP concentration was equal on both sides of the barrier. Since the mechanism of enhancement of drug transport across nasal mucosa of potassium sorbate was less known, the effect of sorbate on in vitro DDAVP transport was investigated. In addition, during the previous experiments it was found that a significant amount of potassium sorbate was transported across the mucosa at the same time as desmopressin. In fact, 74.4 ± 3.0% of potassium sorbate initially present in the donor solution (0.4 mg) diffused into the receptor compartment in 4 h. Given that potassium sorbate is the salt of a short straight-chain fatty acid (2–4 hexadienoic acid), a hypothesis was made that sorbate and desmopressin could form an ion pair in solution with improved permeability. It must be highlighted here that in the commercial sprays desmopressin is in the form of acetate salt, i.e., the peptide is able to form salts with organic acids. This is due to the presence of a d-Arg block in the peptide chain, which is positively charged at pH 5 (pKa > 11). In the literature, the effect of sorbic acid as a counter ion in an aqueous ophthalmic solution was described with respect to the in vivo absorption of timolol maleate upon ocular instillation. In that case, the formation of the ion pair led to improved ocular bioavailability (Higashiyama et al., 2004, 2007).

Fig. 2 – Effect of potassium sorbate on in vitro DDAVP permeation from: 0% (w/v) sorbate, Solution PS-A (full circle, n = 4); 0.05% (w/v) sorbate, Solution PS-B (empty triangle, n = 4), 0.1% (w/v) sorbate, Solution PS-C (empty square, n = 4); 0.2% (w/v) sorbate, Solution PS-D (empty circle, n = 5). Data are expressed as mean ± SEM. Experimental points and theoretical straight lines according to Eq. (1).

Then, additional permeation studies in the presence of sorbate were conducted using different aqueous test solutions (PS-A, PS-B, PS-C, PS-D and PS-E). The all of the first four contained 0.1 mg/ml of desmopressin acetate with potassium sorbate in concentrations ranging from 0 to 0.2% (w/v), whereas the fifth (Solution PS-E) contained only 0.1% (w/v) potassium sorbate and no desmopressin. The results of these transport experiments are shown in Fig. 2. With respect to desmopressin permeation in vitro, the penetration enhancing effect attributed to potassium sorbate was confirmed. For both DDAVP and potassium sorbate, Table 2 summarizes flux values and apparent permeability coefficients (Pe) calculated according to Eq. (1) for each test solution considered. As all test solutions had the same initial DDAVP concentration, likely a change in drug permeability was responsible for the significantly different drug transports in dependence of the presence of sorbate. The amounts permeated from these test solutions were lower than those observed in the experiments with the sorbate-containing nasal spray product, but

Table 2 – Permeation parameters for desmopressin (DDAVP) and potassium sorbate (PS) from experiments with sorbate test solutions (data are expressed as mean ± SEM, n = 3–5). Desmopressin Flux (␮g cm−2 h−1 ) Solution PS-A (0% PS) Solution PS-B (0.05% PS) Solution PS-C (0.1% PS) Solution PS-D (0.2% PS) Solution PS-E (0.1% PS, no DDAVP) ∗

Not significantly different (p > 0.05).

0.28 1.06 0.79 0.88 –

± ± ± ±

0.03 0.20 0.04 0.13

Potassium sorbate

Pe × 10−6 (cm s−1 ) 0.78 2.96 2.19 2.44 –

± ± ± ±

0.09 0.55* 0.10* 0.35*

Flux (mg cm−2 h−1 ) – 0.038 ± 0.091 ± 0.171 ± 0.085 ±

0.008 0.009 0.010 0.002

Pe × 10−6 (cm s−1 ) – 2.11 ± 2.52 ± 2.37 ± 2.37 ±

0.43* 0.24* 0.14* 0.61*

e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

this could be ascribed to the different solvent used in the test solutions. In addition, varying the concentration of potassium sorbate from 0.05% to 0.2% (w/v), the DDAVP permeation profiles were not significantly different (p > 0.05). Considering the molar ratio between desmopressin and sorbate in these test solutions, sorbate was always in large excess (1:40, 1:80 and 1:160 for 0.05%, 0.1% and 0.2%, w/v, respectively). Then, if an ion pair was formed between DDAVP and sorbate, this would occur in the same way at all sorbate concentrations tested, justifying the actual no difference in DDAVP transport in vitro in dependence of sorbate concentration. Similar to what has already been observed with the commercial products, in this second set of experiments a 100% mass balance for DDAVP was never obtained, with a percentage of DDAVP recovered at the fourth hour around 50–60% of the initial loading. On the other hand, while determining desmopressin transport, the permeation profile of potassium sorbate was also followed in comparison with a test solution (Solution PS-E) containing 0.1% (w/v) sorbate without desmopressin (Fig. 3). As Fig. 3 shows, sorbate permeation profiles were rather linear for all test solutions considered and their slopes increased with the concentration of potassium sorbate in the donor. Table 2 also reports the permeation parameters obtained for potassium sorbate in the adopted experimental conditions, where Pe (apparent permeability coefficient) was calculated according to Eq. (1). The measured fluxes of potassium sorbate were directly correlated with the preservative’s concentration in the donor compartment and no significant differences were found in the permeability coefficients calculated for each test solution (p > 0.05). Moreover, by comparing the permeation parameters for the two solutions containing 0.1% (w/v) potassium sorbate


with or without desmopressin, it was clear that the presence of the peptide did not influence sorbate permeation. Based on mass balance calculations, we found that on average about 50% in weight of potassium sorbate diffused across the mucosa in 4 h, but this should not raise toxicity issues as the compound is commonly used as food preservative. In summary, from the analysis of three nasal sprays commercially available for antidiuretic therapy, it can be concluded that benzalkonium chloride and potassium sorbate significantly enhanced the in vitro permeation of the peptide drug desmopressin acetate across rabbit nasal mucosa, whereas chlorobutanol was less efficient. If the same behavior occurred after nasal administration in vivo, one could expect to see differences in drug absorption, raising safety concerns as desmopressin is a potent drug. In this regard, several clinical reports have been published recently in which severe hyponatremia and water intoxication were attributed to desmopressin prolonged half-life and/or too high bioavailability following intranasal administration in patients with nocturnal enuresis (Odeh and Oliven, 2001; Dehoorne et al., 2006). The issue has become even more relevant since primary nocturnal enuresis is a frequent childhood condition, for which desmopressin is a common medication. Indeed, if in the past intranasal desmopressin has been the easiest choice for treating young children, now paediatricians have started to highlight that oral administration might be safer (Robson et al., 2007). Clearly, we found that the three commercial products and the preservative-free test solution here evaluated performed differently in vitro. This should not be taken as a proof of non-bioequivalence in vivo, especially because our data refer to experiments in which the solution remained in contact with the nasal mucosa for a relatively long time. This situation is rather unreal in vivo, where the cilia can clear an aqueous solution away from the nasal cavity in 20–30 min in physiological conditions and this even when the presence of the preservative could compromise the effectiveness of the mucociliary clearance. Hence, focusing on the first 30–60 min of the drug diffusion profiles obtained in vitro, there is basically no difference among the four solutions and this period of time is the most likelihood for drug absorption in vivo. In this regard, in a case study for testing bioequivalence of two desmopressin nasal sprays (Joukhadar et al., 2003), despite the relatively high inter-subject variability, an average time to maximum plasma concentration (tmax ) of about 40 min was measured.


Fig. 3 – In vitro permeation of potassium sorbate from: 0.05% (w/v) potassium sorbate, Solution PS-B (empty triangle, n = 4); 0.1% (w/v) potassium sorbate, Solution PS-C (empty square, n = 3); 0.2% (w/v) potassium sorbate, Solution PS-D (empty circle, n = 5); 0.1% (w/v) potassium sorbate without DDAVP, Solution PS-E (full circle, n = 3). Data are expressed as mean ± SEM.

Bommer, R., Kern, J., Hennes, K., Zwisler, W., 2004. Preservative-free nasal drug-delivery systems. Med. Device Technol. 15, 14–16, 18. Cho, J.H., Kwun, Y.S., Jang, H.S., Kang, J.M., Won, Y.S., Yoon, H.R., 2000. Long-term use of preservatives on rat nasal respiratory mucosa: effects of benzalkonium chloride and potassium sorbate. Laryngoscope 110, 312–317. Colombo, P., Cagnani, S., Sonvico, F., Santi, P., Russo, P., Colombo, G., 2007. Biological in vitro models for absorption by nonoral routes. In: Taylor, J.B., Triggle, D.J. (Eds.), Comprehensive Medicinal Chemistry II. Elsevier Ltd., Oxford, UK, pp. 279–299.


e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 7 ( 2 0 0 9 ) 36–42

Cvetkovic, R.S., Plosker, G.L., 2005. Desmopressin: in adults with nocturia. Drugs 65, 99–107 (discussion 108–109). Das, P., Carcao, M., Hitzler, J., 2005. DDAVP-induced hyponatremia in young children. J. Pediatr. Hematol. Oncol. 27, 330–332. Dehoorne, J.L., Raes, A.M., van Laecke, E., Hoebeke, P., Vande Walle, J.G., 2006. Desmopressin toxicity due to prolonged half-life in 18 patients with nocturnal enuresis. J. Urol. 176, 754–757 (discussion 757–758). Graf, P., 2001. Benzalkonium chloride as a preservative in nasal solutions: re-examining the data. Respir. Med. 95, 728–733. Higashiyama, M., Inada, K., Ohtori, A., Kakehi, K., 2007. NMR analysis of ion pair formation between timolol and sorbic acid in ophthalmic preparations. J. Pharm. Biomed. Anal. 43, 1335–1342. Higashiyama, M., Inada, K., Ohtori, A., Tojo, K., 2004. Improvement of the ocular bioavailability of timolol by sorbic acid. Int. J. Pharm. 272, 91–98. Ho, C.Y., Wu, M.C., Lan, M.Y., Tan, C.T., Yang, A.H., 2008. In vitro effects of preservatives in nasal sprays on human nasal epithelial cells. Am. J. Rhinol. 22, 125–129. Jonsson, K., Alfredsson, K., Soderberg-Ahlm, C., Critchley, H., Broeders, A., Ohlin, M., 1992. Evaluation of the degradation of desaminol,d-arginine8-vasopressin by nasal mucosa. Acta Endocrinol. (Copenh.) 127, 27–32. Joukhadar, C., Schenk, B., Kaehler, S.T., Kollenz, C.J., Bauer, P., Muller, M., Eichler, H.G., 2003. A replicate study design for testing bioequivalence: a case study on two desmopressin nasal spray preparations. Eur. J. Clin. Pharmacol. 59, 631–636. Law, S.L., Huang, K.J., Chou, H.Y., 2001. Preparation of desmopressin-containing liposomes for intranasal delivery. J. Control. Release 70, 375–382. Law, S.L., Huang, K.J., Chou, V.H., 2003. Stability of desmopressin loaded in liposomes. J. Liposome Res. 13, 269–277. Lebe, E., Baka, M., Yavasoglu, A., Aktug, H., Ates, U., Uyanikgil, Y., 2004. Effects of preservatives in nasal formulations on the

mucosal integrity: an electron microscopic study. Pharmacology 72, 113–120. Merkus, F.W., Verhoef, J.C., Schipper, N.G., Marttin, E., 1998. Nasal mucociliary clearance as a factor in nasal drug delivery. Adv. Drug Deliv. Rev. 29, 13–38. Merkus, P., Romeijn, S.G., Verhoef, J.C., Merkus, F.W., Schouwenburg, P.F., 2001. Classification of cilio-inhibiting effects of nasal drugs. Laryngoscope 111, 595–602. Odeh, M., Oliven, A., 2001. Coma and seizures due to severe hyponatremia and water intoxication in an adult with intranasal desmopressin therapy for nocturnal enuresis. J. Clin. Pharmacol. 41, 582–584. Robson, W.L., Leung, A.K., Norgaard, J.P., 2007. The comparative safety of oral versus intranasal desmopressin for the treatment of children with nocturnal enuresis. J. Urol. 178, 24–30. Russo, P., Sacchetti, C., Pasquali, I., Bettini, R., Massimo, G., Colombo, P., Rossi, A., 2006. Primary microparticles and agglomerates of morphine for nasal insufflation. J. Pharm. Sci. 95, 2553–2561. Sacchetti, C., Artusi, M., Santi, P., Colombo, P., 2002. Caffeine microparticles for nasal administration obtained by spray drying. Int. J. Pharm. 242, 335–339. van de Donk, H.J., van den Heuvel, A.G., Zuidema, J., Merkus, F.W., 1982. The effects of nasal drops and their additives on human nasal mucociliary clearance. Rhinology 20, 127–137. Verse, T., Sikora, C., Rudolph, P., Klocker, N., 2003. The tolerability of nasal drugs with special regard to preservatives and physico-chemical parameters. Laryngorhinootologie 82, 782–789. Xiang, J., Fang, X., Li, X., 2002. Transbuccal delivery of 2 ,3 -dideoxycytidine: in vitro permeation study and histological investigation. Int. J. Pharm. 231, 57–66.