A study on the chemical stability of cholesterol-lowering drugs in concomitant simple suspensions with magnesium oxide

Abstract

Background

Difficulty in taking solid medicines is a common issue particularly for the elderly because of a decline in swallowing function, also known as dysphagia. For patients with such a dysfunction, a simple suspension method, in which solid medicines are disintegrated and suspended using warm water, has been developed and widely used in Japanese clinical settings. However, there is little information on drug stability in the simple co-suspension of multiple formulations especially including acidic or alkaline ones. In this study, the chemical stability of typical cholesterol-lowering drugs was investigated in a simple co-suspension with alkaline magnesium oxide (MgO) which is frequently used as a laxative or antacid in Japan.

Methods

A cholesterol-lowering drug (one tablet) was soaked with or without MgO in warm water (55°C), and the vessel was left at room temperature for 10 min or 5 h. The suspensions prepared were then analyzed by high-performance liquid chromatography. Degradation products were analyzed by nuclear magnetic resonance spectroscopy and mass spectrometry for the structural elucidation.

Results

In the simple co-suspension with MgO, no significant degradation was observed for atorvastatin or pravastatin, while a significant decrease of the recovery from the co-suspension was observed for rosuvastatin after 5 h. On the other hand, simvastatin and ezetimibe co-suspended with MgO were partially degraded to simvastatin acid and a pyran compound, respectively.

Conclusions

A simple co-suspension with MgO is feasible for atorvastatin, pravastatin, and rosuvastatin, although the rosuvastatin tablet should not be left soaking for a long time. Further it is inadvisable to suspend simvastatin or ezetimibe together with MgO because of their partial degradation.

Background

A decline in swallowing function is a common issue in the elderly population and it makes it difficult for some patients to ingest solid oral medicines as well as food [1, 2]. This swallowing dysfunction, so-called dysphagia, can also occur in patients with various acute and chronic diseases such as stroke, depression, dementia, and structural or functional abnormalities of the oropharynx. When solid oral medicines are inapplicable for the patients with swallowing dysfunction, other suitable formulations or alternative routes of administration are required. However, such medications are not available in most cases. Thus, it is customary practice to prepare aqueous suspensions of tablets and capsules followed by tube administration [3, 4].

The suspensions have been commonly prepared by a crushing method where tablets are pulverized and capsules are opened. However, there have been reports of issues with the method such as loss of drugs or exposure of clinical staff to drugs [5,6,7]. Thus, a simple suspension method (SSM) has been developed in order to solve these issues [8]. In the SSM, tablets and capsules are soaked in warm water (55°C) and the vessels are left at room temperature for 10 min, by which the formulations are disintegrated and easily suspended after mixing. The suspension formation is facilitated in the warm water, which reduces the risk of clogging feeding tubes. The feasibility of SSM has been extensively tested with respect to ease of suspension and passage through feeding tubes, and found to be applicable to most solid oral medicines [8,9,10,11]. The method is simple and convenient and thus widely used in Japanese clinical settings. However, concerns about the chemical stability of drugs in the SSM have not been eliminated [8]. In particular, little is known about the chemical incompatibility in the concomitant simple suspensions (hereafter simply called ‘co-suspension’) of plural medicines. For example, some drugs may degrade when they are subjected to the SSM with acidic or alkaline drugs. This is an important point because elderly patients usually ingest many kinds of drugs at the same time.

Magnesium oxide (MgO) has been widely used as a laxative as well as an antacid in East Asia [12]. The use of the drug has become more frequent due to an increase in the number of elderly patients with chronic constipation in an aging society [13]. Thus, MgO is often used together with other drugs related to life-style diseases as well as common diseases. MgO is slightly soluble in water to show basicity, and magnesium ions are known to show some interactions such as complex formation with various drugs [14]. Thus, it is necessary to evaluate the chemical stability of drugs concomitantly suspended with MgO under an SSM condition. Aoki et al. [15] reported that no degradation occurred in the co-suspension of MgO and clopidogrel bisulfate, which has an ester moiety. On the contrary, Suryani et al. [16] reported that when ester prodrugs, acemetacin and cefpodoxime proxetil, were subjected to the SSM with MgO, the ester bonds were hydrolyzed. These chemical changes were considered to be serious because of the potential occurrence of gastrointestinal injury or some reduction of drug efficacy.

Cholesterol-lowering drugs (CLDs) such as statins and ezetimibe are most effective for the treatment of dyslipidemia, which is one of the major risk factors for cardiovascular disease [17, 18]. This disease is particularly prevalent in the elderly. Due to the large number of dyslipidemic patients, use of the drugs has become more frequent, which increases the probability of the simultaneous prescription with MgO. Most CLDs possess partial structures, which are susceptible to chemical changes. Their chemical stability under alkaline conditions has been previously investigated [19,20,21,22,23,24,25,26,27,28]. However, the situation is different under an SSM condition with MgO because the drugs are not always fully soluble under the condition, and various additives contained in their formulations may influence chemical stability. Thus, in this paper, the chemical stability of typical CLDs concomitantly suspended with MgO under an SSM condition was investigated.

Methods

Reagents

The following formulations of CLDs were used: Atorvastatin tablets 5 mg (Sawai, Osaka, Japan), Pravastatin sodium tablets 5 mg (Sawai, Osaka, Japan), Rosuvastatin tablets 2.5 mg (DSEP, Tokyo, Japan), Simvastatin tablets 5 mg (Teva Takeda Yakuhin, Aichi, Japan), Ezetimibe tablets 10 mg (DSEP, Tokyo, Japan). Magmitt tablets 330 mg (Nihon Shinyaku, Kyoto, Japan) were used as an MgO containing formulation. The tablets of atorvastatin, pravastatin, rosuvastatin, simvastatin, ezetimibe, and MgO are hereafter referred to as AS, PS, RS, SS, EZ, and MG, respectively. Atorvastatin calcium hydrate (99.8%), simvastatin (100%), and ezetimibe (96.3%) were purchased from Tokyo Chemical Industry (Tokyo, Japan). Pravastatin sodium (99.9%), rosuvastatin calcium (99.7%), distilled water, acetonitrile, methanol, trifluoroacetic acid, and sodium dihydrogen phosphate dihydrate were purchased from Fujifilm Wako Chemicals (Osaka, Japan). Distilled water, acetonitrile, and methanol were of high-performance liquid chromatography (HPLC) grade. An aqueous solution of phosphoric acid (0.5 M) and chloroform were purchased from Kanto Chemical (Tokyo, Japan). Weakly acidic cation exchange resin AMBERLITE® (IRC76) was purchased from ORGANO (Tokyo, Japan). A membrane filter DISMIC®—03 JP (0.50 μm) was purchased from Toyo Roshi Kaisha (Tokyo, Japan). Deuterated chloroform (CDCl₃) containing 0.05% tetramethylsilane was purchased from Eurisotop (Massachusetts, USA). Deuterated dimethyl sulfoxide (DMSO-d₆) containing 0.03% tetramethylsilane was purchased from Acros Organics (Massachusetts, USA).

Preparation of simple suspensions

One CLD tablet was placed with or without MG (one tablet) in a 50-mL plastic centrifuge tube containing 20 mL of warm distilled water (55°C). The centrifuge tube was allowed to stand at room temperature (ca. 25°C) for 10 min or 5 h. The mixture was then vortexed for 1 min to prepare a suspension. The subsequent treatments were immediately performed according to method I or II (see below). In separate experiments, the simple suspension prepared after 10 min soakage was centrifuged and the aqueous supernatant was directly subjected to HPLC analyses and pH measurements using InLab® Routine Pro (Mettler Toledo, Ohio, USA). The suspension preparation was performed three times for each CLD tablet in any of the experiments.

Treatment of simple suspensions

The simple suspensions were treated as follows, and the resulting solutions were analyzed by HPLC.

Method I: The simple suspension was completely transferred into a 100-mL volumetric flask, where the tube was washed by methanol (for AS, PS, RS, and EZ) or acetonitrile (for SS) and the washing solvent was put into the same flask. The mixture in the flask totaled 100 mL with the respective organic solvents, followed by sonication for 1 min unless otherwise noted.

Method II: For SS and EZ, the simple suspension was centrifuged, and the aqueous supernatant (18 mL) was neutralized by passing it through a column of the cation exchange resin. Eighteen milliliters of acetonitrile for SS or methanol for EZ were added to the residue of centrifugation, and the mixture was vortexed for 1 min and centrifuged again. The supernatant was also passed through the same cation exchange column. The same treatment was conducted again for the residue of the centrifugation. The effluents from the column were collected and totaled 250 mL by the respective organic solvents.

HPLC analysis

HPLC analyses were performed using a Shimadzu HPLC system (Kyoto, Japan) as described in our previous paper [29]. The reversed-phase HPLC analyses were performed modifying the conditions in the literature [30]. An Atlantis T3 column (4.6 × 250 mm, 5 μm; Waters, Milford, USA) fitted with a guard column (4.6 × 20 mm, 5 μm) was used. The mobile phases and detection wavelengths used are summarized in Table 1. The HPLC system was operated isocratically at a flow rate of 1.0 mL/min at 25°C. The samples were appropriately diluted and injected after being filtered using the membrane filter. The injection volume of the sample was 10 μL. Chromatograms were obtained within 30 min after the suspension preparation.

Table 1 Reversed-phase HPLC conditions

Validation method of HPLC analysis

Standard stock solutions of atorvastatin calcium hydrate, pravastatin sodium, rosuvastatin calcium, simvastatin, and ezetimibe were prepared in methanol (ca. 100 μg/mL). The solutions for the validation were prepared from stock solutions. The calibration curves were constructed by triplicate analysis of the five concentrations of the samples, whose ranges were one fourth or one eighth to twice the expected drug concentrations in the diluted samples for the HPLC analyses. The expected drug concentrations for atorvastatin, pravastatin, rosuvastatin, simvastatin, and ezetimibe were 2.0, 2.0, 1.0, 2.0, and 4.0 μg/mL, respectively. Quality control samples were prepared at three concentrations (middle, high, low) within the ranges of the calibration curves. The three concentrations were half of the expected drug concentrations, the three-fold concentrations, and 40 to 70% less concentrations. The precision and repeatability of the HPLC methods were determined by intra- and inter-day variations through repeated analyses (N = 5) of quality control samples on three different days. Various diluted samples were analyzed five times, and the limit of detection (LOD) and limit of quantitation (LOQ) were determined based on the concentrations with signal to noise ratios of three and 10, respectively.

Structural elucidation of degradation products

The degradation product from the SS was extracted with chloroform from the aqueous supernatants of the co-suspensions of SS and MG for structural analysis using nuclear magnetic resonance (NMR) spectroscopy. ¹H and ¹H-decoupled ¹³C NMR spectra and two-dimensional NMR spectra were measured with an AVANCE 500 NMR spectrometer (11.75 T, Bruker Japan, Kanagawa, Japan) using 5-mm sample tubes at 300 K. The aqueous supernatants of the co-suspensions of SS and MG were also injected into the HPLC apparatus in order to collect the fractions containing the degradation product. The HPLC conditions were the same as described above, except the mobile phase was a mixture of 0.1% trifluoroacetic acid and acetonitrile (15/85). The fractions were directly analyzed with a micro TOF-Q spectrometer (Bruker Daltonics, Kanagawa, Japan) in order to obtain the high-resolution electrospray ionization mass spectrum (ESI–MS).

The aqueous methanolic solution, which was obtained from the co-suspensions of EZ and MG, was injected into the HPLC apparatus to isolate the degradation product from EZ. The HPLC conditions were the same as described above, except the mobile phase was a mixture of 0.1% trifluoroacetic acid and acetonitrile (40/60). The NMR spectra and ESI–MS of the isolated degradation product were measured using the above spectrometers.

Statistical analysis

All data are expressed as the mean ± standard deviation of at least three experiments. All statistical analyses were performed using one-way ANOVA followed by Tukey’s test to determine significance. Statistical significance was set at p < 0.05.

Results

Preparation of suspensions

The CLDs examined in this study are shown in Fig. 1. These drugs have partial structures such as ester bonds, which are labile to the alkaline conditions, or coordination groups such as carboxy groups. The SSM was applied to their tablets, which were disintegrated with or without MG, and their suspensions successfully prepared. The pH values of the aqueous supernatants of the suspensions without MG were near neutral except for AS, whereas those of the co-suspensions with MG were fairly basic (Table 2).

Fig. 1

Chemical structures of CLDs examined. symbols: A atorvastatin; B pravastatin; C rosuvastatin; D simvastatin; E ezetimibe

Table 2 pH values of supernatants of the simple suspensions prepared after soak for 10 min^a

Stability studies

The simple suspensions were treated by method I and the extracts were analyzed by HPLC. For AS, PS, and RS, only the peaks of the medicinal ingredients were observed with or without MG at any wavelength, as shown in Fig. 2. For SS, one degradation peak was observed at t_R 6.4 min when co-suspended with MG, whereas the degradation peak was negligible without MG, as shown in Fig. 3A-C. The degradation product had a shorter retention time in the reversed-phase column, suggesting that simvastatin was transformed into a more hydrophilic product by hydrolysis of the ester bond or lactone moiety. For EZ, a minute degradation peak was observed at t_R 10.6 min when co-suspended with MG, whereas no degradation peak was observed without MG, as shown in Fig. 3E-G. The EZ degradation product with a longer retention time was considered to be more hydrophobic than ezetimibe. In order to eliminate the possibility of further degradation during the sample treatment by method I, the co-suspension of SS with MG or EZ with MG was analyzed after immediate neutralization by the cation exchange column (method II). Consequently, the degradation peaks were also observed to a similar degree, as shown Fig. 3D and H. In addition, these degradation peaks were also observed when the aqueous supernatants of the co-suspensions with MG were directly injected into the HPLC system, as shown in the insets of Fig. 3B and F. The absorption spectra of these degradation products were almost identical to those of the corresponding parent drugs, indicating that the chromophore structures remained unchanged.

Fig. 2

HPLC chromatograms of AS (left), PS (middle), and RS (right) suspended with or without MG under the SSM condition. The suspensions of the following formulations were prepared and treated by method I followed by HPLC analysis, immediately after disintegration for 10 min. The mobile phases and detection wavelengths; see Table 1. A AS; B AS + MG; C PS; D PS + MG; E RS; F RS + MG. symbols: 1, atorvastatin; 2, pravastatin; 3, rosuvastatin

Fig. 3

HPLC chromatograms of SS (left) and EZ (right) suspended with or without MG under the SSM condition. The suspensions of the following formulations were prepared and treated by method I (A, B, C, E, F, G) or method II (D, H) and the resulting solutions were analyzed by HPLC. Unless otherwise noted, the analyses were performed immediately after disintegration for 10 min. The insets in panels B and F show the chromatograms of the aqueous supernatants of the co-suspensions with MG. The mobile phases and detection wavelengths; see Table 1. A SS; B and D SS + MG; C SS + MG (5 h later); E EZ; F and H EZ + MG; G EZ + MG (5 h later). symbols: 1, simvastatin; 2, ezetimibe; *, degradation product

The amounts of drugs in the extracts were measured based on the HPLC peak areas, and the percentages of the measured values to the labeled amounts of the tablets were calculated (Table 3). In the suspensions without MG, the percentages of the individual drugs were 98 to 105% after 10 min, and no significant decreases of the percentages were observed after 5 h. In the co-suspensions with MG, there were no significant decreases in the percentages of atorvastatin and pravastatin compared with the case without MG, while a considerable decrease of the percentage was observed for rosuvastatin after 5 h.

Table 3 Quantitative determination of the medicinal ingredients in the simple suspensions^a

Structural elucidation of degradation products

The chloroform extract from the co-suspension of SS with MG was confirmed to contain one major HPLC peak due to the degradation product. The ¹H NMR spectra of simvastatin and the extract are shown in Fig. 4. In Fig. 4B, major signals due to the degradation product were observed with minor ones due to the unchanged simvastatin. The signal (H5), due to the lactone moiety of simvastatin, greatly shifted in the degradation product. The accurate mass spectrum of the degradation product showed an ion peak [M + Na]⁺ at m/z 459.2737, which showed that the molecular formula of the degradation product was C₂₅H₄₀O₆. The molecular weight, 436, was 18 (water molecules) greater than that of simvastatin (C₂₅H₃₈O₅). These data indicate the hydrolysis of the lactone moiety of simvastatin, resulting in the formation of simvastatin acid (Fig. 5) as previously reported in the literature [25]. Almost all the major ¹H signals shown in Fig. 4B were assigned to the individual protons of simvastatin acid. Thus, the SS degradation product was identified as simvastatin acid, which was also supported by the ¹³C NMR and two-dimensional NMR spectroscopic data (Supplemental Figs. 1, 2, 3 and 4).

Fig. 4

¹H NMR spectra of simvastatin (A) and the chloroform extract from the co-suspension of SS and MG (B). The spectra were measured in CDCl₃. The chloroform extract was found to mainly contain the degradation product, simvastatin acid, with a small amount of the unchanged simvastatin (with asterisk). The spectrum (B) was considered to also contain some signals due to the additives of formulations. The numeric characters on the spectra (A) and (B) show assignments of the signals to the positions of the chemical structures shown in Fig. 1 (simvastatin) and Fig. 5 (simvastatin acid), respectively

Fig. 5

Chemical structures of the degradation products. I: simvastatin acid, II: (2R,3R,6S)-N,6-bis(4-fluorophenyl)-2-(4-hydroxyphenyl)-3,4,5,6-tetrahydro-2H-pyran-3-carboxamide

When the ¹H NMR spectrum (Supplemental Fig. 5) of the degradation product from EZ was compared with that of ezetimibe, the ¹H signals due to the 5- and 6-positions of ezetimibe (see Fig. 1) were found to be greatly changed in terms of the chemical shifts and signal shapes. The accurate mass spectrum of the degradation product showed an ion peak at m/z 432.1402 [M + Na]⁺, indicating that the molecular formula was C₂₄H₂₁F₂NO₃. These data showed that the product was the structural isomer of ezetimibe, a pyran compound (Fig. 5, compound II) as previously reported in the literature [27]. The ¹H NMR spectrum of the product was substantially identical to the spectrum data of the pyran compound in the literature. The structure was supported by the ¹³C and two-dimensional NMR spectroscopic data (Supplemental Figs. 6, 7, 8 and 9).

Validation of the HPLC method

The HPLC methods were validated for the individual drugs in terms of linearity, precision, LOD, and LOQ as shown in Supplemental Tables 1 and 2. The calibration curves exhibited good linearity in the tested concentrations ranges, with the coefficients of determination (r²) being greater than 0.999. The intra- and inter-day relative standard deviations were less than 1.9% and 2.4%, respectively. The LOD and LOQ were in the range of 16 ng/mL to 65 ng/mL and 47 ng/mL to 262 ng/mL, respectively. All of the validation parameters were within the acceptance ranges.

Discussion

There have been many reports on the chemical stability under aqueous alkaline conditions for the CLDs examined in this study (Table 4). In most cases, relatively strong conditions were applied, and the drugs were more or less decomposed. Thus, the chemical stability for the individual drugs needed to be investigated under the condition of SSM with MG. In clinical settings, the prepared simple suspensions are occasionally left for long time before administration due to business reasons [31]. Thus, in this study, simple suspensions after being soaked for 5 h as well as 10 min were prepared and analyzed, assuming that the maximum time to be left was 5 h to be on the safe side.

Table 4 Degradation studies of cholesterol-lowering drugs under alkaline conditions

From the experimental results, it was shown that atorvastatin and pravastatin were chemically stable under the SSM condition with or without MG. In contrast, the recovery of rosuvastatin from the co-suspensions with MG considerably decreased after 5 h. The decrease seems to be due to adsorption of rosuvastatin to the large amounts of MgO and/or formation of undissolved complex with Mg²⁺ [14]. These putative interactions of rosuvastatin may, at least temporarily, remain unchanged in the stomach when the co-suspension is ingested after meals, because the acidity of the stomach contents considerably decreases (pH ~ 5) due to meals [32], and MgO neutralize the gastric acid. It was reported that the blood concentration of rosuvastatin was markedly decreased when the drug was ingested at the same time with magnesium hydroxide [33]. Phenomena similar to the above ones can be presumed to occur in the stomach, leading to the delayed and reduced absorption of rosuvastatin into the gastrointestinal tract. As shown in Table 3, the percentages of simvastatin to the labeled amount were significantly decreased by the co-suspension with MG regardless of the immersion time and neutralization. Although the percentages of ezetimibe were also decreased by the co-suspension with MG, the decrease was not necessarily statistically significant because of the lower degradation.

When the co-suspensions of SS and MG were initially treated by method I, methanol was used as the washing solvent. The HPLC chromatograms at that time showed two degradation peaks, where one was identified as simvastatin acid with a shorter retention time and the other was observed with a longer retention time (Supplemental Fig. 10). The latter peak was estimated to be due to a simvastatin acid methyl ester, as a result of the NMR spectroscopic and mass spectrometric (accurate mass) analyses of the HPLC fraction corresponding to the peak (data are not shown). Yang et al. [25] have also reported that simvastatin acid methyl ester was formed from simvastatin acid in an alkaline aqueous methanolic solution. These findings indicated that the degradation due to the SSM itself could be overestimated by the treatment with methanol. Thus, the SS suspensions were treated with acetonitrile, by which simvastatin acid was observed as the sole degradation product in the co-suspensions with MG.

In general, statins are known to show a pH-dependent interconversion between the active hydroxy acid form and the lactone one [24, 25, 34, 35]. It has been reported that statins exist as their mixture under strong acidic conditions, whereas under alkaline conditions they exclusively exist as hydroxy acid forms. Simvastatin is used as a lactone form, i.e., a prodrug, and it exerts pharmacological effects after being hydrolyzed to the active hydroxy acid form in the body. This study revealed that simvastatin is partially transformed into a less lipophilic hydroxy acid, simvastatin acid, under the condition of the SSM with MG. Such a transformation before ingestion possibly decreases the efficacy of simvastatin since simvastatin acid is poorly absorbed into the gastrointestinal tract when compared with simvastatin.

It has also been reported that ezetimibe undergoes the cleavage of the β-lactam and the ring transformation into the pyran compound under alkaline conditions [28]. The reactions are dependent on the pH of the solution, where the pyran compound is mainly formed at pH 7 to 12.5. Although the decomposition of ezetimibe into the pyran compound has been demonstrated to occur under the strong conditions in the literature, the same decomposition was found to slightly occur also under the condition of the SSM with MG.

The temperature (55°C) of the water used for the SSM is a rough indication. The temperature was selected because it was maintained at about 37°C after having been left for 10 min at room temperature [8]. In general, the temperature of the water used is not measured in clinical settings, and thus the co-suspensions are possibly prepared at higher temperatures. The degradation can be accelerated in such cases. The degradation observed in this study seems to also occur in the co-suspension with other alkaline drugs such as magnesium hydroxide [14].

This study dealt with the specific generic drugs of CLDs, and it is of interest whether similar results are obtained for other formulations including the original and generic drugs. Although various acidic and basic additives are used, the amounts are considered smaller than the co-suspended MgO (330 mg) in most cases, because relatively small tablets (less than 200 mg) are generally used except some orally disintegrating tablets. Thus, the alkalinity and putative interactions are considered to be largely dependent on MgO. It can therefore be presumed that comparable results with the present ones will also be obtained using the amount of MgO for most of other formulations.

Conclusions

In this paper, the chemical stability of typical CLDs in a co-suspension with a specific amount of MgO was investigated under an SSM condition. We found that atorvastatin and pravastatin can be subjected to the co-suspension without significant chemical changes. However, it is preferable that the co-suspension of rosuvastatin is rapidly prepared and used after having been disintegrated for 10 min. For simvastatin and ezetimibe, the co-suspension with MgO should be avoided because of their partial degradation. Although this study was performed for specific generic drugs, the data should be considered informative for the co-suspensions of other formulations with MgO.