Apparent Solubility and Dissolution Profile at Non-Sink Conditions as Quality Improvement Tools

Stefania Petralito1,*, Iacopo Zanardi2,*, Adriana Memoli1, M. Cristina Annesini3, Vincenzo Millucci4 and Valter Travagli2 1Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza Universita di Roma, 2Dipartimento Farmaco Chimico Tecnologico, Universita degli Studi di Siena, 3Dipartimento di Ingegneria Chimica Materiali Ambiente, Sapienza Universita di Roma, 4Dipartimento di Fisica, Universita degli Studi di Siena, Italy


Introduction
The excipients used during manufacturing as well as the quality of the pharmaceutical product development and preparation are of great importance to dosage form performance. A continuous know-how improvement of both formulation and production process parameters with respect to drug release profiles is a basic aspect of the quality framework for pharmaceutical products. Drug release/dissolution studies from solid dosage forms can be considered among as the most investigated topics in pharmaceutical research (De Castro et al., 2006;Macheras & Iliadis, 2006;Siepmann & Siepmann, 2008). Such a background becomes of paramount relevance in the case of insoluble or poorly soluble drugs, where dissolution represents the most critical factor affecting the rate of systemic absorption, especially in the presence of polymorphism (Snider et al., 2004). Moreover, apart from representing an important element in development and quality control in drug research, dissolution test is proposed to be a surrogate for drug bioavailability evaluation. In fact, in vivo-in vitro relationship represents a useful tool to answer the question about the interchangeability of generic and branded products by revealing differences in dissolution kinetics (Dressmann & Reppas, 2010;Hlinak et al., 2006). In order to increase predictability of these results, several attempts to make in vitro test conditions closer to the physiological ones have been made, for example by adjusting pH or by adding surfactants. However, the so-called "sink conditions" (based on bulk drug solubility i.e. in a system where the solute is present for more than 15% of its maximum solubility have been studied), obtained by using a high concentration of a surfactant in the dissolution medium, may not be a proper approach in developing a bio-relevant dissolution method for a poorly water-soluble drugs (Sirisuth et al. 2002;Tang et al., 2001;Jamzad & Fassihi, 2006). "Non-sink conditions" represented a very discriminating dissolution conditions, acting as a sort of magnifier lens for an in-depth evaluation of the dissolution phenomenology, and dissolution tests under non-sink conditions can be a predictive tool during formulation development as well as for batch-to-batch quality control (Siewert et al., 2003).

Dissolution testing in pharmacopeia
The methods of in vitro dissolution testing can be traced to two general categories: "stirrer beaker method" and "flow through procedure".
From a regulatory standpoint, the legally-binding documents to carry out the dissolution tests are reported in the 7 th edition of European Pharmacopoeia (EP), the 34 United States Pharmacopoeia (USP), and the 15 th edition of Japanese Pharmacopoeia (JP). The World Health Organization (WHO) provides in the 4 th edition of International Pharmacopoeia (IntPh) a more global coverage of issues and strives towards harmonisation among world pharmacopoeia guidance and source material.
The various texts are comparable to the general notions, but differ in the apparatuses described: EP with "Dissolution test for solid dosage forms" (Council of Europe, 2011) and USP with "Dissolution" (United States Convention, 2011a) show four apparatuses: 1 (for Basket method); 2 (for paddle method), 3 (Reciprocating cylinder) and 4 (Flow-through cell), the latter being lacking in the JP "Dissolution test" (Society of Japanese Pharmacopoeia, 2007) monograph. On the other hand, in the IntPh "Dissolution test for solid oral dosage forms" section (World Health Organization, 2011), only the first two devices are indicated. In Table 1, the relation between the dimensions for the first three devices is shown: all the measures are equivalent, and any differences can be noticed only by IntPh, mainly in terms of the significant digits.
Even the choice of the dissolution medium is almost completely overlapped between EP and USP, with a variety of buffers at various pH (e.g. phosphate, acetate, TRIS). Besides, the JP refers to the monographs of specific formulations, but ranging over various possibilities, from pure water to the various buffers. On the other hand, the IntPh indicates eight different points for different pHs of the dissolution media, including simulated gastric fluid (SGF, pH 1.2) and simulated intestinal fluid (SIF); worthy of note is the pH differences of the latter: 7.5 for IntPh vs. 6.8 for both EP and USP.
Moreover, in the latest edition of the EP, as well as in the USP section "The dissolution procedure: development and validation" (United States Convention, 2011b), a chapter entitled "Recommendations on methods for dosage forms testing" (Council of Europe, 2010a) is given, suggesting the use of sink-condition. Sink conditions normally occur in a volume of dissolution medium that is at least 5-10 times the saturation volume, usually by adding surfactants. However, such an approach may be inappropriate in developing a biorelevant dissolution method for a poorly water-soluble drug (Sirisuth et al. 2002;Tang et al., 2001;Jamzad & Fassihi, 2006
Scheme 1. Chemical structure of NIM Its approved indications are the treatment of acute pain, the symptomatic treatment of osteoarthritis and primary dysmenorrhoea in adolescents and adults above 12 years old.
NIM was discovered in 1971 in the U.S. by George G.I. Moore at Riker Laboratories (later acquired by 3M Co.), but in 1980 NIM was licensed by Helsinn Healthcare SA (Switzerland) who proceeded to invest in extensive investigations on the drug (Rainsford, 2006). It was launched in Italy for the first time as Aulin ® in 1985 (Consalvo et al., 2010) and is currently available in more than 50 countries worldwide, among others France, Portugal, Greece, Switzerland, Belgium, Mexico and Brazil. NIM has never been filed for Food and Drug Administration (FDA) evaluation in the United States, where it is not marketed (Traversa et al., 2003) After the expiry of patent protection, a number of other companies have started production and marketing of NIM products.
Controversy regarding NIM toxicity persists due to the fact that clinical series reports and epidemiological trials continue to involve NIM in severe liver damage during the postmarketing studies (Bessone, 2010 Ireland's RTÉ aired an investigative programme highlighting the deadly side effects of NIM and how it has been linked to over 300 cases of liver disease throughout Europe. On September 21, 2007 the EMA released a press release on their review on the liver-related safety of NIM. The EMA has concluded that the benefits of these medicines outweigh their risks, but that there is a need to limit the duration of use to ensure that the risk of patients developing liver problems is kept to a minimum. Therefore the EMA has limited the use of systemic formulations (tablets, solutions, suppositories) of NIM-containing medicinal products to 15 days because of reports of severe hepatic adverse reactions (EMA, 2007;Li et al., 2009).
Singapore Health Science Authority (HSA) suspended NIM containing drugs in June 2007 (Singapore News, 2007;HSA, 2007). Several reports have been made of adverse drug reactions in India (Khan & Rahman, 2004a, 2004b. On Feb 12, 2011, Express India reported that the Union Ministry of Health and Family Welfare had finally decided to ban the pediatric use of the analgesic, NIM suspension. From 2011 onwards, it has been totally banned in India. NIM chemico-physical properties could be summarised as: i) weak acid properties (pKa reported ranging between 5.9 and 6.56 (Singh et al., 1999;Singh et al., 2001;Dellis et al., 2007); ii) values of octanol-water partition coefficient (log P) of 2.38 (Singh et al., 2001); iii) practically insoluble in water (10 g/mL) (Piel et al. 1997); iv) according to the Biopharmaceutical Classification System, BCS, (FDA, 2000), NIM can be classified as a class II drug (low solubility and high permeability), therefore, the drug dissolution may be a ratelimiting step in the drug adsorption process.
Previous studies were carried out for both in vitro (Butler et al., 2000;Rădulescu et al., 2010) and in vivo comparisons among NIM-containing tablets (Hutt et al. 2001;Ilic et al. 2009). However, in the former only a small number of commercial preparations were investigated under sink-condition by means of abnormal surfactant concentration (Butler et al. 2000), and the release rate seems to be critically influenced not by pH value or the concentration of endogenous surfactant, but by the combination of the two characteristics of the in vitro dissolution media (Rădulescu et al., 2010), while in the latter no in vitro and in vivo correlation (IVIVC) was investigated. Moreover, since due to different crystallization processes, crystallographic modification has been recently reported (Kapoor et al. 1998;Di Martino et al. 2007;Moneghini et al. 2007), even though only a single crystal structure has been identified (Dupont et al., 1995). Thus, information on the influence of the different nature and/or amount of excipients as well as of the adopted technological parameters on the in vitro drug release characteristics are reputed of interest.

Dosage form selection
Ten multisource IR NIM tablet formulations (RF for reference formulation, MSF for the multiple-source product formulation, and BF1-BF8 for non-branded bioequivalent formulations) were obtained from the Italian market. They all nominally contain 100 mg of the active ingredient, but greatly differ with respect to the excipient composition. Table 2 summarizes the qualitative excipient composition of the various NIM tablets. www.intechopen.com

Tablet appearance
Each tablet was visually examined for shape and any evidence of physical differences such as weight, thickness and dimension was recorded.

Calibration curve
Calibration curve for NIM reference standard (RS) was obtained by measuring the UV absorption (Perkin Elmer L25 spectrophotometer, max : 392.6 nm) in dissolution medium (Simulated Intestinal Fluid, SIF, pH 6.8) prepared according to EP (Council of Europe, 2010a) except for the absence of pancreatin, in conformity with the aim of this study. This pH value was selected because of the NIM negligible dissolution in acidic conditions. Due to the low aqueous NIM solubility, NIM stock solution was prepared by accurate weight of the substance and subsequent dissolution in 5 mL of ethanol, submitting to ultrasound in a sonicator bath for five minutes, and then diluted to a final volume of 100 mL with SIF, corresponding to 50 g / m L o f N I M . C a l i b r a t i o n s a m p l e s w e r e p r e p a r e d f r o m t h r e e separately weighed stock solutions to obtain sample solutions containing scalar concentrations of NIM. Samples were stored at +4 °C until analysis. The linearity of the calibration curves was confirmed over the range 1-20 g/mL.

Solubility studies
Apparent solubility (S app ) referring to the dynamic solubility (Mosharraf & Nystrom, 2003) of both NIM RS and NIM tablets were determined by stirring an excess amount of the samples in 250 mL of SIF, on a multistirrer thermostatted at 37 ± 0.5 °C for a suitable time in order to achieve equilibrium (max 72 hours). Twenty tablets of the same commercial product were weighed and powdered (particle size ≤ 150 µm, by sieving). An amount corresponding to onefourth of a tablet (equivalent to 25 mg of NIM) was weighed and suspended in 250 mL dissolution medium. In such a way the ratio among active agent, excipients and volume agrees with dissolution studies conditions (see below). The samples were filtered with a 0.45 m nylon membrane filter (Whatman, Maidstone, UK) and the absorbance of the filtrate was measured by UV. Temperature (37 ± 0.5 °C) was carefully maintained constant during all the operations and the amount of drug dissolved was calculated using the calibration curve (see above). All solubility determinations were performed in triplicate.

Dissolution studies
For tablet dissolution tests, apparatus I (rotating basket method) (Council of Europe, 2011) was used employing 1000 mL of SIF at a temperature of 37 ± 0.5 °C and a rotational speed of 100 rpm. Sample solution (5 mL) was withdrawn at appropriated time intervals (5, 10, 15, 30, 45 and 60 min) and the drawn volume was replaced with the same amount of blank dissolution medium from a separate vessel, also held at a temperature of 37 ± 0.5 °C. The samples were filtered with a 0.45 m nylon membrane filter (Whatman, Maidstone, UK) and the absorbance of the filtrate was measured by UV. The amount of NIM was calculated through the calibration curve. All the dissolution tests were conducted on twelve tablets for each formulation.

Mathematical dissolution models
The data obtained from dissolution studies were analyzed using various mathematical models (Table 3), as reported in DDSolver. It is a specialized, freely available software program developed by Zhang et al. with the main objective to provide a tool for facilitating the parameter calculations in dissolution data analysis using nonlinear optimization modeldependent approaches (Zhang et al., 2010). In the present chapter, the selection of the models for fitting dissolution data has been based on their theoretical applicability.

Dissolution model Equation a
First-order b,c  a In all models. F is the concentration (µg/mL) of the drug release in time t. b k 1 is the first-order release constant. c Fmax is the maximum fraction of the drug released at infinite time d k H is the Higuchi release constant e k KP is the release constant incorporating structural and geometric characteristics of the drug-dosage form; n is the diffusional exponent indicating the drug-release mechanism f k HC is the release constant in Hixson-Crowell model g k HB is the combined constant in Hopfenberg model, k HB =k 0 /(C0×a0), where k 0 is the erosion rate constant, C0 is the initial concentration of drug in the matrix, and a0 is the initial radius for a sphere or cylinder or the half thickness for a slab; n is 1, 2, and 3 for a slab, cylinder, and sphere, respectively h k BL is the combined constant in Baker-Lonsdale model, k BL =[3×D×Cs/(r0 2 ×C0)], where D is the diffusion coefficient, Cs is the saturation solubility, r0 is the initial radius for a sphere or cylinder or the halfthickness for a slab, and C0 is the initial drug loading in the matrix i k MB , n, and k are empirical parameters in Makoid-Banakar model (k MB , n, k>0) l k 1 is the constant related to the Fickian kinetics; k 2 is the constant related to Case-II relaxation kinetics; m is the diffusional exponent for a device of any geometric shape which inhibits controlled release m k 1 is the constant denoting the relative contribution of t 0.5 -dependent drug diffusion to drug release; k 2 is the constant denoting the relative contribution of t-dependent polymer relaxation to drug release n k 1 is the constant in Quadratic model denoting the relative contribution of t 2 -dependent drug release; k 2 www.intechopen.com is the constant in Quadratic model denoting the relative contribution of t-dependent drug release o is the scale parameter which defines the time scale of the process; is the shape parameter which characterizes the curve as either exponential ( =1; case 1), sigmoid, S-shaped, with upward curvature followed by a turning point ( >1; case 2), or parabolic, with a higher initial slope and after that consistent with the exponential ( <1; case 3) p Ti is the location parameter which represents the lag time before the onset of the dissolution or release process and in most cases will be near zero q is the scale factor in Logistic 1 and 2 models; is the shape factor in Logistic 1 and 2 models r k is the shape factor in Logistic 3 model; is the time at which F = Fmax/2 s is the scale factor in Gompertz 1 and 2 models; is the shape factor in Gompertz 1 and 2 models t k is the shape factor in Gompertz 3 model; is the time at which F = Fmax/exp(1)≈0.368×Fmax u is the scale factor in Gompertz 4 model; k is the shape factor in Gompertz 4 model v Ф is the standard normal distribution; is the scale factor in Probit model; is the shape factor in Probit model

Statistical analysis
Results were expressed as the mean ± SD of at least six independent measurements. ANOVA one-way performing the Bonferroni post-test (Instat software, version 3.0 GraphPAD Software Inc., San Diego, CA) were used for the statistical analysis of the results. Significance was defined as a p value less than 0.05 (* p < 0.05; ** p < 0.01; *** p < 0.001).

Results and discussion
Tablet weight and dimensions data obtained for the formulations studied are reported in Table 4.  In detail, all the tablets have the same shape and each tablet has a weight of about 400 mg, except BF6 and BF7 that are heavier reaching a weight of about 700 mg. As far as the visual inspection is concerned, BF1 and BF8 show a fracture line along the tablets diameter as well as a different appearance along thickness of BF7 tablets was observed, suggesting lamination due to compression steps (Carstensen et al., 1985).

PRODUCT
From dissolution data analysis it is possible to note that, apart from BF8 representing the lower amount released (6.0 mg, corresponding to the 6% after 1 h), the % of release for the formulations was between 16.1% and 23.0% (Figure 1). Fig. 1. Dissolution profiles of the various tablets at pH 6.8 (CV%<5).
Such differences, that our adopted experimental conditions in the absence of sink-conditions were able to blow up, were attributed to formulation differences and/or manufacturing procedures. For this reason, to better explain the dissolution profile, saturation concentrations obtained from solubility studies in the presence of the various auxiliary substances were also adopted. They are shown in Figure 2. NIM RS has the lowest S app (19.47 µg/mL) while for BF2 an increment of S app value near 30% (24.97 µg/mL) was obtained. With respect to NIM value, statistically significant increase in Sapp was obtained in all cases (p < 0.001, except than NIM vs. BF7 and NIM vs. BF4, for which p < 0.01 and p < 0.05 were observed, respectively).
These data suggested us to use S app value of each formulation as normalization factor for each dissolution evaluation, instead of that of NIM solubility. Results are depicted in Figure  3, from which it appears that also in this case BF8 is not able to reach the 30% of dissolution. Fig. 3. Dissolution profiles of the various tablets at pH 6.8 as obtained with respect to apparent solubility (CV%<5).
As concerns BF4 its profile remains below all other curves. On the contrary, for BF1, after 1h is appreciable the achievement of 100%; such a result were not obtained by the other formulations, except than for RF.
The simplest model to represent a dissolution of a BCS II drug is a first-order model, with a dissolution rate proportional to the difference between the apparent solubility of drug and the drug concentration in the liquid phase. Such a model, along with others, has been applied with DDSolver Software and the obtained results are shown in Table 5.
The corresponding correlation coefficients, R 2 , in most of the adopted equations gave results higher than 0.97, and in all these cases the worst result is obtained in the description of the dissolution of the BF8 formulation.
Further investigation may be done by going to compare the S app experimental data with those determined using some equations, which allows to calculate a value of F max ( Figure 4).
As it is possible to observe, no model is able to estimate properly the experimental S app parameter of the various preparations, except than the Weibull functions. In detail, either under-or overestimations have been observed, with the model Gompertz 2 that provides a www.intechopen.com value always greater than the others. As for BF8, the non-sink conditions allow to show a significant extension of time to reach the maximum value of the dynamic solubility. Fig. 4. Comparison of experimental S app and calculated F max derived from mathematical fitting.

Conclusion
A continuous know-how improvement of both formulation and manufacturing process parameters with respect to drug release modalities is a basic aspect of the quality framework for pharmaceutical products. Such features become of paramount relevance for generic manufacturers in the case of formulation of insoluble or poorly soluble drugs, where dissolution represents the most critical factor. In fact, both the excipients used as well as the manufacture parameters are of great importance to solid dosage form performance. The in vivo solubility behaviour is dependent on many factors and it cannot be fully obtained in vitro. Regardless the worth of "sink conditions" in the achievement of bio-relevant dissolution methods for poorly water-soluble drugs, an approach more suitable as developmental tool as well as for batch quality control in both pre-formulation and formulation stages is of great interest.
For such purposes, various commercial Immediate Release tablets containing a drug, namely NIM, differently banned, used and prescribed in the various European Countries have been chosen. NIM apparent solubilities and dissolution patterns as obtained in "nonsink conditions", i.e. in a system where the solute is present for more than 15% of its maximum solubility, have been studied. "Non-sink conditions" represented a very discriminating dissolution conditions, acting as a sort of magnifier lens for an in-depth evaluation of the dissolution phenomenology, useful since the preformulation stages.
Eventually, in such a situation of drug saturation during tablet dissolution the mathematical approaches far developed usually are not capable of describing the overall profile peculiarity.
For this reason, a more complex dissolution scheme based on the ash layer diffusion control by shrinking core model is under study.