Investigation of biopharmaceutical and physicochemical drug properties suitable for orally disintegrating tablets

The purpose of this study was to evaluate the biopharmaceutical and physicochemical drug properties suitable for orally disintegrating tablets (ODTs). The molecular weight (MW), polar surface area (PSA), hydrogen bond donor (HBD) and acceptor (HBA) numbers, net charge at pH 7.4, log D6.5, the highest dose strength, solubility in water, dose number, and elimination t1/2 of 57 ODT drugs and 113 drugs of immediate-release (IR) formulations were compared. These drugs were classified according to the Biopharmaceutical Classification System (BCS). A lower dose strength and a longer elimination t1/2 have been observed as characteristic properties of ODTs. The proportion of basic drugs was higher in the ODTs than in the IR formulations. A significant difference was not observed between the ODT and the IR formulation for MW, PSA, HBD, HBA, log D6.5, solubility in water, and dose number. The distributions of the ODTs and IR formulations among each BCS class were similar, suggesting that an ODT can be developed regardless of the BCS class of a drug.


Introduction
An orally disintegrating tablet (ODT) improves patient compliance because it can be taken without water, does not cause dysphagia, and can prevent patients from spitting out their medication [1,2].Many pharmaceutical companies are working on the development of ODTs.An ODT should be bioequivalent to a corresponding standard formulation, e.g. an immediate-release (IR) formulation.An ODT disintegrates and dissolves rapidly in the oral cavity within 30 seconds.Therefore, the dissolution rate may differ significantly between ODTs and IR formulations.When the dissolution rates are different between the ODT and the IR formulation, the risk to fail in a bioequivalence (BE) study is high.Therefore, during the development of an ODT, it is important to assess the risk of failing in a clinical BE study.It is preferable to reduce the risk of failing a clinical BE study because such studies are expensive, time consuming, and a burden to healthy volunteers.
However, there has been no research investigating the biopharmaceutical properties of drugs suitable to be an ODT.In the present study, several physicochemical and biopharmaceutical properties were selected and a survey was performed to compare the properties between the two types of oral dose formulations.

Drug list
The physicochemical and biopharmaceutical properties of drugs marketed as ODTs and IR formulations were compared in this study.Currently, the number of drugs developed as ODT is largest in Japan.In addition, all the ODT formulations approved in Japan were proved to be bioequivalent to the IR formulation with and without water intake.Therefore, the ODT formulations in Japanese market were selected in this study.The list of the ODT drugs was obtained from the Pharmaceuticals and Medicinal Devices Agency website (http://www.info.pmda.go.jp/psearch/html/menu_tenpu_base.html)(Supplement Table 1).The list of the IR formulations were selected from the top 200 pharmaceutical products in Japan (Supplement Table 2) [3].Finally, 57 compounds for the ODT and 113 compounds for the IR formulation were selected and analysed.The 25 compounds were overlapping between the lists of the ODTs and IR formulations.

Biopharmaceutical drug properties
The physicochemical and biopharmaceutical properties of drugs related to the formulation design and oral bioavailability were selected [4,5].The molecular weight (MW), polar surface area (PSA), hydrogen bond donor (HBD) and acceptor (HBA) numbers, dissociation constants (pK a ), and n-octanol/water distribution coefficients at pH6.5 (log D 6.5 ) were calculated using ACD Percepta (ACD/Labs Software V 14.0.0 (http://www.acdlabs.com/products/percepta/predictors.php)).The calculated pK a and log D were used for all drugs, as the experimental values were not available for some drugs.
Net charge (NC) at pH 7.4 was represented as the weighted sum of the charge of each species.

NC = (±0
where f 0 is the fraction of the undissociated species, f + is that of +1 charged species, etc.Each fraction was calculated by the pK a and the pH (set to be 7.4 in this study) using the Henderson-Hasselbalch equation [6].The NC equations for acids and bases containing up to three ionization centers are summarized in Supplement Table 3.When a drug was more than 50 % dissociated at pH 7.4, it was classified as an acid (NC < -0.5) or as a base (NC > 0.5).Log D was calculated at pH 6.5 to estimate the permeability, whereas NC was calculated at pH 7.4 to discuss the pharmacokinetics after the absorption.The solubility in water, the highest dose strength, and the elimination t 1/2 were obtained from the prescription information.When a reliable solubility figure was unavailable, a solubility value was assigned based on the solubility category defined by the Japanese pharmacopeia (Supplement Table 4).

Provisional classification according to the Biopharmaceutical Classification System
According to the Biopharmaceutical Classification System (BCS), drugs can be categorized into the four classes, i.e. high solubility/high permeability (class I), low solubility/high permeability (class II), high solubility/low permeability (class III), and low solubility/low permeability (class IV).Moreover, BCS class II drugs can be sub-classified into acid (class IIa), base (class IIb), and undissociated drugs (class IIc) [7].
In the previous studies, solubility in water and calculated log D 6.5 were used as the surrogates of solubility and permeability data to provide provisional BCS class [3,8].The same approach was taken in this study.According to the official BCS guidance, the equilibrium solubility of a drug at the physiological gastrointestinal pH range (namely, pH 1.2 to pH 6.8 or 7.4) is required.However, the pH solubility profile data were not available for many drugs.Therefore, solubility in water reported in the prescription information was used in this study.The dose number (Do) is a dimensionless number expressed by the ratio of the dose and the maximum dissolved amount in the intestine.Do was calculated by using Eq. 2 [5]: where S is the solubility of a drug in water, V is the intestinal fluid volume (set to be 250 mL in this study), and M is the highest dose strength.Drugs were defined as highly soluble when the Do was ≤ 1.
According to the official BCS guidance, to classify the permeability category, the fraction of a dose absorbed (Fa%) in humans or Caco-2 permeability data is required.However, due to the limited availability of these data, the permeability was classified based on the calculated log D 6.5 in this study.The log D 6.5 value of metoprolol was chosen as the criteria for high permeability [Fa% in human, 95 % [9]; human effective permeability, 1.26 × 10 -4 cm/s [9]; log D 6.5 , -0.92 (ACD Percepta)].

Statistical analysis
A student's t-test was used to evaluate the significance of difference between the ODTs and IR formulations in MW, PSA, HBD, HBA, NC at pH 7.4, log D 6.5 , the highest dose strength, solubility in water, Do, and elimination t 1/2 .One-way analysis of variance was used to evaluate the significance of difference between acids, bases, and undissociated drugs in elimination t 1/2 for the ODTs and IR formulations, respectively.A minimum p value of 0.05 was used as the significance level for all tests.Microsoft Excel 2010 (Microsoft) was used for statistical analysis.

Results
MW of free form, PSA, HBD, HBA, NC at pH 7.4, log D 6.5 , the highest dose strength, solubility in water, Do, and elimination t 1/2 are shown in Table 1.Some calculated pK a and log D 6.5 might have a margin of error about 1 log unit (e.g.pK a s of cetirizine (6.7 (B); 7.7 (B) [10]), domperidone (9.0 (B); 7.1 (B) [10]), glimepiride (5.1 (A); 6.2 (A) [11]), log D 6.5 of famotidine (-2.14; -1.3 [11]), glimepiride (1.51; 3.0 [11]) (calculated values; experimental values).NC at pH 7.4 was significantly higher in the ODT than that for the IR formulation (p = 0.02) (Figure 1e).The percentages of acid, base, and undissociated drugs for the ODTs were 13, 41 and 45 %, respectively (2 % unclassifiable).The corresponding percentages in the IR formulations were 30, 30 and 38 %, respectively (2 % unclassifiable).The highest dose strength of the ODTs was significantly lower than that of the IR formulations (p = 0.01) (Figure 1g).The medians of the highest dose strength in the ODTs and IR formulations were 10 mg and 30 mg, respectively.The maximum values of the highest dose strength in the ODTs and IR formulations were 200 mg and 900 mg, respectively (Table 2).The elimination t 1/2 of the ODTs tended to be longer than that of the IR formulations (p = 0.07) (Figure 1j).The medians of t 1/2 in the ODTs and IR formulations were 6.7 h, and 3.3 h, respectively (Table 2).A significant difference was not observed between the ODTs and the IR formulations for MW, PSA, HBD, HBA, log D 6.5 , solubility in water, and Do (Figure 1a, b, c, d, f, h, i).
The distribution of the ODTs and IR formulations among each BCS class were similar (Figure 2).The percentages of BCS class I, class II, class III, and class IV for the ODT were 52, 29, 20 and 0 %, respectively (2 % unclassifiable).The corresponding percentages for the IR formulation were 48, 30, 16 and 4 %, respectively (2 % unclassifiable).The distribution of BCS II subclass was also similar (Figure 2).The percentages of BCS class IIa, class IIb, and class IIc for the ODT were 7, 12 and 9 %, respectively, whereas the corresponding percentages for the IR formulation were 10, 10 and 11 %, respectively.

Discussion
A significant difference was seen in the highest dose strength between the ODT and IR formulation.The ODT would have to have a feasible tablet size and drug loading [13,14].Therefore, a drug with a high dose strength (>200 mg) would be less suitable an ODT.
There was no significant difference in the solubility in water and Do between the ODT and IR formulation.This result suggests that many drugs can be developed as an ODT regardless of their solubility and Do (e.g., bicalutamide: S = 0.01 mg/mL, Do = 32, cilostazole: S = 0.01 mg/mL, Do = 40).Furthermore, an ODT shows BE with a corresponding IR formulation with and without water intake.Therefore, even though it is counterintuitive, water intake may have little effect on the dissolution and oral absorption of low solubility drugs.Previously, Sumesen et al. reported that the oral absorption of danazol was not significantly altered when administered together with 1000 mL of water compared to when administered with 200 mL [15].Danazol, which has poor water solubility (0.2 g/mL) and high permeability (log D 6.5 = 4.5), is a typical BCS class II drug [16].
The drugs that have been developed as ODTs tended to have a longer elimination t 1/2 .Previously, we reported that the elimination t 1/2 of drugs influence the BE of C max [17].For the drugs with high permeability and short elimination t 1/2 , BE of Cmax between two formulations with different dissolution rates would become more difficult to prove.This point has been suggested by several articles [18][19][20][21].A drug with a long elimination t 1/2 might be suitable for an ODT.
The proportion of bases was larger in the ODT than in the IR formulation.It is well known that the basic lipophilic drugs have a large distribution volume, and a long elimination t 1/2 due to wide tissue distribution [22].A significant difference was seen in elimination t 1/2 between acids, bases, and undissociated drugs in the IR formulation (p = 0.04) (Figure 3).The biowaiver schemes (BWS) has been discussed based on BCS proposed by Amidon et al. in 1995 [5].In 2000, the US Food and Drug Administration (FDA) adopted the BCS-BWS [23].The guideline allows BCS class I drugs which show rapid dissolution (>85 % dissolution in 30 min) to waive clinical BE studies.The World Health Organization (WHO) and other regulatory agencies followed the BCS-BWS [24][25][26][27].However, there are some differences among these guidelines, e.g.solubility pH range, criteria for high permeability, and definition of the dose used for the Do calculation [28].One of the most significant differences is about the biowaiver for BCS class III drugs.WHO, European Union, and Canada accept biowaiver for BCS class III drugs which show very rapid dissolution (>85 % dissolution in 15 min), while only BCS class I drugs are eligible for biowaiver in the US FDA and Korea FDA guideline.However, there are several computer simulation and experimental studies on the biowaiver for BCS class III drugs, suggesting that BCS class III drugs are suitable for biowaiver [20,21,[29][30][31][32][33].In addition, it has been pointed out that many BCS class III drugs show BE even when the dissolution profiles are different between the test and reference drugs, for example famotidine, hydrochlorothiazide, and cimetidine [33,34].Moreover, WHO adopts the possibility of biowaiver for BCS class IIa drugs with low solubility at acidic pH and high solubility at neutral pH that are absorbed completely.However, it was reported that, in the case of ibuprofen, the typically BCS class IIa drug, the BE of Cmax is more sensitive to the difference of dissolution rates [35,36].
The BCS class distribution of ODTs and IR formulations may reflect their non-BE risk due to the difference of the dissolution rates.Based on the BCS-BWS, BE is most easily established for BCS class I drugs.However, no difference in the distribution of the BCS classes, including subclasses, between the ODTs and IR formulations was observed in this study.This result may suggest that a BCS class I drug would not necessarily be suitable to show BE.Ramirez et al. reported that in the 124 clinical BE studies there is no difference in the number of subjects in the BE study and the inter-and intra-subject variability for C max or AUC between four BCS classes [37].All the BCS classes drugs have the risk of non-BE.Their results also showed that all of the bioinequivalent parameters in BCS class I drugs was C max , but not AUC.The results of the present study suggest that the elimination t 1/2 would affect the success rate of a BE study more significantly than the BCS class of a drug (e.g.ibuprofen).
In conclusion, drugs with a lower dose strength (<10 mg) and a longer elimination t 1/2 (>6.5 h) were suggested to be more suitable for an ODT.However, the distributions of the ODT and IR formulation among each BCS class were similar, suggesting that the BCS classes are irrelevant to the development risk of ODTs.

Figure 1 .
Figure 1.Box and whisker plots of biopharmaceutical properties of ODT and IR formulation drugs ; the bottom and top of the box are the first and third quartiles, the band inside the box is the median, and the ends of the whiskers are the minimum and maximum.(a) MW of free form.(b) Polar surface area.(c) Hydrogen bond donor number.(d) Hydrogen bond acceptor number.(e) Net charge at pH 7.4.(f) Log D 6.5 .(g) The highest dose strength.(h) Solubility in water.(i) Dose number.(j) Elimination t 1/2 .*p < 0.05.

Figure 3 .
Figure 3. Box and whisker plot of elimination t 1/2 of acid, base, and undissociated drugs

Table 1 .
MW of free form, PSA, HBD, HBA, NC at pH 7.4, log D 6.5 , the highest dose strength, solubility in water, Do, elimination t 1/2 , and BCS class of drugs used in this study.

Table 2 .
Minimum, maximum, and median of biopharmaceutical properties of ODT and IR formulation drugs.Provisional BCS classification of drugs in ODTs and IR formulations.