Over the last decade, there has been much discussion on favorable physicochemical properties for drug candidates.1–4 Nevertheless, formulators must still deal with many challenging compounds, because such problematic compounds, notably poorly soluble compounds, are often judged to be effective in high-throughput screening studies. Reason for the increase of poorly soluble candidates is frequently explained in terms of improvements in synthesis technology, which have enabled the design of very complicated compounds, and a change in discovery strategy from a so-called phenotypic approach to a target-based approach, which occurred around 1990.5 The phenotypic approach is the classical trial-and-error methodology in which candidate compounds are tested against cells, tissues, or whole bodies. Thus, it takes into account various physicochemical and biological factors together that may influence the efficacy of candidates, such as solubility, protein binding, or metabolism. On the other hand, in the target-based approach, candidate compounds are screened against specific targets based on hypotheses on action mechanisms. The great advantage of this approach is that it provides clear idea for molecular design. However, physicochemical and biological problems are frequently ignored in the screening stage.
Various types of the oral special dosage forms including self-emulsifying drug delivery systems, solid dispersions, nano suspensions, and colloidal formulations have been actively developed by both academic and industrial researchers. New administration routes such as pulmonary delivery are also being investigated with the aim of achieving efficient systematic delivery of poorly absorbable drugs. This article describes quick review on their developability as the formulation technology for the early clinical phases, for which the restriction in the developmental timelines is very strict,6 followed by focus on the current status of the amorphous technology, which still has relatively many issues to be proved for the general use. The technical aspect of each formulation except the amorphous formulation will not be discussed in detail in this article, because several excellent reviews7, 8 have already been available on this topic.
TIMEFRAMES FOR FORMULATION DEVELOPMENTAL STUDIES
In major pharmaceutical companies, the typical period from candidate selection to first-in-human is around or less than 1 year. Because the latter half of this 1 year is spent on safety studies and document preparation for IND submission, the physicochemical characterization of the active pharmaceutical ingredient (API) and the most part of the formulation developmental study must be completed in about 6 months. Once the compound for the development is identified, high-throughput polymorph screening is carried out with the aim of preventing unexpected events in the developmental study, such as poor reproducibility of the crystal form or low physical stability of the ongoing form, unless the crystal form has been decided yet. Such trouble may happen even after the launch of the product as in the case of ritonavir. Therefore, an attempt is usually made to find the most stable form for further development. High-throughput salt screening may be conducted as well to improve the physicochemical characteristics of the API including solubility, stability, and crystallinity. Although the co-crystal is also a hopeful option to alter the API characteristics, its screening procedure is still under development. The screening process to determine the crystal form and salt type usually takes about 2 months (1 month for screening experiments plus 1 month for data analysis and scale-up studies), which means that only 4 months can be spent on the formulation developmental study, although minor modifications may be accepted after safety studies have been initiated. It should be noted that extemporaneous preparations are sometimes employed in recent developmental strategies that can significantly shorten the developmental period. This commentary describes the cases where such simple formulations are not acceptable due to problems of API characteristics.
SPECIAL DOSAGE FORMS
No clear definitions of the term “special dosage forms” have been available. In this article, all nonroutine formulations are of interest. Therefore, the only exceptions are immediate-release tablets, capsules, and extemporaneous preparations including powder-in bottles.
Some special dosage forms are easily developed without altering the development timeframes, while extra time becomes required for other formulations. Table 1 classifies special dosage forms based on their applicability within typical timeframes. Class I involves technologies which can easily be integrated into the normal developmental strategies, such as self-emulsifying formulations. Class II technologies, which include amorphous formulations, can also fit into the timeframes. However, revision of timelines may become necessary. For technologies in class III, the formulation strategy significantly affects the timelines, and in some cases, even the molecular design. Therefore, class II and III technologies are basically suitable for candidates, for which proof of concept (POC) has already been established, or for life cycle management (LCM) purposes.
Table 1. Classification of Special Dosage Forms Based on Their Applicability in the Normal Preclinical Timeframes and Major API Problems That May Be Overcome (Right Column)
Dissolution and/or solubility limited absorption.
Includes low crystallinity.
Metabolic and/or pH-dependent stability.
Class I—relatively easy to be integrated in the normal timeframes
Low solubility,a Crystal form control,b Hygroscopicity
Self-(micro)emulsifying drug delivery systems
Low solubility,a Crystal form control,b Hygroscopicity
Controlled release tablets/capsules (such as multiparticulate tablets, enteric coating)
Class III—formulation strategies must be shared before identification of candidates
Other administration routes (such as injection, inhalation, or transdermal therapeutic systems)
Low solubility,a Crystal form control,b Low permeability, Low stability,c Undesired distribution, side effects
Class I Technologies
Liquid dosage forms, such as liquid-filled capsules and self-emulsifying drug delivery systems,9–13 are relatively easily integrated in the normal preclinical timeframe due to their simplicity. The content uniformity is not usually of concern in such formulations. Attention is required for the chemical stability, because the stability is usually lower than the solid dosage forms. The equilibrium solubility must be evaluated carefully, because supersaturated state is easily attained compared to the aqueous systems. However, their evaluation is not time-consuming, and thus decision on the employment of the dosage form is quick. Although a suitable combination and mixing ratio of surfactants and oils must be established for self-emulsifying formulations, this can be done by a screening in advance of candidate identification. While the dissolution of a candidate compound in oil affects the phase diagram, it does not have a significant effect on the absorption behavior, that is, the formation of a (micro)emulsion structure is usually not crucial for absorption enhancement as shown later.
The self-emulsifying formulation is regarded as a powerful option for BCS class II drugs, that is, drug molecules with high permeability and low solubility. However, it is not always effective for class II drugs, and small animal studies have shown that the type of excipients used has a significant effect on absorption behavior. Table 2 shows the absorption of poorly soluble drugs from various self-emulsifying formulations in rats12 and beagle dogs.10 In both cases, significant improvement in the absorption was achieved by using the liquid dosage forms. However, effect of the addition of surfactant including Tween-80 was only marginal, suggesting that the formation of the microemulsion is not a critical factor in this comparison. Remarkable improvement in absorption was achieved by using an HCO (Cremophor)-based formulation, probably due to its gelation behavior on addition of the aqueous phase, which is not a rare observation in water–oil–surfactant systems. Also shown is very low absorption from a C12E9-based formulation in rats. However, such variations in absorption behavior for similar formulations were unclear in beagle dog studies. This difference might be due to the significant difference of the phase transition behavior depending on the intestinal environment (notably volume). Also, it should be noted that the administration volume of the self-emulsifying formulations tends to be very large, and thus the use of the small animals such as rats and mice may not be adequate in many cases. On the other hand, oral administration studies in large animals including dogs may be difficult to carry out in a timely manner due to limited availability of the animals. Further investigation to allow elucidation of absorption behavior in different animal species seems to be required for this type of formulations, if it is regarded as an option for the formulation development in the normal timeframes.
Table 2. Animal Studies of Microemulsion (ME) Formulations for Poorly Soluble Drugs
AUC (µg h/mL)
AUC Ratio (N/F)
C12E9, polyoxyethylene (9) lauryl ether; HCO, hydrogenated caster oil; N, normal condition; F, fasted condition; NT, not tested.
Rat study (nitrendipine, 12 mg/kg)
Dog study (compound X, 5 mg/kg)
If a company has good practice of controlled release formulations, such as multiparticulate systems or the enteric coating technology, it may not be difficult to apply those techniques within the normal timelines. It is relatively easy to manufacture the multiparticulate systems including multitablet capsules and particle-containing tablets, and they enable programmed release of the drug molecules. The controlled release formulations are usually used with the aim of modifying unacceptable pharmacokinetic behavior. On the other hand, liquid dosage forms may be applied for poorly soluble compounds, of which the dissolution process regulates absorption behavior. However, there is no assurance that the special dosage forms improve the absorption or pharmacokinetic behavior of the compound of interest, and thus the use of such formulations must be justified in animal studies. This evaluation may require a few additional weeks, but it should not alter the developmental timelines if the effect of the special dosage forms is significant. When the effect is modest, business decision is required, because the animal studies cannot be repeated many times within the timeframe. Thus, the criteria for a decision to employ a special dosage forms should be established beforehand even for the class I technologies.
Class II Technologies
Amorphous formulations have been of great interest in recent formulation studies,14–17 since their solubility is higher than that of the crystalline formulations. There are two types of representative amorphous formulations, solid dispersions for the oral delivery and the injectable freeze-dried formulations. Although the latter one is out of focus in this article, the similar methodology can be applied for the poorly soluble drugs for the oral delivery using organic solvents including tertiary butyl alcohol as a freeze-drying solvent.18–20 It is even possible to use amorphous API under the normal formulation procedure.
Only a few oral amorphous formulations have been available on the market so far because of various problems including difficulty in mass production. At this moment, the most representative method for producing the oral amorphous formulations is hot melt extrusion technology,17 in which the drug powders are mechanically mixed with the carrier in an extruder. Another unfavorable aspect of the amorphous formulations is their poor chemical and physical stabilities. The chemical stability can be assessed from accelerated studies at higher temperatures, based on the assumption that the temperature dependence of the stability roughly obeys the Arrhenius rule. However, this type of analysis does not exist for the physical stability.21, 22 Much effort has been made to correlate crystallization behavior with various factors, including molecular mobility,20, 23, 24 thermodynamic parameters,22 and preparation methods (history of the material).25, 26 Although crystallization behaviors can partially be explained with reference to these factors, there are no general rules for predicting crystallization kinetics at certain temperatures. Therefore, to exclude the possibility of the crystallization during the storage, the physical stability studies must be done either at room temperature or under other storage conditions that pertain during the clinical studies. It should also be noted that performance of the amorphous formulation, including dissolution rate, may be altered as well with the lapse of time even without crystallization tendency.27 This could be explained in terms of the relaxation phenomena, which is still one of the developing fields in the amorphous research.28, 29
Figure 1 illustrates how the preparation methods affect the thermal history of the amorphous formulations. In solvent-free methods, the crystal lattice structure is destroyed by high temperature in the quenching method and by mechanical stress in the cryo-milling method. In the hot melt extrusion and in normal milling, both of these contribute. One of the main reasons for the occurrence of the crystallization after the preparation is imperfect amorphization, that is, the presence of remaining nuclei or small crystals. In solvent-based methods, the lattice structure is completely destroyed during the preparation process. However, the presence of residual solvent can powerfully enhance the crystallization. The spray-dry and the super critical fluids methods are conducted in a high-temperature environment, while the freeze-dry and the vacuum-dry methods usually provide a low-temperature environment. The high-temperature history may enhance crystal growth, while the low-temperature condition may help the nucleation process. The crystallization kinetics and its temperature dependence of a candidate compound, once elucidated, may become useful for selecting the optimum preparation method based on the expected thermal history. However, such investigation is time-consuming, and thus cannot be considered within the normal developmental timelines.
Table 3 shows the absorption behavior from the amorphous formulations, in which their solubility-enhancing effect within a few hours was shown as well. Unlike the self-emulsifying formulations, the amorphous formulation is effective for relatively various types of the poorly soluble drugs, although the degree of the absorption enhancement is difficult to predict. The solubility within a few hours can sometimes be an indicator to predict the absorption behavior as can be seen in the examples of JNJ-25894934 and carbamazepine. It is difficult to determine the drug/excipients ratio, and thus many studies simply employ the equal amount with the drug. It should be minimized from the viewpoint of the administration volume, however, it may cause physical stability problems. The important notion to solve this problem, the “solubility” of the drug in the solid-state matrix, will be discussed later.
Table 3. Animal Studies of Amorphous Formulations for Poorly Soluble Drugs
Conditions for solubility determination were as follows: *1: 0.1 N HCl at 37°C, 1 h, crystalline drug as reference; *2: pH 6.8 water at 37°C, 20 min, crystalline drug as reference; *3: pH 7 buffer at 37°C, 2 h, crystalline drug as reference; *4: simulated intestinal fluid at 37°C, 1 h, crystalline drug as reference; *5: 0.1 N HCl at 37°C, 90 min, physical mixture as reference; *6: water at 25°C, 30 min, physical mixture as reference; *7: pH 6.8 buffer at 37°C, 1 h, crystalline drug as reference; *8: 0.1 N HCl at 37°C, 2 h, crystalline drug as reference.
Nano suspension is a formulation technology with high potential,38–40 which may require a longer developmental period. Some top-down techniques including the wet-milling are available for obtaining the nano suspensions, and the first requirement for completing development within the normal timeframe is the successful application of the prototype additives and the preparation procedure. Although down-sizing to a few hundreds nanometer is frequently achievable, typical problem is the aggregation after the preparation. Synthetic polymers are the normal choices to keep the dispersity. However, it may significantly increase the viscosity to cause the handling problems. The obtained suspension can be transformed into solid dosage forms such as tablets, which is frequently required to assure the chemical stability. In this case, redispersity may become an issue. This type of formulation has the same hurdle as is experienced for the amorphous formulations, that is, the difficulty of the accelerated evaluation of the physical stability. The temperature dependence of the dispersed state stability is difficult to explain in a quantitative manner. Therefore, careful investigation is necessary for this type of formulation.
Class III Technologies
Drug discovery research is usually performed under the assumption that the drug molecules will be administered orally or via injection. However, other promising routes are of great interest as well for systematic drug delivery. Transdermal therapeutic systems41 enable prolonged action by drug molecules as proved by the nicotine formulation. One route that has come to most attracted in recent years is inhalation,42 which is regarded as important, in particular, for peptide delivery. The first inhalable formulation for systematic drug delivery was that of insulin, Exubera, which was commercialized by Pfizer (New York, NY) in 2006.43 However, this formulation was quickly withdrawn from the market in 2007 due to disappointment sales,44 and some other companies subsequently terminated the development of inhalable insulin formulations. At this moment, the front runner in this field is the company called MannKind (Valencia, CA) with their Technosphere technology,45–47 in which a palm-top device and particle engineering technology are combined. Because it delivers insulin in a monomer form, unlike the injectable insulin which utilizes a hexamer form, the peak blood level is obtained within 12–14 min, while the injectable formulations require 45 min to reach the peak monomer concentration. There are still many issues to be overcome with the inhalation therapy, such as difficulty in using the device, safety concerns, and low bioavailability, and thus much effort is ongoing in this area. Despite the disappointment with regard to the insulin formulations, systematic delivery via the pulmonary route is still promising, notably for poorly permeable drugs such as peptide drugs.
Physicochemical requirements for candidates depend to a large extent on the administration route, and thus this has a significant effect on molecular design. Because these requirements need to be considered during the candidate selection process, accommodation of this kind of technologies is not viable after identification of the candidate.
SELECTION OF THE DOSAGE FORM
Figure 2 shows a proposed decision tree for selecting the dosage forms based on physical characteristics of API and the animal studies. Problems found in preliminary animal studies such as low absorption and short half-life in plasma are usually the reasons to consider the special dosage forms. The problems in the pharmacokinetic profiles may be overcome by the controlled release technology. If the reason of the problem is the low solubility, one simple solution is the use of the salt, which is usually screened in the API definition process in pharmaceutical industry. Employment of the metastable crystal forms is theoretically possible as well. However, this is not a common strategy, partially because of the low solubility ratio between stable and metastable crystal forms48 and difficulty in assuring physical stability of the crystal form. The most simple formulation strategy to be considered is the solubilization to the oils including plant oils and medium chain triglycerides. Addition of surfactants may or may not improve the performance as described previously. However, because the required dose for phase I studies are extremely high in many cases, the formulation volume may become very large. In that case, solid dosage forms should be considered. Some additives including cyclodextrins and surfactants may solve the problem. If not, options for formulators are use of the “Class II technologies” or request for chemists to modify the chemical structure itself. If the class II technologies are selected, the timelines are frequently affected as shown previously. In this selection process, multiple animal studies may become required. However, because it is very time-consuming to repeat the animal studies, simultaneous investigation of multiple formulations, such as combination of the oil solution, the self-emulsifying formulations, and the cyclodextrin formulations, should be the practical approach.
Various factors must be taken into account to determine the developmental timelines in addition to the time required for the formulation development study. The most important one should be project feature and status, such as whether or not POC was achieved, the availability of biomarkers, the status of lead/backup compounds, and the status of pipelines of other companies. Unless POC has been obtained yet, formulators are usually not allowed to spend much time on the formulation development, and thus simple formulations are favored. Whether the disease is acute or chronic significantly affects the formulation strategy, because the required stability is different. Much resource may be invested to first-in-class or best-in-class compounds. Needless to say, the formulation technologies that the company has much experience should be applied relatively easily and vice versa. One of the very difficult aspects may be “risk-taking.” There is no doubt that careful investigation can reduce the possibility of unexpected failures during the clinical studies. However, no matter how much effort is invested, the risk never reaches 0%, and thus a decision must always be made on the balance of risk. This in turn is significantly influenced by culture of the company and the mindset of individuals involved in the project.
Class II and III technologies are not basically suitable for the early phase formulation. However, they are promising candidates for LCM purpose. One good example is Kaletra, a formulation for treatment of HIV, which was originally developed as the self-emulsifying formulation but was later replaced by the hot melt extrude. This modification allowed a significant reduction in the formulation volume, from six capsules to four tablets per day. Also improved was its storage condition, from refrigeration to room temperature. Parallel development of class I and class II/III formulations may worth being considered as well. In this case, the clinical studies of the class II/III formulation can be initiated soon after obtaining POC. If the class II formulations are expected to be the only options for achieving adequate exposure in clinical studies, the development timelines need to be flexible to allow adequate formulation developmental period.
RESEARCH DIRECTION OF THE AMORPHOUS FORMULATION
Current major problems that inhibit the use of the amorphous technology are difficulty in mass production and the poor physical/chemical stability. As for the mass production, although many technologies are already available as shown in Figure 1, most of them including the hot melt extrusion require energy stress that may cause degradation. This is inevitable as far as the crystalline state is used as the starting material. Another unfavorable aspect of the manufacture from the crystalline state is the low system homogeneity that may cause the scale-up problem. Therefore, solution seems to be a better starting material to produce amorphous formulations. Both of the two representative manufacturing methods that start from solution, the spray-dry and the freeze-dry methods, require temperature stress and relatively high operation cost. Thus, novel technologies should be developed to remove solvent under mild condition. One good example is the use of the electrospray technique, where the solutions are sprayed by aid of high voltage.49 This technology can be performed under ambient temperature condition, and thus may be promising for temperature-sensitive ingredients.
Physical stability of the amorphous materials is still a difficult issue to be understood.22, 50 The crystallization theory can be separated into two processes, nucleation and crystal growth. The latter process has frequently been investigated in a quantitative manner for the amorphous formulations using the Avrami equation.51–53 However, the rate-limiting step of the crystallization is usually the nucleation process,53 which is still difficult to observe and predict. In addition to this widely accepted crystallization theory, in which the diffusion is regarded as the kinetic barrier for the crystal growth, the “glass-crystal” mode, in which other molecular motions including the local mobility are regarded to be dominant, has been discussed to explain the fast crystal growth near the glass transition temperature (Tg).54, 55 Because this mode is supposed as important in the temperature region which is of great interest for the pharmaceutical glasses (just below Tg), progress in this new theory may worth being focused.
There have been some attempts to directly observe the nucleation process using synchrotron X-ray scattering.56, 57 This kind of observations needs to be linked to the crystallization studies, in which the change in the crystallization behavior was explained in terms of the nucleation during the annealing.26, 53 Factors that affect the crystallization rate are well summarized in the review of Bhugra and Pikal,22 where molecular mobility has been shown to be an important factor to determine the crystallization behavior. However, other factors need to be considered as well to explain all the investigations in this matter.22 The same idea is valid for the chemical stability as summarized in the review of Yoshioka and Aso.58 The molecular mobility is likely to be a dominant factor that affects the chemical stability in many cases, while other factors including chemical activation energy and the reaction with excipients can rule the stability as well. The physical/chemical stability problem should be overcome by comprehending the mechanism precisely. However, there are still many issues unsolved regarding “molecular mobility” including homogeneity of the relaxation time and its dependence on the amorphous structure. Dynamics of the amorphous structure is a basis to understand each problem, and thus requires extensive investigation.
The discussion above is based on the assumption in which the drug molecules are homogeneously dispersed in the amorphous matrix. However, this is basically achieved when the “solubility” of the drug molecules in the matrix is higher than the prescribed amount. Although it is tentatively achievable as the supersaturated system, the physical stability may become concern in that case. The simple approach to determine the solubility is the investigation of Tg.59, 60 If the system is homogeneously mixed, single Tg should appear between Tg of each component. However, equilibrium state is difficult to obtain because of the slow diffusion, and thus the phase separation may proceed during the storage for long period.59, 60 The precise determination of solubility is difficult even in the liquid systems,61 in which diffusion is much faster than the amorphous systems. Therefore, there is great limitation in the phenomenological reasoning on the solid solubility from the observation of the amorphous formulation itself. Attempt to determine the solubility using thermodynamic approach is in proceeding.62–64 This kind of research should help design of the stable amorphous formulations and may lead to promotion of the amorphous formulations to the “class I” technology.
The special dosage forms that may be integrated in the development strategy of new chemical entities are summarized with an emphasis on the use of the amorphous dosage form. Some formulations, such as the self-emulsifying drug delivery systems, may be developed within normal timeframes, despite their sophisticated formulation design, if the formulators have a good practice in the area. On the other hand, the special dosage forms, for which a protocol for assessing the physical stability has not yet been established, such as the amorphous formulations, are essentially not applicable within the normal developmental timeframes. Some basic investigations including assessment of the molecular mobility are still required for overcoming problems of the amorphous formulations that inhibit their general use in the pharmaceutical industry.