First, as these alternative dosage forms represent a small fraction of the market, for most organizations, even large ones, their development is a rare event and therefore relevant formulation experience is diffuse. Additionally, sensory analysis and flavor system development are not core competencies of the pharmaceutical industry. Many dosage forms and technologies trace their origins to the food industry, which continues to be a rich source of approaches, tools and methods that can be adapted by pharmaceutical scientists to develop palatable drug products.
Before delving into the process for developing palatable drug products, it’s important to define some key terms. To a food or sensory scientist, the term “flavor” refers to the combination of a product’s taste, aroma, mouthfeel and texture, each of which is briefly described below.
Taste refers to those sensations perceived through the stimulation of the receptor cells located in the taste buds on the epithelium of the tongue and oral cavity. There are five distinct tastes, known as basic tastes, that are perceived in the oral cavity—sweet, sour, salty, bitter and umami (savory). The receptor pathways differ across the basic tastes—ion channel receptors for sour (H+) and salty (Na+) and more complex G-protein-mediated signaling pathways for sweet, bitter and umami.
Aromas (odors) are the volatile chemical compounds perceived via the sense of smell (olfaction) through stimulation of receptor cells in the olfactory epithelium located in the upper reaches of the nasal cavity. These volatile aromatics reach the olfactory epithelium via the nose during normal breathing (orthonasal olfaction) or via the nasopharangeal passage during mastication when air is exhaled through the natural action of swallowing (retronasal olfaction).
Mouthfeel refers to sensations (feeling factors), that arise when chemical compounds directly stimulate free nerve endings in the trigeminal (5th cranial) nerve. This is known as chemesthesis, and examples include cooling, numbing and bite/burn.
Textures are the tactile characteristics of a product perceived in the oral cavity. These include those surface attributes measured initially, as well as when a product is deformed through mastication. Notable textural attributes for oral liquid drug products include viscosity, smoothness, and mouth coating. For oral solid dosage forms, important texture attributes may include roughness, hardness, fracturability, and cohesiveness.
Many APIs have a bitter taste or have other aversive sensory attributes, frequently learned anecdotally in the clinic. With little else to go on, often a formulator’s first reaction to taste masking is to add a commercial flavor to the formulation in an attempt to mask the bitterness. Commercial flavors are proprietary blends of aroma chemicals that are perceived by olfaction. As described, taste and smell are comprised of fundamentally different receptors, transduction pathways, and loci of perception in the brain. Just like the sense of sight has no impact on hearing, aromas—orange, grape, chocolate, or mint—cannot mask bitterness.
So how then does a formulator develop palatable a drug product and what is even meant by palatable? The food industry strives to create products that “delight the palate” and provide sustenance and enjoyment. Drugs on the other hand are developed for efficacy and safety, though palatability is important for successful dose administration, particularly with children. Recognizing these important differences between food and drugs, palatable drug products are those in which the aversive sensory attributes have been minimized or eliminated. In other words they are not overly bitter, produce little trigeminal irritation, are smooth, not gritty and have no perceptible malodors.
With this background in mind, we turn to the challenge of taste masking. Figure 1 outlines a structured framework for developing palatable drug products. The framework is a decision-tree that identifies the sequence of questions that need to be answered to guide formulation development. It is built upon two pillars: First, the adage that you can’t improve what you can’t measure; is the formulation challenge bitterness, trigeminal irritation or a malodor? Second, the art and science of taste masking derived from the food and allied industries.
The flowchart is divided into four sections. The top section labeled “Taste Assessment” represents the key questions that need to be answered early to support the development of age-appropriate, palatable drug products. It starts by asking Question 1: Are the aversive sensory attributes of the API known? For example, is the API bitter or produce trigeminal irritation? Does it have an offensive odor or unusual texture? Consumers and patients alike have difficulty accurately describing their perceptions. For example sour and bitter basic tastes are often confused. Regardless of the underlying reason—bitterness, trigeminal irritation, or malodor—patients and untrained healthy volunteers simply describe the drug as “bad tasting.”
It is critically important to collect sensory data early in development ideally at the conclusion of Phase 1 clinical trials. Specifically, this should include both the perceived strength and duration of each aversive attribute (Question 2). Due to the difficulty in deciphering patient responses, trained adult taste panels are frequently used to generate both qualitative and quantitative data. The Flavor Profile method, one of several published (open source) sensory analysis methods, was used to generate the data presented herein.
Most API’s require some form of taste masking to yield a palatable drug product. Some can be effectively masked with a properly constructed excipient system comprised of sweeteners, buffers, taste modifiers, flavoring aromatics and extenders, and sensate materials. If an API has an extremely aversive flavor profile, ameliorating the taste with an excipient system may be insufficient to develop a palatable formulation. In these cases, the API may need to be “sequestered” from the taste receptors using particle coating, adsorption or other technology to yield a palatable drug product.
Sensory time-intensity profiles provide an indication of the overall taste masking challenge and can be used to answer Questions 3: Will an API modification technology be required for taste masking? Figure 2 shows the bitterness profiles of four different APIs. Based on an understanding of perception and recognition thresholds and the limitations of excipients in reducing aversive sensory attributes, approximate decision boundaries can be overlaid as shown by the color-coded boundaries.
When a taste masking technology such as encapsulation or adsorption is required (Compound D in Figure 2), additional questions need to be answered to guide development and understand their technical boundaries as highlighted in the “Technology Guidance” section of the flowchart (Figure 1).
Dose-response sensory analysis is a useful tool for generating quantitative data to guide technology development/optimization by addressing Question 4: Is the maximum free-API concentration known? Maximum refers to the highest “free“ (soluble) API concentration capable of generating a palatable formulation. Figure 3 represents the bitterness profiles for eight concentrations of API from 2mg to 100mg/mL. The same decision boundaries shown in Figure 2 have been applied, indicating that above a free API concentration of approximately 33mg/mL, masking of the API becomes a difficult challenge. With this knowledge, chemical analysis can then be used to support technology development/optimization, reducing free API concentration below 33mg/mL and answer Question 5: Is the free-API concentration below the maximum?
The third section of the flowchart (Figure 1) labeled “Formulation Guidance” outlines a formulation strategy for developing palatable drug products, and consists of two parts. The first part is to minimize or eliminate the aversive (negative) attributes of the API. This is commonly referred to as creating a “neutral” tasting or “white” base. A neutral base exhibits balanced basic tastes, which is a foundational principle of taste masking. In other words, bitterness can be reduced with proper balancing of the complementary basic tastes—sweet, sour and salty—via the mechanism of taste/taste interaction.
The starting point for taste optimization is a preliminary formulation containing the API and the excipients that enable drug delivery in the selected dosage form (functional excipients). The API may be in its native form—free acid, base or salt—or modified for purposes of delivery or taste masking as previously described. Excipients serve many functions: to improve drug solubility (e.g., solvents, co-solvents, surfactants); maintain physical, chemical and microbial stability (e.g., buffers, preservatives, antioxidants, suspending agents); enhance disintegration; control release; and improve manufacturability, to name a few.
Unfortunately, many excipients also have aversive sensory attributes (Figure 4), sometimes more than the API, that need to be considered early in the formulation development process to avoid creating or exacerbating the taste masking challenge. For example, in conducting design-of-experiments for solubility enhancement, it is important to assess the sensory attributes of the excipient blends in an effort to simultaneously optimize for solubility and taste.
The first formulation step is to develop a “mimetic” system to reduce human exposure to drug substances during development. A mimetic system uses Generally Recognized as Safe (GRAS) excipients to replicate the aversive sensory attributes of the API. It is critically important that the mimetic system accurately represent the API bitterness time-intensity profile, for example, as it is much more difficult to mask a lingering bitterness compared to one that is short lived. Bitter mimetics include caffeine, quinine sulfate, and sucrose octaacetate, among others.
Once the mimetic system has been developed, the next step is to develop the “neutral” tasting base. The taste masking process typically begins with the sweetener system. There are numerous candidate nutritive sweeteners and non-nutritive sweeteners including sugar alcohols and high intensity (artificial) sweeteners. The selection of sweeteners is informed by the results of the taste assessment (i.e., the magnitude of the taste masking challenge) as well as technical, regulatory and clinical considerations.
Nutritive sweeteners and sugar alcohols generally have linear dose-response curves, while high intensity sweeteners are generally nonlinear. The nonlinearity of high intensity sweetener dose-response curves makes it impossible to express sweetness as a single value multiple of sucrose, though data on sweetness equivalency is widely reported in the literature without reference to the corresponding sucrose concentration (Figure 5). Additionally, even at the same intensity, sweetness onset, duration and decay profiles are different between sweeteners. The net result is that sweetener systems, like solubility systems, cannot be selected a priori but rather must be determined experimentally.
Following development of the sweetener system, the next step is to explore sourness. When pH has been established as a formulation design criterion, sourness can be adjusted via buffer strength (i.e., titratable acidity). Increasing perceived sourness reduces both the perceived bitterness and sweetness. Lastly, the addition of sodium chloride, or other sodium salts, for saltiness can be beneficial in further reducing bitterness and improving the overall blend, similar to how it works in foods.
Once candidate neutral base formulations have been developed, it is common practice to confirm that these unflavored bases perform as expected when the mimetic is replaced with the API. Question 6: Does the neutral base perform as expected?. Based on the results, adjustments in excipient usage levels may sometimes be warranted.
Some CDMOs confirm the compatibility of the white base components at this stage but most wait until the complete flavor system has been developed. To mitigate technical risk, it is common practice to develop multiple—minimum of two—flavor systems. This increases the likelihood that one will be compatible with the API.
After the aversive attributes have been successfully ameliorated in the neutral base formulation(s), age-appropriate flavoring aromatics may be incorporated. As described herein, these identifying flavors, e.g., orange, strawberry, grape, do not play a role in the reduction of aversive basic tastes, e.g., bitterness. Rather, flavor aromatics are used to mitigate API or excipient malodors that may cause the patient to reject the drug product off the spoon or even upon first opening the bottle. Flavor aromatics are generally the first attributes perceived during dose administration and therefore create a positive initial impression.
Globally, there are hundreds of independent flavor suppliers, which mostly support the food industry, so it is important that the partner of choice be familiar with pharmaceutical requirements. The flavor supplier should be willing to support Drug Master Files—many do not—and have a strong regulatory department to ensure the flavor composition is acceptable in the drug product’s target markets—North America, EU, Japan.
Once patient demographics are considered, samples are requested from selected suppliers and evaluated by the sensory panelists in the mimetic base. It is not unusual to source and screen dozens of flavors from multiple suppliers and themes (e.g., citrus, berry, mixed fruit) to identify the one or two that work best with the specific API and excipient system.
Next, the mimetic is again replaced with the API and the leading sweetened/flavored formulations evaluated by the trained human taste panel to confirm they perform as expected from the mimetic system to answer Question 7: Does the flavor system perform as predicted?
Following taste optimization of the drug product, the formulations are scaled up to Clinical Trial Material (CTM) batch size with requisite compatibility and stability testing. Steps associated with ensuring product palatability are shown in the last section of the flowchart, labeled “Clinical/Commercial Development.” Since the target shelf life for most drug products is two years, the drug product should maintain its palatability over this period. Therefore, it is common practice to assess the sensory attributes of stability samples stored at 30°C/65% RH for 6, 12, 18 and 24 months. Sensory evaluation of samples at higher temperatures is not recommended, as it is not predictive of flavor degradation.
Though necessary, development of palatable oral formulations represents a small fraction of all drug products. Not surprisingly therefore, sensory analysis and taste masking/optimization are not core competencies of the pharmaceutical industry. The structured framework described herein was designed to identify the sequence of questions that need to be answered along the clinical, formulation development and commercialization pathway. By following this process, CDMOs can effectively manage the myriad of technical risks associated with developing and manufacturing palatable drug products for their clients.
Jeff Worthington is President and Founder of Senopsys LLC, a specialty services firm dedicated to the development of palatable drug products. He has 25 years of experience in taste assessment and taste masking of investigational and approved drugs for children and adults. Mr. Worthington is Chair of the Applied Sensory Testing Workgroup of NIH’s Pediatric Formulation Initiative. He can be reached at firstname.lastname@example.org.
David Tisi is the Technical Director of Senopsys LLC. He has particular expertise in applying food technologies to the development of palatable, patient-accepted drug products. Prior to joining Senopsys, David worked as a product development scientist at Nestle and PepsiCo. He can be reached at email@example.com.