Unfortunately for pharmaceuticals manufacturers, drug products can rarely be supplied to patients as the pure active pharmaceutical ingredient (API). Instead, APIs must be formulated into drug products for several reasons, including for consistent and convenient dosing; to deliver multiple strengths or delivery platforms for different patients; to improve bioavailability or stability; to obtain a better drug release profile to ensure efficacy with no side effects – or simply to remove a bad taste or odour.
The need to convert APIs into a dosage form comes at the preclinical stage. In fact, some inactive ingredients – excipients – used in formulations are not always ‘inactive’ and may themselves have daily dosing limits.1 And given that many clinical trials include a placebo arm in their design, the placebo dosage form needs to have the same excipients as the clinical trial material (CTM).
In general, Phase 1 trials involve tens of healthy human volunteers, and only a few hundred CTM dose units. While only 1kg or less of API may be needed, however, two to five times that amount is made to support stability studies to establish the shelf-life of the drug product.
Phase 2 often involves hundreds of targeted patients to set the effective doses and safety limits. The API demand for Phase 2 CTM batches can approach 1000kg, and these studies can last years before moving into Phase 3. Whereas preclinical and Phase 1 studies could be supported from a single batch of API, subsequent studies mean additional batches at current or greater scale.
Phase 3 clinical studies involve many more patients, and the API and drug product are now needed at or near commercial scale. Multiple batches are required, and it is here that fully automated and well-controlled processes are defined. The market presentation of the drug product is nearly set, and all CMC data for both API and dosage form are compiled to support the new drug application (NDA) regulatory filing.
Unless there are significant problems with the API, drug formulations for the preclinical and Phase 1 clinical trials can be prepared by simple dissolution of the API into a buffer for parenteral (non-oral) administration, blending it with a filler to make a powder that is filled into a capsule, or mixing it with a liquid just before ingesting.
The formulation ingredients must not degrade the API to diminish its potency or produce degradation products that pose a threat to patients. The formulator will perform an API-excipient compatibility study to select only those additives that yield a product stable for a year or more. The chemical properties of the API and all potential excipients are accounted for in the study design. For example, any potential formulations of a primary amine-containing API would exclude reducing sugars such as fructose that would yield undesired reaction products. Likewise, a buffered solution for injection with a carboxylic acid-containing API would have a pH where any equilibrium reaction is selected to favour the desired active species.
If all of the steps in the synthesis, purification, and isolation of an API batch yield material with exactly identical physicochemical properties then formulating for larger quantities will be easy. Unfortunately, seemingly insignificant changes can yield big differences from lot to lot, and the effect on the formulation process — and the performance and stability of the drug product — can be disastrous. Moreover, some chemical process changes could yield a new crystalline form or ‘polymorph’ that is a different three-dimensional solid-state arrangement of the API, or a solvate of the API — usually with water — which requires careful examination of its safety and suitability for the intended dosage form.
One example where changing one process parameter led to the disastrous appearance of a polymorph was seen in the production of a marketed depression drug. There were multiple known polymorphs, but most of the forms could only be obtained under somewhat exotic conditions. The manufacturing process produces a hydrate, which is then dried to give another polymorph. During a validation campaign, the heating ramp during the drying cycle was run slowly during the first two batches, and the right polymorph was produced. For the subsequent batches, the heating ramp was accelerated in order to gain efficiencies. This produced another unexpected polymorph. Surprisingly, the heating process in fact turned out to be quite critical late in the development of this drug.
A new solvate or polymorph is viewed as a unique API from a regulatory perspective; thus, the chemistry and physics must be characterised, and controls established. This will likely include repeating some of the preclinical work because properties such as bioavailability can be vastly different from the original form. While some investigators will screen for salt forms and polymorphs shortly after discovering a so-named new molecular entity (NME), this can require significant personnel and financial resources for emerging pharma groups, and it is often deferred until a problem occurs.
The control of chemistry is no small feat, but once the API form is ‘locked in’ and can be made reproducibly many times, the physical properties of the API often present the most headaches. Each batch of bulk powder API has a measurable particle size distribution, crystalline shape, density, equilibrium moisture content, plasticity or brittleness when compressed, and propensity to self-adhere or clump. Formulators will select the best components for the dosage form to account for those unique attributes.
When an attribute changes for the API, the effect on the next batch of product can be a major problem. For example, if the API particle size decreases substantially, then the API and excipients may not mix well or remain homogenous after blending. This can result in doses with incorrect potency or tablets that will not compress properly.
If we focus on solid oral formulations, APIs will be mixed with other powders to prepare capsules or tablets. These dry excipients include bulking agents that dilute the API to a consistent potency for each dose; wetting agents that draw in water so that API particles will disperse and dissolve; flow-aids that enable self-sticking APIs to be transferred by gravity or mechanical means to a process step; binders that enable the API blends to stick together and form a stable tablet when compressed; disintegrants to cause the compressed tablet to break into granules and disperse the API for dissolution when ingested; and functional polymers for coating tablets that dissolve at predefined pH ranges.
Excipients are comprised of starch, modified celluloses, sugars, inorganic salts, and synthetic polymers, such as polyvinylpyrrolidinone and polyethylene glycol. They can present some potential chemical issues — for example, reaction with the API — but like the API combined with them, each excipient varies in particle size distribution, density, and other physical properties that determine how well-mixed the formulation can be prepared.
For many Phase 1 clinical trials, some APIs can be easily formulated as simple blends with one or more excipients, and the blend can be filled into bottles or capsules shells or compressed into tablets. Sometimes, the API’s particle size matches closely that of the excipients, and the blend flows easily and does not segregate when mechanically treated or stored for long periods. However, Phase 1 CTM is rarely close to the final market strength or image, so additional formulation work is required. In addition, greater demand for API and product usually requires an improvement in the physical properties of the API—excipient mixtures so that fully automated and commercially scalable manufacturing equipment can be used.
Another challenge for API formulation beyond Phase 1 is when the release rate into the body must be modified, such as when side effects result from the entire dose being absorbed at once or when a therapeutic effect requires a longer absorption time. In these cases, excipients can be selected such that a tablet erodes slowly or a polymer coating doesn’t dissolve in the intestines until a certain pH is achieved past the stomach. While the dissolved API released from these dosage forms could simply be absorbed, the API may precipitate or degrade in the intestines and impair the efficacy of the drug product. Further formulation studies after Phase 1 are almost always required.
To summarise, the development of an API into a drug product requires knowledge of both the chemical and the physical properties of the molecule and all the components that make up the dosage form. While most formulations in Phase 1 clinical studies are developed as simply as possible, later clinical stages require greater amounts of API and product, and this increases the likelihood of potential changes to the properties of the API and the type of drug product that will be advanced for regulatory approval.
Kevin Kane and Brittany Hayes are applied technology directors at Patheon, 4721 Emperor Blvd, Suite 200, Durham, NC 27703, US.
An early lead
To understand how to advance an API into the clinic, it is helpful to be aware of the overall process. A new molecular entity (NME) is discovered to have a physiological activity that is deemed potentially beneficial. This potential API is characterised for identity, activity, and metabolism in cell cultures and then in animal subjects – this may involve screening quite a few NMEs within a structural class to find the right ’hit‘. Laboratory quantities of 10–200mg are used at this stage, and doses are often delivered as aqueous solutions or suspensions or DMSO-H2O solutions, depending on the NME’s water solubility.
Once the lead NME is selected, scale-up synthesis, purification, and characterisation are needed to support the preclinical toxicity and pharmacokinetic (PK) studies in at least two animal species. These studies are subject to regulatory guidelines, so the API used in toxicity and PK studies is usually manufactured according to current Good Manufacturing Practices (cGMP),2 as these preclinical results will be used to justify clinical trials in humans via an Investigational New Drug (IND) application. Anywhere from hundreds to thousands of grams of API will be consumed in this preclinical stage, and investigators may supply this from several small batches or a single large batch of the molecule.
1 FDA’s Inactive ingredient database is updated periodically at: http://www.fda.gov/Drugs/InformationOnDrugs/ucm113978.htm
2 See Guidance for industry regarding APIs: http://www.fda.gov/iceci/compliancemanuals/compliancepolicyguidancemanual/ucm200364.htm