Stability is the Key

Biopharmaceutical formulation is primarily about stability and the product's mode of use. The route of administration (drug delivery, see next chapter) and the manufacturing process must be considered, along with the molecule's intrinsic stability factors. Because proteins and peptides are such large molecules and exist to interact with their environment, they are somewhat fragile. They must be protected from denaturation and degradation until they can be delivered to their site of action in a patient's body.

The biopharmaceutical development process does not allow enough time to confirm stability requirements for a final formulation (which could take two years) before the company is otherwise ready to apply to market the product. Initial indications should be developed during clinical studies, so formulators must begin with three-, six-, and nine-month tests of their molecule's structure and innate stability using various analytical methods. "Accelerated" stability tests subject products to various stresses: a range of pH values, heat, light, freezing and thawing conditions, additives, and surface materials and interfaces. The test for agitation-induced denaturation is performed by swirling (creating a vortex inside the vials), rotating vials at elevated temperatures, and testing the surface tension of the liquid formulation.

"The experiments necessary to define an acceptable formulation depend on the proposed route of administration," said D.B. Lakings, principal consultant at Drug Safety Evaluation Consulting. Will the drug be injected, inhaled, or delivered some other way? Consider stability in the bloodstream, Lakings suggests. A poorly soluble drug could precipitate out of solution in vivo — not only ruining its chance of doing its job but possibly endangering the patient. "Formulation excipients can possibly be added to prevent degradation," he said. Or the formulation may need to be stored under certain conditions — refrigerated, frozen, freeze dried, or kept in reduced-light yellow or brown vials, for example. The less stringent the storage conditions, the better.

The current market demands one to two years of storage for most products. During preclinical testing, "Samples obtained before shipment (and stored under conditions known to provide stability) and after shipment to a testing laboratory should be analyzed to provide information on potential problems in shipping," Lakings said. A protein that degrades in less than six months can be a problem if stability claims are declared after only three months of testing. Of course, it's normal to want to rush the development process in order to get a product to the market as soon as possible. After three months, a stability protocol and results can be sent to CBER, but it is good business practice to amend that information approximately every three months, extending the product's expiration date as far as possible. Considerations for making formulation decisions follow this order: (1) shelf life, (2) convenience of use, and (3) cost-effectiveness.

What Can Go Wrong

Here's one of the scariest things that can happen at a biopharmaceutical plant: A quality inspector looks at vials of final, liquid-formulated product and sees globs of brown gunk floating or settling to the bottom. An entire batch, worth at least thousands of dollars, has just been wasted. Usually the failure is not so obvious. Formulation scientists work hard to make sure aggregation and precipitation don't happen.

Biopharmaceutical formulators must evaluate protein formulations for degradation products both qualitatively and quantitatively. Qualitative studies include preclinical safety studies and tests that determine how degradation might be prevented by optimizing the formulation. Quantitative evaluation examines temperature dependencies, for example, or a formulation's tendency to convert in vivo to native forms (if it has denatured). Bioassays study protein activity, identity, and critical pathways. Analytical methods are developed in parallel with the product itself, with at least two solid techniques chosen to assess stability.

Chemical degradation changes the primary structure of a protein. Bond cleavage will create an entirely new molecule. Such chemical degradation is usually preceded by a causal physical process, typically unfolding, that makes available residues that are usually inaccessible for chemical reactions with their environment. Physical degradation changes only the higher-order structure (secondary, tertiary, quaternary) of the polypeptide, not necessarily creating a brand new molecule. Such degradation includes aggregation, adsorption, unfolding, and precipitation.

Oxidation. Certain amino acids (tryptophan, methionine, cysteine, histidine, and tyrosine) are susceptible to oxidation. Metal ions (copper, iron) can accelerate the process. Higher pH values, fluorescent lights, and hydrogen peroxide can all cause or contribute to oxidation, but atmospheric O₂ is obviously the main culprit. If the amino acids along a polypeptide chain are deformed by oxidation, the molecule can be irreversibly altered, and the new molecule created may not perform the necessary drug action.

Hydrolysis. Hydrolysis of a side-chain amide on a polypeptide's glutamine or asparagine residues can yield a carboxylic acid. The process, called deamidation, is facilitated by elevated temperature and pH, and its effect is to destroy the activity of most proteins. The peptide bonds that hold amino acids together in the chain can also be severed by hydrolysis — particularly where aspartic acid residues are located. This effect is usually due to heat or to pH levels that are too low.

Crystalline structure of a sugar, often used as a pharmaceutical excipient

Disulfide exchange. Cysteine residues form disulfide bonds, which are important to protein structural integrity. Shuffling of these bonds, where two sulfur atoms from two different amino acid molecules link up, often changes the three-dimensional structure, causing a loss of activity.

The Ingredients and the Recipe

Proteins are most stable in a "glassy," amorphous, and highly viscous matrix (freeze-dried or spray-dried). Technically speaking, glass is just a very hard gel. Liquid solutions must be kept very cold. Maximum stability occurs when conditions restrict molecular interactions and conformational changes. Various technologies offer several options to formulators, and more are being developed all the time. By far the most widely accepted methods of protein stabilization are additives and excipients, cryopreservation, and freeze-drying. Spray-drying, cryogranulation, and undercooling are less proven options.

Additives and excipients. Salts and nonelectrolytes (such as ammonium sulfate and glycerol) help stabilize proteins in high temperatures and low pH when freezing is not an option, but they still require low-temperature storage. Sometimes they must be removed before the drug is used, which can be inconvenient, time-consuming, and expensive. Also, the active ingredient must be diluted, allowing further waste and variability in the final product just as in reconstituting freeze-dried products.

Cryopreservation (freezing) can extend the shelf life of unstable products and improve containment if the freeze-thaw process is consistent. Frozen product can be transported safely for final formulation elsewhere and stockpiled to optimize the fill and finish process, which can reduce processing costs. More detail on this process begins in Chapter 4.

Many formulation scientists believe that proteins cannot be frozen and thawed without damage because no practical methods exist for doing so on a large scale. The usual method, which is both slow and nonreproducible, involves small volumes in bags, bottles, or vials freezing at —20°C or below. A slow freeze alters the physical properties (pH, diffusion, reaction rates) of the aqueous solvent medium or mixture, which can denature proteins, particularly over extended time. The solution is then thawed at room temperature, with some components thawing before others. Ice recrystallization in the thawing process creates mechanical stress. Processes for freezing and thawing protein solutions must, therefore, be well controlled.

Cryogranulation. One answer to the control problem might be cryogranulation. "Standard practice for finished product manufacturing of parenteral dosage forms begins with the bulk drug substance in the solid state, at which point it is dissolved in a suitable solvent during the formulation or compounding stage before filtration. However, some drug substances, particularly proteins, are very difficult, if not impossible, to crystallize," D.J. Schmidt and M.J. Akers wrote in BioPharm (April 1997). Cryogranulation uses a stream of liquid nitrogen to quickly create frozen discrete pellets. The main challenge is controlling thawing and reformulation on the clinical or final product manufacturing side.

Lyophilization (freeze-drying) doesn't always have an "either-or" relationship with cryopreservation; sometimes both are used. Because lyophilization is so often used, there is an expanded discussion later.

Spray-drying is a dehydration process that uses heat from a hot-air stream to evaporate dispersed droplets created by atomization of a continuous liquid feed. Products dry within a few seconds into fine particles (powders). It is similar to freeze-drying. Important data for formulators working with any type of freezing process will be the glass transition temperature (Tg) of each component and the solution (see box, page 18). At that temperature, ice crystal formation decreases to undetectable levels, and the freeze is an amorphous glass from which water will sublime.

Spray-drying offers some advantages over freeze-drying: shorter process times, lower capital investment in equipment, and lower energy input to the solution, which can lessen the chances for protein denaturation. Disadvantages include the brief but high air-temperature exposure during the rapid-drying step (around 100°C) and shear stresses caused by spraying a formulation through a nozzle. The product temperature reaches about 80°C but only for a fraction of a second.

The spray-drying technique is used with classical pharmaceuticals. "Spray-drying is a well-developed technique for creating pharmaceutically elegant powders, and it is readily adaptable for protein pharmaceutical applications," said Douglas Nesta, a development scientist at GlaxoSmithKline, at a meeting on formulations and biopharmaceutical drug delivery. However, he cautioned that a range of analytical methods "must be applied to biopharmaceutical powders in the solid state, as well as after reconstitution, to enable an accurate assessment of product quality."

Particle characteristics and size distribution can be closely controlled in spray-drying by changing process variables such as solution composition and feed rate, atomizing gas pressure, air flowrate, and the inlet and outlet gas temperatures. Physical characteristics of the final product (size, morphology, surface area, density of particles) may be manipulated, as well as biochemical characteristics (purity, potency, solubility, stability of the formulation) and process yields. The goal is a dry, free-flowing powder with well-defined particle characteristics, consistent purity and presentation, and an active, stable, and acceptable dissolution profile, all obtained by as simple a process as possible.

Undercooling. Under certain conditions, a liquid can be cooled to temperatures below its freezing point (supercooling or undercooling). Freezing is usually initiated by particulates, which act as nuclei. The process is similar to the formation of hailstones and how cloud-seeders induce rain out of an apparently dry sky by adding particles around which water condenses. If a solution is separated into droplets, as in a mist or a water-in-oil emulsion, then only those drops with particles in them will freeze if cooled, rather than catalyzing a chain reaction of freezing throughout the solution.

PEGylation: An In Vivo Stability Solution

For protein formulations, the product in aqueous solution can be dispersed as microparticles through an oil-phase carrier (liquid at the mixing temperature, solid if stored at -20°C), so the drops are locked in as liquid. Reconstitution is easier than with a lyophilized or spray-dried product: Users just warm the final product (sometimes with gentle centrifugation) until those phases separate. This undercooling process may be used for even high-concentration protein formulations, requiring no additives like glycerol, which must be filtered out before the product can be used. Undercooling may be a particularly good choice for products that are susceptible to freeze damage.

Excipients

No matter what kind of formulation or stabilization method is chosen, excipients of some kind will be involved. Each has at least one function, and synergistic effects are to be expected. The compound will behave most like its principal constituent. A combination may behave differently than planned, so formulation scientists can spend a lot of time empirically testing the ingredients in combination. Interactions are important: they can make the difference between a formulation that works and one that doesn't. Each ingredient added can change the behavior of the whole mixture.

Sugars (or saccharides) raise Tg and act as stabilizers. Dextran, lactose, maltose, sucrose, and trehalose are used, but the latter two are preferred. Acidic amino acids (glutamic acid, glycine, histidine, threonine) are used to adjust pH in one direction, and alkaline amino acids (arginine, lysine) are used to adjust it the other way. Other pH and buffering agents include HCl, NaOH, PO₄, and acetic acid.

Antioxidants help protect against oxidation by scavenging oxygen for themselves — just as they do in the body. Ascorbic acid is used, but citric acid is preferred, and it can be used as a pH adjuster as well.

A nonionic surfactant (usually a polysorbate) is often added as a finishing touch to inhibit protein aggregation. It is often needed during earlier processing steps anyway, so it may be easier to keep it in the solution than to filter it out (unlike processing contaminants such as peroxides, which must be removed). Surfactants and bulking agents such as mannitol and certain biodegradable polymers are often needed for low-concentration formulations. Clinicians and patients may not feel confident about a vial with hardly anything in it, and tiny amounts of drug substance are difficult to handle.

Lyophilization

A lyophilized biopharmaceutical protein requires protection during both the acute stages (the freezing and drying) and during storage, when excipients must form an amorphous solid (glass) with the protein. Freeze-drying may cause certain stresses on proteins that can result in unfolding, so specific conditions must be determined and stabilizing additives must be appropriate for use during both the freezing and drying stages. Part of a formulation plan is to stabilize the product during each stage: freezing, drying, and dry storage. Some proteins unfold during lyophilization but refold if immediately rehydrated. However, the goals are long-term storage (one or two years), acceptable cake morphology, and adequate dissolution properties.

The process. There are four steps to the freeze-drying process. Freezing forms an amorphous solid of the protein and excipients, with associated water in crystalline form. Annealing, an optional step, increases ice-crystal size and allows crystallization of bulking agents (such as glycine or mannitol), removing them from the amorphous portion and increasing Tg. Primary drying sublimes water ice at temperatures lower than Tg (to avoid collapse of the cake). The higher the Tg, the higher the mixture's temperature and sublimation rate can be. Increasing the protein to excipient ratio will increase Tg. Secondary drying removes water from the amorphous phase by an increase in sample temperature, which still may not exceed Tg. Luckily, as water leaves the amorphous phase, Tg increases. Sample temperatures play a larger role than the duration of secondary drying for determining final water content of the lyophilized cake. Low moisture increases Tg and thus increases the temperature at which the product can be stored.

Eutectic Point (Te) and Glass Transition Temperature (Tg)

A drug solution is first frozen at atmospheric pressure, and then water is removed by a reduction of pressure in the lyophilizer chamber, collecting the water as ice on a condenser. Samples are placed in glass vials and frozen, either before being put in the lyophilizer or on the lyophilizer shelves. The samples contain ice crystals, unfrozen water, amorphous solids (including the therapeutic protein), and crystalline additives. Pressure is reduced, and the ice crystals sublime. That constitutes the primary drying.

It is harder to remove the unfrozen water trapped in an amorphous solid. So after primary drying, a secondary drying stage removes that water by increasing the temperature. The final temperature of this secondary process is the key factor in determining residual moisture in the dried cake. The pressure is kept the same for secondary and primary drying to avoid protein collapse. The ideal result is a porous cake with little residual moisture. Porosity is important in later reconstituting the product.

Sometimes an annealing step (in which the product is kept at a set temperature) is added before the primary drying or near the end of secondary drying to crystallize excipients. This assures that crystallization and moisture release don't happen in an uncontrolled fashion later on during shipping and storage. Phase changes (crystallization of formulation sugars, for example) during shipping or storage can be disastrous. Moisture can even transfer from rubber vial-stoppers during storage unless savvy formulators plan for and prevent it.

The criteria for dried-protein stability include minimum lyophilization-induced unfolding, with proteins native in the dried solid; a powder with Tg higher than the desired storage temperature; low residual moisture (<1%) in the cake; and formulation conditions (such as pH) that inhibit chemical degradation reactions unaffected by glass transition (such as oxidation). The goal is to design the fastest and most robust (acceptable quality even with variations in operating parameters) processing cycle: one that consumes the least amount of energy, does not compromise product quality, and produces a mechanically strong, rapidly dissoluble cake. The cycle must be controlled for reproducibility and rapid correction of any problems that develop.

The challenges. Early lyophilized products were found to be inactive after rehydration, but no one knew where the damage had occurred. Infrared spectroscopy showed that proteins were unfolded in the dried solids. Stabilizers were added to prevent unfolding. Proteins need protection to maintain native structure during lyophilization. Sucrose can protect them during both freezing and drying. Until recently, assessing the necessary additives was impossible until after rehydration. However, Fourier-transform infrared (FTIR) spectroscopy allows the study of protein structures in any state, enabling scientists to discover that all unprotected proteins, except granulocyte colony-stimulating factor, unfold during lyophilization. They either regain their native conformation during rehydration (reversible unfolding) or they aggregate. To prevent aggregation, scientists prevent unfolding by using a stabilizer to retain the native structure in the dried product. Proteins that do refold are also protected by stabilizers from unfolding, which is a critical parameter for successful lyophilization.

Workers load a large-scale lyophilizer.

John Carpenter of the University of Colorado is a recognized expert in protein lyophilization. In their chapter in Biotechnology and BioPharmaceutical Manufacturing, Processing, and Preservation, Carpenter and Byeong Chang, founder of Integrity Biosolution, noted that formulators should watch for contaminants in their excipients. For example, mannitol may crystallize during short-term transient heating. "Think about the temperature of warehouses in tropical countries," they cautioned. "Amorphous mannitol can crystallize on exposure to >45°C."

A New Stability Method for Vaccines