Related Articles
Optimizing Drug Development Strategies

Product & Process Considerations In the Development of Lyophilized



A review of how to identify the critical product and process variables that need to be optimized to formulate a successful lyophilized biopharmaceutical drug product.



Gerald R. Magneson, Karen K. Jette and Thomas R. Kovalcik
Cardinal Health



The development of a parenteral drug product (DP) typically involves stabilizing an active pharmaceutical ingredient (API) that either is unable to meet the desired DP shelflife as a liquid formulation, or is not readily marketed and distributed as a frozen liquid. The API biomolecules that we have encountered range from high molecular weight proteins and peptides to DNA and siRNA. In addition, we have extensive experience in lyophilized synthetic molecules, such as antibiotics and oncology DPs, as well as nuclear imaging agents. While we have employed different drug delivery technologies in developing DPs, this article will focus on the critical product and process variables involved in stabilizing parenteral drug products by lyophilization.

Lyophilization is a process where an aqueous solution is frozen and its water content is largely removed by sublimation of water vapor under reduced pressure and temperature. This technology was developed for large scale commercial manufacturing in the 1930s and demonstrated its therapeutic utility during the production of lyophilized human plasma to treat wounded servicemen during the Second World War.1 The equipment and processing methodology has advanced and matured in its pharmaceutical applications. Although lyophilization technology has been employed to produce stable protein pharmaceuticals for decades, the potential deleterious effect of dehydrating proteins is well documented as a significant problem.2 Since a number of labile biomolecules fail to recover their functional activity upon rehydration, it is imperative to maintain the native state of the biomolecule during the lyophilization process.3

Additionally, there is an increased risk of aggregate formation caused by non-native protein interactions within the lyophilized cake matrix. Hence, the key to successfully developing lyophilized biopharmaceuticals is, first, to identify those variables that may harm the biomolecule's structure. Then, in order to circumvent biomolecular damage, one may formulate with stabilizing excipients or optimize process parameters (e.g., by increasing the freezing rate to decrease mannitol crystallization during the lyophilization cycle).

While alternative technologies such as spray-drying,4 vacuum drying5 and supercritical fluid technology6 have been developed, in part, in an attempt to decrease the extended processing times and expense involved in large-scale manufacturing, lyophilization is a popular choice as a proven technology, because of the difficulties inherent in verifying the safety and equivalence of an alternative process for manufacturing a protein pharmaceutical. Some of the beneficial properties of lyophilized products include: In this article, we will review how to identify the critical product and process variables that need to be optimized to formulate a successful lyophilized biopharmaceutical drug product.

Customer-Specific Development Questions


As soon as a customer's confidentiality needs are met, several questions need to be answered to ensure that product development efforts will satisfy their needs and specifications. Usually, the first question is, what are the structural characteristics of a customer's API? If the API is a synthetic molecule, the customer will likely possess its structure via conventional instrumentation; e.g., NMR, IR and mass spectroscopy. However, for biomolecules such as high molecular weight proteins, one must utilize a multitude of analytical methods that obtain overlapping and complementary information to gain insight about the protein's structural characteristics. When very little structural or stability information on a biomolecule is known, a good place to start is the amino acid sequence, which can be used to explore the fundamental properties of the protein or peptide, such as its molecular weight, theoretical isoelectric point and extinction coefficient. Furthermore, a survey of the scientific literature may provide structural and stability information on this API, or a similar biomolecule purified from different tissues or organisms.

Certain customer information is critical to develop a successful lyophilized dosage form. The clinical indication for this potential drug product is useful in order to know where the DP will be administered (e.g., self-administered at home, in the physician's office or in the hospital) as well as the route of administration (e.g., IV, SC, IM, topical or inhalant). The patient population factors, such as age, disease state and sensitivity to excipients, may significantly influence formulation strategy. Dose requirements for the DP (e.g., single injection, multidose, dosage frequency, fixed versus variable dosing) also influence formulation development. Finally, customers may have specifications for the container/closure system (e.g., vial/stopper) to best accommodate their dosage form in the marketplace *(see Table 1).

Table 1


Performance/Formulation Studies


Once the customer's needs are clear, the next step is to ascertain basic physiochemical properties of a biomolecule. Such information is obtained under aqueous and solid state conditions by conducting screening or pre-formulation studies. Variables in biomolecule formulation that are commonly investigated include: buffers/pH, stabilizers, solubilizers and tonicity modifiers. In the case of a multi-dose (multiple injections) drug product used over several days, one may include preservatives in the variables list.7

In order to test which of the candidate components is best suited for drug product formulation, these variables must be evaluated for their effect on the native structure and function of the API by employing stability-indicating methods. These stability-indicating methods may be acquired from the customer or developed to support formulation studies. These test methods should measure changes in secondary, tertiary and quaternary protein, or peptide structure (e.g., far-UV Circular Dichroism (CD) and Fourier Transform-Infrared Radiation (FT-IR) spectroscopy for secondary structural analyses; near-UV CD and intrinsic protein fluorescence for tertiary structural analyses and static, or dynamic light scattering, with or without Size Exclusion (SEC), or Reversed Phase (RP-HPLC) chromatography for quaternary structural analyses). In addition, one should evaluate the effect of formulation variables on the therapeutic function of the biomolecule by performing a bioassay (e.g., measuring enzymatic activity, determining immunoreactivity of a monoclonal antibody against its epitope, etc.)

Pre-formulation studies help to identify those variables that appear most critical for stabilizing the biomolecule at the desired storage temperature as well as at least one elevated temperature; e.g., 40°C. These typically include the following screening studies: The nature of these screening studies entails using pre-existing stability information, such as observations Figure 1: Design-Expert Plot made during protein or peptide process development, or literature results, to make educated formulation assumptions that are tested by multiple methods in an attempt to identify those components most likely to enhance the stability of the biomolecule. Furthermore, these formulation components and excipients usually are GRAS (Generally Recognized as Safe) chemicals that are sourced for their high purity and low endotoxin levels.

One must be careful not to place too much emphasis on the effect of components on the aqueous solubility of the biomolecule. For example, a protein may be more stable to thermal stress in a high ionic strength formulation in solution, but less stable following multiple freeze-thaw cycles relative to a low ionic strength formulation. Freeze-thaw instability is likely caused by the concentration effect on electrolytes and protein in the interstitial matrix between ice crystals and their interactions in the matrix (i.e., a 25-fold concentration of salt occurs in a 4% (w/v) solids formulation). To better assess these components in a model drug product format, it is preferable to test biomolecule formulations that are freeze-dried using a very mild lyophilization cycle in a research tray dryer.

Figure 1
Figure 1.

Figure 2
Figure 2.

Figure 3
Figure 3


In evaluating biomolecule formulations that are stressed at elevated temperatures, it is important to note that some proteins may retain nearly all of their functional activity, while experiencing extensive structural perturbation. Since there are obvious health risks inherent with the presence of protein aggregates in a parenteral biopharmaceutical, one must be aware of the appearance of non-native structure in the biomolecule as observed by stability-indicating assays. Hence, one must identify those beneficial product variables and their effective concentration range to stabilize biomolecules in lyophilized screening formulations against suitable response factors (e.g., bioassay activity, alterations in secondary structures and percentage soluble aggregates by SEC-HPLC).

After identifying the key product variables for stabilizing the API, the optimization of a lyophilization formulation for the biomolecule may be pursued. If a strong lead formulation is identified from the screening studies, a customer may elect to pursue the scale-up of a screening formulation in order to facilitate the transfer of drug product to manufacturing. Although a biomolecule formulation may appear satisfactory without further optimization, it is important to note that one should verify the reproducibility of preliminary results with one or more additional formulations.

In the event that a lead formulation is not clearly identified from the screening studies, formulation components may be further investigated against response variables by performing experimental design studies, such as using response surface models to better determine indirect or interaction effects on biomolecular stability. Experimental design software (e.g., Design-Expert or JMP) enables one to better visualize the three-dimensional response surface area for variable optimization *(see Fig. 1). As a general rule, it is useful to build additional pilots to test predicted areas of optimal stability to confirm that the model is accurately identifying global maxima. Sufficient formulation vials should be freeze-dried to conduct accelerated stability studies at the desired storage temperature and at least one accelerated storage temperature (e.g., 2-8 and 40∞C for a minimum of three time points for up to six months, as indicated in the ICH guidelines for testing the stability of a new biotechnological/biological drug products).8

Lyophilization Process Design


There are three main process variables that occur during lyophilization: the initial freezing phase, the primary drying phase and the secondary (or terminal) drying phase. These process variables may be optimized to enhance drug product stability as well as to decrease manufacturing costs. These process studies may be performed in parallel with the optimization of screening formulations, but usually will follow the selection of a lead formulation for the biomolecular DP.

Freezing Phase


The primary function of the freezing phase is to ensure that all of the DP vials are completely frozen prior to proceeding to the primary dry phase. Additionally, it is preferable that these vials freeze in a uniform manner. While there are different ways that this can be accomplished, one option is to chill the vials after they are loaded onto the lyophilizer shelves and held for 30-60 minutes prior to initiation of the freezing cycle. It is generally not practical to equilibrate the shelves to a freezing temperature, because of frost accumulation during the filling and loading of the vials.(While the advent of automated freezing/loading equipment may be utilized during manufacturing, it typically is not employed during initial cycle development.)

Figure 4
Figure 4


DP vials must be placed directly on the lyophilizer shelf to ensure uniform contact between the vial and shelf. To achieve this, vials may be manually placed on to the lyophilizer shelves or loaded on bottomless trays and transferred to the shelf surface with a retaining bar. Thus, DP formulations should experience more consistent ice nucleation and freezing as the shelf temperature is ramped to a freezing temperature.

The optimal freezing rate may be determined using a freeze-dry microscope by observing the formation of the ice structure of a formulation. Since the thin film of formulated drug product may not accurately reflect the bulk solution freezing in the vials, it is important to test different freezing ramp rates for their effect. One can determine the degrees of supercooling measured by product thermocouple probes in filled product vials on a lyophilizer shelf (i.e., difference in product temperature before and after water freezes). It is useful to have about 5-15°C degrees of supercooling, indicating that ice has formed with a uniform distribution of its constituents (see Fig. 2a and 2b for examples of uniform and non-uniform lyophilization cakes). One can also confirm the collapse temperature, Tc, of a formulation by freeze-dry microscope. The collapse temperature, Tc, is the temperature at which a formulation exhibits viscous flow of freeze-concentrated liquid and shows a loss of microstructure.9 Structural defects in the lyophilization cake that are caused by exceeding Tc usually give product instability and resultant difficulty in properly drying the cake (see Fig. 2c, d and e for collapse examples). In the event that a formulation needs more structure, one may add a bulking agent; e.g., mannitol. To increase the degree of mannitol crystallinity, one can anneal the frozen formulation by warming it to a temperature beyond its crystallization temperature, about -20∞C, for several minutes before refreezing it prior to starting a primary dry phase.9

Primary Drying Phase


Once the formulation is brought to the desired frozen state, primary drying via sublimation can proceed. The primary dry phase involves the removal of bulk water at a product temperature below the ice transition temperature under a vacuum (pressures typically between 50-150 mTorr). This phase is the most critical one for stabilizing biomolecules. The amorphous glass structure formed by lyoprotectants, such as trehalose and sucrose, must remain rigid throughout this drying phase of the cycle to reduce the likelihood of protein denaturation due to dehydration.

Optimizing the primary freeze-dry phase for a biomolecule formulation is greatly facilitated by a thermal analysis of the DP formulation. The goal of this testing is to identify the glass transition temperature (Tg') for the formulation. The Tg' is the temperature at which there is a reversible change of state between a viscous liquid and a rigid, amorphous glassy state.9 One can measure the Tg' of candidate formulations using a differential scanning calorimeter (DSC), in particular with modulated DSC (see Fig. 3). Generally, the collapse temperature is observed to be about 2-5°C greater than the Tg'. Hence, the shelf temperature is set such that the target product temperature is maintained near or below the Tg' of the formulation throughout the removal of bulk water during the primary dry phase.9

As the bulk water is progressively removed from the formulated vials, the product temperature will approach and reach the shelf temperature since it is no longer cooled by water sublimation. To optimize the duration of the primary dry phase, the removal of bulk water vapor can be tracked using a moisture detector, or by monitoring the decrease in pressure difference between a capacitance manometer and a thermocouple (Pirani) pressure gauge or by a pressure drop measurement.9 The optimization of the primary dry cycle involves the removal of bulk water as quickly as possible without causing cake collapse and subsequent product instability (see Figure 2d).

Secondary (Terminal) Dry Phase


Secondary dry phase is the final segment of the lyophilization cycle where residual moisture is removed from the formulation interstitial matrix by desorption with elevated temperature and/or reduced pressure. The final moisture of a lyophilized formulation, which can be measured by Karl Fisher or other methods, is important to determine because if the cake contains too much residual moisture, the stability of the API in the amorphous glass can be compromised.9 Hence, it is imperative that one achieves a moisture level less than, or equal to 1%.9

To accomplish a low residual moisture, the shelf temperature is typically elevated to accelerate desorption of water molecules. The duration of the secondary dry phase is usually short (e.g., 6-10 hours). When microstructure collapse occurs, the residual moisture is generally significantly greater than desired. One alternative is to purge the sample chamber of the lyophilizer with alternating cycles of nitrogen to facilitate displacement of bound water.10 However, the best solution is to properly formulate the drug product and run an optimal lyophilization cycle.

In general, it is advisable to backfill protein formulations with nitrogen gas to maintain a non-reactive gaseous headspace. To facilitate reconstitution, one may backfill to a slightly sub atmospheric pressure. However, if a formulation does not require an inert gas, the vial can be stoppered either under vacuum or with air.

Conclusions


There are a number of product and process variables that must be considered to produce a stable, viable biomolecular drug product. While it is difficult to envision all of the possible contingencies that one may encounter in optimizing DP formulations, we utilize the decision tree shown in *Figure 4 to develop high-quality lyophilized biopharmaceuticals.

References

  1. Jennings, T.J., “Lyophilization Seminar Notes,” pg. 3, Phase Instruments, Inc. Copyright È 1993.
  2. Carpenter, J.F., Prestrelski, S.J. and Arakawa, T., “Separation of Freezing-and Drying-Induced Denaturation of Lyophilized Proteins Using Stress-specific Stabilization. I. Enzyme Activity and Calorimetric Studies,” Arch. Biochem. Biophys. 303(2), 456-464 (1993).
  3. Prestrelski, S.J., Arakawa, T. and Carpenter, J.F., “Separation of Freezing-and Drying-Induced Denaturation of Lyophilized Proteins Using Stress-specific Stabilization. II. Structural Studies Using Infrared Spectroscopy,” Arch. Biochem. Biophys. 303(2), 465-473 (1993).
  4. Nguyen, X.C., Hersberger, J.D. and Burke, P.A., “Protein Powders for Encapsulation: A Comparison of Spray-freeze Drying and Spray Drying of Darbepoetin Alfa,” Phar. Res. 21(3), 507-514 (2004).
  5. Sharma, V.K. and Kalonia, D.S., “Effect of Vacuum Drying on Protein-mannitol Interactions: The Physical State of Mannitol and Protein Structure in the Dried State,” AAPS PharmSciTech, 5(1), 1-12 (2004).
  6. Ribeiro Dos Santos, I., Richard, J., Pech, B., Thies, C. and Benoit, J.P., “Microencapsulation of Protein Particles within Lipids Using a Novel Supercritical Fluid Process,” Int. J. Pharm. 242(1-2), 69-78 (2002).
  7. Pharmaceutical Biotechnology, Vol. 14, “Development and Manufacture of Protein Pharmaceuticals,” pp. 1-84, (Edited by Nail, S.L. and Akers, M.J.: Kluwer Academic/Plenum Publishers, 2002).
  8. International Conference on Harmonization: Final guideline on stability testing of biotechnological/biological products. Federal Register, July 10, Volume 61, Number 133, pp. 36466-36469 (1996).
  9. Tang, X. and Pikal, M.J., “Design of Freeze-Drying Processes for Pharmaceuticals: Practical Advice,” Pharm. Res. 21(2), 191-200 (2004).
  10. Jennings, T.J., “Lyophilization Seminar Notes,” pg. 145-155, Phase Instruments, Inc. Copyright È 1993.