Low-Risk Lyophilization

March 16, 2009
Developing a successful lyophilization program requires equal parts design, technology, and expertise.

Freeze-drying (lyophilization) of pharmaceuticals has been used successfully for many years to provide stable, sterile injectable dosage forms.  However, as with any technology, lyophilization continues to evolve. Advances in analytical and process sciences allow pharmaceutical manufacturers to further exploit both chemical and physical phenomena that occur during lyophilization. At the same time, energy conservation, safety and complexity of pharmaceutical formulation chemistry continue to push the science of lyophilization well beyond the limits of basic cryogenic and sublimation equipment. This article presents suggested practices for developing a lyophilization program based upon Ben Venue Laboratories' years of experience and technical expertise.  Key points include cycle development for new products, container/closure systems, equipment issues, and risk mitigation. BVL monitors new technologies and strives to remain current with industry and regulatory expectations for optimized development practices based upon Quality by Design (QbD).

BVL continues to be a leader in contract lyophilization of investigational and commercial parenteral pharmaceuticals. We have new and  long term relationships with many domestic and international clients, a longstanding relationship with the National Cancer Institute and a successful generic division,  Bedford Laboratories.  Many of the products we manufacture (which include cytotoxins, genotoxins, biologics and diagnostics) require unusually complex production processes.

The exterior of Ben Venue's
385-square-foot lyo
chamber, its largest.

 BVL operates twenty-nine production lyophilizers and five development lyophilizers (one of which is a production-scale unit). BVL is currently validating a new cytotoxic/genotoxic facility which houses nine of the twenty-nine production lyophilizers.

These will be loaded through the use of laminar flow carts and unloaded into special isolators used to transfer vials to capping equipment. Production chambers range in shelf capacity from 96 square feet to 380 square feet for a total of 7692 square feet. This represents approximately 2.5 million 3cc or approximately 1.25 million 10cc standard tubing vials. BVL presently operates eight vial filling lines, six of which dispense liquid and lyophilized products and two that dispense liquid products only.

The new cytotoxic/genotoxic complex houses three additional filling lines, all of which will dispense both liquid and lyophilized products. Vial size processing capability ranges from 2cc/13mm to 200cc/28mm.

Given our substantial capacity, lyophilization equipment scheduling is logistically complex. For example, as a result of varying cycle lengths from product to product, rescheduling one lot may leave an opening that is either too short or too long for pending products to be scheduled into that slot.  Well developed and defined manufacturing procedures and lyophilization cycles are critical to maintaining a stable schedule.

The interior of the
chamber, with one
shelf loaded
with 2cc vials.

Infrastructure planning for new equipment and upgrades now, more than ever, necessarily involves early interaction between internal departments, outside vendors, regulatory agencies, utility companies and local communities. Development project planning is equally complex.

Presuming that a product’s therapeutic potential has been clearly established, the steps required for further assessment, dosage form development,  and lyophilization cycle development must be carefully mapped and coordinated.

These include scale-up studies to identify resources necessary not only to ensure technical success but also to make certain that safety, regulatory, environmental, cost, and risk guidelines are met.

Development Activities Leading to a Lyophilization Cycle

Beginning with an established target dose, additional information needs to be collected to successfully develop a formulation and suitable lyophilization cycle. This includes:

  • optimum pH
  • solubility - are special excipients indicated? (e.g. ethanol, t-butanol, DMSO, cyclodextrins)
  • stability characteristics - room temperature, refrigerated, freezing (is a cryoprotectant indicated?), light sensitivity
  • headspace lability - does oxygen need to be excluded?
  • administration route - intramuscular, intrathecal, intravenous, (bolus, infusion pump?), instillation?
  • constitution volume - what concentration will be administered (if administered undiluted or diluted into intravenous fluid?
  • constitution diluent/vehicle - is a buffer or special solvent indicated?

A thoroughly-considered development plan is established based on dosage form expectations, pre-clinical data and literature/search information, arriving at one or more formulae for evaluation.  Solutions from small lab batches are examined before and after filtration for their pH, color, clarity, assay, impurity levels and stability.  If no issues are identified, analysis of the liquid using lyophilization development tools can proceed. These include differential scanning calorimetry, moisture measurement methods and freezing-stage microscopy, all of which provide information regarding thermal events and conditions that will ultimately impact lyophilization.  At this point, pilot-scale lyophilization batches are prepared.

In most cases, multiple lyophilization experiments are required to achieve a successful cycle, particularly where unusual or non-optimal product formulae are involved. Lyophilization difficulties may, unfortunately, dictate a return to the formulation “drawing board” with, at minimum, revisions to excipients, concentrations, volumes or other variables.  We recommend that unusual or non-optimal conditions be carefully taken into consideration at the outset of development.  By the time a formula gets to the pilot lyophilizer, there should be a limited number of unknowns.

Liquid formulae that were used for pre-clinical trials may not be suitable for lyophilized products. For example, a buffer system which was used to suspend a protein during early pre-clinical studies may be, at minimum, a significant challenge in the lyophilizer and, if carried through to commercialization, could seriously impact aesthetics, yield, capacity, content, and uniformity.

If the active ingredient, excipient, dosage form, delivery system or other variable ultimately restrict formulation options and lead to lyophilization challenges, the developer must pay particular attention to the boundaries within which this product is formulated. Consider the following:

  • What might happen to an active drug substance that is formulated near its solubility limit at room temperature and within a narrow pH window? Does the API or an excipient precipitate during the initial freezing step of the lyophilization cycle? In this instance, what is the tolerable concentration range of the formulated solution that can be expected to prevent one formulated lot from precipitating while another does not? Is the ionic strength of the formulated liquid so low that the solubility of carbon dioxide in the liquid impacts the pH of the liquid during filtration, filling, loading, or freezing?
  • Does the concentration of the formulated solution require a relatively large fill volume? If so, has an appropriate vial been selected which 1) will provide adequate headspace for the constituting diluent and 2) will not fracture during lyophilization? Larger diameter vials (with the same formulated liquid) translate directly to lower cake height (potentially, but not always, reducing cycle time). They also translate to fewer vials in the lyo chamber. With this example, reduction in production capacity may be a tolerable alternative if cycle time or lyophilized cake characteristics are detrimentally affected by the cake height. BVL generally limits fill volumes to essentially not more than 50% of a vial’s capacity. Can the product be filled at a lower volume (higher concentration) that the constitution volume (lower concentration)?
  • Are the vials and stoppers appropriately matched to the product and to each other? Are there adequate restrictions on their dimensions and processing parameters to prevent stopper misplacement/instability during filling or downward creep during loading and lyophilization? Does the stopper seating pressure or sealing/capping pressure require unique specifications?

Although lyophilization is an option for compounds with limited stability as a liquid, there are a series of measurements and assessments beyond assay and moisture tests that need to be performed prior to and after lyophilization. Although lyophilization may produce a uniform cake that constitutes easily, a dry cake in an amorphous matrix, for example, may be less stable than its crystalline active drug substance. Consequently, a suitable, stability-indicating assay needs to be in-place for proper evaluation. Other formulation and lyophilization steps, including thermal treatment prior to primary drying,  may need to be taken to force a specific cake morphology or avoid other problematic issues such as vial breakage.

Inside the external condenser of Ben Venue’s 385-square-foot
lyo chamber.

If the product is lyophilized from an aqueous liquid, a Karl Fischer moisture test method will likely be qualified for the dried product.  A non-destructive, spectrophotometric moisture test (measuring through the glass) may prove invaluable as a development tool since those same vials may then be analyzed to determine their impurity profile.  While not typically part of a normal QC release regimen, proper equipment and technique may provide a qualifiable, non-destructive test for a large number of vials as a potential part of a development or validation program.

With special isolation equipment, barrier systems or autoloading lyophilizers the placement of thermocouples in individual vials during lyophilization can be impractical (or of little value in the case of vacuum dried liquids or very small fill volumes). Consequently, if lyo cycles are developed with thermocouples, development lots need to be carried out at statistically significant scale and with appropriate frequency of measurements. Coupled with statistically significant final product tests, segment-driven cycles that do not rely on product thermocouples to trigger critical cycle steps can be implemented. Although relatively few product thermocouples are typically used during production cycles, they do provide a simple snapshot of the product profile that can be used to troubleshoot unexpected cycle events.  Monitoring via carefully controlled chamber isolation pressure tests can signal cycle status or completeness, although production equipment behavior under load must be well understood in order to rely completely on this technique. Non-invasive methods of chamber moisture measurement, using  mass spectrometry and Near Infrared techniques, have been developed and may be of use in development and production equipment.

An important consideration for the development stage of a lyophilized product is to consider selection of a stopper that has a coating or film which prevents it from sticking to the shelf above during stoppering. Historically, a silicone emulsion applied during stopper preparation has been used to facilitate machinability during filling as well as reducing stopper adherence to lyophilizer shelves.  For some products, this may result in a perceptible haze or opalescence in the constituted liquid.  Stoppers are available with a fluororesin film or polymerized silicone coating on the top (and other surfaces depending on the stopper). These can reduce or eliminate sticking, prevent silicone-related opalescence in the constituted liquid and make chamber unloading a smoother operation.

Design, Training, and Technical Support

There are several other keys to reduction of risk during formulation and lyophilization development. Above all, a clear set of design criteria are needed. Every opportunity should be taken to involve Quality and Manufacturing personnel in establishing product-specific expectations and design criteria to ensure success and reduce risk from the start of a project. There are times when the need to produce clinical products outweighs the objective of developing an optimized cycle.  In this situation, non-optimal conditions that are introduced into the process may initially be tolerable but will clearly need to be addressed during the continued development of the product. If, for example, active ingredients or excipients are used that do not fully reflect the characteristics of later clinical or commercial materials, the possibility exists that formulation, lyophilization, constituted liquid characteristics or shelf-life stability will be altered. Non-optimal cycles may not only impact product cost, as a result of extended cycle time or unnecessary risk to batches, but may also create product constitution issues or result in unexpected problems with short-term liquid or long-term lyophilized product stability. BVL weighs these issues carefully and provides guidance accordingly during initial product discussions, development, and before transferring technology into production areas.

Documented training for production, quality and support personnel is a regulatory requirement. Training provides a significant opportunity for risk mitigation if properly designed curricula are established and presented. Development personnel, in particular, need to clearly understand the purpose, capabilities, and limitations of Operations facilities and equipment. In-house training and development seminars in lyophilization technology, thermal analysis, analytical method development, data interpretation, and safety are fully supported at BVL and provide an opportunity for staff training and continued education. In addition to internal lyophilization experts Paul Walsh (Vice President, Product & Process Development) and Thomas Kovalcik, (Executive Director, Product & Process Development), BVL has worked directly or through client collaborations with recognized academic and industry consultants including Thomas Jennings, Alan MacKenzie, Steven Nail, Michael Pikal, and Edward Trappler.

The Right Equipment

BVL’s developmental lyophilizer monitoring and control parameters have been specified to achieve temperature and pressure characteristics that mimic our production units. In addition to the scaleable interface and hardware on larger developmental units, BVL has also acquired LyoStar development lyophilizers from FTS Systems that incorporate SMART Technology developed by Drs. Nail and Pikal. SMART is a cycle optimization method that provides a safe, non-invasive batch method of thermal profiling that delivers accurate process data on resistance of the dried layer, ice thickness, heat flow, and mass transfer that can speed the cycle optimization process and reduce batch processing times.

BVL operates state-of-the-art production-scale lyophilizers. Our Hull (SP Industries) lyophilizers, with Multi-Flex and SmartCool refrigeration systems, use fluid condensers and are energy efficient, flexible, and incorporate substantial system redundancy. These units require fewer compressors while offering increased refrigeration capacity, overall efficiency, control precision, and system reliability. Reduction of energy costs during manufacturing results in lower product manufacturing costs, as well as providing environmental advantages.  A key factor in reduction of energy consumption is the use of variable frequency drives (VFD’s) on Hull refrigeration compressors. These units allow compressor speed to be varied based on required refrigeration capacity and load. When refrigeration demand is lower, compressor speed is reduced to save energy.

Other equipment considerations and alternatives for production lyophilization systems:

  • A liquid nitrogen system instead of mechanical refrigeration will reduce power consumption and allow for colder shelf and condenser temperatures. One important consideration is a readily available source of liquid nitrogen. Liquid nitrogen, however, poses a number of other materials and stress issues that may make this approach impractical. BVL has considered a liquid nitrogen system for special lyophilization challenges but has no plans to pursue this technology at the present time.
  • Screw compressors typically have higher initial cost but lower maintenance and operating costs. Screw compressors can achieve refrigeration temperatures as low or lower than reciprocating compressors.
  • Condenser cooling can be achieved via direct expansion or by circulating fluid through a heat exchanger. Fluid systems can be controlled at a given set point. Direct expansion can be less costly but has limited control capabilities.
  • Equipment redundancy. This is especially important for the control and data collection systems, vacuum pumps, generators and heat exchanger circulation pumps.

As with all other aspects of drug development, manufacture and control, success depends on providing the right resources to the best people.

About the Authors
Brian Bucur is BVL's Associate Director, Lyophilization & Sterilization and has been with BVL for 36 years.

Timothy Smith is BVL's Director, Product & Process Development and has been with BVL for 35 years. He is also the Principal Investigator/Project Director for our Contract with the National Cancer Institute.

About the Author

Brian Bucur | Associate Director