Introduction
This document provides guidance on cycle validation for freeze drying, also called lyophilization. The information contained in this document is supplemental to Guidance 108 “Lyophilization”.
This guidance is applicable for sterile drug product; however, the general principles described would be equally applicable to freeze drying of non-sterile drug product or bulk API.
There are three elements to achieving successful validation of a freeze drying cycle:
- A well defined and understood formulation,
- A qualified freeze dryer and a freeze drying cycle that provides the link between a specific formulation and
- A specific freeze dryer.
It is recommended that cycle validation studies shall include a minimum of 3 consecutive, successful lyophilization runs on the worst case load configuration.
This guidance provides an overview of freeze drying and considerations for establishing the cycle validation strategy for the lyophilization cycle developed for a given product. Cycle validation may also be described as product Performance Qualification for a lyophilization cycle.
Recommendations & Rationale
Freeze drying is commonly used to improve the stability of thermally labile molecules. Some active ingredients or drug substances are only stable for a few weeks in liquid formulation but can be stored for years when freeze-dried with the appropriate excipients. Examples of such molecules include peptides, polysaccharides, proteins or even live viruses.
Freeze drying process
A typical freeze drying process consists of the following stages:
- Filling
- The solution is aseptically filled into loosely stoppered vials
- Loading into freeze dryer chamber
- Freezing
- Shelf temperature is reduced at a defined rate and kept for a defined time in order bring the formulation in the solid state
- Primary drying – the chamber pressure is reduced and the shelf temperature is ramped to one or more predefined settings, to allow removal of water vapour by sublimation. At this stage all free water is removed.
- Secondary drying – at reduced pressure the shelves are heated to a specified temperature to remove bound water
- Stoppering – the chamber may be filled with an inert gas (if required for product stability) and then vials are fully stoppered by moving the shelves up or down to push the stoppers into the vials
- Unloading and capping
Freeze drying equipment
A freeze dryer is mainly composed of two chambers, the chamber where the product is freeze-dried and the condenser that stores removed water. Chamber capacity varies from less than 1000 for clinical trials up to 100,000 for commercial operations.
Typical functions include:
- Shelf temperature monitoring and control
- Chamber pressure monitoring and control
- Condenser refrigeration
- Vacuum leak test
- Cleaning-in-place
- Sterilization-in-place
Before commencing cycle validation, the lyophilizer or freeze dryer, should be commissioned using Good Engineering Practice and qualified to verify that the system has been installed correctly and operates to control and monitor the critical functions.
System qualification as well as other components of process validation, such as formulation, aseptic filtration, media fills, cleaning and sterilization validation, and any critical hold times are also outside the scope of this guidance.
Formulation Design and Cycle Development
The design of the formulation is based on the characteristics of the molecule to be freeze dried. Stabilizing agents are chosen and tested. The collapse temperature, or eutectic temperature, is identified and experiments are designed to determine a robust combination of the solution composition and the primary drying temperature. At this stage, the primary packaging (for example, vials, stoppers) are also defined.
Further experiments are carried out to determine effective and robust cycle parameters for the primary drying phase to ensure small variations in the process do not result in defects of the freeze-dried cake at the end of the process.
The appearance of the freeze-dried cake or plug, is a key indicator of whether the freeze dry process has been successful. One of the major concerns for the process and product is meltback. This is a form of cake collapse and is caused by the change from the solid to liquid state. That is, there is incomplete sublimation in the vial. Associated with this problem is a change in the physical form of the drug substance and/or a pocket of moisture. These may result in greater instability and increased product degradation. The residual moisture of the cake is a key measure of the reproducibility of the cycle.
Finally the secondary drying cycle is developed, to ensure that the time, reduced pressure and heat foreseen accomplish that part of the process.
Cycle validation
Due to the sensitivity of the process to variations, scale-up of the lyophilization cycle can be very difficult. Products that have multiple strengths, vial sizes and different batch sizes will often have different cycles for each combination.
It is recommended that each specific product and specific freeze dryer combination is considered separately. Bracketing of equipment can be considered where equivalence of the equipment has been demonstrated, for example, during equipment qualification. However, due to the sensitivity and criticality of the process, the bracketing strategy for multiple, identical freeze dryers should consider evaluation of the cycle in each dryer (for example, placebo runs, engineering or demonstration lots).
Transferring a product to a new freeze dryer should include an evaluation of whether the existing cycle parameters are appropriate for the new combination. A science and risk-based approach may be taken to determine whether any trial batches are required or whether there is sufficient confidence in the applicability of the original cycle parameters that cycle validation may be initiated without trials. Cycle validation requirements for the transfer of an existing commercial cycle, including the number of runs needed, should be based on science and risk. Considerations should include the magnitude of the change, the process history, the robustness of the cycle and the complexity of the formula/container.
Differences when transferring a product between freeze dryers that may be expected to have an impact on the process include:
- Equilibrium product temperature during primary drying
- Chamber dimensions, capacity
- Differences or limitations in pressure control, e.g. Pirani (thermal-conductivity gauge) vs. MKS gauge (capacitance manometer)
- Limitations in temperature control (setpoints, achievable ramp rates)
- Shelf configuration and/or tray type (materials, dimensions)
- Condenser efficiency/capacity
- Batch size
- Load configuration
Lyophilization cycles need to be validated. Typically, this will consist of at least 3 runs using the worst case load. Key acceptance criteria include:
- Moisture content,
- Cake appearance, and
- Reconstitution properties.
Validation of the cycle should also include the establishment of validated loading patterns. It is recommended that the load configurations are described and diagrammed and include the following:
- Description of container design, shape, size, and fill volume for drug products;
- Description of the loading configuration within the chamber; and
- Identification of the number and placement of temperature probes, if used, in drug product containers and throughout the load for APIs.
Identification of “worst case” locations in the load Temperature mapping of the empty chamber (across and between shelves) is performed during qualification for the freeze dryer.
During cycle validation, temperature probes should ideally be placed at the “worst case” locations (coolest and warmest) in the load based on the empty chamber data. The location of the probes should be documented in the protocol, ideally using a combination of description and diagram. At minimum, the key positions of product in the load (corners, middle, identified “worst case” locations) in the load should also be documented to allow correlation of any variations in the cake appearance or other key product acceptance criteria with the temperature distribution in the chamber.
An alternative practice to identify “worst case” sampling locations would be to use a “sublimation mapping” technique to identify warm and cold spots. Instead of using temperature probes, a partial lyophilization cycle is run using a placebo formulation. The weight loss in vials at specified locations is measured to determine the areas in the dryer where sublimation is occurring fastest and slowest (usually edges and centers of shelves, respectively).
Partial Loads
Batch sizes for freeze-dried products are generally defined by the capacity of the freeze-dryer. Cycle times are typically long and therefore the processing of partial loads is inefficient and unusual.
It is recommended that a risk-based evaluation of the impact of running a partial load is included in the protocol. In the majority of cases a full load will be the worst case – that is, the water load is heaviest.
The evaluation should include whether the cycle parameters used for the full-load can be applied to the partial load without significant impact on the product.
However, in the event that partial loads are foreseen and the load is evaluated as “worst case”, the load must be defined and validated.
Assessment of Normal Operating Ranges
It is recommended that the cycle validation runs be performed at the target control values. However, an approach that can be considered is to perform the cycle validation runs across the allowable range for temperature and pressure, as defined in the regulatory filing. The cycle validation would then include one run each of low temperature/low pressure, high temperature/high pressure, and the target temperature and pressure for the primary drying phase. If this approach is taken it is recommended that the secondary drying phase is performed at the target values for temperature and pressure, as it is the
primary drying process that is most sensitive to variation in cycle parameters.
If the allowable ranges for pressure and/or temperature are comparable to the known performance capability of the freeze dryer, only the target values should be used during cycle validation. It is recommended that assessment of more than one lot of API is included in the cycle validation, where practical.