Cleaning freeze dryers

March 12, 2010
Pharmaceutical Technology Europe

Freeze dryer internal surfaces become contaminated, post cycle, by the ablation of API and excipient from the freeze?drying cake.

Freeze dryer internal surfaces become contaminated, post cycle, by the ablation of API and excipient from the freeze drying cake. Although the freeze dryer is, logically, a non‑contact part of the process chain (gas flow is predominantly away from the product and only reverses for a short time during aeration), regulatory authorities tend to regard these internal surfaces as direct contact and, therefore, GMP mandates cleaning. There is no GMP requirement for clean in place (CIP), but CIP is increasingly the method of choice for freeze dryer cleaning because of validation requirements, access within larger freeze dryers, the handling of cleaning materials in aseptic areas and the vagaries of operator efficiency and their protection.

TACT — the cleaning mnemonic
CIP by water spray fulfils the requirements of the cleaning mnemonic TACT:

  • Time is a programmable variable
  • Action is generated by the scouring effect of the spray and may be enhanced by increased spray pressure or by shelf actuation
  • Chemical Concentration is a choice, although most freeze dryer CIP cycles merely use purified water and water for injection (WFI) to avoid issues of cleaning agent clearance
  • Temperature is generated by the use of hot WFI. Recirculation is not recommended because it redistributes any removed soil and may block spray nozzles. To minimise endotoxins, the final rinse should always be by WFI.

Water usage is short term and high (up to 5 m3/CIP cycle), and needs addressing during facility design. It should be noted that CIP is unlikely to put enough heat into the system to achieve post‑CIP dryness and CIP is invariably followed by steam in place (SIP).

CIP is not a universal panacea for cleaning issues. As most freeze dried products are water soluble, CIP is likely to clear product from the freeze dryer internals, but unlikely to remove stopper silicone from the shelf undersides. In addition, the clearance of any external contamination is unknown. CIP will also not clear fallen stoppers, vials or glass shards — a sieve should be incorporated into the freeze dryer drain to facilitate removal of these items. Because of the above reasons, CIP should always be preceded by a visual inspection and regular manual scrubbing of the shelf undersides, with a suitable detergent or solvent, should be included within the area SOPs.

CIP validation
All cleaning activities require validation and CIP is no exception. Initial spray coverage should be determined by the clearance of riboflavin (15 mg/L) and post‑cycle residuals visualised by ultraviolet light. Visualisation should be performed in the wet state because dry riboflavin does not fluoresce. Once spray coverage is established, the cleaning cycle may be developed — mainly by adjusting the spray times rather than the spray flow/pressure, which would require requalification of coverage. The assessment of cleaning efficacy should always be conducted by swabbing, with credence being taken of the analytical method detection limits. A drain conductivity probe assumes the soil is soluble and the only practical use for such a probe is to determine cleaning agent clearance.

Validation acceptance criteria for cleaning efficacy is a subject of debate. Visually clean, water break and the 10 ppm rule may be unlikely to survive regulatory scrutiny, but the 0.1% of the minimum therapeutic dose (MTD) criterion is common. The swab area contamination is extrapolated to the total freeze dryer internal surface area and then assumed to be distributed equally amongst all the product containers in the next batch. Alternatively (and less stringently), a modified MTD within a hypothetical box between the shelves may be used.

In conclusion, CIP reproducibly provides a cleaning process that can be validated, is closed, is automatic, allows the use of appropriate cleaning fluids (if necessary), reduces labour cost and variability, and prevents contamination spreading to the aseptic area or to operators. However, CIP should not be seen as the only means of cleaning a freeze dryer because the CIP process does not have a feedback mechanism that verifies the clean effectiveness and cannot remove solid objects. For these reasons, CIP should always be prefaced with a visual inspection to remove solid objects from the chamber and to verify the effectiveness of the clean after the CIP cycled is completed. Swab verification of the clean is also a useful periodic adjunct.

Based on a contribution by Kevin Murgatroyd, Senior Pharmaceutical Consultant for SciTech Engineering Ltd.