Leak Rate Testing for Freeze Dryers

1 Leak Rate Testing for Freeze Dryers : A Scientific Approach Leak rate Testing for Freeze Dryers Assurance of the vacuum integrity of Freeze - Background Dryers used for the manufacture of sterile pharmaceutical products is essential for Class 100 (FDA) and Grade A (EU). GMP operations; however, there is currently specifications concerning active microbial no generally accepted scientific rationale for counts are similar - an action level of one establishing acceptance criteria for such colony forming unit (cfu) for each cubic Testing [1]. meter of air space for Class 100 [3] and an 3. average limit of <1 cfu/m , with appropriate As it stands, current acceptance criteria are alert and action limits for Grade A [4]. generally based on equipment capability . perhaps proposed by the manufacturer, or FDA states, Facility design should ensure from data collected during qualification of a that the area between a filling line and the new Freeze - dryer . The targeted specification lyophilizer provide for Class 100 [3].

2 Is often one cited by The Parenteral Society, however, it does not make reference to A frequently specified leak rate for new, requirements for the inside of the dryer . EU. clean, dry, and empty Freeze - Dryers would guidance is clearer: Partially stoppered -2. be 2 x 10 mbar-liter/sec (15 m (mTorr) Freeze drying vials should be maintained liters per sec). The leak rate should not under Grade A conditions at all times until change significantly during the life of the the stopper is fully inserted [4]. Freeze - dryer .. [2]. Process control is demonstrated with media A study was performed to use a more fills, but regulatory guidance is sparse, and scientific approach to establish leak rate sometimes seemingly contradictory. FDA. acceptance criteria, by answering the states, Media fill studies should closely question: What is the maximum volume of simulate aseptic manufacturing operations . air that can leak into the Freeze - dryer without For lyophilization operations, FDA.

3 Violating the Class 100 / Grade A microbial recommends that unsealed containers be specifications? A method was developed to exposed to partial evacuation of the measure the exact void volume of a dryer , chamber in a manner that simulates the that is, the total volume of air space present process. Vials should not be frozen, and in chambers, condensers, and connecting ensure that the medium remains in an ducts. A worst-case bioburden value for air aerobic [3]. Although these that could leak in was determined, and then recommendations simulate the air the maximum allowable leak rate was turbulence associated with Freeze -drying, calculated. The leak rate was then while still maintaining the viability of the converted to the maximum allowable growth medium, they create conditions that pressure rise over the duration of a leak test. do not simulate lyo operations, per the FDA. definition: Lyophilization - Drying in which the water vapor sublimes from the product after freezing [5].

4 2. EU media fill guidelines contain the following For this study, the Testing was performed recommendation for all media fills (although and results evaluated under the following nothing specific for lyophilized products), three assumptions: The process simulation test should imitate as closely as possible the routine aseptic 1. Although leaks can occur in manufacturing process and include all the numerous places in a Freeze -drying critical subsequent manufacturing steps system, to create a worst-case [4]. scenario the leak is considered to be from a source(s) large enough to For vacuum integrity (leak rate) Testing of admit microorganisms, and located Freeze Dryers , no official limits are dictated in a part of the system housed in an per FDA. A general discussion states, unclassified mechanical room (with Leakage into a lyophilizer may originate no set specifications and no routine from the atmosphere into the vessel monitoring performed). is necessary to monitor the leak rate 2.

5 Risk to the product from a system periodically to maintain the integrity of the leak is limited to the time from the system Should the leak rate exceed end of primary drying to stoppering specified limits, determine the actual leak of the vials ( , all of secondary site for purposes of repair. Thus, it would be drying). During primary drying beneficial to perform a leak test at some product risk is minimal, because the time after sterilization The time and vigorous rate of mass transfer of frequency will vary and will depend on the water vapor from the product out data developed during the cycle validation. through the open slot in the lyo- The pressure rise found acceptable at stopper during sublimation creates a validation should be used during positive pressure in the vial production. [6] . headspace. Matter that might enter the system from the outside will EU guidance makes one leak test reference likely travel to the area of lowest (pertaining to moist heat sterilizers), There pressure, the ice on the condenser, should be frequent leak tests on the and stay there.

6 Since it is difficult to chamber when a vacuum phase is part of determine when exactly positive the cycle. The only indication that headspace pressure ceases, specifications should be established is this duration of risk is defined as the guidance, The efficacy of any new entire time the product dwells in the procedure should be validated, and chamber from the end of primary ..verified at scheduled intervals based on drying to stoppering of the vials. performance history [4]. 3. The leak rate remains approximately constant over the time course of secondary drying. 3. Materials and Methods was added to the average worst-case room results, for a bioburden maximum with a The Dryers in this study were manufactured 95% confidence level. The calculated limit by IMA Edwards and have the same was doubled for an extra margin of safety, external condenser design. Each have and this was the value used as the identical footprints but contain shelf area of maximum potential bioburden load from the 2 2.

7 Either 98 ft (5 shelves) or 215 ft (11 mechanical rooms. shelves). The chamber is loaded and unloaded via a slot door in the main A canister made from a closed section of chamber door, which opens into a stainless steel tubing was attached to a Class100/Grade A area of the production spare port on the chamber of each Freeze suites. The condensers, at the rear of the dryer . The pressure inside the chamber was units, are housed in unclassified mechanical decreased and stabilized, then a valve rooms. Three Dryers were installed in 2001, inside the canister was opened, and the air and three in 2006. in the canister entered the chamber. The change in internal chamber pressure was Each system was initially verified, after a measured and used to calculate the void typical production sequence of cleaning, volume of the Freeze - dryer (that is, the steaming, and drying, to achieve leak rate volume not occupied by shelves, condenser test results of not more than mbar (20 coils, and other hardware).)

8 MT) pressure rise over 30 minutes. Post- cycle leak tests, performed to ensure that aseptic integrity was not lost during the cycle, continue to require the same results for a passing test. Measurements and calculations based on the Ideal Gas Law, pV = n RT, were performed to arrive at the maximum allowable pressure rise (leak rate) value that would maintain Class 100 / Grade A. conditions for each dryer . The worst-case potential bioburden for leaked air was determined by using an air IDEAL unit, operating according to the ISO/DIS 14698-1 recommended impaction principle, to measure the microbial load of The following page provides an overview of air in the two mechanical rooms behind the all measurements and calculations Freeze Dryers . Twice the standard deviation performed. 4. Measure Measure pressureand temperature By measuring By measuring thethe change in Calculate Calculate how much how Calculate Calculate how fastfast how the air the pressure of air, and useof it to pressure of the chamber/ air at worst-case flow from the leak could and temperature change in pressure of condenser when the air in much air at condition of 300.

9 Air flow from the leak be over the entire course calculate # of moles of air in air, and use it to the chamber/. the canister is introduced worst-case cfu/m 3. could enter could of be over 2 drying to keepthe the a canister of known volume calculate # of moles condenser when to the chamber, the the # of condition (300. chamber/condenser entire course of 2 . total microbial load below moles of air that causes and still 3keep it the maximum allowable of air in a canister of air in the canister is cfu/m ) could drying to keep the total that pressure change can below the maximum value for cfu/m3. known volume introduced to the be used in a calculation to enter chamber/. allowable value for microbial load below Pressure rise result is chamber, thevoid determine the # ofvolume condenser cfu/m 3. and the maximum calculated from the flow moles of air that of the chamber/condenser still keep it rate and voidvalue allowable volumeforof the dryer3. causes that pressure below the cfu/m ; Pressure rise change can be used in maximum result is then a calculation to allowable value calculated from the determine the void for cfu/m3 flow rate and void volume of the volume of the dryer chamber/condenser 5.

10 Results and Discussion conditions of lowest expected barometric pressure and an inflated room bioburden Calculations were performed over a range of load. This tool can be used as an aid to possible secondary drying times for all assess the potential microbial impact to any Dryers tested, with representative results for cycle that exceeds the established post-test one large dryer and the smaller dryer shown leak rate acceptance criteria. in the table. On the days of Testing , all Dryers meeting the leak rate limit of no more Max. allowable Pressure rise that than mbar pressure rise over 30 leak flow rate to corresponds to minutes would have been able to maintain Hours maintain Class max. flow rate Class 100 / Grade A microbial standards for 100/Grade A. 2 dry (mbar per 30 min). up to 64 hours of secondary drying, even (L mbar/min). with an elevated assumed air bioburden of 3 2 2 2 2. 300 cfu/m . 215 Ft 98 Ft 215 Ft 98 Ft Note that, using this approach, the allowable 1 pressure rise is independent of the size of the dryer , despite the fact that the allowable 2 leak rate is a function of dryer size.