Barrier Isolation Technology: A Safe and Effective Solution for Providing Pharmaceutical Development Facilities
By Daniel Liberman, Christopher Lockwood, Mary McConnell-Meachen, Eugene McNally, Hank Rahe, Kevin Shepard, and Glenn Snow
Introduction
In the last 10 years, the pharmaceutical industry has been discovering and developing increasingly more potent drugs. This increase in potency of new drug substances has reduced patient dosage for new drug approvals. While this trend is a positive outcome for the patients requiring the medication, it has created increasing problems for individuals involved with the development of these potent New Chemical Entities (NCEs).
The trend toward increased potency of the active drug substance not only includes cytotoxics, but also most major categories of pharmaceutical compounds. With this increase in potency of NCEs comes the need for greater safety measures to protect workers developing the compounds. Traditionally, formulation and process development work with NCEs has performed in open laboratories with scientists wearing personal protective clothing and respirators to guard against skin contamination and inhalation exposure during dusty processing operations. However, this dependence on personal protective equipment may be inappropriate when the exposure limit for some of these compounds is in the sub-microgram region.
Boehringer Ingelheim (BI) has adopted a strategy that does not rely on Personal Protective Equipment (PPE) to protect workers. They have adopted the philosophy of containing the potent material at its source, the processing equipment itself. Using barrier isolators, contaminants are confined to the equipment that generates the contaminants. In this way, very low exposure levels can be achieved, and the dependence on PPE to protect workers can be minimized or even eliminated.
This article will describe an approach for processing potent compounds using pharmaceutical solid dosage form equipment. Isolating the worker from the potent compound prevents worker exposure without relying on PPE. The isolator systems, the work practices necessary to maintain containment of the potent material, and the validation data generated that demonstrate containment has been achieved will be described.
Background
Like many of the major pharmaceutical companies, BI developed a classification system for describing the level of potency of materials. Table A summarizes a five-category classification system for NCEs and the criteria used in assigning a classification to a compound. All relevant data about the compound are reviewed. This includes chemical, physical, pharmacological, and toxicological data from both human and animal subjects. A hazard assessment is conducted and a determination is made as to which Hazard Category is the "best fit". Team members collectively apply their expertise in industrial hygiene, toxicology, pharmacology, occupational medicine, and clinical medicine to review the data for the pharmaceutical active ingredient and make the hazard category assignment. Both acute and chronic data are considered, and the assignment relies on professional judgment. To assess potential acute effects, both the toxicity and pharmacological activity of the compound are evaluated. The type of pharmacological effect(s) expected, the mechanism of action, and the dose required to produce these pharmacological effects are important considerations, as is the severity of acute (life threatening) effects. This latter assessment is a determination of whether medical intervention might be required and how rapid the response must be if an overexposure occurs. This information in conjunction with the results of acute toxicity studies in animals provides the likelihood that the compound may produce immediate adverse effects. Compounds with a high order of acute toxicity and poor delayed warning properties are of concern.
A determination is made on the likelihood and severity of possible chronic effects. This weight-of-evidence evaluation is based on the results of genotoxicity assays in cell culture, in vitro experiments, in vivo studies in laboratory animals that are designed to determine the potential for the material to produce target organ effects, reproductive or developmental toxicity, cancer, or other chronic effects. Where possible, the results of clinical studies in humans are used as well. A key piece of information is the dose required to produce these effects, or preferably, the highest dose that does not produce a toxic or pharmacological effect (i.e., the No Observable Effect Level or NOEL). In the event that a NOEL cannot be defined, we attempt to define a No Toxic Effect Level or NOTEL.
A judgment is made regarding the severity of chronic effects and whether they may have disabling consequences or the potential to cause early death. A very important consideration is whether effects are reversible or irreversible. The procedure that we follow in developing Occupational Exposure Level (OEL) consists of the five steps outlined in Table B. Often the hazard category assignment is conservatively based on the most sensitive health effect endpoint, especially when there is potential for life threatening, disabling, or other irreversible chronic effects. During the early discovery phase of a compound, toxicological profiling is limited by the small quantities available. In the absence of valid toxicological information, research compounds are provisionally assigned to Category 3, unless the molecular structure or other indicators suggest a higher or lower classification. As additional data defining the risk of exposure becomes available, the categorization of the new chemical entity can be changed.
The equipment and containment devices that will be described have been designed to handle material in Categories 1-3 without the use of PPE. Category 4 compounds will require additional evaluation of both the compound and the containment devices to determine if they can be processed safely. Additional decontamination techniques, the use of PPE, and potential personnel exposures and allow for the safe processing of compounds in Category 4. The derivation of OELs for therapeutic substances is not a precise scientific exercise; it is a matter of judgment involving medical, toxicological, and industrial hygiene disciplines.
Processing of Potent Compounds
At larger scale (50 kg and above), pharmaceutical equipment can be purchased with containment options that limit worker exposure to dust during processing; transport of materials between unit operations can be accomplished by vacuum transfer between closed bins. However, such containment options are not available on smaller capacity laboratory equipment used when marking the first solid dosage form for a NCE at the 200g - 1 kg scale. At this small scale, past practice has been to work in a chemical fume hood, or in an open lab wearing some form of respirator to prevent inhalation of dust. However, the goal was to be able to work with small-scale equipment in the lab in either a contained or non-contained manner without the PPE and the facility clean up issues. This will allow the development of all levels of compounds in house.
When performing analytical testing, potent compounds need to be treated as hazardous and work with them in a properly contained environment. This means that for processing operations that generated considerable amounts of dust (i.e., dispensing active, mixing, granulating, milling, tabletting), containment strategies need to be developed in order to do prototype formulation development work. We found that we needed to develop criteria to guide our decisions. These criteria evolved as we learned the advantages and disadvantages to various isolation/containment approaches. The initial criteria were:
- Safety of personnel working with the potent material and those scientists in adjacent labs not involved with these materials
- Flexibility for using lab for both potent and non-potent compounds
- Ability to work on a scale of 200 grams to 2 kg
- Ability to adopt new process equipment into the lab without redesigning containment practices
- Allow development on compounds with an OEL on the order of 0.1 micrograms/cubic meter of air
Stategy Development
A discussion of how our containment strategy evolved is informative since it identifies the approaches that were evaluated and abandoned due to their impracticality.
Our first generation concept was to construct a down draft booth that would be used in conjunction with PPE. It was felt that this would be a quick solution for handling potent compounds; however, based on achievable particle capture, this approach would not be useful for compounds with OELs of 50ug/cubic meter or below without using PPE. When the operation was complete, the entire laboratory potentially would have to be cleaned, and we would need to engineer a means of entry and exit into the lab for workers to safely remove their PPE.
The second generation approach was adopted to decrease our reliance on PPE and to minimize the need to clean up a large laboratory space upon completion of an operation. We envisioned that all work would take place in a HEPA filtered lab fitted with airlocks and mist showers to contain the material within a smaller lab space. This would allow personnel to exit and then remove their PPE safely. All material transfers and dust generating operations would be performed in a glove box within the HEPA filtered laboratory. One large glove box would accommodate the operations of weighing active drug, granulating, milling, blending, and tabletting. An initial cleaning of residual powder would be performed inside the glove box and then the equipment would be removed for final cleaning and breakdown during which lab personnel would be wearing PPE. As we worked through the logistics of this approach, several problems were identified. First, the concept that one glove box could support all of the equipment in a standard wet granulation process was impractical. The ergonomics associated with operating and cleaning a tablet press, a high shear granulator, a mill, a fluid bed granulator/dryer/coater, etc, were unique enough that one glove box could not accommodate all of these operations. Several stopping points would have to be designed into the usual workflow of the granulation process to clean equipment, move it out of the lab, and pull new equipment in. Depending on the OEL of the drug being processed, this movement of equipment in and out would necessitate a cleaning of the entire laboratory prior to equipment change over. These potential stopping points in the process to clean the lab and move equipment in and out contradicted our preferred work practice of cleaning after completing the entire manufacturing process.
Our third and final concept was to contain the potent compound at source in the equipment. We developed an isolator system, which included a HEPA filtered air design component such that all of the engineering containment controls would be designed into the isolator and not into the laboratory as in our earlier approaches. All material transfers would take place in a closed environment, either inside the barrier isolators or within a HEPA filtered weighing hood. This allowed the equipment to be operated in normal laboratory space and did not require air locks and exit showers.
Description of Equipment and Isolators
Figure 1 depicts a typical wet granulation process for producing pharmaceutical tablets. The process is a series of weighing, mixing, granulating, drying, and compression operations all of which are contained in isolators or performed inside a HEPA filtered weighing hood. No material transfers take place outside of a HEPA filtered environment. The weighing hood is connected to a HEPA filter and is used to weigh out the active ingredient, and to dissolve this material in water before being added to the binder solution. The general isolator is designed to contain multiple pieces of equipment one at a time. It is used for operating an oscillating mill, a high shear co-mill, and a high shear granulator. In addition, it can be used for charging and discharging mixing totes, that once loaded, can be cleaned, removed, and mixed in the open lab using a drive unit that is not contained. The fluid bed isolator is used to operate a bench top, fluid bed unit that can be used for granulation, drying and Wurster coating. The third isolator is dedicated to operation of a bench top 10 station instrumented rotary tablet press. All of the isolators are complete with HEPA filters that re-circulate air back into the isolator with a small amount of the filtered air being vented into the room to keep the unit under negative pressure. For the purpose of this article, we will focus on describing the process by which tablets are made inside the containment isolator dedicated to the tablet press. Ergonomically, this is the most challenging operation to achieve inside an isolator and illustrates all of the handling practices used through a wet granulation process using all of the isolators.
Operation of Tablet Press Inside Isolator
The first step is to tool up the press as needed, connect the electrical cables, misting wand, interior vacuum hose, and transfer all materials required for compression into the isolator at this time. The tablet pres isolator is shown in Figure 2. If an isolator panel has been removed, it is more convenient to tool up and make all connections prior to reinstallation of the panel. Once this isolator panel has been reinstalled, the Ultra Low Penetration Air Filters (ULPA) equipped vacuum is connected to the isolator manifold - Figure 3. Sealed polyethylene sleeves are installed on the bag out ports. Once these installations are complete, the isolator is turned on and the chamber air pressure is checked via a magnetic gauge located on the front panel - Figure 4. This ensures the unit is sealed and not leaking. To start the tabletting operation, the powered hopper is filled with granulation and the manifold is adjusted to turn on the vacuum for local dust collection within the tablet press. The tablet press force monitoring system is turned on and the press is started using the externally mounted press control panel - Figure 5. The vacuum is adjusted for the proper amount of turret dust extraction using a butterfly valve connected to the isolator manifold. The isolator chamber air pressure is adjusted to compensate for the turret vacuum by adjusting a slide gate valve located on top of the isolator - Figure 6. Initial tablet samples are collected by passing tablets out the tablet chute directly into a polyethylene bag for evaluation of tablet weight, thickness, and hardness. The press is then adjusted as needed and granulation can be added to the hopper as required by filling from inside the isolator or bagging material in from the outside via the sealed polyethylene sleeve located just above the hopper. When compression is complete, the product can be removed through a bag out port or through the pass through chamber - Figure 7. The tablet press and isolator are now ready for cleaning.
Cleaning Strategy
The term cleaning means to reduce the concentration of the active ingredient on surfaces of the machine and the enclosure to an established acceptable level. Cleaning is divided into deactivation and cleaning. Deactivation is the operation that is performed to reduce or immobilize the potent compound to the point that the material cannot produce an airborne concentration exceeding the exposure guideline for the compound. This is accomplished by a number of steps and should be validated for individual compounds because each has different characteristics in terms of particle size, solubility, density, surface characteristics, etc. The validation of cleaning is required for the protection of employees when working with potent compounds.
Description of Cleaning Process
After processing is complete, an initial vacuuming (ULPA filtered Vacuum) is performed to remove gross powder. Parts routinely cleaned in a sink can be "bagged-out" for cleaning. These bags are opened in the sink under water to minimize dust exposure. Larger pieces of the equipment are cleaned inside the isolator by wiping first with a cleaning solution in which the compound is soluble, followed by rinsing with water. Based on the sampling results described later, a misting wand was attached to the manifold for a direct supply of water. For hard to reach places a series of special tools such as squeegees are useful. Testing of the deactivated equipment inside the isolator is routinely performed as a part of the validation process that will be described later. In the case of non-GMP operations, if the deactivation step is being used as a cleaning step, a final rinse may be used. All deactivation materials are bagged out of the isolator and disposed of in a safe manner. Once the unit has dried, a visual inspection is conducted to determine if any visible powder can be detected. If the compound is highly potent at levels of less than one microgram, it is good practice to collect a swab from several locations within the enclosure and have it analyzed to determine that the potent compound is not present above the exposure limit before opening the isolator.
Once the isolator has been cleaned, it can be opened and the equipment can be further cleaned if GMP cleaning clearance is needed to release the equipment prior to the production of a future GMP batch. Such cleaning can be performed in accordance with the standard operating procedures for a particular piece of equipment.
Validation
Our approach has been to use lactose as a surrogate for the drug in the first round of validation experiments. Lactose was selected due to its dustiness and the availability of sensitive analytical methods. The low detection limit allowed sample collection intervals to coincide with the normal work activities: set up, compression, tablet testing, and cleaning. Prior to safety batch operations, background samples were collected in the lab during routine activities. During safety batch operations, personal sampling pumps equipped with 25mm glass fiber filters were placed on both operators, downstream of the HEPA filter exhaust from the isolator and next to the ULPA filtered vacuum. The filters were changed out at various stages of the process to quantify exposures specific to each part of the process. Upon completion of testing, the filters were sent to an independent testing laboratory for HPLC analysis. The personal sampling results were used to generate either hour Time Weighted Average (TWA) exposures. These TWA exposures were then compared to the OEL established for the compound of interest. In addition, a P TrackTM Ultra fine Particle Counter was used to identify possible points of containment release. Upon completion of the cleaning operation, a sample was collected inside the isolator.
The initial personal and area samples results are shown in Table C. Measured concentrations ranged from 0.12 ug/m3 to 4.4 ug/m3. Using the particle counter, leakage from the ULPA filtered vacuum at a level of 19,000 particles/cm3 was measured at one location at the filtered exhaust. The leak was a likely contributor to the high than expected personal exposures. As a result, the ULPA filter was changed and tested again prior to use in the next batch. The sample collected inside the isolator after cleaning had measurable levels of lactose (0.44 ug/m3). To reduce residual airborne dust, the cleaning procedure was revised to include covering the press with a polyethylene bag after the press was cleaned and misting the interior of the isolator using a misting wand.
The next simulation batch was sampled as described previously. Results from air monitoring are shown in Table D. Personal sampling results ranged from 0.21 ug/m3 to 0.91 ug/m3 for Operator 1 and were below the limit of detection for Operator 2. The highest concentration measured for Operator 1 occurred during cleanup. While collecting this sample, the seal on an out port bag ruptured possibly resulting in a significant release of lactose. Equipment was heat and tie sealed when being bagged out of the isolator to prevent bag ruptures in all future batches. Area samples were below the OEL with the exception of one sample collected near the vacuum (2.3 ug/m3).
Due to the continuing high concentration of lactose measured near the vacuum, we replaced the ULPA filtered vacuum and isolated the vacuum from the lab. Exposures were higher for Operator 1 who removed the tablets from the isolator and performed tablet testing in the vented weighing hood than for Operator 2 who worked at the isolator only - Table D. Due to these differences, tablets were removed from the isolator in bags or covered weigh boast. In addition, PPE was upgraded to double gloves and sleeves for tablet testing, and the operator performing testing removed and bagged the outer layer of PPE inside the hood before taking his hands out of the hood. Results from the third validation batch are shown in Table E. These results were all below the detectable limit of the lactose assay. The experiment was replicated to demonstrate repeatability and similar results were obtained and shown in Table F.
Discussion/Conclusions
The primary objective of this work was to develop an approach that would provide adequate worker safety during small scale processing of potent compounds classified as Category 3. The approach was to include a strategy for protecting both the operators working directly with the potent materials and other laboratory personnel throughout the facility. The approach described above provides processing capability under the high level of worker protection appropriate for Category 3 while allowing processing in general laboratory space. This feature minimized the need for costly facility modifications such as airlocks, showers, and changing rooms, which would be required to contain material to a given laboratory and effectively protect the surrounding facility in the absence of a barrier isolator. While this approach minimizes the need for worker PPE for processing most Category 3 compounds, one can still work with more toxic Category 4 materials by adding the appropriate PPE and administrative controls as additional protection. Lastly, the use of isolator technology is consistent with the OSHA directive to minimize reliance of PPE controls and to maximize the role of engineering controls as the main basis of worker protection when handling potent compounds.
Barrier isolation is not an instantaneous fix for handling potent compounds. The proper design of the isolator for a particular piece of equipment or process is a nontrivial task. It requires an intimate knowledge of the manufacturing process and equipment operation. The development of each isolator has been a multi step process. A detailed mock-up of the isolator to determine the optimal configuration of gloves, sample ports, and equipment utilities is needed. This mock up also allows the user to address ergonomic issues of the process. Evaluation of each stage of the design by skilled operators with detailed knowledge about the process is crucial to developing an effective isolator that is easy to use. Identification of the optimal location of controls to minimize exposure of electronic and pneumatic controls required considerable consideration and evaluation of the functional operation of the equipment to obviate cleaning and contamination issues. This ensured that the equipment was safe for operation outside of the isolator using Category 1 and 2 compounds. Next, after the design was structurally sound, a prototype isolator was constructed. Performance testing and validation were then performed with lactose to determine the best operating practice to minimize contamination and maximize ease of use prior to working with potent material. As more experience was gained working with the isolator, minor modifications were made to the prototype to arrive at the final working configuration.
The use of an easily detectable surrogate compound (lactose) during the initial stages of the isolator validation proved very useful. We were able to test the integrity of the isolators as well as the adequacy of procedures for operating the equipment and handling of potent materials without operator exposure. After achieving acceptable air monitoring and cleaning level result using lactose, active batches may be tested using the procedure developed with the lactose batches.
During initial batch development, workers should be in PPE and the room closed for use until containment of the compounds of interest could be demonstrated using analytical results from swabbing the equipment, including areas outside of the isolator for the potent compound. This type of validation procedure should be completed for each new compound used inside the isolator.
It is important to note that the design of any containment system for working with potent compound should not be "cookie-cutter" approach. The system must be designed with the needs of the operators, the existing facility, and the process in mind. Advantages and disadvantages exist for all of the containment options available. Not all processes are amenable to every containment technology. As indicated previously, barrier isolators were only one of the approaches considered for the containment of potent compounds. Several others, as discussed above, were evaluated for our particular situation, and several design iterations were completed before we selected this particular solution for our facility. For instance, we found containment of potent material in an existing tray dryer. Table G outlines some of the advantages and disadvantages for the barrier isolator systems described in this work. After potent materials at the source, and the flexibility desired, we concluded that an isolator system was the best alternative to meet out particular needs.
About the Authors
Daniel F. Liberman, PhD, is the Associate Director for Environmental Affairs and Safety for Boehringer Ingelheim Pharmaceuticals, Inc. Liberman is responsible for managing the environmental health and safety programs at BIPI, and serves as an internal consultant on biological and chemical safety to the other Boehringer Ingelheim Companies in the U.S. He has a BS and an MS in biology and a PhD in radiation biology and biophysics. He has served as the Associate Editor of Chemical Health and Safety, an American Chemical Society (ACS) Journal devoted to chemical safety and has edited three books on the management of health and safety hazards in the workplace. Currently, Liberman is working with colleagues in the US and Germany to harmonize occupational health and safety practices for the development of active pharmaceutical ingredients.
Boehringer Ingelheim, 900 Ridgebury Rd., Ridgefield, CT 06877.
Christopher E. Lockwood is presently a Senior Scientist in Pharmaceutics at Boehringer Ingelheim. He received his BS in chemical engineering (1993) and PhD in medicinal chemistry and pharmaceutics (1997) from the University of Kentucky. He currently works in formulation and process development for oral solid dosage forms. His research interests also have included protein reformulation, formulation, and novel separation techniques. The author or co-author of a number of professional publications, abstracts, and scientific presentations, he is a member of the American Association of Pharmaceutical Scientists.
Boehringer Ingelheim, 900 Ridgebury Rd., Ridgefield, CT 06877
Mary McConnell-Meachen, CIH, CSP, is the Senior Industrial Hygienist at Boehringer Ingelheim. She is responsible for the identification, evaluation and control of employee exposure to hazardous substances and active pharmaceutical ingredients. She manages the Industrial Hygiene Program for the Connecticut campus, and provides technical and program support to other Boehringer Ingelheim US locations. She is a Certified Industrial Hygienist and Certified Safety Professional with 23 years of safety and industrial hygiene experience. She has worked in the insurance, environmental, chemical, and pharmaceutical industries. Her bachelors degree in environmental sciences was obtained at the University of Massachusetts at Amherst.
Boehringer Ingelheim, 900 Ridgebury Rd., Ridgefield, CT 06877
Eugene J. McNally is presently an Associate Director, Department of Pharmaceutics, Boehringer Ingelheim. He received his BS in pharmacy from Duquesne University (1984), and an MS (1987) and PhD (1989) in pharmacy from the University of Wisconsin-Madison. At Boehringer Ingelheim, he is responsible for preformulation/formulation development of recombinant proteins, delivery of macromolecules, and formulation and process development for oral solid dosage forms. His areas of research interest have included the characterization of protein instability, protein conjugation, and macromolecule drug delivery. The author or co-author of a number of professional publications, abstracts, and scientific presentations, he is a member of the American Association of Pharmaceutical Scientists.
Boehringer Ingelheim Pharmaceutical, Inc., Mail Drop R1-3, 175 Briar Ridge, Ridgefield, CT 06877.
Hank Rahe is Director Technology at EnGuard Systems. He is a frequent speaker and course leader for ISPE and has served as president of the Society. He writes a monthly column for Cleanrooms Magazine and has authored more than 20 articles on containment of potent compounds. Academic credentials include a BSIM and MSE from Purdue University where he served as a guest lecturer for four years.
EnGuard Systems, 10329 Vandergriff Rd., Indianapolis, IN 46239
Kevin Shepard is currently a Senior Research Technician at Boehringer Ingelheim in the Department of Pharmaceutics. He is responsible for evaluation and testing of small scale manufacturing equipment and the fabrication of change parts to adapt commercial equipment to development activities. His responsibilities also include manufacturing and evaluation of development batches and other formulation development experiments.
Boehringer Ingelheim, 900 Ridgebury Rd., Ridgefield, CT 06877
Glenn Snow is presently a Senior Research Associate, Department of Pharmaceutics, Boehringer Ingelheim. He received his BA in chemistry from Western Connecticut State University (1994). At Boehringer Ingelheim, he is responsible for preformulation/formulation development for oral solid dosage forms. He is a member of the American Chemical Society.
Boehringer Ingelheim, 900 Ridgebury Rd., Ridgefield, CT 06877
