Wyeth Carolina Uses PAT to Monitor an Oxygen-Sensitive New Parenteral

Dec. 20, 2005
Close teamwork, and incorporating validation criteria into the FAT test, helped Wyeth’s team meet a tight drug approval and commercialization timeline.
One of our key products at Wyeth Carolina is a new parenteral that is highly sensitive to oxygen and can degrade from continued exposure. During manufacture, great care must be taken to prevent oxygen ingress. Thus, for example, the lyophilized product is sealed under partial vacuum in a nitrogen atmosphere. During the vacuum break cycle, nitrogen is back-flushed into the lyophilization chamber to 860 mb pressure and the stopper is then set.Last year, our team saw an opportunity to use Process Analytical Technology (PAT) to expand our capabilities to monitor oxygen with this product. PAT’s overall goal, as defined by FDA, is to design, analyze and control manufacturing by measuring critical quality and performance attributes during manufacturing. If designed right, a PAT application would allow us to monitor and control a critical product attribute and ensure the safety and efficacy of the drug from manufacture through distribution and sale.There are many possible ways that air can enter a vial before sealing or capping. A stopper can stick to the lyophilizer shelf (due to shelf finish, defective stoppers, stopper siliconization, or defective stopper finish material) and become unseated when the shelf rises, or it can become partially dislodged during manual transfer from the lyophilizer to the sealing area. The vial can be defective, cracked, or out of specification.Air ingress can increase headspace pressure in the vial by up to 1 atmosphere, causing oxygen content in the product to increase to 3%, well above acceptable levels. (The pressure value reference used is 1,013 mb at sea level. The 3% value is based on available headspace volume and ingress of 173 mbar of air.) To prevent this from occurring, we needed some way to monitor oxygen levels within the vial headspace. A non-destructive technologyThe traditional method of doing this is through a destructive laboratory analysis involving gas chromatography. Since we had used this test successfully for batch release and it met all component testing requirements, we had no intention of replacing it. However, we knew that we needed a technology to support our oxygen monitoring efforts. Other technologies were available, but each had its own limitations:
  • Dye-leak tests
  • are not suitable for 100% in-line inspection.

  • Pressure decay tests
  • offer a leak detection limit of less than 5 microns.

  • Pressure decay, dye leak and helium tests
  • require that the container leak at time of measurement.

  • Spark tests
  • might impact product.

  • Moisture/pressure tests
  • may prove inaccurate since moisture amounts can vary between lots.
After evaluating the alternatives, we opted for a new technology based upon laser technology and Frequency Modulation Spectroscopy. The automated oxygen analyzer we chose was the Wilcomat R4 VR, manufactured by Wilco AG (Wohlen, Switzerland).We had a very tight timeline for the project, and had to meet target dates for FDA approval and product launch. Wilco’s technology offered a number of benefits, while the vendor had domain experience, good technical support, and documentation and validation experience.Wyeth Carolina bought the Wilcomat machine as an in-line inspection that could identify and/or measure critical material or process attributes for each lot of product. The Wilcomat machine met the required definitions of PAT, since it is a process analyzer system and provides non-destructive measurements that deliver information on critical product attributes. The ability to measure relative attribute differences within each lot provides useful product information for process control and/or process feedback.Our team decided to install the analyzer in the packaging area, to inspect for oxygen content in that part of the plant, rather than immediately after vial sealing. We had two very practical reasons for doing this:
  • We wanted to inspect the product as late as possible in the process, so that we could maximize inspection opportunities and remove any defects that might be introduced during the sealing process (e.g., cracks under the seal finish);

  • Space was limited in the sealing area, so there wasn’t enough room to allow a complete cosmetic and oxygen content inspection to be performed.
The machine is a four-head station precision tester. Each head contains a laser unit that passes a laser beam through the upper headspace area of each vial as shown in Figure 1 (to access the figures accompanying this article, click the Download Now button at the end of the page).Frequency Modulation Spectroscopy uses a modulating wavelength-calibrated low-energy laser beam (whose midpoint was tuned at approximately 760 nm, the absorption frequency of oxygen), which is directed through the vial’s headspace. The laser beam detects changes in this energy when increased levels of oxygen molecules are detected within a container. The principle behind this measurement is that the energy from the laser diode changes the laser wavelength. The comparison, peak to amplitude, changes with variation of oxygen content (Figure 2, accessible by clicking the Download Now button at the end of this article).During each vial inspection, the laser transmission power is monitored. If it is below or above set levels, the data result is deemed “unknown,” and the vial is directed to a special collection area for evaluation and possible re-testing. Transmission errors can result from laser beam impairment or blockage (i.e., when powder particles are on the vial wall). The pre-set level includes a safety factor so that the minimum transmission limit is above the minimum molecule sampling point required for accuracy.
The oxygen analyzer used at Wyeth Carolina, the Wilcomat R4 VR. Photo courtesy of Wilco AG.

The inspection machine also monitors the laser head operation and the air environment around the exterior of the vial. During the inspection cycle, the target vial to be inspected is automatically inserted upwards into the laser head with minimum clearance, but there remains some minimal air between the beam emitter and receiver and the vial glass. This external air is monitored with a certified 0% oxygen calibration vial inside the head, which is simultaneously moved into the inspection position when the target vial is moved out of the head. Readings are recorded and the data are used to keep the laser head in calibration with the external environment and within pre-configured operational settings so that measurements remain in laser head specification limits.The analyzer features multiple alarm functions that stop the equipment if inspection operations are not within pre-set function criteria.The inspection machine’s detection limit is between 0.5 and 3% oxygen, depending on the vial diameter, the vacuum level and level of nitrogen back-flush. Lower vacuum, starting at about 700 mbar, may overcome the possible noise levels of nitrogen, which improves the limit of detection.The measurement time can be configured to either 0.3 seconds to 1.2 seconds, depending on the sample time required to obtain reliable data. We have programmed the machine to not accept vials containing elevated oxygen contents of 3% or higher.During early development testing, air was injected to simulate 2.5% oxygen content (chosen due to the elevation of the Swiss testing location), and readings were taken with a Wilco benchtop analyzer.The in-line system was calibrated using 0% and 3% oxygen. The actual inspection set point is configured at a lower value to ensure a 99.99% confidence that vials with 3% oxygen or higher would not be accepted.Teaming up for FAT and SATWyeth’s team worked closely with Wilco at every point in the project’s development life cycle, from the project scope and User Requirement Specifications (URS) stage up to the Qualification Process (Commissioning, IQ, OQ, PQ). For both parties, the goals were an implementation that would achieve consistent, reproducible readings, and that was 99.99% certain to reject any vials with air ingress. Factory Acceptance and Site Acceptance Tests (FAT, SAT) provided the grounds for the machine’s development phases and performance evaluation.The Factory Acceptance Test (FAT) was written with the development process in mind. No historical data were available for these particular inspection criteria. Set point development was a critical part of the FAT process. Normally, a FAT is conducted entirely against the function and design documents designed to meet the URS.In this case, though, since the product’s launch date was fast approaching, development had to occur within the FAT framework. This process succeeded, as all design and function tests were carried out.
Wyeth’s FAT Team on its visit to Wilco in Wohlen, Switzerland.

As material handling tests began, data were gathered in parallel. This approach allowed the FAT team (with Wilco technician’s support) to use equipment time most effectively. It also eased the transition into performance testing. These tests were specifically designed to make use of some development data, and yet challenge the system based on the configuration settings as they’d been developed.Due to time constraints, the machine was shipped and installed before conclusion of the FAT. This qualification process carried forward from the initial FAT through an on-site FAT/SAT/commissioning after the installation. Additional adjustments were made upon installation, for completion of the FAT requirements.Incorporating IQ, OQ, and PQ test similarity into the FAT allowed the system to be challenged against validation type testing while at the manufacturing vendor’s location.Wyeth Carolina’s FAT team included staff from engineering, technology, and maintenance departments. Wyeth ensured that these personnel fully understood the Wilco system and, thus, that PQ improvements had been incorporated during the FAT, so the on-site commissioning, qualification and validation exercises were completed efficiently and within the project’s timeframe. Close teamwork helped make the project a success.Now that the PAT installation is up and running at Carolina, data from the system will be continuously evaluated to verify the equipment’s performance. We are confident that this use of PAT will help our site maintain product quality, and customer service as well.References
  1. FDA Guideline. General Principles of Process Validation, May 1987.

  2. FDA Guideline. Container and Closure Integrity Testing in Lieu of Sterility Testing as a Component of the Stability Protocol for Sterile Products, March 18, 2003.

  3. FDA Guidance for Industry. Container Closure Systems for Packaging Human Drugs and Biologics, May 1999.

  4. FDA Guidance for Industry. Sterile Drug Products Produced by Aseptic Processing — Current Good Manufacturing Practice, September 2004.

  5. FDA Guide to Inspections of Lyophilization of Parenterals, July 1993.

  6. FDA Guide to Inspections of Sterile Drug Substance Manufacturers, July 1994.

  7. FDA Guidance for Industry, PAT — A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance, September 2004.
About the Authors

John Walsh
is a Technology Manager, Global Technology, for Wyeth in Collegeville, Pa., supporting new technologies and validation applications at various manufacturing sites. He has been with Wyeth for over 15 years.Dave Bertsch is a Principal Equipment Engineer at Wyeth corporate headquarters, supporting manufacturing and packaging sites on technical equipment applications. He has been with Wyeth for 27 years.Vivianne Colón is a Technology Scientist at Wyeth Carolina, Puerto Rico, currently working in the support of technology development projects which include the transfer of products, processes, and equipment technology for parenteral products. She has been with Wyeth for four years.
About the Author

Vivianne Y. Colón | John E. Walsh