Evaluation of the Microbial Growth Potential of Pharmaceutical Drug Products and Quality by Design

Examples and Case Studies

1. Moist heat sterilization—type of cycle and validation approach: Instead of employing moist heat sterilization cycles with high heat input, low heat input, bioburden-based validation approaches can be followed for a product that does not support microbial growth. Currently, two different approaches exist for sterilizing products on the market. There are multiple products sterilized using an “overkill” approach or cycles delivering excess lethality and validated using biological indicators with many orders of magnitude higher resistance than common bioburden. Alternatively, there are commercial drug products rendered sterile by moist heat terminal sterilization using bioburden-based or bioburden-biological indicator combination approaches delivering low but adequate lethality to assure sterility and safety.

   Terminal sterilization assures a lower probability for a nonsterile unit in a defined load under defined parameters and is less costly than aseptic processing. However, there are commercial products that are aseptically filled even though they can be terminally sterilized without any adverse effects on their stability profile. One reason is that only the high heat input approach was evaluated during development without consideration of bioburden or bioburden-biological indicator validation approaches. By understanding the microbial growth potential of a product formulation, the most heat-resistant bioburden can be selected and used in sterilization cycle development or be compared to the biological indicator. There are approved sterilization cycles that have been developed using biological indicators other than G. stearothermophilus.

   Exploring all options for moist heat sterilization during development is QbD because it leads to a scientific and risk-based sterilization method and cycle type that is validated using the most reasonable and risk-based approach. The most appropriate sterilization cycle is developed and validated for commercial manufacture.

   It is worth noting that there are several commercial drug products including large volume parenterals that are sterilized using moist heat and low heat input cycles—for example, dextrose and salt solutions sterilized with minimum lethality (F₀) of less than 10 min—that have been approved for parametric release. For these products, sterility testing of each lot has been replaced with meeting defined sterilization process parameters (11). If it is possible to sterilize a growth-promoting solution such as dextrose using very low heat input and parametric release, then there are many possibilities for solutions that are not growth-promoting.

2. Increased batch size and holding times for aseptic processing or multiple campaigns: A longer bulk solution holding time can be applied if the product solution does not support microbial growth. However, good manufacturing practices still must be followed along with appropriate process control. This information should not be used to justify and hide inappropriate practices.

   In this example, the holding time of a sterile-filtered bulk solution between the end of filtration and start of filling was increased by more than 300% and up to five filling campaigns (consecutive batches) were approved. This was an interesting case study and may not be common due to extenuating circumstances of the formulation and manufacturing process: the product was highly bactericidal and toxic with a low pH and water content and was manufactured using distillation and redundant sterile filtration. Filling took place inside an isolator. The holding time and multiple campaign changes were validated by media simulations. The case study reveals that process improvements can be achieved if the microbial growth potential of the formulation is known and the aseptic manufacturing process is adequately controlled and validated. Had the firm understood the microbial growth potential of the formulation, these changes may have been incorporated into the manufacturing process at the time of submission of the original application.

3. Preservatives: Preservatives are added to multiple-dose drug product formulations for nonsterile products and during use of sterile products for the purpose of maintaining their microbiological quality over their shelf life (USP 〈51〉) (8). However, it is known that exposure to a preservative may lead to adverse events and allergic reactions in users (12–14). Consequently, for patient safety, it is desirable to use as low a concentration of the given preservative in the formulation of a drug product as possible. This is presented from a microbiological point of view and not a chemistry or stability point of view, as it is well known that preservative amounts may decrease significantly over time depending on storage conditions.

   An understanding of the drug product's microbial growth potential can aid the drug developer in determining both the selection of the preservative system and the safest preservative concentration to incorporate into the drug formulation, while still allowing the drug product to meet the preservative effectiveness acceptance criteria in USP 〈51〉 (8). This approach to preservative system design is different than simply selecting the preservative concentration that was used in the last approved product for the same indication as the drug under development, a seemingly common practice in NDA product applications. Alternatively, the use of a preservative system has also been noted in single-dose products approved in the past. Current thinking and knowledge would not allow a preservative in a single-dose formulation without a valid rationale. Furthermore, a preservative should never be used to justify inappropriate, unhygienic, or nonvalidated manufacturing practices. For example, a single-dose sterile liquid product should not be preserved to justify a poorly controlled aseptic manufacturing process where there are media fill failures and possibly contaminated product. In this case, the process and practices should be revisited and modified instead of adding a preservative to the formulation.

   It is worth noting that propofol, a drug used for general anesthesia, is a drug product with very high microbial growth potential. Adverse events were initially attributed to improper practices such as preparing syringes in batches for use throughout the day, reusing solutions or infusion pump lines on different patients, failure to wear gloves or disinfect the vial stoppers of the 50 mL and 100 mL vials prior to use (15, 16). The propofol formulation did not contain a preservative. Later, it was found that even single-dose vials were mishandled and used in multiple patients. Following these adverse events, sodium metabisulfite, disodium edetate, and other compounds have been added to the formulation to reduce the rate of microbial growth. However, even with these changes, the formulation does not meet the USP criteria for preservative effectiveness. Propofol is an emulsion that consists of soybean oil and/or egg yolk phospholipids that are highly conducive to microbial proliferation following introduction of microorganisms into the sterile product. For this drug, microbial growth studies performed during product development may have led to a different initial formulation, or to approval of the product with the current, strict handling instructions in the package insert and product label. The label has been revised multiple times, including as recently as this year (2010). While mishandling and inappropriate aseptic practices by hospital personnel contributed to the clinical adverse events, perhaps some of these events could have been prevented if the label included the current strict provisions upon approval of the drug, if a different formulation had been developed, or if vials containing low volumes of drug had been marketed so as not to allow for multiple use of a single vial. Subsequent microbial growth studies were conducted, and now the label instructs users that the formulation is inhibitive to microorganisms for up to 12 h in the event of external contamination.

4. In-use storage period prior to administration: Two case studies are presented here where microbial challenge studies resulted in additional product knowledge and optimization with labeling consequences.

   • a) A lyophilized product that contained a significant amount of lactose was to be reconstituted and further diluted with sterile saline to a final concentration of 1 mg/mL and stored for 24 h at room temperature prior to administration. Reconstituted product was inoculated with very low counts of Pseudomonas aeruginosa, Escherichia coli, and Staphylococcus aureus. The comparison controls were tryptic soy broth (TSB) with product (1 mg/mL), TSB without product, saline without product, and negative controls (samples that were not inoculated). Two temperature conditions were evaluated: 2–8 °C and room temperature. The bacterial count was determined at initial time, 4, 6, 8, 12, 20, and 24 h. Product in saline was bactericidal to P. aeruginosa and bacteriostatic to E. coli and S. aureus at both temperature conditions. Product in TSB at 2–8 °C resulted in <0.5 log₁₀ increase for all organisms. At room temperature, two of the challenge bacteria (P. aeruginosa and E. coli) increased >3 log₁₀ after 24 h in the TSB product. TSB without product was bacteriostatic at 2–8 °C. Similar results were obtained with saline without product. As a result of these studies, the proposed in-use storage time of 24 h at room temperature was approved when reconstituting the product with saline.

   • b) A lyophilized product was to be reconstituted with different diluents such as sterile water for injection (WFI), 0.9% sodium chloride injection, and Lactated Ringer's injection, stored at 2–8 °C for 5 days, and then further diluted and stored at room temperature for 48 h prior to administration. Data were provided to support the 5 day storage period at refrigerated conditions. Additional studies were performed to support storage at room temperature as requested by the NDA reviewer. One product vial was diluted with 100 mL of each diluent and inoculated with approximately 10–50 colony-forming units per milliliter (CFU/mL) of each microorganism (Aspergillus niger, Candida albicans, E. coli, P. aeruginosa, and S. aureus). Aliquots were withdrawn and tested at initial time, 24, 36, 48, 60, 72, 84, and 96 h. Similar results were obtained for sterile WFI and saline. A. niger remained relatively stable over the course of the study. Bacterial counts steadily decreased to the point of no detection. C. albicans was not detected in sterile WFI after 72 h, while there were survivors in saline. Similar results were obtained for the yeast and mold in product diluted with Lactated Ringer's. However, the results were different for the bacterial organisms. S. aureus numbers decreased to 1 CFU/mL at 24 h and the organism was not detected again until 96 h, at which point very high growth was obtained. E. coli started growing before the 48 h interval and reached high numbers after 72 h. P. aeruginosa started growing after 60 h. The study was repeated with only E. coli and P. aeruginosa and high growth was obtained for both organisms after only 24 h. As a result of these studies, only sterile WFI and 0.9% sodium chloride injection were selected for drug product reconstitution and dilution. Lactated Ringer's was removed from the label as a recommended diluent.

5. Extended pharmacy bulk holding times: Pharmacy bulk packages contain multiple single doses of a given sterile product. The purpose of the pharmacy bulk package is to dispense product into empty, sterile syringes in a clinical pharmacy. According to USP 〈1〉, the pharmacy bulk package closure may be penetrated only once. A sterile transfer device is then used for the distribution of each single use dose, and the pharmacy bulk package is exempt from the requirement that the product “contain a substance or suitable mixture of substances to prevent the growth of microorganisms.” Because the product does not contain a preservative system, the post penetration holding period during which the product may be dispensed into sterile syringes is limited by FDA during the review of the product application.

   A product with a pharmacy bulk package-approved, post-penetration holding time of 4 h was subsequently approved for an extended holding time of 10 h after submission of data in a supplemental application demonstrating that the product does not support microbial growth. Challenge microorganisms (S. aureus, P. aeruginosa, E. coli, C. albicans, and A. niger) satisfied the USP 〈51〉 definition of no increase in growth when inoculated in the product and held at the intended storage conditions over the proposed extended holding period. The submission of this study led to approval of the extended post-penetration holding period of 10 h.

6. Therapeutic drug proteins and process design: This section presents a case study from an actual application of employing the knowledge gained from microbial growth studies to design the manufacturing process. Biological products are generally growth-promoting because they are derived from living systems. This is especially true of proteins in buffer matrices. However, understanding the level of growth promotion can result in the design of a more robust process. In this case, microbial studies were performed using the bulk solution and different concentrations of protein in the buffer matrix. During routine manufacture, the formulation is compounded and filtered after thawing and pooling of the drug substance. The holding time of the resulting bulk solution is a critical parameter of the aseptic process.

   A rather high inoculum was used for the microbial challenge studies in the range of 10⁴–10⁵ CFU/mL. Different drug substance concentrations were tested. The microbial count was evaluated at initial time, 6 h, 1, 3, 7, and 14 days for P. aeruginosa, S. aureus, C. albicans, and A. niger. A. niger counts remained static but there was formation of mycelia after 7 days. C. albicans had increased in numbers by Day 1 and continued to increase >10-fold after 3 days. S. aureus numbers remained static up to 3 days and then started decreasing. P. aeruginosa increased >10-fold after 1 day of incubation. As a result of these studies, the maximum holding time from start of compounding until end of sterile filtration was limited to ≤12 h. These data demonstrated that the solution is highly growth-promoting and led to the design of a process with a pre-filtration step followed by sterile filtration. Additional holding time limits were built into the process for drug substance thawing and start of compounding to end of pre-filtration. It is worth noting that the applicant conducted the studies during development, effectively applying the QbD concept.