Abstract
Environmental monitoring and aseptic process simulations represent an integral part of the microbiological quality control system of sterile pharmaceutical products manufacturing operations. However, guidance documents and manufacturers practices differ regarding recommendations for incubation time and incubation temperature, and, consequently, the environmental monitoring and aseptic process simulation incubation strategy should be supported by validation data. To avoid any bias coming from in vitro studies or from single-site manufacturing in situ studies, we performed a collaborative study at four manufacturing sites with four samples at each location. The environmental monitoring study was performed with tryptic soy agar settle plates and contact plates, and the aseptic process simulation study was performed with tryptic soy broth and thioglycolate broth. The highest recovery rate was obtained with settle plates (97.7%) followed by contact plates (65.4%) and was less than 20% for liquid media (tryptic soy broth 19% and thioglycolate broth 17%). Gram-positive cocci and non-spore-forming Gram-positive rods were largely predominant with more than 95% of growth and recovered best at 32.5 °C. The highest recovery of molds was obtained at 22.5 °C alone or as the first incubation temperature. Strict anaerobes were not recovered. At the end of the five days of incubation no significant statistical difference was obtained between the four conditions. Based on these data a single incubation temperature at 32.5 °C could be recommended for these four manufacturing sites for both environmental monitoring and aseptic process simulation, and a second plate could be used, periodically incubated at 22.5 °C. Similar studies should be considered for all manufacturing facilities in order to determine the optimal incubation temperature regime for both viable environmental monitoring and aseptic process simulation.
LAY ABSTRACT: Microbiological environmental monitoring and aseptic process simulation confirm that pharmaceutical cleanrooms are in an appropriate hygienic condition for manufacturing of sterile drug products. Guidance documents from different health authorities or expert groups differ regarding recommendation of the applied incubation time and incubation temperature, leading to variable manufacturers practices. Some recent publications have demonstrated that laboratory studies are not relevant to determine the best incubation regime and that in situ manufacturing site studies should be used. To solve any possible bias coming from laboratory studies or single-site in situ studies, we conducted a multicenter study at four manufacturing sites with a significant amount of real environmental monitoring samples collected directly from the environment in pharmaceutical production during manufacturing operations with four solid and liquid nutrient media. These samples were then incubated under four different conditions suggested in the guidance documents. We believe that the results of our multicenter study confirming recent other single-site in situ studies could be the basis of the strategy to determine the best incubation regime for both viable environmental monitoring and aseptic process simulation in any manufacturing facility.
- Environmental monitoring
- Aseptic process simulation
- Primary isolation of microorganisms
- Incubation conditions
- Incubation temperature
- Microbiology
- Cleanroom
Introduction
The evaluation of the microbiological quality of sterile pharmaceutical products is mainly based on sterility tests, environmental monitoring (EM), and aseptic process simulations (APS), but the incubation program of culture media used for EM and APS can differ depending on recommendations or manufacturers, although nothing is indicated in the current EU Annex 1 (1). The sterility tests performed following the harmonized monograph 2.6.1 of European Pharmacopoiea (2) must be incubated at 20–25 °C for tryptic soy broth (TSB) and at 30–35 °C for thioglycolate broth (THIO), at least 14 days each. For APS, the Pharmaceutical Inspection Co-operation Scheme (PIC/S) recommendation (3) is to incubate first for 7 days at 20–25 °C and then for 7 days at 30–35 °C although other guidance documents recommend only incubation at a single temperature between 20 °C and 35 °C (4, 5). For EM, recommendations are made to incubate the sample at a single temperature between 20 °C and 35 °C (6) or at two temperatures, 20–25 °C and 30–35 °C (4, 7, 8) although only a few manufacturers take two samples at the same location, one incubated at 20–25 °C and the other at 30–35 °C (9). However, the incubation of the same sample at two different temperatures may result in a thermal shock, reducing the total recovery rate (10, 11). In PDA Technical Report No. 13 (12) incubation conditions for microbiological EM samples are not addressed.
Recent studies performed by Symonds et al. (13) and Gordon et al. (14) on incubation regimes of EM as well as Pharmaceutical and Healthcare Sciences Society (PHSS) monograph 20 (11) and the 2012 A3P survey (9) showed that:
– the highest quantitative results were obtained when using two samples at the same location, one incubated at 20–25 °C and the other at 30–35 °C;
– there is no significant difference between tryptic soy agar (TSA) and Sabouraud dextrose agar (SDA) for the recovery of molds even if SDA was found to be slightly superior (14);
– only one sample incubated at two successive temperatures, either 20–25 °C and then 30–35 °C or the reverse, was found to be less effective;
– even if laboratory strains of molds can grow at different temperatures, including 30–35 °C, at least some environmental strains of molds in primary isolation prefer 20–25 °C.
The A3P Annex 1 Common Interest Group would like to include in its Annex 1 modification proposals (15) recommendations for the incubation program of EM and APS, and this strategy was approved during the roundtable of the 2014 international meeting (16). In order to base these recommendations on experimental data, an in situ multicenter study was proposed for the incubation regimes both for EM with solid culture media (EM study) and for APS with a liquid medium (settle tube study = ST study). In addition, the presence of strict anaerobic bacteria in pharmaceutical manufacturing environments was assessed with a liquid thioglycolate medium.
Material and Methods
1. EM Study
Media:
Media used provided by bioMérieux with conform growth promotion tests were 90 mm TSA settle plates (SPs) bioMérieux and 55 mm TSA contact plates (CPs) containing four neutralizing agents: lecithin, Tween 80, sodium thiosulphate, and L-histidine).
Incubation Conditions:
– temperature a: 22.5 °C (20 to 25°C) for 5 days.
– temperature b: 32.5 °C (30 to 35 °C) for 5 days.
– temperatures c: 22.5 °C for 3 days then 32.5 °C for 2 days for a total of 5 days.
– temperatures d: 32.5 °C for 2 days then 22.5 °C for 3 days for a total of 5 days.
A first reading L1 was made after 2 or 3 days at 32.5 °C and 22.5 °C, including plates incubated only at one temperature.
Samples:
– 10 sequences of 3 samples were made in quadruple replicates in order to incubate each sample at 4 incubation regimes, thus 12 samples were taken for SPs or CPs in each sequence.
– SPs were exposed during exactly 4 h in Grade C, D, or controlled non-classified (CNC) rooms and supportive zones mid-shift during normal operating conditions.
– CPs were applied with a bioMérieux device during 10 s with a 500 g pressure in Grade C, D, or CNC rooms and supportive zones on walls, surfaces, or floors.
Recommendations were made to arrange the samples in a pattern designed to minimize any bias that their location would have on the capture of environmental microorganisms as shown by Symonds et al. (13). For this, each sampling sequence grouped 3 samples in quadruplicate with a total of 12 plates arranged in 4 columns of 3 plates, and the opening and closing of plates was performed in order to avoid that any part of the operator could be over an open SP.
Data:
Any bacterial colony-forming units (CFUs) detected at the first reading L1 or at the end of the incubation period, second reading, L2, were categorized as Bacillus spp., Gram-negative (B–) or non-spore-forming Gram-positive rods (B+) rods, Gram-negative (C–) or Gram-positive (C+) cocci, molds, or yeasts. All data were reported as total numbers of CFUs recovered for each category with each medium SP or CP per incubation condition.
Good Manufacturing Practice Requirements:
This study was performed in addition to the routine regulatory EM and thus the results could not have been followed by investigations and corrective actions in case of nonconformities detected.
2. Settle Tube Study
Media and Samples:
Eleven millimeter diameter tubes of liquid media provided by bioMérieux with conform growth promotion tests were 4 × 30 TSB and 4 × 30 THIO. The surface exposed with these settle tubes is around 70 times less than with 90 mm SPs.
Grade C, D, or CNC rooms and supportive areas known to have some bacteria and fungi were sampled for exactly 4 h with 8 tubes at each location, 4 TSB and 4 THIO, with the same recommendations as for solid media to avoid any possible bias coming from the sampling.
Incubation Conditions:
a = 14 days minimum at 22.5 °C (20–25 °C).
b = 14 days minimum at 32.5 °C (30–35 °C).
c = 7 days minimum at 22.5 °C then a first reading followed by 7 days minimum at 32.5 °C.
d = 7 days minimum at 32.5 °C then a first reading followed by 7 days minimum at 22.5 °C.
A first reading L1 was made after 7 days of incubation, including regimes at only one temperature.
Microbial Characterization:
Microbial Characterization of Molds, yeasts, bacteria (rods or cocci, Gram-positive or Gram-negative), Bacillus, and strict anaerobes was performed.
Results
1. EM Study
The CFUs recovered for each incubation regime and each manufacturing site are given for each group of microorganisms for SPs and for CPs in Tables I and II, respectively. The overall numbers of plates with growth was 97.7% for SPs, with two sites at 100%, and only 65.4% for CPs. Gram-positive cocci (C+) were highly predominant, representing more than 80% for SPs and 87% for CPs, followed by non-spore-forming Gram-positive rods (B+), with more than 15.5% for SPs and 8% for CPs. Thus, human commensals (total C+ and B+) were found in more than 95% for SPs and 97% for CPs. Excluding Gram-negative rods (B–) found mainly at one site (139 CFUs in a total of 180 for the four sites), the other groups of microorganisms were found only in a few occasions, with less than 0.5% for each group, or never found.
Settle Plate (SP) Detailed Results in CFU Numbers per Site and Microorganism Categories
Contact Plate (CP) Detailed Results in CFU Numbers per Site and Microorganism Categories
The four manufacturing sites found a high majority of human commensals. However, important differences were observed between sites in the overall CFU numbers. For C+ the results were 477 to 1618 with SPs and 273 to 1244 with CPs, and for B+, 91 to 406 with SPs and 3 to 92 with CPs. The sites with the highest CFU numbers were not the same based on media SP or CP or on groups of bacteria C+ or B+.
Statistical Analyses for Settle Plates:
A highly significant difference was observed (Table III) between the two incubation conditions, with the most important recovery rate for 32.5 °C during 2 days compared to 22.5 °C during 3 days (first reading L1), and this was coming from B+ and C+. For the overall results for the four sites after 5 days incubation (last reading L2), the Kruskal-Wallis test did not demonstrate a significant difference between the four incubation conditions, but some differences were observed depending on sites and microorganism groups. The difference between readings L1 and L2 was not statistically significant and was observed only for C+ for two of the four sites and for 22.5 °C alone or followed by 32.5 °C.
Settle Plates First Reading L1. Statistical Differences between 22.5 °C for 3 Days and 32.5 °C for 2 Days
Statistical Analyses for Contact Plates:
For the first reading L1 (after 3 days at 22.5 °C or 2 days at 32.5 °C) a statistical difference was observed for B+ only for one site and for C+ for all sites. For the last reading L2, the Kruskal-Wallis test did not show a significant difference, at a 5% α risk, between the four incubation conditions. The difference between L1 and L2 showed a nonsignificant increase for L2 only for C+ and only for two of the four sites.
Molds and Bacillus:
A statistical analysis was difficult due to the low numbers of CFUs. However, Tables I and II show that of the 26 molds found (17 SPs + 9 CPs), 23 were detected at 22.5 °C alone or at the beginning. For 40 Bacillus detected (32 SPs + 8 CPs), no significant difference was observed between the four conditions a, b, c, or d, with a total SPs + CPs of 7, 12, 8, and 13, respectively.
2. Settle Tube Study
The overall growth was very low, with less than one of the five tubes with growth (TSB 19%, THIO 17%) even after 4 h exposure in Grade C, D, or CNC during normal operations, and statistical analyses were not possible due to these low numbers of tubes with growth. The difference after 7 days and 14 days of incubation was very low with less than 2%. Gram-positive cocci were highly predominant, representing more than 70% with THIO and more than 84% with TSB. Yeasts, molds, Gram-negative cocci, and strict anaerobes were never found and only one tube showed Bacillus. One site has recovered 75% of Gram-positive rods. Human commensals (B+ and C+) were observed in 95.9% and the growth was better at 32.5 °C alone or as the first incubation temperature for TSB as well as for THIO (Table IV), with 63.8% together, around two times better than at 22.5 °C alone or as the first incubation temperature. The growth of these human commensals was almost equivalent between TSB (50.3%) and THIO (49.7%).
Gram-Positive Rods (B+) without Bacillus and Gram-Positive Cocci (C+) Growth Depending on Incubation Conditions and Liquid Medium
Discussion
There is no consensus regarding the incubation conditions used for EM of cleanrooms (6⇓–8, 10, 11) and for the APS (3⇓–5). This multicenter study was performed in order to try to propose a strategy for the determination of the best incubation condition in a given environment or for specific processes while also minimizing the possible bias coming from a single-site in situ study. The two recent studies (13, 14) performed with in situ samples did not give rise to the same conclusion for incubation conditions at two temperatures because there was no difference in incubation conditions at a single temperature for Gordon et al. (14) but was found less appropriate for Symonds et al. (13). Our study could not observe a statistically significant difference between the four incubation conditions used at the end of the 5 days of incubation for the total numbers of CFUs, according to Gordon et al. (14).
As for these two previous studies (13, 14), for bacteria of human origin (non-spore-forming Gram-positive rods and Gram-positive cocci) the incubation conditions at 32.5 °C allow for better growths after the first reading after 2 or 3 days for solid media (SPs and CPs) and also for liquid media (TSB and THIO) after 7 or 14 days of incubation (Table IV). This confirms that in liquid media as well as in solid media human commensals are better recovered after incubation at 32.5 °C, as suggested by Symonds et al. (13). These human commensals were highly predominant and this was the case for the four manufacturing sites according to previous studies, and this was regarding the important operator activities. Consequently, except for one site with Gram-negative rods in a washing room, the other categories of microorganisms were not found, such as yeasts and strict anaerobes, or only in very few occasions, such as Gram-negative cocci, molds, or Bacillus.
Our multicenter study confirmed also that the primary isolations from manufacturing environments of molds are never obtained after incubation at 32.5 °C alone but only after 22.5 °C alone or at the beginning, and this deficiency could not be rectified by subsequent incubation at 20–25 °C, as shown previously (14). As already observed by Horn et al. (10) and Gordon et al. (14), a first incubation period for only 2 or 3 days at 32.5 °C is able to inhibit further growth for a great majority of mold strains in primary isolation. This inhibited growth is also observed in our settle tube study for human commensals because we observed more than two times the growth at 32.5 °C (57 tubes) than at 22.5 °C, followed by 32.5 °C (27 tubes), and thus the PIC/S recommendation (3) to incubate the TSB-filled units first 7 days at 20–25 °C followed by 7 days at 30–35 °C is not appropriate for these environments and processes.
Different media (13, 18, 19) and temperatures (13, 14, 17) were evaluated for the recovery of yeast and molds and, because the results are still not conclusive, supplementary in situ studies are needed to determine the optimal conditions for the primary isolation of fungi in pharmaceutical environments. If a temperature between 20–25 °C and 30–35 °C could be recommended, the impact on the incubation temperatures of the sterility tests should be discussed.
For Bacillus our study showed only a trend toward better growth at 32.5 °C alone or at the beginning, although for Symonds et al. (13), 20–25 °C was better at least in the first part of their study.
It should be noted that the total numbers of growth are linked with the diameter of the exposure or contact because for 90 mm SPs almost 100% was obtained instead of only 66% for 55 mm CPs, and less than 20% was obtained with 11 mm diameter liquid media tubes, the exposed surface of which being around 60 times less than for 90 mm SPs. As expected, the manufacturing operations with the highest risk of microbiological contamination are operations including large open manual phases, and these operations must be performed in barrier systems (isolators or closed restricted access barrier systems in operations) or closed systems.
Finally, our study confirmed the suitability of the sterility test (2) because for human commensals the incubation temperature of 32.5 °C is important and not related to the culture medium used, TSB or THIO.
Conclusion
Recommendation for the Choice of an Incubation Regime for EM and APS
The EM of cleanrooms and the APS are looking for the same microbial contaminants and thus it seems justified to use the same incubation condition for the culture media engaged for their recovery. As for the four manufacturing sites in our study, the cleanrooms with important human activities should have a highly predominant human commensals–based flora, which should be better recovered after incubations of TSA or TSB at 30–35 °C. During the initial qualification of these cleanrooms, and periodically or after important changes or at risk events, the other categories of microorganisms such as yeasts, molds, or Bacillus should be confirmed as either not recovered or recovered only in few occasions with a second medium incubated at 20–25 °C. Thus, in contrast with Symonds et al. (13), incubation conditions at two different temperatures, 20–25 °C followed by 30–35 °C with TSA settle and CPs for EM, could be possible in such environments. However, for consistency with TSB-filled units for APS, a single incubation condition at 32.5 °C for both EM and APS could be recommended.
Conflict of Interest Declaration
The authors declare that they have no competing interests.
Acknowledgments
We thank A3P Services and the board of directors of the A3P Association for supporting our multi-center study.
- © PDA, Inc. 2017
