Increased attention to microbiological contamination forces a greater focus on air sampling technology
By Christopher Mach, Pall Life Sciences
Improvements in environmental air monitoring technology are becoming increasingly important to the pharmaceutical industry. The latest ISO 14698 standards have served to underline the increased focus on air monitoring being shown by regulatory bodies around the world. Current monitoring technologies do not, for the most part, meet these challenges because their efficiency-the number of microbial particles they capture compared to those that are actually present-isn’t high enough to meet the new ISO standards. Another concern with existing technology is its inability to distinguish between contaminants captured from cleanroom air and those from externally introduced sampling. These increasing requirements have intensified the attention paid to advancements in air sampling technology. This article will outline some of the basic issues involved.
Increased focus on air monitoring
The increasing emphasis on environmental air monitoring in aseptic processing areas was apparent even before the British Medicines and Healthcare products Regulatory Agency (MHRA) suspended manufacturing at Chiron’s plant in Liverpool, England, for five months in 2004 and 2005. Chiron had been expected to produce half of the United States’ 100-million-dose flu vaccine supply for the 2004-05 season but the MHRA suspended the plant’s manufacturing license because of contamination issues. Inspectors found bacterial contamination of vaccine batches, microbial growths on equipment used to fill vaccine vials, high levels of bacterial toxins in vaccine preparations and contaminant in viral seed cultures, among other issues. Chiron’s shares fell more than 30 percent with news of the plant suspension and its 2004 earnings fell by half from the previous year.
Similar concerns have caused regulators around the world to focus increasing amounts of attention on ensuring that airborne microorganisms are within acceptable levels. Tight control over the cleanroom used for aseptic processing of pharmaceutical products is the primary safeguard against contamination. Nearly as important is the ability to evaluate bacterial contamination associated with airborne particles so that manufacturing personnel can promptly intervene in the event that values outside acceptable levels are detected. The increased awareness of microbiological contamination has forced a greater focus than ever before on bioaerosol sampling technology designed to accurately quantify the total number of viable microorganisms per unit volume of air.
The attention being paid to air sampling is further amplified by the Food and Drug Administration’s Process Analytical Technology (PAT) initiative, which aims to spearhead a shift from reliance on final product quality control to in-process control at key points in the manufacturing process. The ability to monitor quality and identify microbial contamination events during processing enables companies to diagnose root causes and take preventive action quickly in order to ensure product safety and avoid economic losses.
Limitations of conventional methods
As pharmaceutical manufacturers have begun focusing more attention in this area, the limitations of conventional bioaerosol sampling methods have become more apparent. The basic challenge of bioaerosol sampling is the trade-off that is required between particle collection efficiency, which requires moving microorganisms at high speed into the sampler, and avoiding severely stressing the microorganisms during the sampling process, which can conceal air contamination. This trade-off has become more challenging than ever since ISO released the 14698-1:2003 standard, which establishes tough sampling efficiency requirements that cannot be achieved by the vast majority of existing bioaerosol samplers. Another critical concern is the ability of the sampler to distinguish between contamination of the cleanroom air and random contamination from other sources such as the assay media. Finally, the critical locations in which samplers are placed lead to the requirement that they must not contaminate the environment.
The new ISO standard highlights the limitations of traditional sedimentation technology, which, because of its simplicity, has long been the most popular environmental air sampling method. Sedimentation sampling relies on the force of gravity and air currents to cause airborne microorganisms to settle onto plates filled with media. After exposure, the plates are incubated. The problem with this method is that its efficiency is low and it is particularly inefficient at capturing smaller particles, which constitute a relatively high proportion of microbiological contaminants. For these reasons, Annex A.3.3 of the ISO standard states that settle plates should not be used to attempt to measure the total number of viable particles in the air but should be limited to measuring the rate at which particles settle on surfaces. The standard explicitly describes the use of active sampling methods in risk zones as essential to the accurate measurement of microbiological air quality.
Movement towards active methods
For these reasons, there has been steady movement in the industry towards active sampling methods. Impingement methods entrap airborne microorganisms in a liquid medium as the air is transmitted through the fluid. Centrifugal samplers spin an aerosol at a high rate of speed and rely on centrifugal force to propel particles against the settling plate. Filtration samplers collect airborne microorganisms by impelling air against a filter, such as gelatin membranes or cellulose fibers, that can then be placed onto an agar surface for quantification. Impaction samplers utilize a vacuum to draw microorganisms onto an agar surface. All active samplers must overcome the difficult trade-off between efficiency and overstressing organisms to comply with ISO standards. High air velocity runs the risk of overstressing microorganisms to the point they will not be detectable, while low air velocity reduces collection efficiency.
But even the vast majority of active air samplers are incapable of meeting the ISO standards, which enumerate several important factors to consider in selection of an air sampler. The effective sampling rate of the instrument, duration of sample acquisition and physical attributes of the sampling device all have the ability to strongly influence the viability of the microorganisms that are collected. Since there are many microbial air sampling systems available on the market today, ISO 14698-1 recommends users consider, as a minimum:
■ type and size of viable particles to be sampled
■ sensitivity of the viable particles to the sampling procedure
■ expected concentration of viable particles
■ capability of detecting high or low levels of biocontamination
■ appropriate culture media
■ time and duration of sampling
■ ambient conditions in the environment being sampled
■ disturbance of unidirectional airflow by the sampling apparatus
How active samplers work
Understanding these requirements demands an examination of the complicated physical forces that affect the motion of microbial particles in air. The force of gravity causes particles to accelerate based on their weight. Brownian motion causes random movements. Bioparticles have an electrical charge that influences their velocity and is strongly influenced by humidity. Vapors are absorbed on the surface of biological particles, which may cause coagulation of proteins. Acoustic forces may exert accelerations depending on their frequency and intensity. Adhesion forces are also important in determining particle behavior. For particles of 10 microns and below, Van der Waals forces are dominant. However, as particle size approaches 100 microns, gravity becomes the dominant force.
Inertial impact samplers have benefited from some of the most significant technical advances made in recent years. An inertial sampler is based upon the principle that when a stream of gas undergoes a sharp change in direction, the particles it transports will tend to continue in their original direction to a degree that is proportional to the ratio of their mass to their linear dimensions. Particles that have different dimensions and densities will follow different trajectories and can be collected separately. Inertial impact samplers capture particles by accelerating a jet of air in a nozzle. The particles transported in the jet are carried at the same speed as the fluid and follow its flow lines. Then, at the nozzle output, the fluid flow lines rapidly change direction, while particles tend to run in a straight line-into a surface to which they adhere-and are captured.
The impact efficiency (in other words, the probability that the particle will be captured by the impact plane) depends upon the particle linear dimensions, particle speed at the acceleration nozzle output, air viscosity, impactor geometry, and adhesion of the particle to the impact plane. Particle efficiency can be expressed in terms of a dimensionless parameter called the Stokes number, which takes into account the physical laws that govern the motion of the particles moving in a fluid under laminar flow conditions. These methods make it possible to calculate the efficiency of a given impactor design.
Stokes = —- C
C = pressure dependent correction factor (atmospheric pressure C=1)
Φp = particle density
Dp = particle equivalent diameter
V= particle speed at nozzle output
μ= air viscosity
W= slot width for rectangular slot impactors
Improving sampler design
There are two potential problems that must be considered in the design of the impact surfaces. If the particles do not adhere well to the impact surface, they can be resuspended in the fluid owing to the motion of the air. On the other hand, if the impact with the impact plane is elastic, particles may bounce off it and fall back into the fluid or break up. In order to avoid these potential problems, the impact surface must be able to absorb the particle’s kinetic energy and the kinetic energy imparted to the particle must be minimized.
One new design approach is to use laser technology to form rectangular slits that serve as nozzles to improve collection efficiency while avoiding severe stress to microorganisms. The slits also create a recognizable pattern of microorganisms on the assay media surface, which makes it possible to distinguish between microbial contamination in the clean air and particles introduced from other sources (see Fig. 2).
Evaluating sampler performance
New-design test samplers operating at 25 and 50 liters per minute were evaluated for physical and biological efficiency using the techniques described in ISO 14698. A conventional slit sampler was operated at 30 liters per minute during the biological and background efficiency testing for comparison purposes. Physical efficiency measures the sampler’s ability to collect various sizes of particles regardless of their nature. Biological efficiency goes one step further by measuring the ability of the sampler to capture viable particles-that is, particles that contains one or more living microorganisms-by taking into account the survival of the microorganisms during the collection process and the ability of the collection medium to support their growth. Viable particles generally range in size from 0.2 micron to 30 microns.
Figure 3. Settle plate exposed to ambient on left; plate used with the new-design impact sampling system on right.
The physical efficiency of the sampler was measured through the particle range of 1 to 10 microns by creating aerosols of an aerostable bacterial spore Bacillus subtilis var. niger. The size of the spores containing particles was varied by aerosolizing them from a range of concentrations of potassium iodide solution. Physical efficiency testing was carried out in a cleanroom with 28 cubic meters of volume, supplied with a horizontal flow of clean air through HEPA filter banks. During aerosol sampling, the ventilation was switched off. A spinning top aerosol generator (STAG) was used to generate aerosol drops. A peristaltic pump was used to supply the STAG with the spray suspension at a flow rate of 0.5 mL/min. The STAG was placed above a standard room ventilation fan to allow dissemination of the aerosol. The samplers were placed 1.0 meter from the STAG next to the two filter samplers. The microbial aerosol was generated in on/off pulses of ten seconds each while the samplers were operated.
Figure 4. Impacted media plate with colonies defined in the impaction plane. The circled colony represents false positive growth outside the impaction area
At the same time, a Cascade sampler was operating to measure the particle size of the aerosols used in the physical efficiency testing. The final particle size was controlled by the use of 80 percent ethanol as a solvent to allow rapid evaporation of the fluid, and potassium iodide as a bulking agent to govern the final size. A three-jet nebulizer was used to generate the mixed microbial aerosol for biological efficiency testing.
Comparing efficiency to conventional sampler
The biological efficiency of the sampler was compared to a standard slit sampler using mixed aerosols of the aerostable spore and a common air contaminant, Staphylococcus epidermidis. The ratio of the number of S. epidermidis cells collected to the aerostable spore gives an indication of the biological efficiency of the sampler. This microorganism was chosen as a test strain because, being derived from human skin cells, it is a common contaminant in cleanroom air. The biological efficiency testing was undertaken in a Class III microbiological safety cabinet with an internal volume of 0.865 cubic meters. The cabinet generates six air changes per minute when the fan unit is operated. The spray suspension was aerosolized for 30 seconds and then the samplers were operated for one minute. After 24 hours of incubation, the colonies of S. epidermidis and B. subtilis var niger were counted individually.
In the physical efficiency tests, the 50 LPM model was shown to have a d50 (smallest particle size of which it can capture 50 percent or more) of less than 1 micron. The 25 LPM model had a d50 value of less than 2 microns. The 50 LPM sampler collected 79 percent of spores aerosolized from 0.0 percent potassium iodide compared to 54 percent reported by Benbough et al. (1993) for an earlier model. The submicron d50 values of the 50 LPM sampler are considerably lower than d50 values reported in the past (Lach, 1985). The 50 LPM sampler had an average biological efficiency of 95.3 percent and 82.4 percent for a one-minute and 20-minute sampling period respectively, which is higher than a conventional slit sampler. The total net impact surface is less than 1 percent of the total plate surface. This implies that false positives can be excluded with a statistical probability exceeding 99 percent.
The efficiency and reliability of environmental air monitoring systems need to increase in order to address the requirements of pharmaceutical manufacturers as they move to meet the increasing scrutiny of regulatory agencies, as well as the challenges of the new ISO standard. Advanced air monitoring technologies are the key to gaining tighter control over manufacturing processes and providing greater assurance of product safety. The performance of microbiological air samplers plays an important role in the environmental monitoring program. The latest ISO 14698 standards provide a needed impetus for improvements by providing techniques for the evaluation and qualification of the efficiency of air samplers.
1. International Organization for Standardization. ISO 14698-1:200i, Cleanrooms and associated controlled environments-Biocontamination control.
2. International Organization for Standardization. ISO 14698-2:2003, Evaluation and interpretation of biocontamination data.
3. Benbough, J.E., A.M. Bennett, and S.R. Parks. “Determination of the collection efficiency of a microbial sampler,” Journal of Applied Bacteriology, 74: 170-173, 1993.
4. Lach, V. “Performance of the Surface Air System air samplers,” Journal of Hospital Infection, Vol. 6, pp. 102-107, 1985.
Christopher Mach is a biotechnology marketing manager for Pall Life Sciences. He can be reached at [email protected]