ASEPTIC PROCESSING: TWO - Risk Management Applications in Pharmaceutical and Biopharmaceutical Manufacturing - page 263

at each process step and populating the associated risk components and risk
factors with data that truly quantitatively describes the probability of microbial
ingress realize the Q8-Q9-Q10 paradigm. Tidswell and McGarvy [19,27] inno-
vated, developed, and successfully applied this strategy to quantitatively describe
parenteral product sterility during aseptic manufacturing. Successful exploita-
tion of this strategy is based on three fundamental tenets. First, that a single
microorganism accessing an aseptically manufactured or manipulated product is
unacceptable. The notion that species type per se is not imperative to patient
infection and therefore salient to nonsterility is legitimate; fundamentally, any
microorganism has the potential to cause patient harm and is salient to assess-
ment of patient risk. Although a recent compelling discourse has pragmatically
evaluated the risk from a limited number of microorganisms present within asepti-
cally manufactured products [64], this contradicts the fundamental aim of aseptic
manufacture—the generation of products devoid of microorganisms. Ironically,
we continually seed our own bloodstream with microorganisms originating from
our own inherent microflora and accessing at distinct locations. For example, the
oral and buccal cavity contains large numbers of microorganisms accessing our
bloodstream via small lesions and imperfections in our gums and oral mucosa.
Any attempt to justify a low level of microorganisms within a parenteral prod-
uct must be founded upon a proven clinical rationale, highly impractical and
uneconomical to attempt. As previously described in terms of assessing risk to
an aseptically manufactured product from the ingress of microorganisms, the
fundamental risk is described by the probability of ingress and does not include
evaluation of severity as a consequence of the realization of risk (Eq. 10.2).
The second fundamental tenet is that measurement uncertainty exists in the
enumeration of microorganisms; this is especially the case when growth-based
technologies are implicit in the sterility assurance program. It is appropriate,
however, to account for this uncertainty and variability by the description of
the magnitude of the microbial hazard by probabilistic population distributions.
What is important to know is the maximum and minimum values possible, and
the likely value. These describe the spread, shape, and likely bounds of any
quantity of microbial hazard present. Traditionally, uncertainties implicit in the
logic chain of quantitative microbial risk assessment methodologies have been
accounted for using Monte Carlo simulations [65]; however, Bayesian belief
networks may represent more sophisticated means of accounting for stochasticity
[66]. These tools permit a systematic and repeated generation of random values
for the microbial risk factor variables to populate probability distributions. In this
manner, all possible permutations are evaluated to permit a thorough evaluation
of risk, rather than rely on a single value that is likely inaccurate, somewhat
subjective, and may bias the determination of risk. Simultaneous interpolation of
all contributing permutations far exceeds our own mental acuity and does demand
the use of software to generate the quantitative assessment of risk.
Finally, the third fundamental tenet is that the transfer of microorganisms
from a source or location and any consequential ingress into an aseptically
manufactured product can be defined, measured, and validated. Any microbial