ASEPTIC PROCESSING: TWO - Risk Management Applications in Pharmaceutical and Biopharmaceutical Manufacturing - page 260

to product risk assessment is of paramount importance. Disappointingly, uti-
lizing common culturing technology and culture media only 10-50% of the
human microflora (depending upon anatomical location) is recoverable by growth
[43,44]. In contrast to nonculturability (because of suboptimal growth conditions),
microbial dormancy is part of an essentially ordered developmental program such
as sporulation [54,55]. In gram-positive microorganisms, there is evidence that
the coordinated flux in metabolic and biochemical functions suggests dormancy is
active and programmed [56]. Recent data have illustrated that this phenomenon is
a characteristic of clinically relevant species of microorganisms. Sachidanandham
and Gin [57] revealed that dormancy is a mechanistically ordered process per-
mitting Klebsiella pneumonia , Escherichia coli ,and Enterobacter sp . to endure
adverse environmental conditions. In addition to human-borne microorganisms,
the recovery and growth of microorganisms from clean room and aseptic process-
ing environments are crucial considerations in the evaluation and quantification
of risk. La Duc [45] has illustrated that the magnitude of the microbial hazard
with the potential to contaminate a product processed within clean rooms and
aseptic conditions may be far larger than previously considered. Indeed, when
we consider that dormancy is a preferentially opted physiological state, adopted
by microorganisms to endure suboptimal conditions, common to species of the
human micro flora, and species causing infections, we must question the ade-
quacy of our reliance on growth-based technology platforms for sterility testing
and environmental monitoring, and risk evaluation.
Bryce [58] recognized the fundamental and unavoidable constraint of sample
size that limits the sterility testing of finished articles to solely a determination of
batches “sterility.” Furthermore, for any batch contamination event, a relationship
exists between the mean number of microorganisms per unit and the frequency of
units within a batch containing at least one microorganism—see Tables 10.7 and
10.8 [59]. The relative frequency can be calculated using the following equation:
Q = 100 1 − e − m

where:

Q = percentage of units containing at least one microorganism
e = 2.7182818
m = the average number of microorganisms per unit
Applying this calculation, if one batch of a product contains on average one
microorganism per individual unit (i.e., vial, container, or device), then only 63%
of all units within that batch are likely to contain at least one microorganism.
With sampling a finite ( n = 20) number of units used in any testing strategy, there

remains a significant risk of failing to identify a nonsterile unit; no refinement or
optimization of sample size, sample choice, or frequency provides a satisfactory
or adequate level of sterility assurance by end product sterility testing. The risk of
reporting a false negative (Type 2 error), that is, the failure to identify a nonsterile
unit by finished product testing is calculable from the following equation [60]: