Quality Control Procedures

QUALITY CONTROL PROCEDURES

While there is general agreement on the need to control total microbial levels in non-sterile medicines and to exclude certain species that have previously proved troublesome, the precision and accuracy of current methods for counting (or even detecting) some microbes in complex products are poor. Pathogens, present in low numbers, and often damaged by processing, can be very difficult to isolate. Products showing active spoilage can yield surprisingly low viable counts on testing. Although present in high numbers, a particular organism may be neither pathogenic nor the primary spoilage agent, but may be relatively inert, e.g. un-germinated spores or a secondary contaminant which has outgrown the initiating spoiler. Unevenly distributed growth in viscous formulations will present serious sampling problems. The type of culture medium (even different batches of the same medium) and conditions of recovery and incubation may significantly influence any viable counts obtained from products.

An unresolved problem concerns the timing of sampling. Low levels of pseudomonads shortly after manufacture may not constitute a spoilage hazard if their growth is checked. However, if unchecked, high levels may well initiate spoilage.

The European Pharmacopoeia has introduced both quantitative and qualitative microbial standards for nonsterile medicines, which may become enforceable in some member states. It prescribes varying maximum total microbial levels and exclusion of particular species according to the routes of administration. The British Pharmacopoeia has now included these tests, but suggests that they should be used to assist in validating GPMP processing procedures and not as conformance standards for routine end-product testing. Thus, for a medicine to be administered orally, the total viable count (TVC) should not be more than 10³ aerobic bacteria or 10² fungi per gram or millilitre of product, and there should be an absence of Escherichia coli. Higher levels may be permissible if the product contains raw materials of natural origin, as in the case of herbal products where the TVC should not exceed 10⁵ aerobic bacteria, 10⁴ fungi and 10³ Enterobacteria and Gram-negatives, with the absence of E.coli/gram or millilitre and Salmonella/ 10 gram or millilitres.

Most manufacturers perform periodic tests on their products for total microbial counts and the presence of known problem microorganisms; generally these are used for in-house confirmation of the continuing efficiency of their cGPMP systems, rather than as conventional end-product conformance tests. Fluctuation in values, or the appearance of specific and unusual species, can warn of defects in procedures and impending problems.

In order to reduce the costs of testing and shorten quarantine periods, there is considerable interest in automated alternatives to conventional test methods for the detection and determination of microorganisms.

Although not in widespread use at present, promising methods include electrical impedance, use of fluorescent dyes and epifluorescence, and the use of ‘vital’ stains. Considerable advances in the sensitivity of methods for estimating microbial ATP using luciferase now allow the estimation of extremely low bioburdens. The recent development of highly sensitive laser scanning devices for detecting bacteria variously labelled with selective fluorescent probes enables the apparent detection even of single cells.

Endotoxin (pyrogen) levels in parenteral and similar products must be extremely low in order to prevent serious endotoxic shock on administration. Formerly, this was checked by injecting rabbits and noting any febrile response. Most determinations are now performed using the Limulus test in which an amoebocyte lysate from the horseshoe crab (Limulus polyphemus) reacts specifically with microbial lipopolysaccharides to give a gel and opacity even at very high dilutions. A variant of the test using a chromogenic substrate gives a coloured end point that can be detected spectroscopically. Tissue culture tests are under development where the ability of endotoxins to induce cytokine release is measured directly. Sophisticated and very sensitive methods have been developed in the food industry for detecting many other microbial toxins. For example, aflatoxin detection in herbal materials, seed stuffs and their oils is performed by solvent extraction, adsorption onto columns containing antibodies selective for the toxin, and detection by exposure to ultraviolet light.

Although it would be unusual to test for signs of active physico-chemical or chemical spoilage of products as part of routine product QC procedures, this may occasionally be necessary in order to examine an incident of anticipated product failure, or during formulation development. Many of the volatile and unpleasant-tasting metabolites generated during active spoilage are readily apparent. Their characterization by high performance liquid chromatography or gas chromatography can be used to distinguish microbial spoilage from other, non-biological deterioration. Spoilage often results in physico-chemical changes which can be monitored by conventional methods. Thus, emulsion spoilage may be followed by monitoring changes in creaming rates, pH changes, particle sedimentation and viscosity.