FTIR for Estimating Viable Counts in Beef

I found a very interesting paper describing the use of FTIR to estimate total viable counts on the surface of beef. The paper was published in Food Microbiology and the full reference and abstract are provided below.

A comparison of artificial neural networks and partial least squares modelling for the rapid detection of the microbial spoilage of beef fillets based on Fourier transform infrared spectral fingerprints. Panagou EZ, Mohareb FR, Argyri AA, Bessant CM, Nychas GJ. Agricultural University of Athens, Department of Food Science, Technology and Human Nutrition, Laboratory of Microbiology and Biotechnology of Foods, Iera Odos 75, 118 55 Athens, Greece. Food Microbiol. 2011 Jun; 28(4): 782-790.

A series of partial least squares (PLS) models were employed to correlate spectral data from FTIR analysis with beef fillet spoilage during aerobic storage at different temperatures (0, 5, 10, 15, and 20 °C) using the dataset presented by Argyri et al. (2010). The performance of the PLS models was compared with a three-layer feed-forward artificial neural network (ANN) developed using the same dataset. FTIR spectra were collected from the surface of meat samples in parallel with microbiological analyses to enumerate total viable counts. Sensory evaluation was based on a three-point hedonic scale classifying meat samples as fresh, semi-fresh, and spoiled. The purpose of the modelling approach employed in this work was to classify beef samples in the respective quality class as well as to predict their total viable counts directly from FTIR spectra. The results obtained demonstrated that both approaches showed good performance in discriminating meat samples in one of the three predefined sensory classes. The PLS classification models showed performances ranging from 72.0 to 98.2% using the training dataset, and from 63.1 to 94.7% using independent testing dataset. The ANN classification model performed equally well in discriminating meat samples, with correct classification rates from 98.2 to 100% and 63.1 to 73.7% in the train and test sessions, respectively. PLS and ANN approaches were also applied to create models for the prediction of microbial counts. The performance of these was based on graphical plots and statistical indices (bias factor, accuracy factor, root mean square error). Furthermore, results demonstrated reasonably good correlation of total viable counts on meat surface with FTIR spectral data with PLS models presenting better performance indices compared to ANN.

Latest RMM papers from European Pharmaceutical Review

The current issue of European Pharmaceutical Review contains two new papers focused on rapid methods. I have published the second of six articles on RMM technologies, and the team at Novartis have provided an overview of using RMMs for microbiological quality control. Below please find introductory excerpts from both papers; however, you will need to subscribe to EPR (either online or in print) to access the entire text.

Overview of rapid microbiological methods evaluated, validated and implemented for microbiological quality control.

Oliver Gordon, Jennifer C. Gray, Hans-Joachim Anders, Alexandra Staerk & Oliver Schlaefli, Novartis Pharma Stein AG and Gunther Neuhaus, University of Freiburg.

The risk for patients through spoiled or otherwise adulterated pharmaceuticals has been acknowledged for many centuries and led to the establishment of Good Manufacturing Practice (GMP) and pharmacopoeial guidelines. Besides chemical purity, pharmaceuticals also have to meet microbiological standards, the latter primarily depending on the administration route. Drug products which are injected directly into blood vessels or tissues or that are applied directly into eyes and ears represent a greater infection risk than products which are administered orally or onto intact healthy skin. While parenteral drug products are required to be free from any viable microorganism (USP <71>, Ph. Eur. 2.6.1), oral and topical products are not required to be sterile, but are subject to strict guidelines limiting the number and types of acceptable microorganisms (USP <61> and <62>, Ph. Eur. 2.6.12 and 2.6.13).

Article 2: Direct detection of microorganisms using viability-based technologies.

Michael J. Miller, President, Microbiology Consultants, LLC

This is the second in a series of articles on rapid microbiological methods that will appear in European Pharmaceutical Review during 2011. In my last article, I provided an overview of growth-based rapid microbiological methods (RMMs). This was a good place to start my review of RMM technologies, as most of us continue to use conventional agar and liquid medium for the growth of micro – organisms. In the current article, I will significantly depart from growing microorganisms to the direct detection of microorganisms using viability-based technologies, which will include flow cytometry and solid phase cytometry.

Michael Miller Discusses False Positive Testing and Highlights a Case Study Using an Environmental Monitoring RMM

During the final rapid method session at this year's PDA Annual meeting, I provided an overview of the compendial, regulatory and technical reasons why interference testing, including false positive and false negative assessments, should be conducted during the validation of RMMs. False positives occur when a positive detection signal and/or apparent microbial count is obtained when viable microorganisms are not actually present. This can occur for different types of rapid methods but can be more apparent when using a technology that does not rely on microbial growth. This may result in an overestimated of viable counts. Alternatively, a false negative occurs when no detection signal and/or microbial count is observed when viable microorganisms are actually present. This may occur, for example, when test sample components interfere with the detection and/or counting or microorganisms. In this case, an underestimation of viable counts can occur.

I finished my presentation with an overview of a case study on false positive testing using BioVigilant's IMD-A, a real-time environmental monitoring technology that simultaneously detects, sizes and enumerates both total particles and viable particles in the air. Cleanroom and sterility test materials, such as sterile gowns, paper, Tryptic soy broth and sterile silicone, were evaluated for their ability to produce a false positive response under certain conditions. Recommendations for reducing or eliminating the potential for false positives when using these materials were also provided.

Claudio Denoya Discusses the Pall GeneDisc during the PDA Annual Meeting

Claudio Denoya (Pfizer) provided an overview of how his team is validating the Pall GeneDisc rapid microbiology system. Specificity testing was conducted using purified DNA from 6 different organisms represented by the system's Specified Microorganism detection plate, and excellent linearity was observed. They also evaluated the system's ability to detect the presence of the target DNA in the presence of a variety of excipients, such as magnesium stearate, mannitol and propylene glycol, and demonstrated very good linearity and a low limit of detection during these studies. Ruggedness was also assessed using different lots of reagents and multiple operators, and robustness was evaluated using frozen filters and nucleic acids held overnight under refrigeration.

Validation using stressed organisms and statistics discussed at the PDA Annual Meeting

During the 2011 Annual PDA meeting in San Antonio, two sessions were dedicated to rapid methods. The first session focused on the advances and challenges for rapid methods. Sven Deutschmann (Roche Diagnostics) discussed the validation of the Rapid Micro Biosystems Growth Direct system for in-process control and water testing. As part of their validation strategy, they challenged the system with slow growing organisms and determined that a 33 hour incubation was suitable for the detection of microorganisms. Using eight different types of microorganisms, accuracy data between the rapid system and conventional plates was greater than 70%, and the precision of the RMM was within the expected acceptance criteria as recommended in USP 1223. Next, they used stressed organisms to challenge the system in a similar fashion. Three organisms were chosen and they incubated these at 60 degrees C until the viability of each was reduced by 50% as compared with untreated organisms. These stressed organisms were also shown to be detected in the RMM within 33 hours. A separate set of experiments were conducted with pH-stressed organisms. They exposed the organisms to a pH 3 environment and determined the exposure time that gave at least a 90% reduction in viability was sufficient to generate stressed organisms for the same studies. These organisms were then tested as previously described and it was also determined that a 33 hour incubation would detect these stressed organisms as well.

Julie Schwedock (Rapid Micro Biosystems) then discussed the use of statistics that have been recommended in the USP, Ph. Eur., and PDA TR33. She provided a very understandable review of current expectations when using statistics as well as their limitations, and offered alternative statistical models for use depending on the expected bacterial population and the validation criteria that was being evaluated. A discussion of the use of the T test and the F test was provided, and actual data was presented that demonstrated the current compendial methods may not actually meet the statistical requirements of the compendia themselves! This was a very informative presentation and the audience applauded Julie's ability to discuss statistics in a manner that was understandable by all who attended.

Rapid Methods at the PDA Annual Meeting

Once again, I will be blogging from the PDA annual meeting next week from San Antonio. There are only a few rapid method presentations this year but I will be blogging live to bring you the latest in what's being discussed!