Environmental monitoring using a rapid nondestructive automated compendial method Andrew Sage, Principle Scientist Rapid Micro Biosystems New England PDA 16 May, 2008 1
Overview overview of the automated compendial rapid microbial enumeration technology- the Growth Direct system application to environmental testing in manufacturing facilities: water air surface 2
The business problem: high cost of culture-based QC microbiological testing in pharmaceutical manufacturing $$$$ cost of materials regulatory risk: gold standard skills required sensitivity (for culturable bugs) time to results cost of labor cost of held inventory cost of product scrap cost of plant downtime cost of cleanup 3
Goals in automating the compendial method Improve accuracy & decrease time-to-results replace human eye with digital imaging facilitate system validation use same procedures and method principles as traditional culture save labor & improve compliance automate analysis and documentation 4
Automating the compendial method by replacing the human eye with sensitive digital imaging- a better set of eyes Growth Direct system visual plate counting ~100 cells ~5x10 6 cells 5
Using large area non-magnified digital imaging to detect microscopic microcolonies lamp photosensitive pixel CCD chip autofluorescing microcolony 6
How the image analysis software enumerates growing microbes 7
Image analysis using Growth Direct software make a stack of images from the various time points 8
Image analysis using Growth Direct software find objects on each image using image analysis software 9
Image analysis using Growth Direct software align images 10
Image analysis using Growth Direct software trace all objects backwards through time 11
Image analysis using Growth Direct software identify growing objects (intensity increases over time) 12
Image analysis using Growth Direct software ignore debris (objects that do not grow over time) 13
Image analysis using Growth Direct software 5 growing microcolonies report number of growing objects 14
Accuracy: by analyzing image time series system counts growing colonies and ignores inanimate fluorescent debris 0.5 mm 10 11 12 13 14 15 16 17 18 2 hrs 3 4 5 6 7 8 9 P. aeruginosa 15
The work flow of the automated compendial test 16
Labor savings and improved compliance from an automated compendial test labor savings data acquisition is automated documentation is electronic, and easily transferred to data management systems increased compliance fewer data management errors greater reproducibility 17
Automating the compendial test preserves its advantages while addressing its weaknesses captures the positive features of the compendial tests non-destructive ultra-sensitive (1 CFU ) breadth of testing applications enumerates replicating cells high throughput no added reagents industry standard media, membranes addresses the limitations of the compendial tests automation: labor, compliance, reproducibility speed: saves days, generally ~50% faster 18
Bacteria detected by cellular autofluorescence Acidovorax delafieldii Curtobacterium sp. Proteus vulgaris Acidovorax sp. Deinococcus proteolyticus Pseudomonas aeruginosa Acidovorax temperans Dermacoccus nishinomiyaensis Pseudomonas fluorescens Acinetobacter junii Enterococcus faecalis Pseudomonas putida Afipia broomeae Escherichia coli Pseudomonas stutzeri Arthrobacter sp. Geobacillus stearothermophilus Ralstonia pickettii Bacillus cereus Hydrogenophagea sp. Rhodococcus erythropolis Bacillus clausii Hyphomicrobium sp. Roseomonas gilardii Bacillus fusiformis Kocuria kristinae Roseomonas sp. Bacillus gibsonii Kocuria rhizophila Salmonella enterica Bacillus licheniformis Kytococcus sedentarius Serratia marcesens Bacillus megaterium Macrococcus caseolyticus Sphingomonas paucimobilis Bacillus pumilus Methylobacterium extorquens Sphingomonas spp. Bacillus sp. Methylobacterium radiotolerans Sphingomonas terrae Bacillus subtilis Microbacterium luteolum Staphylococcus aureus Bacillus vortex Microbacterium maritypicum Staphylococcus capitis Bacteriodes fragilis Microbacterium sp. Staphylococcus epidermidis Brachybacterium sp. Micrococcus luteus Staphylococcus equorum Bradyrhizobium spp. Moraxella osloensis Staphylococcus haemolyticus Brevibacterium sp. Myxococcus xanthus Staphylococcus hominis Brevundimonas diminuta Neisseria sp. Staphylococcus saccharolyticus Burkholderia cepacia Paenibacillus lautus Staphylococcus sp. Caulobacter leidyii Paenibacillus sp. Staphylococcus warneri Cellulomas sp. Pantoea agglomerans Streptococcus sp. Chromobacterium violaceum Paracoccus sp. Streptomyces chrysolmalus complex Clostridium sporogenes Porphyromonas gingivalis Streptomyces coelicolor Corynebacterium sp. Prevotella melaninogenica Streptomyces sp. Corynebacterium xerosis Propionibacterium acnes Vibrio natriegens Corynebacterium pseudodiptheriticum 19
Fungi detected by cellular autofluorescence Alternaria alternata Alternaria geophila Arthrinium sacchari Aspergillus flavus Aspergillus fumigatus Aspergillus niger Aspergillus sp. Aspergillus versicolor Aureobasidium pullulans Candida albicans Candida parapsilosis Chaetomium globosum Cladosporium herbarum Epicoccum nigrum Fusarium solani Penicillium camemberti Penicillium chrysogeneum Penicillium corylophylum Penicillium notatum Penicillium roquefortii Rhizopus oligosporus Saccharomyces cerevisiae Schizophyllum commune Schizophyllum fasciatum Schizosaccharomyces pombe Sporidiobolus johnsonii Sporotrichum pruinosum Trichoderma asperellum Zygosaccharomyces rouxii 20
Time savings: the system detects microscopic microcolonies (scanning EM images) 1 micron Escherichia coli (~120 cells) Candida albicans (~12 cells) 21
The automated compendial method saves days for slow growing strains Methylobacterium extorquens Bacteroides vulgatis Mycobacterium chelonae Proionibacterium acnes Deinococcus proteolyticus Mycoplasma bovis Aspergillus versicolor Ralstonia picketii Aspergillus niger Clostridium sporogenes Growth Direct (days) 2.6 0.9 1.9 0.9 1.6 1.3 1.5 1.1 0.8 0.6 Visual (days) 17.2 7 6.7 3.6 4 3.7 3.6 3 2.4 1.8 Days saved 14.6 6.1 4.8 2.7 2.4 2.4 2.1 1.9 1.6 1.2 22
Time savings is greatest for slow growing microbes 23
Water testing 24
Rapid detection of water microbes: autofluorescent detection detects the same colonies that later become visible by eye Growth Direct microcolonies visual plate counting 2.5 days 5 days sample: purified water from a pharmaceutical facility 25
Correlation of Growth Direct and visible counts in pharma water samples Growth Direct colonies at 68 hr Growth Direct (68 hr) vs visible counts 350 300 250 200 150 100 R2A, 32.5ºC 50 y = 1.02x 0 R 2 = 0.90 0 50 100 150 200 250 300 350 visible colonies at 120 hr 26
Accuracy: resolving at the microcolony stage colonies that are uncountable by traditional visible plate visual plate counting (5 days) Growth Direct (1.5 days) 27
Air monitoring 28
Rapid detection of airborne microbes at a pharma plant Growth Direct microcolonies visual plate counting 20 hr 72 hr 29
Rapid detection of diverse airborne microbes at a pharma facility 18 hrs 39 hrs 21 hrs 12 hrs 12 hrs 72 hrs (TSA, 32.5ºC) 30
Air monitoring: co-trending of rapid (1.5 day) and traditional (3 day) tests at a pharma facility Growth Direct colonies (36 hr) 60 40 20 0 32.5 ºC, TSA y = 1.105x + 0.41 R 2 = 0.7899 4-Sep-07 19-Sep-07 2-Oct-07 15-Nov-07 0 20 40 60 visible colonies (72 hr) 31
Air monitoring: co-trending of rapid (1 day) and traditional (2 day) tests at a pharma facility 10 Naked visual agar plate (48 count hours) at 48 hr Growth Direct (18 test hours) at 20 hr colonies detected 8 6 4 2 0 2 17 3 Oct 24 4 Oct 9 Nov 5 6 Dec 6 21 Dec 7 16 8Jan 8 Mar 9 1 10 May 11 date 32
Comparing recovery on membranes and agar in air testing using a half membrane strategy 33
Various air samplers, half-membrane experiments show equivalent recovery on membrane and agar membrane (CFU) 180 160 140 120 100 80 60 40 20 0 Air Ideal (90 mm) y = 1.04x - 5.1, R 2 = 0.96 membrane (CFU) 200 150 100 50 0 Air Ideal (55 mm) y = 0.93x - 1.7, R 2 = 0.87 0 50 100 150 agar (CFU) 0 50 100 150 200 agar (CFU) SAS 180 SMA P201 membrane (CFU) 250 200 150 100 50 0 y = 0.94x - 0.93, R 2 = 0.97 0 50 100 150 200 250 agar (CFU) membrane (CFU) 70 60 50 40 30 20 10 0 y = 1.00x + 2.1, R 2 = 0.77 TSA medium 0 20 40 60 agar (CFU) 34
Surface monitoring 35
Rapid detection of microbes on surfaces at a pharma site Growth Direct microcolonies visual plate counting 15 hr 72 hr 36
Surface testing: co-trending of rapid (1 day) and traditional (2 day) tests at a pharma facility Naked visual agar plate (48 count hours) at 48 hr 20 Growth Direct Direct (24 hours) test at 24 hr colonies detected 15 10 5 0 Oct 17 Oct 24 Nov 9 Dec 6 Dec 21 Jan 16 Mar 8 May 1 2 3 4 5 6 7 8 9 10 11 date TSA, 32.5 C 37
Comparing recovery on membranes and agar in surface testing using capture efficiency (Whyte et al, 1989) Efficiency (E) = fraction recovered of total microbes/ replicate Sample multiple times on same location (e.g. 5 replicates) Incubate Count each plate 38
Comparing efficiency of recovery for surface contact plates: membrane Vs agar 1000 Efficiency (E) = fraction recovered of total microbes/ replicate E = [1 log (slope)] E membrane = 0.34 E = agar 0.28 colonies recovered 100 10 agar membrane 1 0 2 4 6 8 10 replicate contact plate number Whyte et al, 1989. J. Hosp. Inf., 13: 33-41 39
Surface contact testing: equivalent capture efficiencies on membranes vs. agar average capture efficiency Surface stainless steel 12 sites glass 10 sites tyvek 8 sites plexiglass 10 sites latex gloves 9 thumbs membrane 0.40 ± 0.13 0.38 ± 0.12 0.32 ± 0.13 0.40 ± 0.08 0.26 ± 0.09 agar 0.38 ± 0.11 0.49 ± 0.07 0.31 ± 0.10 0.40 ± 0.09 0.32 ± 0.10 40
Validation Question - Growth Direct System, New Technology? Growth Direct is not an alternative technology it is based on standard USP growth based membrane filtration methods the results are given as CFU s. The novel Growth Direct is an automated compendial method: the system is an Automated colony counter and can be linked to the USP Chapter <16> Automated Methods of Analysis. validation requires proof that the camera sees as many micro-colonies as the eye would see colonies on the membrane surface. Performance Qualification would follow standard requirements in chapter <1227>, <61> etc. other validation components are standard incubator and software validation protocols. 41
Summing up Advantages of the an automated compendial enumeration method: addresses same broad spectrum of QC applications as the compendial method sensitive digital imaging detects microcolonies non-destructive, compatible with microbial ID equivalent counts to current method Autofluorescence-based detection offers equivalent results with substantial time savings for environmental applications: water air surfaces 42
Environmental monitoring using a rapid non-destructive automated compendial method 43