Post on 07-May-2018
2nd Int. Congress on Quality of Fishery Products, 17-18 Nov., Bilbao
Use of predictive microbiology in the assessment p gy
of health risks and shelf life in fishery products
Paw Dalgaard
Seafood & Predictive Microbiology
National Food Institute (DTU Food) at o a ood st tute ( U ood)
Technical University of Denmark (DTU)
pada@food dtu dk pada@food.dtu.dk
Use of predictive microbiology in the assessment ofhealth risks and shelf life in fishery products
Outline of presentation:
health risks and shelf life in fishery products
Outline of presentation:
• Predictive microbiology
• Human health risks – assessment and management
• Histamine in marine finfish
• Listeria monocytogenes in ready-to-eat seafood
• Shelf-life of fishery products• Shelf life of fishery products
• Predictive models and software
Ti i i • Time-temperature integration tags
• Conclusions and perspectives
DTU Food 2/32
Use of predictive microbiology in the assessment ofhealth risks and shelf life in fishery products
/g)
health risks and shelf life in fishery products
og c
fu/ Spoilage microorganisms
Pathogenic microorganisms
sms
(L
Critical concentration of
Shelf-life
oorg
anis Critical concentration of
spoilage microorganisms
of m
icro
'Safe shelf-life'
Con
c. o Critical concentration of
pathogenic microorganisms
DTU Food 3/32Storage time
C
Predictive microbiology and seafood complexity
• Temperaturep
• pH
• NaCl/water activity
• Smoke components (phenol)
• Nitrite
• CO2• CO2
• Acetic acid
• Benzoic acid
12 parameters
• Citric acid
• Diacetat
Lactic acid• Lactic acid
• Sorbic acid
• Interactions between all these parameters
DTU Food 4/32
p
Predictive microbiology and seafood complexity
910
6789
u/g)
Growthrate
3456
Log
(cfu
012
Lag time
Storage time
Primary model Secondary model
A simplified approach is needed to model the effect of the many relevant
product characteristics and storage conditions that influence growth and growth
DTU Food 5/32
boundary of pathogenic and spoilage microorganisms in seafood
Use of predictive microbiology in the assessment ofhealth risks and shelf life in fishery products
Outline of presentation:
health risks and shelf life in fishery products
Outline of presentation:
• Predictive microbiology
• Human health risks – assessment and management
• Histamine in marine finfish
• Listeria monocytogenes in ready-to-eat seafood
• Shelf-life of fishery products• Shelf life of fishery products
• Predictive models and software
Ti i i • Time-temperature integration tags
• Conclusions and perspectives
DTU Food 6/32
Histamine in marine finfish
• Histamine fish poisoning is responsible for more foodborne incidents
of disease than any other hazard in fish and shell fishof disease than any other hazard in fish and shell-fish
Free histidine Histidine decarboxylase HistamineFree histidine Histidine decarboxylase Histamine
• Significant growth is required more than 1-10 million bacteria/g
• Toxic histamine concentrations (> 500 mg/kg) can be formed by:
• Mesophilic bacteria at above 7–10˚C• Mesophilic bacteria at above 7–10 C
• Psychrotolerant bacteria at above ~0˚C
Toxic histamine concentrations can be formed in marine finfish • Toxic histamine concentrations can be formed in marine finfish
when these are chilled in agreement with EU regulations
DTU Food 7/32
Histamine in marine finfish
Morganella psychrotolerans can grow and is able to produce toxic
t ti f hi t i t 0°Cconcentrations of histamine at 0°C
10Growth Histamine
6000
7000
8000
9000
pm)7
8
9
)
3000
4000
5000
6000
stam
ine
(pp
20°C3
4
5
6
Log(
cfu/
g
20°C15°C
0
1000
2000
3000
His 15°C
10°C 5°C 0°C
0
1
2
3 15°C 10°C 5°C 0°C
0 5 10 15 20 25 30 35 40 45 50
Days0 5 10 15 20 25 30 35 40 45 50
Days
0
DTU Food 8/32Emborg & Dalgaard (2008a)
Histamine in marine finfish
Both mesophilic and psychrotolerant bacteria have beenresponsible for incidents of histamine fish poisoning
Seafood Bacteria Place and time
Fresh tuna Morganella morganii Japan, 1955
Fresh tuna Morganella morganii Japan, 1965
Fresh tuna Hafnia sp. Prauge, 1967
Fresh tuna Raoultella planticola (Klebsiella California, 1977es tu a aou te a p a t co a ( ebs e apneumoniae)
Ca o a, 9
Dried Sardine Photobacterium phsophoreum Japan, 2002
Tuna in chilisauce Morganella psychrotolerans or Denmark, 2003Tuna in chilisauce Morganella psychrotolerans orPhotobacterium phosphoreum
Denmark, 2003
Cold smoked tuna Photobacterium phosphoreum Denmark, 2004
Cold smoked tuna Morganella psychrotolerans Denmark 2004Cold smoked tuna Morganella psychrotolerans Denmark, 2004
Tuna in flexible packaging
Morganella morganii Denmark, 2004
Fresh tuna Photobacterium phosphoreum Denmark 2006
DTU Food 9/32
Fresh tuna Photobacterium phosphoreum Denmark, 2006
Dalgaard et al. 2008
Histamine in marine finfish
Predictive models for growth and histamine formation by both
M. psychrotolerans and M. morganii have been developed and validated
1 2
1.4 : Morganella psychrotolerans: Morganella morganii
1.0
1.2
x, h
-1)
0.6
0.8
Sqrt
(µm
ax
0.2
0.4
S
-5 0 5 10 15 20 25 30 35 40 450
DTU Food 10/32
Temperature (°C)
Emborg & Dalgaard (2008b)
Histamine in marine finfish Histamine formation by M psychrotolerans can be predicted for vacuum Histamine formation by M. psychrotolerans can be predicted for vacuum packed fresh tuna and it is markedly faster at 4.4˚C compared to 2.0˚C
DTU Food 11/32Emborg & Dalgaard (2008b) - sssp.dtuaqua.dk
Histamine in marine finfish
• Combined model for M. psychrotolerans and M. morganii predicts
histamine formation for a wide range of storage temperaturesg g
• The model allows the effect of delayed chilling to be predicted
Delayed chilling: 25˚C for 17 h 25˚C for 22 hThen chilled storage at: 5 ˚C 5˚C
DTU Food 12/32Emborg & Dalgaard (2008b) – http://sssp.dtuaqua.dk
Use of predictive microbiology in the assessment ofhealth risks and shelf life in fishery products
Outline of presentation:
health risks and shelf life in fishery products
Outline of presentation:
• Predictive microbiology
• Human health risks – assessment and management
• Histamine in marine finfish
• Listeria monocytogenes in ready-to-eat seafood
• Shelf-life of fishery products• Shelf life of fishery products
• Predictive models and software
Ti i i • Time-temperature integration tags
• Conclusions and perspectives
DTU Food 13/32
Listeria monocytogenes
• Present in low concentrations in many fresh and lightly preserved
aquatic foods
• Causes listeriosis with high mortality (20-30%)
• Is rapidly inactivated at > 70 75°C (cooking and hot smoking)• Is rapidly inactivated at > 70 - 75°C (cooking and hot smoking)
• Growth can be difficult to prevent in chilled foods
(Psychrotolerant and halotolerant)
• Listeriosis has been caused by various ready-to-eat (RTE) foods Listeriosis has been caused by various ready to eat (RTE) foods
DTU Food 14/32
Listeria monocytogenes - microbiological criteria
EU regulation (EC 2073/2005):
• RTE seafoods able or unable to support growth
• Predictive microbiology models can be used to document
control of Listeria monocytogenes in seafood
Ready-to-eat f d
Critical limit Comments foods
C t ca t Co e ts
Support growth None in 25 g - When produced
S t th 100 f / It t b d t d th t Support growth 100 cfu/g - It must be documented that 100 cfu/g is not exceeded within the storage period
bl f /Unable to support growth
100 cfu/g - Documentation - pH ≤ 4,4 or aw ≤ 0,92 - pH ≤ 5,0 and aw ≤ 0,94
DTU Food 15/32
- Shelf-life below 5 days
Quantitative microbiological risk assessment(QMRA)
Prevalence and conc. of Listeria
Product characteristics
Storage conditions
Storage time (shelf-life)
Predictive microbiology models
osu
ssm
ent
(Deterministic and stochastic)
Expo
a
sses Output: Predicted concentration in seafood at
the time of consumption
n
Consumption patterns and dose response models
zard
re
rizat
ion dose-response models
DTU Food 16/32
Haz
char
axr
Output: Cases per year FAO/WHO (2004)
Listeriosis QMRA for RTE seafood in Navarra, Spain
• Risk estimate: < 0.3 cases/year in Navarra due to smoked fish
• Risk can be markedly reduced by limiting storage time to 7 daysy y g g y
and storage temperature to 4.5°C
DTU Food 17/32Garrido et al. (2010)
Listeria monocytogenes – product validation of complex predictive growth rate model
0.50Line of perfect match
0.40
cted
0.30
e - p
redi
c
0.20
row
th ra
te
Meat (n = 390)Seafood (n = 145)Poultry (n 47)
0.10G
Bias/accuracy factors = 1.0/1.5 (n = 635)
Poultry (n = 47)Dairy (n = 53)
0.00 0.10 0.20 0.30 0.40 0.50
0.00
DTU Food 18/32
Growth rate - observedMejlholm et al. 2010
S ft di t th f L t i id f f d
Listeria monocytogenes in ready-to-eat seafood
Software can predict growth of L. monocytogenes in a wide range of seafood
DTU Food 19/32DIFRES
Listeria monocytogenes in ready-to-eat seafood
Software can predict growth of L. monocytogenes (and lactic acid bacteria, LAB) for constant and dynamic temperature storage conditions
The SSSP software is available for free at http://sssp.dtuaqua.dk
DTU Food 20/32
Listeria monocytogenes in ready-to-eat seafood
The SSSP software predicts combinations of product characteristics
that prevent growth of L. monocytogenes
d
MIC sorbic acid
p g y g
rbic
aci
d
ψ = 1
ψ = 2.0
No-growth
hase
so M
IC benzo
ψ = 0
ψ = 1.0
1.5ψ = 1.25
Wat
er p
ψ = 0.5
oic acid
0.75
% W Growth
DTU Food 21/32
% W ater phase benzoic acidMejlholm & Dalgaard 2009
Listeria monocytogenes in ready-to-eat seafood
Product development Quality control
(Taget characteristics) (Acceptable variation)
Validated
3.54.0 % LactateTemperature:8°CpH:6.0Phenol:10.0ppmPr edicted growth boundaries - effect of w
Predictive model
Customers Authorities
(Documentation) (Documentation)
DTU Food 23/32
Use of predictive microbiology in the assessment ofhealth risks and shelf life in fishery products
Outline of presentation:
health risks and shelf life in fishery products
Outline of presentation:
• Predictive microbiology
• Human health risks – assessment and management
• Histamine in marine finfish
• Listeria monocytogenes in ready-to-eat seafood
• Shelf-life of fishery products• Shelf life of fishery products
• Predictive models and software
Ti i i • Time-temperature integration tags
• Conclusions and perspectives
DTU Food 24/32
Growth of spoilage bacteria in fresh MAP cod fillets
910
: Total microfloraPh t b t i h h
78
)
: Photobacterium phosphoreum
56
g (c
fu/g
)
34Lo
g
12
0 2 4 6 8 10 12 14 16 18Storage period (days at 0 oC)
0
DTU Food 25/32
Storage period (days at 0 C)
Shelf-life prediction - models and software
SSO Product Freeware
H2S-producing Fresh seafood - Seafood Spoilage and Safety Predictor
Shewanella Fresh seafood - Seafood Spoilage and Safety Predictor
Pseudomonas spp. Fresh seafood - Combase Predictor - Fish Shelf Life Prediction Fish Shelf Life Prediction
Photobacterium phosphoreum
Fresh marine MAP fish and shell-fish - Seafood Spoilage and Safety Predictor
Lactic acid bacteria Fresh and lightly preserved products
- Seafood Spoilage and Safety Predictor
Brochothrix thermosphacta
Fresh and lightly preserved products
- Combase Predictor
• Seafood Spoilage and Safety Predictor (http://sssp.dtuaqua.dk )
• Combase Predictor (http://www.combase.cc)
DTU Food 26/32
• Fish Shelf Life Prediction (http://www.azti.es/...)
Shelf-life prediction –d l d ft
Software predicts the effect of
models and software Temperature p
measured or theoretical product temperature profiles on growth of
f
(°C)
specific spoilage organisms (SSO) and on the remaining shelf-life of products:products:
Example with Photobacterium phosphoreum in fresh MAP salmonphosphoreum in fresh MAP salmon
DTU Food 27/32
http://sssp.dtuaqua.dk
Time temperature integration (TTI) tags
The French company CRYOLOG produces the microbial TTIs TRACEO® and (eO)®
Predictive models for lactic acid bacteria inside the tags is used to set gtheir response time for different foods
DTU Food 28/32http://www.cryolog.com
Conclusions and perspectives 1
• Predictive microbiology models are available to predict human health
risk and shelf-life for several seafood:micro-organisms combinations
• Predictive microbiology models can be used to:
- Evaluate and document the microbial growth in specific seafoodsEvaluate and document the microbial growth in specific seafoods
(Human health risk and shelf-life assessment)
Identify combinations of product characteristics that reduce - Identify combinations of product characteristics that reduce
growth of specific microorganisms
( d t d l t d t ti ) (product development, documentation)
DTU Food 30/32
Conclusions and perspectives 2
• The number successfully validated predictive models is increasing but
we are still lacking:we are still lacking:
- Shelf-life models for various lightly preserved seafoods
M d l f i bi l i l i d h i l h - Models for microbiological, enzymatic and chemical changes
- New/improved safety models (e.g. Clostridium, Vibrio and viruses)
• Software has stimulate the practical application predictive
models within the seafood sector
• The benefits of predictive models are far from being fully exploited
within the seafood sector (including both industry and authorities)
DTU Food 31/32