Weak points of HPP difficult to inactivate spores Margosch, et al. - - PowerPoint PPT Presentation

Predictive modelling for the recovery of Listeria monocytogenes on sliced cooked ham after high pressure processing (HPP)

• S. Koseki and K. Yamamoto

National Food Research Institute, Japan

5th ICPMF 16-19 Sep. 2007 in Greece

Weak points of HPP

 difficult to inactivate spores

Margosch, et al. (2006): C. botulinum spores in pH 5.5 buffer were not completely inactivated by pressure 600 - 1400 MPa at 120 - 140°C.

 recovery after HPP during storage

Bull et al. (2005): Recovery of L. monocytogenes following HPP of milk.

Bacterial recovery after HPP

2 4 6 8 20 40 60 80 100 120

• E. coli (log CFU/ml)

Stored@25ºC

Immediately after

550 MPa treatment Time (h)

Objectives

Food processors need to know how long is safe of product? not only the number of specific microorganisms, but also the risk or safety level (probability) of the product

to develop a predictive model for bacterial recovery

• n sliced cooked ham after HPP

by probabilistic approach

Materials and methods

◆ Bacteria：Listeria monocytogenes ATCC19117 ◆ Food：sliced cooked ham (pasteurized, initial flora <10 CFU/g) ◆ Inoculum level：3 and 5 log CFU/g ◆ HPP: 400, 450, 500, 550, and 600 MPa (5 conditions), 1, 5, 10, 20, 30, 45, and 60 min (7 conditions) (25°C) ◆ Storage： at 10°C for up to 70 days ◆ Sampling： each 14 days (0, 14, 28, 42, 56, 70 days)

Model development

1 2 3 4 5 6 7 8 10 20 30 40 50 60 70 80

• L. monocytogenes (log10 CFU/g)

time@10°C (days)

immediately after HPP ND ND ND

recovery no recovery

At the day of sampling during storage, the number of L. monocytogenes is ≥102 CFU/g ⇒ recovery (1) <102 CFU/g ⇒ no recovery (0) Definition recovery / no recovery

Logit(P) = a0 + a1Press+ a2 log10 Time

( ) + a3Storage + a4IC

Logistic regression

ai: parameters Press: pressure applied Time: pressure-holding time Storage: duration of storage at 10°C IC: inoculum level (log CFU/g) Total 1260 data points

(70 conditions × 6 sampling days × 3 replicates)

Recovery of L. monocytogenes on sliced cooked ham after HPP

1 2 3 4 5 6 7 8 10 20 30 40 50 60 70 80

500MPa-10 min 600MPa-10 min

• L. monocytogenes (log10 CFU/g)

time@10°C (days)

immediately after HPP ND ND ND No treatment

Recovery / No recovery interface of

• L. monocytogenes on sliced cooked ham

after HPP during storage

Inoculum level 3 log CFU/g Inoculum level 5 log CFU/g

Logit(P) =16.8423 0.0722Press 7.8899log10 Time

( ) + 0.1508Storage + 4.9037IC

different responses among the three triplicate trials Recovery No recovery All the three triplicate trials are same

10 20 30 40 50 60 70 400 500 600 Pressure (MPa) Pressure-holding time (min)

~ ~

for 28 days 0.1 10 20 30 40 50 60 70 400 500 600 Pressure (MPa) Pressure-holding time (min)

~ ~

for 28 days 0.1

Prediction of recovery

600 MPa 550 MPa 10 20 30 40 50 60 70 10 20 30 40 50 60 70 80 500 MPa pressure holding time (min) storage time@10ºC (days)

Inoculum level 5 log CFU/g

(b) 450 MPa

Inoculum level 5 log CFU/g

550 MPa 500 MPa 450 MPa 10 20 30 40 50 60 70 10 20 30 40 50 60 70 80 400 MPa pressure holding time (min) storage time@10ºC (days)

Inoculum level 3 log CFU/g

(a)

Inoculum level 3 log CFU/g time taken for the recovery (P = 0.5; 50% probability of recovery)

Prediction of recovery probability

20 min 0.2 0.4 0.6 0.8 1 10 20 30 40 50 60 70 80 10 min storage time@10ºC (days)

Inoculum level 3 log CFU/g treated@500MPa

probability of recovery (a) 5 min 1 min

Inoculum level 3 log CFU/g treated by 500 MPa

60 min 45 min 30 min 0.2 0.4 0.6 0.8 1 10 20 30 40 50 60 70 80 10 min storage time@10ºC (days) probability of recovery

Inoculum level 5 log CFU/g treated@500MPa

(b) 20 min

Inoculum level 5 log CFU/g treated by 500 MPa

probability of recovery during storage after HPP at 500 MPa for 1 - 60 min.

Summary

• 1. A probabilistic predictive model was developed, which

can estimate the recovery of L. monocytogenes on sliced cooked ham after HPP.

• 2. The model calculate appropriate shelf lives and

processing conditions, taking into account for bacterial recovery on the products.

• 3. The model can estimate the risk of recovery of

bacteria during storage.

Future work

To obtain more broad estimation of the model parameters

Hieratical Bayesian statistical modelling

Thank you for your attention!

• S. Koseki and K. Yamamoto

National Food Research Institute, Japan

5th ICPMF 16-19 Sep. 2007 in Greece