European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 1 of 16

Report of the NRL Proficiency Test to detect Trichinella spiralis larvae in pork and/or horse meat

samples according to the EU directive 2075/2005

March-April, 2010

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 2 of 16

Table of contents

1 Introduction 3

2 Scope 3

3 Time frame 3

4 Test material 3

5 Instructions to participants 4

6 Statistical analysis 4

7 Z-score 4

8 Participating laboratories 4

9 Results 5 9.1 NRL description 5 9.2 Delivery of the package to NRLs 5 9.3 Digestion methods 5 9.4 Type of meat samples 5 9.5 Descriptive statistic of the results 5 9.6 Not consistent results 6 9.7 Z-score 6 9.8 NRL experience 6

10 Conclusions 6 11 References 7 Annex 1 8 Annex 2 11 Annex 3 14 Annex 4 15 Annex 5 16

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 3 of 16

1 Introduction Nematode worms of the genus Trichinella are zoonotic parasites circulating in most of the European countries in both wild and domestic animals (Pozio et al., 2009). Humans acquire the infection by the consumption of raw or undercooked meat from pigs, horses, wild boars and other game animals (Pozio et al., 2003). According to the Commission Regulation (EC) No 2075/2005, all animals which are potential carrier of Trichinella spp. infective larvae should be tested at the slaughterhouse according to one of the approved test.

2 Scope One of the core duties of the EURL for Parasites is to organise proficiency tests, as it is stated in the Regulation (EC) No 882/2004 of the European Parliament and of the Council (Commission Regulation No. 882/2004). The scope of this comparison is to test the competence of the appointed NRLs to detect and to count Trichinella larvae in pork and/or horse meat samples according to one of the approved methods reported in the Annex 1 of the Commission Regulation EC No 2075/2005. 3 Time frame The proficiency test (PT) was announced to NRLs by email on February 12, 2010 and the dead line to send the participation form was February 28, 2010. On March 15, 2010, the samples were dispatched to participants by an international courier. Reporting deadline was April 7, 2010. 4 Test material

The test material forwarded to each laboratory, consisted of 10 meat balls made with 100 ± 5 grams (or 35 ± 2 grams) of minced pork or 5 meat balls made with 100 ± 5 grams (or 35 ± 2 grams) of minced pork plus 5 meat balls made with 100 ± 5 grams (or 35 ± 2 grams) of minced horse meat. Nine meat balls contained in the core a known number of viable Trichinella spiralis larvae; whereas, one meat ball, which did not contain any larva, was the negative control. To evaluate the NRL competence and skill and the sensitivity of the digestion method in each participating laboratory, the number of larvae added into the meat balls was of three different sizes: 1) in 35 g samples they were 1, 2 or 3 larvae; 2) in 100 g samples they were 2, 4, 6 or 9. Larvae were obtained by a partial artificial digestion of T. spiralis-infected mice (isolate code ISS03), according to the EURLP Guideline. Larvae were counted under a stereo-microscope using a watch glass of 2 cm of diameter and transferred to the meatball rinsing the watch glass with PBS. To ensure that no larva remained on the glass, it was examined twice under the stereo-microscope and rinsed with PBS allowing the PBS to reach the core of the ball (see Annex 1). Each meat ball was close in a plastic bag under vacuum, a code was added on the envelop and the same code with the number of larvae was reported in an Excel file. Each envelop containing the meat ball was then stored at +4°C until the forwarding. The 10 meat balls under vacuum were forwarded in a polystyrene box containing ice boxes. In the package, the ice boxes were separated from the meat samples by a cardboard separator to avoid a direct contact between meat samples and ice boxes. The packages were forwarded according to the international forwarding regulations by an international courier. An envelop with the hard copies of the 5 forms (see Annex 1) and an accompanying letter was also included in the box. To check the material stability during the time, and to estimate the suitability of the packing and forwarding conditions under which the meat balls were forwarded, two groups of 10 meat balls each, were stored in the package (and the package was stored at room temperature) as those that were forwarded, and tested at the EURLP three and five days after the forwarding. The hard copies of the forms which should be filled in and send back to EURLP were:

1) check of the package content and its condition of preservation (form 1, Annex 2);

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 4 of 16

2) laboratory description (form 2, Annex 2); 3) instruments, reagents and materials required for the detection of Trichinella larvae in pork

according to the Commission Regulation No 2075/2005 (form 3, Annex 2); 4) results of pork sample digestion (form 4A, Annex 2) or of horse meat sample digestion (form

4B, Annex 2); 5) code assigned to the laboratory (form 5); 6) an accompanying explanatory letter (Annex 2).

5 Instructions to participants Practical instructions were given to all the participants in the form 3 that accompanied the samples (Annex 2). A list of instruments as well as a list of chemicals and disposable material required to perform the digestion procedures, were also included. To make comparable the results obtained by the different laboratories involved in the proficiency tests, all the participants should follow the protocol step by step or, on the contrary, describe the variation. 6 Statistical Analysis The analysis of data was performed using STATA software, the statistical significance of the obtained results was confirmed by Kruskal-Wallis test . 7 Z-score A Z-score was tentatively established to evaluate the laboratory performance. The evaluation parameters were: 1) at least the 70% of the number of larvae present in the sample should be recovered for samples with more than 9 larvae; 2) 50% of larvae present in the sample should be recovered for samples with a number of larvae ranging from 4 to 9; and 3) at least one larva should be detected in samples with 1-3 larvae, but no Z-score value can be applied in this case. If the Z-score value is ≤ 2.0, the result is positive. If the Z-score value is 2.0 < Z ≤ 3.0, the result is a warning signal; and if the Z value is >3.0, it is an action signal.

8 Participating laboratories Of the 27 MS, Luxembourg appointed the Belgium NRL for parasites, all the other NRLs participated at the PT. In addition, NRLs from Croatia, Serbia and Switzerland agreed to participate. Consequently, a total of 29 labs were involved.

Table 1 – National Reference Laboratories (NRL) participating at the proficiency test for Trichinella.

Participating NRL for parasites

Country

Institut für Veterinärmedizin, Innsbruck Austria Institute of Tropical Medicine, Antwerp Belgium National Diagnostic and Research Veterinary Institute, Sofia Bulgaria State Veterinary Laboratory, Nicosia Cyprus National Reference Laboratory for Trichinella, Zagreb Croatia University of Veterinary and Pharm Sciences, Brno Czech Rep Danish Food and Veterinary Institute, Copenhagen Denmark Estonian Veterinary and Food Laboratory, Tartu Estonia Finnish Food Safety, Evira, Oulu Finland Lab. études et recherches en pathol animale et zoonoses, AFSSA, Maison Alfort France Federal Institute for Risk Assessment, BfR, Berlin Germany Centre of Athens Veterinary Institutions, Athens Greece

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 5 of 16

Laboratories for Parasitology, Fish and Bee Diseases, Budapest Hungary Central Meat Control Laboratory, Celbridge, County Kildare, Ireland Istituto Superiore di Sanità, Rome Italy Laboratory of Food and Environmental Investigations (LFEI), Riga Latvia National Veterinary Laboratory, Vilnius Lithuania National Veterinary Laboratory, Albertown, Marsa Malta National Institute of Public Health and the Environment, RIVM, Bilthoven Netherlands National Veterinary Research Institute, Pulawy Poland Laboratório Nacional de Investigação Veterinária, Lisboa Portugal Hygiene and Public Veterinary Health Institute, Bucharest Romania Institute for Laboratory Diagnostic INEP, Belgrade Serbia State Veterinary and Food Institute, Bratislava Slovak Rep. National Veterinary Institute, Ljubljana Slovenia Centro Nacional de Alimentación, Agencia Española de Seguridad Alimentaria y Nutrición, Majadahonda

Spain

Statens Veterinärmedicinska Anstalt, Uppsala Sweden Institute of Parasitology, Vetsuisse Faculty, University of Bern, Bern Switzerland Veterinary Laboratories Agency, Weybridge UK

9 Results 9.1 NRL description

Of the 29 NRLs which agreed to participate at the proficiency test, 23 (79%) are accredited according to ISO/IEC 17025:2005 and have validated and/or accredited the digestion test according to the Commission Regulation 2075/2005. Out of the six non-accredited labs, five are NRLs of EU and one is from a non-EU country. The experience of the laboratories in the digestion test (considered as number of samples analyzed during the previous year) was very variable (from zero to several thousands).

9.2 Delivery of the package to NRLs

Out of the 29 packages, 21 were delivered within 24 h, 3 within 48 h, 1 within 72h, 2 after 5 days and 2 after 7 days. All the packages were delivered to NRLs within 24-72 hours. At the delivery, the internal temperature of 27 packages was less than 10°C, whereas in one package delivered after 24 h and in one package delivered after 7 days the internal temperature was of 14°C and 12°C, respectively. The time elapsed from the arrival of the package at the NRL and its control was within 1 hour for 25 NRLs, of 2 hours for 3 NRLs, and of 6 days for 1 NRL, but the parcel was stored in a refrigerator at +4°C.

9.3 Digestion methods

The magnetic stirred method for the pooled sample digestion was the most used (25 NRLs, 86%); the mechanically assisted pooled sample digestion method/sedimentation technique was used in 4 NRLs (14%).

9.4 Type of meat samples

Out of 29 participating laboratories, 13 labs (45%) requested 5 pork samples and 5 horse meat samples of 100 ± 5 g, and 3 labs (10%) requested 5 pork samples and 5 horse meat samples of 35 ± 2 g. Eleven labs (38%) requested 10 pork samples of 100 ± 5 g; 1 lab (3.4%) requested 10 pork samples of 35 ± 2 g.; and 1 (3.4%) lab requested 5 pork samples of 100 ± 5 g.

9.5 Descriptive statistic of the results

The descriptive statistic of the difference between expected and reported counts by laboratory, shows a statistically significant variation among laboratories on the difference between expected and

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 6 of 16

reported count according to the Kruskal-Wallis test (p=0.001) (Annex 3). The descriptive statistic of the difference between the expected and reported larval count by laboratory and the comparison between 2007, 2008, 2009 and 2010 results are shown in the Annex 4. The box-plots of the absolute difference between expected and reported count by laboratory is shown in the Annex 5. Ten labs (38.5%, out of 26 for which data from 2007 to 2010 were available) show a constant trend toward improvement in the digestion test during the 2007-2010 period; 7 labs (27%, out of 26 labs) show small fluctuations in the results; and the other 9 labs (35%, out of 26 labs) show greater fluctuations in their results. Three labs participated for the first time at the PT. The comparison of the overall mean difference (i.e. the mean of relative difference values considering all samples and all laboratories) shows a general improvement in the PT results during the 2007-2010 period.

Table 2 - Overall mean difference comparison for 2007-2010 period

Year 2007 2008 2009 2010 Relative difference values 0.40 0.22 0.14 0.12

9.6 Not consistent results

Two NRLs (7%) detect one false positive sample each. Seven labs (24%) found false negatives (Table 2). All samples stated negatives, contained only one larva. 9.7 Z-score Out of 27 labs for which the Z-score was applied (for 2 labs, the Z-score did not apply to any case, i.e. the samples tested contained only 1-3 larvae), 15 labs (55.5%) obtained only positive signals, 1 lab (3.5%) reported only a warning signal and 11 labs (41%) reported warning and/or action signals.

Table 3. No. of larvae in the sample considered to be negative by NRLs

Laboratory code 1 12 16 18 24 35 36 No. of larvae in the samples considered negative

1 2 6 4 4 1 1

2

9.7 NRL experience

The NRL “experience” based on the number of sample analyzed in the previous years (2007-2009), does not show any influence on the score value of the NRL. No difference was observed between accredited and non-accredited laboratories.

10 Conclusions

More homogeneous results among NRLs were observed in the 2010 PT, compared to the 2007-2009 PTs. Ten labs (38.5%) of 26 performed better than in 2009. The presence of laboratories reporting a number of larvae greater then the real number of larvae spiked into the sample could suggest: 1) a problem to count the larvae present in the sediment; and 2) cross contamination from one digestion process to the next. No false positive has been found, whereas 8 (31%) laboratories found false negatives (last year the value was 37.5%). The use of the suggested Z-score seems to be a suitable method to evaluate single laboratory results for the artificial digestion in the course of PTs. However, according to the EC Regulation 2075/2005, technicians must be able to identify positive samples irrespective to the amount of larvae contained in the sample.

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 7 of 16

11 References

Pozio E., Rinaldi L., Marucci G., Musella V., Galati F., Cringoli G., Boireau P., La Rosa G. (2009) Hosts and habitats of Trichinella spiralis and Trichinella britovi in Europe. Int. J. Parasitol. 39:71-79. Pozio E., Gomez Morales M.A., Dupouy Camet J. (2003). Clinical aspects, diagnosis and treatment of trichinellosis. Expert Review of Anti-infective Therapy 1:471-482. Commission Regulation (EC) No 2075/2005 of 5 December 2005 laying down specific rules on official controls for Trichinella in meat. OJ L338/60-82. Commission Regulation (EC) No. 882/2004 of the European Parliament and of the Council of 29 April 2004 on official controls performed to ensure the verification of compliance with feed and food law, animal health and animal welfare rules. OJ L191/1-59. Community Reference Laboratory for Parasites (2007). Guideline for the detection of Trichinella larvae at the slaughterhouse or connected laboratory in a Quality Assurance System, pp. 1-14, www.iss.it/crlp/docu/.

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 8 of 16

Annex 1

Meatball samples (35g and 100g) of minced horse meat and pork

are prepared

An hollow is made in the center of each meatball to house the larvae

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 9 of 16

Larvae were counted under a stereo-microscope using a watch glass and transferred to the meatball rinsing the watch glass

with 200 µl of PBS

The glass was examined under the stereo-microscope to ensure that no larva remained on it

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 10 of 16

Each sample is under vacuum sealed and labeled with a numeric code

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 11 of 16

Annex 2

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 12 of 16

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 13 of 16

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 14 of 16

Annex 3

Descriptive statistics of the absolute difference between expected and reported count, by laboratory for all samples (horse meat and pork) in 2010

Laboratory N mean sd

1 9 0.11 0.60

2 9 0.00 0.50

3 9 0.66 0.87

4 9 0.66 0.50

5 9 0.88 1.05

6 9 0.66 0.70

7 9 0.11 0.60

8 9 0.55 0.73

9 9 0.00 0.50

10 9 2.00 1.41

11 9 0.88 0.78

12 9 1.22 1.30

13 9 1.22 1.20

14 9 0.66 1.41

15 9 1.33 1.50

16 9 1.33 1.93

17 9 0.11 0.33

18 9 1.77 1.85

19 9 0.77 0.67

20 9 0.11 0.33

21 9 0.66 2.87

22 9 0.77 1.39

23 9 0.00 0.70

24 9 0.88 1.83

25 9 1.33 1.87

26 9 0.44 1.01

34 8 1.00 3.38

35 9 0.77 0.66

36 5 1.80 0.84

All 256 0.70 1.41

Kruskal-Wallis test, p = 0.001

European Union Reference Laboratory for Parasites

Istituto Superiore di Sanità

page 15 of 16

ANNEX 4

Descriptive statistic of the difference between the expected and reported larval count by laboratory and comparison among 2007, 2008, 2009 and 2010.

Laboratory Mean 2007 Mean 2008 Mean 2009 Mean 2010 Overall mean

1 0.36 0.21 0.08 0.04 0.17

2 0.14 0.22 0.07 0.03 0.11

3 0.47 0.30 0.23 0.10 0.27

4 0.18 0.12 0.03 0.12 0.11

5 0.14 0.00 0.08 0.12 0.08

6 0.16 0.08 0.04 0.10 0.09

7 0.20 0.05 0.03 0.03 0.08

8 0.13 0.08 0.03 0.12 0.09

9 0.13 0.15 0.06 0.00 0.08

10 0.17 0.45 0.28 0.41 0.33

11 0.33 0.21 0.20 0.22 0.24

12 0.72 0.10 0.08 0.26 0.29

13 1.00 0.29 0.17 0.25 0.43

14 0.75 0.09 0.20 0.06 0.27

15 0.64 0.64 0.42 0.25 0.48

16 0.40 0.11 0.13 0.23 0.22

17 0.86 0.26 0.61 0.05 0.44

18 0.26 0.27 0.23 0.36 0.28

19 0.08 0.12 0.03 0.16 0.10

20 - 0.19 0.08 0.01 0.10

21 0.40 0.25 0.02 0.40 0.27

22 0.36 0.37 0.30 0.12 0.29

23 0.48 0.38 0.17 0.01 0.26

24 0.45 0.27 0.37 0.10 0.30

25 0.69 0.43 0.75 0.21 0.52

34 - - 0.23 0.08 0.15

Istituto Superiore di Sanità

European Union Reference Laboratory for Parasites

page 16 of 16

Annex 5

Box-plots of the absolute difference between expected and reported count, by laboratory all samples (horse meat and pork)

-10

-50

510

Overall mean

difference (0.7 larvae)

1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526343536

Laboratory code