The occurrence of Listeria monocytogenes in cheese from a manufacturer associated with a case of listeriosis

International Journal of Food Microbiology, 10 (1990) 255-262

255

Elsevier FOOD 00310

T h e occurrence of Listeria monocytogenes in cheese from a manufacturer associated w i t h a case of listeriosis J. McLauchlin 1, Melody H. Greenwood 2, Pia N. Pini 3 I Division of Microbiological Reagents and Quality Control, The Central Public Health Laboratory, London, U.K., 2 Public Health Laboratory, Poole General Hospital, Poole, U.K., and 3 Food Hygiene Laboratory, The Central Public Health Laboratory, London, U.K.

(Received 28 August 1989; accepted 27 November 1989)

A case of listeriosis was associated with the consumption of a soft cheese produced in England. Goats cheese and other products from the same food manufacturer were examined for the presence of Listeria over the following 11 months. Listeria monocytogenes was isolated from 16 of 25 cheese samples on retail sale, 12 of 24 cheese samples obtained directly from the factory, and from shelving within the plant. Phage-typing of 68 isolates of L. monocytogenes from cheese samples and the factory showed that 66 (97%) were indistinguishable from the strain isolated from the patient's cerebrospinal fluid and stool. L. monocytogenes was not isolated from seven goats milk or two yoghurt samples. Listeria innocua was isolated from 10 cheese samples, two of which contained no other species of Listeria. Levels of L. monocytogenes shortly after production were low ( < 10/g), but were higher (105-107cfu/g) in six of the 16 cheese samples obtained from retail outlets. Multiplication of L. monocytogenes was demonstrated in cheeses contaminated at the factory and held at 4 ° C in the laboratory. Key words: Listeria monocytogenes; Foodborne infection; Epidemiology; Soft cheese

Introduction I n F e b r u a r y 1988, a p r e v i o u s l y h e a l t h y 4 0 - y e a r - o l d w o m a n w a s a d m i t t e d t o a hospital in London with meningitis. Listeria monocytogenes serovar 4b was isolated f r o m h e r c e r e b r o s p i n a l f l u i d ( C S F ) a n d s t o o l . 24 h b e f o r e t h e o n s e t o f s y m p t o m s , s h e h a d c o n s u m e d a p p r o x i m a t e l y 85 g o f a g o a t s m i l k ' A n a r i ' t y p e s o f t c h e e s e produced in England. The pack of cheese from which she had eaten was not available for examination. However, four unopened packs from the same retailer w e r e h e a v i l y c o n t a m i n a t e d w i t h L. m o n o c y t o g e n e s s e r o v a r 4 b ( A z a d i a n et al., 1989). Following this case of meningitis, cheese samples were obtained from the same m a n u f a c t u r e r ( P r o d u c e r X ) as w e l l as g o a t s m i l k s u p p l i e d t o t h i s f a c t o r y . T h e s e

Correspondence address: J. McLauchlin, Division o~ Microbiological Reagents and Quality Control, The Central Public Health Laboratory, Colindale, Londbn, NW9 5HT, U.K.

0168-1605/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

256 samples plus all further batches of cheese made by Producer X were examined for the presence of Listeria until production ceased in January 1989. Batches of cheese in which L. monocytogenes was detected were destroyed. The investigations of the products from this manufacturer form the basis of this study.

Materials and Methods

Producer X Producer X operated a one-man, off-farm dairy making cheese and yoghurt from goat's and sheep's milk. Two types of goats cheese, Halloumi and Anari, accounted for the most of his cheese production. Halloumi cheese was a full-fat soft unripened cheese produced from unpasteurised goat's milk and vegetarian rennet which was subsequently heated in hot whey at about 8 5 ° C for 10-15 min, and then lightly pressed and brined. Anari was a low fat cheese formed after the whey was further heated to 85 ° C. Before distribution from Producer X, the cheese was portioned and vacuum-packed on site. No recommendations were made concerning temperature of storage. Food and environmental samples Four samples of Anari cheese made by Producer X were obtained from a retail outlet in February 1988. These four vacuum packed cheeses were identified by the family of the patient to be from the same producer and of identical type as that eaten 10 days earlier; they had been stored by the retailer in a refrigerated display unit (8°C) and each had a sell-by date of June 1988. A further 16 cheese and two yoghurt samples from Producer X were obtained from retail outlets during February 1988. Five further cheese samples on retail sale were obtained in August 1988 (two samples), October 1988 (two samples) and in November 1988 (one sample). Sell-by dates ranged from 1 to 6 months after the date of the purchase. Food samples were also obtained directly from Producer X from February 1988 to January 1 9 8 9 : 2 4 cheese (seven types) and seven goats milk samples (Table I). Twenty of these cheese samples were tested within 5 days of manufacture. Environmental swabs were taken from walls, floors, shelves, working surfaces and utensils within the plant. Microbiological testing Samples (10 g or 25 g) were examined for the presence of Listeria by the method of Lovett et al. (1987), with or without pre-enrichment at 3 0 ° C for 18 h in buffered peptone water (Oxoid, Basingstoke, U.K.). Enrichment broths were subcultured after 24 and 48 h onto at least two of the following selective agars: Modified McBride's agar (Lovett et al., 1987), Acriflavine-Ceftazidime agar (Bannerman and Bille, 1988), Oxford Listeria Selective agar (Curtis et al., 1989), and 5% (v/v) horse blood in Columbia agar plus 10000 IU/1 polymyxin and 40 m g / l nalidixic acid (sodium salt). Enumeration was performed by surface plating known volumes of the

257 homogenized sample and its decimal dilutions on to selective agars. Environmental swabs were pre-enriched in 10 ml of buffered peptone water at 30 o C for 18 h, after which 1 ml was subcultured to enrichment broths and examined for Listeria as described above. A homogenate of selected cheeses (1 g in 10 ml of 0.1% peptone water) was used to test for the presence of coliforms, Escherichia coli, Staphylococcus aureus, Bacillus cereus, and Clostridium perfringens with: Violet Red Bile agar (Oxoid, product code CM107), Tryptone Bile agar (Oxoid, product code CM595), Baird-Parker agar (Oxoid, product code CM275), Bacillus cereus agar (Oxoid, product code CM617 plus supplement SR99) and Tryptose Sulphite Cycloserine Agar (Oxoid, product code CM587, plus supplement SR88), respectively. The presence of Salmonella was also sought in 25 g samples by pre-enrichment in buffered peptone water (Oxoid), enrichment in Rappaport-Vassiliadis (Oxoid, product code CM669) and Selenite broths, and subcultured onto Desoxycholate citrate (Lab M, product code 65) and Bismuth sulphite agars (Gibco, product code 152-00560), Presumptively positive colonies of all organisms were identified by means of conventional techniques. The p H of selected cheese samples was measured with indicator paper. Identification and serotyping of Listeria Isolates of Listeria were identified and serotyped with the methods previously described (Seeliger and Hohne, 1979; McLauchlin, 1987). Phage-typing was carried out in Tours, France, under code (Audurier et al., 1979; McLauchlin et al., 1986a).

Results

L monocytogenes serovar 4b was isolated at levels of > 1 0 7 c f u / g from all four Anari cheese packs obtained from the retailer identified by the patient's family. L. monocytogenes serovar 4b was isolated from 24 additional cheese samples, 12 obtained directly from the factory and 12 from retail outlets (Table I). All contaminated cheese samples obtained directly from the manufacturer and 10 retail samples contained levels of L. monocytogenes at < 10 cfu/g. The other two contaminated retail samples, which were purchased 10 weeks before their sell-by date, contained levels of L. monocytogenes at > 105 cfu/g. In eight of the samples containing L. monocytogenes, L. innocua was also isolated. L. innocua only was detected in two cheeses. Listeria were not detected in the remaining 19 cheese, seven goats milk, or two yoghurt samples (Table I). L. monocytogenes serovar 4b and L. innocua were isolated from shelving inside the cheese factory. L. innocua was also isolated from the floor and a knife used to cut portions of cheese, A total of 73 isolates of L. monocytogenes were obtained from the cheese samples (multiple subcultures were obtained from some samples), and all these subcultures were serovar 4b. Phage-typing showed that 65/67 of these isolates were indistinguishable from the strain of L. monocytogenes obtained from the patient's CSF and stool samples (i.e. reacted with phages 2425 and 2671). The isolate of L. monocytogenes from the factory environment was also indistinguishable from the

258 TABLE I The detection of Listeria spp. in food samples from Producer X Product type Goats Milk Yoghurt Cheeses Anari Halloumi Cheddar (Goat) Cheddar (Sheep) Fetta Gjestost Soft chive Total

Total number

Number of

Results of isolation of Listeria (numbers of samples)

of samples

retail samples

L. monocytogenes detected

L. innocua detected

7 2

0 2

ND ND

ND ND

23 11 8 2 3 1 1 49

14 1 4 2 3 1 0 25

14 ~ 8b 4c 0 0 1d 1d 29

6 2 2 0 0 0 0 10

N D = not detected. The levels of L. monocytogenes present in the cheese samples were: a 4 samples at 3-5 x 10V/g, 10 samples at < 1 0 / g ; b8 samples at < 1 0 / g ; c2 samples at 0.5-4.5 x l 0 6 / g , 2 samples at < 1 0 / g ; d 1 sample at < 10/g.

patient's strain. Coliforms were detected in 12 of 24 samples, 10 of which contained 1 0 2 - 1 0 6 c f u / g (Table II). E. coli was detected in 15 of 31 samples, 12 of which contained 102-103 c f u / g (Table II). S. aureus was detected in one of 24 cheese samples. B. cereus was found in three of 22 samples at levels of 2.0-3.0 × 10 2 cfu/g. Salmonella sp. and Cl. perfringens were not detected in eight and 14 samples, respectively.

TABLE II Coliform and E. coli counts from 31 cheese samples from Producer X Cheese

Number of

Results of culture

Numbers of specimens with a

type

samples

for L. monocytogenes

Coliforms

E. coil

Anari

4 5 3 5 2 2 5 3 1 1

present b present ~ absent present absent present absent absent present present

Not 2/4 2/2 3/4 1/2 Not 3/5 1/2 Not 0/1

4/4 3/5 2/3 1/5 0/2 2/2 2/5 0/3 1/1 0/1

Halloumi Cheddar Fetta Gjestost Soft chive

done (2/4) (1/2) (2/4) (1/2) done (3/5) (1/2) done (0/1)

(4/4) (1/5) (1/3) (1/5) (0/2) (2/2) (2/5) (0/3) (1/1) (0/1)

Numbers in parenthesis indicate numbers of cheese samples containing > 102 coliforms or E. coli/g. b Associated with infection, contained, L. monocytogenes at > 107/g. c Contained L. monocytogenes at < 10/g. a

259 The p H values of cheese samples were: Anari p H 5.0-7.0, Halloumi p H 6.0, Cheddar p H 5.5 and Fetta p H 4.0-5.0. The Anari cheeses that contained L. monoeytogenes at > 10 v c f u / g were all p H 6.0. Levels of L. monocytogenes in four cheese samples (one Anari and three Halloumi) from Producer X were determined 2 - 3 days after manufacture and after storage in the laboratory at 4 ° C. In all four samples, L. monocytogenes was initially detected at < 10 cfu/g. The L. monocytogenes level in the Anari sample had increased to 8 x 104 c f u / g after 5 weeks, and in one of the Halloumi samples to 1 X 10 6 c f u / g after 4 weeks. Counts did not increase in samples from the remaining two Halloumi cheeses after 4 - 8 weeks.

Discussion

With the possible exceptions of instances of neonatal cross-infection (McLauchlin et al., 1986b), and three sporadic cases associated with the consumption of soft cheese (Bannister, 1987), cooked chicken and vegetable rennet (Kerr et al., 1988), the source of L. monocytogenes in cases of human listeriosis in Britain remains unknown. In addition, there remains some doubt as to the true source of infection in these three sporadic cases (Bannister, 1987; Kerr et al., 1988) because the implicated strains were isolated only from single opened packs of food which had been used by the patient. It is possible that these implicated foods were contaminated either from other unknown sources that were common to the patient (for example, by other products in the same refrigerator) or from the patients themselves. In the case described here (Azadian et al., 1989), clear evidence of bacterial meningitis was demonstrated, and L. monocytogenes was isolated both from a single colony grown from the CSF and from the faeces. This strain was isolated from faeces on direct culture and, although levels were not enumerated, was present in high numbers. Although the pack of cheese that had been partially eaten by the patient was not available for examination, unopened packs together with a site within the factory manufacturing these products yielded L. monocytogenes which were indistinguishable from those isolated from the patient. Hence the association of this case of listeriosis with an English-produced soft cheese provided stronger evidence for food-borne listeriosis in Britain than had previously been obtained. In addition, it provided a rare opportunity to study the distribution of a potentially pathogenic strain of L. monocytogenes and relate this to the epidemiology of human listeriosis. Of the 49 cheeses that had been obtained from this same producer over the subsequent 11 months, 28 (57%) were contaminated with L. monocytogenes (Table I). Phage-typing, which is highly discriminatory (McLauehlin et al., 1986a), showed that almost all of these isolates of L. monocytogenes as well as a subculture from a site within the factory were indistinguishable from those from the patient. The manufacturing process for both the Anari and the Halloumi cheeses described in this study involved a cooking stage (heating to 85 o C) followed by vacuum packing on the premises. The cheese samples obtained directly from the plant were found to

260 be contaminated with this strain of L. monocytogenes soon after production. Thus, it seems most likely that contamination by L. monocytogenes occurred between cooking and packing, and not from the ingredients of the cheeses or after distribution. The presence of coliforms and E. coli in some samples also supports the concept of post-processing contamination. It has been noted previously (Pini and Gilbert, 1988) that L. monocytogenes may be present in cheese in the absence of other indicators of poor manufacturing practices. In this study, L. monocytogenes was present in cheeses where either E. coil (five samples) or coliforms (two samples) were not detected (Table II). L. innocua was also isolated from 10 samples, two of which were in the absence of L. rnonocytogenes (Table I). Within the genus Listeria, L. monocytogenes alone causes significant human morbidity; the remaining species have either never or very rarely been associated with infection (McLauchlin, 1987). However, since the physiology and distribution of the different species within the genus Listeria are so similar (Seeliger and Jones, 1986), the detection in food of species other than L. mono~ytogenes is likely to indicate an increased risk of contamination by L. monocytogenes. Subcultures of L. monocytogenes from 533 cases of listeriosis occurring in Britain during 1987-1988 were collected, and of these, 378 (71%) were serovar 4b (Division of Microbiological Reagents and Quality Control, Colindale (DMRQC), unpublished data). Phage-typing was performed on isolates from 277 cases, and seven were indistinguishable from the strain of L. monocytogenes associated with Producer X. Of these seven cases, one patient was interviewed and reported consuming both goats milk and goats milk products, but not those from Producer X. It was not possible to ascertain if the other six patients had eaten food from Producer X (Dr S. Hall, Communicable Disease Surveillance Centre, Colindale, U.K.), personal communication). Producer X made approximately 150 l b (70 kg) and 100 lb (45 kg) of cheese per week during 1987 and 1988, respectively, which was distributed to health food shops and delicatessens in the South of England. As the cheese samples from the plant were frequently contaminated with L. monocytogenes, many people are likely to have ingested this potentially pathogenic strain. However, phage-typing results show that this strain was responsible for a low proportion ( < 3%) of the listeriosis cases occurring in Britain during 1987-1988 where strains were sent to D M R Q C . The single case of listeriosis associated with the consumption of cheese from this producer was probably exposed to an extremely high dose ( > 108 cfu) of L. monocytogenes. Since other samples of cheese from this producer, but not associated with infection, also contained high numbers ( > 105 c f u / g ) of this strain (Table I), it is highly likely that other individuals were exposed to an equivalent dose but serious disease did not result. The colonisation of the producer's premises described here show similarities with the recent outbreaks of listeriosis in Switzerland (Bille and Glauser, 1988) and in California (Linnan et al., 1988). In all these three instances transmission was associated with soft cheese, and the products from the manufacturers were contaminated with the epidemic strain for extended periods ( > 6 months). L. monocytogenes shows m a n y properties which favour transmission through food, but it is relatively acid intolerant (McLauchlin, 1987). However, as demonstrated in this

261 study, the p H of soft cheese is frequently within the range of conditions which can support the growth of Listeria, and the growth of L. monocytogenes in various dairy products has been reported (Ryser and Marth, 1987, 1988; Rosenow and Marth, 1987; Marshall and Schmidt, 1988). In 'naturally' contaminated products made by Producer X, the strain of L. monocytogenes associated with a case of listeriosis could multiply in at least two of the cheese types. Assuming a zero lag phase and original levels of L, monocytogenes between 1 and 9 organisms/g, this strain showed a doubling time at 4 ° C of 47-56 h in the Anari cheese, and 32-37 h in one of the Halloumi cheeses. These growth rates correspond well to those of Rosenow and Marth (1987), who reported doubling times of between 29 and 45 h for four strains of L. monocytogenes at 4 ° C in artificially contamined liquid milk products and cream. The ability and rate of growth of L. monocytogenes in food may be of importance in the pathogenic potential of different strains. The shelf-life assigned to cheese may also be significant: sell-by dates on retail samples from this producer were at least 3 months after distribution. This would allow ample time for low levels of L. monocytogenes to increase when stored at 4 ° C . Strict hygiene at the point of production is critical in the prevention of contamination of soft cheese with L. rnonocytogenes. Improved temperature control after manufacture and shorter shelflives may also be required to prevent the proliferation of L. monocytogenes in soft cheese to levels which pose an unacceptable hazard to human health. Codes of practice have been produced by both the Creamery Proprietors Associations (19 Cornwall Terrace, London, NW1 4QP, England; " G o o d Hygienic Practice in the Manufacture of Soft and Fresh Cheeses", August 1988), and the Milk Marketing Board (Thames Ditton, Surrey, K T 7 0 E 1 , England; " G o o d Hygienic Practice in the Manufacture of Soft and Fresh Cheeses in Small and Farm Based Production Units", February 1989), to help address the problems highlighted here in the manufacture of soft cheeses.

Acknowledgements We gratefully acknowledge the support of: Mr D. Roe and Mr A. Weldycz, The Environmental Health Department, New Forest Council; Dr R.J. Gilbert, Food Hygiene Laboratory, Central Public Health Laboratory, London; Dr Susan M. Hall, Communicable Disease Surveillance Centre, London; Dr W.L. Hooper, Public Health Laboratory, Poole; Mr M. Jacob, Department of Health and Social Security, London; Dr A.G. Taylor, Division of Microbiological Reagents and Quality Control, Central Public Health Laboratory, London; Professor A. Audurier, H6pital Trousseau, Tours, France for phage typing; and technical assistance from A m a n d a Scott, Poole Public Health Laboratory.

References Audurier, A., Chatelain, R., Chalons, F. and Pi~chaud, M. (1979) Lysotypie de 823 souches de Listeria monocytogenes isol6es en France de 1958 h 1978. Ann. Microbiol. (Paris) 130B, 179-189.

262 Azadian, B.S., Finnerty, G.T. and Pearson, A.D, (1989) Cheese-borne Listeria meningitis in an immunocompetent patient. Lancet i, 322-323. Bannerman, E.S. and Bille, J. (1988) A new selective medium for isolating Listeria spp. from heavily contaminated material. Appl. Environ. Microbiol. 54, 165-167. Bannister, B.A. (1987) Listeria monocytogenes meningitis associated with eating soft cheese. J. Infect, 15, 165-168. Bille, J. and Glauser, M.P. (1988) Zur Listeriose-Situation in der Schweiz. Bull. Bundesamtes Gesundh. 3, 28-29. Curtis, G.D.W., Mitchell, R.G., King, A.F. and Griffin, E.J. (1989) A selective differential medium for the isolation of Listeria monocytogenes. Lett. Appl. Microbiol. 8, 95 98. Kerr, K.G., Dealler, S.F. and Lacey, R.W. (1988) Materno-foetal listeriosis from cook-chill and refrigerated food. Lancet ii, 1133. Linnan, M.J., Mascola, L., Lou, X.D. et al. (1988) Epidemic listeriosis associated with Mexican-style cheese. New Engl. J. Med. 319, 823-828. Lovett, J., Francis, D.W. and Hunt, J.M. (1987) Listeria monocytogenes in raw milk: Detection, incidence and pathogenicity. J. Food Protection. 50, 188-192. McLauchlin, J., (1987) Listeria monocytogenes, recent advances in the taxonomy and epidemiology of listeriosis in humans. J. Appl. Bacteriol. 63, 1-11. McLauchlin, J., Audurier, A. and Taylor, A.G. (1986a) The evaluation of a phage-typing system for Listeria monocytogenes for use in epidemiological studies. J. Med. Microbiol. 22, 357-365. McLauchlin, J., Audurier, A. and Taylor, A.G. (1986b) Aspects of the epidemiology of h u m a n Listeria monocytogenes infections in Britain 1967-1984; The use of serotyping and phage typing. J. Med. Microbiol. 22, 367-377. Marshall, D.L. and Schmidt, R.H. (1988) Growth of Listeria monocytogenes at 1 0 ° C in milk preincubated with selected pseudomonads. J. Food Protect. 51, 277-282. Pini, P.N. and Gilbert, R.J. (1988) The occurrence in the U.K. of Listeria species in raw chickens and soft cheeses. Int. J. Food Microbiol. 6, 317-326. Rosenow, E.M. and Marth, E.H. (1987) Growth of Listeria monocytogenes in skim, whole and chocolate milk and in whipping cream incubated at 4, 8, 13, 21 and 35°C. J. Food Protect. 50, 452 459. Ryser, E.T. and Marth, E.H. (1987) Fate of Listeria monocytogenes during the manufacture and ripening of Camembert cheese. J. Food Protect. 50, 372-378. Ryser, E.T. and Marth, E.H. (1988) Growth of Listeria monocytogenes at different p H values in uncultured whey or whey cultured with Penicillium camemberti. Can. J. Microbiol. 34, 730-734. Seeliger, H.P.R. and Hohne, K. (1979) Serotyping of Listeria monocytogenes and related species. In: T. Bergan and J.R. Norris (Eds.), Methods in Microbiology, Vol. 13. Academic Press, London, pp. 31-49. Seeliger, H.P.R. and Jones, D. (1986) G e n u s Listeria. In: P.H.A. Sneath, N.S. Mair, M.E. Sharpe and J.G. Holt (Eds.), Bergey's Manual of Systematic Bacteriology, Vol. 2. Williams and Wilkins, Baltimore, pp. 1235-1245.