Anda di halaman 1dari 6

European Journal of Epidemiology 16: 913918, 2000.

2001 Kluwer Academic Publishers. Printed in the Netherlands.

An outbreak in Italy of botulism associated with a dessert made


with mascarpone cream cheese
P. Aureli1, M. Di Cunto2, A. Maei2, G. De Chiara2, G. Franciosa1, L. Accorinti3,
A.M. Gambardella3 & D. Greco4
1

Reparto di Microbiologia degli Alimenti, Centro Nazionale di Riferimento per il Botulismo, Istituto Superiore di Sanita;
Dipartimento di Epidemiologia e Prevenzione, ASL CE/2; 3Centro di Rianimazione, Presidio Ospedaliero ``G. Jazzolino'',
Vibo Valentia; 4Laboratorio di Epidemiologia e Biostatistica, Istituto Superiore di Sanita, Rome, Italy

Accepted in revised form 8 December 2000

Abstract. In the late 1996, an outbreak of botulism


aected eight young people (age of patients ranged
from 6 to 23 years) in Italy. The onset of the illness
was the same for all of these patients: gastrointestinal
symptoms (nausea and vomiting) followed by neurologic symptoms. The most common neurologic
symptoms were dysphagia, respiratory failure
(100%), diplopia (87%), dysarthria, ptosis (75%) and
mydriasis (50%). All patients required mechanical
ventilation. Botulinum toxin was detected from two
of respectively ve sera and six stool samples analysed, while spores of Clostridium botulinum type A
were recovered from all patient' faeces. The epidemiological investigation led to suspect a commercial
cream cheese (`mascarpone') as a source of botulinum
toxin: indeed, it had been eaten by all the patients
before onset of the symptoms, either alone or as the

(uncooked) ingredient of a dessert, `tiramisu'. Botulinum toxin type A was found in the `tiramisu'
leftover consumed by two patients and in some
mascarpone cheese samples collected from the same
retail stores where the other patients had previously
bought their cheeses. A break in the cold-chain at the
retail has likely caused germination of C. botulinum
spores contaminating the products, with subsequent
production of the toxin. One of the patients died,
while the others recovered very slowly. Prompt international alerting and recall of the mascarpone
cheese prevented the spread of the outbreak due to
the wide range of distribution, demonstrating the
importance of a rapid surveillance system. None of
the people complaining of symptoms after the public
alert resulted positive for botulinum spores and toxin.

Key words: C. botulinum type A, Dessert, Foodborne botulism, Inadequate refrigeration, Refrigerated cheese
Introduction
Botulism is a relatively rare, occasionally fatal neuromuscular disease caused by an extremely powerful
toxin which, in addition to other neurological
symptoms, causes descending symmetrical accid
paralysis within several hours of poisoning. The etiological agent of the disease is Clostridium botulinum,
a gram-positive obligately anaerobic spore-forming
organism [1]. To date, the species that have been
recognised to be associated with foodborne botulism
are C. botulinum types A, B, E, and F [2] and
C. butyricum type E [3, 4].
In Italy, a total of 204 cases have been notied to
the National Botulism Reference Centre (Food Microbiology Laboratory, Istituto Superiore di Sanita)
in the past 5 years (i.e.,19941998) [5]. The most
commonly implicated food has been home-preserved
vegetables.
In late August and early September 1996, eight
cases of botulism caused by C. botulinum type A were
reported in two regions of southern Italy (Campania

and Calabria); the implicated food was a type of


cream cheese, which had been industrially produced
[6]. We describe the results of the epidemiological
investigation conducted, demonstrating the importance of rapid and accurate diagnoses and of immediately alerting the public.
The outbreak
The rst cluster
On the morning of 12 August 1996, a woman in the
Campania Region (southern Italy) prepared a family
lunch, which included `tiramisu' (a refrigerated uncooked dessert made with a type of cream cheese
known as `mascarpone'). The woman had purchased
the mascarpone several days earlier in a local food
store and had properly stored it in the refrigerator
until preparing the tiramisu.
Once prepared, the `tiramisu' was stored at room
temperature for approximately 6 hours before

914
consumption. Four family members (the woman, her
husband, and their two daughters) were present at the
lunch. Two more family members [the woman's
24-year old son (R.C.) and her 9-year-old granddaughter (A.P)] were present only for dessert. The
dessert was rst tasted by the woman who prepared
the lunch and by one of her daughters; they both
decided not to eat the dessert because it tasted butyric
acid o-avour. The woman's husband and the other
daughter neither tasted nor ate the dessert. Only the
son and granddaughter ate entire portions. The leftover dessert was thrown away.
Approximately 7 hours after tasting the dessert,
the woman and her daughter began to complain of
acute abdominal pain and nausea, yet they did not
seek medical care: One hour later, the granddaughter
began to have serious intestinal disturbances
(repeated vomiting, diarrhoea, and serious asthenia),
and the family brought her to the hospital in the
nearby town of Caserta (12 August). Upon hospitalisation, the girl presented with drooping eyelids
and diplopia. In the meantime, the symptoms of the
woman and her daughter began to spontaneously
dissipate until reaching complete resolution that
night. At 6:00 a.m. (13 August), the son (R.C.) suffered a sudden acute attack of gastroenteric syndrome accompanied by neurologic symptoms
(diplopia) and was immediately hospitalised in the
same hospital. In the week following hospitalisation,
symptoms tended to worsen in both patients: the girl
(A.P.) showed dyspnea, dysphagia, lipothymia, cranial and facial nerve dysfunction, tetra hypostenia,
drooping eyelids, diculty in speaking, but the patient was alert; and the son showed ptosis, lingual
paresis, dysphagia, and respiratory failure.
The hospital sta thus decided to transfer both
patients to hospitals in Naples with better intensivecare facilities (20 August); the girl (A.P.) was transferred to the Cardarelli Hospital and the son (R.C.)
to the Policlinico Hospital. In both cases, the patients' symptoms led hospital sta to suspect botulinum intoxication. Blood and faeces samples were
submitted to laboratory conrmation and trivalent
anti-botulinum serum (750 ml) was administered.
The second cluster
The second cluster of cases also occurred in the
Campania Region. On the morning of 30 August,
two brothers [a 15 year-old (N.S.) and a 12 year-old
(G.S.)] invited their 14 year-old neighbour (P.F.) to
their home. The boys prepared a tiramisu for lunch;
the mascarpone used in the dessert's preparation had
been produced by the same rm as that implicated in
the rst cluster; once purchased, it had been properly
stored before preparation of the dessert. The three
boys ate the entire dessert at around 12:00 p.m. At
8:00 p.m., the 12 year-old (G.S.) and the 14 year-old
(P.F.) began to suer from nausea and vomiting and

were immediately hospitalised at the Policlinico


Hospital in Naples; both boys were in a comatose
state upon arrival. In the very early morning of the
following day, the 15 year-old (N.S.) began to suer
similar symptoms and was hospitalised at the Santobono Hospital in Naples. The patients' symptoms
led the sta of the two hospitals to suspect botulinum
intoxication. Blood and stool samples were taken for
the detection of toxins and C. botulinum spores, and
trivalent anti-botulinum serum (750 ml) was administered.
The third cluster
On 3 September, in the town of Vibo Valentia (Calabria Region, southern Italy), two brothers (S.O.,
18 years old, and V.O., 15 years old) consumed a
lunch consisting of tuna sh, cured meat, and tiramisu, which they had prepared that same morning
with mascarpone purchased in a local food store. The
mascarpone had been properly stored after purchase.
Approximately 12 hours after having eaten the lunch,
the two brothers began to suer nausea, vomiting,
and diplopia and, upon the advice of their family
doctor, were admitted (4 September) to the hospital
in the town of Vibo Valentia. A relative of the
brothers, who was also present at the lunch and who
had eaten everything but the tiramisu, suered no
symptoms. Upon hospitalisation, the brothers presented with dysphagia, dysarthria, mydriasis, asthenia, drooping eyelids, and dryness of mouth. The
younger brother also presented with respiratory failure and was in a comatose state upon hospitalisation.
Botulism was suspected; samples were taken for
laboratory conrmation, and the trivalent serum was
administered (750 ml).
An isolated case
On 21 August, a 6-year old boy (P.Z.) was hospitalised at the Catanzaro Hospital in the Calabria Region. During the 2 days prior to hospitalisation, the
child had suered cephalea, vomiting, pharyngodynia, dysphagia, dizziness, ptosis, asthenia and sleepiness accompanied by lipothymia. Upon admittance
to the hospital, the child was not completely oriented
and presented torpor, dysarthria, muscular hypotonia. Viral encephalitis was suspected and therapy was
begun. The child's clinical conditions worsened, with
the onset of bradydyspnea, bradycardia, and cianosis
of face and estremities. Cardiac massage and intubation were then performed. Peristalsis was present.
Following a national alert issued on 3 September for
a brand of mascarpone associated with cases of botulism, the child's father requested that the treating
physicians consider botulism, reporting that on 18
August the child had eaten the same brand of mascarpone. On 7 September, the treating physicians sent

915
serum and stool samples to National Reference
Centre for laboratory examination.
Methods
Epidemiological investigation
The epidemiological investigation consisted of a cohort study in which both the hospitalised patients and
the other individuals present at the implicated meals
were directly interviewed. For the minors involved in
the outbreaks, parents were also interviewed. To reconstruct the history of the purchase, preparation,
and consumption of the implicated foods, we interviewed the persons responsible for their preparation.
To evaluate the production, distribution, and storage
of the implicated foods, we interviewed individuals
working at the production plants, local warehouse
managers, and vendors.
The case denition of foodborne botulism adopted
in this study was the presence of a clinically compatible symptom, conrmed by laboratory analysis
(detection of spores and/or toxins in biological samples) or present in persons associated with laboratory-conrmed cases, which had onset 17 days after
the ingestion of a preserved food.
At least one of the following clinical symptoms had
to be present: diplopia, dryness of mouth, blurred
and/or doubled vision, dysphagia, cranial nerve dysfunction, muscular hypotonia, respiratory failure,
and compatible electromyographic alterations.
Laboratory and environmental investigation
Detection of toxins and spores
A total of 30 serum samples and 37 stool samples or
rectal swabs were taken from patients before treatment with trivalent botulinum antiserum (750 UI
botulinum type A antitoxin, 500 UI botulinum type
B antitoxin, and 50 UI botulinum type E antitoxin
Behringwerke AG, Marburg, Germany, 250 ml vial)
and sent to the National Reference Centre for Botulism, Istituto Superiore di Sanita, Rome, for the
detection of toxins and C. botulinum spores. Laboratory investigations were conducted using standardised methods [7]. In addition to the epidemic cases,
we received also the 29 individuals samples who were
not related or living in the same household presented
with various disturbances in emergency rooms in
Napoli and Caserta (Campania Region) in the days
immediately following the warning issued to the
public. The National Reference Centre also examined
39 samples of mascarpone from the stores where
the patients had purchased the product and at the
warehouses of the distributors that had provided the
product to the stores; the analysed samples were of
the same brand and weight as those consumed by the
patients and were received at the Reference Centre in

their original packaging. The Reference Centre also


examined the leftover tiramisu consumed by the two
boys constituting the third cluster (Calabria Region)
using the same analytical methods. The quantity of
toxin and the counting of C. botulinum spores was
conducted as described previously [8].
Inspection of production plants
Sta members of the Ministry of Health and of the
National Reference Centre inspected the production
plant to examine the production process and to collect environmental samples from the suspected production line and samples of containers used for
packaging and of mascarpone belonging to the same
lots as the incriminated mascarpone and to the lots
produced immediately afterwards.
Results
Clinical and epidemiological data
Twenty nine individuals reported that symptoms had
begun following the consumption of mascarpone or
foods prepared with this product, which had been
purchased in various stores. Each individual was
treated with one vial of trivalent antibotulinal serum
(250 ml) after biological samples had been taken for
detection of toxins and spores; all patients were kept
under observation for 24 hours before being discharged.
Nonetheless, of the 37 individuals hospitalised for
gastroenteric and/or neurologic problems after having eaten mascarpone produced by the same plant,
only eight met the case denition (Table 1). None of
the 24 individuals who had eaten meals with these
eight individuals, yet without consuming tiramisu or
mascarpone, showed any of the symptoms included
in the case denition.
The age of patients meeting the case denition
ranged from 6 to 23 years; six of the eight patients
were males. In four of these patients, onset of the
disease 12 hours before hospitalisation occurred in a
moderate form (i.e., only gastrointestinal disturbance
with diplopia or with cephalea and asthenia, lipothymia), whereas the other four patients presented
with a serious form (gastroenteric symptoms with
ocular signs, weakness of upper limbs, and respiratory muscles involvement). The interval between
consumption of the suspected food and onset of
symptoms for these eight patients ranged from 8 to
72 hours (median 12 hours). The onset of the illness was the same for all of these patients: gastrointestinal symptoms (nausea and vomiting) followed by
neurologic symptoms. All patients manifested more
than one symptom consistent with botulism (Table 1)
and among these, the accid paralysis of muscles
innervated at least by one cranial nerve. The most

916
Table 1. Clinical, epidemiologic and laboratory aspects of eight cases of foodborne botulism in Italy
Case

Region

Age-Sex

Botulinum
toxin detected
Serum

Stool

Stool culture

Symptoms

Fatal

Nausea, diplopia, dysphagia,


ptosis, facial diplegia,
ataxia, tetra paresis,
respiratory failure
Nausea, diarrohea, asthenia,
diplopia, ptosis, dysarthria,
dysphagia, tetra paresis,
respiratory failure
Diplopia, dysarthria, dysphagia,
tetra paresis, respiratory
failure
Diplopia, dysarthria, dysphagia,
tetra paresis, respiratory
failure
Vomiting, diplopia, dysarthria,
disphagia, midriasis, tetra
paresis, respiratory failure
Vomiting, nausea, diplopia,
ptosis, ophtalmoplegia,
dysarthria, dysphagia,
midriasis, dry
mouth, respiratory failure,
paresis upper limbs
Vomiting, diplopia, ptosis,
ophtalmoplegia, dysarthria,
dysphagia, midriasis,
dry mouth, respiratory
failure, paresis upper and
lower limbs
Vomiting, disarthria,
ptosis, dysphagia, dyspnea,
hypotonia, respiratory
failure

No

Campania

23-M

C. botulinum
type A

Campania

9-F

C. botulinum
type A

Campania

14-M

n.d.

C. botulinum
type A

Campania

15-M

n.d.

C. botulinum
type A

Campania

12-M

C. botulinum
type A

Calabria

16-M

C. botulinum
type A

Calabria

18-F

C. botulinum
type A

Calabria

C. botulinum
type A

7-M

common neurologic symptoms were dysphagia, respiratory failure (all patients) diplopia (seven patients
out of eight), dysarthria, ptosis (six out of eight) and
mydriasis (four out of eight). Intubation and ventilated respiration were performed on all eight patients
meeting the case denition, including the four who
manifested respiratory diculties at a later time.
Botulism was the only disease diagnosed in six of
these patients, whereas for the other two, this diagnosis had been preceded by other clinical diagnoses
(polyradiculoneuritis, viral encephalitis). One of the
eight patients (N.S., 15 years of age, second cluster,
Campania Region) died 37 days after being hospitalised, despite the fact that he had been undergoing
mechanical respiration and had received the trivalent
anti-botulinum serum immediately upon hospitalisation. The other seven patients were also administered
the trivalent serum, some of them immediately upon
hospitalisation and others later.
The duration of hospitalisation for the seven
surviving patients was extremely long (medi-

No

No
Yes
No
No

No

No

an 56 days; range 3778 days). These patients were


interviewed 24 months after discharge, revealing that
the following symptoms were still present: exercise
intolerance (4/7 patients), general weakness (5/7 patients), shortness of the breath (4/7 patients), diculty swallowing (1/7 patients), diculty speaking
clearly (1/7 patients).
Microbiological conrmation of suspected clinical
diagnoses
Of the patients meeting the case denition, only the
sera from N.S. and G.S. (second cluster, Naples)
contained botulinum toxin type A. This serotype
toxin was also found in the stool samples of two other
patients (P.F., and V.O., respectively of second and
third cluster); spores of C. botulinum type A were
found in the stool samples or rectal swabs of all eight
patients. Biological samples from the women of the
rst cluster who had tasted the dessert were not
taken. The serum samples from patients A.P., S.O.,

917
and V.O were not analysed either because they arrived at the Botulism Reference Centre in unsuitable
conditions or because they were not taken before the
administration of the antiserum.
Serum and stool samples from all 29 patients not
matching the case denition were negative for toxin
and C. botulinum spores.
Identication of the source of intoxication
Botulinum toxin and C. botulinum spores were found
in two samples of mascarpone taken from the same
food stores where the mascarpone eaten by R.C. (rst
cluster) and G.S. (second cluster), respectively, had
been purchased. Clostridium botulinum spores were
also found in ve of the 34 samples taken at the store
in Villa Literno (a small town near Napoli), at the
warehouse of the distributor for the two regions, and
at the warehouse of the manufacturer, only were
found.
Botulinum toxin and C. botulinum type A spores
were also found in three packages of the product
taken from the stores in Vibo Valentia (third cluster)
[7] and Catanzaro (sporadic case P.Z.).
Inspection of the plant
The results of the environmental analyses failed to
reveal the presence of botulinum spores in the production line, in the waste area adjacent to the plant,
or in the plastic containers used for packaging.The
production cycle included the production of 24,000
250 g packages of mascarpone and 12,000 500 g
packages. The nal contents of the product were as
follows: 50% fat content, 4.5% protein, and 4222%
water, with a nal pH of 6.16.2. The lot of mascarpone (aw ranged from 0.945 to 0.988) involved in
the outbreak (a total of 7000 kg of the 250 and 500 g
packages) was produced in the second to last week of
July and distributed in 11 of Italy's 21 regions.

Discussion
The neuromuscular disease that aected eight individuals in three dierent cities in southern Italy had
all of the clinicalepidemiological characteristics of
an outbreak of foodborne botulism. Although
electromyography is considered to be useful in conrming clinical diagnoses, it was not available for two
of these cases. For the six cases for which EMG diagnosis was available the electromyograms showed a
presynaptic neuromuscular block compatible with
botulism.
The clinical picture of the botulinum intoxication
was conrmed by the microbiologic results. The stool
samples of all patients contained type A C. botulinum
spores. The serum sample of two of the patients

contained botulinum toxin, whereas for all other


cases, it was not possible to perform the test. The fact
that toxin was not found in the sera of most patients
(three out of ve) was attributed to various factors
and is not surprising; in fact, the literature reports
that toxin is found in only 1328% of serum samples
taken 2 days after ingestion [9]. Thus the nding of
C. botulinum spores in patients with compatible
clinical signs is considered to be sucient for diagnostic conrmation.
It has been observed that the severity of disease is
correlated with the type and quantity of toxin ingested [10]. Nonetheless, in the serum of the two
positive patients, it was not possible to determine the
quantity of toxin because of the scarcity of the material received. However, it is possible to estimate the
quantity ingested on the basis of the quantity
detected (2495 LD50/g) in a sample of mascarpone
or in the leftover tiramisu (125 LD50/g) eaten by some
of the patients [6].
The recovery of all patients was quite long, although most of them had received physiotherapy;
however, the literature describes cases in which the
course of clinical recovery was even longer than that
observed in the present study: in any case, the treating physician must take this into account when preparing the patient who is recovering from botulism.
The recovery of botulinum toxin poisoning occurs
through regeneration of the terminal and subterminal
unmyelinated nerve twigs [11, 12].
In Italy, cases of botulism have been prevalently
associated with the consumption of home-prepared
foods, especially those of vegetable origin [5]. Prior to
the episode described here, no case had ever been
associated with dairy products. In other countries
(i.e., the US, Argentina, France, and the UK), cases
of botulism associated with the consumption of dairy
products have been reported, though in the past
50 years, the number of cases associated with dairy
products seems to be quite low in industrialised
countries (seven outbreaks with 105 cases) [13].
Spores are inactivated by time/temperature parameters used for the production of sterilised milk
(UHT milk and sterile milk). In fact, thermal inactivation of non-proteolytic spores can be obtained with
less stringent combinations of time and temperature
(D100C 0.20.3), whereas for proteolytic spores,
combinations with higher values are needed
(D100C 25) [14].
Toxin is produced by the microorganism in anaerobiosis when the water activity (aw) of the medium
is higher than 0.93 and the acidity is >4.6. Toxin
production is also inuenced by storage temperature.
Specically, proteolytic species of C. botulinum and
C. butyricum are not capable of multiplying or of
producing toxins at temperatures lower than 10 C,
whereas non-proteolytic species of C. botulinum do
not grow at less than 3.3 C [14]. The ubiquity of
botulinal spores results in the potential contamina-

918
tion of many types of foods. Milk can also act as a
vehicle for the spore [15] and it has been estimated
that the level of contamination can be 1 spore/l [14].
In fact, we recently demonstrated that, of 35 milk
samples taken from the dairies that provided milk to
the plant that produced the mascarpone implicated in
the present outbreak, two contained at least 1 type B
spore per 100 ml.
Studies that have experimentally inoculated samples of mascarpone [15] have shown that the production of toxin is favoured by a great abuse in the
storage temperature. Even if refrigeration alone can
be sucient for controlling C. botulinum in mascarpone, several studies on the storage temperature of
retail food products have shown that having a storage
temperature greater than 10 C is not unusual. On
the other hand, inadequate refrigeration still has a
high incidence (19.3%) as the causal factor in foodborne infections and intoxications in Europe [16]. For
this reason, it is necessary that the production of
mascarpone include thermal treatment capable of
inactivating spores or treatments capable of removing
spores from the basic ingredients and/or combined
factors capable of inhibiting germination [17, 18].
Acknowledgements
We thank Dr Franco Santanastasi, Osservatorio
Epidemiologico Regione Campania, Dr Pasquale
Campanile, Servizio Veterinario Area 3 Dipartimento
di Prevenzione ASL CE2, Dr Talesa A, Pallone G,
Miceli M, Consoli D, Grillo P Presidio Ospedaliero
USSL n.8 di Vibo Valentia for their contribution to
the epidemiological and clinical investigations.
References
1. Smith L, Sugiyama H. Botulism: The Organism, its
Toxins, the Disease. Springeld: Charles C. Thomas,
II. 1988.
2. Hatheway CL. Clostridium botulinum and other clostridia that produce botulinum neurotoxin. In: Hauschild AHW, Dodds KL (eds), Clostridium botulinum.
Ecology and Control in Food. New York: Marcel
Dekker, 1992; 320.
3. Meng X, Karasawa T, Zou K, et al. Characterization
of a neurotoxigenic Clostridium butyricum strain isolated from the food implicated in an outbreak of foodborne type E botulism. J Clin Microbiol 1997; 35:
21602162.
4. Chaudhry R, Dhawan B, Kumar D, et al. Outbreak of
suspected Clostridium butyricum botulism in India. EID
1998; 4: 506507.
5. Aureli P, Fenicia L, Franciosa G. Classic and emergent
forms of botulism: The current status in Italy. Eurosurveillance 1999; 4: 79.

6. Aureli P, Franciosa G, Pourshaban M. Foodborne


botulism in Italy. Lancet 1996; 348: 1594.
7. Hatheway CL. Botulism. In: Balows A, Hausler WJ,
Jr., Ohashi M, Turano A (eds), Laboratory Diagnosis
of Infectious Diseases: Principles and Practice, vol 1.
New York: Springer-Verlag, 1988; 111133.
8. Kautter DA, Solomon HM, Lake DE, Bernard DT,
Mills DC. Clostridium botulinum and its toxins. In:
Vanderzant C, Splittstoesser DF (eds), Compendium of
the Methods for the Microbiological Examination
of Foods, 3rd edition, Washington: APHA, 1992;
605621.
9. Woodro BA, Grin PM, McCroskey LM, et al.
Clinical and laboratory comparison of botulism from
toxin type A, B, and E in the United States, 19751988.
J Infect Dis 1992; 166: 12811286.
10. Donadio JA, Gangarosa EJ, Faich GA. Diagnosis and
treatment of botulism. J Infect Dis 1971; 124: 108112.
11. Angaut-Petit D, Molgo J, Comella JX, Faille L, Tabti
N. Terminal sprouting in mouse neuromuscular
junctions poisoned with botulinum type A toxin:
Morphological and electrophysiological features.
Neuroscience 1990; 37: 799808.
12. Mann JM, Martin S, Homan R, Marrazzo S. Patient
recovery from type A botulism: Morbidity assessment
following a large outbreak. Am J Public Health 1981;
71: 266269.
13. Collins-Thompson DL, Wood DS. Control in dairy
products. In: Hauschild AHW, Dodds KL (eds),
Clostridium botulinum. Ecology and Control in Food,
New York: Marcel Dekker, 1992; 261277.
14. Skovgaard N. Facts and trends in microbial contamination of dairy products. International Dairy Federation Bulletin 1990; 250: 3133. International Dairy
Federation, Square Vergote 41,1040 Bruxelles, Belgium.
15. Franciosa G, Pourshaban M, Gianfranceschi M, et al.
Clostridium botulinum spores and toxin in mascarpone
cheese and other milk products. J Food Prot 1999; 62:
867871.
16. WHO surveillance programme for control of foodborne infections and intoxications in Europe. Sixth
Report 19901992. Schmidt K (ed.) Federal institute
for health protection of consumers and veterinary
medicine, Berlin 1995.
17. ter Steeg PF, Cuppers HGAM. Growth of proteolytic
Clostridium botulinum in process cheese products: II.
Predictive modeling. J Food Prot 1995; 58: 11001108.
18. ter Steeg PF, Cuppers HGAM, Hellemons JC, Rijke G.
Growth of proteolytic Clostridium botulinum in process
cheese products: I. Data acquisition for modeling the
inuence of pH, sodium chloride, emulsifying salts, fat
dry basis, and temperature. J Food Prot 1995; 58:
10911099.

Address for correspondence: D. Greco, Laboratorio di Epidemiologia e Biostatistica, Istituto Superiore di Sanita,
Viale Regena Elena 299, 00161, Rome, Italy
Phone: +39-06-49903390; Fax: +39-06-49387069
E-mail: greco@iss.it