Joint FAO/WHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotics III Food London, Ontario,

Canada, April 30 and May I,2002

Food and Agriculture Organization of the United Nations

World Health Organization

Guidelines for the Evaluation of Probiotics in Food

Report of a Joint FAOiWHO

Working Group on

Drafiting Guidelines for the Evaluation of Probiotics in Food

London Ontario, Canada April 30 and May I,2002

The opinions expressed in this report are those of the participants of the Working Group and do not imply any opinion on the part of FAO and WHO

Joint FAOlWHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotics in Food London, Ontario, Canada, April 30 and May 1,2002

1. Introduction
The Joint FAOAVHO Expert Consultation on Evaluation of Health and Nutritional Properties of Probiotics in Food held in Cordoba, Argentina from 1-4 October, 2001 recognized that there is a need for guidelines to set out a systematic approach for the evaluation of probiotics in food leading to the substantiation of health claims. Consequently, a Working Group was convened by FAOrWHO to generate guidelines and recommend criteria and methodology for the evaluation of probiotics, and to identify and define what data need to be available to accurately substantiate health claims. The aims of the Working Group were to identify and outline the m inimum requirements needed for probiotic status. Consequently, guidelines were prepared to meet this objective.

2. Scope
The report of the Joint FAOBVHO Expert Consultation on Evaluation of Health and Nutritional Properties of Probiotics in Food addressedprobiotics in food, and excluded reference to the term biotherapeutic agents, and beneficial m icroorganisms not used in food (httn://www..fao.or.&s/ESN/Probio/probio.htm). The Working Group adopted the same scope as the Expert Consultation, and specifically excluded genetically modified organisms (GMOs).

3.

Guidelines for Probiotics

In order to claim that a food has a probiotic effect, the guidelines set forth in this report should be followed. A scheme outlining these guidelines for the evaluation of probiotics for food use is shown in Fig. 1. This was the basis for discussions and details are specified in the fohowing sections of this report. 3.1. Genus/species/strain

It was recognized that it is necessary to know the genus and species of the probiotic strain. The current state of evidence suggests that nrobiotic effects are str& specific. _ Strain identity is important to link a strain to a specific health effect as well as to enable accurate surveillance and epidemiological studies. A possible exception is the ability in general of S. themojyhilus and L. delbrueckii ssp. bulguricus to enhance lactose digestion in lactose intolerant individuals. In this case, or i &er cases where there is s&a& SC subs&&a~f health benefits that are not strain specific, individual strain identity is not critical. Speciation of the bacteria must be established using the most current, valid methodology. It is recommended that a combination of phenotypic and genetic tests be used. /

Joint FAOlWHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotics m Food London, Ontario, Canada, April 30 and May 1, 2002

Figure 1. Guidelines for the Evaluation of Probiotics for Food Use

l Genus, species, strain

Safety assessment (Detailed in

Double blind, randomized. placebo-controlled (DBPC) phase 2 human trial other appropriate or design with sample size and primary outcome appropriate to determine if strain/product is efficacious (Detailed in Section 3.4)

1 mysecond ent DBPC n.-.e.4z-

J
Phase 3, effectiveness trial is appropriate to compare probiotics with standard treatment of a specific condition
I I

Probiotic Food
I

Labeling (Detailed in Section 3.5) *Contents - genus, species, strain designation

l M inimum numbers of viable bacteria at end of shelf-life

l Corporate contact details for

consumer information.

Jomt FAO/WHO Working Group Report on Draftmg Guidelmes for the Evaluation of Probiotics tn Food London, Ontario, Canada, April 30 and May I, 2002

Nomenclature of the bacteria must conform to the current, scientifically recognized names. Protracted use of older or m isleading nomenclature is not acceptable on product labels. The use of incorrect names does not properly identify the probiotic bacterium in the product and forces consumers and regulatory agencies to make assumptions about the identity of the real bacterium being sold. Current nomenclature can be retrieved as follows:
->
l

-3

Approved Lists of Bacterial Names (Int. J. Syst. Bacterial, 1980,30:225-420) also available in http://www.bacterio.cict.fr/ Validation Lists, published in the International Journal of Systematic and Evolutionary M icrobiology (or International Journal of Systematic Bacteriology, prior to 2000)

DNA-DNA hybridization is the reference method to specify that a strain belongs to a species; however, as it is time consuming and beyond the resources of many laboratories, requiring a large collection of reference strains, the use of DNA seauences encodinp IhS ,rRNA is suggesl . In this case, it is recommended that this genotypic technique be combined with phenotypic tests for confirmation. Patterns generated from the fermentation of a range of sugars and final fermentation products obtained from glucose utilization are key phenotypes that should be investigated for identification purposes. Strain typing has.to be performed with a reproducible genetic method or using a unique phenotypic trait. Pulsed Field Gel Electrophoresis (PFGE) is the gold standard. Randomly A m lified Polymorphic DNA (RAPD) can also be used, but is less reproducible. eDetermination of the presence of extrachromosomal genetic elements, such as plasmids can contribute to strain typing and characterizationi It is recommended that all strains be deposited in an internationally recognized culture collection. 3.2. In vitro lests to screen potential probiotics.

In vitro tests are critical to assessthe safety of probiotic m icrobes (see Section 3.3). In addition, in vitro tests are useful to gain knowledge of strains and the mechanism of the probiotic effect. However, it was noted that the currently available tests are not fully adequate to predict the functionality of probiotic m icroorganisms in the human body. It was also noted that in vitro data available for particular strains are not sufficient for describing them as probiotic. Probiotics for human use will require substantiation of efficacy with human trials. Appropriate target-specific in vitro tests that correlate with in vivo results are recommended. For example, in vitro bile salts resistance was shown to correlate with gastric survival in vivo (Conway et al., 1987). A list of the main currently

Joint FAO/WHO Working Group Report on Draftmg Guidelines for the Evaluation of Problotics in Food London, Ontario, Canada, April 30 and May I,2002

used in vitro te,stsfor the study of probiotic strains is shown in Table 1. All of these tests require validation, however, with in vivo performance. Table 1. Main currently used in vitro tests for the study of probiotic strains Resistance to gastric acidity Bile acid resistance Adherence to mucus an&or human epithelial cells and cell lines Antimicrobial activity against potentially pathogenic bacteria Ability to reduce pathogen adhesion to surfaces Bile salt hydrolase activity Resistance to spermicides (applicable to probiotics for vaginal use) Safety considerations: Requirementsfor proof that a probiotic strain is safe and 3.3. without contamination in its delivery form Historically, laccobacilli and bifidobacteria associated with food have been considered to be safe (Adams & Marteau, 1995). Their occurrence as normal commensals of the mammalian flora and their established safe use in a diversity of foods and supplement products worldwide supports this conclusion. However, probiotics may theoretically be responsible for four types of side-effects (Marteau, 2002): Ii 1. Systemic infections 2. Deleterious metabolic activities 3. Excessive immune stimulation in susceptible individuals 4. Gene transfer Documented conrelations between systemic infections and probiotic consumption are few and all occurred in patients with underlying medical conditions. The following is a list (including some m icrobes used in non-food applications) of infections reported to be associated (although not necessarily proven) with the consumption of commercial products: Two cases of I.. rhamnosus traced to possible probiotic consumption (Rautio et al., 1999; Mackay et al., 1999). 9 Thirteen case:; of Saccharomyces fungemia due to vascular catheter contamination (Hennequin et al., 2000). Bacillus infections linked to probiotic consumption include three reports (Spinosa et al., 2000; Oggioni et al., 1998; Richard et al., 1988) detailing seven cases of R. subtilis bacteremia, septicemia and cholangitis, all in patients with underlying disease. -4N o cases of infections from Bifidobacterium have been reported. Enterococcus is emerging as an important cause of nosocomial infections and isolates are increasingly vancomycin resistant. The Working Group recognizes that some strains of Enlerococcus display probiotic properties, and may not at the point of inclusion in a product display vancomycin resistance. However, the onus is on the producer to prove that any given probiotic strain is not a significant risk with regard to transferable antibiotic resistance or other opportunistic virulence properties.
n n

II

Joint FAO/WHO

Working Group Report on Drafting Guidelines for the Evaluation of Probiotics in Food London, Ontario, Canada, April 30 and May I,2002

In recognition of the importance of assuring safety, even among a group of bacteria that is Generally Recognized as Safe (GRAS), the Working Group recommends that probiotic strains be characterized at a m inimum with the following tests: J d J 1. Determination of antibiotic resistance patterns 2. Assessment of certain metabolic activities (e.g., D-lactate production, bile salt deconjugation) 3. Assessment of side-effects during human studies 4. Epidemiological surveillance of adverse incidents in consumers (post-market) 5. If the strain under evaIuation belongs to a species that is a known mammalian toxin producer, it must be tested for toxin production. One possible scheme for testing toxin production has been recommended by the EU Scientific Committee on Animal Nutrition (SCAN, 2000) 6. If the strain under evaluation belongs to a species with known hemolytic potential, determination of hemolytic activity is required Assessment of lack. of infectivity by a probiotic strain in immunocompromized animals would add a measure of confidence in the safety of the probiotic. 3.4. In vivo studies using animals and humans In some cases, animal models exist to provide substantiation of in vitro effects and determination of probiotic mechanism. Where appropriate, the Working Group encourages use of these prior to human trials. The principal outcome of efficacy studies on probiotics should be proven benefits in human trials, such as statistically and biologically significant improvement in condition, symptoms, signs, well-being or quality of life; reduced risk of disease or longer time to next occurrence; or faster recovery from illness. Each should have a proven correlation with the probiotic tested. Probiotics have been tested for an impact on a variety of clinical conditions (see Expert Consultation Report, Section 5.3). Standard methods for clinical evaluations are cornprized of Phase 1 (safety), Phase 2 (efficacy), Phase 3 (effectiveness) and Phase 4 (surveillance). Phase 1 studies focused on safety are discussed in Section 3.3 above. Phase 2 studies, generally in the form of randomized, double blind, placebo-controlled (DBPC) design, measure efficacy compared with placebo. In addition, phase 2 studies measure adverse effects. A general recommendation for the testing of probiotic foods is that the placebo would be cornprized of the food carrier devoid of the test probiotic. Sample size needs to be calculated for specific endpoints. Statistically significant differences must apply to biologically relevant outcomes. Probiotics delivered in food generally are not tested in Phase 3 studies, which are concerned with comparison with a standard therapy. When a claim is made for a probiotic altering a disease state, the claim should be made based on sound scientific evidence in human subjects.

Joint FAO/WHO W orking Group Report on Drafting Guidelines for the Evaluation of Probiotics in Food London, Ontario, Canada, April 30 and May 1,2002

In Phase 2 and 3 studies, the Working Group recognizes the value of validated quality of life assessment tools, It is recommended that human trials be repeated by more than one Center for confirmation of msults. No adverse effects related to probiotic administration should be experienced when food is considered. Adverse effects should be monitored and incidents reported. The Working Group recommends that information accumulated to show that a strain(s) is a probiotic, including clinical trial evidence be published in peer-reviewed scientific or medical journals. Furthermore, publication of negative results is encouraged as these contribute to the totality of the evidence to support probiotic efficacy. Further information on the generation and use of clinical information to substantiate health effects can be found at www.ftc.gov/bcp/conline/pubs/buspubs/dietsupp.htm#IIb
3.5. Health claims and labeling

Currently in most countries, only general health claims are allowed on foods containing probiotics. The Working Group recommends that specific health claims on foods be allowed relating to the use of probiotics, where sufficient scientific evidence is available, as per the guidelines set forth in this report. Such specific health claims should be permitted on the label and promotional material. For example, a specific claim that states that a probiotic reduces the incidence and severity of rotavirus diarrhea in infants would be more informative to the consumer than a general claim that states improves gut health This would better comply with Codex General Guidelines on Claims (CAC/GL . I- 1979 (Rev. I- 19!J1) to avoid m isleading information. It is recommended that it be the responsibility of the product manufacturer that an independent third party review by scientific experts in the field be conducted to establish that health claims are truthful and not m isleading. The Working Group recommends that the following information be described on the label: Genus, species and strain designation. Strain designation should not m islead consumers about the functionality of the strain 8 M inimum viable numbers of each probiotic strain at the end of the shelf-life The suggested serving size must deliver the effective dose of probiotics related to the health claim = Health claim(s) . Proper storage conditions . Corporate contact details for consumer information
n

Joint FAO/WHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotlcs in Food London, Ontario, Canada, April 30 and May I,2002

4.

Recommendations
Live m icroorganisms which when 1. Adoption of the definition of probiotics as administered in adequate amounts confer a health benefit on the host . 2. Use and adoption of the guidelines in this report should be a prerequisite for calling a bacterial strain probiotic . 3. Regulatory framework to allow specific health claims on probiotic food labels, in cases where scientific evidence exists, as per the guidelines set forth in this report. 4. Promotion of these guidelines at an international level. 5. Good manufacturing practices (GMP) must be applied in the manufacture of probiotic foods with quality assurance,and shelf-life conditions established. 6. Further development of methods (in vitro and in viva) to evaluate the functionality and safety of probiotics.

Joint FAOlWHO Working Group Report on Draftmg Guidelines for the Evaluation of Probiotlcs in Food London, Ontario, Canada, April 30 and May I,2002

5.

List of Abbreviations
Codex Alimentarius Commission/General Guidelines on Claims

CAUGL:

DBPC: Double blind, randomized, placebo-controlled DNA: Deoxyribonucleic Acid FAO: Food and Agriculture Organization of the United Nations GMO: Genetically Modified Organism GMP: Good manufacturing practices GRAS: Generally Recognized as Safe PFGE: Pulsed Field Gel Electrophoresis RNA : Ribonuclei Acid c RAPD: Randomly Amplified Polymorphic DNA SCAN: EU Scientific Committee on Animal Nutrition WHO: World Health Organization

Joint FAO/WHO Working Group Report on Drafting Guidelmes for the Evaluation of Probiotics in Food London, Ontario, Canada, April 30 and May 1, 2002

6. References
Adams MR, Marteau P (1995): On the safety of lactic acid bacteria. Int J Food M icro, 27: 263-264. Conway PL, Gorbach SL, Goldin BR (1987): Survival of lactic acid bacteria in the human stomach and adhesion to intestinal cells. J Dairy Sci, 70: l-12. Hennequin C, Kauffmann-Lacroix C, Jobert A, Viard JP, Ricour C, Jacquemin JL, Berche P (2000): Possible role of catheters in Saccharomyces boulardii&ngemia. Eur J Clin M icrobial Infect Dis, 19: 16-20. Mackay AD, Taylor MB, Kibbler CC, Hamilton-Miller JMT (1999): Lactobacillus endocarditis caused by a probiotic organism. Clin M icrobial Infect, 5: 290-292. Marteau P (2002): Safety aspects ofprobioticproducts. Stand J Nutr, (In Press).

Oggioni MR, Pozzi G, Balensin PE, Galieni P, Bigazzi C (1998): Recurrent septicemia in an immunocompromised patient due to probiotic strains of Bacillus subtilis. J Clin M icrobial, 36: 325-326. Rautio M , Jousimies-Somer H, Kauma H, Pietarinen I, Saxelin M , Tynkkynen S, Koskela M (1999): Liver abscess due to a Lactobacillus rhamnosus strain indistinguishable from L. rhamnosus strain GG. Clin Infect Dis, 28: 1159-60. Richard V, Auwem P, Snoeck R, Daneau D, Meunier F (1988): Nosocomial bacteremia caused by Bacillus species. Eur J Clin M icrobial Infect Dis, 7: 783-785. SCAN (2000): Report of the Scientific Committee on Animal Nutrition on the Safety of Use of Bacillus Species in Animal Nutrition. European Commission Health & Consumer Protection Directorate-General. > http://europa.eu.int/comrn/food/fs/sc/scan/out4 1.pdf Spinosa MR, Wallet F, Courcol RJ, Oggioni M R (2000): The trouble in tracing opportunistic pathogens: cholangitis due to Bacillus in a French hospital caused bv a strain related to an Italian probiotic. 7 M icrob Ecol Health Dis, 12: 99- 10 1.

10

Joint FAO/WHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotics In Food London, Ontario, Canada, April 30 and May I, 2002

Annex 1

List of Participants
ARAYA, Magdalena, Institute of Nutrition - Macul, Santiago, Chile Tel: +56 2 678 1468 Fax: +56 2 221 4030 E-mail: maravaO,uec.inta.uchile.cl MORELLI, Lorenzo, Istituto Piacenza, Italy Tel: +39 0523 599248 Fax: +39 0523 599246 E-mail: morelli@pc.unicatt.it and Food Technology (INTA), Macul 5540

di M icrobiologia UCSC, Via Emilia Partnense 84,29 100

REID, Gregor, Lawson Health Research Institute, 268 Grosvenor St, London, Ontario N6A 4V2, Cana (Chairperson) da Tel: +1 519 64616100 x 65256 Fax: +I 519 646 6110 E-mail: gregor@uwo.ca SANDERS, Mary Ellen, Dairy and Food Culture Technologies, 7119 S. Glencoe Ct., Centennial, CO 80122, USA Tel: +l 303 793 9974 Fax: +I 303 771 6201 E-mail: mesanders@,msn.com STANTON, Catherine, Teagasc, Dairy Products Research Centre, Moorepark, Fermoy, Co Cork, lreland (Rapporteur) Tel: +353 25 42442 Fax: +353 25 42340 E-mail: cstanton@,moorepark.teagasc,ie FAO/WHO SECRETARIAT PINEIRO, Mayal, Food Quality and Standards, Food and Nutrition Division, FAO, Via Delle Terme di Caracalla, Rome 00 100, Italy Tel: +39 06 570 53308 Fax: +39 06 570 54593 E-mail: mava.pineiro@,fao.org BEN EMBAREX, Peter, Food Safety Program, WHO, 20, Avenue Appia, CH- I2 11 Geneva 27, Switzerland Tel: +41 22 791 42 04 Fax: +41 22 791 48 07 E-mail: benembarekp@,who.int