1 PHLN SUMMARY LABORATORY DEFINITION
1.1.1 Definitive Criteria
- Isolation of toxigenic Corynebacterium diphtheriae or C. ulcerans (confirmed by specialist laboratory).
1.1.2 Suggestive Criteria
- Isolation of C. diphtheriae (toxin production unknown).
1.1.3 Guide for Use
Nucleic acid tests have been developed in house for the detection of diphtheria toxin genes in clinical specimens but are not acceptable for the purpose of notification in Australia.
Serological assays, including EIA for diphtheria antitoxin antibodies are available in Australia but are used for immune status studies not for diagnosis of diphtheria.
Classical diphtheria is an acute infectious disease of the upper respiratory tract caused by toxigenic Corynebacterium diphtheriae
. The bacterium multiplies in the pharynx and produces an exotoxin, diphtheria toxin, which is encoded on a lysogenic bacteriophage. Diphtheria toxin is readily absorbed through mucous membranes and disseminated through the body, targeting the heart and peripheral nervous system. C. diphtheriae
may also cause skin infections, particularly in tropical climates, and invasive disease including endocarditis 17
, bacteraemia, septic arthritis and splenic abscesses1
. Most strains causing skin and invasive disease are non-toxigenic unless associated with an outbreak of pharyngeal diphtheria.
Top of page
The highly effective diphtheria toxoid was licensed, and school based immunisation campaigns commenced, in the early 1930’s in Australia and had an immediate impact on disease rates. Small foci of classical diphtheria persisted in disadvantaged areas of Australian capital cities and rural areas until the early seventies. The last known outbreak in Australia occurred in Alice Springs and Katherine in NT during 1991–1992 9,15
, involving 23 cases with one death. Since then occasional cases have occurred in travelers entering Australia from endemic areas overseas.
Since immunisation with diphtheria toxoid does not prevent carriage of C. diphtheriae
, it remains endemic in Australia and is occasionally cultured from upper respiratory and wound swabs, particularly from Aboriginal people. These isolates and those causing endocarditis, which also occurs infrequently in Australia, are nearly always non-toxigenic.
Toxigenic strains of C. diphtheriae
have been isolated from respiratory and skin infections of horses and cattle, particularly when outbreaks of human disease are occurring but it is unknown whether the cattle are infected by humans or whether animals may be a reservoir for human disease11
The veterinary pathogens C. ulcerans
and C. pseudotuberculosis
can be lysogenised by the C. diphtheriae
bacteriophage and thus produce diphtheria toxin. Toxigenic C. ulcerans
can cause classical diphtheria. The majority of human isolates of C. ulcerans
referred to the Health Protection Agency (formerly PHLS) in UK between 1986 and 2002 were toxigenic 18
. Ribotypes of C. ulcerans
strains isolated from cats were indistinguishable from those of human strains18
The diagnosis of pharyngeal diphtheria usually depends on clinical suspicion and a request for throat swab culture on appropriate media. In Australia, where doctors are unfamiliar with the clinical characteristics of diphtheria and few, if any, laboratories routinely culture throat swabs on selective media for C. diphtheriae
, the diagnosis is likely to be delayed. If there is a strong clinical suspicion of diphtheria treatment should not wait for laboratory confirmation. Infections with toxigenic strains of both C. diphtheriae
and C. ulcerans
are notifiable in Australia.
Further information may be found in the WHO publication: Manual for the Laboratory Diagnosis of Diphtheria http://whqlibdoc.who.int/euro/1994-97/ICP_EPI_038_(C).pdf
Although C. diphtheriae
and C. ulcerans
are not fastidious bacteria and grow well on media routinely used in a medical microbiology laboratory, special media containing tellurite are used to aid in distinguishing them from other flora present in swabs from non-sterile sites. Pai (egg based) or Loeffler (serum based) slants should also be inoculated. Smears made from growth after 18 hr incubation at 35 oC on these lipid rich media will show the distinctive cell morphology of C. diphtheriae
when stained with Loeffler’s alkaline methylene blue or Neisser’s stain. Respiratory specimens should also be inoculated onto blood agar to detect haemolytic streptococci which can cause symptoms which may be confused with diphtheria.
3.1.1 Suitable and unsuitable specimens
Routine throat swabs, nasopharyngeal aspirates or swabs, membrane tissue, wound swabs sterile site tissue. Swabs may be placed in bacterial transport media such as Stuart’s if transport is likely to be delayed although the organism is not fastidious and has been shown to survive on fomites for up to 6 months. However, special media are usually required for the culture of C. diphtheriae
and most microbiologists have had little or no experience in the diagnosis of diphtheria. Therefore the laboratory should be alerted that a specimen for diphtheria is on its way so that media can be prepared or the specimen referred to a reference laboratory.
Two formulations of tellurite media are commonly used for the primary isolation of C. diphtheriae
: Tinsdale’s which contains tellurite and cystine and Hoyle’s which contains tellurite without cystine. Both media should be freshly prepared and have a shelf life of approximately 7 days. Colonies of C. diphtheriae
are greyish black (gunmetal grey) and range in size from 1–3mm on cystine tellurite agar after 24–48 hours incubation.
3.1.3 Test sensitivity
Unknown, but C. diphtheriae
and C. ulcerans
are readily cultured on solid media from throat swabs and membrane, despite some inhibition of C. diphtheriae
by tellurite, albeit less than of normal flora. Sensitivity may be improved by culture on Loeffler’s or Pai media.
Top of page
3.1.4 Test specificity
Although tellurite is inhibitory for much of the normal upper respiratory tract flora, some non-corynebacteria will grow on media containing tellurite. Suspect colonies should be Gram stained to confirm diphtheroid morphology and subcultured to Loeffler’s or Pai slants to demonstrate the presence of the distinctive metachromatic granules and to provide inoculum for biochemical identification tests and toxin detection.
3.1.5 Predictive values and relevant populations
High during outbreaks but may be missed in sporadic cases if the laboratory is not prepared for investigating a case of diphtheria. Although C. diphtheriae
grows well on routine laboratory media it may be difficult to distinguish among normal respiratory flora and especially by inexperienced microbiologists. The laboratory must be notified and given a full clinical history if diphtheria is suspected clinically.
3.1.6 Suitable test acceptance criteria
Non-motile, Gram positive rods. C. diphtheriae
colonies show typical morphology on tellurite media. Typical cell morphology in methylene blue stained smears of growth on lipid rich media. Ferment glucose and maltose, reduce nitrate, urease negative, do not hydrolyse arginine. C. ulcerans
has similar colony morphology but does not produce metachromatic granules, is urease positive and does not reduce nitrate.
3.1.7 Suitable internal controls
A control strain of C. diphtheriae
must be cultured along with the unknown strain.
3.1.8 Suitable external QAP programme
RCPA QAP (rarely).
3.1.9 Special considerations
Most microbiologists will not have investigated a case of diphtheria although they may have isolated C. diphtheriae
or C. ulcerans
from a cutaneous lesion or heart valve. If a clinician suspects diphtheria the laboratory must be notified so they can prepare special isolation media and obtain a control organism if possible. The specimens may need to be referred to a more specialized laboratory.
Presumptive isolates must be tested for diphtheria toxin production by phenotypic and/or molecular methods. Multiple colonies should be subcultured to Loeffler’s or Pai slants for further testing since toxigenic and non-toxigenic colonies can be isolated from the same specimen
3.2 Diphtheria Toxin Detection
Traditional methods for toxigenicity testing are the in vitro plate method (Elek Plate), cell culture method in Vero cells and in vivo method in guinea pigs. All rely on a supply of high quality antitoxin, usually raised in rabbits, and are tedious and slow to perform, usually taking 1–4 days for a result. Only tertiary laboratories in Australia and elsewhere are likely to have the experienced staff and facilities required to obtain reliable results.
The guinea pig method remains the gold standard assay for toxigenicity and relies on the ability of diphtheria antitoxin to protect a guinea pig from a lethal injection of the test isolate which should kill an unprotected guinea pig if the isolate is toxigenic. If both challenged animals survive the isolate is non-toxigenic; if both die within 1–4 days the isolate may not be C. diphtheriae
. This test is rarely used these days.
Rapid serological methods including EIA have been developed in house for detecting toxin in isolates (Engler et al 00). An immunochromatographic strip test method (ICT), introduced in 2002 8
, is sensitive and specific for detecting toxin in specimens and isolates. These methods are not available commercially.
The Elek test is the most commonly performed phenotypic test for toxin detection and is discussed below. Further information about other tests is available 3
and WHO Manual for Laboratory Diagnosis of Diphtheria http://whqlibdoc.who.int/euro/1994-97/ICP_EPI_038_(C).pdf
In 1991 Pallen 14
described a PCR based assay for detecting diphtheria toxin genes in isolates. PCR has now largely superceded phenotypic testing for toxigenicity and has been adapted to direct detection in specimens to provide a rapid presumptive laboratory diagnosis 13,12
3.2.1 In vitro toxin test (Elek Plate)
The original assay was first described by Elek in 1948 4
and modified in 1949 5
. It virtually replaced the guinea pig in vivo test in most laboratories, but has itself now been largely replaced by molecular testing except in countries with a relatively high rate of diphtheria. More recent modifications to the conventional Elek test have reduced the time required to perform the assay from 48 hours to 16–24 hours 3
Top of page
18.104.22.168 Suitable and unsuitable specimens
Single colonies should be used as inocula, although earlier protocols state that mixed cultures and the original swab can be used to inoculate an Elek plate. Several colonies from each primary isolation plate should be tested as both toxigenic and non-toxigenic isolates can be present in a single specimen.
22.214.171.124 Test details
The assay is an immunoprecipitation test performed in clear peptone agar containing foetal calf serum in Petri dishes. In the original assay a strip of filter paper impregnated with diphtheria antitoxin is laid onto the centre of the dish and the test and control isolates are inoculated in single streaks up to the antitoxin filter paper strips (WHO manual). A modification 3,6
uses filter discs impregnated with antitoxin and single point inocula 7mm from them. Diphtheria toxin produced by toxigenic strains diffuses into the agar and forms a line of precipitation when it meets antitoxin diffusing from the impregnated strips.
126.96.36.199 Test sensitivity
Highly variable and is influenced by the experience of the staff, the test protocol and the quality of the reagents. The components of the Elek base medium must be standardised, prepared with extreme care and tested with toxigenic controls. The optimal dilution of each batch of antitoxin must be determined. The depth of the agar and distance of inocula from filter paper impregnated with antisera will influence the results 3
188.8.131.52 Test specificity
Purity of the antitoxin will influence the number of non-specific precipitin lines and must be obtained from a reliable source.
184.108.40.206 Predictive values and relevant populations
UK surveys in 1993 and 1995 showed that laboratories were more likely to obtain correct results for toxigenic survey strains than for non-toxigenic strains – possibly owing to misinterpretation of non-specific precipitation lines 3
220.127.116.11 Suitable test acceptance criteria
Toxigenic control strains produce expected lines of precipitation.
18.104.22.168 Suitable internal controls
Each plate must be inoculated with known toxigenic and non-toxigenic control strains.
22.214.171.124 Suitable original test validation criteria
Fully referenced method.
126.96.36.199 Suitable External QAP Programme
RCPA QAP (rarely).
3.2.2 Nucleic acid detection
PCR assays were introduced originally to confirm the toxigenic status of isolates 14
. In 1997 Nakao and Popovic 13
described a PCR assay to amplify toxin A and B subunit fragments directly from clinical specimens with the products being visualized on agarose electrophoresis gels. In 2002 Mothershed et al 12 adapted the assay to real time PCR technology to provide a more rapid result.
Most primers to confirm toxigenicity status in isolates have been designed to target the biologically active (fragment A) subunit although other targets have been used successfully. For real time direct detection assays primers should be designed to target at least 2 gene fragments (usually A and B subunits of the toxin gene).
There are no commercial PCR kits available.
188.8.131.52 Suitable and unsuitable specimens
Clinical specimens include throat and nasopharyngeal swabs and membrane tissue for direct detection assays. Several colonies of C. diphtheriae
/ C. ulcerans
/ C. pseudotuberculosis
from each isolation plate should be tested to confirm their toxigenicity status since both toxigenic and non-toxigenic isolates may be present.
184.108.40.206 Test details
Refer to published literature for appropriate protocols.
Top of page
220.127.116.11 Test sensitivity
Very high correlation with the gold standard Elek plate assay for toxigenic isolates of C. diphtheriae
. Real time PCR assays may not detect the presence of toxigenic C. ulcerans
owing to variability in toxin genes (tox
) of this species. Standard PCR assays which do not rely on probe hybridization for detection are recommended for C. ulcerans 2,16
The efficiency of the nucleic acid extraction method is important for direct detection of tox in specimens.
18.104.22.168 Test specificity
Highly specific for diphtheria toxin genes – however, toxin gene positive strains may not produce toxin.
22.214.171.124 Predictive values and relevant populations
Negative predictive value for non-toxin producing strains of C. diphtheriae
is very high, lower for strains of C. ulcerans
which produce toxin but whose toxin genes may not be detected by the particular PCR assay.
Positive predictive value is influenced by the ability of the tox positive strains to produce detectable biologically active toxin.
126.96.36.199 Suitable test acceptance criteria
All controls within expected ranges.
188.8.131.52 Suitable internal controls
As recommended in NPAAC guidelines: Laboratory Accreditation Standards and Guidelines for Nucleic Acid Detection and Analysis.
Controls should be designed to detect sample inhibitory activity.
Control strains of toxigenic and non- toxigenic C diphtheriae and C. ulcerans
should be included.
184.108.40.206 Suitable original test validation criteria
In accordance with NPAAC guidelines: Requirements for the Validation of In House In vitro Diagnostic Devices. Refer to published papers for specific data.
220.127.116.11 Special considerations
PCR is a useful test to screen isolates for toxigenicity but tox
positive strains should be referred to a specialist laboratory for confirmation by molecular and biological tests and for subtyping.
PCR for direct detection of tox
in clinical specimens should only be used by experienced laboratories and, preferably, only during outbreaks when the assay has been shown to detect the causal strain/s.
Diagnosis of diphtheria by direct PCR on specimen is not notifiable in Australia although a positive result should be communicated by telephone to public health authorities if diagnosis by culture techniques is delayed.
4 Typing and Subtyping Methods
There are four biotypes of C diphtheriae
, var. mitis
, var. belfanti
and var. intermedius
) classified on morphologic and biochemical features. Laboratories rarely report biotype data these days as the information is of limited use for epidemiological investigations and experienced staff are rarely available to determine the biotype.
Ribotyping has now replaced biotyping and provides good discrimination for epidemiological studies. An international nomenclature for C diphtheriae
ribotypes has been published 10
Top of page
5 References 1.
Belko J, Wessel MD and Malley MD. 2000. Endocarditis caused by Corynebacterium diphtheriae
: case report and review of the literature. Paediatr Infect Dis J 19
Cassidy PK, Pawlooski et al. 2008 Analysis of toxigenic Corynebacterium ulcerans strains revealing potential for false-negative real-time PCR results. J Clin Microbiol 46
Efstratiou A, Engler KH et al. 2000. Current approaches to the laboratory diagnosis of diphtheria. J Infect Dis 181 (Suppl 1)
Elek SD. 1948. Title unknown. BMJ 1
Elek SD. 1949. The plate virulence test for diphtheria. J Clin Pathol 2
Engler KH, Glushkevich T et al. 1997. A modified Elek test for detection of toxigenic corynebacteria in the diagnostic laboratory. J Clin Microbiol 35
Engler KH, Efstratiou A. 2000. A rapid enzyme immunoassay for the detection of diphtheria toxinamong clinical isolates of Corynebacterium spp. J Clin Microbiol 38
Engler KH, Efstratiou A et al. 2002. Immunochromatographic strip test for rapid detection of diphtheria toxin: description and multicenter evaluation in areas of low and high prevalence of diphtheria. J Clin Microbiol 40
Gidding HF, Burgess MA, Gilbert GL. Diphtheria in Australia, recent trends and future prevention strategies. Comm Dis Intell 2000;24:165-7.
Grimont PA, Grimont F et al. 2004. International nomenclature for Corynebacterium diphtheriae
ribotypes. Res Microbiol 155
Henricson B, Segarra J et al. 2000. Toxigenic Corynebacterium diphtheriae
associated with an equine wound infection. J Vet Diagn Invest 12
Mothershed E, Cassidy P et al. 2002. Development of real-time fluorescence PCR assay for rapid detection of the diphtheria toxin gene. J Clin Microbiol 40: 4713-4719.
Nakao H, Popovic T. 1997. Development of a direct PCR assay for detection of the diphtheria toxin gene. J Clin Microbiol 35
Pallen MJ. 1991. Rapid screening for toxigenic Corynebacterium diphtheriae
by the polymerase chain reaction. J Clin Pathol 44
Patel M, Morey F, Butcher A, et al. The frequent isolation of toxigenic and non-toxigenic Corynebacterium diphtheriae
at Alice Springs Hospital. Commun Dis Intell 1994; 18: 310-311.
Sing A, Hogardt M et al. 2003. Detection of differences in the nucleotide and amino acid sequences of diphtheria toxin from Corynebacterium diphtheriae
and Corynebacterium ulcerans causing extrapharyngeal infections. J Clin Microbiol 41
Tiley SM, Kociuba KR et al. 1993. Infective endocarditis due to nontoxigenic Corynebacterium diphtheriae
: report of seven cases and review. Clin Infect Dis 16
De Zoysa A, Fry N et al. 2003. Characterisation of toxigenic Corynebacterium ulcerans strains isolated in the UK from humans and domestic cats. Clin Microbiol and Infect 9 (Suppl 1)