Screening tests for the presumptive identification of Corynebacterium diphtheriae in a diagnostic laboratory.

We were concerned to read the recent article by Pennie et al. (7) on the misidentification of toxigenic Corynebacterium diphtheriae as a Corynebacterium sp. with low virulence in a child with endocarditis. The paper raises several issues relating to current methods for the microbiological diagnosis of diphtheria. Firstly, specific guidelines on laboratory diagnosis have been issued by the World Health Organization (WHO) (5) and there have been several recent publications relating to this area in view of the resurgence of diphtheria in Eastern Europe and the emergence of other infections caused by non-toxin-producing strains (1, 3, 4). Secondly, the fermentation of sucrose is not regarded as an essential test in the laboratory diagnosis of C. diphtheriae infection. Pennie et al. do not state the methodologies in use within the originating laboratory in Malaysia. It is stated by the WHO that screening tests for the presumptive identification of toxigenic C. diphtheriae are essential within the diagnostic laboratory. We refer in particular to tests for the enzymes pyrazinamidase and cystinase. It is likely that use of these simple screening tests would avoid misidentification of C. diphtheriae. C. diphtheriae (all biotypes), C. ulcerans, and C. pseudotuberculosis do not produce the enzyme pyrazinamidase but do, however, produce the enzyme cystinase. “C. xerosis” and other corynebacteria are usually pyrazinamidase positive and cystinase negative (5). The WHO manual makes recommendations for the use of positive and negative controls for these tests. The definitive identification of C. diphtheriae to species and biotype levels relies upon biochemical tests, fermentation of sugars, hydrolysis of urea, and nitrate reduction, in addition to the detection of toxigenicity. Lastly, a recent publication from Funke and colleagues states that the majority of “C. xerosis” strains reported in the literature may have been misidentified as C. amycolatum. They further emphasize that from their data, “C. xerosis” is rarely encountered in clinical specimens (6). Data from Coyle and colleagues (2) clearly show the diversity among “C. xerosis” organisms: they appear to comprise six taxonomic groups, one of which is indistinguishable from C. striatum. In view of the immense public health significance attached to the isolation of toxigenic C. diphtheriae, we fully support all attempts to ensure that accurate methodologies are in use within diagnostic and reference microbiology laboratories.