Background.We present 2 case reports in the United States and investigations of diphtheria-like illness caused by toxigenic Corynebacterium ulcerans. A fatal case occurred in a 75-year-old male Washington resident who was treated with clindamycin but did not receive equine diphtheria antitoxin. A second, nonfatal case occurred in a 66-year-old female Tennessee resident who received erythromycin and diphtheria antitoxin. Methods.Both case patients and close human and animal contacts were investigated by their respective state health departments. Results.C. ulcerans isolated from the patient who died was resistant to erythromycin and clindamycin. For both isolates, conventional polymerase chain reaction results were positive for A and B subunits of diphtheria toxin gene tox, and modified Elek tests confirmed toxin production. The source of infection remained undetermined for both cases. Neither patient was up-to-date with diphtheria toxoid vaccination. Conclusion.These case reports highlight the importance of early treatment with diphtheria antitoxin, the selection of effective antimicrobial agents, and prevention through up-to-date vaccination.

To the Editor—In a recent report about diphtheria-like illness due to toxigenic Corynebacterium ulcerans, Tiwari et al. [1] highlighted the importance of this emerging but often-neglected pathogen. Unfortunately, because of the pathogen’s novelty (the first valid molecular description of the species dates only to 1995 [2]), uncertainties related to diagnosis, therapy, and prevention of illness due to toxigenic C. ulcerans still exist. Two issues that were discussed by Tiwari et al. [1] are related to the sequence of C. ulcerans diphtheria toxin that we recently described [3, 4]. First, in our opinion, the efficacy of the commonly used diphtheria toxoid vaccine against diphtheria due to C. ulcerans is not known. For instance, among the 15 C. ulcerans–infected patients who had an obtainable vaccination history listed by Tiwari et al. [1] (including their own 2 patients), 4 and 7 patients were fully or partly vaccinated, respectively, whereas 4 patients had not received any diphtheria vaccination. More promisingly, although older, inconclusive studies were probably hampered by the presence of phospholipase D in C. ulcerans (reviewed by Wong and Groman [5]), more-recent laboratory studies that used toxigenic C. ulcerans and diphtheria antitoxin in cytotoxicity assays [3] or in an animal model [6] demonstrated a protective effect of diphtheria antitoxin, although the protective effect was determined on the basis of only 2 analyzed human-pathogenic strains. In conclusion, even if a protective effect of diphtheria toxoid vaccine against C. ulcerans is assumed (i.e., attenuated clinical symptoms are observed), one should remember from studies of Corynebacterium diphtheriae that vaccination probably does not prevent colonization by toxigenic corynebacteria but only prevents overt disease. Although we agree that, because of the lack of a proven vaccine against C. ulcerans, diphtheria toxoid vaccine might be a reasonable alternative, especially during reconvalescence, we feel that the evidence that “up-to-date immunization with a diphtheria toxoid vaccine will prevent diphtheria and diphtheria-like illness caused by C. ulcerans” [1, p. 399] is scanty. Second, Tiwari et al. [1] reported that the real-time C. diphtheriae diphtheria toxin–encoding tox gene PCR, described by Mothershed et al. [7], revealed atypical amplification of subunit A and no amplification of subunit B. Similarly, another study reported that real-time PCR was insufficient for the detection of the C. ulcerans tox gene [8]. This is not unexpected, because a previous study revealed major nucleotide sequence differences between the C. diphtheriae tox and the C. ulcerans tox genes, as well as between tox genes from different C. ulcerans strains. Three more sequences of C. ulcerans diphtheria toxin have been published in GenBank, and an additional sequence was reported that was identical to 1 of our published strains [9]. Careful probe design is needed to detect the C. ulcerans tox gene by PCR, because considerable variations in the sequence of the gene have already been described. Taken together, these 2 uncertainties in the diagnosis and the treatment of diphtheria-like illness due to toxigenic C. ulcerans highlight the need for a better understanding of the molecular epidemiology and characteristics of C. ulcerans and its diphtheria toxin–like toxin.

To the Editor—We appreciate the opportunity to respond to Schuhegger et al. [1], who highlight 2 issues in their letter. First, Schuhegger et al. [1] raise questions about the effectiveness of diphtheria toxoid vaccine in protecting against toxinproducing Corynebacterium ulcerans. It is known that there are genetic variations in the diphtheria toxin gene (tox) of Corynebacterium diphtheriae and C. ulcerans, as well as differences in the amino acid sequences of the diphtheria toxins they produce. However, given that the functional consequences of these differences have not been fully established for clinical diphtheria or diphtheria-like illness due to C. ulcerans, we believe that the use of diphtheria toxoid vaccine as a preventative measure, and especially the use of diphtheria antitoxin as part of the treatment for diphtheria-like illness due to C. ulcerans, is justified. In the case we reported, the patient who had extensive membrane in the pharynx and received equine diphtheria antitoxin recovered fully; the patient who did not receive the antitoxin had a fatal outcome. Second, we are in agreement that a careful probe design may be useful for specifically detecting unique tox sequences, but we believe that our study has clearly documented the presence of toxin-producing C. ulcerans. We reported that the real-time PCR for detection of the C. diphtheriae tox gene [2] revealed atypical amplification of subunit A and no amplification of subunit B for both patients. Consequently, test results were reported as negative for the tox gene. However, C. ulcerans was isolated from both patients, and conventional PCR and a modified Elek test results were positive for the detection of the tox gene and the diphtheria toxin itself, respectively [3]. In summary, we support the need for a better understanding of the molecular epidemiology and characteristics of C. ulcerans and the diphtheria toxin that it produces, because the spectrum of severity of this uncommon disease is unknown, and the diagnosis of mild infections may be missed completely, especially in a non– diphtheria outbreak setting. C. ulcerans produces an exotoxin that is very similar, if not identical, to the diphtheria toxin produced by C. diphtheriae, and it can also produce typical membranous lesions in the throat and complications that are clinically indistinguishable from those of respiratory diphtheria. Most commonly, C. ulcerans is isolated as an etiologic agent while investigating for diphtheria. Because respiratory diphtheria may be severe or fatal, treatment should be immediate and should not be delayed until laboratory results are available. Specifically, diphtheria antitoxin should be administered promptly to patients with clinical respiratory diphtheria to prevent serious complications and death, even in cases in which C. ulcerans, rather than C. diphtheriae, is isolated [3].