Ten Years of Diphtheria Toxin Testing and Toxigenic Cutaneous Diphtheria Investigations in Alberta, Canada: A Highly Vaccinated Population

Abstract

Background

Respiratory diphtheria is a potentially fatal toxin-mediated disease that is rare among highly vaccinated populations. Cutaneous infections with toxigenic Corynebacterium diphtheriae are most commonly linked to travel to an endemic region. Corynebacterium ulcerans has emerged as a predominant, locally acquired cause of respiratory and cutaneous diphtheria in Western Europe. Recently, public health agencies from several highly vaccinated regions expanded their guidelines to investigate toxigenic cutaneous diphtheria regardless of travel history. With relatively unknown epidemiology of C diphtheriae in North America, and increasing diphtheria toxin testing over the last decade, this change could lead to substantial increases in public health investigations with unclear benefits.

Methods

This study examined the diagnostic and public health benefits of toxigenic cutaneous diphtheria investigations in the highly vaccinated population of Alberta, Canada, where travel history is not required for cutaneous diphtheria investigations. All C diphtheriae isolates collected between 2010 and 2019 were reviewed for specimen source, toxigenicity, biovar, and associated clinical and public health data.

Results

Of these, 5% of C diphtheriae isolates were toxigenic and 82% were isolated from cutaneous sites. Three cases of toxigenic cutaneous disease were identified, none from patients with recent travel. Contact tracing identified asymptomatic C diphtheriae colonization among 0%–26% of close contacts, with identical isolate profiles among colonized contacts and primary cases.

Conclusions

Cutaneous diphtheria in nonendemic regions warrants public health investigation regardless of travel history and overall vaccination levels. This study underscores the importance of including C ulcerans in public health guidelines to assess the overall prevalence and epidemiology of toxigenic corynebacteria.

Corynebacteria are part of the healthy human microbiome and are frequently found in clinical specimens. However, some species can cause severe disease, such as toxigenic strains of Corynebacterium diphtheriae, Corynebacterium ulcerans, and Corynebacterium pseudotuberculosis, which can cause classic, respiratory diphtheria. When toxigenic corynebacteria infect the respiratory tract, the diphtheria toxin kills local cells and forms a pseudomembrane that can obstruct the airway and cause fatal suffocation [1]. Respiratory diphtheria is a medical and public health emergency, because it can be transmitted efficiently among unvaccinated individuals; therefore, the World Health Organization (WHO) recommends surveillance of respiratory diphtheria [2]. Vaccination with inactivated diphtheria toxin is correlated with extremely low rates of respiratory diphtheria [3, 4].

Cutaneous diphtheria is not a toxin-dependent disease and therefore can occur in fully vaccinated individuals. Cutaneous disease manifests as a chronic, ulcerating wound, and is endemic to tropical and subtropical regions [5]. In countries with high vaccination rates, cutaneous diphtheria makes up a more substantial proportion of diphtheria cases. Cutaneous diphtheria cases in high-income countries are most commonly associated with recent travel to endemic regions and less commonly with injection drug use, poverty, and homelessness [6–16]. In particular, toxigenic cutaneous cases are almost always associated with travel, with the exception of a 1972 C diphtheriae outbreak among an urban homeless population in Seattle, Washington [11], and C ulcerans, for which the most common risk factor is exposure to domestic animals [6, 7, 17]. The WHO only recommends investigation of suspected cutaneous diphtheria for people in (i) close contact with a confirmed diphtheria case, (ii) endemic tropical/subtropical regions, or (iii) regions with diphtheria outbreaks [2].

In recent studies, several national public health agencies in highly vaccinated regions—including the United States and Australia—expanded diphtheria case definitions to include toxigenic cutaneous disease regardless of travel history [9, 18, 19]. This expanded case definition would increase the number of public health investigations for diphtheria. Moreover, Canada has observed a recent increase in the number of C diphtheriae isolates identified from clinical samples over the last 15 years [20]. These factors, individually or taken together, could greatly impact the demand on both laboratory and public health resources required for toxin testing and investigations into all cases of toxigenic cutaneous diphtheria.

Furthermore, the broader epidemiology of toxigenic corynebacteria in North America is unestablished in the vaccine era, including the prevalence of toxigenic strains and the transmissibility of cutaneous toxigenic corynebacteria among highly vaccinated individuals. Much of the current understanding of diphtheria transmission was established in the prevaccine era, in endemic regions, or from outbreaks among unvaccinated clusters, although this is no longer representative of highly vaccinated temperate regions [5, 21–26]. From these studies, cases of cutaneous C diphtheriae disease have preceded respiratory illness among unvaccinated groups in endemic and nonendemic regions [21, 25]. C diphtheriae transmission dynamics among highly vaccinated populations in nonendemic areas remains unstudied, although the diphtheria toxin provides no known advantage in cutaneous colonization or disease transmission [27]. Overall, the benefits of these resource-intensive investigations are unknown.

To address these gaps, we examined a decade of laboratory and public health surveillance data for toxigenic cutaneous diphtheria across the province of Alberta, Canada: a highly vaccinated population where these protocols have been in place for over a decade.

MATERIALS AND METHODS

Laboratory Testing of C diphtheriae Isolates

Clinical specimens for routine microbiological testing were submitted to acute diagnostic laboratories across Alberta for identification. All C diphtheriae isolates were sent for confirmation to the Provincial Public Health Laboratory (ProvLab; Edmonton, Alberta, Canada) and subsequently referred to Canada’s National Microbiology Laboratory ([NML] Winnipeg, Manitoba, Canada) for diphtheria toxin testing and biotyping (available until early 2019). Public health specimens were collected by swab and shipped to ProvLab in Amies transport medium. Tinsdale, Hoyle’s, and 5% (v/v) sheep blood agar plates were inoculated and incubated for 48 hours at 35°C. After 24 and 48 hours, isolates with suggestive growth and evidence of cystinase and tellurite reductase activity [28] were identified by Gram stain and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (Vitek 2; BioMérieux). The mass spectrometer was installed with Clinical Use Knowledge Base v2.0, which includes C diphtheriae, C ulcerans, and C pseudotuberculosis. Confirmed C diphtheriae isolates were referred to the NML for diphtheria toxin testing and biotyping according to methods described previously: biovars were determined by standard biochemical tests, and diphtheria tox gene carriage and toxin production were determined by polymerase chain reaction and the modified Elek test, respectively [29]. Historic laboratory records were compiled for all diphtheria toxin tests performed on isolates collected in Alberta since January 2010, including the toxin test result and the source of each isolate (eg, cutaneous, respiratory).

Public Health Investigations of Toxigenic Cutaneous Diphtheria

Diphtheria is a notifiable disease in Alberta; all public health investigations were performed according to provincial guidelines [30–32]. History of travel or recent illness was assessed for each case, their close contacts (including household members), and any individuals who were found to be colonized with C diphtheriae during the course of the investigation. The criteria used to identify close contacts are summarized in Supplementary Table 1. All specimens associated with a public health investigation were compiled under a unique investigation identifier provided by ProvLab. Diphtheria public health investigations from January 2010 to December 2019 were identified within provincial laboratory archives and reviewed, including specimen collection date, growth of C diphtheriae (if any), and toxin test results for specimens collected from the primary case and all contacts identified by Public Health.

Survey of Canadian Regional Guidelines for Cutaneous Diphtheria

Diphtheria surveillance guidelines for the 13 Canadian provinces and territories were found on each region’s public health website [30–41]. British Columbia and Alberta conducted diphtheria testing for Yukon territory and the Northwest Territories, respectively. No public information was found for Prince Edward Island.

Statistics

Changes over time were assessed with a two-tailed, non-parametric Spearman correlation. Diversity among C diphtheriae isolate sources and biovars was assessed by two-way analysis of variance with Tukey’s post hoc analysis. Toxigenicity was evaluated with a two-tailed Wilcoxon matched-pairs signed-rank test.

RESULTS

Diphtheria Toxin Testing Significantly Increased From 2010 to 2019

When C diphtheriae was identified by routine laboratory testing of a respiratory, systemic, or cutaneous infection, diphtheria toxin testing was automatically performed as per provincial public health guidelines (summarized in Supplementary Table 1). Toxigenic isolates triggered a public health investigation to identify contacts of the primary case and perform follow-up tests for anyone from whom toxigenic C diphtheriae was isolated, requiring further diphtheria toxin testing. The process involved is illustrated in Figure 1.

Figure 1.

Summary of Alberta’s public health guidelines for cutaneous diphtheria, 2010–2019. Any C diphtheriae identified from a clinical specimen was sent for diphtheria toxin testing (tox PCR then a modified Elek [mElek] test). For any positive (toxigenic) isolates, a public health investigation was opened, beginning with contact tracing. Screening was routinely performed by nasal and throat swabs cultured on Tinsdale and Hoyle’s agar; for any patient with a cutaneous lesion, a skin swab of the lesion was also collected. Any resulting C diphtheriae growth was tested for the diphtheria toxin. All cases and contacts were rescreened twice after antibiotic treatment unless they were shown to be noncolonized in the initial screens. Further details are described in Supplementary Table 1.

From 2010 to 2019, C diphtheriae was identified from 145 specimens and each was sent for toxin testing. The total number of C diphtheriae identifications increased significantly over this time, although the proportion of toxigenic isolates did not (P = .0001 and P = .8611, respectively) (Figure 2A). Routine diagnostics accounted for 84.1% of all C diphtheriae isolates referred in the decade (n = 122) and the number of isolates ordered by Public Health increased over the decade (Figure 2B), correlating with the increase in reported cutaneous diphtheria cases (P = .0167) (Figure 2C). Each investigation for these 3 cases of laboratory and clinically defined toxigenic cutaneous disease prompted contact tracing and screening, which accounted for the remaining 15.9% of all isolates for diphtheria toxin testing (n = 23). Therefore, unsurprisingly, toxin test increases were significantly correlated with the increased number of cases (P = .0167). Zero isolates of C ulcerans or C pseudotuberculosis were referred for toxin testing, although this was not mandatory throughout the study period. No cases of respiratory diphtheria were reported in this time.

Figure 2.

Trends in diphtheria testing and cases in Alberta from January 2010 to December 2019. ***, P = .0001; *, P < .05 (Spearman correlation). (A) The number of C diphtheriae isolates sent for diphtheria toxin testing by year of original specimen collection. (B) C diphtheriae isolates sent to toxin testing according to the reason for collection (orange, routine clinical testing; red, follow-up for a diphtheria public health investigation). (C) Cases of diphtheria disease reported to public health.

To assess the diversity of clinical C diphtheriae isolates across Alberta, we compared the specimen source, biovar, and toxigenicity of all isolates from routine clinical testing (Figure 3). C diphtheriae was most commonly identified from cutaneous clinical specimens (82.4% of isolates), significantly more often than respiratory or invasive sources (12.6% [P = .0001] and 2.5% [P < .0001], respectively) (Figure 2A). When public health isolates were also considered, the reflexive toxin testing from cutaneous isolates accounted for 85.5% of the province’s total diphtheria toxin testing over the past decade. Isolates belonged to the mitis biovar significantly more often than gravis or belfanti (P = .0011), and the intermedius biovar was never detected (Figure 3B). Finally, 4.9% of clinical isolates were toxigenic—significantly fewer than predominant non-toxigenic C diphtheriae strains (P = .0078)—none of which encoded a nonfunctional tox gene (Figure 3C). Overall, routine isolates from across Alberta demonstrated a significant predominance of non-toxigenic C diphtheriae biovar mitis.

Figure 3.

Properties of C diphtheriae isolates collected from routine clinical diagnostics versus public health investigations in Alberta. (A) Specimen source from which C diphtheriae was isolated. (B) Biovar and (C) toxigenicity of each isolate. ****, P < .0001; ***, P = .0001; **, P < .01; *, P < .05.

Three Public Health Investigations Occurred in Alberta for Toxigenic Cutaneous Diphtheria With No Recent Travel

Between 2010 and 2019, public health investigations occurred for 3 separate cases of toxigenic cutaneous diphtheria (2017, 2018, 2019) (Figure 4A). It is notable that the 2017 and 2019 investigations occurred in the same primary case, and the diagnostic specimen from the 2019 diagnosis had concurrent growth of Staphylococcus aureus and Streptococcus pyogenes. Both patients had underlying skin conditions or wounds at the site from which C diphtheriae was isolated. Both patients’ immunizations were up-to-date and neither had recent travel—or close contact with anyone who had recently traveled—to a classical C diphtheriae-endemic area in the last 6 months, as identified by extensive public health investigations and contact tracing. Provincial public health guidelines were modified in July 2010, January 2011, April 2018 (before the 2018 case), and September 2019 (after the 2019 case), although no changes were made in how toxigenic cutaneous diphtheria was reported (Supplementary Table 1).

Figure 4.

Properties of C diphtheriae isolates collected from public health investigations after a case of toxigenic cutaneous diphtheria. (A) The 3 cases of toxigenic cutaneous diphtheria in Alberta from January 2010 to December 2019. (B) The number of close contacts identified for each case and whether each was found to be colonized (yellow) or not colonized (gray), according to Supplementary Table 1. (C) The number of infected cases and colonized contacts in each public health investigation and the specimen source, toxigenicity, and biovar. Noncolonized contacts are not shown.

To assess C diphtheriae colonization among case contacts, we compiled each investigation’s laboratory results. No additional cases of cutaneous or respiratory diphtheria were identified. Asymptomatic colonization was detected in 0.0% (n = 0/13), 25.6% (n = 9/35), and 14.6% (n = 6/41) of all contacts in the 3 investigations (Figure 4B). The most common site of C diphtheriae colonization among contacts was the respiratory tract, particularly the throat (Figure 4C). All case and contact isolates were toxigenic and biovar mitis, until this testing was discontinued. It is notable that non-toxigenic C diphtheriae was not detected in any of the 92 contacts screened, suggesting a low prevalence of non-toxigenic C diphtheriae strains among case contacts, in contrast with predominant non-toxigenic C diphtheriae among routine clinical isolates. Taken together, this suggests low C diphtheriae heterogeneity among contacts. Overall, 3 public health investigations identified a minority of contacts asymptomatically colonized with toxigenic C diphtheriae.

To examine the laboratory burden of provincial public health guidelines for diphtheria, we assessed test volumes from each investigation (Figure 5). In total, the 3 investigations identified and sampled 92 contacts, collected 315 specimens for testing, and required 23 diphtheria toxin tests. On average, each cutaneous diphtheria investigation required identifying 29.7 contacts, collecting 105.0 laboratory specimens for C diphtheriae screening, and testing 7.7 isolates for diphtheria toxin testing (Figure 5B).

Figure 5.

Summary of the laboratory and public health requirements for cutaneous diphtheria investigations in Alberta. (A) The number of swabs and screening tests required for each stage of the investigations in 2017, 2018, and 2019. (C) An overview of the mean and range of selected resources and outcomes of all 3 investigations combined. Diamonds indicate the exact numbers from individual investigations.

Public Health and Testing Guidelines for Cutaneous Diphtheria Were Similar Across Canada

To determine how representative Alberta’s public health and testing guidelines were of other Canadian regions, we compared the diphtheria guidelines of all 13 Canadian provinces and territories (Supplementary Figure 1). Similar to Alberta, many Canadian regions recommended toxin testing (100.0%), biotyping (100.0%), contact tracing (77.8%), chemoprophylaxis of contacts (88.9%), vaccination (100.0%), and test of cure (100.0%) for toxigenic cutaneous diphtheria. Overall, Alberta’s guidelines for cutaneous diphtheria were largely representative across Canada.

DISCUSSION

The diphtheria toxin plays a significant role in respiratory diphtheria and is therefore targeted by strategies for diphtheria prevention (vaccination), treatment (antitoxin administration), and public health management. In recent studies, several organizations from highly vaccinated countries have explicitly added toxigenic cutaneous diphtheria to diphtheria case definitions [9, 19]. This 10-year review of toxigenic cutaneous diphtheria cases in a highly vaccinated population demonstrates the yield of these public health investigations and the scale at which public health and laboratory resources were utilized with this response. Furthermore, we show that toxigenic cutaneous diphtheria investigations contributed to an overall increase in diphtheria toxin testing. Finally, we highlight a gap in C ulcerans detection and report 3 cases of cutaneous disease caused by toxigenic C diphtheriae but with no established link to travel to an endemic area, supporting the removal of travel history as a condition of public health investigation in highly vaccinated areas.

In this time period, 3 cases of toxigenic cutaneous diphtheria were reported in Alberta, Canada, all caused by C diphtheriae and occurring in fully vaccinated individuals with pre-existing skin conditions and no recent travel. Before these 3 cases of toxigenic cutaneous C diphtheriae infection, North America’s most recent published case with no travel history was in 1972 in Seattle [11]. This differs from other reports from high-income countries: extensive retrospective surveillance studies in the United Kingdom and Belgium (1986–2017) identified the greatest risk factor for toxigenic cutaneous C diphtheriae infection as recent travel to a classical endemic region [6, 7, 17], as highlighted by several international case reports demonstrating toxigenic C diphtheriae acquisition after travel to an endemic area [8–10, 42]. Pragmatically, this finding supports reflexive testing of potentially toxigenic Corynebacterium species, as well as contact tracing for toxigenic cutaneous cases regardless of an identifiable travel history, as performed in this study.

A second notable difference between the extensive surveillance data from Western Europe versus this study is the lack of reported cases caused by C ulcerans, which has become the predominant cause of cutaneous diphtheria in Western Europe and can cause respiratory or cutaneous diphtheria without a travel history (the most common risk factor is exposure to domestic animals) [6, 7, 17]. In the United Kingdom and Belgium, C ulcerans accounts for 45%–71% of toxigenic corynebacteria infections, increasing in recent decades and often outnumbering C diphtheriae infections [6, 7, 17]. Although this has not been studied in North America, C ulcerans has been identified before in both Canada and the United States from human clinical samples, and it has even been detected in Alberta in the years leading up to this study [29, 43]. The lack of C ulcerans in this study therefore suggests either a markedly different epidemiology of toxigenic corynebacteria in North America compared with Europe or a limitation in detecting or reporting these non-diphtheriae pathogens.

We suggest two reasons why neither C ulcerans nor C pseudotuberculosis were identified in this study. Most importantly, laboratory surveillance of these species was in place only briefly in Alberta, and neither were included in Alberta’s diphtheria case definition. Therefore, toxin testing of other corynebacteria was not mandatory and may represent a lack of referrals rather than a true absence of infections. Although neither reporting C ulcerans infections nor referring isolates for toxin testing was not required, C ulcerans has been detected in Alberta in the years just before this study [29]. Therefore, this study cannot definitively exclude toxigenic C ulcerans infections occurring in Alberta over this period. Futhermore, these “other” corynebacteria are less well recognized, and toxin testing may not have been considered. C pseudotuberculosis human infections are exceedingly rare (only 1 case was reported across the United Kingdom between 1986 and 2017 [6, 7]), and the importance of C ulcerans in human disease has only been established relatively recently. Nonetheless, omitting C ulcerans and C pseudotuberculosis from the diphtheria case definition is an important oversight by which laboratories may fail to recognize, interpret, and adequately report clinically relevant and potentially fatal infections. Finally, the methods to identify corynebacteria changed over the time of this study, with the introduction of mass spectrometry in 2012 replacing biochemical strips and nucleic acid sequencing as the primary technique to identify C diphtheriae. More importantly, the first version of the mass spectrometer database did not include C ulcerans or C pseudotuberculosis (only C diphtheriae), although all 3 are listed in subsequent versions. This did not affect this study, because the instruments were introduced to Alberta laboratories with the second database version installed; however, it highlights a crucial area in which clinical microbiologists must be attentive.

It is notable that the introduction of mass spectrometers in Alberta’s 2 largest clinical microbiology laboratories coincided with the year when the number of C diphtheriae identifications began to significantly increase in Alberta, with zero respiratory cases. This trend was observed more broadly across all of Western Canada [20]. More importantly, the proportion of toxigenic isolates from Alberta did not significantly change over the decade, remaining under 5% of all C diphtheriae isolates identified from routine clinical isolates across the province. Although this was not a systematic study assessing colonization, this suggests low diphtheria toxin prevalence among the population.

Although current guidance on laboratory diagnostics for diphtheria focus on respiratory disease [28], more than 85% of toxin tests over the last decade in this study originated from cutaneous isolates. This highlights an opportunity to steward laboratory resources by judiciously reviewing test utilization, ensuring that subsequent testing is appropriately focused on specimens from sterile sites, or with coryneform bacteria that are pure, predominant, or in the presence of S pyogenes or S aureus. This practice could substantially reduce testing, curtail costs, and steward downstream public health resources while maintaining the important roles of laboratories in diphtheria diagnosis [28] and a robust, ongoing response to respiratory and cutaneous diphtheria, if vaccination levels remain high and travel history can be used reliably as an indicator of toxigenic infection. As discussed above, this practice may not be the case.

Downstream public health investigations are also resource-intensive. In this study, contact tracing and specimen collection were required for dozens of contacts, with over 100 specimens collected over multiple days of follow-up, with few case contacts identified as colonized. No further cases of disease were identified, although prophylaxis was administered to close contacts. Other published reports have included contact tracing of up to 250 contacts with no further cases identified [6]. However, in the nature of public health work, effective interventions promote the absence of disease.

In contrast with the diversity and rare toxigenicity of isolates submitted for routine clinical testing, isolates from colonized contacts had the same toxigenicity and biovar (when available) as the primary case, suggesting but not proving transmission among contacts. Unfortunately, because the cases were not tested for C diphtheriae colonization at respiratory sites when they presented with cutaneous disease, there is no data on where the primary cases were colonized at the beginning of each investigation. Nonetheless, both respiratory and cutaneous swabs for contact were required to identify all colonized contacts.

CONCLUSIONS

There is substantial pressure on laboratories and public health. Resource allocation must be continually reassessed. Should travel history be considered in investigating cases of toxigenic cutaneous diphtheria? This study analyzes the benefits of doing so among highly-vaccinated populations. While nontoxigenic C diphtheriae vastly predominated in clinical specimens, and subsequent investigations for toxigenic cases were resource-intensive, this approach identified three cases of toxigenic cutaneous diphtheria that may not have been identified under current WHO guidelines due to the absence of recent travel. Furthermore, complementary strategies should be emphasized. This includes ongoing vaccination programs, the elimination of barriers to care, education of physicians who see patients before or after travel, and careful review of microbiology testing algorithms to detect all potentially toxigenic Corynebacterium species capable of causing human disease.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.