Clostridium difficile infection: an Australian clinical perspective

The scale of the problem now posed by Clostridium difficile infection (CDI) is becoming frighteningly clear. Since 2001, a dramatic increase in the incidence and severity of CDI has occurred, particularly, in North America, the United Kingdom and Europe, associated with the emergence of a fluoroquinolone-resistant clone known as restriction endonuclease type BI, pulsed field type NAP1 or PCR ribotype 027 (RT027) Clostridium difficile (CD)^1–3. CD is now the most commonly identified nosocomial pathogen in the USA^4–6 and in 2011 there were approximately 450 000 incident cases of CDI in the USA and 29 300 deaths at day 30 post diagnosis⁶. Using an estimated attributable mortality rate of 50%, approximately 15 000 deaths due to CDI occurred in the USA in 2011.

Australian data have been slow to emerge, but a coordinated survey of lab diagnosis of CDI in 450 Australian public hospitals in the 2-year period 2011–2012 has provided the first comprehensive nationwide profile of CDI. The study identified 12 683 cases of CDI and the incidence of disease increased by 26% over the period of the study⁷, confirming that CDI is responsible for substantial morbidity and presumably mortality in Australia.

There is clear evidence that CD RT027 causes more severe disease⁸ and also has a reduced cure rate and an increased risk of relapse⁹. Other emergent strains of CD have contributed to the increasing burden of disease and also the changing the pattern of CDI^10,11. CD RT078 for example, shares genetic virulence characteristics with CD RT027, has been shown to cause severe disease and is associated with community-acquired infection¹⁰.

Australia was forewarned of the possibility for arrival of CD RT027¹², although the fact that routine diagnostics are based on toxin detection, and isolates are not routinely cultured, hampered surveillance efforts. The first CD RT027 strain identified in Australia occurred in a patient returned from hospitalisation in North America¹³ but did not apparently result in secondary cases. Locally acquired infection was then reported in Melbourne in early 2010¹⁴, although subsequent surveys of Australian CD isolates indicate that RT027 has not become endemic within Australia^15,16.

In this setting of heightened awareness and understanding of the potential for outbreaks of severe CDI, two patients at Monash Health developed clinically severe CDI in mid-2011. While not unusual, we were actively surveying for the appearance of RT027 using the Xpert C. difficile PCR (Cepheid, Sunnyvale, CA, USA) as the confirmatory diagnostic test for CDI. This assay includes a detection probe for the tcdC nucleotide 117 deletion (tcdCΔ117) characteristic of CD RT027 strains, although this deletion is present in some other CD nonRT-027 strains¹⁷. Faecal samples from both patients signaled as presumptive CD RT027 but subsequent analysis found the strain to be PCR ribotype 244 (RT244) and fluoroquinolone susceptible. We eventually identified 12 patients at our institution over a period of 9-months with RT244 CDI and a retrospective case-control study found that RT244 CDI was more severe, and death was 13-times more likely in patients infected with this strain compared to non-RT244 CDI controls¹⁸. One-third of infections were community acquired. In October 2011 CDI due to RT244 was reported also in New Zealand and was also found to cause more severe disease and was also frequently community acquired¹⁹. Whole genome sequencing of Australian outbreak strains found that RT244 CD was from the same clade as RT027 but significantly divergent from it. A whole-genome sequence comparison of 15 Australian CD RT244 strains from Victoria, New South Wales and Western Australia confirmed the CD RT244 strains formed a highly related clone with 14 isolates having 8 or fewer single nucleotide polymorphisms²⁰. The reasons for the sudden appearance of this virulent strain, and its subsequent disappearance are unknown.

Genomic analysis of CD RT027 isolates from around the world has identified two distinct epidemic lineages with evidence of intercontinental transmission, including repeated introductions from North America to the United Kingdom²¹. Australian CD RT027 isolates were shown to have been imported from North America and the transmission of the Australasian RT244 clone found in a patient who developed CDI on return to the UK²⁰ emphasises the potential for rapid international dissemination of virulent clones.

To further emphasise the potential for CDI to pose new challenges, we found that RT244 produces a variant Toxin B that has an altered N-catalytic domain more similar to C. sordellii lethal toxin and produces atypical cytotoxicity on Vero cells in vitro¹⁸. This was the first report of this variant toxin being associated with a severe CDI cluster, but the role this variant toxin may play in the virulence of RT244 CDI has not yet been established.

A particular challenge for clinicians is the identification and treatment of relapses that occur in 20–25% of patients following the initial episode, but increase following each relapse and exceeds 50% following multiple recurrences. In the USA, 83 000 (21%) of CDI patients experienced first recurrences in 2011 and the high associated morbidity and costs associated with relapses have lent strong impetus to the development of new therapeutics that may prevent their occurrence⁶. Fidaxomicin is a non-absorbable macrolide antibacterial recently licensed for CDI. Although substantially more expensive and no more effective than oral vancomycin, fidaxomicin reduces relapse rates in non-RT027 CDI²² and has therefore found a role in some CDI treatment guidelines^23,24. A number of new antibacterials are currently in Phase III trial including the quinolonyloxazolidinone cedazolid, and the lipopeptide analogue of Daptomycin, CB-183,315 as well as repurposed agents such as the non-absorbable rifamycin, rifaximin and nitazoxanide²⁵.

CDI is associated with antibacterial induced alteration in the gut microbiome and recurrent disease is associated with persistent gut dysbiosis. Restoration of the gut microbiome by administration of donor-derived faecal suspension is highly effective in preventing recurrence and restoration of gut microbial diversity²⁶. Theoretical concerns as to the lack of knowledge of long-term effects, transmission of pathogens and lack of standardisation of donor bacterial flora as well as the logistic difficulties have hampered access to this treatment. Efforts to define a minimal set of bacterial strains that could provide standardised on-demand treatment seem likely to be the way forward²⁷. A recent Phase II study offers the prospect of highly targeted bacteriotherapy using non-toxigenic CD for prevention of recurrent infection²⁸ and another current study in a mouse model found that Clostridium scindens offered protection from CDI by altering bile acid metabolism in the gut so as to inhibit CD spore germination²⁹.

Further hope is offered by immune mediated protection to CDI. Absence of anti-toxin antibody is associated with disease and recurrence and passive immunisation by administration of bivalent monoclonal anti-Toxin A and anti-Toxin B antibodies reduced CDI recurrence from 25% to 7%³⁰. A Phase III study examining the relative protective effect of anti-Toxin A, anti-Toxin B or the bivalent combination has been completed and results are anticipated shortly³¹.

Intensively applied control measures including anti-microbial stewardship, enhanced identification and reporting of cases and implementation of mandated infection control measures can reduce CDI¹¹; however, the huge burden of CDI, and the emergence of community acquired disease, challenges how further CDI control is to be achieved. An international Phase III placebo-controlled vaccine trial with a bivalent Toxin-A and Toxin-B toxoid vaccine is now underway in immunocompetent at-risk patients over 50 years of age, including at a number of Australian sites and if successful, offers an alternative and complementary strategy to reduce the global problem presented by CDI³².