American Association for Clinical Chemistry
Better health through laboratory medicine
January 2009 Clinical Laboratory News: C. difficile Infections on the Rise

CLN Banner Logo

January 2009: Volume 35, Number 1


C. difficile Infections on the Rise
Is it Time to Return to Culturing?
By Gina Rollins

A prevalence study of Clostridium difficile released in November by the Association for Professionals in Infection Control and Epidemiology (APIC) heralded grim news for the healthcare industry. The increasingly potent pathogen is 6.5 to 20 times more common than previously identified, with 13 per 1,000 inpatients either infected or colonized. Extrapolating that rate to the average number of patients hospitalized in the U.S., the study estimated that C. difficile infection (CDI) affects 7,178 inpatients on any given day and causes the deaths of about 301 patients per day. The care provided to these individuals results in an extra 40,197 days of hospitalization on average, at a cost of about $32.1 million. At the same time, detecting CDI remains a challenge because available diagnostic methods all have drawbacks and there isn’t an industry-wide diagnostic algorithm.

“CDI is an escalating issue in our nation’s healthcare facilities,” said William Jarvis, MD, the study’s principal investigator and president of Jason and Jarvis, a healthcare epidemiology consulting firm. “Clearly, preventing the development and transmission of it should be a top priority for every healthcare institution.”

The Bad News

The APIC analysis amplified prior incidence studies and highlighted the most troubling aspects of C. difficile. This spore-forming, gram-positive, anaerobic bacillus, first identified in the 1930s, has become a leading cause of healthcare-acquired infections and is adapting quite hardily to its environment. Of particular concern, a hypervirulent, antibiotic-resistant strain has developed within the past 8 years. As it has taken hold, laboratorians and infectious disease professionals have discovered to their dismay that diagnostic techniques and protocols are inadequate in terms of providing rapid and accurate diagnosis, typing the strain, and testing its antibiotic susceptibility. Of particular concern, the most sensitive test, anaerobic culture, with reported sensitivity of 95% or higher, is rarely performed. Indeed, the APIC study found that just 4.2% of participating institutions routinely performed cultures for C. difficile, and less than 2% of C.difficile-infected patients were identified by culture. Other available tests, though faster and less labor intensive, have a wide range of reported sensitivities, with most between 70% and 80%.

“It’s pretty clear the most sensitive test was available 30 years ago,” said Dale Gerding, MD, associate chief of staff for research and development at the Hines Veterans Affairs Hospital in Hines, Ill., and professor of medicine at Loyola University in Chicago. “Since then we’ve sacrificed sensitivity for rapid testing and less use of technician time. The development and use of a rapid, accurate diagnostic is a major goal for improved C. difficile surveillance and treatment.” Gerding is on the forefront of C. difficile research; his lab contains a library of between 7,000 and 8,000 isolates and was instrumental in identifying the hyper-virulent outbreak strain.

Aside from the sensitivity of culture in detecting C. difficile, other information to be gained from culture, namely having isolates available for antibiotic susceptibility and strain identification, also has been lost in the shuffle. That circumstance was reflected in the APIC study, according to Jarvis. “With the higher-than-expected prevalence, we’re missing something, not only in terms of epidemiology but also in testing susceptibility. We’re totally dependent on someone [at an organization] suspecting there’s a problem,” he noted.

The Hypervirulent Strain

The C. difficile strain causing the most concern is BI/NAP1/027, which is referred to differently based on the type of analysis used to characterize the bacterium. By pulsed-field gel electrophoresis (PFGE), the strain is called North American Pulse-field type 1 (NAP1), by PCR-ribotyping, as 027, by restriction endonuclease analysis (REA), as BI, and by toxinotyping as toxinotype III. Ribotyping and the 027 moniker are more common in Europe; in the U.S., PFGE and REA predominate, as do the names NAP1 and BI.

Nomenclature aside, the outbreak strain has several properties that have made it a force to be reckoned with. NAP1 strain produces both the toxins TcdA (an enterotoxin) and TcdB (a cytotoxin), which are encoded on a pathogenicity of locus within the chromosome of the organism. Of note, it produces more of both toxins than other strains: 16 times higher concentration of TcdA and 23 times higher concentration of TcdB in comparison to toxinotype 0, historically the most predominate toxinotype of C. difficile and the source of about 75% of infectious strains. The mechanism of this accelerated production is thought to be deletion at position 117 of the tcdC gene that results in a non-functional TcdC protein, which normally downregulates expression of TcdA and TcdB. Without the gene, there is less suppression of toxin production, resulting in higher levels in the stationary phase of growth. There also are C. difficile strains that lack TcdA but are TcdB positive. Originally these were thought to be benign, but recent analyses suggest these strains can cause severe disease too, according to Gerding.

Another unique property of the NAP1 strain that may play a role in its virulence is the presence of a binary toxin, consisting of CdtA and CdtB, which is present outside the pathogenicity locus and has been associated with increased disease severity. This formerly uncommon toxin is similar to other binary toxins like the iota toxin, which is responsible for virulence in some types of C. perfringens. NAP1 also has a propensity to hypersporulate and this may be the reason it establishes itself so quickly and successfully during outbreaks. “My personal opinion is that’s what increases the quantitative contamination,” noted Robert Owens, Jr., PharmD, co-director of the antimicrobial stewardship program at Maine Medical Center in Portland. “It’s not surprising why it’s so hard to get out of an environment. With the hypersporulation it really anchors itself in your hospital.”

More Virulence Factors

Perhaps of most concern, the NAP1 strain is highly resistant to fluoroquinolones such as gatifloxacin and moxifloxacin. This characteristic of the strain developed sometime between the early 1990s and 2000, according to Gerding. His laboratory tracked its existence in a less virulent form to 1984. Canadian researchers noted a marked increase in the incidence of CDI in 2004, and the strain was first reported in the literature in 2005 based on analysis of isolates dating back to 2000 (NEJM 2005; 353: 2433-2441). A complete picture of the strain’s evolving pathogenesis in the 1990s is not available because of lack of isolates from this period, an issue that continues today. “It’s very challenging in the U.S. to know what’s going on because no one is culturing,” noted Jarvis.

Another factor thought to impact the spread of C. difficile is the use of alcohol-based hand hygiene products instead of soap and water handwashing, as these solutions are not sporicidal and don’t remove spores effectively from the hands. Likewise, the most common hospital cleaning agents, like quaternary ammonium-based solutions, also appear ineffective in eradicating spores.

The Infection and Its Spread

CDI causes a spectrum of illness from mild diarrhea to toxic megacolon. The NAP1 strain is associated with increased mortality and morbidity, including higher rates of pseudomembranous colitis and colectomies. Mortality from CDI increases with increasing age, but old age is not the only risk factor for the disease. Antimicrobial usage, especially over prolonged periods, is a definitive cause of susceptibility; proton-pump inhibitors and H2 antagonists are putative causes. Recently there have been reported cases of particularly severe CDI in the peripartum period, among women who took prophylactic antibiotics during delivery (Am J Obster Gynecol 2008; 198: 635). Other analyses have found the disease to be present in community-based patients without recent antibiotic exposure, and in a variety of animals, with toxinotypes common to humans.

Following outbreaks in North America, the NAP1 strain now has a presence in at least 16 European countries, although molecular subtyping is demonstrating divergence of the North American and European strains over time, according to Ian Poxton, PhD, DSc, professor of microbial infection and immunity at the University of Edinburgh College of Medicine and Veterinary Medicine. Poxton also is chair of the European Study Group for C. difficile. Already, aggressive efforts in the Netherlands to rein in the strain, though successful, have been followed by the emergence of another strain, Ribotype 078, which is now the prevalent strain in that country, according to Poxton. In the U.S., NAP1 is known to exist in 40 states and “we have every reason to believe it’s in all states,” said Brandi Limbago, PhD, team lead in bacterial characterization, typing and identification in the division of healthcare quality promotion at CDC.

CDC and Gerding’s lab collaborated to identify the NAP1 strain originally and continue to provide assistance in strain typing and resistance testing to organizations hit with outbreaks. There is a C. difficile module in CDC’s National Healthcare Safety Network, but it is not a reporting requirement of all states. At the present time, there is no standardized nationwide method of C. difficile surveillance.

Detection Methods: Pros and Cons

The arsenal for detecting C. difficile has grown over time, but all methods have pros and cons, so there is no single ideal test and no industry-wide testing protocol (see tables, below). Basic bacterial culture detects the presence of C. difficile, with a reported sensitivity of 95% or higher, but it does not distinguish between toxigenic and nontoxigenic strains, takes 2 or more days for results, is labor intensive, and requires effort to maintain proficiency with specialized growth media.

Table 1
Detection Properties of C. difficile Diagnostic Assays

Assay

Target

Advantages

Limitations

Imaging

Disease pathology

Can assess disease severity and extent

Cannot be used to specifically determine C. difficile infection

Anaerobic culture

C. difficile presence

Directly measures presence of organism with highest sensitivity

Slow turnaround; labor intensive; low specificity due to detection of non-toxigenic isolates

Toxigenic culture

C. difficile presence and toxin B presence

Highest sensitivity

Slow turnaround; labor intensive

CTX assay

C. difficile toxin B presence

Directly measures toxin presence; highly sensitive

Slow turnaround; labor intensive

Toxin A or Toxin A/B (ELISA or EIA)

C. difficile toxins A and B presence

Abillity to batch samples; same day results

Toxin A-only EIA will not detect A-/B+ isolates; lower sensitivity with higher variability reported compared to other diagnostic tools

GDH or Common Antigen

GDH protein presence

GDH constituitively produced at high levels; has high NPV; same-day results

May cross-react with other organisms; GDH is not a virulence factor, but is a constituitively produced protein; positive samples require further testing to identify toxin producing strains

Polymerase Chain Reaction (PCR)

Presence of target gene (usually toxin)

Ability to batch samples, same-day results; comparable sensitivity to CTX assay

Does not detect toxin presence; may be limited by nucleic acid extraction

Reprinted with permission from IHP Healthcare Digest.

Toxigenic culture, which measures both C. difficile and toxin TcdB presence, has acceptable sensitivity and specificity and is considered a reference standard for new diagnostic assays. However, it also takes days to process. “Culture is not fast enough for routine patient care,” said Limbago. “It takes 2 days to culture and up to another 2 to get the toxinigenicity, so the entire process can take up to 4 days.”

Cytotoxicity assay, which detects toxin TcdB, is both sensitive and specific, and has reported sensitivity in the range of 74% to 90%, although some studies have found lower sensitivity compared with toxigenic culture. It also is somewhat slow and dependent on the lab’s proficiency in maintaining cell lines.

EIA, which detects both toxins TcdA and TcdB using antibodies, is the test of choice today, mainly because it is readily available through easy-to-use kits, and can be run in a matter of hours, and batch-processed. The APIC study found that nearly 90% of participating institutions use EIA to detect C. difficile. But the test is not as accurate as culture or cytotoxicity assay and has reported sensitivity ranging from 44% to 99%, and specificity from 75% to 100%.

Table 2
Properties of C. difficile Diagnostic Assays

 

Directly detects C. difficile presence

Measures pathology

Detects toxin A

Detects toxin B

Differentiates toxigenic vs. non-toxigenic infection

Can bundle samples

Same-day results

Directly measures toxin presence

Easy use

Imaging

 

X

 

 

 

 

 

 

 

Anaerobic culture

X

 

 

 

 

 

 

 

 

Toxigenic culture

X

 

 

X

X

 

 

 

 

CTX assay

 

 

 

X

X

 

 

X

 

Toxin EIA

 

 

X

X

X

X

X

X

X

GDH

 

 

 

 

 

X

X

 

X

PCR

 

 

 

X

X

X

X

 

X

Reprinted with permission by IHP Healthcare Digest.

With this relatively low sensitivity, clinicians tend to repeat-order the test during outbreaks when results don’t match their suspicions about a patient. For instance, during Maine Medical Center’s original NAP1 outbreak in 2002, clinicians panicked. “It was chaos. We were doing tests three times per day per patient,” recalled Owens. Aside from over-taxing already busy labs, that type of repeat testing only “increases the probability of false-positives. It becomes a confounder,” noted Gerding.

Another immunoassay, glutamate dehydrogenase (GDH), uses antibodies to detect the enzyme from C. difficile. Like EIA, the GDH assay offers speedy results and ease of use, but it too has a wide range of reported accuracies. Significantly, since it detects both toxigenic and nontoxigenic organisms and the antibody against C. difficile, and also can detect GDH from other organisms, GDH assays generally have been used as a first step to rule-out C. difficile, followed by another test, such as EIA, to determine if there is a toxigenic strain. Two-step, GDH-EIA assays are available and have been reported to have high sensitivity, but not outstanding specificity. Other research indicates that the sensitivity of the GDH portion of at least one assay is only 76%, making it “far too low for use as a screening test for negative specimens,” according to Gerding.

New on the scene are commercial PCR assays, which are available in Europe and expected soon to enter the U.S. market. These assays detect the presence of toxin genes. However, they don’t all detect the same gene or use the same primers, so sensitivities and specificities vary, but reported sensitivities have been in the 86% to 100% range. Like EIA, PCR offers the advantages of fast results and batch processing. Once adopted widely in the U.S., results from the assay will support epidemiologic efforts in identifying strains, but will not provide information about antibiotic susceptibility.

Some PCR assays are being touted to detect characteristics of C. difficile, such as presence of the binary toxin and deletion of the tcdC gene, but the benefits of that type of data are unclear, at least for clinical labs. “A lab’s concern is whether C. difficile is present and whether it is a toxin-producer. I’m not sure how useful it is to know about specific characteristics because there aren’t specific infection control guidelines around a specific strain,” noted Limbago.

Gerding predicts PCR eventually will become the standard C. difficile assay. “Clinical labs have become familiar with the technology because of MRSA, so they can just drop in a new cassette to test for C. difficile,” he explained.

New Guidelines Forthcoming

Both the European Study Group for C. difficile and a joint effort of the Infectious Diseases Society of America (IDSA) and the Society for Healthcare Epidemiology of America (SHEA) are in the process of updating diagnostic and treatment guidelines for C. difficile. Current IDSA-SHEA guidelines, issued in 1995, recommended EIA or stool culture in symptomatic patients only. The new guidelines are expected next summer, according to Steve Baragona, communications and public affairs officer for IDSA. The European Study Group for C. difficile is closer to finalizing its revision, with new guidelines expected early in 2009, according to Poxton. He indicated that “a lot of different algorithms are under consideration,” but that the guidelines committee may recommend the GDH assay as a front-line test, followed by another to establish whether a toxigenic strain is present.

For now, typing of the strains continues for the most part to be the domain of only selected labs. A recent analysis compared the discriminating ability and typeability of seven techniques, along with their agreement in grouping isolates by allele profile A through F, which are defined by toxinotype, presence of binary toxin gene, and deletion in the tcdC gene. The methods included PCR-ribotyping, PFGE, REA, surface layer protein A gene sequence typing, multilocus variable-number tandem-repeat analysis, amplified fragment length polymorphism, and multilocus sequence typing. All were found to be capable of detecting outbreak strains, but only REA and MLVA had enough discrimination to distinguish strains from different outbreaks (J Clin Microbiol 2008; 46:431-437).

What Should Labs Do Now?

Even without updated guidelines, there are a number of measures laboratorians can take to improve C. difficile diagnostic capabilities and help clinicians stay on top of any outbreaks. First and foremost, is to “know the limitations of the assay you’re using,” advised Limbago. With that in mind, if one of the methods known to have less-than-ideal sensitivity produces a negative result, “if there is any concern based on clinical symptoms, use another test, including culture,” said Poxton.

Jarvis suggested that laboratorians, infection control, and infectious diseases staff collaborate closely to “flag and do additional follow-up on patients that are unusual —that is, who do not have known risk factors for C. difficile—or who have more severe disease or have had treatment failures.” Such a group’s focus should be on developing a testing and treatment algorithm that is based foremost on clinical criteria, said Owens. “If you’re using a test with 80% sensitivity, you want to make sure you’re only testing people who are clinically appropriate for testing. Clinicians just want confidence so that if a test comes back yes or no, they can trust the result,” he observed. That type of multidisciplinary approach was essential in Maine Medical Center’s ability to get a grip on its NAP1 outbreak, according to Owens. Even so, CDI rates at the institution seem now to have settled into a new “normal”, at about 15 patients per day, down from 30-to-50 per day at the height of the outbreak.

Regardless of which of the newer assay methods are used, Limbago suggested that laboratorians develop a system for routinely culturing for C. difficile, by, for example, performing a certain number each month as a matter of course. “It will help maintain proficiency with the culture, which is not necessarily easy, and ensure that isolates are available should an outbreak develop,” she explained. Culturing for C. difficile has made a comeback in Western Europe, such that many facilities now are performing it “quite routinely,” according to Poxton.

Time invested now in systems and protocols will position labs to respond well to C. difficile outbreak strains as the organism continues to evolve. “NAP1 is the strain currently of interest, but that won’t be the case in the future. There will be another,” Gerding predicted.

Disclosures: Gerding holds patents for the treatment and prevention of CDI licensed to ViroPharma, and is a consultant for BD GeneOhm, Genzyme, GOJO, Optimer, Salix, Merck, Cepheid, Schering-Plough and ViroPharma. He holds research grants from the US Dept of Veterans Affairs, Cepheid, GOJO, Massachusetts Biological Laboratories, Optimer, and ViroPharma.