Diagnosis of C. difficile – Why So Difficult?

With no single best approach, laboratories must tailor testing based upon preanalytical factors, their patient populations, and other factors unique to their institutions
Clostridioides (previously Clostridium) difficile has received much attention in the past couple of decades due to its rapid spread and rising virulence. C. difficile increasingly has been reported outside of acute care facilities in nursing homes and community home settings.
In 2013, the Centers for Disease Control and Prevention (CDC) listed C. difficile as one of three antimicrobial-resistant urgent threats—of the highest level of concern to human health (2). Also in 2013, the Centers for Medicare and Medicaid Services (CMS) mandated that healthcare institutions track and report C. difficile infection (CDI) rates to the National Healthcare Safety Network, which enables the public to compare rates between similar hospitals.
Debates about the most appropriate diagnostic method for CDI, as well as the investment needed to accurately capture CDI rates, have been heated within the laboratory and clinical community. This article will review factors affecting the choice of laboratory diagnostic methods for diagnosing C. difficile in light of the recent Infectious Diseases Society of America (IDSA)/Society for Healthcare Epidemiology of America (SHEA) C. difficile guidelines (1).
CLINICAL DISEASE STATES
C. difficile can cause life-threatening diarrhea and colitis (i.e., inflammation of the colon) primarily in individuals who recently have taken antibiotics and recently had contact with the medical system. An anaerobic Gram-positive bacillus, C. difficile possesses spores highly resistant to alcohol gels and many disinfectants that persist on inanimate surfaces for months if those surfaces are inadequately cleaned.
Once ingested, the spores can germinate in the intestine and produce toxins. The toxins can cause fluid accumulation in the bowel, leading to diarrhea or gastrointestinal complications. Infection control precautions must be instituted to curb the spread of the spores; thus, accurate and rapid diagnosis of C. difficile is critical.
C. difficile can exist within our bodies in various states. Consider the following scenario: A 92-year old nursing home resident was admitted with altered mental status, after having recently completed a 3-week course of levofloxacin for urinary tract infection. She had not had bowel movements for the past 2 days.
On admission, her peripheral white blood cell (WBC) count was 90,100 cells/mm3 (normal range 3,600–11,100 cells/mm3) with 24% bands (normal range 0–6%), or immature neutrophils. An abdominal computed tomography scan revealed that she had diffuse mucosal and bowel wall thickening, consistent with megacolon. A C. difficile test performed on rectal swab by toxin A/B enzyme immunoassay was positive. Subsequently this patient had a colectomy and started treatment with metronidazole.
Representative photos of a colectomy specimen obtained from a different patient with CDI demonstrate pseudomembranes visible on gross examination (Figure 2A) as well as microscopically (Figure 2B), which resembled this patient’s specimen.
This scenario illustrates severe disease due to C. difficile. CDI may present with a range of clinical findings from diarrhea to pseudomembranous colitis to toxic megacolon. Sigmoidoscopy or histopathology may show thick, adherent layers of inflammatory cells and mucus, also known as pseudomembranes.
The most significant risk factor for CDI is antibiotic exposure. Although clindamycin, broad-spectrum cephalosporins, ampicillin, and fluoroquinolones typically are implicated in CDI, any antibiotic can cause this condition. When a person takes an antibiotic, beneficial bacteria in the intestine are destroyed or impaired for a period of time, thus increasing the likelihood that C. difficile may lead to infection.
Patients also can be colonized with or carry C. difficile without symptoms. This colonization state complicates clinical diagnosis, since the organism can be detected but isn’t necessarily causing disease. Colonizations with toxigenic (ie. toxin-producing) strains occurs in up to 15% of patients, and approximately 6% of patients are colonized with nontoxigenic strains. (3).
Consequently, merely detecting the C. difficile organism or its toxins within intestinal contents—stool or otherwise—does not mean they are causing disease. A patient’s clinical state at the time of sample collection determines whether CDI should be considered. 
Pediatric CDI is more difficult to diagnose. Infant intestinal cells do not appear to have receptors for C. difficile toxins, so neonates may have detectable C. difficile in their stools without necessarily manifesting with disease. This makes the results of C. difficile testing difficult to interpret in children with diarrhea, and some laboratories do not perform testing for C. difficile for children younger than 1 or 2 years of age.
LABORATORY DIAGNOSTIC METHODS
Several C. difficile diagnostic methods are available (Table 1). Test targets include the C. difficile organism itself or its toxins.
C. difficile produces a variety of toxins. Toxin A, encoded by the tcdA gene, is an enterotoxin that causes diarrhea. Toxin B, encoded by the tcdB gene, causes cellular distortion in vitro and presumably affects cells in vivo. The tcdC gene regulates toxin A and B production. Genes encoding for toxins A and B are present on the pathogenicity locus (PaLoc).
The NAP1/BI/027 strain was discovered in the early 2000s to produce a newly recognized toxin—the binary toxin (4). Although the function of the binary toxin has not been well elucidated, the cdtA and cdtB genes that encode it are located near the PaLoc. Assays target a variety of these genes and/or toxins.
Enzyme immunoassays (EIAs) detect toxins A and/or B, and some assays detect glutamate dehydrogenase (GDH), a conserved metabolic enzyme—a common antigen—secreted by C. difficile. Presence of GDH confirms presence of C. difficile but not necessarily its toxins. GDH EIAs detect the enzyme produced by both toxigenic and/or nontoxigenic strains of C. difficile. GDH is produced at much higher levels than toxins A and B.
A key advantage of GDH testing is its rapid turnaround time. Recent studies of this method also have demonstrated fairly high sensitivities (5). The disadvantage of GDH testing is that it can’t be used as a stand-alone test for CDI, and a confirmatory test for the presence of toxin is needed. These assays have been used in multistep and algorithmic approaches to diagnosis in combinations of testing involving toxin EIA, molecular assays, or other assays.
Toxin A/B EIAs target the toxins rather than C. difficile itself but are less sensitive than other methods in diagnosing CDI. However, these tests are rapid and may be combined with GDH targets to increase sensitivity.
Toxigenic culture is a highly sensitive method of recovering the organism when selective culture media are used. Recovering C. difficilestrains allows for further molecular typing studies, such as to compare strains’ relatedness, or for antimicrobial susceptibility testing when indicated.
However the organism is recovered, presence of the toxin must be confirmed. Newer commercially available chromogenic media are also available to aid in organism recovery. Cell culture cytotoxicity assays may be performed to detect the toxin directly from stool. Time to results by this method is typically 24 to 48 hours.
Finally, molecular-based assays such as nucleic acid amplification tests (NAATs) detect the genes encoding for the toxins rather than the toxins themselves. Various NAATs are commercially available to laboratories, and these assays are highly sensitive.
LABORATORY DIAGNOSTIC TESTING CONSIDERATIONS
NAATs rose in popularity over the past 10 years based on their high sensitivity in detecting CDI. However, given concerns that NAATs may be detecting colonization states, some groups have advocated multistep or algorithmic approaches to diagnosis that make use of EIAs rather than relying solely upon molecular assays. No one answer provides the single, best approach for all institutions and hospitals; rather, laboratory approaches will vary based upon preanalytical issues of specimen type, the patient population under assessment, and the intended use of the test.
The type of specimen a laboratory receives is important when interpreting results of an assay. In most situations, only unformed stools should be tested when assessing for CDI. Formed stools may be tested, however, in cases of ileus or toxic megacolon when stool is not passed often, such as in the case we presented earlier, in which the patient was no longer producing stool due to infection-driven intestinal stasis. Given the colonization state associated with C. difficile organisms or its toxins, a positive C. difficile test in a person without diarrhea does not imply CDI.
Some laboratories encounter more stool collection-related issues than others. Diarrhea is defined as three or more unformed stools passed within the past 24 consecutive hours. This definition can be somewhat difficult to apply to certain patients who may not be able to relay this information to healthcare providers. Besides, the difference between unformed stools and formed stools may also be difficult to discern. Laboratory guidelines address what constitutes a formed stool—generally defined as a stool which takes the shape of its container (6).
However, many laboratories now utilize liquid-based stool collection devices, which make it difficult to ascertain the solidity of a stool. These devices are projected to become ever more popular as clinical microbiology laboratories embrace laboratory automation. Laboratories that utilize liquid collections must rely on clinicians to submit only loose stool specimens.
Electronic decision-making tools have been explored to decrease the probability that laboratories receive unacceptable specimens. Some institutions use an electronic order check that cross-references patients’ medical chart for use of laxatives or administration of other agents known to cause loose stools (7).
The recent 2017 IDSA/SHEA clinical practice guidelines for CDI suggest different approaches to laboratory testing based on specimens’ pre-test probabilities (1).
For instance, if clinicians institution-wide agree to screen carefully for clinical symptoms associated with CDI (i.e., at least three loose stools or unformed stools within 24 hours with a history of antibiotic exposure), then a highly sensitive test such as NAAT or a multistep algorithm may be best. Conversely, NAAT alone would not be the best choice for laboratories that operate without institution-wide criteria for stool submission and that typically receive unscreened stools for testing. These labs would do well to  consider a multistep algorithm such as GDH plus toxin testing or NAAT plus toxin testing.
Laboratories may choose to utilize an algorithmic or multistep approach to CDI diagnosis with more than one testing methodology.
One common approach includes GDH testing combined with a toxin test, arbitrated by NAAT. Laboratories have many different approaches, but it is important to consider not only test performance but also other factors such as complicated algorithms for generating results when multistep testing methodologies are employed, as well as turnaround time (1). Table 2 lists some of the most commonly employed laboratory approaches to CDI diagnosis along with their advantages and disadvantages.

Labs also need to consider their patient populations when deciding which assays to use. Clinicians caring for patients with underlying gastrointestinal disorders that affect motility of the gastrointestinal tract have questioned the utility of relying only on a highly sensitive NAAT, since most of their patients suffer from loose stools at baseline.
The reasoning behind using a less sensitive assay for diagnosing CDI in this patient population presumably is that a toxin EIA would be less likely to detect C. difficile colonization than would a NAAT assay due to the lower sensitivity of the EIAs. The future of application of various C. difficile tests in different patient populations remains to be determined.
The final factor to consider when determining which assays to utilize for diagnosing C. difficile is the intent of testing.
Applying C. difficile testing as a screening test may require using a more sensitive detection platform than that used to diagnose C. difficiledisease. Active screening for asymptomatic carriers has been explored recently owing to the belief that these patients may be reservoirs of bacteria, potentially transferring the organisms to other patients (8). Timely environmental cleaning and contact precautions are important steps if institutions use this approach.
TREATMENT AND CONTROL OF CDI
The most significant risk factor for CDI is antibiotic exposure. Thus, providers begin therapy with an appropriate antimicrobial agent such as metronizadole or oral vancomycin and discontinue antimicrobial agents that may be predisposing to CDI. Best practices also call for contact precautions to be continued in patients with CDI for at least 48 hours after their diarrhea has resolved.
There is no clinical value in repeat CDI testing to establish whether a patient has been cured, as more than half remain positive for C. difficileeven after successful treatment and resolution of diarrhea (1).
The Food and Drug Administration (FDA) approved in 2011 a macrolide antimicrobial, fidaxomicin, for the treatment of CDI, making it only the second agent after vancomycin approved for this purpose. The recent IDSA/SHEA guidelines for CDI treatment recommend using either vancomycin or fidaxomicin rather than metronidazole for an initial episode of CDI (1). If, however, the initial episode is not severe, metronidazole may be considered if access to vancomycin or fidaxomicin is limited.
More recently, FDA approved in 2017 bezlotoxumab, a human monoclonal antibody that binds to and neutralizes C. difficile toxin B, to prevent CDI recurrence in adults receiving antibiotic treatment. A single intravenous infusion of this compound was associated with a significantly lower rate of recurrent infection in clinical trials.