Atypical presentation of Clostridioides difficile pseudomembranous colitis with laboratory rejection of stool specimen

Abstract

Clostridioides (formerly Clostridium) difficile is a major cause of nocosomial infection in the USA and worldwide. It has a wide spectrum of presentation, ranging from an asymptomatic carrier state to fulminant colitis. Pseudomembranous colitis is a manifestation of severe C. difficile infection (CDI), typically with progressive symptoms including watery diarrhoea, abdominal cramping and fevers and elevated white cell count and/or creatinine. It is diagnosed on three levels, including clinical assessment, stool assays and visualisation of the colonic mucosa. Laboratories will reject stools that do not meet criteria for testing. In the era of molecular testing for the presence of toxigenic C. difficile DNA, which only indicates the potential for infection, it is vital to use clinical evaluation in the diagnosis of CDI. We present an atypical case of pseudomembranous colitis affecting the right colon in a patient whose stools were rejected multiple times for C. difficile testing.

Background

Clostridioides (formerly Clostridium) difficile is an anaerobic Gram-positive bacterium that is the leading cause of healthcare-associated diarrhoea in the USA and a major cause of healthcare-associated infection globally.¹ From 2000 to 2008, there was a 3.5-fold increase in US hospital stays with a primary diagnosis of Clostridioides difficile infection (CDI).² According to the CDC, in 2011, C. difficile caused 453 000 infections in patients throughout the USA.³ In the USA that year, there were an estimated 29 300 deaths within 30 days of the initial diagnosis of CDI.³ One meta-analysis estimated that between 2005 and 2015, the total annual cost related to CDI was $6.3 billion US dollars.⁴

To address the burden of C. difficile in healthcare, measures such as diagnostic stewardship have become widely used. Diagnostic stewardship programmes aim to encourage evidence-based usage of diagnostic tests to ultimately improve the quality of care and decrease costs.⁵ They have been shown to reduce the amount of inappropriate inpatient diagnostic testing.⁶ As more commercial diagnostic testing for CDI have been developed, the optimal choice for diagnostic test is still up for debate. While the gold standard for diagnosing CDI is cytotoxicity assay or cyotoxogenic culture, these are impractical due to longer turnaround times and higher labour intensity and thus are not widely used.⁷ Commonly used commercial testing methods include nucleic acid amplification tests (NAATs) and enzyme immunoassays (EIA).^{8 9} EIA has high specificity, up to 99%, but lower sensitivity from 29% to 86%.^{9 10} NAAT has a higher sensitivity of 96% and specificity of 94% compared with the cell cytotoxicity assay.⁸ The lower sensitivity of EIA increases the chance for false negatives, while the higher sensitivity of NAATs suggests the possibility of false positives in asymptomatic carriers of C. difficile. In one study, up to 55% of inpatients with a positive NAAT had negative clinical toxin tests, suggesting they were only colonised with C. difficile and may not have required treatment.¹¹

Some hospitals have implemented computerised clinical decision support (CCDS) tools to decrease the number of unnecessary tests. The University of Virginia health system developed a CCDS tool for C. difficile that presents a series of questions to the provider upon ordering the C. difficile test. The tool was designed to guide appropriate testing based on the 2017 IDSA guidelines for C. difficile.¹² In the most recent IDSA guidelines, the preferred population for CDI testing were patients with new and unexplained ‘unformed stools’, or a 6–7 on the Bristol stool scale, with a frequency of three or more of such stools in a 24-hour period. These patients should not have diarrhoea that is attributable to other conditions such as chemotherapy, laxatives or irritable bowel disease (IBD). They recommend laboratories reject stools that are not liquid or soft (conform to shape of a container) to improve specificity.¹³

While reducing the amount of unnecessary testing may decrease cost and improve care, the underutilisation of tests may potentially lead to infections being underdiagnosed and undertreated. We present an atypical presentation of pseudomembranous colitis (PMC) that led to lab rejection of stool specimens.

Case presentation

A 61-year-old man presented to his primary care doctor with 5 weeks of intermittent diarrhoea, following a hospitalisation for a new diagnosis of epilepsy less than 3 months prior. He described his stool as foul smelling, varying from loose and watery to semiformed. He noticed that some stools floated and an occasional sheen on the toilet water. He was averaging 6–12 bowel movements per day and was experiencing nocturnal symptoms, waking up 5–6 times per night. He described associated abdominal cramping, streaks of blood in his stool and low-grade fevers at night. The patient reported losing about 20–30 pounds following his hospital discharge 3 months prior. He tried loperamide twice daily without improvement.

Recent medication changes included starting levetiracetam and completion of a course of doxycycline, both about 3 months prior to presentation. There was no other antibiotic exposure. He had never smoked and drank two glasses of wine per night. His family history was negative for inflammatory bowel disease and colon cancer. A colonoscopy 6 years prior was unremarkable.

The patient was noted to be afebrile, normotensive, saturating well on room air and with heart rate within normal range. His cardiovascular, pulmonary and abdominal exams were benign. Preliminary lab work included a complete blood count, which was significant for leucocytosis at 12.7×10⁹/L, mild normocytic anaemia with a haemoglobin of 132 g/L and normal platelets. His erythrocyte sedimentation rate was normal, but C reactive protein was elevated at 7.9 mg/dL (normal range <0.5 mg/dL). Serum creatinine was normal. The stool sample sent by the primary care provider (PCP) for C. difficile testing was rejected as it was too formed. The patient was referred for a gastroenterology consultation.

Investigations

Both the PCP and the gastroenterologist considered CDI the most likely diagnosis based on the history of hospitalisation and antibiotic exposure and their temporal relationship to symptom onset. They both ordered the only confirmatory test available to them in their system, C. difficile NAAT polymerase chain reaction (PCR) confirmatory testing. C. difficile PCR was not performed, however, because two separate stool samples were rejected by the laboratory for being too formed. The first ordered by PCP and later the second ordered by the gastroenterologist. Stool community gastrointestinal PCR was negative for common causes of community acquired diarrhoeal illnesses including diarrhoeagenic Escherichia coli, Campylobacter, norovirus or adenovirus. Stool ova and parasites were negative. Faecal occult blood was positive. Calprotectin was elevated at 580.0 µg/g (>120 µg/g is abnormal). Amylase, lipase, zinc, copper, elastase, celiac panel, antinuclear antibodies, rheumatoid factor and thyroid stimulating hormone were all normal.

The patient was scheduled for an oesophagogastroduodenoscopy (OGD) and colonoscopy with random duodenal and colon biopsies. OGD revealed a normal stomach and duodenum, and biopsies were normal. Colonoscopy was significant for a diffuse areas of moderately congested and pseudomembrane-covered mucosa throughout the entire colon, more severe in right than left colon (figure 1). Biopsies were taken with cold forceps for histology. A specimen for C. difficle was obtained from the right colon after washing the area of diffuse exudate. The specimen obtained from the right colon showed acute inflammatory infiltrate and formation of inflammatory pseudomembrane. C. difficile PCR from samples obtained during the colonoscopy were positive for C. difficile.

Figure 1.

These scattered yellow/white plaques seen in the ascending colon are pathognomonic for pseudomembranous colitis.

Differential diagnosis

The initial differential diagnosis for the patient’s heme-positive diarrhoea was broad. Infectious causes such as C. difficile or other common community associated diarrhoeal illnesses were certainly high on the differential. Other considerations included inflammatory bowel disease, pancreatic insufficiency, coeliac disease and irritable bowel syndrome. Colonoscopy and microbiology ultimately yielded the diagnosis of PMC due to C. difficile.

Treatment

The patient was treated with fidaxomicin 200 mg twice per day for a total 10-day course. Fidaxomicin is a bactericidal macrocyclic lactone antibiotic that acts by inhibiting RNA polymerase, thereby preventing transcription. Phase III clinical trials have shown that fidaxomicin is non-inferior to vancomycin in terms of clinical response to therapy and superior to vancomycin for achieving a sustained clinical response and preventing recurrent CDI.¹⁴

Outcome and follow-up

Approximately 1 month later, the patient reported that the intermittent diarrhoea had resolved, but he was still experiencing chronic fatigue, intermittent low-grade fever and foul-smelling stool. His stool was sent for repeat testing and was negative for C. difficile.

Discussion

This is an atypical presentation of PMC where multiple stools were rejected for C. difficile diagnostic testing.

CDI can present with a wide range of clinical manifestations. These can range from an asymptomatic carrier state to fulminant colitis.¹⁵ Anywhere from 7% to 15% of healthy adults are colonised with C. difficile, and most of these people are asymptomatic.^{14 15} Only a small minority of individuals develop clinical signs or symptoms of C. difficile infection.¹⁶

PMC is a less common manifestation of full-blown C. difficile colitis, with a frequency of about 1%–5%.¹⁷ These patients often have more severe symptoms than those patients without pseudomembranes.¹⁵ The frequency of diarrhoea is significant, occurring in about 99% of patients, while fever (29%), leucocytosis (61%) and abdominal pain/cramping (33%) are also common symptoms.¹⁸ Our patient interestingly did not have persistent diarrhoea with his presentation, as his stools were intermittently formed.

The most well-known risk of developing PMC is exposure to broad-spectrum antibiotics. Clindamycin, fluoroquinolones, broad-spectrum penicillins and cephalosporins have been associated with the highest risk.^{17 19} However, the use of any antibiotic can also increase the risk of C. difficile colonisation.¹⁵ Other risk factors include patients 65 years or older, those on gastric acid suppressive medication or with comorbid illnesses.¹ Our patient had received doxycycline less than 3 months prior to presentation, and this antibiotic exposure and recent hospitalisation were his biggest risk factors for developing CDI.

Diagnosis of PMC involves clinical evaluation, stool or tissue assay for a causative pathogen and visualisation of colonic mucosa.¹⁷ On endoscopic evaluation, pseudomembranes appear as scattered raised yellow plaques (figure 1), ranging from 2 mm to 10 mm in diameter.¹⁴ In one prospective study, about 91% of patients with PMC had involvement of the rectosigmoid area, but only 9% were found to have PMC in the more proximal colon.²⁰

In a time where diagnostic stewardship and antimicrobial stewardship programmes are becoming more prevalent, many laboratories have started rejecting stools that do not ‘take the shape’ of the container.¹³ The evidence to determine the optimal number of diarrhoeal episodes to justify CDI testing is weak, and the number of unformed stools to justify testing has changed over the years. As described in the current Infectious Diseases Society of America (IDSA) guidelines, patients with three or more new and unexplained unformed stools in a 24-hour period are the preferred population to test. However, one study found that only 64% of patient whose stools were submitted for C. difficile testing met the clinical definition of diarrhoea. They used the definition of ≥3 watery bowel movements, type 6 or 7 on the Bristol stool scale or diarrhoea plus abdominal pain.²¹ In one study, the negative predictive value of having no significant diarrhoea or antibiotic exposure was 94%, which means that only 6% of patients who lacked one or both criteria had positive C. difficile cytotoxin assays.²² Our patient’s stools were rejected on two separate occasions due to being too formed for testing, although he reported stools ranging from liquid to soft semiformed stools, suggestive of Bristol stool scale 5, 6 or 7.

In summary, our patient’s case was unusual in that he presented with stools too formed for laboratory testing for C. difficile and was found to have PMC in the right colon. This case illustrates the broad spectrum of presentation of C. difficile and emphasises the importance of using clinical judgement in the diagnosis of C. difficile. If one’s clinical suspicion for CDI is high enough, contacting the laboratory to run the sample regardless of the character of stool may be important to achieve the correct diagnosis.

Learning points.

• Clostridiodes difficile has a wide spectrum of presentation, ranging from asymptomatic carrier state or mild self-limiting diarrhoea to fulminant colitis.

• While implementing diagnostic stewardship to reduce the amount of unnecessary testing may decrease cost and improve care, the underutilisation of tests may potentially lead to serious infections being underdiagnosed and undertreated.

• It is important to use one’s clinical judgement when considering the diagnosis of C. difficile and test patients for C. difficile infection (CDI) when one’s clinical suspicion is high.

• If one’s clinical suspicion for CDI is high enough, we encourage the provider to call the laboratory to test the sample regardless of stool characteristics.