Avoiding Colectomy during Surgical Management of Fulminant Clostridium difficile Colitis

Abstract

Background:

Clostridium difficile is the most common cause of nosocomial diarrhea in adults. Over the last decade, there has been a substantial increase in the disease-associated morbidity and mortality rate from this infection accompanied by identification of new hypervirulent strains. Fulminant colitis, a severe and complicated form of the disease that frequently necessitates surgical intervention, occurs in 3–8% of patients infected with C. difficile. The postoperative mortality rate for fulminant colitis continues to be dire, ranging from 34–57%.

Methods:

Review of the literature to offer insight into the dilemma associated with the surgical management of fulminant C. difficile colitis and provide alternatives to total abdominal colectomy for treatment.

Results:

Several recent studies have elucidated factors that contribute to the unacceptably high postoperative mortality rate: Surgical intervention too late in the course of the disease, lack of clearly defined guidelines for patient selection, and difficulty in predicting the clinical course of the disease. Perforation, need for vasopressor support, and end-organ damage all affect the postoperative mortality rate negatively.

Conclusion:

A high clinical suspicion and careful patient selection for colectomy is imperative to improve postoperative survival. An alternative surgical strategy for fulminant C. difficile colitis is laparoscopic creation of an ileostomy with total colonic washout.

C lostridium difficile , a spore-forming, anaerobic, gram-positive bacillus, is the most common cause of nosocomial diarrhea in adults. First discovered in 1935, C. difficile infection today accounts for up to 20% of all antibiotic-associated diarrheas [1] and can be found in as many as 30% of asymptomatic hospitalized patients [2,3]. Increases in both the number of infections and the severity of disease have made C. difficile a growing concern.

Epidemiology and Pathogenesis

In 2001, the number of reported C. difficile infections in U.S. hospitals increased from 30–40 cases per 100,000 to nearly 50 per 100,000. The number of cases continued to increase at an alarming rate, subsequently reaching 84 per 100,000 cases in 2005 [4]. As the incidence of CDI has continued to increase over the last 20 years, so has the severity of the disease, as seen in an increase in disease-associated morbidity and deaths.

The greatest medical concern regarding CDI is the progression of disease in some patients to fulminant colitis, a severe and complicated form of the disease that can be refractory to medical management and frequently necessitates surgical intervention for treatment. Fulminant colitis occurs in 3–8% of patients infected with C. difficile and can progress quickly to shock and death even despite appropriate initial medical treatment. Reports from the U.K. implicated CDI as the primary cause of death in 499 patients in 1999, a number that increased to an astounding 3,393 by the year 2006 [5]. Once believed to be a disease afflicting only elderly and institutionalized patients, C. difficile-associated disease (CDAD) can now be seen in young and previously healthy people [6]. Recent sporadic outbreaks of CDI (CDAD) in U.S. and international hospitals also have raised great concern. Most notable was the 2003 outbreak in Québec, Canada, where the number of reported (CDI) nearly quadrupled from 2002 to 2003, and was associated with a significant increase in disease severity [7]. During the peak of this epidemic, emergency colectomies for fulminant C. difficile colitis were performed nearly 20 times more often than in the pre-epidemic period [8]. In the U.S., McDonald et al. studied isolates of C. difficile associated with outbreaks in eight major health care facilities across six states from 2002–2003 [9]. A single strain accounted for as many as one-half the isolates in five facilities. Strikingly, this same strain was found in 82% of stool samples collected from patients during the Québec outbreak and has now been identified as the hypervirulent North American Pulsed Field type 1, polymerase chain reaction ribotype 027 (NAP-1/027).

The major virulence determinant of C. difficile is the production of toxins A and B. Toxin A is an enterotoxin that causes the excretion of inflammatory fluid from the epithelial cells of the colon [10], whereas toxin B is a cytotoxin. Binding of the B toxin to receptors on the surface of intestinal epithelial cells results in receptor-mediated endocytosis, glycosylation of Rho-GTPases, and subsequent disaggregation of the cytoskeleton and cell death [11]. Importantly, toxin-negative strains are non-pathogenic. Patients who are colonized with non-toxin-producing strains of C. difficile become asymptomatic carriers but do not develop disease unless they become infected with a toxin-producing strain also.

The startling increase in the mortality rate observed in recent clinical outbreaks of C. difficile infection secondary to complications from fulminant colitis has been linked in part to the emergence of the NAP-1/027 C. difficile strain. Further investigations into the unique properties of this strain revealed several factors that help account for its higher association with fulminant colitis. First, the greater virulence of the NAP-1/027 strain can be attributed in part to increased production of both toxins A and B. Isolates carry a mutation of tcdC, a regulatory gene responsible for inhibition of toxin transcription during the early exponential growth phase of the bacteria [12]. A deletion mutation in this gene leads to a greater than 10-fold increase in toxin production [13], which can account for its association with greater morbidity and a higher mortality rate. The second factor is the production of binary toxin, C. difficile transferase (CDT). The precise role of this actin-ADP-ribosylating toxin in C. difficile pathogenicity remains unclear, although recent studies suggest that its expression increases C. difficile cellular adherence by inducing the formation of long microtubule-based protrusions on the surface of colonic epithelial cells [14]. As many as 35% of C. difficile isolates produce binary toxin, which frequently is co-expressed with toxins A and B in hypervirulent strains [15]. Interestingly, strains producing CDT in the absence of toxins A and B do not appear to cause colitis, suggesting that CDT acts synergistically with toxins A and B [16,17]. Finally, recently identified strains of NAP-1/027, as well as other hypervirulent isolates, have high-level resistance to fluoroquinolones, an infrequent characteristic before 2001.

Despite changes in C. difficile pathogenicity over the last decade, the most important risk factor for CDI continues to be recent (<2 mos) exposure to antibiotic therapy, which is found in more than 96% of patients who develop the disease [18]. Antibiotic use alters the gastrointestinal microflora, which can facilitate C. difficile colonization and proliferation. The antibiotics implicated most often are clindamycin, ampicillin, amoxicillin, and cephalosporins, although C. difficile disease has been linked to almost every antibiotic, including metronidazole and vancomycin, two of the drugs used most commonly for treatment of CDI. Other important risk factors include a hospital stay longer than one week, immunosuppression, renal insufficiency, advanced age (65 years or older), and a recent gastrointestinal operation [19]. Patients who have undergone vascular or cardiac surgery also are at greater risk of developing C. difficile colitis.

Diagnosis

The diagnosis of C. difficile colitis with physical examination alone is unreliable. Patients can exhibit a variety of findings, ranging from mild abdominal distention to severe generalized peritonitis, which are all non-specific for various gastrointestinal pathologies. Therefore, clinicians have come to rely almost exclusively on laboratory and radiographic testing to make a diagnosis.

There are several laboratory tests available to verify the presence of toxin-producing CDI in the colon. The gold standard continues to be the cell culture cytotoxin assay, which detects toxin in the stool. Fecal filtrates are inoculated onto tissue culture plates and incubated for 24 and 48 h, evaluating at each time point for cell rounding to indicate a cytotoxic effect. The presence of toxin is then confirmed through cytotoxic neutralization with an antitoxin. This test has a reported sensitivity of 94%–100% and a specificity of 99%, but has been criticized for its turnaround time of 24–48 h [18]. In an attempt to provide quicker laboratory results, many centers have adopted an enzyme-linked immunosorbent assay to detect both toxins A and B in stool samples. Although results can be obtained within 2–6 h, the test has a sensitivity of only 70–90% when used alone, requiring serial repetitions to decrease the rate of false-negative results [18]. Anaerobic culture techniques also have been employed, although not routinely in the U.S. Culture provides greater sensitivity and the ability to determine the strain and its antimicrobial susceptibilities. However, cultures can take as long as five days to grow, are technically difficult to perform, and are not specific for toxigenic strains [18]. In a study performed by Delmée et al., 10,552 stool samples were analyzed over seven years using both fecal cytotoxin assay and anaerobic culture to determine if this protocol improved sensitivity and specificity for diagnosing disease-associated CDI (Table 1) [20]. If a sample was culture positive but cytotoxin assay negative, the colonies grown in culture underwent an in vitro assay to detect toxin A. The investigators found that of the 1,058 culture-positive samples, more than one-half (56%) had a negative direct cytotoxin assay result. Of these 598 cytotoxin-negative samples, 355 (33% of the total culture-positive samples) produced toxin A in the in vitro assay. Overall, the direct cytotoxin assay was able to detect only 57% of the toxigenic isolates. These results underline the problems inherent in current laboratory testing for toxigenic C. difficile strains: Lack of sensitivity of the cytotoxin assay alone compared with culture, plus the lack of specificity of anaerobic culture alone in detecing toxin-producing strains.

Table 1.

Summary of Cytotoxin Assay Results on Clostridium difficile Strains from Stool Cultures

Culture ^a	Cytotoxin assay	Toxin A positive	N (%)
−	−		9,427
+	+		460 (44)
+	−		598 (56)
+	−	−	243
+	−	+	355

Total culture-positive samples = 1,058.

Data from Delmée M, Van Broeck J, Simon A, et al. Laboratory diagnosis of Clostridium difficile-associated diarrhoea: A plea for culture. J Med Microbiol 2005;54:187–191 [20].

Inherent to the success of all laboratory assays in detecting C. difficile toxin is collection of an adequate stool sample from the patient. This can be a challenge in patients with fulminant colitis, where absence of diarrhea and stool is often the case, leading to crucial delays in confirmation of the diagnosis. In several studies examining patients with pathology-confirmed fulminant C. difficile colitis, toxin assays had a false-negative rate of 8%–12.5% [21 –23]. The number of C. difficile bacteria and the concentration of toxin in the fecal sample are linked to the probability of obtaining a positive cytotoxin assay result [20], which may explain why diagnosing C. difficile with stool studies becomes increasingly difficult as fulminant colitis develops and progresses.

The time delays and the lack of sensitivity and specificity of laboratory testing become more important when a patient develops fulminant colitis. In these patients, the use of abdominal computed tomography (CT) scans has emerged as a valuable tool for detection of CDI before toxin assay results or cultures become available. The major advantages of the abdominal CT scan are that it is non-invasive and can be obtained quickly and with relative ease, gives rapid real-time results, and can be repeated to monitor disease progression or resolution. Intravenous contrast medium does not have to be administered to yield an adequately sensitive examination. Colonic wall thickening of greater than 4 mm with edema can be seen with early CDI and is the most common sign found on CT scans [24 –26]. In a study performed by Ash et al. in 2006 [24] examining CT findings in adults with CDI, colonic wall thickening often was segmental, with the descending colon affected more frequently than the ascending colon (75% vs. 58% of all cases, respectively). Ascites, pericolonic fat stranding, and, less often, colonic distention also can be seen. The “accordion sign”—oral contrast material trapped between nodular thickened folds of bowel—is uncommon but is believed to be specific for CDI. The positive predictive value of CT scans alone for diagnosing C. difficile colitis is as high as 88% [25]. Computed tomography scans also can rule out other causes of abdominal pain and may play a role in monitoring the efficacy of medical treatment.

Endoscopy has been used in some cases to help distinguish colitis caused by CDI infection from other causes of diarrhea, although with limited success. The presence of pseudomembranes on the colonic mucosa—elevated yellow-white plaques composed of fibrin, inflammatory cells, and cellular debris—is pathognomonic for CDI. However, pseudomembranes are seen in only 50–60% of patients with C. difficile colitis [27]. Other findings include edema, erythema, and increased friability of the mucosa, as well as non-specific ulcers. Endoscopy is less valuable when the disease involves the right colon only, and the high incidence of rectal sparing (60–70%) makes less invasive proctosigmoidscopy examination inadequate for diagnosis. The false-negative rate for endoscopy can be as high as 25% [23]. The lack of sensitivity combined with higher costs, patient discomfort, and concern for bowel perforation during the examination have made endoscopy a less attractive diagnostic option. Most clinicians prefer CT scan to endoscopy for its timely results, ease of completion, and sensitivity for detection of early disease.

Spectrum of Disease

The signs and symptoms of CDAD caused by toxin-producing strains often are non-specific, with the spectrum of the disease ranging from mild diarrhea to severe, life-threatening fulminant colitis and death. Crampy abdominal pain with diarrhea, abdominal distention, and fever are the most common signs and symptoms of mild-to-moderate disease, usually occurring 4–9 days after the start of antibiotic exposure. Symptoms typically resolve within 10 days after the initiation of treatment [28]. Fulminant colitis is the most severe form of CDAD, with a mortality rate exceeding 40% [29]. Although there are no firm clinical criteria outlining the disease state, fulminant C. difficile colitis is defined broadly as colitis with systemic signs of sepsis. The systemic signs, including fever, tachycardia, and hypotension necessitating supportive care, are caused by the release of toxin-induced inflammatory mediators into the colon, not by bacteremia [30]. Hypotension is often a late and ominous sign of severe disease that can become refractory to volume and vasopressor support, leading to end-organ damage and subsequent death.

Although diarrhea is the most common presenting symptom of CDI, it can be strikingly absent in as many as 20% of patients with fulminant colitis [10], making early diagnosis of the disease a challenge. Physical examination can reveal abdominal distention and generalized tenderness with guarding, with radiographic evaluation showing megacolon, but progression to colonic perforation is rare. Patients who develop fulminant colitis often exhibit profound leukocytosis (30–50 × 10⁹/L) with bandemia, which usually precedes hypotension and end-organ dysfunction [21]. It remains unclear why some patients develop fulminant colitis, although inability to mount a sufficient antibody-mediated response to clostridial toxins has been suggested [31].

Management

Medical management

Most cases of CDI can be treated by ceasing all previous antibiotic therapy and starting metronidazole or vancomycin. Oral metronidazole and vancomycin have been used since the first descriptions of CDI in the early 1970s and continue to be the mainstay of medical treatment. Once treatment is initiated, symptoms typically resolve within 3–5 days in patients with mild-to-moderate CDI [32].

Studies in the 1980s showed no difference in efficacy between oral metronidazole and oral vancomycin [33,34]. Metronidazole became the preferred initial treatment mainly because of its low cost and the risk that vancomycin may induce colonization by vancomycin-resistant enterococci. With the recent increase in the incidence and severity of the disease, however, the failure rate of metronidazole therapy has increased also from 2.5% to approximately 18.2% [35], reigniting the debate over whether vancomycin should be first-line therapy in patients with severe infection. In 2007, Zar et al. performed a randomized trial comparing the efficacy of oral metronidazole vs. oral vancomycin in 172 patients with various degrees of disease severity [36]. The two drugs had the same efficacy in mild disease. In patients with severe disease (i.e., pseudomembranous colitis or intensive care unit admission), however, vancomycin was significantly more effective (97% vs. 76% for metronidazole). This observation can be attributed primarily to the more favorable pharmacologic profile of vancomycin. This drug is absorbed poorly as it passes through the gastrointestinal tract and is found in the gut lumen at concentrations more than 100-fold higher than the highest minimum inhibitory concentration. Metronidazole, on the other hand, is highly absorbed by the gut and is nearly undetectable in solid feces [37,38]. When diarrhea is present, however, the mean concentration of metronidazole in stool exceeds the minimum therapeutic concentrations for C. difficile [38]. This is likely attributable to an increase in gastrointestinal transit time, leading to incomplete absorption of the drug.

Understanding the pharmacologic profiles of vancomycin and metronidazole could be crucial in initiating the best initial medical treatment for fulminant C. difficile colitis. The severe ileus, bowel wall edema, and absence of diarrhea commonly observed in fulminant colitis would lead to greater absorption of metronidazole than of vancomycin, with less drug reaching the clostridial spores adherent to colon mucosa. The combination of increased toxin burden secondary to less efficient extrusion from the bowel and suboptimal drug delivery to the bowel lumen may lead to rapid progression of the disease and clinical deterioration. In patients who are not able to take medication orally, vancomycin delivered in enema form is recommended and may be supplemented with intravenous metronidazole. It is important to appreciate that the enema solution may not be able to reach all of the C. difficile spores. As fulminant colitis develops, the gut lumen becomes narrowed due to bowel wall edema, subsequently restricting passage of contents.

In addition to antibiotic therapy, fluid replacement, and electrolyte normalization, all antiperistaltic agents, including narcotics, should be discontinued. Antiperistaltic drugs inhibit the elimination of luminal toxin, allowing accumulation of toxin in the colon and exacerbating the disease.

Surgical management

In a subset of patients with CDI, fulminant colitis with severe systemic toxicity can develop despite appropriate medical management, or can be present on initial patient presentation. In either case, surgical intervention may be necessary. It is estimated that surgical intervention occurs in 0.17%–3.5% of patients with CDAD [39]. Partial colectomy with removal of the grossly affected portion of the colon was performed in the past in an attempt to decrease the morbidity of surgery. Currently, this practice has been rejected because affected bowel was often left behind, leading to relapses or progression of the disease. The current surgical standard of care for fulminant CDI is total abdominal colectomy with end-ileostomy. The mortality rate after colectomy for fulminant CDI is an astounding 34%–57% [8,21 –23,39,41], a statistic that has improved only marginally since the first recognition of CDI and has become a great source of frustration for surgeons with the recent increase in severe colitis cases. Why the postoperative mortality rate continues to be unacceptably high is an open question. Several studies published over the last decade have shed light on the answer, offering three critical factors that contribute to this surgical dilemma: (1) Surgical intervention frequently is delayed, occurring too late in the course of the disease; (2) patient selection is incorrect because of a lack of clearly defined guidelines; and (3) the clinical course of the disease is difficult to predict.

Delays in operating can be attributed to several factors, beginning with difficulties in diagnosis the disease. Early recognition of fulminant colitis has continued to be a challenge, especially in patients who, because of severe ileus or colon wall edema, do not have diarrhea. Fulminant colitis occurs along a spectrum ranging from mild diarrhea and abdominal discomfort to peritonitis, shock, and death, making it clinically more difficult to recognize. Without the hallmark episodes of diarrhea, the initial clinical presentation may be interpreted as an unexplained, non-specific gastrointestinal illness or an overall deterioration in medical condition, often in a patient with multiple acute and chronic medical problems. Stool sample testing is delayed because of low clinical suspicion for C. difficile, and can take as long as 48 h to produce results once the samples finally are sent. There also may be difficulty in obtaining appropriate samples if the patient does not have diarrhea. Moreover, physicians have become overly reliant on toxin assays to determine patient management, which are significantly less reliable in cases of fulminant CDI. When the diagnosis of fulminant colitis eventually is made, too often the patient's clinical status already has deteriorated substantially. In a retrospective study performed by Dallal et al. that evaluated the death rate and colectomy indications for patients with fulminant C. difficile colitis at the University of Pittsburgh Medical Center from 1989–2000, the correct diagnosis was made only at autopsy examination in an astounding 35% of their patients [21]. This observation underlines the difficulty of diagnosing fulminant CDI accurately and suggests that the death rate attributed to this disease likely is underestimated. Importantly, patients who died of fulminant CDI without surgical intervention were twice as likely to have a false-negative toxin assay result, reportedly as high as 12.5%. The findings from this study stress the crucial importance of maintaining a high clinical suspicion for fulminant CDI and highlight the error of heavy reliance on toxin assay results alone to guide management. Computed tomography scans of the abdomen in patients who are suspected of having fulminant CDI can help reveal severe inflammation of the colon and lead to earlier surgical intervention.

Another factor contributing to the postoperative mortality rate is the lack of clearly defined guidelines indicating if and when surgery should be performed for fulminant CDI. Although there is some variation among authors, common surgical criteria include perforation, toxic megacolon with impending perforation, peritonitis, severe sepsis or septic shock, need for vasopressor support, end-organ dysfunction, and colitis refractory to medical management [8,21,26]. The optimal timing of colectomy is still a topic of debate. Perforation is an obvious surgical indication, but is surprisingly rare with fulminant CDI. Moreover, if surgical intervention follows colonic perforation, the postoperative mortality rate can easily surpass 60% [26]. Several studies have shown that the requirement for vasopressor support increases significantly postoperative deaths, and is now considered widely to be a predictor of death after colectomy [21,23,26,40]. Colectomy after preoperative initiation of vasopressors nearly quadruples the perioperative death rate compared with colectomy before vasopressors are needed [21]. Similar results have been found with respect to postoperative mortality rates for patients who have preoperative end-organ dysfunction. Koss et al. performed a retrospective study of patients undergoing colectomy for fulminant CDI in an attempt to identify factors affecting surgical outcomes [41]. In addition to preoperative vasopressor support, they found that development of multiple organ dysfunction correlated strongly with an increase in the postoperative mortality rate. Sixty-seven percent of patients with end-organ failure died within 30 days after colectomy vs. 12.5% of patients without end-organ dysfunction. Unfortunately, the ominous turn toward end-organ failure can be difficult to recognize. Mental status changes are one heralding sign. In a retrospective study by Byrn et al. evaluating 73 patients who underwent emergency colectomy for fulminant CDI, preoperative mental status changes correlated significantly with a higher postoperative mortality rate, often occurring before liver or kidney dysfunction was evident [23]. The results of these studies clearly indicate that waiting to perform a colectomy until after there is perforation, the patient requires vasopressor support, or develops end-organ damage increases drastically postoperative mortality rates. These factors should be viewed as predictors of death rather than indications for surgery. When they are present, surgical intervention can do little to improve the clinical outcome.

Surgical decision-making is perhaps at its most difficult when interpreting the indication of “refractory” to medical management because the clinical course of the disease is so difficult to predict. There is neither an established time frame that defines how long medical management should be attempted nor identified factors that should trigger surgical intervention before perforation, shock, and end-organ damage occur. The challenge of predicting which patients will improve with medical management alone and which will suffer shock and death without a colectomy has remained a crucial issue in the surgical management of fulminant CDI. It has become increasingly evident that a narrow time window exists for therapeutic surgical intervention. Progression to hemodynamic instability is difficult to foresee and can occur within hours to days after the initial presentation. In an attempt to improve surgical outcomes, several studies have investigated what clinical factors may be used as clues to predict which patients will progress to shock if colectomy is not performed. Various laboratory parameters have been found to be associated closely with poor post-colectomy outcomes, offering guidance for surgical timing and patient selection. In a retrospective study by Lamontagne et al. examining patients with CDAD requiring intensive care unit admission from 2003–2005 in Québec, Canada, 23% (38 of 161 patients) underwent colectomy [22]. Colectomy was more beneficial with leukocytosis <50 × 10⁹/L or a lactate concentration <5 mmol/L. Similar findings were reported by Pepin et al., who examined 130 patients requiring emergency colectomy for fulminant C. difficile colitis from 1994–2007 (Table 2)[8]. In this retrospective study, predictors of postoperative death included a preoperative lactate concentration of ≥5 mmol/L (75% mortality rate), leukocytosis of ≥50 ×10⁹/L (73% mortality rate), and albumin <15 g/L (52% mortality rate). Selected laboratory parameters, therefore, can be used to predict disease progression. A rising white blood cell count or lactate concentration are continuous, quantitative parameters that can be used to track the disease along the spectrum of its evolution, and act as triggers for aggressive surgical intervention before the patient's clinical condition passes outside the therapeutic window. Careful patient selection for colectomy is imperative to improve survival. If colitis is allowed to progress beyond the therapeutic window, the surgical results are dismal.

Table 2.

Preoperative Factors Associated with Higher Postoperative Mortality Rates after Colectomy for Fulminant Clostridium difficile Colitis

Factor	Mortality rate (%)	P value
White blood cell count (× 10⁹/L)
≥50	73	0.006
<20	32
Lactate concentration (mmol/L)
≥5	75	0.001
≤2.1	26
Vasopressor support
Yes	50	0.005
No	36

Data from Pepin J, Vo T, Boutros M, et al. Risk factors for mortality following emergency colectomy for fulminant Clostridium difficile infection. Dis Colon Rectum 2009;52:400–405 [8].

Alternative surgical strategy

Laparotomy with total colectomy and end-ileostomy is currently the only surgical procedure that can be offered to patients with fulminant C. difficile colitis as the standard of care. However, this procedure has continued to be associated with a dire postoperative mortality rate. In addition to the problems of inappropriate patient selection and a lack of defined guidelines for surgical intervention, early colectomy may not be accepted at first by the patient, his or her family, or the primary medical physician. The reluctance to accept an extensive, invasive, morbid procedure early in the disease course before obvious impending death can lead to operating too late, when death is almost unavoidable. An alternative surgical strategy that is less invasive and can avoid the morbidity of a total colectomy could improve patient acceptance of early surgical treatment and improve the postoperative mortality rate.

A crucial understanding of the disease process is necessary to propose a suitable solution. Clostridium difficile colitis is caused by cytotoxin-mediated cell death and enterotoxin-mediated secretion of inflammatory cytokines into the bowel lumen. Fulminant colitis develops when life-threatening systemic symptoms occur secondary to excessive toxin accumulation in the bowel. The disease itself is not an invasive process; rather, the colon serves as a reservoir for the toxin-producing bacteria. One of the major hurdles in treating fulminant C. difficile colitis is colonic stasis, attributable either to ileus or to severe edema of the bowel wall, causing significant narrowing of the lumen, which results in inefficient drug delivery to the infected colon. Oral antibiotics may be unable to reach the distal colon, and enemas can be unsuccessful in reaching the proximal colon, leading to a significant portion of bowel that is not exposed to antibiotic therapy. The toxin burden in the colon can become considerable as stool excretion slows or stops, exacerbating the systemic response. Total abdominal colectomy has become the surgical treatment of choice because most efforts to remove the offending toxins, including enema therapy, are unsuccessful. In the majority of cases in which perforation or bowel ischemia has not occurred, a therapy that could remove the toxin burden in the bowel as well as deliver treatment to the entire colon could obviate removal of the colon. One possible alternative is laparoscopic creation of a double-barrel ileostomy for distal colonic washout. The concept of intestinal washout is not new and has been shown to improve sepsis in animal and human studies, although only limited data exist [42 –44]. It can be delivered safely if there is close monitoring. Moreover, the minimally invasive approach combined with the possibility of ileostomy reversal could encourage earlier acceptance of surgical intervention by the patient. In patients who have deteriorated to a point where end-organ damage is present or vasopressor support is necessary, a less-morbid laparoscopic operation may be better tolerated. A colectomy may be avoided or, alternatively, colonic washout may improve the clinical condition to permit a later colectomy. Table 3 summarizes the proposed indications for ileostomy with antegrade colonic lavage. In all cases, it is possible that an improvement in the postoperative mortality rate can be achieved.

Table 3.

Proposed Indications for Ileostomy with Antegrade Colonic Lavage

Suspected or established diagnosis of CDI plus one or more of the following:

• Need for intensive care unit admission

• Hypotension necessitating vasopressor support

• Neurologic changes

• Respiratory failure necessitating mechanical ventilation

• Increasing white blood cell count ≥20 × 10⁹/L

• Lactate concentration ≥5 mmol/L

• Other signs of end-organ damage

Conclusion

Not all patients with fulminant C. difficile colitis require surgery, but for those patients who fail to improve with medical intervention alone, the disease can be fatal without proper patient selection and prompt surgical intervention. Recent retrospective studies have shown many of the current indications for colectomy to be invalid, leading to surgical intervention too late in the spectrum of the disease (i.e., after perforation, end-organ dysfunction, or cardiovascular collapse) and inappropriate patient selection, resulting in an unacceptably high postoperative mortality rate. Increasing leukocytosis or serum lactate concentration in a patient with mental status changes or an unexplained deterioration in clinical condition should elicit a high level of suspicion for fulminant CDI. A CT scan of the abdomen can aid in determining the diagnosis and guiding surgical decision-making. Early surgical intervention is crucial to achieve a favorable outcome; however, it can be met with resistance by the patient and the family, as well as by the treating physician. A total abdominal colectomy is a considerable invasive procedure associated with substantial morbidity. The laparoscopic creation of an ileostomy for total colonic washout provides an alternative strategy for surgical treatment of fulminant CDI and may decrease postoperative mortality.

Footnotes

Author Disclosure Statement

No competing financial interests exist for any author.

Presented at the 29th Annual Meeting of the Surgical Infection Society, Chicago, Illinois, May 6–9, 2009.

References

Kelly

, Pothoulakis

, LaMont

. Clostridium difficile colitis. N Engl J Med, 1994; 330:257–261.

Settle

, Wilcox

, Fawley

. Prospective study of the risk of Clostridium difficile diarrhea in elderly patients following treatment with cefotaxime or piperacillin-tazobactam. Aliment Pharmacol Ther, 1998; 12:1217–1223.

Wiesen

, Gossum

, Preiser

. Diarrhoea in the critically ill. Crit Care, 2006; 12:149–154.

Kelly

, Lamont

. Clostridium difficile—More difficult than ever. N Engl J Med, 2008; 359:1932–1940.

Office for National Statistics, UK Statistics Authority. National Statistics UK. www.statistics.gov.uk

McDonald

, Owings

, Jernigan

. Clostridium difficile infection in patients discharged from US short-stay hospitals, 1996 to 2003. Emerg Infect Dis, 2006; 12:409–415.

Pepin

, Valiquette

, Alary

. Clostridium difficile-associated diarrhea in a region of Québec from 1991 to 2003: A changing pattern of disease severity. CMAJ, 2004; 171:466–472.

Pepin

, Vo

, Boutros

et al. Risk factors for mortality following emergency colectomy for fulminant Clostridium difficile infection. Dis Colon Rectum, 2009; 52:400–405.

McDonald

, Kilgore

, Thompson

. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med, 2005; 353:2433–2441.

10.

Hurley

, Nguyen

. The spectrum of pseudomembranous enterocolitis and antibiotic-associated diarrhea. Arch Intern Med, 2002; 162:2177–2184.

11.

Reineke

, Tenzer

, Rupnik

. Autocatalytic cleavage of Clostridium difficile toxin B. Nature, 2007; 446:415–419.

12.

Warny

, Pepin

, Fang

. Toxin production by an emerging strain of Clostridium difficile associated with outbreaks of severe disease in North America and Europe. Lancet, 2005; 366:1079–1084.

13.

MacCannell

, Louie

, Gregson

. Molecular analysis of Clostridium difficile PCR ribotype 027 isolates from Eastern and Western Canada. J Clin Microbiol, 2006; 44:2147–2152.

14.

Schwan

, Stecher

, Tzivelekidis

et al. Clostridium difficile toxin CDT induces formation of microtubule-based protrusions and increases adherence of bacteria. PLoS Pathogenes, 2009; 5,10:1–14.

15.

Goncalves

, Decre

, Barbut

, Burghoffer

, Petit

. Prevalence and characterization of a binary toxin (actin-specific ADP-ribosyltransferase) from Clostridium difficile. J Clin Microbiol, 2004; 42:1933–1939.

16.

Barbut

, Decre

, Lalande

. Clinical features of Clostridium difficile-associated diarrhoea due to binary toxin (actin-specific ADP-ribosyltransferase)-producing strains. J Med Microbiol, 2005; 54:181–185.

17.

Geric

, Carman

, Rupnik

et al. Binary toxin-producing, large clostridial toxin-negative Clostridium difficile strains are enterotoxic but do not cause disease in hamsters. J Infect Dis, 2006; 193:1143–1150.

18.

Kelly

, LaMont

. Clostridium difficile infection. Annu Rev Med, 1998; 49:375–390.

19.

Clabots

, Johnson

, Olson

. Acquisition of Clostridium difficile by hospitalized patients: Evidence for colonized new admissions as a source of infection. J Infect Dis, 1992; 166:561–567.

20.

Delmée

, Van Broeck

, Simon

et al. Laboratory diagnosis of Clostridium difficile-associated diarrhoea: A plea for culture. J Med Microbiol, 2005; 54:187–191.

21.

Dallal

, Harbrecht

, Boujoukas

et al. Fulminant Clostridium difficile: An underappreciated and increasing cause of death and complications. Ann Surg, 2002; 235:363–372.

22.

Lamontagne

, Labbe

, Haeck

et al. Impact of emergency colectomy on survival of patients with fulminant Clostridium difficile colitis during an epidemic caused by a hypervirulent strain. Ann Surg, 2007; 245:267–272.

23.

Byrn

, Maun

, Gingold

et al. Predictors of mortality after colectomy for fulminant Clostridium difficile colitis. Arch Surg, 2008; 143:150–154.

24.

Ash

, Baker

, O'Malley

Jr . Colonic abnormalities on CT in adult hospitalized patients with Clostridium difficile colitis: Prevalence and significance of findings. AJR Am J Roentgenol, 2006; 186:1393–1400.

25.

Kirkpatrick

, Greenberg

. Evaluating the CT diagnosis of Clostridium difficile colitis: Should CT guide therapy? AJR Am J Roentgenol, 2001; 176:635–639.

26.

Longo

, Mazuski

, Virgo

et al. Outcome after colectomy for Clostridium difficile colits. Dis Colon Rectum, 2004; 47:1620–1626.

27.

Bergstein

, Kramer

, Wittman

et al.

Pseudomembranous colitis: How useful is endoscopy?

Surg Endosc, 1990; 4:217–219.

28.

Kyne

, Merry

, O'Connell

. Factors associated with prolonged symptoms and severe disease due to Clostridium difficile. Age Ageing, 1999; 28:107–113.

29.

Poutanen

, Simor

. Clostridium difficile-associated diarrhea in adults. Can Med Assoc J, 2004; 171:51–58.

30.

Flegel

, Muller

, Daubener

. Cytokine response by human monocytes to Clostridium difficile toxin a and b. Infect Immun, 1991; 59:3659–3666.

31.

Kelly

. Immune response to Clostridium difficile infection. Eur J Gastroenterol Hepatol, 1996; 8:1048–1053.

32.

Wilcox

, Howe

. Diarrhoea caused by Clostridium difficile: Response time for treatment with metronidazole and vancomycin. J Antimicrob Chemother, 1995; 36:673–679.

33.

Cherry

, Portnoy

, Jabbari

et al. Metronidazole: An alternate therapy for antibiotic-associated colitis. Gastroenterology, 1982; 82:849–851.

34.

Teasley

, Gerding

, Olson

. Prospective randomized trial of metronidazole versus vancomycin for Clostridium difficile-associated diarrhea and colitis. Lancet, 1983; 2:1043–1046.

35.

Musher

, Aslam

, Logan

. Relatively poor outcome after treatment of Clostridium difficile colitis with metronidazole. Clin Infect Dis, 2005; 40:1586–1590.

36.

Zar

, Bakkanagari

, Moorthi

, Davis

. A comparison of vancomycin and metronidazole for the treatment of Clostridium difficile-associated diarrhea, stratified by disease severity. Clin Infect Dis, 2007; 45:302–307.

37.

Aslam

, Hamill

, Musher

. Treatment of Clostridium difficile-associated disease: Old therapies and new strategies. Lancet Infect Dis, 2005; 5:549–557.

38.

Bolton

, Culshaw

. Faecal metronidazole concentrations during oral and intravenous therapy for antibiotic associated colitis due to Clostridium difficile. Gut, 1986; 27:1169–1172.

39.

Hermsen

, Dobrescu

, Kudsk

. Clostridium difficile infection: A surgical disease in evolution. J Gastrointest Surg, 2008; 12:1512–1517.

40.

Sailhamer

, Carson

, Chang

et al. Fulminant Clostridium difficile colitis: Pattern of care and predictors of mortality. Arch Surg, 2009; 144:433–439.

41.

Koss

, Clark

, Sanders

et al. The outcome of surgery in fulminant Clostridium difficile colitis. Colorectal Dis, 2005; 8:149–154.

42.

Videla

, Vilaseca

, Guarner

. Polyethylene glycol lavage prevents inflammatory lesions induced by trinitrobenzenesulfonic acid in the rat colon. Gastroenterology, 1994; 106:A790.

43.

Wellmann

, Fink

, Benner

, Schmidt

. Endotoxaemia in active Crohn's disease: Treatment with whole gut irrigation and 5-aminosalicyclic acid. Gut, 1986; 27:814–820.

44.

Alverdy

, Piano

. Whole gut washout for severe sepsis: Review of technique and preliminary results. Surgery, 1997; 121:89–94.