Epidemiology, Diagnosis and Treatment of Clostridium Difficile Infection

Matteo Bassetti; Giovanni Villa; Davide Pecori; Alessandra Arzese; Mark Wilcox


Expert Rev Anti Infect Ther. 2012;10(12):1405-1423. 

In This Article


Treatment of CDI represents a challenge to the clinician. As recommended by the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) guidelines,[126] a key measure for treating CDI ideally includes the discontinuation of concurrent antibiotic therapy to allow the intestinal flora to restore itself; in practice, the patient's clinical condition does not always allow this. Another important issue is represented by recurrences. Present and future therapeutic strategies must therefore share two goals: first, eradicating the infection despite the continuation of concomitant antibiotic therapy and second, reducing the incidence of recurrence.

Metronidazole and vancomycin have been the mainstay treatment options for CDI. However, before 2007, only two prospective randomized trials had compared these antibiotics, and these were not blinded or placebo-controlled and were underpowered to detect a difference in efficacy. Teasley et al. randomized 94 patients to receive vancomycin or metronidazole therapy and demonstrated a 95% cure rate for patients who received metronidazole therapy and a 100% cure rate for patients who received vancomycin; the results did not reach statistical significance.[127] In the study of Wenisch et al. (31 patients randomized to receive metronidazole or vancomycin therapy), the cure rates in both groups was 94%.[128]

In 2007, Zar et al. published the first prospective, randomized, double-blind, placebo-controlled trial comparing vancomycin and metronidazole for the treatment of CDI, stratified by disease severity.[129] They enrolled 172 patients with 150 successfully completing the trial. They were stratified into mild and severe disease groups according to a severity assessment score, incorporating six clinical variables, specifically developed for the study. Among patients with mild CDAD, treatment with metronidazole or vancomycin resulted in clinical cure in 90 and 98% of the patients, respectively (p = 0.36). Clinical cure was reached in 76 and 97% in patients with severe CDAD, respectively, with metronidazole or vancomycin (p = 0.02). In this study, the superior efficacy of vancomycin was demonstrated, but this benefit is limited only for severe cases. These results were endorsed by ESCMID guidelines, where vancomycin is recommended for first-line treatment of severe CDI (A-I recommendation).

Pépin et al. found that metronidazole was not inferior to vancomycin for treating patients with a first recurrence of CDI.[130] In this observational study, the authors reviewed the outcomes for patients with a first recurrence of CDI in their institution between 1991 and mid-2004. When metronidazole was given as the treatment of a first recurrence of CDI, patients did not experience second recurrences more frequently than those receiving vancomycin, regardless of which of the two agents had been used to treat the initial episode. Independent predictors of a second recurrence were age and above all the duration of hospitalization after the first recurrence. This suggests that a significant proportion of these patients experienced new infections rather than true relapses. The authors concluded that the decision to administer vancomycin as treatment for a first recurrence of CDI should be based solely on the presence of markers of severe disease at the time of symptom recurrence (PMC, hypotension, white blood cell count of 15,000 cells/ml or higher or a rising serum creatinine level).

Second or later recurrence of CDI should be treated with vancomycin therapy, using a tapered and/or pulse regimen, as suggested by ESCMID guidelines (B-II recommendation). Kelly and LaMont suggested an approach for the treatment of the second recurrence as indicated: 125 mg 4 times a day (q.i.d.) for 14 days, 125 mg twice a day (b.i.d.) for 7 days, then 125 mg once a day (q.d.) for 7 days, 125 mg once every 2 days for 8 days (four doses) and lastly 125 mg once every 3 days for 15 days (five doses).[131] However, this approach has not been examined in a robust comparative study and various regimens have been proposed.

Rifaximin is a molecule, derivative of rifamycin, which has been proposed as a rescue option in the treatment of second and later recurrences. C. difficile usually shows good in vitro susceptibility to rifaximin, although high-level resistant strains have emerged.[132] Small case series showed reduction of recurrence rates with rifaximin treatment after initial treatment with vancomycin. Johnson et al. reported an approach using rifaximin immediately after vancomycin treatment for the most recent episode of CDAD. Eight women who each experienced at least four episodes of CDAD were given a 2-week course of oral rifaximin at a dosage of 400–800 mg daily (in two or three divided doses) when they were asymptomatic, immediately after completing vancomycin therapy; all of them had not responded to multiple, previous treatment regimens. Seven out of the eight patients had no further recurrence of symptoms, with duration of follow-up ranging from 51 to 431 days. The eighth patient responded to a further course of rifaximin but a rifaximin-resistant C. difficile was recovered post-treatment.[133] In a further study by the same group of authors, four out of six patients with multiple CDI recurrences did not have another episode following the rifaximin regimen, with follow-up ranging from 4 to 26 months.[134] A recent randomized, double-blind, placebo-controlled pilot study found that patients given rifaximin 400 mg t.i.d. for 20 days, given immediately after finishing standard anti-CDI antibiotics, had a decreased incidence of recurrent diarrhea. While these results are interesting, the intensive antibiotic use in this approach raises concerns about prolonged flora disturbance and the potential for the selection of rifamycin resistance.[135]

Ramoplanin is a new lipoglycodepsipeptide which has been compared with vancomycin in one single clinical trial, showing noninferiority in cure and relapse rate.[136] The authors concluded that further data are required, but ramoplanin, in contrast to vancomycin, does not promote clinically relevant resistances in gut microbiota.

Nitazoxanide, a nitrothiazolide compound with good antimicrobial activity against a range of helminthic and protozoal parasites, is active against anaerobic bacteria, included C. difficile .[137] Two small clinical trials demonstrated noninferiority of nitazoxanide, compared with metronidazole and vancomycine, in the treatment of CDI.[138,139] Further and larger studies are required to assess the safety profile.

Fidaxomicin (previously known as OPT-80) is a macrocyclic, bactericidal antibiotic, more active than vancomycin in vitro against clinical isolates of C. difficile, including NAP1/BI/027 strains. Citron et al. reported the MIC90 for fidaxomicin was 0.125 µg/ml, with no significant difference observed between NAP1 and non-NAP1 strains.[140] The mechanism of action is through inhibition of the bacterial sigma subunit involved in protein synthesis. Fidaxomicin has a narrow antimicrobial spectrum, with activity against Gram-positive anaerobes and most strains of staphylococci and enterococci (except for Staphylococcus aureus and Enterococcus faecalis), but negligible activity against Gram-negative bacteria and fungi. There is minimal systemic absorption of fidaxomicin, and although high fecal concentrations are achieved, there is limited inhibition of normal gut flora.[141] Fidaxomicin was approved by the FDA in May 2011 for the treatment of CDI in adults. In 2011, Louie et al. published the results of the first Phase III prospective, multicenter, randomized, controlled, parallel-group, noninferiority study comparing fidaxomicin with vancomycin in 629 patients with CDI (596 patients were included in the modified intention-to-treat analysis and 548 [87.1%] were evaluated for the per-protocol analysis [PP]).[80] Patients could not received antibiotics active against C. difficile for more than 24 h prior to enrollment. Patients received medication every day for 10 days, every 6 h: this was either 125 mg of vancomycin every 6 h or 200 mg of fidaxomicin every 12 h with intervening matching doses of placebo. Clinical cure, the primary end point, was defined as the resolution of symptoms and no need for further therapy from the second day after the end of treatment; recurrence (diarrhea with a positive stool toxin test within 4 weeks after treatment) and global cure (cure without recurrence) represented secondary end points.

Fidaxomicin was noninferior to vancomycin in terms of clinical cure: this was demonstrated in both the modified intention-to-treat analysis (88.2% with fidaxomicin and 85.8% with vancomycin) and in the PP (92.1 and 89.8%, respectively). However, fidaxomicin was superior to vancomycin in terms of rate of recurrences. The modified intention-to-treat analysis showed a rate of recurrence for fidaxomicin of 15.4%, compared with 25.3% of vancomycin (p = 0.005); the PP confirmed these data (13.3 vs 24.0%; p = 0.004). However, the hypervirulent strains NAP1/BI/027 did not prove to be more susceptible to fidaxomicin in this context, being the rate of recurrence substantially the same between the groups treated with fidaxomicin and the one which received vancomycin.

No differences between the two groups were demonstrated in terms of adverse events or serious adverse events. The most common adverse events that were related to the study treatments were mild GI symptoms, and they had similar rate in the two groups (9.7 and 9.0% for fidaxomicin and vancomycin, respectively).

The same design of the above-mentioned study was used in the second Phase III multicenter, double-blind, randomized, noninferiority clinical trial.[142] Cornely et al. enrolled 535 patients from 45 hospitals in Europe and 41 hospitals in the USA and Canada. Of these, 509 were included in the modified intention-to-treat analysis (mITT), with 451 in the PP. In the latter population, 198 out of 216 (91.7%) patients in the fidaxomicin group achieved clinical cure, compared with 213 out of 235 (90.6%) in the vancomycin group, proving that fidaxomicin is not inferior to vancomycin. Noninferiority was also proved for clinical cure in the mITT population, with 221 out of 252 (87.7%) patients treated with fidaxomicin and 223 out of 257 (86.8%) treated with vancomycin that reached clinical cure. By analyzing population subgroups (age, inpatient status, previous CDI, infection with NAP1/BI/027 strain, severity, exposure to concomitant antibiotics and region) there was not a statistically significant difference in clinical cure with the two drugs for any subgroup, with the exception of patients exposed to concomitant antibiotics during treatment; this subgroup had a higher cure rate with fidaxomicin (46 out of 51; 90.2%) than with vancomycin (33 out of 45; 73.3%; p = 0.031).

The secondary end points for the mITT population and the PP population were very similar to those of the other Phase III trial, with decreased rates of recurrence and improved global cure in the fidaxomicin group. When recurrence was assessed for the NAP1/BI/027 strain, there was a nonsignificant decrease in recurrence in the fidaxomicin group compared with vancomycin (12 out of 54 [22.2%] vs 18 out of 47 [38.3%]; p = 0.08). Regarding safety, incidences of treatment adverse events were not significantly different between the groups; adverse events possibly or definitely related to the study treatment were primarily GI symptoms in both.

Despite guidelines recommending discontinuing all implicated drugs at the onset of CDI, individuals frequently require antibiotic treatment during CDI therapy to manage concurrent systemic infections. Mullane et al. carried out a post-hoc analysis (of the above two Phase III studies) to determine the efficacy of fidaxomicin versus vancomycin in individuals taking concomitant antibiotics (CAs) for other concurrent infections.[143] Rates of cure, recurrence and global cure (cure without recurrence) were assessed for subgroups of subjects defined by CA use and treatment group. CAs was prescribed for 27.5% of 1164 patients during study participation (at the same time as CDI treatment or during 4 weeks of follow-up). The use of CAs concurrent with CDI treatment was associated with a lower cure rate (84.4 vs 92.6%; p < 0.001) and an extended time to resolution of diarrhea (97 vs 54 h; p < 0.001). Subjects who achieved clinical cure were analyzed for recurrence within 28 days of completing treatment: CA use during the follow-up was associated with more recurrences (24.8 vs 17.7%; p= not significant). CA administration at any time was associated with a lower global cure rate (65.8 vs 74.7%; p = 0.005). In the absence of CA use, fidaxomicin and vancomycin were equivalent in achievement of clinical cure by the end of treatment (92.3 vs 92.8%, respectively; p = 0.8). When subjects received at least one CA concurrently with study drug, fidaxomicin was superior to vancomycin in achieving clinical cure: 90.0 versus 79.4%; p = 0.04. In subjects receiving CAs during treatment and/or follow-up, treatment with fidaxomicin compared with vancomycin was associated with 12.3% fewer recurrences (16.9 vs 29.2%; p = 0.048).

In a letter to the editor published in the New England Journal of Medicine on May 2011,[144] Linsky et al. reported the results of their retrospective, cohort study. They observed that patients who did not receive PPIs had a lower rate of recurrence of CDI within 90 days compared with patients treated with PPIs during treatment for CDI, independent of the treatment regimen.[145]

Louie et al. replied to this comment that the use of PPIs and H2-receptor antagonists was similar in the vancomycin and in the fidaxomicin group, showing data from both Phase III trials. No difference was reported in the rate of recurrence of CDI among patients with or without exposure to PPIs and H2-receptor antagonists. In the fidaxomicin group, the rate of recurrence was 15.8% with exposure versus 13.0% without exposure (p = 0.38); in the vancomycin group, the rate of recurrence was 27.0% with exposure versus 24.8% without exposure (p = 0.42). Adjusting for age, markers of severity of illness and exposure to PPIs and H2-receptor antagonists, data showed that the only significant predictor of recurrence was the molecule used for the treatment of CDI.

Tigecycline is a glycylcycline derivative of minocycline, which inhibits bacterial protein synthesis. Tigecycline achieves fecal concentrations well above the MIC for C. difficile, because of primary biliary excretion of unchanged drug; it also does not induce C. difficile toxin production in vitro.[146] These attributes support a potential use of tigecycline in CDI, although it is not licensed for this indication and there are no randomized trials. There are case reports showing full recovery from severe CDI following tigecycline treatment for 7–24 days.[147] Previous standard therapy had failed for three out of the four patients, whereas one patient was treated primarily with tigecycline. None of these patients experienced recurrence during follow-up of various lengths. Recently, El-Herte et al. reported a case of C. difficile colitis that had not responded to a combination therapy with metronidazole and oral vancomycin.[148] The patient's abdominal conditions worsened and required a surgical intervention, which the patient refused; he was then successfully treated with a combination of tigecycline and rifaximin. In 2010, a case report was reported showing failure of tigecycline in combination with vancomycin, metronidazole and intravenous immunoglobulin (IVIG) in the treatment of severe CDI in a 70-year-old man: during tigecycline treatment, the patient developed a Proteus mirabilis bacteremia, was colonized with Acinetobacter baumannii resistant to tigecycline and finally died.[149] The authors urged caution against indiscriminate use of tigecycline for off-label indications.

To access safety and efficacy of the molecule, a randomized controlled trial is needed; tigecycline is therefore currently recommended only where other standard options have failed.

Since pathogenicity of C. difficile is largely mediated by its toxins, agents with the potential of binding these targets have been tested. The most researched compound in this class is tolevamer, a soluble, high-molecular-weight, anionic polymer that neutralizes C. difficile toxins A and B. Tolevamer has no antimicrobial activity, so it does not interfere with the normal intraluminal bacterial flora. Tolevamer also neutralizes toxins produced by specific strains of the NAP1/BI/027 clone.[150] In two Phase III studies, one in North America[151] and one in Europe, Australia and Canada,[152] tolevamer showed lower cure rates (42–46%), if compared with vancomycin and metronidazole (46–81 and 72–72%, respectively), despite a lower recurrence rate (3–6% for the polymer, 18–23% and 19–27% for vancomycin and metronidazole, respectively). From these results it can be observed that tolevamer represents a weaker option of treatment if compared with standard molecules, but it might keep a potential role for the treatment of recurrent conditions as supplemental therapy. Further studies are needed to prove this hypothesis.

Synsorb 90 is an inert support (developed by Synsorb Biotech, AB, Canada) carrying a specific trisaccharide receptor called Galα1-3Galβ1-4GlcNAc, usually located on intestinal cells, which binds C. dfficile toxin A. Despite encouraging results of a Phase II study demonstrating a reduction of the C. difficile relapse rate, development of this drug was abandoned.[153]

Vaccination against C. difficile appears feasible and, for example, a toxoid vaccine has been shown to induce serum anti-toxin A and B IgG antibody levels that are associated with protection against recurrent CDI.[154] Sougioultzis et al. described three patients with recurrent CDI who were vaccinated with intramuscular administration of toxin A and B. None of patients developed recurrence of disease after vaccination within the 6-months follow-up.[155] Recently, a Phase I dose-finding study of an adjuvanted C. difficile toxoid vaccine has been reported.[156] Fifty healthy adult (18–55 years) and 48 elderly (≥65 years) volunteers were randomized to receive a candidate C. difficile toxoid vaccine (2, 10 or 50 µg) or placebo on days 0, 28 and 56. No one in the placebo group developed specific antibodies. For toxin A, seroconversion by day 56 (post-dose 2) was observed in 100% of volunteers aged 18–55 years in all dose groups and in 50, 89, and 100% of elderly participants in the 2, 10 and 50 µg dose groups, respectively. For both age groups, seroconversion for toxin B was lower than toxin A. There were no safety concerns; the most commonly reported vaccine-related adverse events were injection site reactions. Since the highest responses were in the 50 µg dose group, this is the dose that will be evaluated in further studies. Among novel treatment modalities for CDI, use of IVIG has been proposed. Bovine colostrum contains high concentration of immunoglobulins and has been proposed as a treatment for CDI.[157,158]

IVIG has been utilized off-label to treat both refractory and fulminant CDI despite the lack of large randomized controlled trials. As recently described in a paper by Abougergi and Kwon,[159] fifteen small, mostly retrospective and nonrandomized reports documented successful treatment of protracted, recurrent or severe CDI with IVIG. The mechanism of action is through neutralization of mainly toxin A by IgG anti-toxin A antibodies. Anti-toxin B antibodies play an adjunctive role in conferring immunity against CDI when added to anti-toxin A antibodies, but they do not have any significant role on their own. Aldeyab et al. recently published a case–control study about the impact of a single-dose IVIG regimen in the treatment of CDI.[160] The objective was to assess the impact of administration of a single-dose of IVIG (400 mg/kg) on the following patient outcomes: length of stay in the hospital until discharge following the first positive CDI toxin test result, 30-day clinical outcomes (recovered/recovering, ongoing infection), and requirement for surgery. Cases involved 18 patients who received a combination of the standard treatment (metronidazole and/or vancomycin) and IVIG treatment; controls were 18 patients who received only the standard treatment. There were no statistically significant differences in outcomes in this small study.

On this matter, a randomized, double-blind, placebo-controlled study was designed by Lowy et al., where two fully human monoclonal antibodies were used, one against C. difficile toxin A (CDA1), the other toxin B (CDB1).[161] They were administered together in patients with symptomatic CDI that were receiving a treatment either with metronidazole or with vancomycin. A single infusion of 10 mg/kg each was used. The primary outcome was the documented rate of recurrence within the first 84 days following the administration of either the antibodies or the placebo. Overall, recurrence rate was lower in the group treated with the antibodies (7 vs 25%; p < 0.001); patients infected by BI/NAP1/027 strains also experienced a lower recurrence rate (8 vs 32%; p = 0.06). Where recurrences were more than one, the antibodies showed superior results, and recurrence rates were 7% in the treatment group compared with 38% of placebo (p = 0.006). Given these results, monoclonal antibodies may represent an interesting option for the treatment of severe and/or recurrent CDI.

As discussed above, a critical role in the pathogenesis of CDI is determined by the alterations in the intestinal flora the antimicrobial therapy determines. Therefore, eradication of CDI might be achieved through the restoration of the appropriate gut microbiota. This can be obtained by use of probiotics (especially Saccharomyces boulardii and Lactobacilli) or by stool instillation. A Cochrane review assessed the efficacy of probiotics in the treatment of CDI.[162]

This review included studies that were randomized and prospective, were probiotics were administered with or without conventional-antibiotics. Only four studies met the inclusion criteria, and all had problems in terms of size and methodology. Out of these four, only three studies included recurrent cases of CDI, whereas both first episodes and recurrent cases were considered in the fourth. In the latter, it was found that the administration of S. boulardii improved the outcome of both patients with first episode of CDI and the one of those with recurrent disease, if compared with placebo (recurrent rate 34.6% in the S. boulardii group vs 64.7% in the placebo group; p = 0.04).[163] The other studies of the review did not show adjunctive benefit if probiotics were used.

Lactobacilli have been tested too for their possible role in treating and preventing CDI. Studies conducted so far, such as the trial of Wullt et al.,[164] are too small and results cannot be considered statistically relevant.

Therefore, the evidence in the literature is not sufficient to recommend the use of probiotics for the treatment of first episode or recurrent cases of CDI, neither alone nor associated with antibiotics.[165] Moreover, treatment with probiotics represents a risk factor for fungemia in patients with immunodeficiency, critically ill or with CVC.[166]

There is great interest about prevention of AAD and CDAD with the use of probiotics, especially S. boulardii, but large and well-controlled studies are needed to confirm the efficacy of this approach. An Italian group of authors recently reported results of a single-center, randomized, double-blind, placebo-controlled, parallel-group trial.[167] Two hundred and seventy five patients were randomized to receive a capsule containing S. boulardii or an indistinguishable placebo twice daily within 48 h from the beginning of antibiotic therapy, and then continued the treatment for 7 days after antibiotic withdrawal. They were subsequently followed for 12 weeks after ending antibiotic treatment (204 patients completed the follow-up). Onset of AAD was reported in 13 out of the 98 patients who received the placebo (13.3%) and in 16 out of the 106 patients who were treated with S. boulardii (15.1%). The AAD rate was not statistically different between these groups (p = 0.71). Five patients with diarrhea had a C. difficile toxins test positive: two cases were in the placebo group (2.0%) and three in the S. boulardii group (2.8%; p = 1.00 between groups). No difference was observed in mortality rates (12.7 vs 15.6%; p = 0.60). According to these data, S. boulardii was not effective in preventing the development of AAD and CDI.

Stool instillation, known as 'fecal transplantation', can restore a normal microbiota in the gut. This process is characterized by the instillation of a liquid suspension of stool into the patient's upper GI tract through a nasogastric/nasoduodenal catheter. Another approach requires the instillation of the stool suspension into the colon of the patient through colonoscopy or a rectal catheter. The stool suspension comes from a healthy donor. The first authors that reported a successful treatment with this approach were Eiseman et al., who used fecal enemas to treat four patients with PMC,[168] while Tvede and Rask-Madsen successfully cured six patients with recurrent CDI thanks to stool instillation in 1989.[169] In the latter report, one patient received the stool suspension from two family members, while for the other five a mixture of ten different bacterial species, facultative aerobic and anaerobic species, were used. Lately, this approach to the treatment of relapsing CDI has been used in northern Europe and in the USA, with positive results.

In summary, current data on fecal transplantation are very promising.[170] A prospective, randomized trial of fecal transplants for recurrent CDI has taken place in The Netherlands (FECAL trial).[205] This project aims to compare the efficacy of oral treatment with vancomycin, oral vancomycin and bowel lavage and the two combined with fecal transplantation for the treatment of relapsing cases. For the time being, until final results are published, fecal transplantation is reserved to the treatment of recurring CDI, where all other approaches have failed.