Advertisement

Can we identify patients at high risk of recurrent Clostridium difficile infection?

  • C.P. Kelly
    Correspondence
    Corresponding author: C. P. Kelly, Gastroenterology Division, Beth Israel Deaconess Medical Center, DA 601, 330 Brookline Ave, Boston, MA 02215, USA
    Affiliations
    Gastroenterology Division, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, USA
    Search for articles by this author

      Abstract

      Although most patients with Clostridium difficile infection (CDI) can be managed effectively with discontinuation of prescribed antibiotics and additional treatment with oral metronidazole or vancomycin, up to 25% experience disease recurrence, usually within 30 days of treatment. Failure to mount a systemic anti-toxin antibody response differentiates patients with CDI and recurrent CDI from symptomless carriers of toxinogenic C. difficile. The immunological senescence that accompanies ageing may lead to impaired immune responses to C. difficile and contribute to the significant association between advancing age and increased risk of CDI recurrence. Inadequate immunity may also explain why previous episodes of recurrence constitute a significant risk factor for further CDI recurrences. Other risk factors for recurrent CDI include concurrent use of antibiotics for non-C. difficile infections (which perpetuate the loss of colonization resistance), proton-pump inhibitors, and other gastric acid anti-secretory medications, prolonged hospitalization, and severe underlying illness (as reflected by a high Horn index score). Prominent risk factors have been examined to develop and validate a clinical prediction tool for recurrent CDI, with three factors (age >65 years, severe underlying disease (by the Horn index score), and continued use of antibiotics for non-CDI infections) being highly predictive of CDI recurrence. Such simple clinical prediction rules have the potential to identify patients at high risk of recurrent CDI, and can alert the treating physician to the need for prompt recognition, confirmatory diagnosis and treatment with regimens ideally designed to mitigate the risk of subsequent recurrences.

      Keywords

      Introduction

      First identified in 1935, Clostridium difficile emerged as an important pathogen in the late 1970s, following the introduction of broad-spectrum antibiotics [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile infection.
      ]. Because most patients experienced few complications at that time [
      • Kelly CP.
      • Pothoulakis C.
      • LaMont JT.
      Clostridium difficile colitis.
      ], C. difficile infection (CDI) was generally regarded as a mildly troublesome side effect of antibiotic use. The marked increase in the incidence and severity of CDI seen in the past decade means that CDI is now viewed as a serious healthcare-associated infection linked to high rates of morbidity and mortality [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile—more difficult than ever.
      ,
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Loo VG.
      • Poirier L.
      • Miller MA
      • et al.
      A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality.
      ].
      Most patients who develop CDI today require specific antibiotic therapy to eliminate C. difficile from the colon, but nonetheless some can be managed conservatively. This entails stopping all antimicrobial therapy that may have been implicated in the onset of diarrhoea, requesting stool tests for C. difficile or its toxins, monitoring patients' progress, and then starting specific antibiotic therapy if symptoms and signs of CDI worsen or persist. Patients with mild to moderate CDI generally receive first-line treatment with oral metronidazole; oral vancomycin is often reserved for more severe cases [
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
      ,
      • Cohen SH.
      • Gerding DN.
      • Johnson S
      • et al.
      Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA).
      ]. These antibiotics are considered to be effective for the treatment of a first episode of CDI, although metronidazole is coming under increased scrutiny because of growing reports of treatment failure. Prior to 2000, reported treatment failure rates were <10%, whereas failure rates approaching 20% have been recorded for metronidazole since 2000 [
      • Aslam S.
      • Hamill RJ.
      • Musher DM.
      Treatment of Clostridium difficile-associated disease: old therapies and new strategies.
      ].
      Irrespective of which antibiotic is used as first-line therapy for an initial acute episode of CDI, a significant minority of patients experience disease recurrence following cessation of therapy. The evidence suggests that rates of CDI recurrence are increasing from those seen in the 1980s and 1990s [
      • Aslam S.
      • Hamill RJ.
      • Musher DM.
      Treatment of Clostridium difficile-associated disease: old therapies and new strategies.
      ], and that the emergence of hypervirulent strains of C. difficile, such as ribotype 027, may have been a contributory factor. Data from a retrospective review of patient records in Canada showed that, coincident with the emergence of the 027 strain outbreak in Quebec, the 60-day probability of CDI recurrence with metronidazole increased significantly from 20.8% (between 1991 and 2002) to 47.2% (in 2003-2004; p <0.001), suggesting a possible link with the 027 strain [
      • Pépin J
      • Alary M-E
      • Valiquette L
      • et al.
      Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada.
      ]. Among patients infected with the 027 strain of C. difficile in the two randomized controlled trials (RCTs) that compared fidaxomicin with vancomycin, rates of recurrence were significantly higher than those seen in patients infected with other strains (27.4% vs. 16.6%, respectively; p 0.002) [
      • Petrella LA.
      • Sambol SP.
      • Cheknis A
      • et al.
      Decreased cure rate and increased recurrence rate for Clostridium difficile infection caused by the epidemic C. difficile BI strain.
      ]. In fact, infection with the 027 strain emerged as a significant risk factor for recurrence in a multivariate analysis of these data (OR 1.57; 95% CI 1.01-2.45; p 0.046) [
      • Petrella LA.
      • Sambol SP.
      • Cheknis A
      • et al.
      Decreased cure rate and increased recurrence rate for Clostridium difficile infection caused by the epidemic C. difficile BI strain.
      ]. Data from the individual trials showed that the significant reductions in recurrence observed in the fidaxomicin treatment arm vs. vancomycin occurred primarily in patients infected with strains of C. difficile other than 027 [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Cornely OA.
      • Crook DW.
      • Esposito R
      • et al.
      Fidaxomicin versus vancomycin for infection with Clostridium difficile in Europe, Canada, and the USA: a double-blind, non-inferiority, randomised controlled trial.
      ]. There was no significant difference in rates of recurrence between the two treatments for patients infected with the 027 strain of C. difficile [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Cornely OA.
      • Crook DW.
      • Esposito R
      • et al.
      Fidaxomicin versus vancomycin for infection with Clostridium difficile in Europe, Canada, and the USA: a double-blind, non-inferiority, randomised controlled trial.
      ]. Overall, the results from these trials showed that, in the vancomycin comparator arm, c. 25% of patients experienced recurrent CDI in the 30-day follow-up period [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Cornely OA.
      • Crook DW.
      • Esposito R
      • et al.
      Fidaxomicin versus vancomycin for infection with Clostridium difficile in Europe, Canada, and the USA: a double-blind, non-inferiority, randomised controlled trial.
      ]. Such recurrence rates are similar to those observed for metronidazole (19%) and vancomycin (18%) in an RCT of the toxin-binding agent tolevamer [
      • Bouza E.
      • Dryden M.
      • Mohammed R
      • et al.
      Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
      ], and rates of 23% and 32%, respectively, in the antibiotic treatment arms of RCTs of adjunctive anti-toxin monoclonal antibody therapy [
      • Lowy I.
      • Molrine DC.
      • Leav BA
      • et al.
      Treatment with monoclonal antibodies against Clostridium difficile toxins.
      ].
      Thus, recurrent CDI represents one of the most difficult and increasingly common challenges that we face in the management of CDI. This problem is exacerbated by the dramatic and sustained increase in the incidence of primary CDI in both North America and many parts of Europe [
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
      ,
      • Kelly CP.
      • Kyne L.
      The host immune response to Clostridium difficile.
      ]. This review will look at the factors that predispose patients to CDI recurrence, examine risk factor indices that might be used to improve CDI management, and evaluate the practicalities of applying such risk factor indices in the clinical setting.

      Pathophysiology and Frequency of CDI Recurrence

      The key stages in the pathogenesis of CDI are illustrated in Fig. 1. CDI is most likely to develop when a patient is exposed to either the bacterium or its spores and, at the same time, the normal colonic microflora is disturbed following a course of antimicrobial therapy. Depletion of Bacteroides and other colonic bacteria leads to a decrease in the overall diversity of the colonic microbiota [
      • Rea MC.
      • O'Sullivan O.
      • Shanahan F
      • et al.
      Clostridium difficile carriage in elderly subjects and associated changes in the intestinal microbiota.
      ,
      • Chang JY.
      • Antonopoulos DA.
      • Kaira A
      • et al.
      Decreased diversity of the fecal microbiome in recurrent Clostridium difficile-associated diarrhea.
      ] and the subsequent loss of colonization resistance. This allows opportunistic infection with C. difficile to flourish [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile infection.
      ,
      • Tvede M.
      • Rask-Madsen J.
      Bacteriotherapy for chronic relapsing Clostridium difficile diarrhoea in six patients.
      ] and, where richness of the colonic microflora is markedly depleted, for CDI to recur [
      • Chang JY.
      • Antonopoulos DA.
      • Kaira A
      • et al.
      Decreased diversity of the fecal microbiome in recurrent Clostridium difficile-associated diarrhea.
      ]. Primary infection is facilitated by the fact that C. difficile is resistant to many commonly used antibiotics, with the exception of those used to treat CDI [
      • Rupnik M.
      • Wilcox MH.
      • Gerding DN.
      Clostridium difficile infection: new developments in epidemiology and pathogenesis.
      ]. CDI develops in patients who acquire toxinogenic strains of C. difficile, which typically produce both toxin A (encoded by TcdA) and toxin B (encoded by TcdB), and who lack protective immunity. The evidence suggests that patients who become colonized with toxinogenic C. difficile and are also able to mount an adequate systemic immune response to C. difficile and its toxins are less likely to develop CDI [
      • Kelly CP.
      • Kyne L.
      The host immune response to Clostridium difficile.
      ]. Many subjects colonized by toxinogenic strains of C. difficile remain asymptomatic, and this is associated with high serum levels of specific anti-toxin IgG antibodies, which are lacking in those with symptomatic CDI [
      • Kyne L.
      • Warny M.
      • Qamar A.
      • Kelly C.
      Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
      ]. Furthermore, when symptomatic patients mount an anti-toxin immune response early in the course of infection with toxinogenic C. difficile, they become less likely to experience recurrent CDI [
      • Kelly CP.
      • Kyne L.
      The host immune response to Clostridium difficile.
      ].
      Figure thumbnail gr1
      FIG. 1.Outcomes following exposure to toxinogenic strains of Clostridium difficile [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile infection.
      ]. Adapted, with permission, from the Annual Review of Medicine, Volume 49 © 1998 by Annual Reviews http://www.annualreviews.org.
      Today, patients with CDI generally receive treatment with either a 10-14-day course of oral metronidazole or vancomycin, depending on the severity of the disease [
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
      ]. The majority of patients will respond to such therapy with complete resolution of symptoms within 7-10 days [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile infection.
      ,
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
      ,
      • Teasley DG.
      • Gerding DN.
      • Olson MM
      • et al.
      Prospective randomised trial of metronidazole versus vancomycin for Clostridium-difficile-associated diarrhoea and colitis.
      ,
      • Zar FA.
      • Bakkanagari SR.
      • Moorthi KM
      • et al.
      A comparison of vancomycin and metronidazole for the treatment of Clostridium difficile-associated diarrhea, stratified by disease severity.
      ]. However, as we have seen, up to 25% of patients will experience disease recurrence within 30 days following cessation of treatment with metronidazole or vancomycin [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Bouza E.
      • Dryden M.
      • Mohammed R
      • et al.
      Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
      ,
      • Lowy I.
      • Molrine DC.
      • Leav BA
      • et al.
      Treatment with monoclonal antibodies against Clostridium difficile toxins.
      ] (Fig. 2). Recurrence of CDI usually manifests as diarrhoea and other signs and symptoms of CDI, and is confirmed by the recurrence or persistence of a positive stool test result for C. difficile or its toxins [
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
      ].
      Figure thumbnail gr2
      FIG. 2Frequency of recurrent Clostridium difficile infection (CDI) following an initial episode and first and second recurrence [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Bouza E.
      • Dryden M.
      • Mohammed R
      • et al.
      Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
      ,
      • Lowy I.
      • Molrine DC.
      • Leav BA
      • et al.
      Treatment with monoclonal antibodies against Clostridium difficile toxins.
      ,
      • McFarland LV.
      • Elmer GW.
      • Surawicz CM.
      Breaking the cycle: treatment strategies for 163 cases of recurrent Clostridium difficile disease.
      ,
      • McFarland LV.
      • Surawicz CM.
      • Greenberg RN
      • et al.
      A randomized placebo-controlled trial of Saccharomyces boulardii in combination with standard antibiotics for Clostridium difficile disease.
      ].
      Most recurrences occur within the first 30 days of completing a course of anti-CDI antibiotic therapy, and may arise from a resumption of the primary infection (relapse) or from a new exposure to C. difficile (re-infection). The term ‘recurrent CDI’ will be used to denote either relapse or re-infection, as the diagnosis and management of both forms of recurrence are similar, and the two mechanisms are rarely differentiated in clinical practice. Relapse and re-infection may occur with similar frequencies [
      • Poutanen SM.
      • Simor AE.
      Clostridium difficile-associated diarrhea in adults.
      ,
      • Barbut F.
      • Richard A.
      • Hamadi K
      • et al.
      Epidemiology of recurrences of Clostridium difficile-associated diarrhea: relapse versus reinfection.
      ], although continued exposure to C. difficile is necessary for re-infection to occur. Ironically, as metronidazole and vancomycin both alter the colonic microflora, either drug may perpetuate a loss of colonization resistance to C. difficile, and may therefore predispose to recurrence [
      • Kelly CP.
      • LaMont JT.
      Clostridium difficile infection.
      ]. Resistance to metronidazole and vancomycin has, to date, not been identified as a significant clinical problem in CDI management, and is not therefore considered to be a contributor to CDI recurrence [
      • DuPont HL.
      The search for effective treatment of Clostridium difficile infection.
      ]. However, it is important to note in this context that standard antibiotic therapy for CDI is ineffective against C. difficile spores in the gut [
      • Gerding DN.
      • Muto CA.
      Owens RC Jr. Treatment of Clostridium difficile infection.
      ].
      Recurrent CDI places a heavy burden on patients, from increased morbidity and diminished quality of life associated with repeated episodes of diarrhoea [
      • Johnson S.
      Recurrent Clostridium difficile infection: a review of risk factors, treatments, and outcomes.
      ]. Furthermore, mortality occurs in recurrent CDI as it does in primary infection. Patients with recurrent CDI also serve as a reservoir of infection that can lead to secondary infection in other vulnerable patients [
      • DuPont HL.
      The search for effective treatment of Clostridium difficile infection.
      ]. Thus, reducing the risk of recurrent CDI has the potential to improve patients' quality of life through reductions in CDI-associated morbidity and mortality as well as by reducing the risk of person-to-person transmission [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ]. Reducing recurrent CDI can also lead to commensurate reductions in inpatient and outpatient costs associated with treating additional episodes of CDI and from the need to isolate patients.

      Risk Factors for Recurrent CDI

      Once patients have experienced one recurrence of CDI, they are at significantly increased risk of further recurrences [
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ]. In fact, studies have shown that the risk of recurrence more than doubles after two or more recurrences [
      • McFarland LV.
      • Elmer GW.
      • Surawicz CM.
      Breaking the cycle: treatment strategies for 163 cases of recurrent Clostridium difficile disease.
      ,
      • McFarland LV.
      • Surawicz CM.
      • Greenberg RN
      • et al.
      A randomized placebo-controlled trial of Saccharomyces boulardii in combination with standard antibiotics for Clostridium difficile disease.
      ] (Fig. 2). Some patients experience multiple recurrences, leading to repeated bouts of diarrhoea and long courses of antibiotic therapy [
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ]. Continued disruption of the normal colonic microflora by repeated cycles of antibiotic therapy used to treat recurrent CDI perpetuates the risk of repeated recurrences.
      Efforts to discover why some patients are at greater risk of experiencing CDI recurrence have identified a number of predisposing factors in addition to prior recurrence. The risk of recurrence increases with increasing age, and patients aged ≥65 years are at significantly greater risk of recurrent CDI than younger patients [
      • Pépin J
      • Alary M-E
      • Valiquette L
      • et al.
      Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada.
      ,
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ]. One study showed a two-fold increase in recurrence rates for those aged ≥64 years as compared with those aged <64 years [
      • Pépin J
      • Alary M-E
      • Valiquette L
      • et al.
      Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada.
      ] (Fig. 3). A systematic analysis of 12 studies (refined from a total of 1215 studies initially identified), which included 1382 patients with CDI, has confirmed these findings, and shown that older age (≥65 years) is significantly associated with increased risk of recurrent CDI (p 0.0012) [
      • Garey KW.
      • Dao-Tran TK.
      • Jiang ZD
      • et al.
      A clinical risk index for Clostridium difficile infection in hospitalised patients receiving broad-spectrum antibiotics.
      ]. This systematic analysis, for which rigorous eligibility criteria were used, also showed that patients were at increased risk of recurrent CDI if they were receiving concomitantly administered non-C. difficile antibiotics following a diagnosis of CDI, or were receiving gastric acid-suppressing medications, especially proton-pump inhibitors [
      • Garey KW.
      • Dao-Tran TK.
      • Jiang ZD
      • et al.
      A clinical risk index for Clostridium difficile infection in hospitalised patients receiving broad-spectrum antibiotics.
      ]. However, in this systematic analysis, potential risk factors were evaluated only if studied in at least three independent publications that met the quality inclusion criteria. For this reason, several prominent risk factors for recurrent CDI were unable to be studied, e.g. underlying disease severity (as measured by the Horn index) and anti-toxin immune response.
      Figure thumbnail gr3
      FIG. 3The association between advancing age and risk of recurrent Clostridium difficile infection (CDI) (data from 2003 to 2004 during an outbreak of severe C. difficile infection in Quebec, Canada) [
      • Pépin J
      • Alary M-E
      • Valiquette L
      • et al.
      Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada.
      ]. Reproduced, with permission, from Oxford University Press.
      Given that the development of a protective antibody-mediated immune response to C. difficile toxins appears to be critical with regard to whether or not patients become symptomless carriers of C. difficile or develop symptomatic CDI, the immune status of the host would be expected to influence susceptibility to CDI recurrence. Studies have indeed shown an association between serum IgG anti-toxin antibody responses at day 12, the point at which specific anti-CDI antibiotic therapy will usually be discontinued, and protection against recurrent CDI (Fig. 4) [
      • Kyne L.
      • Warny M.
      • Qamar A.
      • Kelly CP.
      Association between antibody response to toxin A and protection against recurrent Clostridium difficile diarrhea.
      ]. This suggests that an impaired immune response to C. difficile toxins also contributes to an increased risk of disease recurrence.
      Figure thumbnail gr4
      FIG. 4Association between serum IgG antibody response to Clostridium difficile toxins and protection against recurrent C. difficile infection (CDI) [
      • Kyne L.
      • Warny M.
      • Qamar A.
      • Kelly CP.
      Association between antibody response to toxin A and protection against recurrent Clostridium difficile diarrhea.
      ].
      Hospitalization constitutes a major risk factor for the acquisition of CDI, as it brings together multiple risk factors, including an increasingly elderly population, exposure to antibiotics, and an environment contaminated by C. difficile spores. Prolonged hospital stay is associated with the risk of developing recurrent CDI [
      • Johnson S.
      Recurrent Clostridium difficile infection: a review of risk factors, treatments, and outcomes.
      ,
      • Pépin J
      • Routhier S.
      • Gagnon S.
      • Brazeau I.
      Management and outcomes of a first recurrence of Clostridium difficile-associated disease in Quebec, Canada.
      ]; this finding was confirmed in a recent retrospective analysis of 1678 patients with CDI that also found past hospital exposure and, especially, previous admission to a gastrointestinal ward to be significant predictors of first recurrence [
      • Eyre DW.
      • Walker AS.
      • Wyllie D
      • et al.
      Predictors of first recurrence of Clostridium difficile infection: implications for initial management.
      ]. Patients are also at increased risk of recurrent CDI if they have severe or extremely severe underlying disease, as indicated by a modified Horn index score of 3 or 4 [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]. Specific comorbidities that have been found to be associated with an increased risk of recurrent CDI include a history of chronic renal impairment [
      • Do AN.
      • Fridkin SK.
      • Yechouron A
      • et al.
      Risk factors for early recurrent Clostridium difficile-associated diarrhea.
      ], as patients are often in poor health and require prolonged hospitalization, a compromised immune system [
      • Cohen MB.
      Clostridium difficile infections: emerging epidemiology and new treatments.
      ], and inflammatory bowel disease [
      • Kelsen JR.
      • Kim J.
      • Latta D
      • et al.
      Recurrence rate of Clostridium difficile infection in hospitalized patients with inflammatory bowel disease.
      ]. Infection with the hypervirulent 027 strain of C. difficile may also, as discussed earlier, convey an increased risk of recurrence [
      • Petrella LA.
      • Sambol SP.
      • Cheknis A
      • et al.
      Decreased cure rate and increased recurrence rate for Clostridium difficile infection caused by the epidemic C. difficile BI strain.
      ], although this has not been borne out in all recent analyses of risk factors for CDI recurrence [
      • Eyre DW.
      • Walker AS.
      • Wyllie D
      • et al.
      Predictors of first recurrence of Clostridium difficile infection: implications for initial management.
      ].

      Validation of a Clinical Prediction Rule for Recurrent CDI

      Predicting which patients, or patient subsets, may experience recurrent CDI would be clinically useful. On the basis of a prospective study of 44 patients with CDI (derivation cohort), we undertook a multivariate logistic regression analysis to develop a clinical prediction rule for recurrent CDI, and then prospectively evaluated the performance of the rule in an independent cohort of hospitalized patients with CDI (validation cohort). Age >65 years, the presence of severe underlying disease (modified Horn index score of 3 or 4) and the use of additional antibiotics after discontinuation of CDI therapy emerged as factors that are independently associated with an increased risk of recurrence (Table 1) [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]. All three of these had previously been identified as risk factors for recurrent CDI, either in the systematic analysis described above or in another study [
      • Kyne L.
      • Sougioultzis S.
      • McFarland LV.
      • Kelly CP.
      Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea.
      ]. The clinical prediction rule effectively discriminated between patients with and without recurrent CDI, with 77.3% and 71.9% of patients in the derivation and validation cohorts, respectively, being correctly classified as having a single episode of CDI vs. recurrence (Table 2) [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]. Although patients who fail to mount a protective antibody-mediated response to C. difficile toxins are at increased risk of CDI recurrence [
      • Kyne L.
      • Warny M.
      • Qamar A.
      • Kelly CP.
      Association between antibody response to toxin A and protection against recurrent Clostridium difficile diarrhea.
      ], the inclusion of day 12 anti-toxin A serum IgG antibody measurements did not increase diagnostic accuracy in the derivation cohort, and so was not used in the final prediction rule [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]. Moreover, to be of clinical use, realtime measurement of antibody levels would be required, which is not feasible at the present time.
      TABLE 1.Clinical prediction rule for recurrent Clostridium difficile infection [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]
      Risk factorScore
      Age >65 years1
      Severe underlying disease (Horn index score of 3 or 4)
      The Horn index rates the severity of underlying disease into one of four categories on the basis of clinical judgement: 1 = mild (single mild illness); 2 = moderate (more severe illness but uncomplicated recovery expected); 3 = severe (major complications or multiple conditions requiring treatment); 4 = fulminant (catastrophic life-threatening illness).
      1
      Additional antibiotic use1
      Add one point for each risk factor present to determine score (range 0-3)
      a The Horn index rates the severity of underlying disease into one of four categories on the basis of clinical judgement: 1 = mild (single mild illness); 2 = moderate (more severe illness but uncomplicated recovery expected); 3 = severe (major complications or multiple conditions requiring treatment); 4 = fulminant (catastrophic life-threatening illness).
      TABLE 2.Risk of recurrent Clostridium difficile infection (CDI) in comparison with score on the clinical prediction rule [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]
      Recurrence
      Derivation cohort n = 44; 50% recurrenceValidation cohort n = 64; 20% recurrence
      Scoren%n%
      00/700/90
      15/1533.36/3616.7
      210/1471.45/1631.3
      37/887.52/366.7
      Patients with scores >2 were classified as being at high risk of recurrent CDI. The point estimate of the accuracy of a score >2 for predicting recurrence is 71.9% (95% CI 59.2-82.4%) [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ]. Other point estimates of performance include a sensitivity of 53.8%, a specificity of 76.5%, a positive predictive value of 36.8%, and a negative predictive value of 86.7% [
      • Hu M.
      • Katchar K.
      • Kyne L
      • et al.
      Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
      ].
      The clinical prediction rule described above is believed to be a useful contribution to clinical practice. However, it has some important limitations, including small sample size. In a recent systematic review of prediction tools for unfavourable outcome in CDI that included recurrent disease, Abou Chakra et al. [
      • Abou Chakra CN.
      • Pepin J.
      • Valiquette L.
      Prediction tools for unfavourable outcomes in Clostridium difficile infection: a systematic review.
      ] argue the case for the development of evidence-based tools through appropriate prospective cohorts in preference to empirically developed rules, but also stress the need for further validation of existing tools. The clinical prediction rule described here certainly merits further validation in a larger cohort of patients, as this would help to improve its precision and robustness.

      Using Clinical Prediction Rules to Guide Interventions

      The prediction rule for recurrent CDI described above appears to be simple, reliable, and accurate [
      • Garey KW.
      • Dao-Tran TK.
      • Jiang ZD
      • et al.
      A clinical risk index for Clostridium difficile infection in hospitalised patients receiving broad-spectrum antibiotics.
      ]. In both the derivation and validation cohorts, there was a stepwise increase in the percentage of patients experiencing recurrent CDI as the score increased from 0 to a maximum of 3. Thus, in clinical practice, it has potential to discriminate between patients at high risk of developing recurrent CDI from those at low risk. By heightening awareness of a risk of recurrent CDI, it can be used to alert the treating physician and the patient to the need for prompt recognition, confirmatory diagnosis, and treatment, should recurrence develop. Patients who meet the criteria for a high risk of CDI recurrence may benefit from interventions with the potential to mitigate the risk of recurrence, such as fidaxomicin [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Cornely OA.
      • Crook DW.
      • Esposito R
      • et al.
      Fidaxomicin versus vancomycin for infection with Clostridium difficile in Europe, Canada, and the USA: a double-blind, non-inferiority, randomised controlled trial.
      ]. Where clinically possible, other preventive measures might include avoidance of concomitant antibiotic therapy, judicious choice of lower-risk agents where concomitant antibiotics are needed, and the avoidance of acid anti-secretory medications such as proton-pump inhibitors.
      As we have seen, a simple and validated clinical prediction rule can be used to identify patients at risk of recurrent CDI. Validated prediction tools would be equally beneficial in identifying patients at risk of primary CDI, as well as those at risk of severe disease and adverse outcomes. Identifying patients at high risk of primary CDI is especially important at a time of rising incidence, and can be used to test novel preventive strategies and interventions. Validated prediction tools are also needed for severe or complicated CDI to inform decisions on antibiotic choice for CDI therapy.
      Similarly to the clinical prediction rule for recurrent CDI, a prospective cohort study has identified underlying disease severity (Horn index score) as an independent predictor of acquisition of nosocomial CDI [
      • Kyne L.
      • Sougioultzis S.
      • McFarland LV.
      • Kelly CP.
      Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea.
      ]. Among hospitalized patients receiving antibiotic therapy and expected to stay in hospital for ≥2 days, the probability of acquiring CDI was 27% in those with a Horn index score of 3 or 4 (severe or fulminant underlying disease), as compared with only 4% in those with scores of 1 or 2 [
      • Kyne L.
      • Sougioultzis S.
      • McFarland LV.
      • Kelly CP.
      Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea.
      ].
      Several clinical factors have been linked to CDI of increased severity and adverse clinical sequelae, with older age again emerging as an important risk factor [
      • Loo VG.
      • Poirier L.
      • Miller MA
      • et al.
      A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality.
      ,
      • Cohen SH.
      • Gerding DN.
      • Johnson S
      • et al.
      Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA).
      ]. Thus, advancing age is predictive of acquisition of primary CDI, the development of severe, complicated CDI, and increased risk of recurrence. Factors that are strongly suggestive of severe CDI include an elevated peripheral white blood cell count (>15, 000 cells/μ L), with counts above 50, 000 cells/μ Lbeing considered a warning of likely mortality [
      • Lamontagne F.
      • Labbé A-C
      • Haeck O
      • et al.
      Impact of emergency colectomy on survival of patients with fulminant Clostridium difficile colitis during an epidemic caused by a hypervirulent strain.
      ], and a rising serum creatinine level [
      • Pépin J
      • Valiquette L.
      • Alary ME
      • et al.
      Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity.
      ], which may be indicative of severe diarrhoea, dehydration, and reduced renal function [
      • Cohen SH.
      • Gerding DN.
      • Johnson S
      • et al.
      Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA).
      ].
      Although further prospective studies are needed to confirm the utility of currently developed clinical prediction rules, such tools will be increasingly important in helping to identify at-risk patients, so that care can be tailored for individual patients as additional preventive and treatment options become available.
      Recurrent CDI affects up to one-quarter of patients who respond to initial CDI therapy with metronidazole or vancomycin, and is the most common management problem with current treatments. Identifying patients at high risk of CDI recurrence would allow treating physicians to intervene more promptly and, possibly, more effectively. Several patient risk factors for recurrent CDI have been identified, and have led to the development of a simple, validated, clinical prediction tool for recurrent CDI. Similar prediction tools for identifying primary CDI and severe CDI are in development. Further validation studies and, possibly, the incorporation of additional criteria can help to strengthen these tools and improve their diagnostic accuracy. Applying such prediction tools in clinical practice will allow for the rational application of interventions that can help to lower recurrence rates and, in the longer term, eventually prevent the development of primary, severe and complicated CDI.

      Acknowledgements

      The author wishes to thank Elements Communications Ltd (Westerham, UK) for medical writing assistance, funded by Astellas Pharma Europe Ltd (Staines, UK).

      Provenance

      The development of the content and the printing of this supplement has been funded by Astellas Pharma Europe Ltd. This supplement was created in collaboration with the faculty from the Astellas-sponsored symposium at the 2012 ECCMID Congress.

      Transparency Declaration

      C. P. Kelly has received funds for speaking, consultancy, advisory board membership or travel from Astellas, Claremont BioSolutions, CSL Behring, Cubist, GSK, Merck, Optimer, Pfizer, Sanofi-Pasteur, and ViroPharma; and received research funding support from Claremont BioSolutions, CSL Behring, Cubist, Merck, Optimer, and Sanofi-Pasteur.

      References

        • Kelly CP.
        • LaMont JT.
        Clostridium difficile infection.
        Annu Rev Med. 1998; 49: 375-390
        • Kelly CP.
        • Pothoulakis C.
        • LaMont JT.
        Clostridium difficile colitis.
        N Engl J Med. 1994; 330: 257-262
        • Kelly CP.
        • LaMont JT.
        Clostridium difficile—more difficult than ever.
        N Engl J Med. 2008; 359: 1932-1940
        • Louie TJ.
        • Miller MA.
        • Mullane KM
        • et al.
        Fidaxomicin versus vancomycin for Clostridium difficile infection.
        N Engl J Med. 2011; 364: 422-431
        • Loo VG.
        • Poirier L.
        • Miller MA
        • et al.
        A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality.
        N Engl J Med. 2005; 353: 2442-2449
        • Bauer MP.
        • Kuijper EJ.
        • van Dissel JT.
        European Society of Clinical Microbiology and Infectious Diseases: treatment guidance document for Clostridium difficile infection.
        Clin Microbiol Infect. 2009; 15: 1067-1079
        • Cohen SH.
        • Gerding DN.
        • Johnson S
        • et al.
        Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA).
        Infect Control Hosp Epidemiol. 2010; 31: 431-455
        • Aslam S.
        • Hamill RJ.
        • Musher DM.
        Treatment of Clostridium difficile-associated disease: old therapies and new strategies.
        Lancet Infect Dis. 2005; 5: 549-557
        • Pépin J
        • Alary M-E
        • Valiquette L
        • et al.
        Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada.
        Clin Infect Dis. 2005; 40: 1591-1597
        • Petrella LA.
        • Sambol SP.
        • Cheknis A
        • et al.
        Decreased cure rate and increased recurrence rate for Clostridium difficile infection caused by the epidemic C. difficile BI strain.
        Clin Infect Dis. 2012; 55: 351-357
        • Cornely OA.
        • Crook DW.
        • Esposito R
        • et al.
        Fidaxomicin versus vancomycin for infection with Clostridium difficile in Europe, Canada, and the USA: a double-blind, non-inferiority, randomised controlled trial.
        Lancet Infect Dis. 2012; 12: 281-289
        • Bouza E.
        • Dryden M.
        • Mohammed R
        • et al.
        Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
        Clin Microbiol Infect. 2008; 14: S103-S104
        • Lowy I.
        • Molrine DC.
        • Leav BA
        • et al.
        Treatment with monoclonal antibodies against Clostridium difficile toxins.
        N Engl J Med. 2010; 362: 197-205
        • Kelly CP.
        • Kyne L.
        The host immune response to Clostridium difficile.
        J Med Microbiol. 2011; 60: 1070-1079
        • Rea MC.
        • O'Sullivan O.
        • Shanahan F
        • et al.
        Clostridium difficile carriage in elderly subjects and associated changes in the intestinal microbiota.
        J Clin Microbiol. 2011; 50: 867-875
        • Chang JY.
        • Antonopoulos DA.
        • Kaira A
        • et al.
        Decreased diversity of the fecal microbiome in recurrent Clostridium difficile-associated diarrhea.
        J Infect Dis. 2008; 197: 435-438
        • Tvede M.
        • Rask-Madsen J.
        Bacteriotherapy for chronic relapsing Clostridium difficile diarrhoea in six patients.
        Lancet. 1989; 1: 1156-1160
        • Rupnik M.
        • Wilcox MH.
        • Gerding DN.
        Clostridium difficile infection: new developments in epidemiology and pathogenesis.
        Nat Rev Microbiol. 2009; 7: 526-536
        • Kyne L.
        • Warny M.
        • Qamar A.
        • Kelly C.
        Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
        N Engl J Med. 2000; 342: 390-397
        • Teasley DG.
        • Gerding DN.
        • Olson MM
        • et al.
        Prospective randomised trial of metronidazole versus vancomycin for Clostridium-difficile-associated diarrhoea and colitis.
        Lancet. 1983; 2: 1043-1046
        • Zar FA.
        • Bakkanagari SR.
        • Moorthi KM
        • et al.
        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
        • Poutanen SM.
        • Simor AE.
        Clostridium difficile-associated diarrhea in adults.
        CMAJ. 2004; 171: 51-58
        • Barbut F.
        • Richard A.
        • Hamadi K
        • et al.
        Epidemiology of recurrences of Clostridium difficile-associated diarrhea: relapse versus reinfection.
        J Clin Microbiol. 2000; 36: 2386-2388
        • DuPont HL.
        The search for effective treatment of Clostridium difficile infection.
        N Engl J Med. 2011; 364: 473-474
        • Gerding DN.
        • Muto CA.
        Owens RC Jr. Treatment of Clostridium difficile infection.
        Clin Infect Dis. 2008; 46: S32-S42
        • Johnson S.
        Recurrent Clostridium difficile infection: a review of risk factors, treatments, and outcomes.
        J Infect. 2009; 58: 403-410
        • Bauer MP.
        • Notermans DW.
        • van Benthem BH
        • et al.
        Clostridium difficile infection in Europe: a hospital-based survey.
        Lancet. 2011; 377: 63-73
        • McFarland LV.
        • Elmer GW.
        • Surawicz CM.
        Breaking the cycle: treatment strategies for 163 cases of recurrent Clostridium difficile disease.
        Am J Gastroenterol. 2002; 97: 1669-1775
        • McFarland LV.
        • Surawicz CM.
        • Greenberg RN
        • et al.
        A randomized placebo-controlled trial of Saccharomyces boulardii in combination with standard antibiotics for Clostridium difficile disease.
        JAMA. 1994; 271: 1913-1918
        • Garey KW.
        • Dao-Tran TK.
        • Jiang ZD
        • et al.
        A clinical risk index for Clostridium difficile infection in hospitalised patients receiving broad-spectrum antibiotics.
        J Hosp Infect. 2008; 70: 142-147
        • Kyne L.
        • Warny M.
        • Qamar A.
        • Kelly CP.
        Association between antibody response to toxin A and protection against recurrent Clostridium difficile diarrhea.
        Lancet. 2001; 357: 189-193
        • Pépin J
        • Routhier S.
        • Gagnon S.
        • Brazeau I.
        Management and outcomes of a first recurrence of Clostridium difficile-associated disease in Quebec, Canada.
        Clin Infect Dis. 2006; 42: 758-764
        • Eyre DW.
        • Walker AS.
        • Wyllie D
        • et al.
        Predictors of first recurrence of Clostridium difficile infection: implications for initial management.
        Clin Infect Dis. 2012; 55: S77-S87
        • Hu M.
        • Katchar K.
        • Kyne L
        • et al.
        Prospective derivation and validation of a clinical prediction rule for recurrent Clostridium difficile infection.
        Gastroenterology. 2009; 136: 1206-1214
        • Do AN.
        • Fridkin SK.
        • Yechouron A
        • et al.
        Risk factors for early recurrent Clostridium difficile-associated diarrhea.
        Clin Infect Dis. 1998; 26: 954-959
        • Cohen MB.
        Clostridium difficile infections: emerging epidemiology and new treatments.
        J Pediatr Gastroenterol Nutr. 2009; 48: S63-S65
        • Kelsen JR.
        • Kim J.
        • Latta D
        • et al.
        Recurrence rate of Clostridium difficile infection in hospitalized patients with inflammatory bowel disease.
        Inflamm Bowel Dis. 2011; 17: 50-55
        • Kyne L.
        • Sougioultzis S.
        • McFarland LV.
        • Kelly CP.
        Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea.
        Infect Control Hosp Epidemiol. 2002; 23: 653-659
        • Abou Chakra CN.
        • Pepin J.
        • Valiquette L.
        Prediction tools for unfavourable outcomes in Clostridium difficile infection: a systematic review.
        PLoS ONE. 2012; 7: e30258
        • Lamontagne F.
        • Labbé A-C
        • Haeck O
        • 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
        • Pépin J
        • Valiquette L.
        • Alary ME
        • et al.
        Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity.
        CMAJ. 2004; 171: 466-472