Clostridium difficile in clinical practice : Increasing rates , more virulent organisms and new therapies on the horizon

(CDAD) in Canada have called attention to this important adverse outcome of antibiotic prescribing. Yet, beyond diagnosis by testing for C difficile toxins and prescribing of metronidazole or vancomycin as treatment, health care providers are aware that there are few other proven treatment options for nonresponders or patients with recurrent illness. The present paper provides a summary of the current issues.

Clostridium difficile-associated diarrhea has recently come under increasing scrutiny because of outbreaks of disease in eastern Canada in association with a hypervirulent ribotype 027 (NAP1) strain.Hospitals in the United States, the United Kingdom and Europe have been affected as well.Epidemiological studies are underway to determine the presence of the more virulent strain of C difficile in hospitals across Canada.Because this strain is highly quinolone resistant, it is suspected that overuse of that class of antimicrobial agent may be a selection factor for this strain.Although C difficile-associated diarrhea is usually considered a nosocomial infection, approximately one-quarter of cases occur in the community.Recurrence following successful resolution of diarrhea occurs in one in five patients treated with metronidazole or vancomycin; relapse is more common in patients needing retreatment.New therapies that may have equivalent or higher response rates and lower recurrence rates are being investigated.It is hoped that these new treatments will be a substantial improvement over current therapies.In the interim, prudent antimicrobial use and attention to infection control practices are the main preventive strategies.Prompt testing for C difficile toxin in patients with diarrhea after antibiotic exposure, discontinuation of unnecessary antibiotic therapy and rapid initiation of treatment should minimize complications.With the appearance of more toxigenic strains of C difficile, careful monitoring of patients to detect suboptimal clinical response is necessary.
absence of defined outbreaks.The Centers for Disease Control and Prevention estimate that there are 178,000 cases per year in the United States, affecting one in 1700 persons per year -a case rate double that of a decade ago (3).The United Kingdom has tracked the annual incidence of CDAD for more than a decade and has documented a five-to sevenfold increase in case rates since the early 1990s.In 2004, 44,000 new cases were reported to the Health Protection Agency in the United Kingdom, double the number in 2001 (4).These rates appear to precede the emergence of the ribotype 027 organism as an outbreak strain.
Outbreaks of disease have resulted in case rates that are fiveto 10-fold higher.Most outbreaks investigated in the past decade have been proven to be clonal (ie, spread of a particular strain) (5).In Calgary in 2000 to 2001, during an outbreak involving a clindamycin-resistant clone (selected for by excessive clindamycin prescribing) that affected all three adult general hospitals, the peak case rate was sevenfold higher than before the outbreak.At that time, 77% of isolates were of one clone.Restriction of clindamycin use rapidly controlled the outbreak, and over the past four years, the prevalence of the outbreak strain has been markedly reduced (6).In 2002 to 2004, a multihospital outbreak in Quebec involving a single clone has been associated with case rates of 14 to 35 cases per 10,000 patientdays (7).A hypertoxigenic ribotype 027, toxinotype III, North American pulso-type clone (NAP1), which is closely related to strains previously known to be present in Europe, has accounted for the majority of strains in affected Quebec hospitals (8).This clone has an abnormal toxin repressor gene that causes it to make 16 to 20 times as much C difficile toxin A and B in vitro as do 'normal' strains, and produces toxin immediately in the log phase growth curve rather than later in the stationary phase, as has been the case in 'normal' strains (9).Infection has been associated with more fulminant disease, resulting in death rates more than three times higher than expected for matched patient control subjects in the elderly population (10).Concurrently, a number of American hospitals have been investigated for outbreaks involving the same or a similar clone (11).The outbreak strain is highly quinolone-resistant, and it is suspected that overuse of quinolones is a driving force behind these outbreaks.Ongoing investigations of outbreaks in Europe are supportive of the notion that quinolones as a class are a factor in selecting these outbreak strains (12).
Is the Quebec outbreak strain or similar strains present in other Canadian communities?A recently completed survey (13) of nosocomial CDAD cases involving 34 hospitals across Canada from November 2004 to May 2005, and which included culture confirmation of the pathogen, showed that this strain is present in seven provinces.This strain has been present in Calgary, without a defined outbreak, since 2001, and accounts for 16% of community strains and 2% to 5% of hospital strains (14).It appears that the presence of the strain does not mean that an outbreak is imminent, and that other factors or processes have allowed for the development of an outbreak.
Physicians need to be aware that outbreaks of infection are on the rise and the causes may be multifactorial.In addition to strain virulence, crowded hospitals with high occupancy, increased or continued antibiotic selection pressure, transmission-prone hospital environs, inadequate infection control conditions or practices, an increasing aged population with increased susceptibility to fulminant infection and possible cofactors such as widespread proton pump inhibitor use are suspected reasons for the increasing rates of infection (15)(16)(17).
Therefore, physicians need to prescribe antimicrobial agents sparingly and to maintain infection control practices (for all patients, not only for identified cases) because these two factors repeatedly emerge as successful control measures against CDAD.Improved infection control resources have halved the case rates of CDAD in Quebec.Additional measures are required in the face of rising case rates.

DIAGNOSIS OF CDAD
Detection of C difficile toxins is the main diagnostic end point.Patients who have diarrhea in association with prior or concurrent antibiotic exposure should be routinely screened for the presence of C difficile toxin in fecal samples.Generally, approximately 8% to 10% of submitted stool samples are found to be positive.Testing should be available on a daily basis with as short a turnaround time as is practicable.Enzyme immunoassays (EIAs) are the most commonly used tests in the clinical laboratory due to their low cost, ease of use, less than one-day turnaround times and high specificity.Because controlled studies have shown that EIA tests are 55% to 89% sensitive in detecting toxin (18)(19)(20), repeat tests should be ordered in highly suspected cases.The tissue culture toxin assay, long used as the reference standard, detects lower concentrations of toxin B, but is less commonly used now because of its higher cost and slower turnaround time.Endoscopy is occasionally required to confirm a diagnosis of presumed colitis in the face of negative test results (this is estimated to occur in about 2% of instances, so it is not rare).Culture for C difficile alone is not adequate for diagnosis because one-quarter of strains are nontoxigenic (and therefore not considered pathogenic).In addition to cytotoxicity testing, culture to recover C difficile with confirmation of toxigenicity has been found to increase the number of cases found by 15% to 77% (21,22), but this is not commonly done.In the event that test results are not to be available for 48 h to 72 h, it is prudent to consider empirical therapy of florid cases (ie, known prevalence of CDAD on the ward; antibiotic exposure; older patients with comorbid conditions; or presence of diarrhea, fever or leukocytosis) that are clinically compatible with CDAD (23).
Once a case is diagnosed, it is not necessary to retest with additional EIA tests to guide further management.A common error on hospital wards is to order repeat EIA tests to determine when to discontinue isolation.Patients should remain in contact isolation with separate toilet facilities until diarrhea has resolved and the patient is proficient at maintaining bowel continence for approximately 48 h.Despite a negative toxin test, stools of patients with CDAD continue to have infectious spores during and following therapy.Alternatively, firm stools may still yield a positive toxin test.
In patients who have recurrence of diarrhea following a successful response to initial therapy, the EIA toxin test may be negative initially, despite a clinical diagnosis of relapsed CDAD.The toxin test should be repeated and, in the absence of an alternative cause, treatment may be considered.

IS IT NECESSARY TO TREAT ALL PATIENTS?
With increased virulence and with the inability to determine which patients will abruptly deteriorate, there is now a movement to consider treating all patients.However, this may vary by hospital, region and clinical circumstance.If the patient is on the inducing antibiotic and it is not necessary to continue treatment, stopping antibiotic therapy is the first priority.
Observational studies have shown that stopping the antibiotic leads to cessation of diarrhea in approximately 25% of patients within approximately four days.The Cochrane review of CDAD (24) also concludes that there is likely a selfresolution rate that needs to be considered in therapeutic trials.A common office practice scenario is the screening of patients who develop diarrhea following or during antibiotic therapy.The antibiotic is discontinued and a request for toxin testing is made.Depending on testing parameters, often 48 h to 72 h elapses before results are returned to the office.In these milder cases, an inquiry should be made regarding persistence of diarrhea.If patients have fewer than three bowel movements per day, and they are otherwise well, it is acceptable to observe them and provide follow-up care.This is because current therapies are likely to suppress the residual normal gut flora and may predispose patients to relapse.In the future, when therapies with a lesser likelihood of relapse become available, it may become more convenient to treat all patients.

THERAPY
Metronidazole 500 mg three times daily or 250 mg four times daily by mouth for 10 to 14 days remains the standard treatment for CDAD (Figure 1).During the outbreak in Quebec, vancomycin was preferentially used in some of the affected hospitals.Metronidazole works by being absorbed in the upper gastrointestinal tract, and is leached through the inflamed bowel mucosa into the intestinal lumen.Maximal concentrations of metronidazole are in the order of 7 µg/mL in fecal filtrates, and minimum inhibitory concentrations are usually 1 µg/mL or lower.Resistance to metronidazole is rarely observed.Nausea, abdominal cramps, altered taste and headache may occur in up to one-half of patients receiving metronidazole; however, only approximately 10% of patients require discontinuation of the drug.A higher than expected rate of relapse after metronidazole therapy were observed by physicians in Quebec during the recent outbreak (25).
Vancomycin 125 mg four times daily for 10 days is also a standard regimen.Fewer than 10% of patients require a higher dose to attain a response.It was previously shown that patients randomly assigned to 125 mg or 500 mg four times daily did not have major differences in diarrhea resolution (26).Therefore, a higher dose of 250 mg to 500 mg four times daily should not be routinely used as the initial therapy.The failure to respond to the 125 mg dose is unexplained because vancomycin concentrations in fecal filtrates with the standard dose are several hundredfold higher than the minimum inhibitory concentration of the pathogen (1 µg/mL to 2 µg/mL).
At present, cost differences and the higher likelihood of selection for vancomycin-resistant enterococci favour the use of metronidazole in routine clinical practice.Metronidazole should remain the first-line choice, with vancomycin as the alternative.The response to metronidazole or vancomycin is achieved in 4±2 days in most patients, with the clinical impression that vancomycin works somewhat faster.It may take seven to nine days before diarrheal symptoms resolve in some patients.In the absence of worsening diarrhea, increased abdominal pain, fever and leukocytosis, these patients may be closely followed.Vancomycin 125 mg capsules four times daily for 10 days costs approximately $300.An increasingly common practice is the use of the intravenous form, given orally both in hospital practice and in an outpatient setting, provided that the pharmacy is willing to reconstitute the vancomycin.Less than 0.5% of vancomycin is absorbed.Vancomycin damages the normal bowel flora (27), and it is presumed that metronidazole may have a similar effect, possibly accounting for relapses.
Prompt and effective therapy is essential to avoid disease progression.Should the patient develop nausea, vomiting, abdominal pain and distention, ileus or pancolitis with clinical toxicity, therapeutic approaches include intensive care unit should not be shared between patients.‡ The criteria for discontinuing isolation are as follows: patient is symptom free for 48 h and has had at least one normal or formed stool, and if the patient has underlying chronic diarrhea (for other reasons) or is given a bowel routine or treatment that causes loose frequent stools, it may be necessary to repeat toxin testing to decide whether the CDAD has resolved.§ While recent studies have shown that alcohol-based hand rubs have not been associated with increased rates of CDAD, handwashing with soap (or 2% chlorhexidine gluconate) and water is preferred.Hand hygiene before and after all patient contact as part of standard practice should be the 'culture' of physicians and other health care providers.For each proven case of nosocomial infection, there are several others that are undetected.¶ The best treatments have not been well investigated in clinical trials.The combination of oral or nasogastric tube instillation of vancomycin 500 mg every 6 h, intravenous (IV) metronidazole 500 mg every 6 h, rectal instillation of 500 mg of vancomycin every 6 h in 100 mL of normal saline and IV gamma globulin (IVIG) 0.3 g/kg to 0.5 g/kg (in either one or two doses) has been used in small studies or clinical experiences by treating physicians (61).Early consultation with surgery to determine if a colectomy is needed is recommended.Postsurgical intensive care unit (ICU) support is usually required.QID Four times daily; TID Three times daily admission, intravenous administration of metronidazole, intravenous immunoglobulin 0.3 g/kg to 0.5 g/kg (28)(29)(30), rectal instillation of vancomycin, and colectomy for threatened perforation, pancolitis and sepsis.Mortality in these circumstances is in the order of 30% to 40%.Additional alternative therapies, all with less clinical experience, include bacitracin, fucidin, rifampin, teicoplanin and rifaximin (last two not available in Canada) (31)(32)(33)(34).

UPCOMING NEW THERAPIES
It is expected that new therapies will be forthcoming in the next two to four years with the aim of achieving greater than 90% response rates with a low rate of relapse.There are several novel therapies being examined.Tiacumicin B, OPT-80/PAR-101 Tiacumicin B, OPT-80/PAR-101 (Optimer Pharmaceuticals, USA, and Par Pharmaceuticals, USA) is a nonabsorbed, macrocyclic antibiotic that has a partial lactone ring structure but is not a macrolide, and is eight-to 10-fold more active than vancomycin against C difficile.It possesses modest activity against enterococci but otherwise is nearly inactive against the normal flora.It was previously shown to durably cure the hamster model of C difficile diarrheal disease (no relapse or death after treatment stops) (35).A phase 2A study of 45 patients has just been completed showing a high response rate followed by a low relapse rate (36), with relative sparing of components of the normal gut flora (37).The results were sufficiently promising such that a phase 2B/3A study of 500 patients, comparing 200 mg twice daily for 10 days and vancomycin 125 mg four times daily for 10 days, is now underway.

Tolevamer
Tolevamer (Genzyme Corporation, USA) is a soluble binder of C difficile toxins A and B. The objective of toxin binding is to neutralize toxins A and B, and to cure or prevent disease by allowing the normal flora to recover over time.Toxin-binding therapy was shown to durably cure hamsters with CDAD in 2001 (38).A phase 2 study in 2002 to 2003 studied 287 patients in three groups.The 6 g per day of toxin-binding therapy resulted in a 79% response rate (noninferior to vancomycin) followed by a 7% recurrence rate, whereas a 91% response rate to vancomycin was followed by a 20% recurrence rate (39).Two large, phase 3 trials are underway to compare a 9 g daily dose of toxinbinding therapy to metronidazole or vancomycin.

Other candidate therapies
Other candidate therapies include ramoplanin, a lipoglycodepsipeptide (Oscient Pharmaceuticals, USA); rifaximin (Salix Pharmaceuticals, USA), a poorly absorbed rifamycin (40,41); and nitazoxanide, an anticryptosporidia agent (Alinia, Romark Pharmaceuticals, USA).Immunobiological approaches that are in the early stages of development include vaccines (42), ingestion of nontoxigenic C difficile (43,44) and ingestion of antibodies (45) that would bind elements of the toxin or adherence factors.Definitive clinical trials are required to determine the usefulness of these new treatments.

CDAD-INDUCING ANTIBIOTICS
Any antimicrobial that can suppress the normal microbial floral balance, particularly intestinal anaerobes, can induce CDAD.Exact risks are not available because the number of persons colonized is usually not determined, infectious inocula are variable, and because of varying determinants of disease, such as time of acquisition, host humoral antibody response (46), age, underlying disease and ecological effects of antibiotics.Nevertheless, it is generally appreciated that aminoglycosides, co-trimoxazole and intravenously administered vancomycin rarely induce CDAD.The main inducers are clindamycin, cephalosporins and amoxicillin or other beta-lactams.Quinolones, tetracyclines and probably macrolides can also be inducers.In the Quebec outbreak, the dominant pathogen was highly quinolone resistant, and it is suspected that the high use of quinolones in hospital and in community medicine was a factor.While there is a class effect by quinolones in selecting for C difficile disease, the risk may be higher with the 8-methoxy agents, gatifloxacin and moxifloxacin (12,47).At present, antibiotic control in institutions should still be focused on restricting clindamycin and cephalosporins.
Narrow-spectrum agents are preferred in office practice (ie, penicillin V potassium is preferred over cephalosporins for the treatment of streptococcal pharyngitis; trimethoprimsulfamethoxazole [co-trimoxazole] still covers approximately 90% of outpatient strains of Escherichia coli for treatment of urinary tract infection and could reduce the overuse of quinolones; and cloxacillin remains a first-choice oral therapy of staphylococcal infections [methicillin-sensitive Staphylococcus aureus]).Pivampicillin, a highly absorbed (greater than 95%) alternative to amoxicillin or ampicillin, is associated with less diarrhea and theoretically should be less likely to induce disease.It is not possible to completely avoid the induction of CDAD, but it is possible to use antibiotics so that the induction of CDAD is minimized.Limiting the duration of exposure (ie, shorter courses of therapy) to all agents is desirable.

INFECTION CONTROL ISSUES
Due to the high frequency of use of antimicrobials in the hospital population combined with a suboptimal physical plant and difficult practice conditions, CDAD is a universal infection control problem.Up to 30% of patients have been observed in hospital surveys to acquire C difficile in their stool over time.The spore persists in the environment for months, and endemicity in hospitals is perpetuated by the vicious cycle of diarrheal illness and environmental contamination.For this reason, the medical staff's prescribing habits have a major role in maintaining or, alternatively, in stopping this cycle.We have found that most patients have 10 3 to 10 4 spores/g of feces during and following treatment.Ideally, all patients on medical and surgical floors, where CDAD risk is highest, should have a bed-to-toilet ratio of 1:1 to limit transmission of nosocomial pathogens.However, the reality is that the majority of acute care beds in Canada have shared toilet facilities.Therefore, prudent antimicrobial use, patient personal hygiene, avoidance of shared toilet facilities with ill patients and health care worker adherence to the principles of standard practice (hand hygiene, and appropriate barrier precautions before and after all patient contacts) are cornerstones for the prevention of CDAD.Patients with diarrheal stools and incontinence are to be isolated and maintained on contact precautions.Handwashing with soap and water (because alcohol gel does not kill the spores) is recommended.However, widespread application of alcohol hand rubs has not been observed to increase CDAD rates (48).On wards with high rates of endemicity, hand hygiene by all patients should be taught and enforced.The most difficult clinical issue in the management of CDAD is what to do about the chronic relapsing patient.Generally, most patients respond to a repeat course of therapy with metronidazole or vancomycin, although subsequent recurrence rates remain high (20% to 40%).Quantitative cultures of the normal flora during CDAD show severe depletion of major genera of the normal flora before the initiation of therapy.There is no proof that probiotics, given as one, two or three organisms as a means of replenishing the normal microbiota, are efficacious in preventing relapse (49).
There are also no data to support that probiotics prevented CDAD in the Quebec outbreak (M Miller and K Weiss, personal communication).However, probiotics have been found to be of some benefit in antibiotic-induced diarrhea.Saccharomyces boulardii has been shown to be modestly helpful in reducing the likelihood of relapse (50,51).Tapering the dosage of vancomycin is commonly practised, although there are no controlled studies to examine the strategy; an observational substudy conducted during a probiotic study (51) supported tapering (Table 1).Fecal enemas to introduce the microflora of donors has been practised since 1958 (52,53) and is highly effective but aesthetically displeasing.In Calgary, from 1996 to 2006, 26 patients with chronic relapsing disease for longer than six months were treated with fecal enemas, with a greater than 95% response rate (authors' unpublished observation).One investigator in Duluth, Minnesota, USA, used a nasogastric tube to deliver intestinal microbes (54); however, this approach runs the risk of colonizing the small bowel with coliforms.To simplify microfloral replacement, Tvede and Rask-Madsen (55) used a 10-microflora replacement cocktail for relapsing CDAD, given rectally.It is hoped that new therapies will reduce the incidence of recurrent disease.There is an urgent need to study the immunology and microbiology of relapsing CDAD patients and to investigate new therapies for this patient subset.

SUMMARY
CDAD is an antibiotic-induced infection that is common both in and out of hospital.It is hoped that more selective therapies will lead to less relapsing disease and that the new therapies will allow more therapeutic options.Combinations of new modalities should also be considered.Current control strategies include prudent antibiotic use both in and out of hospital, prompt screening of patients with antibiotic-induced diarrhea, attention to infection control practices, physical plant improvement when the opportunity arises and expeditious therapy.Periodic surveillance of strains should be considered on a local and nationwide basis to help physicians choose less inductive therapies for the treatment of infectious diseases.Treatment of recurrent Clostridium difficile-associated diarrhea

CONFLICT OF INTEREST
• Re-treat for 10 days with previous effective regimen for symptom control.
• Start pulsed doses of vancomycin over three weeks (eg, 125 mg to 500 mg every three days for a total of seven doses) (56), or start tapering doses of vancomycin over three to four weeks (eg, 125 mg three times daily to two times daily to once daily) (56), stepping down by one capsule per week.A lower dose is possible with the intravenous solution given orally (eg, 62.5 mg as a 1.25 mL solution once daily in the fourth week).
• No clear benefit of using probiotics has been shown, although probiotics have no observed adverse outcomes (49).
• Cholestyramine has not been shown to be effective (38).
• A modest benefit has been shown for Saccharomyces boulardii (51).
• Rifaximin or rifampin may be tried; there have been no randomized controlled trials for this indication.Rifaximin is available in many countries but not in Canada.Rifampin plus vancomycin has been used in one small set of patients for therapy (58); with the knowledge that vancomycin suppresses the normal bowel flora, including Bacteroides species, use of rifampin alone needs to be investigated.
• Await new therapies with neutralization of toxins A and B or treatment with new, potentially more selective therapies.
Most cases of relapse are with the same strain, but in an environment where different C difficile spores are present, infection by a different strain is possible; generally, there is no difference in approach Clostridium difficile in clinical practiceCan J Infect Dis Med Microbiol Vol 17 Suppl B May/June 2006 21B Figure1) Management of adult patients with Clostridium difficileassociated diarrhea (CDAD).*Do not repeat stool tests once a patient has a positive C difficile stool toxin test unless the patient was successfully treated and subsequently has recurrent diarrhea (repeat stool testing is used to confirm that the relapsing diarrhea is due to C difficile [see § ]).Medical management is determined with the first positive test, and treatment is not altered with repeat positive tests.Discontinuation of isolation is based on cessation of diarrhea.A negative toxin is not required to discontinue precautions.† A private bathroom is preferred, if possible.After patient is transferred to a private room, the old room and old bathroom should undergo terminal cleaning with bleach or activated hydrogen peroxide agent.Commodes are potential vehicles for transmission and

:
TJL is the medical director of the Infection Prevention and Control Program, Calgary, Alberta, and is an investigator for Genzyme Corporation and Optimer Pharmaceuticals.
Clostridium difficile in clinical practiceCan J Infect Dis Med Microbiol Vol 17 Suppl B May/June 2006 23B