Clinical spectrum of Clostridium difficile Infection (CDI) and the emergence of hypervirulent strains Ed J. Kuijper National Reference Laboratory for Clostridium difficile Leiden University Medical Center
S. Johnson, D. Gerding: Clin Infect Dis 1998:26;1027-36
Clostridum difficile infection (CDI) Asymptomatic carriership Hospital-acquired diarrhoea mild self-limiting diarrhea cholera-like pseudomembranous colitis Community-acquired diarrhoea Complications toxic megacolon (mortality 65%) colonic perforation protein-losing enteropathy relapses (20%) reactive arthritis
Risk factor for C difficilediarrhoea Antibiotic administration; overestimated, at the LUMC 16 (28%) of 57 patients diagnosed with CDAD in a two-year period, did not use any antibiotic for treatment or prophylaxis. Elderly age Duration of hospital (ICU) stay Renal failure Chemotherapy Cancer (solid tumours, leukemia or lyphoma) Gastrointestinal disease or intestinal surgery Antacids, proton pump inhibitors Diabetes mellitus COPD Liver cirrhosis
Augustus 2004 Promed; In Canada during a period of 18 months, 189 patients died due to consequences of severe CDAD. Hospitals, nursing homes, community acquired (25%) USA: outbreaks in 7 hospitals in 6 states 2001-2004; > 50% toxinotype III with 18 bp deletion in TcdC and fluoroquinolone resistant C. difficile NAP1/toxinotype III, REA BI, PCR Ribotype 027
New Emerging Clostridium difficile Clinical Mortality within 30 days: 4.7% to 13.8% Complications: 7.1% to 18.2% Relapse rate: 20.8% to 47.2% less response to metronidazole Characteristics of the strain tcda and tcdb tcdc tcdc ermb Binary toxin Positive 18 bp deletion Deletion at 117, frameshift Negative Positive Genotyping of the strain B4 2 B1 R202 91 19 17 M246 C37 L1 3 2 V6-81 C44 V6-35 V6-80 5 17 0 M216 B6 2 M278 9 PCR-ribotyping 027 REA group BI PFGE NAP1 Toxinotyping III Further subtyping REA and MLVA 2 7 - V6-44 V6-142 2 1ML 1 1 C1 2191cc C4 C8 3ML 0 3 13 ML 4108 I
Complications: toxic megacolon, perforation, colectomy, septic shock.
Response to metronidazole Pepin et al, CMAJ 2005:40: 1591-7. Response to metronidazole decreased to 25% in 2003-2004 compared to 9% in the periode 1991-2002. Musher, CID 2005: 40:1586-90. 22% persistent symtoms after 10 days of treatment. Al-Nassir et al CID 2008:47:56-62. 9 months prospective observation study, vancomycin and/or metronidazole. At day 5, vancomycine more effective in eradication of C. difficile and resolution of diarrhoea. No difference occurred at the end of treatment of 3 months. Recommendation: Assess the response daily and switch when symptoms have not resolved at day 5
Response to metronidazole Levels in feces to 9 ugr/gram wet weight in semisolid or watery samples: lower dosage than 500 mg every 8 hours is unlogical hydroxymetabolite similar concentrations inactivation of metronidazole by enterococci modest increase in MIC values might result in ineffectiveness: UK strains of Type 001 collected in 2005-2006 had increased MIC values to 3.5 mg/l (agardilution and spiral gradient endpoint testing
Severity of CDI General: age, peak leukocytosis and serum creatine level Zar et al. CID 2007:45;302-7. Algorithm for severe diarrhoea and treatment withn vancomycin. Score, based on age, temperature, serum albumin level, periphertla WBC, endoscopic evidence of PMC, treatment at ICU. No prospective study with severity markers including a weight for patient comorbidity Recommendation: WBC> 15,000 cells/ul, acutely rising serum creatinine level (e.g 50% above baseline), elevated serum lactate concentration, temperature >38.5 C, evidence of severe colitis (abdominal sign, radiology)
Clinical definition of severe CDI admission with CDI at the hospital transfer to ICU (eg shock) surgery for CDI (toxic megacolon, perforation, refractory colitis) death within 30 days with CDI primary or contributive factor
Outbreaks due to Type 027 Sporadic cases due to Type 027 Kuijper E, et al. Euro Surveill. 2008;13: July 31th
ESCMID and ECDC Background information Guidance documents for infection control, diagnosis and treatment Pan-european surveillance study; Baseline incidence of hospital-acquired and community-acquired C. difficile infections To build a network of laboratories with links to national surveillance institutes Standardization of typing techniques and distribution of reference strains
Admission Discharge 48h 4 weeks Healthcare-onset (*) 8 weeks time Community-onset Healthcare-associated Unknown Community-associated (*) : - may be community- or healthcare-associated, depending on case s history. - if healthcare-associated, may have been acquired in the same facility or imported from another.
Why now? First isolate of 027 in 1988 (France, patient with severe CDAD) Historical database (>6000 strains) of Dale Gerding: 1984-1993, 14 patients with FQ susceptible isolates of 027 Historical isolates and new isolates shared 18bp deletion and binary toxin genes No gradually increase: new epidemic strain since 2002 outbreaks in Canada, USA and Europe No replacement of existing types
Rapid spread of type 027 Antibiotic misusage: role of fluoroquinolones and resistance of 027 Age and underlying diseases of the patients Change of attack-rate Change of cleaning of hospital environment Handwashing with ethanol Lack of facilities for isolation and cohort isolation Lack of clinical recognition and late diagnostics Combatting CDAD has insufficient priority in hospital management Transfer of patients with non-diagnosed CDAD between hospitals and countries Alteration of the strain of sporulation, toxin production
The hospital St. Jansdal hospital in Harderwijk General hospital, 341 beds Number of patients admitted in 2004: 25,625 The laboratory Before march 2005 testing for CDAD referred to Meander Medical Centre, Amersfoort (cytotoxicity test) March: local introduction of a new diagnostic test in St Jansdal: Rapid enzyme-immuno-assay (ICTAB, Meridian)
Ribotypering: 027
number of patients with CDAD 18 16 14 12 10 8 6 2003 20 04 v no se p l ri ap ja n ju ne n ja ja n 0 ju ne 2 2005 Periode April until September 1th, 2005: n=45 patients (58.4/10,000) 9 (20%) died of which 3 (7%) directly to complications of CDAD 10 (22%) patients suffered from relapses
Symptoms and signs < 2 weeks Diarrhoea (100%) Fever (53.3%) Abdominal pain (20%) Bloody stools only in 3 patients (6.7%)
Laboratory parameters < 2 weeks High white blood cell count: 10-20*109 cells/ml : 47% > 20*109 cells/ml : 23.7% High ESR: 10 100 mm/hr : 80% >100 mm/hr : 14% High serum creatinin level: 45% 100 200 micromol/l : > 200 micromol/l : 17.5% Low serum albumin level: < 20 g/l : 7.7%
Outcome: Recurrence R/ vancomycine and /or metronidazole Recurrence of diarrhoea: 10 patients (22%) Recurrence more often seen in patients with: Peak white blood cell count > 20*109 cells/ml (p = 0.002; OR = 16, 95% CI 2.8 90.4) Peak serum creatinin level > 200 micromol/l (p = 0.03; OR = 7.1, 95% CI 1.3 40.2)
Outcome: Mortality Mortality within 30 days after diagnosis: 9 patients (20%) 3 (7%) as a direct result of CDAD Predictor of mortality: Peak white blood cell count > 20*109 cells/ml within the first 2 weeks following onset of diarrhoea (p = 0.01; OR = 7.8, 95% CI 1.5 39.1).
Epidemic ended at St. Jansdal Early and rapid diagnostics, introduction of 3-days rule Strict hand hygiene with water and soap Gloves and apron Cohorting of patients with CDAD Effective environmental cleaning with chlorine containing desinfectants Complete banning of fluoroquinolones and restriction of cephalosporins
Course of the epidemic and dynamics of antibiotic use in the St.Jansdal hospital Start infectioncontrol measuremen ts Reintroduction of FQ use Ban of all fluoroquinolones (FQ) 10 9 DDD/100 bed-days per month 8 7 6 75 5 4 3 2 1 0 0 jan feb march april 2005 cefuroxim iv may june july aug sept months ciprofloxacin po + iv oct nov dec jan* feb 2006 incidence CDAD CDAD incidence per 10 000 admissions per month 150
Case-control study: 3 different study groups Group I: 45 patients with CDAD Risk factors for CDAD Risk factors for diarrhea due to other cause Group II: 90 randomly selected controls without diarrhoea Group III: 109 patients with noncdad diarrhoea Difference between CDAD and non-cdad diarrhea
Crude and adjusted odds ratios for development of diarrhoea, according to demographic, clinical, and pharmaceutical characteristics Clostridium difficile associated diarrhoea Diarrhoea due to other causes Crude odds ratio (95% CI) Adjusted odds ratio (95% CI) * Crude odds ratio (95% CI) Adjusted odds ratio (95% CI) * Age, years 18-64 65 1 (reference) 2.6 (1.0 5.7)1 1 2.6 (1.0 5.7)1 1 0.8 (0.5 1.5) 1 0.8 (0.5 1.5) Antibiotics: Any antibiotic Cephalosporines Macrolides Quinolones 15.3 (4.4 53.2)3 7.0 (3.1 15.7)3 4.9 (2.0 12.3)3 11.6 (3.1 43.6)3 12.5 (3.2 1.1 3 48.1) 0.5 5.7 (1.8 0.3 3, a 18.6) 2.0 2.4 (0.7 8.7) b 15.3 (2.7 1 p< 0.05 ; 2 p< 0.01 ; 3 p < 0.001 84.6)2, c * = adjusted for differences in age, duration of hospital stay, co-morbidity (ICD-10), level of care and co-medication. a = additional adjustment for concommitant use of macrolides and quinolons b = additional adjustment for concommitant use of cephalosporines and quinolones c = additional adjustment for concommitant use of cephalosporines and macrolides (0.6 (0.3 (0.1 (0.5 2.0) 1.0) 1.2) 7.8) 1.8 (0.9 0.9 (0.3 0.3 (0.1 2.2 (0.5 10.6) 3.7) 2.1) 1.6)
Risk for developing CDAD (OR) 70 57,5 60 50 40 28,8 30 20 10 0 7,8 1 1 2 1 = no CE or FQ use 2 = CE therapy 3 4 3 = FQ therapy only 4 = combination therapy
Population attributable risk percent (PAR%) Cephalosporin therapy: 31% of all antibiotics prescribed PAR% = 56% Fluoroquinolone therapy: 9.5% of all antibiotics prescribed PAR% = 33%
Fluoroquinolones Ciprofloxacin lacks in vitro potency against many important anaerobic bacteria Levofloxacin, trovafloxacin, sparfloxacin: good acivity against anaerobic bacteria Methoxyfluoroquinolones (gatifloxacin and moxifloxacin, similar to that of trovafloxacin against a broad spectrum of anaerobic bacteria) Clinical isolates of C. difficile with decreased susceptibility to fluoroquinolones exhibited mutations in either gyra or gyrb Until 2000 no relation of CDAD with ciprofloxacin and oflaxacin
Year hospital Quinolon e Design Multivariate ICHE 2001:22;5 62-575 1998 300-bed USA tertiary care hospital ciprofloxac in Retrospective case-control study (27/54) Ciprofloxacin OR=9.5 Cephalosporine s OR=6.7 EID 2003:9;73 0-33 2001 778-bed, VAMHCS, Baltimore Levofloxaci n ciprofoxaci n gatifloxaci n Retrospective Case-control study (30/60) CID 2004:38;6 40-645 2002 173-bed aute care USA hospital Switch levofloxaci n to gatifloxaci n Retrospective case-control study (37/59) Fluoroquinolon es OR=12.7 Cephalosporine s OR=0.4 Clindamycin OR=2.2 Clindamycin OR=7.7 Duration of gatifloxacin OR=11 ICHE 2005:26;2 73-280 2002001 Pittsburg h, USA levofloxaci n Retrospective case-control (203/203) Levofloxacin OR=2.0 Ceftriaxon OR=5.4 Clindamycin OR=4.8 31% 6.7% 10% CID 2005: 41;125460 2003 2004 Teaching hospital, Sherbroo k, Canada ciprofloxac in levofloxaci n gatifloxaci n moxifloxaci n Switch levofloxaci n to moxifloxaci n Retrospective cohort (293/5619) Fluoroquinolon es OR=3.44 Cephalosporine s OR=1.56-1.99 Clindamycin OR=1.77 Moxifloxacin OR=3.14 Cephalosporine s OR =ns Clindamycin OR=ns 35.9% 10% 1.5% Submitted 2003 320, acute care, nonteaching USA hospital Matched case control study (50/100) Etiologi c fraction s
CID price best article 2005
Up to December 2007 Until november 2006: 23 HCF Until december 2007: 34 HCF Outbreaks in 14 HCF Sporadic cases in 20 HCF
New Emerging Clostridium difficile Clinical - II Attributable mortality within 30 days: 3.8% Complications: 9.6% Relapse rate: 15.8% Severe diarrhoea as 027, but affects younger patients Characteristics of the strain tcda and tcdb tcdc tcdc ermb Binary toxin Positive 39 bp deletion mutation at 184, stopcodon Negative Positive The Netherlands, Northern Ireland, Genotyping of the strain (Scotland, Belgium) PCR-ribotyping 078 Toxinotyping V Further subtyping MLVA?
Type 027 outbreak Type 078 outbreak Endemic Type 027 Endemic Type 078 A B
New aspects of Clostridum difficile CDI in healthy children Klein et al. (CID 2006; 43;807-13) found an unexpectedly high rate of 6.7% CDI in children with diarrhoea who presented to a pediatric emergency department Pregnant women Rouphael NG, et al. Clostridium difficile-associated diarrhea: an emerging threat to pregnant women. Am J Obstet Gynecol. 2008 Jun;198(6):635.e1-6. Using Emerging Infection Disease network 10 women with severe CDI found Adults without known risk factors
High incidence of Clostridium difficileassociated diarrhoea with a community onset in a hyperendemic region in Germany (HP Weil et al : ECCMID, 2007) August 2006-December 2006 in North West Germany with 103 GP s in a 1500 km2 area with 1.7 million inhabitants: Patients (>40 yrs of age) with diarrhoea attending a GP were investigated for CDAD Standardized interview, strain characterization Definitions of ECDC/CDC/ESGCD 1133 patients of which: 113 (10%) had CDI 47 (4.2%) S.enterica 31 (3.7%) campylobacter
High incidence of Clostridium difficileassociated diarrhoea with a community onset in a hyperendemic region in Germany Of 113 CDAD episodes, 31 (27%) were not healthcare-associated and not associated with prior antibiotic treatment. CO-CA-CDAD patients were significantly younger, had less underlying diseases, and the course of CDAD was less severe with fewer relapses. Most of CO-HA-CDAD patients (79%) lived in a southern subregion where hospitals with increased CDAD incidence rates were located, whereas COCA-CDAD patients were equally distributed in northwest Germany.
Number of tests Number of positive tests
Of 2,000 randomly selected fecal samples, 2.1% was positive for C. difficile cytotoxin Approximately one-third neither had antibiotic exposure nor recent hospitalisation
Clinical characteristics of communityonset Clostridium difficile infection in The Netherlands. Martijn Bauer, Jaap van Dissel, Ed Kuijper Uncontrolled prospective study Of 2,423 patients included, 37 patients (1.5%) CDI 23% had no risk factor for CDI 13 different PCR ribotypes Bauer MP, Goorhuis A, Koster T, et al. Community-onset Clostridium difficile-associated diarrhoea not associated with antibiotic usage. Two case reports with review of the changing epidemiology of Clostridium difficile-associated diarrhoea. Neth J Med 2008; 66(5):207-11.
Antimicrobial susceptibility pattern of Type 027 Clindamycin resistant isolates in Switzerland, Ireland and France Fluoroquinolones susceptible isolates in Sweden Erythromycin susceptible isolates in Germany and Denmark
IR8 IR17 3 34 IR9 FR7 IR20 11 IR27 FR1 FR2 IR3 IR6 IR7 IR19 2 IR21 IR24 2 IR26 2 24 ZW3 19 IR14 39 1 4 2 IR13 IR5 2 1 IR22 IR25 2 IR18 21 9 IR16 4 ZW16 IR12 32 ZW17 ZW20 2 ZW7 ZW2 ZW8 ZW10 ZW13 16 1 ZW5 ZW6 ZW9 1 38 ZW4 1 Single Locus Variance Double Locus Variance Triple Locus Variance Quadruple Locus Variance ZW21 ErmB positive strains ZW1 1 ZW11 ZW14 ZW19 FR4 16 IR15 FR8 39 ZW12 ZW15 FR5 2 4 IR10 1 FR3 7 ZW18 1 1 38 1 41 IR11 Irish strains Swiss strains French strains FR9 8 FR6
Fluoroquinolone resistant Type 027 FP6 Project LSHE-CT-2006-037870 European approach to combat outbreaks of Clostridium difficile associated diarrhoea by development of new diagnostic tests P. Spigaglia and P. Mastrantonio (Istituto Superiore di Sanità, Rome, It); C. difficile lacks genes for topoisomerase IV; alterations in the QRDR of either GyrA or GyrB 20 Type 027 strains from the European surveillance study (2005): All 20 contained one amino-acid substitution in GyrA (Thr82 to Ile) Of 20 isolates, 1 had a MIC=6 for moxifloxacin, and 9 had a MIC>32 mg/l
Conclusions C difficile type 027 has now affected HCFs in 16 European countries: 9 countries have reported outbreaks and 7 European countries have only reported sporadic cases Antimicrobial pattern to macrolides, clindamycin and fluoroquinolones is not predictable anymore Two countries reported outbreaks due to clindamycin resistant Type 027 New emerging types are likely to arrive in the near future, such as Type 078 CDI in animals merits more attention