A Prospective Study On Risk Factors For Toxigenic Clostridium Difficile Colonization And Infection Among Inpatients With Hematologic Disease And Analysis Of Drug Resistance Of Clinical Isolates

Objective: Clostridium difficile-associated diarrhea (CDAD), as the mainmanifestation of the intestinal infection, is caused by excessive proliferation oftoxigenic C.difficile and released toxins in the gut with clinical featuresranging from mild diarrhea to severe pseudomembranous colitis. Since the21st century, with the large-scale outbreaks of nosocomial infection caused bythe hypervirulent strain (ribotype027/NAP1/toxinotype Ⅲ) in North Americaand Europe, the increasing incidence and severity of hospital-acquired andcommunity-acquired C. difficile infection has attracted worldwide attention.Especially elderly patients with recent history of hospitalization or long-termcare facilities prone to have C.difficile infection. Few studies on risk factorsfor C.difficile colonization and infection were reported in mainland China.Patients with hematologic diseases usually have complex underlying disease,long-term receipt of proton pump inhibitors (PPIs), chemotherapy andantibiotics, whether are likely infected with C.difficile due to immunedeficiency or long-term hospitalization is still unknown. This study aims toclarify the risk factors for toxigenic C.difficile colonization and infectionamong new patients with hematologic disease, meanwhile investigate theribotypes and antimicrobial susceptibility of clinical isolates for monitoringdrug resistance.Methods: A7-month prospective study on new admissions withhematologic disease was conducted in the Second Hospital of Hebei MedicalUniversity from March2013to September2013. Stool samples were obtainedfrom patients within48hours of admission, weekly during hospitalization andat the time of diarrhea happening. Stool samples were transported to the labor- atory within2hours after collection. Strains isolated through Bacillusenrichment method were confirmed to be C.difficile based on the typical odor,appearance of colonies, microscopic morphology with Gram stain, andspecific biochemical reactions. Bacterial genomic DNA extraction kit wasused for pure colony DNA extraction. In addition, a5-plex polymerase chainreaction was done to confirme the presence of tcdA, tcdB, cdtA, cdtB,16SrDNA while screening for the existence of the binary toxin genes.Conventional PCR was conducted for ribotyping of toxigenic isolates bymeans of analysis of16~23S internal spacer region polymorphism. The bandswere compared using Quantity One software. MICs of45toxigenic isolateswere done for metronidazole, vancomycin and fidaxomicin using agar dilutionmethod. In this study, diarrhea was defined as three loose stools within at leastone24h-period. Toxigenic C.difficile colonization (tCDC) was defined as apositive stool culture for toxigenic C.difficile in the absence of diarrhea. C.difficile-associated diarrhea(CDAD) was defined as the presence of diarrheawithout an alternative explanation and a positive toxigenic culture. Allenrolled patients were followed until discharge or death, the primary outcomewas the occurrence of CDAD. No stool collection if the patient wasreadmitted.Results: Of180patients enrolled,105patients had finished the study.And45toxigenic C.difficile isolates was isolated from29patients. MultiplexPCR found that the tcdA and tcdB genes were positive in all the isolates whilethe binary-toxin gene were negative.21(20.0%) patients had tCDC at the timeof admission, of which14(66.7%) had history of hospitalization within theprior two months,13(61.9%) had antibiotic use history within30days. Afterhospitalization8(7.6%) occurred hospital-acquired tCDC, compared withpatients had no tCDC until discharged, the age (58.2vs53.2years old, P=0.00), length of hospital stay (29.0vs15.0days, P=0.003), and the use ofcarbapenems (37.5vs7.9%, P=0.037) or carbapenems combinedfluoroquinolone antibiotics (25.0vs1.3%, P=0.023) were statisticaldifference. During hospitalization,6(5.7%) occurred CDAD, compared with patients without CDAD, the CDAD patients were more likely to have a historyof tCDC before CDAD, prolonged hospitalization (mean32.0days), in use oftwo or more antibiotics, or prescribed with glycopeptides or glycopeptidecombined β-lactam antibiotics.45C.difficile is divided into12ribotypes,epidemic strains PR001and PR002accounted for66.7%. The hypervirulentstrain (ribotype027/NAP1/toxinotype Ⅲ) was not detected in this study yet.Currently we didn’t found the strains resistant against metronidazole,vancomycin and fidaxomicin.Conclusions:1For hematologic patients, hospitalization within2months beforeadmission, history of antibiotic use are risk factors for tCDC at admission.Patients who were elderly, had long-term hospitalization, using carbapenemsor carbapenems combined fluoroquinolone antibiotics prone to havehospital-acquired tCDC. A history of tCDC, prolonged hospitalization, in useof two or more antibiotics, or prescribed with glycopeptides or glycopeptidescombined β-lactam antibiotics were correlated with the incidence of CDAD.2The hypervirulent strain (ribotype027/NAP1/toxinotype Ⅲ) was notyet detected in this study.3Currently C. difficile strains isolated from this study were stillsusceptible to metronidazole, vancomycin and fidaxomicin.