Extraintestinal Clostridium difficile Infections: A Single-Center Experience

ORIGINAL ARTICLE

Extraintestinal Clostridium difficile Infections: A Single-Center Experience Arjun Gupta, MBBS; Robin Patel, MD; Larry M. Baddour, MD; Darrell S. Pardi, MD, MS; and Sahil Khanna, MBBS, MS Abstract Objectives: To evaluate the clinical burden of extraintestinal Clostridium difficile infection (CDI) seen at a single institution and to characterize the management and outcomes of these rare infections. Patients and Methods: A retrospective medical record review was conducted to identify patients with isolation of C difficile from extraintestinal sites from January 1, 2004, through December 31, 2013. Medical records were reviewed and data, including demographic characteristics, microbiology, clinical associations, management, and infection outcomes, were abstracted. Results: Overall, 40 patients with extraintestinal CDI were identified: 25 had abdominopelvic infections, 11 had bloodstream infections, 3 had wound infections, and 1 had pulmonary infection. C difficile was isolated with other organisms in 63% of cases. A total of 85% of infections were nosocomial. Factors associated with extraintestinal CDI included surgical manipulation of the gastrointestinal tract (88%), recent antibiotic exposure (88%), malignant tumors (50%), and proton pump inhibitor use (50%). Diarrhea was present in 18 patients (45%), 12 of whom had C difficile polymerase chain reaction (PCR)e positive stool samples. All isolates tested were susceptible to metronidazole and piperacillin-tazobactam. Management included both antimicrobial therapy and guided drainage or surgical intervention in all but one patient. The infection-associated mortality rate was 25%, with death a median of 16 days (range, 1-61 days) after isolation of C difficile. Conclusion: Extraintestinal CDI is uncommon and often occurs in patients with surgical manipulation of the gastrointestinal tract and well-recognized risk factors for intestinal CDI. Management of extraintestinal CDI includes both antimicrobial and surgical therapies. Extraintestinal CDI is characterized by poor outcome with high mortality. ª 2014 Mayo Foundation for Medical Education and Research

C

lostridium difficile is the most common cause of infectious diarrhea in hospitalized patients and may lead to pseudomembranous colitis.1,2 Studies have found increasing incidence, severity, and recurrence rates of C difficile infection (CDI),3,4 and it has surpassed methicillin-resistant Staphylococcus aureus (MRSA) as the most common hospital-acquired infection in the United States.5 C difficile was isolated from outside the gastrointestinal (GI) tract before its recognition in 1978 as an enteric pathogen that causes colitis.6 The burden of C difficile is highest as a cause of diarrhea, with reports of it causing extraintestinal infections being scarce and heterogeneous.7-12 We performed a retrospective review of patients with extraintestinal CDI at our institution during the past decade to better characterize the syndrome.

n

Mayo Clin Proc. 2014;89(11):1525-1536

PATIENTS, MATERIALS, AND METHODS Electronic medical records from the clinical microbiology laboratory were searched to identify isolates of extraintestinal C difficile from January 1, 2004, through December 31, 2013. Medical records of cases with recovery of this organism from extraintestinal sites were reviewed. Patients were followed up until March 1, 2014, death, or relapse or reinfection. The study was approved by the institutional review board of Mayo Clinic’s campus in Rochester, Minnesota. Descriptive statistics were used to summarize the demographic, microbiological, clinical, treatment, and outcome details using JMP statistical software, version 9.01 (SAS Institute Inc).

From the Division of Infectious Diseases (A.G., R.P., L.M.B.), Division of Gastroenterology and Hepatology (D.S.P., S.K.), and Division of Clinical Microbiology (R.P.), Department of Medicine, Mayo Clinic, Rochester, MN.

Identification of C difficile Isolates C difficile was isolated from extraintestinal sites using conventional anaerobic culture techniques. Blood for bacterial or Candida culture

Mayo Clin Proc. n November 2014;89(11):1525-1536 n http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org n ª 2014 Mayo Foundation for Medical Education and Research

1525

MAYO CLINIC PROCEEDINGS

was inoculated into blood culture media bottles (BD Biosciences), with 2 BACTEC Plus Aerobic/F bottles and a BACTEC Lytic Anaerobic/F routinely inoculated for each culture. Fluids for anaerobic culture were either directly inoculated into blood culture bottles or placed into anaerobe transport vials that contained prereduced anaerobically sterilized peptone yeast extract with inoculation into blood culture bottles and/or processing for culture as described subsequently in the laboratory. Tissues for anaerobic culture were collected in anaerobic culture tubes. Extraintestinal specimens were inoculated into thioglycolate broth and onto CDC anaerobic sheep blood, laked kanamycin or vancomycin blood, and phenylethyl alcohol blood agar. Toxin testing was not performed on extraintestinal isolates. Metronidazole, clindamycin, penicillin, piperacillin-tazobactam, and ertapenem susceptibility testing was performed using the Epsilometer test method. Before July 2007, the Premier Toxins A & B (Meridian Diagnostics Inc) enzyme immunoassay was used for testing stool for C difficile toxin. From July 2007 onward, C difficile toxin testing of stool was performed using a previously described PCR assay.13 The total number of patients with intestinal CDI during the same time period was calculated using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code of 008.45. ICD9-CM codes were not used to identify extraintestinal CDI cases, and all extraintestinal CDI cases were laboratory confirmed.

was isolated, with 25 (63%) being abdominopelvic infections, 11 (28%) being bloodstream infections (BSIs), 3 (8%) being wound infections, and 1 (3%) a lung infection (Table 2). Some patients had C difficile isolated on more than one occasion (7 patients had 2 and 2 patients had 3 positive cultures from the same anatomical region); no patient had multiple C difficile isolates from specimens from anatomically distinct regions. Table 3 provides individual clinical features and outcomes in these patients. Clinical Associations Health Care Exposure. A total of 34 of the 40 patients (85%) developed infection during hospitalization, and 5 patients (13%) had been hospitalized or had an outpatient visit in the antecedent 2 months. Only one patient did not report recent antecedent health care exposure. Antimicrobial Exposure. Systemic antibiotic exposure in the antecedent 2 months was present in 35 of 40 patients (88%). All patients with BSI and wound infections had preceding antibiotic exposure. Proton Pump Inhibitor Use. Twenty of the 40 patients (50%) were receiving proton pump inhibitors (PPIs) with maximum use in patients with BSI (9 of 11 patients [82%]). Indications included gastroesophageal reflux disease, peptic ulcer disease, and stress ulcer prophylaxis in the intensive care unit.

RESULTS Baseline Characteristics During the 10-year study period, C difficile was isolated from extraintestinal sources in 51 samples from 40 patients; 19 of the 40 patients (48%) had organisms recovered from 2011 through 2013, and 8 (20%) had organisms isolated in 2013. During this period, 6525 patients with intestinal CDI were identified. Thus, 40 of 6525 (0.6%) of all CDIs were extraintestinal. The median patient age was 56.5 years (range, 6-88 years), and 19 (48%) were male. Patient characteristics are summarized in Table 1. Sites of Infection The CDIs were categorized into 3 groups according to the site from which the organism 1526

Mayo Clin Proc.

n

GI Surgery A total of 24 of the 25 patients (96%) with abdominopelvic infections had a GI surgical procedure performed as an elective or an emergency procedure for an acute abdomen a median of 25 days before C difficile isolation. One patient who did not undergo a surgical procedure before C difficile isolation had diffuse large B-cell lymphoma treated with intensive chemotherapy and had developed diverticulitis and a pelvic abscess treated with computed tomographic drainage and antibiotics. These procedures involved the colon in most patients; 2 patients had surgical procedures that did not involve the colon, including one patient who developed a large abdominal abscess after peptic ulcer perforation and another with

November 2014;89(11):1525-1536

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org

EXTRAINTESTINAL C DIFFICILE INFECTIONS

TABLE 1. Patient Characteristicsa,b Characteristic

Entire cohort (N¼40)

Abdominopelvic infection (n¼25)

Bloodstream infection (n¼11)

Wound or lung infection (n¼4)

Age (y), mean (range) Male/female sex Inpatient at time of culture Outpatient health care exposure (if not inpatient) Any health care exposure History of diarrhea Clostridium difficile stool PCR positive MRSA colonization Vancomycin-resistant enterococci colonization Diabetes mellitus Congestive heart failure Peripheral vascular disease Chronic obstructive pulmonary disease Cirrhosis or chronic liver disease Chronic kidney disease Alcohol use History of stroke Peptic ulcer disease Connective tissue disease Malignant tumor Immunocompromising therapyc Inflammatory bowel disease History of GI surgery Antibiotic exposure Proton pump inhibitor use

56.5 (6-88) 19/21 34 (85) 5/6 (83) 39 (98) 18 (45) 12/18 (67) 8 (20) 9 (23) 8 (20) 11 (28) 7 (18) 4 (10) 4 (10) 4 (10) 15 (38) 4 (10) 6 (15) 1 (3) 20 (50) 23 (58) 8 (20) 35 (88) 35 (88) 20 (50)

57 (16-84) 15/10 21 (84) 1/ 4 (25) 24 (96) 7 (28) 6/7 (86) 0 (0) 5 (20) 4 (16) 5 (20) 4 (16) 1 (4) 2 (8) 3 (12) 8 (32) 2 (8) 4 (16) 0 (0) 15 (60) 16 (64) 5 (20) 24 (96) 20 (80) 10 (40)

51 (27-88) 1/10 10 (91) 1/1 (100) 11 (100) 10 (91) 5/10 (50) 7 (63) 3 (27) 2 (18) 3 (27) 2 (18) 2 (18) 2 (18) 0 (0) 6 (55) 1 (9) 1 (9) 0 (0) 4 (36) 5 (45) 3 (27) 11 (100) 11 (100) 9 (82)

61.5 (6-67) 3/1 3 (75) 1/1 (100) 4 (100) 1 (25) 1/1 (100) 1 (25) 1 (25) 2 (50) 3 (75) 1 (25) 1 (25) 0 (0) 1 (25) 1 (25) 1 (25) 1 (25) 1 (25) 1 (25) 2 (50) 0 (0) 0 (0) 4 (100) 1 (25)

GI ¼ gastrointestinal; MRSA ¼ methicillin-resistant Staphylococcus aureus; PCR ¼ polymerase chain reaction. Data are presented as No. (percentage) of patients unless otherwise indicated. c Immunocompromising therapy was corticosteroid use and/or anticancer and/or antimetabolite chemotherapy. a

b

chronic liver disease and hepatocellular cancer after chemoembolization, who presented with monomicrobial C difficile peritonitis. All 11 patients with BSI had undergone a recent GI surgical procedure. Overall, 36 of the 40 patients (90%) had either a recent GI surgical procedure or an intestinal perforation; the remaining 4 had wound or lung infections. Intestinal CDI and Bacterial Colonization A total of 18 of the 40 patients (45%) reported diarrhea in the antecedent 4 weeks, and all but TABLE 2. Sites of Infection Site of infection Abdominopelvic infection Blood stream infection Wound infection Lung infection

No. (%) of patients (N¼40) 25 11 3 1

(60) (28) (8) (3)

Mayo Clin Proc. n November 2014;89(11):1525-1536 www.mayoclinicproceedings.org

n

one of the patients with BSI had diarrhea. Stool samples from 12 of these 18 tested positive for C difficile toxin by PCR. Eight patients tested positive for MRSA, and the rectal swabs of 9 patients were positive for vancomycinresistant enterococci. Malignancy Malignancy was present in 20 of the patients: 15 of 25 patients with abdominopelvic infection, 4 of 11 patients with BSI, and 1 of 4 patients with wound or lung infections. Seventeen patients had a solid organ tumor, 2 had lymphoma, and 1 had acute lymphoblastic leukemia. Fourteen of the 17 patients with solid tumors had a primary tumor in the abdominopelvic region, with colorectal cancer (5 cases) predominating. Inflammatory Bowel Disease Inflammatory bowel disease (IBD) was present in 8 of the 40 patients (20%); 5 of 25 patients

http://dx.doi.org/10.1016/j.mayocp.2014.07.012

1527

1528

TABLE 3. Extraintestinal Clostridium difficile Infections Sample Age No. (y)/Sex

Underlying disease

Clinical setting

Abdominopelvic infections 1 56/M Metastatic gastrinoma Previous liver transplant, now diverticulitis to liver, C difficile colitis Previous total 2 26/M Severe ulcerative colitis, failed medical proctocolectomy and ileostomy management

Mayo Clin Proc.

Appendicitis

November 2014;89(11):1525-1536

Previous emergency appendectomy Severe C difficile colitis Previous exploratory with ileus laparostomy and ileostomy Nonesmall cell lung Appendicitis after cancer laparotomy Previous hemicolectomy, Retroperitoneal intraoperative liposarcoma involving liver radiotherapy Appendiceal carcinoma Previous debulking, ileostomy

4

29/M

5

77/F

6

56/F

7

71/M

8

16/F

Crohn disease on maximal medical management

9

52/M

Sigmoid diverticulitis

10

56/M

11

69/M

12

33/M

Previous hemicolectomy, stricturoplasty, ileocolotomy

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org

Previous sigmoid resection Mantle cell lymphoma, Previous exploratory appendicitis laparotomy, pancreatic leak, enterocutaneous fistula C difficile colitis leading Previous total colectomy to toxic megacolon and ileostomy in setting of IBD Acute lymphoblastic Previous right leukemia, now cecal hemicolectomy, perforation ileostomy

Other organisms isolated Organisms isolated from from same sample other samples Surgical management

Pelvic fluid

Enterococcus species

Abscess fluid

Escherichia coli

Abscess fluid (2 samples) Peritoneal fluid

Peritoneal fluid Pelvic fluid

. Candida species

E coli, Bacteroides species .

Abscess fluid

.

Abdominal fluid (2 samples)

Candida species

Peritoneal fluid

E coli

Prevotella species, Pancreas, Enterococcus species, Abdominal fluid P aeruginosa, (2 samples) P aeruginosa Peritoneal fluid, C albicans, Abdominal wall Clostridium (2 samples) clostridioforme, C albicans Peritoneal fluid .

Abdominal washout Escherichia coli, Pseudomonas aeruginosa, Clostridium perfringens CT aspiration and Enterococcus species, drain placement viridans group Streptococcus species, Candida albicans . CT aspiration and drain placement . Abdominal washout, drain placement .

Abdominal washout

Medical management

Outcome

Metronidazole, linezolid

Resolved

Levofloxacin, amoxicillinclavulanic acid

Resolved

Ciprofloxacin, metronidazole Vancomycin, meropenem, caspofungin Ertapenem

Resolved

E coli, Clostridium species, Enterobacter species

CT drain placement, Metronidazole, t-PA injection piperacillintazobactam E coli, Enterococcus species, CT drain placement Metronidazole, Eggerthella lenta piperacillintazobactam . CT drain placement Metronidazole, piperacillintazobactam, caspofungin E coli, Enterococcus species, CT drain placement PiperacillinClostridium species tazobactam CT drain placement Vancomycin, Enterococcus species, P meropenem aeruginosa, Lactobacillus species

Resolved

Resolved Resolved

Resolved

Resolved

Resolved Died 18 days later

C albicans, MRSE

CT drain placement, Vancomycin, abdominal washout cefepime, caspofungin

Died 29 days later

E coli, C perfringens, Bacteroides species

CT drain placement

Resolved

Ertapenem

Continued on next page

MAYO CLINIC PROCEEDINGS

51/M

n

3

Source

Sample Age No. (y)/Sex

Underlying disease

Clinical setting

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012

Abdominopelvic infections, continued Previous 13 57/M Hepatocellular chemoembolization, carcinoma, ESRD, now ascites C difficile colitis 14 63/F Metastatic endometrial Previous TAH/BSO, now cecal perforation after cancer, C difficile hemicolectomy colitis 15 76/M Peptic ulcer disease Perforation after laparotomy and repair 16 84/M Bladder cancer Previous radical cystoprostatectomy, sigmoid urinary diversion 17

57/M

Rectal cancer

18

40/M

19

22/F

Alcohol abuse, appendicitis Crohn disease with appendicitis

20

57/F

21

35/F

22

63/F

23

79/F

24

77/M

Source

Other organisms isolated Organisms isolated from from same sample other samples Surgical management

Peritoneal fluid

.

Vancomycin, metronidazole

Died 1 day later

.

CT drain placement

Meropenem, daptomycin, caspofungin Cefepime, metronidazole Vancomycin, piperacillintazobactam, ciprofloxacin Vancomycin, ertapenem Piperacillintazobactam Metronidazole, levofloxacin

Died 61 days later

Abscess fluid Peritoneal fluid (2 samples)

E coli, P aeruginosa, . CT drain placement Klebsiella pneumoniae Enterococcus species E coli, Enterococcus species Paracentesis Enterococcus species

MRSE

Abscess fluid

E coli, E lenta

Previous laparotomy

Pelvic fluid

Clostridium cadaveris, viridans group Streptococcus species E coli, Bacteroides species, Parvimonas micra E coli, P aeruginosa, Enterococcus species, Porphyromonas levii E coli, Bacteroides fragilis

Diffuse large B-cell Diverticulitis with pelvic lymphoma while abscess taking methotrexate Previous pelvic Recurrent bladder cancer metastatic to exenteration, colostomy, urostomy pelvic side walls Crohn disease, colon Previous liver transplantation, carcinoma, PSC, colectomy with C difficile colitis ileosigmoid anastomosis Colonic Previous hemicolectomy adenocarcinoma with enterocutaneous fistula Recurrent colonic Previous right/transverse adenocarcinoma colectomy and hernia repair

Paracentesis

P aeruginosa, Candida tropicalis

Pelvic fluid

Abscess fluid

Left iliac fossa tissue Abdominal wall abscess

Abdominal wall mesh Abdominal wall

Klebsiella sp, Enterococcus species, P aeruginosa .

Outcome

.

Peritoneal fluid

Previous low anterior resection Previous laparotomy

Medical management

.

CT drain placement

E coli, S anginosus, Candida CT drain placement species E coli, Klebsiella species CT drain placement

.

CT drain placement

.

Extensive debridement

.

Drainage

P aeruginosa, Enterococcus Debride, mesh species explanted .

US drain placement

Metronidazole, vancomycin, cefepime Linezolid, meropenem

Resolved Died 2 days later

Resolved Resolved Resolved

Resolved

Resolved

Ampicillinsulbactam

Resolved

Meropenem, tigecycline

Resolved

Metronidazole

Resolved

Continued on next page

EXTRAINTESTINAL C DIFFICILE INFECTIONS

Mayo Clin Proc. n November 2014;89(11):1525-1536 www.mayoclinicproceedings.org

TABLE 3. Continued

1529

1530

TABLE 3. Continued Sample Age No. (y)/Sex

Underlying disease

Clinical setting

Abdominopelvic infections, continued 25 82 /F Sigmoid colon rupture Previous sigmoid colectomy, colostomy

Source Peritoneal fluid

Other organisms isolated Organisms isolated from from same sample other samples Surgical management

E coli, P aeruginosa

.

Candida tropicalis

.

Enterococcus species, Candida species, Klebsiella species Clostridium orbiscindens, ., Clostridium orbiscindens .

30

47/F

31

79/F

32

80/F

33

51/F

Blood (2 samples) Enterococcus species, Crohn disease, multiple Self-inflicted abdominal Clostridium ramosum, wounds, suspicion for suicide attempts Bacteroides species factitious contamination after self-stab to (both samples) abdomen leading to liver laceration Colorectal cancer after Ovarian cyst after Blood . resection, C difficile oophorectomy, colitis postoperative confusion, ascites DM, CHF, COPD, Diverticulitis after Blood . stroke laparotomy . Ileal neuroendocrine Anastomotic breakdown Blood and postoperative fever tumor, Crohn disease after ileal and sigmoid resection, C difficile colitis

.

E coli

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org

.

Outcome

Metronidazole, piperacillintazobactam

Resolved

Metronidazole, cefepime, ciprofloxacin Abdominal washouts Meropenem

Resolved

Paracentesis

Resolved

US drain placement

.

Metronidazole, cefepime

Died 17 days later

Debridement of groin Meropenem, infection linezolid

Resolved

Wound debridement Piperacillintazobactam

Resolved

Paracentesis

Vancomycin, piperacillintazobactam

Died 7 days later

Abdominal washout

Cefepime

Anastomotic takedown, colostomy, washout

Levofloxacin, metronidazole

Died 6 days later Resolved

Continued on next page

MAYO CLINIC PROCEEDINGS

Alcohol abuse after Recurrent groin cellulitis inguinal hernia repair

n

41/F

November 2014;89(11):1525-1536

.

29

n

Mayo Clin Proc.

Bloodstream infection C difficile colitis, lower GI Blood B fragilis 26 88/F Peptic ulcer disease bleed after partial gastrectomy Cecal impaction and Blood (2 samples) 27 75/F Squamous cell carcinoma of mouth rupture after laparotomy after resection 28 46/F Hepatic adenoma after Alcoholic hepatitis and Blood resection ascites Blood (3 samples)

.

E coli, E lenta

Medical management

Sample Age No. (y)/Sex

Underlying disease

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012

Bloodstream infection, continued 34 35/M Congenital pancreatic duct abnormality after pancreatectomy, splenectomy, C difficile colitis 35 56/F COPD, concurrent C difficile colitis, SIBO after adhesiolysis 36 27/F Crohn disease, recurrent C difficile colitis Wound infections 37 6/M Supracondylar fracture humerus after closed reduction and pinning 38 62/M Myxoid liposarcoma of right groin after wide excision 39 61/F SLE with DM, stroke, PAD after BKA of right lower extremity due to nonhealing ulcers, C difficile colitis Lung infection 40 67/M COPD, CHF after mitral valvuloplasty, CT revealing mediastinal lymphadenopathy, loculated effusions

Clinical setting

Source

Other organisms isolated Organisms isolated from from same sample other samples Surgical management

Recurrent polymicrobial bacteremia and skin abscesses

Blood

.

Abdominal compartment syndrome, surgical wound infection Previous right hemicolectomy and ileostomy

Blood

.

Blaudia coccoides, K pneumoniae, E coli

.

Medical management

Skin debridement

Meropenem, linezolid

Resolved

Wound debride

Metronidazole, vancomycin

Resolved

Resolved

Blood

Citrobacter species, Streptococcus anginosus

Streptococcus anginosus

Anastomotic Metronidazole, takedown, washout vancomycin, ertapenem

Drainage and distal humerus osteomyelitis

Elbow sinus tract

E coli, P aeruginosa, Bacteroides species

Other 3 samples yielded only MSSA

Necrosis and cellulitis of surgical site

Thigh

MRSE

Extensive debridement, antibiotic bead placement Extensive debridement

E coli, MRSE, Enterococcus species, Propionibacterium acnes . Debridement

Drainage and infection of Knee (3 samples) stump

Dyspnea

Lung biopsy

., Enterobacter species, Clostridium tertium

.

.

Outcome

Decortication

Cefepime, metronidazole

Resolved

Vancomycin, metronidazole, levofloxacin Metronidazole, ertapenem

Resolved

Died 40 days later

Cefepime, levofloxacin

Died 34 days later

BKA ¼ below knee amputation; CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; CT ¼ computed tomography; DM ¼ diabetes mellitus; ESRD ¼ end-stage renal disease; IBD ¼ inflammatory bowel disease; MRSE ¼ methicillin-resistant Staphylococcus epidermidis; PAD ¼ peripheral arterial disease; PSC ¼ primary sclerosing cholangitis; SIBO ¼ small intestinal bowel obstruction; SLE ¼ systemic lupus erythematosus; US ¼ ultrasonography.

EXTRAINTESTINAL C DIFFICILE INFECTIONS

Mayo Clin Proc. n November 2014;89(11):1525-1536 www.mayoclinicproceedings.org

TABLE 3. Continued

1531

MAYO CLINIC PROCEEDINGS

TABLE 4. Microbiology of Extraintestinal Clostridium difficile Infections

Monomicrobial C difficile isolation from extraintestinal specimen Monomicrobial C difficile isolation from extraintestinal specimen but other microbes isolated from specimens collected within a day Monomicrobial C difficile isolation after excluding patients in above row

Entire cohort (N¼40)

Abdomino-pelvic infection (n¼25)

Bloodstream infection (n¼11)

Wound or lung infection (n¼4)

15 (37.5)

6 (24)

8 (73)

1 (25)

8/15 (53) 7 (17.5)

3/6 (50) 3 (12)

5/8 (63) 3 (27)

0/1 (0) 1 (25)

Data are presented as No. (percentage) of patients.

with abdominopelvic infections and 3 of 11 patients with BSI had IBD. Four of the 8 patients with IBD had concurrent intestinal CDI at the time of diagnosis of extraintestinal CDI.

across the cohort had infections that would be considered monomicrobial (3 of 25 abdominopelvic infections, 3 of 11 BSI, 0 of 3 wound infections, and 1 of 1 lung infection).

Immunocompromising Medications A total of 23 of 40 (58%) across the cohort had received anticancer and/or antimetabolite chemotherapy or corticosteroids for IBD or connective tissue disorders in the antecedent year.

Antimicrobial Susceptibility Testing. Susceptibility test results were available for isolates from 35 of the 40 patients (88%) (Table 5). In the case of multiple specimens testing positive for C difficile, susceptibilities were the same as the first isolate or were only performed on the first isolate. All tested isolates were susceptible to metronidazole and piperacillin-tazobactam. In contrast, no isolate tested was susceptible to penicillin. Only 3 of 35 isolates were susceptible to clindamycin, whereas 13 of 35 isolates were not susceptible to ertapenem (3 resistant and 10 intermediate).

Microbiology Polymicrobial Infection. C difficile was isolated with other microorganisms in 25 of the 40 patients (63%). Monomicrobial isolation of C difficile occurred in 6 of 25 abdominopelvic infections, 8 of 11 BSIs, and 1 of 1 lung infection (15 of 40 patients [33%]) (Table 4). Among the 8 patients with an initial monomicrobial C difficile BSI, 5 had other organisms cultured from different blood cultures on the same day or within a day. Similarly, among the 6 patients who initially had monomicrobial abdominopelvic infections, 3 had other organisms isolated from specimens from the same anatomical site collected within a day of collection of the specimen, which yielded C difficile. If these patients are included in the polymicrobial infection group, only 7 (18%) patients

Management All patients received antibiotics, and 39 of 40 (98%) also underwent image-guided drainage and/or a surgical procedure (Table 6). Abdominopelvic Infections. Depending on the location and extent of abdominopelvic infection, patients underwent abdominal washouts, drain placement, paracentesis, or debridement. Most patients developed intra-abdominal

TABLE 5. Antimicrobial Susceptibilities (Available for 35 Isolates Unless Otherwise Stated)a,b Antibiotic

Susceptible

Intermediate

Resistant

Clindamycin Ertapenem Metronidazole Penicillinc Piperacillin-tazobactamd

3 (9) with MIC of 2 mg/mL 22 (63) with MIC 4 mg/mL 35 (100) with MIC 0.5 mg/mL e 30 (100) with MIC 16/4 mg/mL

14 (40) with MIC of 4 mg/mL 10 (28) with MIC of 8 mg/mL e 1 (3) with MIC of 1 mg/mL e

18 (51) with MIC 8 mg/mL 3 (9) with MIC of 16 mg/mL e 34 (97) with MIC 2 mg/mL e

MIC ¼ minimum inhibitory concentration; e, no data available. Data are presented as No. (percentage) of patients unless otherwise indicated. c One isolate with intermediate susceptibility to penicillin G was susceptible to both clindamycin and ertapenem. d Susceptibility analysis was available for 30 isolates for piperacillin-tazobactam. a

b

1532

Mayo Clin Proc.

n

November 2014;89(11):1525-1536

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org

EXTRAINTESTINAL C DIFFICILE INFECTIONS

abscesses after postoperative intestinal anastomotic leakage, which initially required drainage (Table 6). Bloodstream Infections. Patients with BSI underwent procedures depending on the clinical scenario and proposed source of BSI; skin abscesses and cellulitis were managed with debridement and drainage, whereas BSI due to a postoperative abdominal abscess required anastomotic takedown and abdominal washout (Table 6). The only patient in the cohort who received only antibiotics was an 88-year-old woman who had several episodes of polymicrobial BSI with feculent flora after an episode of intestinal CDI. A possible gateway of infection was not located after an intensive search, including colonoscopy, imaging, and an indium scan, and she was treated with intravenous cefepime and ciprofloxacin and oral metronidazole. Wound and Lung Infections. Standard wound care and debridement were the primary surgical management means used. A 6-year-old boy with distal humeral osteomyelitis was successfully managed with debridement and antibioticimpregnated bead placement. The patient with the lung infection was managed with decortication and antibiotics (Table 6). Treatment Response and Outcomes. In 30 of the 40 patients (75%), infection resolved with initial antibiotic treatment and surgical intervention. Ten of these patients required intensive care unit admission during or in the immediate period after extraintestinal CDI was detected, owing to concurrent issues. No recurrences or relapse of extraintestinal disease occurred. Ten of the 40 patients died within a median duration of 16 days (range, 1-61 days) after detection of extraintestinal CDI, usually in the same hospital admission. The immediate extraintestinal CDI-associated mortality rate was 20% (n¼5/25) in abdominopelvic, 27% (n¼3/11) in BSI, 33% (n¼1/3), in wound infection, and 100% (n¼1/1) in lung infection. All these patients required intensive care unit admission, and their death was attributed to uncontrollable CDI rather than their underlying comorbidities. The remaining 30 patients were discharged successfully after treatment, Mayo Clin Proc. n November 2014;89(11):1525-1536 www.mayoclinicproceedings.org

n

TABLE 6. Initial Management Management

No. of patients

Medical and surgical therapy Medical therapy alone Initial antimicrobial regimen b-lactam alone Metronidazole plus b-lactam Vancomycin plus b-lactam Metronidazole plus fluoroquinolone b-lactam plus linezolid Metronidazole plus vancomycin b-lactam plus fluoroquinolone Metronidazole plus vancomycin plus b-lactam Metronidazole alone Metronidazole plus linezolid b-lactam plus daptomycin Vancomycin plus fluoroquinolone plus b-lactam Metronidazole plus fluoroquinolone plus b-lactam Vancomycin plus metronidazole plus fluoroquinolone b-lactam plus tigecycline Additional management Drain placement Abdominal washout Abdominal washout and drain placement Paracentesis Debridement Decortication

39 1 8 8 5 3 3 2 2 2 1 1 1 1 1 1 1 17 6 2 4 9 1

Initial therapy of 4 patients also included caspofungin to cover Candida species isolated in clinical specimens.

and there were no other deaths within 1 year of diagnosis. We have assumed that these patients were cured of their infection. No recurrent infections were noted in the follow-up period. For 5 patients presenting in 2013, the median duration of follow-up is only 8 months (range, 2-11). However, all these patients have been discharged from hospital and were doing well as of their last follow-up. The median duration of follow-up for patients who did not have an infection-associated death was 29.5 months (range, 2-116 months). DISCUSSION To our knowledge, this report describes the largest cohort of extraintestinal CDI described to date. Extraintestinal CDI was rare, accounting for 0.61% of all CDIs. Most infections were in the abdominopelvic region, and there were high rates of health care exposure, antibiotic and PPI use, GI surgery, and malignant tumors. Management was determined on the basis of the

http://dx.doi.org/10.1016/j.mayocp.2014.07.012

1533

MAYO CLINIC PROCEEDINGS

clinical setting and included a variety of procedures and antimicrobial agents. Early infectionassociated mortality was high, but patients tended to do well after an initial 2-month period. Twenty percent of cases were reported in the last year of the 10-year survey. This increase might be mirroring the overall increase in intestinal CDI disease burden and severity. The disease burden outside the intestine might be underestimated because this organism is a fastidious anaerobe, often present in polymicrobial infections; toxin detection assays are not used on nonstool samples in regular clinical practice. In recent years, CDI enteritis14-16 and reactive arthritis after CDI17 have been extensively studied. The literature suggests that toxin-negative strains may be responsible for some cases of extraintestinal CDI.8 Zheng et al18 found that 24% of extraintestinal C difficile isolates analyzed were both toxin A and B negative. Toxinnegative strains have been isolated from monomicrobial CDI of the brain and blood.19,20 Perhaps undefined nontoxin virulence factors are operative in extraintestinal CDI. Identical C difficile isolates have been cultured from the stool and splenic abscess of a patient.21 More recently, monomicrobial C difficile isolates from blood were found to be indistinguishable from enteric isolates from the same patient.22 All 3 monomicrobial C difficile BSIs in our cohort had concurrent intestinal CDI, although typing was not performed to compare intestinal and extraintestinal C difficile strains. The skin contamination rate in carriers and patients with CDI is high.23 With strict adherence to infection control practices, the growth of skin contaminants in blood cultures is usually low.24 In our study, the frequent polymicrobial natures of infections with other enteric organisms support gut bacterial translocation as a mechanism of C difficile bacteremia pathogenesis. An increase in intestinal CDI has been reported in surgical patients, with a maximum increase in patients undergoing emergency operations and intestinal tract resections.25 It is plausible that anatomical disturbance of the lower GI tract allows a path for C difficile and other colonic bacteria to set up an infectious focus in the abdominopelvic region. Our findings of polymicrobial abdominopelvic abscesses and relation to history of intestinal 1534

Mayo Clin Proc.

n

perforation, GI surgical procedures, and frequent anastomotic leakage support his hypothesis of spread of C difficile and are in line with earlier findings.11,12 C difficile was likely introduced into the bloodstream by the above mechanism in our cohort; all the patients with BSI in our study had previously undergone GI operations, whereas none of the patients with wound infections had undergone GI operation or had major GI disease, supporting some other route of spread. Traditional factors associated with intestinal CDI, including antibiotic exposure, hospitalization, and possibly PPI use, have been described in earlier reports of extraintestinal CDI11,12,26 and were present in our cohort. Most patients in our cohort had a diagnosis of cancer or IBD, which has been described previously.10,11,26 They possessed multiple risk factors for intestinal CDI, including frequent hospitalization, antimicrobial exposure, immunocompromised status, and surgical intervention. Both parenteral metronidazole and vancomycin have been used to treat C difficile BSI in the past.7 No specific guidelines for the treatment for extraintestinal CDI exist, in contrast to specific intestinal CDI guidelines.27,28 The antimicrobial(s) chosen to treat infections in our cohort varied widely. Several patients had polymicrobial infections, and therapy was directed to cover other microorganisms. Caspofungin was added to cover Candida species in 4 patients. Some patients were already receiving antibiotics as treatment for other ongoing infections or as prophylaxis for their immunocompromised status; others required expansion of coverage owing to the development of new infections, such as ventilatorassociated pneumonia. Surgical techniques were also commonly used to clear foci of infection, especially when abscesses were present. Although antimicrobial resistance is being reported increasingly among C difficile isolates,29,30 all our isolates were susceptible to metronidazole and piperacillin-tazobactam, which were often used to treat extraintestinal CDI. Vancomycin was also used frequently, although susceptibility testing was not performed to this agent. Interestingly, one-third of isolates were not susceptible to ertapenem. Given these susceptibility findings and the high rates of MRSA colonization and frequent coisolation of Enterococcus species and

November 2014;89(11):1525-1536

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org

EXTRAINTESTINAL C DIFFICILE INFECTIONS

Pseudomonas aeruginosa (which are intrinsically resistant to ertapenem) in our study, it is not surprising that ertapenem was infrequently used as a first-line antibiotic. The risk of CDI after antimicrobial exposure increases if C difficile is resistant to that antibiotic.31 Higher rates of CDI after ertapenem use have been reported compared with other antibiotics,32 and some patients in our cohort had received ertapenem before extraintestinal CDI isolation. In these patients, it is difficult to understand outcomes attributable to CDI and factors that influence those outcomes. Treatment protocols varied widely, depending on the clinical picture, and several patients were critically ill. Indeed, we observed high short-term mortality (25%) related to the underlying infection a median of 16 days (range, 1-61 days) after C difficile isolation. Patients who survived the initial 2-month period after infection did well or died years later from unrelated causes. No recurrent cases of extraintestinal CDI were observed in our cohort even after a long median follow-up of 29 months, although a single case of recurrent monomicrobial abdominal abscess has been reported in the literature.33 Limitations of our study include its retrospective design; the incidence of extraintestinal CDI could not be ascertained. Strain typing and toxin production assays were not performed, so an analysis of C difficile virulence factors was not possible. The diagnosis of intestinal CDI was made by ICD-9-CM code rather than by positive stool assay results because we did not have information for all 10 years. Nevertheless, the ICD-9-CM code for CDI has been found to correlate well with the results of C difficile toxin assay.34,35 Extraintestinal CDI cases were not identified by ICD-9-CM codes but rather by laboratory testing. CONCLUSION Extraintestinal CDI is uncommon and predominately occurs in patients with recent surgical manipulation of the GI tract. An aging, immunocompromised population with high exposure to both antibiotics and PPIs and who undergo more frequent hospitalizations and surgical procedures is susceptible to extraintestinal CDI. Virulence factors responsible for extraintestinal infections remain to be defined. Management options for extraintestinal CDI Mayo Clin Proc. n November 2014;89(11):1525-1536 www.mayoclinicproceedings.org

n

often include combined therapy with antibiotics and surgery. Despite aggressive management, the mortality rate was high. ACKNOWLEDGMENT We thank Nicolynn Cole for her efforts in study data collection. Abbreviations and Acronyms: BSI = bloodstream infection; CDI = Clostridium difficile infection; GI = gastrointestinal; IBD = inflammatory bowel disease; ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Modification; MRSA = methicillin-resistant Staphylococcus aureus; PCR = polymerase chain reaction; PPI = proton pump inhibitor Correspondence: Sahil Khanna, MBBS, MS, Division of Gastroenterology and Hepatology, Department of Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905 ([email protected]).

REFERENCES 1. Bartlett JG, Chang TW, Gurwith M, Gorbach SL, Onderdonk AB. Antibiotic-associated pseudomembranous colitis due to toxin-producing. Clostridia. N Engl J Med. 1978; 298(10):531-534. 2. Kelly CP, LaMont JT. Clostridium difficileemore difficult than ever. N Engl J Med. 2008;359(18):1932-1940. 3. Khanna S, Pardi DS. The growing incidence and severity of Clostridium difficile infection in inpatient and outpatient settings. Expert Rev Gastroenterol Hepatol. 2010;4(4):409-416. 4. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med. 2005;353(23):2442-2449. 5. Magill SS, Edwards JR, Bamberg W, et al. Multistate pointprevalence survey of health care-associated infections. N Engl J Med. 2014;370(13):1198-1208. 6. Levett PN. Clostridium difficile in habitats other than the human gastro-intestinal tract. J Infect. 1986;12(3):253-263. 7. Feldman RJ, Kallich M, Weinstein MP. Bacteremia due to Clostridium difficile: case report and review of extraintestinal C. difficile infections. Clin Infect Dis. 1995;20(6):1560-1562. 8. Garcia-Lechuz JM, Hernangomez S, Juan RS, Pelaez T, Alcala L, Bouza E. Extra-intestinal infections caused by Clostridium difficile. Clin Microbiol Infect. 2001;7(8):453-457. 9. Lee NY, Huang YT, Hsueh PR, Ko WC. Clostridium difficile bacteremia, Taiwan. Emerg Infect Dis. 2010;16(8):1204-1210. 10. Libby DB, Bearman G. Bacteremia due to Clostridium difficilee review of the literature. Intl J Infect Dis. 2009;13(5): e305-e309. 11. Mattila E, Arkkila P, Mattila PS, Tarkka E, Tissari P, Anttila VJ. Extraintestinal Clostridium difficile infections. Clin Infect Dis. 2013;57(6):e148-e153. 12. Wolf LE, Gorbach SL, Granowitz EV. Extraintestinal Clostridium difficile: 10 years’ experience at a tertiary-care hospital. Mayo Clin Proc. 1998;73(10):943-947. 13. Sloan LM, Duresko BJ, Gustafson DR, Rosenblatt JE. Comparison of real-time PCR for detection of the tcdC gene with four toxin immunoassays and culture in diagnosis of Clostridium difficile infection. J Clin Microbiol. 2008;46(6):1996-2001. 14. Dineen SP, Bailey SH, Pham TH, Huerta S. Clostridium difficile enteritis: a report of two cases and systematic literature review. World J Gastrointest Surg. 2013;5(3):37-42.

http://dx.doi.org/10.1016/j.mayocp.2014.07.012

1535

MAYO CLINIC PROCEEDINGS

15. Navaneethan U, Giannella RA. Thinking beyond the colonsmall bowel involvement in Clostridium difficile infection. Gut Pathog. 2009;1(1):7. 16. Williams RN, Hemingway D, Miller AS. Enteral Clostridium difficile, an emerging cause for high-output ileostomy. J Clin Pathol. 2009;62(10):951-953. 17. Prati C, Bertolini E, Toussirot E, Wendling D. Reactive arthritis due to Clostridium difficile. Joint Bone Spine. 2010;77(2):190-192. 18. Zheng L, Citron DM, Genheimer CW, et al. Molecular characterization and antimicrobial susceptibilities of extra-intestinal Clostridium difficile isolates. Anaerobe. 2007;13(3-4):114-120. 19. Elliott B, Reed R, Chang BJ, Riley TV. Bacteremia with a large clostridial toxin-negative, binary toxin-positive strain of Clostridium difficile. Anaerobe. 2009;15(6):249-251. 20. Gravisse J, Barnaud G, Hanau-Bercot B, et al. Clostridium difficile brain empyema after prolonged intestinal carriage. J Clin Microbiol. 2003;41(1):509-511. 21. Stieglbauer KT, Gruber SA, Johnson S. Elevated serum antibody response to toxin A following splenic abscess due to Clostridium difficile. Clin Infect Dis. 1995;20(1):160-162. 22. McGill F, Fawley WN, Wilcox MH. Monomicrobial Clostridium difficile bacteraemias and relationship to gut infection. J Hosp Infect. 2011;77(2):170-171. 23. Sethi AK, Al-Nassir WN, Nerandzic MM, Bobulsky GS, Donskey CJ. Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection. Infect Control Hosp Epidemiol. 2010;31(1):21-27. 24. Roth A, Wiklund AE, Palsson AS, et al. Reducing blood culture contamination by a simple informational intervention. J Clin Microbiol. 2010;48(12):4552-4558. 25. Zerey M, Paton BL, Lincourt AE, Gersin KS, Kercher KW, Heniford BT. The burden of Clostridium difficile in surgical

1536

Mayo Clin Proc.

n

26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

patients in the United States. Surg Infect. 2007;8(6): 557-566. Daruwala C, Mercogliano G, Newman G, Ingerman MJ. Bacteremia due to Clostridium difficile: case report and review of the literature. Clin Med Case Rep. 2009;2:5-9. 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(5):431-455. Surawicz CM, Brandt LJ, Binion DG, et al. Guidelines for diagnosis, treatment, and prevention of Clostridium difficile infections. Am J Gastroenterol. 2013;108(4):478-498:quiz 499. Baines SD, O’Connor R, Freeman J, et al. Emergence of reduced susceptibility to metronidazole in Clostridium difficile. J Antimicrob Chemother. 2008;62(5):1046-1052. Huang H, Weintraub A, Fang H, Nord CE. Antimicrobial resistance in Clostridium difficile. Int J Antimicrob Agents. 2009;34(6): 516-522. Johnson S, Samore MH, Farrow KA, et al. Epidemics of diarrhea caused by a clindamycin-resistant strain of Clostridium difficile in four hospitals. N Eng J Med. 1999;341(22):1645-1651. Owens RC Jr, Donskey CJ, Gaynes RP, Loo VG, Muto CA. Antimicrobial-associated risk factors for Clostridium difficile infection. Clin Infect Dis. 2008;46(suppl 1):S19-S31. Bedimo R, Weinstein J. Recurrent extraintestinal Clostridium difficile infection. Am J Med. 2003;114(9):770-771. Dubberke ER, Reske KA, McDonald LC, Fraser VJ. ICD-9 codes and surveillance for Clostridium difficile-associated disease. Emerg Infect Dis. 2006;12(10):1576-1579. Dubberke ER, Butler AM, Nyazee HA, et al. The impact of ICD-9-CM code rank order on the estimated prevalence of Clostridium difficile infections. Clin Infect Dis. 2011;53(1):20-25.

November 2014;89(11):1525-1536

n

http://dx.doi.org/10.1016/j.mayocp.2014.07.012 www.mayoclinicproceedings.org