Jpn. J. Infect. Dis., 64, 520-524, 2011 Short Communication Clinical Characteristics, Antimicrobial Susceptibilities, andoutcomesofpatientswithchryseobacterium indologenes Bacteremia in an Intensive Care Unit Deng-Wei Chou 1,2 *, Shu-Ling Wu 2, Chao-Tai Lee 3, Fan-Ting Tai 4,andWen-LiangYu 5,6 1 Department of Critical Care Medicine, 3 Department of Clinical Laboratory, and 4 Committee of Infection Control, Tainan Municipal Hospital, Tainan; 2 Department of Nursing, Chung-Hwa University of Medical Technology, Tainan; 5 Department of Intensive Care Medicine, Chi-Mei Medical Center, Tainan; and 6 Department of Medicine, Taipei Medical University, Taipei, Taiwan (Received June 8, 2011. Accepted September 2, 201 SUMMARY: Ten patients with intensive care unit (ICU)-acquired Chryseobacterium indologenes bacteremia between January 2004 and December 2008 were studied. The primary site of infection was unknown for 80z of the cases. The known primary sites of infection were empyema (10z) and catheter-relatedbacteremia (10z). Eight patients (80z) had polymicrobialbacteremia, spent morethan 21 days in the ICU, and received more than 14 days of broad-spectrum antibiotic therapy prior to the onset of C. indologenes bacteremia. All isolates were 100z susceptible to minocycline and trimethoprim/sulfamethoxazole. Vancomycin, imipenem, piperacillin/tazobactam, ciprofloxacin, and levofloxacin exhibited 0z, 10z, 20z, 30z, and30z, respectively, susceptibility against this pathogen. All isolates were 100z resistant to ceftazidime, cefepime, meropenem, piperacillin, and amikacin. The 14-day mortality rate was 40z. Our findings suggest that this pathogen should be included among the causes of ICU-acquired bacteremia, especially in patients with a prolonged stay in an ICU or who had received long-term broad-spectrum antibiotic therapy. Extended-spectrum penicillins, third- and fourth-generation cephalosporins, and quinolones had very little or no effect against this pathogen. Therefore, choosing an appropriate antibiotic therapy for this pathogen is very difficult. Chryseobacterium indologenes is a non-motile, oxidase-positive, indole-positive, and glucose-nonfermentative Gram-negative rod that is widely distributed in nature (. The microorganism had not been reported as a cause of bacteremia in humans until 1996. Hsueh et al. described 12 patients with C. indologenes bacteremia over a 3-year period (2). Since that report, there have been relatively few reports of C. indologenes bacteremia in humans (3 10), and most of them have occurred in Taiwan (2,5,11,12). Clinical manifestations include nosocomial pneumonia, biliary tract infection, peritonitis, surgical wound infection (2), intravascular catheterrelated bacteremia (3,9), cellulitis (4), and primary bacteremia (2,5 8,10). C. indologenes infections that are associated with various indwelling devices have been reported (12). Although C. indologenes is a relatively uncommon human pathogen, we have observed an increasing incidence of C. indologenes bacteremia in our intensive care unit (ICU) over the last few years (0 in 2003, 0.079 cases per 1,000 patient-days in 2004, 0.155 in 2005, 0.154 in 2006, 0.154 in 2007, and 0.227 in 2008). The incidence of bacteremia caused by other pathogens was 4.4 cases per 1,000 patient-days in 2003, *Corresponding author: Mailing address: Department of Critical Care Medicine, Tainan Municipal Hospital, No. 670, Chung-Te Road, East District, Tainan, Taiwan 701. Tel: +886-6-2609926 ext. 21396, Fax: +886-6-2606351, E-mail: choudw@gmail.com 4.3 in 2004, 4.6 in 2005, 5.1 in 2006, 6.8 in 2007, and 5.6 in 2008. However, thorough examination of existing literature revealed no publications that specifically focused on ICU-acquired C. indologenes bacteremia. Here, we elucidate the clinical characteristics, antimicrobial susceptibilities, and outcomes of patients with ICU-acquired C. indologenes bacteremia. This study was conducted at Tainan Municipal Hospital, which is a referral teaching hospital located in southern Taiwan. The mixed medical-surgical ICU is a 38-bed unit with individual rooms and about 2,000 annual admissions. Admissions are two-thirds medical and one-third surgical. All patients hospitalized in the ICU during the study period, which was between January 2004 and December 2008, with proven positive blood cultures for C. indologenes were included. Collected patient data included age, gender, any immunocompromised diseases, condition while in the ICU, use of indwelling devices (including central venous catheter, invasive mechanical ventilation, and urinary catheter), length of stay in the ICU prior to the onset of infection, other sites of C. indologenes isolation, primary site of infection, polymicrobial bacteremia, antibiotic therapy, and hospital outcome. ICU-acquired C. indologenes bacteremia was defined when a patient had C. indologenes bacteremia after staying in the ICU for more than 48 h. Catheter-related bacteremia caused by C. indologenes was defined when a patient had a vascular catheter in place for longer than 48 h and who experienced an unexplained fever. In ad- 520
dition, at least one peripheral blood culture and one culture of the catheter tip had to be positive for C. indologenes. Clinical diagnoses of sepsis and septic shock were based upon the criteria specified by the American College of Chest Physicians and the Society of Critical Care Medicine (13). Polymicrobial bacteremia was defined as the identification of more than one microorganism from a single set of blood cultures (14). Appropriate antibiotic therapy was defined as one of the prescribed antibiotics with minimal inhibitory concentrations (MICs) below the corresponding breakpoint for non- Enterobacteriaceae in vitro. Ten isolates of C. indologenes were identified by the commercial API 20NE identification system (biomáerieux, Marcy-l'Etoile, France) during the study period. Results were interpreted using the API LAB ID computer software. The positive blood culture isolates were stored at -709C and subcultured before further confirmation. All isolates were confirmed as C. indologenes by the BD Phoenix Automated Microbiology System (software version 4.01; Becton Dickinson, Sparks, Md., USA). The individual MICs of the antimicrobial agents for C. indologenes were determined by the Phoenix system for antimicrobial susceptibility testing. The tested antimicrobial agents from the Phoenix panels were gentamicin, amikacin, ciprofloxacin, levofloxacin, piperacillin, piperacillin/tazobactam, ceftriaxone, ceftazidime, cefepime, imipenem, meropenem, and trimethoprim/sulfamethoxazole (TMP/SMZ). Vancomycin and minocycline were tested with Etest strips (AB Biodisk, Solna, Sweden). The percentage of isolates that were susceptible to these antimicrobial agents was determined by applying the Clinical and Laboratory Standards Institute susceptibility breakpoints used for non- Enterobacteriaceae (15). Our study included 10 patients (1 patient in 2004, 2 in 2005, 2 in 2006, 2 in 2007, and 3 in 2008). C. indologenes accounted for 2.95z of the total number of ICU-acquired bloodstream infections over the 5-year study period. Table 1 shows the clinical characteristics, antibiotic therapy, and outcome of these patients. Five patients were men (50z), and 5 were women (50z). The mean age was 71.1 years (range, 48 84 years). Six patients (60z) had underlying immunocompromised diseases, 4 patients (40z) had malignancies, 3 (30z) had diabetes mellitus, and 2 (20z) had cirrhosis. The primary site of C. indologenes infection was unknown in 80z of the cases. The known primary sites of infection were empyema (10z) and catheter-related bacteremia (10z). Eight patients (80z) had polymicrobial bacteremia (2 patients with Acinetobacter baumannii, 2 patients with Klebsiella pneumoniae, 1 patient with Chryseobacterium meningosepticum, 1 patient with Enterococcus faecalis, 1patientwithEnterobacter cloacae, 1patientwithBurkholderia cepacia, 1patientwith Staphylococcus aureus, and1patientwithcandida albicans). Eight patients (80z) hadstayedintheicufor more than 21 days. Eight patients (80z) had received more than 14 days of broad-spectrum antibiotics prior to the onset of C. indologenes bacteremia. The MIC ranges and the MICs at which 50z and 90z for the 10 isolates of C. indologenes are shown in Table 2. All isolates were 100z susceptible to minocycline and TMP/SMZ. Vancomycin, imipenem, piperacillin/tazobactam, ciprofloxacin, and levofloxacin exhibited 0z, 10z, 20z, 30z, and30z, respectively, susceptibility against this pathogen. All isolates were 100z resistant to ceftazidime, ceftriaxone, cefepime, meropenem, piperacillin, gentamicin, and amikacin. No patients had received appropriate antibiotic therapy prior to C. indologenes isolation. All patients presented with sepsis. Four patients (40z) developed septic shock and died within 6 days. Bacteremia was determined to be the cause of death in all of these cases. The 14-day mortality rate in our study was 40z. There were no overlapping dates of hospitalization among our patients, and they were treated by different health care workers. Therefore, all patients had sporadic infections. Although they had all received invasive mechanical ventilation, only one patient had C. indologenes isolated from the pleural effusion. None of the C. indologenes was isolated from the tracheal aspirate or bronchoalveolar lavage. The SENTRY Antimicrobial Surveillance Program showed that the quinolones, including levofloxacin (100z susceptible) and ciprofloxacin (85.0z susceptible), had the highest potency against C. indologenes. Piperacillin/tazobactam (90.0z susceptible), piperacillin (85.0z susceptible), ceftazidime (85.0z susceptible), and cefepime (85.0z susceptible) were the most active agents among the b-lactams (16). Here, C. indologenes was uniformly resistant to third- and fourthgeneration cephalosporins, piperacillin, meropenem, and aminoglycosides. Piperacillin/tazobactam, imipenem, ciprofloxacin, and levofloxacin exhibited little effect against this pathogen. Our findings confirm that the isolates collected from patients hospitalized in the ICU had much lower susceptibilities to various antimicrobial agents than that suggested by previous investigations. Most of our patients had received more than 14 days of broad-spectrum antibiotics, which might have led to selective pressure for resistance in this pathogen. As indicated here, it is very difficult to choose an appropriate antibiotic therapy for this pathogen. Although minocycline and TMP/SMZ had the highest potency against this pathogen in vitro, clinical treatment of C. indologenes bacteremia with minocycline and TMP/SMZ awaits further research. In addition, a high proportion of our patients had polymicrobial bacteremia, including Gram-negative bacilli, Gram-positive cocci, or Candida. The above results severely complicate the choice of an effective antibiotic therapy for ICU-acquired C. indologenes bacteremia. Four of our patients who were free from underlying immunocompromised diseases (patients 1, 3, 5, and 7) recovered well, despite the initial administration of antibiotics that lacked activity against C. indologenes. This outcome suggests a low virulence that is relatively ineffective in immunocompetent patients. In contrast, four immunocompromised patients (patients 2, 4, 6, and 10) who had not received appropriate antibiotic therapy died of septic shock within 6 days. All of them also had polymicrobial bacteremia. Thus, it is not surprising that this pathogen resulted in a high mortality rate in immunocompromised patients. It is well known that an infected patient's immune system plays the most 521
Table 1. Summary of 10 patients with Chryseobacterium indologenes bacteremia Case Age (y)/ gender Immunocompromised disease(s) Condition(s) in ICU Indwelling devices Onset (day) Primary site of infection Polymicrobial bacteremia Antimicrobial agent(s) before C. indologenes isolation Antimicrobial agent(s) after C. indologenes isolation Outcome Day of mortality 3) 522 1 83/F None Pneumonia IMV, CVC Ã21 Unknown Enterococcus faecalis Cefmetazole+ Imipenem+ Recovered Klebsiella pneumoniae amikacin 2 80/F Colon cancer, PMV 2) IMV, CVC À21 Unknown Enterobacter cloacae Imipenem Imipenem+ Died Day 5 DM Burkholderia cepacia levofloxacin 3 82/M None Stroke, PMV IMV, À21 Unknown Acinetobacter baumannii Imipenem+ Imipenem Recovered 4 72/F Cirrhosis SBP, PMV IMV, CVC À21 Unknown Klebsiella pneumoniae Ceftazidime+ Imipenem+ Died Day 5 levofloxacin 5 84/F None Pneumonia, PMV IMV, CVC À21 Catheter-related None Levofloxacin Levofloxacin Recovered bacteremia 6 63/F Gastric cancer Pneumonia, PMV IMV, CVC, À21 Unknown Chryseobacterium Meropenem Meropenem Died Day 3 meningosepticum 7 48/M None Head trauma IMV, CVC, Ã21 Unknown Acinetobacter baumannii Imipenem+ Imipenem Recovered 8 69/M Colon cancer, Pneumonia, PMV IMV, CVC À21 Empyema Candida albicans Imipenem+ Imipenem+ Recovered DM fluconazole 9 55/M DM Stroke, PMV IMV, À21 Unknown None Imipenem+ Imipenem Recovered 10 75/M Cirrhosis, HCC SBP, PMV IMV, CVC, À21 Unknown Staphylococcus aureus Imipenem+ Imipenem+ + levofloxacin Died Day 6 : Length of stay in intensive care unit prior to the onset of C. indologenes bacteremia. 2) : PMV was defined as ventilator support for more than 21 days. 3) : Day of mortality after the onset of C. indologenes bacteremia. ICU, intensive care unit; DM, diabetes mellitus; HCC, hepatocellular carcinoma; PMV, prolonged mechanical ventilation; SBP, spontaneous bacterial peritonitis; IMV, invasive mechanical ventilation; CVC, central venous catheter;, urinary catheter.
Table 2. In vitro antimicrobial susceptibility of various antibiotics against 10 clinical isolates of Chryseobacterium indologenes Antimicrobial agent MIC range (mg/ml) MIC 50 MIC 90 Susceptibility breakpoint (mg/ml) z Susceptible z Resistant Ceftazidime À16 À16 À16 Ã8 0 100 Ceftriaxone À32 À32 À32 Ã8 0 100 Cefepime 8 À16 À16 À16 Ã8 0 100 Imipenem 2 À8 À8 À8 Ã4 10 90 Meropenem 8 À8 À8 À8 Ã4 0 100 Piperacillin À64 À64 À64 Ã16 0 100 Piperacillin/tazobactam Ã4/4 À64/4 À64/4 À64/4 Ã16/4 20 50 Gentamicin À8 À8 À8 Ã4 0 100 Amikacin À32 À32 À32 Ã16 0 100 Ciprofloxacin Ã0.5 À2 À2 À2 Ã1 30 70 Levofloxacin Ã1 À4 4 À4 Ã2 30 50 Minocycline 0.38 2 1 1.5 Ã4 100 0 TMP/SMZ Ã0.5/9.5 1/19 Ã0.5/9.5 Ã0.5/9.5 Ã2/38 100 0 Vancomycin 8 128 12 32 Ã2 0 40 :MIC 50 and MIC 90, the MICs at which 50z and 90z of the isolates were inhibited. TMP/SMZ, trimethoprim/sulfamethoxazole. important role in predicting survival rate. In addition, the condition of polymicrobial bacteremia can contribute to an increased mortality rate of immunocompromised patients. Our study has at least two notable limitations. First, the patient sample size was relatively small by statistical standards. Therefore, we cannot identify the risk factors for ICU-acquired C. indologenes bacteremia with statistical significance. The results of the antimicrobial susceptibilities of various antibiotics against C. indologenes were performed in only one institution. More case numbers in other hospitals are needed to generalize our findings with greater confidence (although C. indologenes bacteremia was, and still is, very rare). Second, patient data was collected retrospectively. A culture of the catheter tip had not been performed for three patients (patients 2, 4, and 6). Therefore, the case numbers of catheter-related bacteremia may have been underestimated. The basic findings of our study are as follows. First, this pathogen should be included among the potential causes of ICU-acquired bacteremia, especially in patients who had stayed in the ICU for longer than 21 days, or who received broad-spectrum antibiotic therapy for more than 14 days. Second, extended-spectrum penicillins, third- and fourth-generation cephalosporins, and quinolones had very little effect against this pathogen. In addition, because this pathogen occurred simultaneously with polymicrobial bacteremia, choosing an appropriate antibiotic therapy for this pathogen is greatly complicated. More epidemiological studies are needed to elucidate the acquisition of C. indologenes bacteremia in the ICU and to develop efficacious preventive measures. Acknowledgments We gratefully thank the editorial assistance of Marc Grenier, MEG International English Center, Tainan, Taiwan. Conflict of interest None to declare. REFERENCES 1. Schreckenberger, P.C., Daneshvar, M.I., Weyant, R.S., et al. (2003): Acinetobacter, Achromobacter, Chryseobacterium, Moraxella, and other nonfermentative Gram-negative rods. p. 749 779. In Murray, P.R., Baron, E.J., Jorgensen, J.H., et al., Manual of Clinical Microbiology. 8th ed. Washington, D.C. 2. Hsueh, P.R., Hsiue, T.R., Wu, J.J., et al. (1996): Flavobacterium indologenes bacteremia: clinical and microbiological characteristics. Clin. Infect. Dis., 23, 550 555. 3. Nulens, E., Bussels, B., Bols, A., et al. (200: Recurrent bacteremia by Chryseobacterium indologenes in an oncology patient with a totally implanted intravascular device. Clin. Microbiol. Infect., 7, 391 393. 4. Green, B.T. and Nolan, P.E. (200: Cellulitis and bacteraemia due to Chryseobacterium indologenes. J. Infect., 42, 219 220. 5. Lin, J.T., Wang, W.S., Yen, C.C., et al. (2003): Chryseobacterium indologenes bacteremia in a bone marrow transplant recipient with chronic graft-versus-host disease. Scand. J. Infect. Dis., 35, 882 883. 6. Christakis, G.B., Perlorentzou, S.P., Chalkiopoulou, I., et al. (2005): Chryseobacterium indologenes non-catheter-related bacteremia in a patient with a solid tumor. J. Clin. Microbiol., 43, 2021 2023. 7. Cascio,A.,Stassi,G.,Costa,G.B.,etal.(2005):Chryseobacterium indologenes bacteraemia in a diabetic child. J. Med. Microbiol., 54, 677 680. 8. Ray, P., Sharma, K. and Gautam, V. (2005): Chryseobacterium indologenes bacteremia: a case report. J. Commun. Dis., 37, 259 260. 9. Akay, M., Gunduz, E. and Gulbas, Z. (2006): Catheter-related bacteremia due to Chryseobacterium indologenes in a bone marrow transplant recipient. Bone Marrow Transplant., 37, 435 436. 10. Bayraktar, M.R., Aktas, E., Ersoy, Y., et al. (2007): Postoperative Chryseobacterium indologenes bloodstream infection caused by contamination of distillate water. Infect. Control Hosp. Epidemiol., 28, 368 369. 11. Hsueh, P.R., Teng, L.J., Yang, P.C., et al. (1997): Increasing incidence of nosocomial Chryseobacterium indologenes infections in Taiwan. Eur. J. Clin. Microbiol. Infect. Dis., 16, 568 574. 12. Hsueh, P.R., Teng, L.J., Ho, S.W., et al. (1996): Clinical and microbiological characteristics of Flavobacterium indologenes infections associated with indwelling devices. J. Clin. Microbiol., 34, 1908 1913. 13. Bone, R.C., Balk, R.A., Cerra, F.B., et al. (1992): Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest, 101, 1644 1655. 14. Roberts, F.J. (1989): Definition of polymicrobial bacteremia. 523
Rev. Infect. Dis., 11, 1029 1030. 15. Clinical and Laboratory Standards Institute (2009): Performance standards for antimicrobial susceptibility testing. 19th information supplement, 29, M100-S19. Clinical and Laboratory Standards Institute. Wayne, Pa. 16. Kirby, J.T., Sader, H.S., Walsh, T.R., et al. (2004): Antimicrobial susceptibility and epidemiology of a worldwide collection of Chryseobacterium spp.: report from the SENTRY Antimicrobial Surveillance Program (1997 200. J. Clin. Microbiol., 42, 445 448. 524