Infections post transplant Corynebacterium jeikeium bacteremia in bone marrow transplant patients with Hickman catheters

(2001) 27, 445 449 2001 Nature Publishing Group All rights reserved 0268 3369/01 $15.00 www.nature.com/bmt Infections post transplant Corynebacterium jeikeium bacteremia in bone marrow transplant patients with Hickman catheters CC Wang 1, D Mattson 2 and A Wald 1,3 Departments of 1 Medicine and 3 Epidemiology, University of Washington and 2 Program in Infectious Diseases, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Summary: Prior studies suggest that Corynebacterium jeikeium bacteremia in immunocompromised patients results in frequent morbidity that may be decreased by prompt removal of the indwelling catheter. To summarize recent experience, charts of 53 bone marrow transplant recipients with Hickman catheters and C. jeikeium bacteremia were reviewed. Forty-one patients were treated with vancomycin without catheter removal and 10 patients underwent catheter removal with subsequent vancomycin therapy. No patient in either group died with C. jeikeium bacteremia as the proximate cause. Salvage of the intravascular catheter was successful in 38 of 41 (93%) attempts. Three patients (7%) in the cathetersalvage group and one patient (10%) in the catheterremoval group experienced recurrent bacteremia. In both catheter-salvage and catheter-removal groups, median time to negative blood culture was 2 days. Thus, time to clearance of bacteremia and patient clinical outcome did not differ between treatment groups. In many patients with Hickman catheters, C. jeikeium bacteremia may be treated successfully with vancomycin and without removal of the catheter. (2001) 27, 445 449. Keywords: Corynebacterium jeikeium; catheter-indwelling; bacteremia Correspondence: Dr CC Wang, Harborview Medical Center, 325 Ninth Avenue, Box 359 909, Seattle, WA 98104 2499, USA Received 2 May 2000; accepted 8 December 2000 Corynebacterium jeikeium is an important nosocomial pathogen, particularly in neutropenic patients. Risk factors for C. jeikeium bacteremia in patients with neoplastic diseases include an indwelling intravascular catheter, prolonged neutropenia, and treatment with multiple antibiotics. 1,2 The overall mortality of C. jeikeium bacteremia in neutropenic hosts can be as high as 34%. 3 Of additional clinical concern is the fact that C. jeikeium is typically resistant to multiple antibiotics, and can be treated reliably only with vancomycin. 4,5 Indwelling intravascular catheters are one of the main- stays of the therapeutic approach in bone marrow transplant centers. However, infectious complications of indwelling catheters occur commonly, including exit site infections, tunnel infections, bacteremia, and fungemia. 6 Catheterrelated infections occur at an estimated rate of 10.3 infections per 1000 catheter days, and catheter-related bacteremias occur at an estimated rate of 5.3 bacteremias per 1000 catheter days. 7 Management of catheter-related bacteremia is controversial. For some organisms, catheter removal is recommended in addition to antibiotic treatment to eradicate infection. 8 However, replacement of indwelling catheters is associated with morbidity, and may not be necessary in all cases. 8 To develop guidelines for the management patients with Hickman catheters who develop positive blood cultures for C. jeikeium, we reviewed charts of patients with Hickman catheters and concurrent C. jeikeium bacteremia at a major bone marrow transplant center. Methods Study patients and case identification This retrospective study included bone marrow transplant patients at the Fred Hutchinson Cancer Research Center in Seattle between 1 January 1994 and 31 December 1998. Standardized procedures were used for conditioning, transplant, and management of infectious complications. 9 Blood was routinely drawn for culture from patients with temperatures greater than 38.5 C or with clinical evidence of bacteremia or soft tissue infection. Furthermore, any patient treated with high-dose corticosteroids (intravenous methylprednisolone, 2 mg/kg/day or more) had blood cultures drawn three times a week when an inpatient and once weekly when an outpatient. Bacteremia in patients with Hickman catheters was initially treated empirically with a regimen including intravenous vancomycin. In general, catheter removal was recommended for patients with evidence of local catheter insertion site infection or persistent bacteremia. The microbiology database was searched for all positive blood cultures for C. jeikeium during the study period. Patients were included in the study group if they had C. jeikeium bacteremia without signs of local catheter site infection. Culture data from all sites within a 6-day window

446 period (extending from 72 h before to 72 h after the positive blood culture) were also extracted. In addition, all blood culture data for the duration of the patients clinical care after the C. jeikeium bacteremic episode were reviewed. Recurrent bacteremia was defined as C. jeikeium bacteremia occurring after a documented negative blood culture in a patient treated with vancomycin. Demographic and clinical information were abstracted from a standardized clinical database and patient charts were reviewed to determine reasons for blood sampling, management of bacteremia, and patient outcome. Clinical decisions regarding removal of the catheter were made by the attending physician. If a decision was made to administer vancomycin for line-associated C. jeikeium bacteremia without removal of the catheter, the patient was considered to be in the catheter-salvage group. If the line was removed or dislodged within 10 days of documented bacteremia, the patient was considered to be in the catheter-removal group. Reason for death was abstracted from patient charts by two independent reviewers (DM, CW). Microbiology Blood was drawn into two standard bottles designed to contain 10 cc of blood per bottle (PML Microbiologicals, Tualatin, OR, USA) and an isolator tube containing a red blood cell lytic agent (DuPont, Wilmington, DE, USA). The aerobic bottle contained 40 ml of brain heart-infusion broth supplemented with 0.035% sodium polyanethol, carbon dioxide, and powdered activated charcoal. The anaerobic bottle contained 40 ml of tryptic soy broth with 0.035% sodium polyanethol, carbon dioxide, a reducing agent, and oxygen-free nitrogen. Both bottles were incubated for 5 days using the Organon Technika Bact-Alert (Organon Technika, Durham, NC, USA) system. The contents of the isolator tubes were centrifuged, and the sediment inoculated onto agar plates which were assessed daily for 5 days. C. jeikeium was identified by Gram-stained films of blood cultures demonstrating characteristic Gram-positive rods, and confirmed using the API Coryne biochemical identification kits (API-bioMérieux, Hazelwood, USA). Results From 1 January 1994 to 31 December 1998, 4139 (8.6%) of 48187 blood cultures were positive for microorganism growth. Ninety-three (2.2%) of the 4139 positive blood cultures were identified as C. jeikeium in this 5-year period. Fifty-three (2.6%) of 1963 patients admitted to the Fred Hutchinson Cancer Research Center during the study period had at least one blood culture positive for C. jeikeium without signs of local catheter-site infection. The 5-year study period was chosen to provide feasibility of chart review and an assessment of recent trends. Demographic and clinical characteristics of the 53 patients are presented in Table 1. Five patients (50%) in the catheter-removal group and 15 (36%) patients in the catheter-salvage group had at least one other organism isolated from the same blood culture positive for C. jeikeium. The proportion of mixed cultures Table 1 Demographic and clinical characteristics of 53 patients with positive blood cultures for Corynebacterium jeikeium Characteristic Number (%) a Age, years mean (range) 40 (3 74) Sex Male 41 (77) Female 12 (23) Diagnosis Acute lymphocytic leukemia 5 (9) Acute myelocytic leukemia 17 (32) Chronic myelogenous leukemia 10 (19) Lymphoma 7 (13) Myelodysplastic syndrome 8 (15) Multiple myeloma 3 (6) Breast cancer 1 (2) Neuroblastoma 1 (2) Severe combined immunodeficiency 1 (2) Type of transplant b Allogeneic 41 (77) Autologous 8 (15) High-dose steroids for GVHD c Yes 16 (39) No 25 (61) Reason for blood draw Fever 46 (87) Surveillance 6 (11) Unknown 1 (2) ANC d on day of culture 0 14 (27) 500 16 (30) 500 23 (43) Days after transplant median (range) e 17 (1 375) a Unless otherwise specified. b 49 of 53 patients underwent bone marrow transplant. c High-dose steroids defined as methylprednisolone 2 mg/kg/day in 41 patients undergoing allogeneic bone marrow transplant and at risk for graft-versus-host disease. d Absolute neutrophil count. e 45 patients with positive blood cultures on or after day of transplant. was not statistically different in the two groups (P = 0.6). Nineteen patients had urine cultures taken within 72 h of the positive blood culture for C. jeikeium; none of these cultures were positive. One of 17 patients with lower respiratory cultures obtained within the defined window period had C. jeikeium isolated. This patient died of pulmonary mucormycosis. All patients had negative oral biopsy cultures and fecal cultures for C. jeikeium. One of five skin cultures grew C. jeikeium; this patient had severe cutaneous manifestations of graft-versus-host disease. Outcome and management of the 53 patients are summarized in Figure 1. Forty-one patients received a median 10 days (range, 1 19) of vancomycin, without removal of the catheter. Ten patients were managed with removal of the catheter and a median 14 days (range, 10 30) of vancomycin. Two patients did not receive specific antibiotics for C. jeikeium. Of the 41 patients in the catheter-salvage group, 20 were discharged from hospital and 21 died. Three discharged patients experienced recurrent bacteremia. In one, recurrent bacteremia prompted replacement of the catheter. Interestingly, C. jeikeium bacteremia recurred in the setting of this second catheter, but responded to further therapy with vancomycin. In the other two patients, treatment with vanco-

53 patients with C. jeikeium bacteremia 447 41 patients treated with vancomycin without catheter-removal 17 discharged from FHCRC 10 died <14 days* 11 died >14 days* 3 recurrent bacteremia 10 patients treated with vancomycin and catheter-removal 5 discharged from FHCRC 4 died >14 days* 1 recurrent bacteremia 2 patients not treated with vancomycin 1 died at day 41 1 discharged from FHCRC *14 days from initial bacteremic episode; causes of death: polymicrobial sepsis (2), cytomegalovirus pneumonia (1), Fusarium pneumonia (1), Aspergillus pneumonia (1), Mucor pneumonia (1), non-bacterial endocarditis (1), veno-occlusive disease (1), EBV lymphoproliferative syndrome (1), and pulmonary hemorrhage (1). Figure 1 Management of 53 patients with C. jeikeium bacteremia and without evidence of local infection (Fred Hutchinson Cancer Research Center 1994 1998). mycin was repeated with good clinical response and negative follow-up cultures. None of the patients developed complications of C. jeikeium infection, such as endocarditis or meningitis. Of the 21 patients in the catheter-salvage group who died, 10 did so within 14 days of C. jeikeium bacteremia and 11 died a median of 62 (range, 17 782) days after the bacteremia. Reasons for death in patients who died early are listed in Figure 1. Of seven patients who died of infection, all had potentially fatal pathogens isolated other than C. jeikeium bacteremia, and two patients with clinical sepsis had documented negative blood cultures for C. jeikeium bacteremia prior to death. Therefore, C. jeikeium infection appears to be an unlikely cause of death in these patients. Thus, treatment of C. jeikeium bacteremia with vancomycin and without catheter removal was successful in 28 of 41 (68%) attempts and was possibly successful in an additional 10 (25%) attempts, although early death precluded definitive assessment of catheter outcome. In 10 patients, catheters were removed within 14 days of the positive blood culture for C. jeikeium. Reasons for removal included catheter malfunction (n = 3), and simultaneous bloodstream infection with Candida species or coagulase-negative Staphylococci (n = 6). In only one instance was C. jeikeium bacteremia documented as the reason for catheter removal. One patient in the catheterremoval group developed recurrent bacteremia with the second Hickman catheter which was successfully treated with vancomycin. None of the patients in the catheterremoval group died within 14 days of bacteremia. Four patients died a median of 47 days (range, 24 82) after C. jeikeium bacteremia of transplant-related complications. Duration of bacteremia could not be precisely defined for most patients because blood cultures were not obtained on consecutive days after the initial positive culture for C. jeikeium. Eight patients in the catheter-salvage group and three in the catheter-removal group had documented clearance of bacteremia in 1 day. Sixteen of 41 patients in the catheter-salvage group and five of 10 patients in the catheter-removal group had multiple blood cultures positive for C. jeikeium prior to completing a course of vancomycin. Thirty-five patients in the catheter-salvage group and nine in the catheter-removal group had follow-up negative blood cultures without recurrent bacteremia. In both the catheterremoval and catheter-salvage group, the median time to negative blood cultures was two days (range, 1 10). Overall, recurrent bacteremia developed in four patients a median of 19 days (range, 10 54 days) following the initial isolate. No differences were noted in the clinical characteristics of patients with and without recurrent bacteremia, although small numbers limited the usefulness of such comparisons (data not shown). Discussion In this study of 53 patients with C. jeikeium bacteremia, salvage of Hickman catheters with vancomycin therapy was successful in 38 (93%) of 41 attempts. Only three (7%) patients with attempted catheter salvage had recurrent bacteremia and in two of these patients, a repeat course of vancomycin resolved the infection. In both groups the median time to negative blood cultures was 2 days. From these data, C. jeikeium bacteremia appears to clear quickly and recurs infrequently with vancomycin therapy, even when the intravascular catheter is maintained. C. jeikeium bacteremia was not the main cause of death in any patient. A potentially fatal infection distinct from C. jeikeium bacteremia was documented in the seven patients in the catheter-salvage group who died within 14 days of bacteremia. In the presence of multiple infections, the specific contribution of C. jeikeium bacteremia to patient death was difficult to establish. Since the use of indwelling catheters in cancer patients became routine in the early 1980s, many studies have addressed the management of catheter-related bacteremia in oncology patients. Such studies emphasize the importance of considering the specific organism involved in a catheter-related infection when making therapeutic

448 decisions. For Gram-negative catheter infections, particularly P. aeruginosa, catheter removal is usually required. 10 Similarly, catheter infections with S. aureus, 11 Bacillus species, 12 and fungi 13 may respond poorly to antibiotic therapy alone and often require catheter removal for eradication of infection. In contrast, many authors have reported successful treatment of coagulase-negative Staphylococci catheter infections with antibiotics, obviating the need to remove the catheter. 14,15 Few studies exist addressing the appropriate management of C. jeikeium catheter-associated infection. In a 1979 study of 32 cases of C. jeikeium bacteremia, Stamm et al 2 reported that of nine patients who retained their catheters and received antibiotics as treatment for their bacteremia, five died. In contrast, all eight patients who had their Hickman catheters removed survived. Other authors have reported the characteristic slime production of this organism, possibly interfering with penetration of antibiotics into a site of infection and reducing efficacy of antimicrobial drugs. 16 As a result, some experts recommend that catheters associated with C. jeikeium bacteremia be removed. 8 This study is limited by the sole use of catheter-drawn cultures to diagnose line-related infection. The Centers for Disease Control definition of catheter related bloodstream infection requires positive catheter-segment culture or paired central peripheral quantitative cultures demonstrating five-fold greater growth from the centrally obtained culture compared to the peripherally obtained culture. 17 However, in patients undergoing bone marrow transplant, peripheral cultures are rarely done because of underlying thrombocytopenia, risk of introducing infection, and considerations of patient discomfort. None of our patients had peripheral cultures drawn. However, catheter-related bacteremia seemed probable since most patients were febrile at the time of culture and had negative cultures from the lower respiratory tract, the urinary tract, and other sites. Although C. jeikeium has been described as the causative agent in pneumonia, 4 osteomyelitis, 18 and otitis media, 19 it is most commonly associated with the presence of an indwelling catheter. 20 The evaluation of patients in our study reflects the clinical setting of bone marrow transplantation. Our study demonstrates that catheter-related C. jeikeium bacteremia in the absence of local infection may be successfully treated with vancomycin and without removal of the catheter in most patients. Recurrent bacteremia may also be successfully treated. Patient outcome and time to negative blood culture did not differ in the catheter-removal and catheter-salvage group, and no patient suffered severe sequelae of the infection. Thus, in this population of severely immunocompromised bone marrow transplant patients, C. jeikeium bacteremia appeared to be associated with low morbidity. Treatment of catheter-related C. jeikeium infection with vancomycin can result in catheter salvage, sparing patients repeated catheter insertions. Furthermore, while patient numbers were too small to provide definitive guidelines, the data suggest that the vast majority of catheterrelated infections treated with vancomycin alone resolve within 3 days. C. jeikeium bacteremia persisting beyond 3 days may be an indication for catheter removal. Acknowledgements We would like to thank Lawrence Corey, Ajit Limaye, Thomas Novicki, and Walter Stamm for their guidance in the preparation of manuscript. We would like to thank Christopher Dorling for his assistance in retrieving information from the patient database, and Caroline McKallor and Madeline Keston for their expertise with the microbiology database. We are also grateful to Bruce Ulness for the information he provided from the microbiology laboratory. This work was supported in part by grants from the National Institutes of Health (No. PO1-AI-33118); Center for AIDS Research (NIAID AI27757); University of Washington STD Cooperative Research Center (NIAID AI31448). Chia Wang was a fellow in the UW STD/AIDS Postdoctoral and Predoctoral Research Training Program (NIAIDAI07140). References 1 Rozdzinski E, Kern W, Schmeiser T, Kurrle E. Corynebacterium jeikeium bacteremia at a tertiary care center. Infection 1991; 19: 201 204. 2 Stamm WE, Tompkins LS, Wagner KF et al. Infection due to Corynebacterium species in marrow transplant patients. Ann Intern Med 1979; 91: 167 173. 3 van der Lelie H, Leverstein Van Hall M, Mertens M et al. Corynebacterium CDC group JK (Corynebacterium jeikeium) sepsis in haematological patients: a report of three cases and a systematic literature review. Scand J Infect Dis 1995; 27: 581 584. 4 Brown AE. Other Corynebacteria and Rhodococcus. In: Mandell GL, Bennett JE, Dolin R (eds). Principles and Practice of Infectious Diseases, 5th edn. Churchill-Livingstone: New York, 2000, pp 2198 2208. 5 Soriano F, Zapardiel J, Nieto E. Antimicrobial susceptibilities of Corynebacterium species and other non-spore-forming gram-positive bacilli to 18 antimicrobial agents. Antimicrob Agents Chemother 1995; 39: 208 214. 6 Raad II, Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect Dis 1992; 15: 197 208. 7 Elishoov H, Or R, Strauss N, Engelhard D. Nosocomial colonization, septicemia, and Hickman/Broviac catheter-related infections in bone marrow transplant recipients. A 5-year prospective study. Medicine (Baltimore) 1998; 77: 83 101. 8 Greene JN. Catheter-related complications of cancer therapy. Infect Dis Clin North Am 1996; 10: 255 295. 9 Storb R, Leisenring W, Deeg HJ et al. Long-term follow-up of a randomized trial of graft-versus-host disease prevention by methotrexate/cyclosporine versus methotrexate alone in patients given marrow grafts for severe aplastic anemia (letter). Blood 1994; 83: 2749 2750. 10 Benezra D, Kiehn TE, Gold JW et al. Prospective study of infections in indwelling central venous catheters using quantitative blood cultures. Am J Med 1988; 85: 495 498. 11 Raad II, Sabbagh MF. Optimal duration of therapy for catheter-related Staphylococcus aureus bacteremia: a study of 55 cases and review. Clin Infect Dis 1992; 14: 75 82. 12 Blue SR, Singh VR, Saubolle MA. Bacillus licheniformis bacteremia: five cases associated with indwelling central venous catheters. Clin Infect Dis 1995; 20: 629 633. 13 Lecciones JA, Lee JW, Navarro EE et al. Vascular catheterassociated fungemia in patients with cancer: analysis of 155 episodes. Clin Infect Dis 1992; 14: 875 883. 14 Press OW, Ramsey PG, Larson EB et al. Hickman catheter infections in patients with malignancies. Medicine (Baltimore) 1984; 63: 189 200.

15 Clarke DE, Raffin TA. Infectious complications of indwelling long-term central venous catheters. Chest 1990; 97: 966 972. 16 Bayston R, Compton C, Richards K. Production of extracellular slime by coryneforms colonizing hydrocephalus shunts. J Clin Microbiol 1994; 32: 1705 1709. 17 Pearson ML. Guideline for prevention of intravascular devicerelated infections. Part I. Intravascular device-related infections: an overview. The Hospital Infection Control Practices Advisory Committee. Am J Infect Control 1996; 24: 262 277. 18 Munian MA, Martinez-Martinez L, Suarez AI et al. Corynebacterium jeikeium osteomyelitis successfully treated with teicoplanin (letter). J Infect 1997; 35: 325 326. 19 de Miguel-Martinez I, Ramos-Macias A, Martin-Sanchez AM. Otitis media due to Corynebacterium jeikeium. Eur J Clin Microbiol Infect Dis 1999; 18: 231 232. 20 Riebel W, Frantz N, Adelstein D, Spagnuolo PJ. Corynebacterium JK: a cause of nosocomial device-related infection. Rev Infect Dis 1986; 8: 42 49. 449