ABSTRACT. Background: The antibiotic-lock technique has been suggested to treat catheter-related sepsis in parenteral nutrition and to avoid catheter removal. Methods: To determine the incidence of catheter-related sepsis, the bacteria involved, and the efficacy of the antibioticlock technique with teicoplanin, all patients (n = 263) undergoing parenteral nutrition from January 1997 to December 1999 in one center, with patients at the hospital (n = 209) and at home (54) were retrospectively studied. The antibiotic-lock technique with teicoplanin was systematically used in all suspected infections and maintained in staphylococcus epidermidis (SE) infections. Results: A total of 21 of 263 patients had 34 infections (0.11/patient per year): 12 of 209 hospitalized and 9 of 54 home patients. A total of 10 of 34 infections were due to non-SE, and the catheter was immediately removed. The
other 24 of 34 infections were due to SE; in 5 of 24, the catheter was removed after 48 hours of the antibiotic-lock technique because of persistent fever or thrombosis. A total of 5 of 12 patients had 2 or more infections on the same catheter. The antibiotic-lock technique prevented short-term catheter removal in these cases, but a second infection occurred within a median of 50 days. In 4 of 5 cases, a third infection occurred in a mean delay of 90 days so that the catheter was removed. In 3 of 5 patients, bacteria was analyzed with pulsed field gel electrophoresis, which showed that recurrent infections were due to the same strain in all cases. Conclusions: In this study, the incidence of catheter-related sepsis was low and mostly related to SE. Our results do not support the use of the teicoplanin antibiotic-lock technique in SE infections. (Journal of Parenteral and Enteral Nutrition 26:109-- 113, 2002)
Catheter-related sepsis is a serious complication of long-term central venous access, particularly for parenteral nutrition (PN). The true incidence of catheterrelated sepsis and the prevalence of staphylococcus epidermidis (SE) in catheter-related sepsis remain controversial. It has been suggested that the antibioticlock technique (lock) might cure catheter-related sepsis and prevent catheter removal.1-3 However, the efficacy and safety of lock are controversial.4,5 Vancomycin and amikacin are usually used for lock. Recently, teicoplanin has been suggested.3,6 The aim of this study was to determine the incidence of catheter-related sepsis, the prevalence of SE infection, and the efficacy of teicoplanin-lock in SE-catheter-related sepsis during PN.
The finding that the antibiotic-lock technique with teicoplanin is not highly effective in the treatment of catheter-related sepsis related to SE does not support the use of this drug for the antibiotic-lock technique. Moreover, these results raise questions about the efficacy of the technique in catheter-related sepsis regardless of the antibiotic used.
MATERIALS AND METHODS
This retrospective study was performed from January 1997 to December 1999 in the Hepatology-Gastro-- enterology and Nutrition Unit in Hospital Charles Nicolle, Rouen, France. A total of 263 patients received nocturnal cyclic PN through a subclavian subcutaneously tunneled catheter. All patients had gastrointestinal disease. They received PN either at the hospital (n = 209) or at home (n = 54).
Diagnosis of Catheter-Related Sepsis
Diagnosis of catheter-related sepsis was made on semiquantitative drawn-back and peripheral blood cultures for aerobic, anaerobic, and fungi. A catheter culture was performed if the catheter was removed. Criteria for catheter-related sepsis were positive cultures from peripheral and drawn-back blood with the same infectious organism and a colony count from drawn-- back blood at least 5 times greater than in cultures from peripheral blood or a positive culture in the catheter after removal.7,8
Management of Catheter-Related Sepsis
During the study period, all patients were treated according to the same written protocol that included prospectively all suspected and proven cases of catheter-related sepsis in our unit. Suspected cases were first treated with teicoplanin-lock until culture results were available. PN was stopped for 48 hours and started again in the absence of fever after a 48-hour lock. Teicoplanin-lock was prepared as a 2-mL saline solution at a concentration of 5 mg/2 mL. It was instilled every morning into the catheter lumen and rinsed after 12 hours. Nocturnal PN was then infused.
After blood culture results, the therapeutic strategy depended on the nature of the isolated infectious organism. In case of non-SE-catheter-related sepsis, the catheter was immediately removed. In case of SE-catheter-related sepsis, if fever persisted after a 48-hour lock, the catheter was removed and lock was considered to have failed. For all other SE-related cases, the lock was used for a 14-day period. In cases of bacteriemia, systemic antibiotics adapted to the isolated infectious organism were added. After 3 infections with SE in the same patient on the same catheter, the catheter was systematically removed. Recurrent infections were infections occurring on the same catheter after a first infection regardless of the delay between the infectious episodes. Five of 12 patients with SE-catheter-related sepsis had 2 or more infections on the same catheter. To determine if these infections were due to the same SE strain, in 3 of these 5 patients (2 of 5 could not be analyzed because of technical problems) pulsed field gel electrophoresis after Sma 1 macrorestriction was performed as previously described9 in all isolates collected for each infection. The results were interpreted according to the criteria of Tenover et al.10 To evaluate the impact of teicoplanin-lock on bacteriological resistance, minimum inhibitory concentrations of teicoplanin were determined on the same isolates by a macrodilution agar method using 2-fold dilutions from 0.06 to 64 (mu)g/mL.
Because of the chemical and physical incompatibility between heparin and teicoplanin, during the lock, the usual heparin-lock was replaced by a daily bolus injection of heparin solution before cyclic PN.
X^sup 2^ test was used to compare home patients with hospitalized patients.
The results of drawn-back and peripheral blood cultures and the catheter culture after removal are described in Table I.
During the study period, 21 of 263 patients had 34 cases of catheter-related sepsis (0.11 catheter-related sepsis/patient per year): These patients were 12 of 209 hospitalized patients and 9 of 54 home patients (p
A non-SE infecting organism was the cause of 10 of the 34 cases of catheter-related sepsis (in 9 patients): 5 Staphylococcus aureus, 1 Candida, 1 Bacillus circulans, 1 Pseudomonas, 1 Streptococcus B, and 1 corynebacterium. The other 24 of 34 cases were SE-related in 12 patients (19 meti S and 5 meti R; Table II).
In 4 of 34 cases of catheter-related sepsis, the catheter was removed immediately because it was the scheduled time for the end of treatment (n = 3) or because the patient came from an intensive care unit (n = 1). The blood and catheter cultures showed that these 4 cases were SE-related (n = 3) and corynebacterium-related (n = 1). In all other cases (n = 30), the lock was initially used before the blood culture results. The catheter was removed later if blood culture results showed a non-SE infecting organism (n = 9).
A new catheter was inserted in 4 of 10 cases (1 after immediate removal and 3 for non-SE-catheter-related sepsis). In 5 of 24 SE-related cases, the catheter was removed after 48 hours of lock because fever persisted 48 hours after the lock had begun (n = 2; 1 with SE meti S and 1 with SE meti R infection), because of thrombosis in the subclavian vein (n = 2; 1 with SE meti S infection and 1 with SE meti R infection), or because the catheter was displaced (n = 1). Finally, the catheter was not removed after 48 hours of lock in 15 of 34 cases of catheter-related sepsis.
Five of the 12 patients with SE-related sepsis had 2 or more SE-related sepsis on the same catheter. The first SE infections (median delay, 60 days; range, 7 to 110 days) were SE meti S infections in 5 of 5 cases occurring in home PN. The lock allowed short-term continued use of the catheter in 5 of 5 cases. The median delay for development of a second infection was 50 days (range, 43 to 315 days). The second infection was due to meti R SE (2 cases) and meti S SE (3 cases). In the 2 meti R SE second infections, lock did not prevent a third infection after 18 months of follow-up in 1 patient, and the efficacy of lock could not be determined for the second patient who had final-stage cancer and died.
In 3 patients, the second infection in the same catheter was SE meti S infection, which was successfully treated with lock. However, in these 3 patients, a third SE meti S infection occurred within a mean of 90 days (range, 20 to 140 days).
Minimum inhibitory concentrations of teicoplanin in the blood cultures of the 3 patients with a second infection after a first SE-catheter-related sepsis who could be evaluated are described in Table III. It is interesting to note that two-thirds of the isolates in these SE-related cases first had an intermediate sensitivity to teicoplanin and showed stable minimum inhibitory concentrations despite teicoplanin-lock use. In 1 case, the isolates responsible for the third infectious episode showed an increased minimum inhibitory concentration and switched from S to I categorization. Pulsed field gel electrophoresis typing confirmed a relapse with a similar SE strain in 3 of 3 patients according to the criteria of Tenover et al.10
A total of 22 of 34 cases of catheter-related sepsis were associated with bacteriemia and resulted in the use of systemic antibiotics associated with the technique.
In this study, the incidence of catheter-related sepsis was lower (0.11 catheter-related sepsis/patient per year) than that reported in other series, which varied from 0.5/patients per year11 to 1.5/patients per year.12 The incidence at home was also lower in this study than that reported in other published studies.11-13 The literature concerning home PN indicates a mean annual incidence of catheter-related sepsis ranging from 0.36 to 0.50.14,15 The largest series from a single institution in home PN patients reported an overall infection rate of 0.37/patient per year.13 However, this retrospective study was performed in patients with home PN between 1973 and 1991 and the protocols for manipulation of the catheter may explain these differences. In our study, the incidence of catheter-related sepsis in hospitalized patients was significantly higher than at home. Most of these hospitalized patients came from intensive care units, and catheter-related sepsis probably occurred in the first days after insertion of the catheter due to the increased risk of contamination in these units. The reinfection rate in hospitalized patients is probably underestimated compared with home patients. The lower reinfection rate in the hospitalized patient group may be due both to a shorter duration of PN and more frequent ablation of the catheter.
The lower incidence of catheter-related sepsis in the present study compared with other studies cannot be explained by the criteria to define catheter-related sepsis. The criteria in this study are more rigorous than those in the literature.2,3,13,16,17
Most cases of catheter-related sepsis have been shown to be related to gram-positive bacteria, especially SE.2,12-14,18 This study confirms these data. However, to our knowledge, no other studies have investigated meti S and meti R Staphylococcus, and most SE in this study were meti S.
The antibiotic-lock technique is a well-accepted method of treatment for catheter-related sepsis. The concentration of antibiotics inside the catheter is 40-fold that of peak-blood levels of antibiotics given during conventional parenteral delivery.1 Although some authors also use the lock to prevent catheter infections,19 we only used it for SE-related sepsis. This strategy is controversial; some authors use the lock for non-SE infections17 and even for Candida,16,20 but others have shown that this technique does not successfully treat Candida sepsis.2
Vancomycin and amikacin are the most frequently investigated antibiotics in locks.17 We chose teicoplanin because as in other studies,3 preliminary tests with vancomycin-lock showed a precipitate when heparin saline was added. Teicoplanin is a complex glycopeptide with specific activity against gram-positive cocci and comparable antibacterial activity to vancomycin.21 The teicoplanin-lock in this study contained 5 mg/2 mL, corresponding to 300 times the minimum inhibitory concentrations (50 times the seric peak). These concentrations are lower than those used by McCarthy et al3 in 11 children with cancer and in some cases with neutropenia (400 mg/3 mL; ie, 4000 to 20,000 times the minimum inhibitory concentrations). These authors reported a short-term success rate of 100%. However, 36% of the patients developed another infection within a mean delay of 13 weeks after the first infection. The relatively low concentration of teicoplanin used in this study may explain the less successful short-term results compared with McCarthy et al.3 The concentration of teicoplanin used was similar to that used by other authors with other antibiotics,1 except McCarthy et al,3 which was 4000 to 20,000 times the minimum inhibitory concentrations because their patients were immunosuppressed.
Criteria of reinfection also differ in the literature: any infection with the same putative organism occurring within 8 weeks13 or 3 months16 after a first catheter-related sepsis was considered recurrent infection in 2 previous studies,13,16 whereas Messing et al1 did not consider 7 successive infections on the same catheter in the same patient to be recurrent infections.
Our results do not suggest that teicoplanin-lock is highly effective for SE-catheter-related sepsis. Although this technique prevented immediate catheter removal (9 of 12 patients), 5 of 9 of these patients showed 1 or more SE recurrent infections on the same catheter. Moreover, SE meti S became meti R in 2 of 5 patients.
To determine if the same bacterial strain was involved in the 3 patients who had SE-recurrent infections, we determined their pulsed field gel electrophoresis patterns. One of the most interesting results showed a bacterial relapse with the same strain in 3 of 3 patients by pulsed field gel electrophoresis typing. The inefficacy of teicoplanin-lock in preventing recurrent infection with the same strain could be due to the adherence of SE to the polyurethane catheter. Moreover, the increase in minimum inhibitory concentrations of teicoplanin observed in one-third of the patients suggests that teicoplanin-lock caused the development of SE resistance against this antibiotic, which could compromise later treatment with this antibiotic.
As most of the patients (22 of 34) received combined systemic antibiotherapy because of bacteriemia, the results probably would have been worse if patients only received the lock.
These relatively poor results confirm previous negative reports2,20 but differ from others. McCarthy et al3 had better results and showed a 100% short-term efficacy and 64% efficacy within a mean delay of 13 weeks after the first infection. Messing et al1,14 used amikacin, vancomycin, and minocycline for lock. In their first study,1 15 cases of catheter-related sepsis were observed in 6 of 11 patients. The delay between the episodes of recurrent infections was not specified. In a second study,14 9 of 19 patients had 27 catheter-related sepsis, 15 of which were recurrent infections with a median delay of 152 days after the first lock. The same bacteria was observed in 8 of 15 cases.
In our study, the incidence of catheter-related sepsis was low. It was lower at home than in the hospital. Catheter-related sepsis was mostly related to SE. In most patients with long-term PN, teicoplanin-lock did not prevent a second or third infection. When an analysis could be performed, reinfections were shown to be due to the same bacterial strain with an increase in teicoplanin resistance. Although this study is not prospective, it does not support the use of teicoplanin-lock in SE-catheter-related sepsis in patients undergoing long-term PN. ACKNOWLEDGMENTS
The authors thank Dale Roche for editorial assistance and the nurses in their unit for their collaboration.
Received for publication, October 18, 2000.
Accepted for publication, December 4, 2001.
1. Messing B, Peitra-Cohen S, Debure A, et al: Antibiotic lock technique: A new approach to optimal therapy for catheterrelated sepsis in home parenteral nutrition patients. JPEN 12:185-189, 1988
2. Krzywda EA, Andris DA, Edmiston CE, et al: Treatment of Hickman catheter sepsis using antibiotic lock technique. Infect Control Hosp Epidemiol 16:596-598, 1995
3. MacCarthy A, Byrne M, Breathnach F, et al: "In situ" teicoplanin for central venous catheter infection. Ir J Med Sci 164:125-127, 1995
4. Kentos A, Struelens MJ, Thys JP: Antibiotic-lock technique for the treatment of central venous catheter infections. Clin Infect Dis 23:418-419, 1996
5. Gaillard JL, Merlino R, Pajot N, et al: Conventional and nonconventional modes of vancomycin administration to decontaminate the internal surface of catheters colonized with coagulase-negative staphylococci. JPEN 14:593-597, 1990
6. Widmer AF: Management of catheter related bacteriemia and fungemia in patients on total parenteral nutrition (TPN). Nutrition 13(Suppl):185-255, 1997
7. Fan ST, Teoh-Chan CH, Lau KF: Evaluation of central venous catheter sepsis by differential quantitative blood culture. Eur J Clin Microbiol Infect Dis 8:142-144, 1989
8. Yagupsky P, Nolte FS: Quantitative aspects of septicemia. Clin Microbiol Rev 3:269-279,1990
9. Maslow JN, Slutsky AM, Arbeit RD: Application of pulsed field gel electrophoresis to molecular epidemiology. IN Diagnostic Molecular Microbiology: Principles and Applications, Persing DH, Smith TF, Tenover FC, et al (eds). American Society of Microbiology, Washington DC, 1993, pp 563-572
10. Tenover FC, Arbeit RD, Goering RV, et al: Interpreting chromosomal DNA restriction patterns produced by pulsed field gel electrophoresis: Criteria for bacterial strain typing. J Clin Microbiol 33:2233-2239, 1995
11. Vargas JH, Ament ME, Berquist WE: Long-term home parenteral nutrition in pediatrics. Ten years experience in 102 patients. J Pediatr Gastroenterol 6:24-37, 1987
12. Schmidt-Sommerfield E, Snyder G, Rossi T, et al: Catheter related complications in 35 children and adolescents with gas
trointestinal disease on home parenteral nutrition. JPEN 14:148-151, 1990
13. Buchman AL, Moukarzel A, Goodson B, et al: Catheter related infections associated with home parenteral nutrition and predictive factors for the need for catheter removal in their treatment. JPEN 18:297-303, 1994
14. Messing B, Man F, Colimon R, et al: Antibiotic lock technique is an effective treatment of bacterial catheter related sepsis during parenteral nutrition. Clin Nutr 9:220-225, 1990
15. Messing B: Catheter related sepsis during home parenteral nutrition. Clin Nutr 14:46-51, 1995
16. Benoit JL, Carandang G, Sitrin M, et al: Intraluminal antibiotic treatment of central venous catheter infection in patients receiving parenteral nutrition at home. Clin Inf Dis 21:1286-1288, 1995
17. Messing B: Catheter related sepsis during home parenteral nutrition: Use of the antibiotic-lock technique. Nutrition 14:466-468, 1998
18. Capdevila AJ, Segara A, Planes AM, et al: Long term follow up of patients with catheter related bacteremia treated with out catheter removal. Clin Microbiol Infect 4:472-476, 1998
19. Jurewitsch B, Lee T, Park J, et al: Taurolidine 2% as an antimicrobial lock solution for prevention of recurrent catheter-related bloodstream infections. JPEN 22:242-244, 1998
20. Faintuch J, Waitzburg DL, Bertevello PL, et al: Conservative management of septic parenteral nutrition catheters JPEN 19:428-429, 1995
21. Brogden RN, Peters DH: Teicoplanin. A reappraisal of its antimicrobial activity, pharmacokinetics properties and therapeutic effects. Drugs 5:823-854, 1994
C. Guedon, MD, PhD*; M. Nouvellon, MD^; O. Lalaude, MD*; and E. Lerebours, MD*
From the *Hepato-Gastroenterology and Nutrition Unit and ^Microbiology Unit, Hospital Charles Nicolle, Rouen Cedex, France
Correspondence and reprint requests: Dr Claire Guedon, MD, PhD, Service d'Hepatogastroenterologie et Nutrition, Hopital Charles Nicolle, 76031 Rouen Cedex, France. Electronic mail may be sent to firstname.lastname@example.org.
Copyright American Society for Parenteral and Enteral Nutrition Mar/Apr 2002
Provided by ProQuest Information and Learning Company. All rights Reserved