• 
• 

Three Part Question

In [a stable child with a central venous catheter colonised by bacteria] is [antibiotic line lock] useful [to eradicate the colonisation]?

Clinical Scenario

You are the specialist registrar in paediatrics in a district general hospital which provides shared care for paediatric oncology patients. A girl on cyclical maintenance chemotherapy for acute lymphoblastic leukaemia was admitted 10 days ago with neutropenic sepsis. Blood culture from her Hickman line at admission had grown Staphylococcus epidermidis. She was treated according to the sensitivity pattern of the organism. She is now free of signs of systemic infection. However, the most recent blood culture from her Hickman line continues to grow S epidermidis. Concurrent peripheral venous blood cultures are sterile.

At this point, you are keen to remove the central line. However, her parents want to avoid line removal as venous access has always been extremely difficult for her. You consult the oncologist at her referral centre, who suggests a trial of antibiotic lock of the Hickman line. The microbiologist at your hospital is not in favour of this approach and wants the colonised line to be removed immediately. You are unsure about the best therapy in this situation. You decide to do a literature search on the benefits and risks of antibiotic line lock technique (ALLT) and critically appraise the evidence.

Search Strategy

Secondary sources
Cochrane library, Best Evidence and Clinical Evidence were searched. No relevant articles were found.

MEDLINE (1966 to date) and OLDMEDLINE (1950–1965) were searched by the PubMed interface on 4 May 2008.


The search terms "Catheter" AND "(antibiotic lock)" with limits (English, Human) yielded 116 results. The search was not restricted to the paediatric population in view of the lack of any difference in the biological mechanisms involved when compared to adults. We focused our analysis on the treatment of line colonisation in patients receiving cancer chemotherapy. Papers relating to long term catheter use in non-oncology patients (such as those on haemodialysis, total parenteral nutrition (TPN) and antimicrobial therapy for HIV infection) were excluded as were in vitro studies and the prophylactic use of antibiotic line locks in non-colonised central venous catheters.

Search Outcome

Seven articles were selected for final analysis. Four articles were found to be of direct relevance. In addition, three further studies were included by manual review of reference lists from identified studies and systematic reviews.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Fortun et al, 2006,	Patients on cancer chemotherapy or on TPN (48 episodes of CRB in 39 patients)	Historically controlled trial (level 3b)	Primary end point defined as "failure to cure the CRB"	CRB episodes 19 vs 29 (cases vs controls).Treatment success 16/19 (84%) vs 19/29 (65%), p = 0.27Mortality 1/19 (5%) vs 2/29 (7%)	Cases and controls had similar baseline characteristics. Absence of randomisation or blinding limits validity of conclusions.Some patients with complicated CRB excluded.Better results in ALLT group, but not statistically significant.
Rijnders et al, 2005,	174 patients with bacteraemia and CVAD were recruited.46 patients found fit for final analysis, including 34 from haemato-oncology and 12 from other (gastroenterology and renal) units. 44 proceeded to enter the study.	Double blind randomised controlled trial (level 2b)	Primary end point was "failure to cure the CRB".Outcome measures were:(1) prevention of primary endpoint until 24 weeks after inclusion;(2) death due to underlying disease;(3) removal of catheter (as no longer needed)	ALLT group vs placebo group:CRB episodes 21 vs 23Treatment success 14/21 (67%) vs 10/23 (43%), p = 0.1	Well conducted study with clear definitions and methodology. Strict exclusion criteria to avoid all possible confounding variables.Failed to achieve minimum sample size for adequate statistical power.Excluded Staphylococcus aureus CRB as clinicians were reluctant to preserve line.Global and subgroup analysis indicate potential benefit for ALLT; however results not statistically significant.
Sanchez-Munoz et al, 2005,	14 episodes of "uncomplicated" CRB in non-neutropenic patients with solid tumours	Prospective case series (level 4)	"Treatment successful" if afebrile and culture negative in 72 h."Catheter salvage successful" if catheter in place >3 months with no recurrence of bacteraemia or death.	Treatment and catheter salvage successful in 12 cases.Catheter removed in the remaining 2 patients due to growth of resistant bacteria or multiple organisms.	Limited sample size with no controls.Quantitative blood cultures not done.Most CRB episodes were caused by Staphylococcus epidermidis. No episode due to S aureus or Gram negative bacteria.Avoided patients with neutrophils less than 0.5x109/l.Anecdotal evidence in favour of ALLT in S epidermidis CRB.
De Sio et al, 2004,	10 children on chemotherapy for solid tumours with Gram positive bacteraemia	Prospective case series (level 4)	Negative blood cultures from catheter, avoidance of device removal.	ALLT resulted in negative cultures in all cases.Catheter survival time 11–334 days; median 90 days.Authors claim 100% success.	Limited sample size.Lack of controls and limited details on catheter survival times makes authors’ conclusions questionable.Weak anecdotal evidence in favour of ALLT.
Longuet et al, 2001,	16 adults with HIV (n = 10) or cancer (n = 6) with CRB	Prospective case series (level 4)	Clinical resolution of sepsis within 48 h and negative blood cultures within 5 days of ALLT.	ALLT completely successful in 5 (31%), partial response in 2 (13%) and failure in 9 (56%) patients.	Limited sample size.Main aim was to establish the diagnostic role of reservoirs in CRB.Clear definitions and stringent outcome measures.Duration of catheter survival not assessed as an end point.
McCarthy et al, 1995,	11 paediatric oncology patients with Gram positive infection of CVAD (19 infective episodes)	Prospective case series (level 4)	Negative blood cultures from catheter, avoidance of device removal.	ALLT successful in 100% of episodes, with average duration of first negative culture 6 days.Mean catheter survival time 136 days (55–262 days).All CVADs remained in situ until either chemotherapy completion or death.	Small study attempting to extrapolate previous work by Rao et al to Gram positive CVAD infections.No control group.Teicoplanin used as empirical antibiotic of choice.Validity of conclusions limited by sample size.
Rao et al, 1992,	11 paediatric oncology patients in a tertiary centre (14 infective episodes)	Prospective case series (level 4)	Negative blood cultures from catheter, avoidance of device removal.	ALLT successful in 100% of episodes, with average duration of first negative culture 8 days.Mean catheter survival time 118 days (51–249 days).	Innovative and pioneering study on efficacy of ALLT in paediatric population.Limited sample size with lack of controls.Simple methodology and clear definitions. Amikacin chosen as initial antibiotic.

Comment(s)

Bacterial colonisation of central venous access devices (CVADs) is a major cause of morbidity and potential cause of mortality in children receiving cancer chemotherapy. Catheter related bacterial infections can occur in the form of exit site infection (erythema, tenderness or swelling with positive skin swab), tunnel infection (erythematous tracking along the catheter path), colonisation of the line (positive blood culture from CVAD or fever/rigor with line flush, peripheral venous cultures sterile) or true catheter related bacteraemia (bacterial isolation from central and peripheral blood cultures)(Bishop). Device removal is frequently advised to manage the infection. However, it can result in significant morbidity, which includes the need for general anaesthesia during removal and reinsertion of a new line, and delay in chemotherapy. Particular problems arise with recurrent infections following CVAD colonisation. These concerns prompted early researchers to try alternative measures using intra-catheter (in situ) antibiotics, in an attempt to avoid catheter removal and minimise the exposure to systemic antibiotics (Longuet, McCarthy).Formation of a microbial biofilm matrix within the CVAD is a crucial step in the pathogenesis of line colonisation. In vitro studies have demonstrated the efficacy of topical high concentration antibiotic solutions in penetrating this biofilm, thus providing biological plausibility and rationale for the use of this intervention (Droste).

This intervention, now called the antibiotic line lock technique (ALLT), involves instilling a concentrated antibiotic solution into a CVAD and maintaining it within the lumen for a sufficiently long duration to eradicate its colonisation by bacteria (Mermel). The solution is instilled in a volume adequate to fill the lumen of the catheter and the clamp is applied to "lock" the solution in place. The CVAD is then left unaccessed for the desired duration. Treatment is considered successful if subsequent blood cultures from the treated lumen are negative, thus leading to salvage of the CVAD.

In recent years, "locks" other than antibiotics have also been described to manage line infection. This includes the use of solutions with an antiseptic property, such as taurolidine, ethanol and hydrochloric acid (Pagani). Most studies on these agents have been on the prevention of colonisation rather than the treatment of established infections. Some clinicians advocate administration of systemic antibiotics through the infected line without "locking" the solution in situ. There have been no trials comparing ALLT with the above options. There is a paucity of studies evaluating these alternative approaches and further discussion is beyond the scope of this analysis.

Practice and opinion regarding the efficacy and safety of ALLT in the treatment of a colonised CVAD vary considerably among paediatricians, haemato-oncologists and microbiologists (Bishop). In spite of theoretical advantages and in vitro studies favouring ALLT, doubts have been raised concerning its applicability in clinical practice. Potential for progression to fulminant sepsis and mortality remains a valid concern, precluding widespread adoption of ALLT as opposed to device removal. In clinical practice, a haemodynamically compromised child with confirmed CRB will require immediate removal of the infected line. CVAD infections caused by highly pathogenic organisms such as candida and certain Gram negative bacilli will also need line removal. In the more common scenario of a stable child with a CRB or line colonisation due to organisms with low pathogenicity, available evidence and risk benefit analysis justify the use of ALLT as a valid option before considering line removal. There is lack of scientific data regarding the best therapy for infection of multilumen CVADs and treatment therefore needs to be individualised. When infection is confined to one lumen, it seems reasonable to limit ALLT solely to that lumen. In case of infection of more than one lumen, ALLT will be required either simultaneously or alternately for all infected lumens.

A review of existing literature reveals some evidence, albeit low grade, in favour of ALLT (table 2). The choice, dose and duration of antibiotic as well as the need for adjuncts such as heparin remains to be standardised. The United Kingdom Children’s Cancer and Leukaemia Group (CCLG) is currently co-ordinating a multicentre study to compare the efficacy of teicoplanin lock/infusion versus conventional bolus administration in coagulase negative Staphylococcus line infections (Portfolio database). Further large randomised trials are required to provide more concrete evidence on the usefulness of ALLT as well as to clarify its specific efficacy against individual organisms.

Editor Comment

ALLT, antibiotic line lock technique; CRB, catheter related bacteraemia; CVAD, central venous access device; TPN, total parenteral nutrition.

Clinical Bottom Line

There is a lack of high quality evidence to recommend antibiotic line lock as a standard approach for the treatment of colonised central lines.

An open clinical trial in the UK is evaluating the efficacy of antibiotic line locks in paediatric oncology patients with coagulase negative Staphylococcus line infections.

In selected patients, judicious use of antibiotic line lock is a safe technique to eradicate bacterial colonisation and avoid catheter removal (Grade C).

References

1. Fortun J, Grill F, Martin-Davila P, et al.. Treatment of long-term intravascular catheter-related bacteremia with antibiotic-lock therapy. J Antimicrob Chemother 2006;58:816–21.
2. Rijnders BJ, van Wijengaerden E, Vandecasteele SJ, et al. Treatment of long-term intravascular catheter-related bacteremia with antibiotic lock: randomized, placebo-controlled trial. J Antimicrob Chemother 2005;55:90–4.
3. Sanchez-Munoz A, Aguado JM, Lopez-Martin A, et al. Usefulness of antibiotic-lock technique in the management of oncology patients with uncomplicated bacteremia related to tunneled catheters. Eur J Clin Microbiol Infect Dis 2005;24:291–3.
4. De Sio L, Jenkner A, Milano GM, et al. Antibiotic lock with vancomycin and urokinase can successfully treat colonised central venous catheters in pediatric cancer patients. Pediatr Infect Dis J 2004;23:963–65.
5. Longuet P, Douard MC, Arlet G, et al. Venous access port-related bacteremia in patients with acquired immunodeficiency syndrome or cancer: the reservoir as a diagnostic and therapeutic tool. Clin Infect Dis 2001;32:1776–83.
6. McCarthy A, Byrne M, Breathnach F, et al. In-situ teicoplanin for central venous catheter infection. Ir J Med Sci 1995;164:125–7.
7. Rao JS, O’Meara A, Harvey T, et al. A new approach to the management of Broviac catheter infection. J Hosp Infect 1992;22:109–16.
8. Bishop L, Dougherty L, Bodenham A, et al. Guidelines on the insertion and management of central venous access devices in adults. Int J Lab Hematol 2007;29:261–78.
9. Droste JC, Jeraj HA, MacDonald A, et al. Stability and in-vitro efficacy of antibiotic heparin lock solutions potentially useful for treatment of central venous catheter related sepsis. J Antimicrob Chemother 2003;51:849–55.
10. Mermel LA, Farr BM, Sherertz RJ, et al. Guidelines on the management of intravascular catheter-related infections. Clin Infect Dis 2001;32:1249–72.
11. Pagani JL, Eggimann P. Management of catheter-related infection. Expert Rev Anti Infect Ther 2008;6:31–7.
12. UK Clinical Research Network. Portfolio database. Available from http://public.ukcrn.org.uk/search/StudyDetail.aspx?StudyID = 1158 (accessed 24 April 2009).