Eur opean Rev iew for Med ical and Pharmacol ogical Sci ences 2015; 19: 2440-2445 Catheter-related bloodstream infections by opportunistic pathogens in immunocompromised hosts

A. MIRIJELLO 1, M. IMPAGNATIELLO 1, V. ZACCONE 1, G. VENTURA 2, L. POMPA 2, G. ADDOLORATO 1, R. LANDOLFI 1, on behalf of Internal Medicine Sepsis Study Group*

1Department of Internal Medicine, Catholic University of the Sacred Heart, Rome, Italy 2Clinic of Infectious Diseases, Catholic University of the Sacred Heart, Rome, Italy *Internal Medicine Sepsis Study Group: A. Adamo, C. Kadhim, G. Passaro, A. Tosoni, C. Vallone, V. Veca

Abstract. Catheter-related bloodstream in - and the potential for coagulation disorders 4. fections (CRBI) represent a frequent complica - Moreover, when clinical situation allows, the cost tion of immune-compromised hosts with a high of a new procedure supports catheter salvage 1,4 . mortality rate. In this setting, opportunistic Antibiotic lock therapy (ALT) consists in the pathogens can create a on implanted devices, being the source of infection. infusion of a concentrated antibiotic solution in a We provide a mini-review of the literature, small volume, in order to completely fill the lu - starting from the description of two cases of men of the catheter. The rationale of this tech - CRBI by opportunistic pathogens in poly-mor - nique is to expose the interior walls of the bid patients, successfully treated by antibiotic catheter to high antibiotic concentrations, in or - lock-therapy. der to penetrate the biofilm, and to achieve bacte - rial eradication 5. ALT at present represents an ap - Key Words: proved indication for coagulase-negative staphy - Sepsi, Opportunistic pathogens, Central venous 6 catheter, Biofilm. lococci and Gram-negative bacilli . In case of Gram-positive vancomycin represents the antibiotic of choice, while, in case of Gram- negative bacilli, ciprofloxacin represents the most used treatment 7. It has been demonstrated Introduction that the use of ALT in addition to intravenously administered antibiotics reduces the failure rate Catheter-related bloodstream infections (CRBI) to cure the CRBI, although the optimal ALT regi - represent a significant cause of morbidity and mens are still matter of research 8, in particular in mortality in patients treated with long-term vas - the treatment of opportunistic-germ related infec - cular access devices 1. As well as other types of tions. medical implants, venous devices are associated Here we provide a mini-review of the litera - with the development of biofilm on the internal ture, starting from the description of two cases of lumen 2. Consequently, CRBI represent a typolo - catheter-related bloodstream infections by oppor - gy of sepsis in which the vascular device acts as tunistic pathogens in immunocompromised hosts source and reservoir of infection 3. successfully treated by ALT in combination to As general rule, the treatment of bloodstream systemic treatment. infections is based on the removal of the source of sepsis and on the administration of intra - Case One venous antibiotics 1,3 . When CRBI is diagnosed, A 47-year-old woman was admitted to our In - the most effective therapeutic strategy is to re - ternal Medicine inpatient Unit because of a 10- move the catheter, in order to remove the source days fever (peak 38°C) resistant to ceftriaxone of infection and to collect microbiological sam - treatment. The patient was affected by Systemic ples for culture 1. On the other hand, for certain Lupus Erythematous (SLE) with end-stage renal patients, catheter removal could represent a high- disease treated by hemodialysis via a silicone, risk procedure since the lack of venous access double-lumen, tunneled, right subclavian central

2440 Corresponding Author: Antonio Mirijello, MD; e-mail: [email protected] Catheter-related bloodstream infections by opportunistic pathogens in immunocompromised hosts venous catheter (CVC) due to a previous throm - ment with meropenem was continued for 10 bosis of the artero-venous fistula. At physical ex - days, until PCT normalization. Antibiotic lock amination patient’s clinical condition were dis - therapy with ciprofloxacin was prescribed for 14 crete: blood pressure, pulse and respiratory rate days. The patient showed a complete recovery. and peripheral oxygen saturation were normal Blood cultures from CVC collected 15 days after (Table I). A complete blood count showed mild ciprofloxacin discontinuation were negative. thrombocytopenia without leukocytosis. Chest X-ray was normal. At the first febrile peak, pe - ripheral and CVC blood cultures were collected Case Two and an empirical antibiotic treatment with An 81 -year-old man was admitted to our In - meropenem was started . Procalcitonin (PCT) ternal Medicine inpatient Unit because of the [Thermo Fischer Scientific (B-R-A-H-M-S PCT- rapid onset of fever, dyspnea and blood-traced Q) ] levels were high , beta-D-glucane was nega - cough after chemotherapy infusion. He had tive (Table II). Trans-thoracic echocardiography been treated with left hemicolectomy because of (TTE) did not show signs of endocarditis. Blood colon adenocarcinoma, and he was now on ad - cultures were positive for pickettii juvant chemotherapy because of lung and liver (Burkholderia pickettii ) with a differential time metastases. A few hours after the infusion of to positivity of 12 h (6h from CVC vs. 18h from chemotherapy through the CVC (Port-a-Cath), peripheral blood ). Ralstonia pickettii is a non -fer - the patient showed nausea, vomiting, chills , menting gram-negative bacillus known to have a fever, dyspnea and blood-traced cough and relatively low virulence and to be often associat - came to the emergency department . Chest X-ray ed with pseudobacteremia or asymptomatic colo - showed the presence of multiple bilateral pul - nization of patients 9. However, it has been report - monary hypodiaphanias, associated to bilateral ed to create a biofilm on implanted devices 9. An - pleural effusion. Blood analysis showed elevat - tibiotics’ susceptibility is shown in Table I . Treat - ed serum creatinine, anemia and neutrophilic

Table I. Clinical characteristics and blood tests of case 1.

VITAL PARAMETERS Blood pressure 130/70 mmHg Normal Pulse rate 90 beats/min Normal Respiratory rate 18 breaths/min Normal

Peripheral oxygen saturation 97% in 0.21 FiO 2 Normal

LABORATORY TESTS

Blood count: • HB 12.5 g/dl (n.v. 12.0; 16.0 g/dl) • WBC 4.05 x 10^9/l (n.v. 4.10; 9.80 x 10^9/l) • PLT 138,000/mm 3; ( n.v. 140 ,000; 450, 000/mm 3)

OTHER EXAM Procalcitonin PTC 4.97 ng/ml (n.v. 0.00; 0. 05 ng/ml) B-D-glucane negative

Antibiotics’ susceptibility spp MIC (mcg/ml) Ralstonia pickettii

Piperacillin/tazobatam Sensitivity < =4 Ciprofloxacin Sensitivity < =0.2 Cefepime Sensitivity < =1 Meropenem Sensitivity 1 Gentamicin Resistance > =16 Ceftazidime Resistance 16 Amoxicillin/clavulanic acid Resistance 16 Amikacin Resistance

2441 A. Mirijello, M. Impagnatiello, V. Zaccone, G. Ventura, L. Pompa, G. Addolorato, R. Landolfi

Table II. Clinical characteristics and blood tests of case 2.

VITAL PARAMETERS Blood pressure 90/60 mmHg Low Pulse rate 95 beats/min regular High Respiratory rate 24 breaths/min High

Peripheral oxygen saturation 85% in 0.5 FiO 2 Low

LABORATORY TESTS

Blood gas:

• pH 7.414, pO 2 85.9 mmHg (pO 2/FiO 2 170), • pCO 2 35.5 mmHg, • HCO 3- 22.0 mEq/l.

Blood count: • HB 11.0 g/dl ( n.v. 12.5;16 g/dl) • WBC 11.28 x 10^9/l (n.v. 4.10; 9.80 x 10^9/l) • NG 9.64 x 10^9/l (n.v. 1.90; 7.00 x 10^9/l) • PLT 142,000/mm 3; (n.v. 140,000; 450,000/mm 3)

OTHER EXAM Serum creatinine 1.64 mg/dl (n.v. 0.67; 1.17) Clcr Cockroft-Gault 39.00 mL/min NT-Pro BNP 9753 pg/ml (n.v. 0; 450) Procalcitonin 2.55 ng/ml (n.v. 0.00; 0.00) C-Reactive Protein 81.5 mg/mL (n.v < 3) B-D-glucane Negative

Antibiotics’ susceptibility Agrobacterium radiobacter MIC (mcg/ml) Rhizobium radiobacter Imipenem Sensitivity < = 0.25 Ciprofloxacin Sensitivity < = 0.25 Meropenem Sensitivity < = 0.5 Gentamicin Resistance 4

leukocytosis (Table II). Patient’s clinical condi - devices 11 . Antibiotics’ susceptibility is shown in tions at admission were poor: hypotension, Table I . Treatment with piperacillin/tazobactam tachycardia and tachypnea were present (Table was continued for 10 days, until PCT nor maliza - II). Fluid therapy was started via CVC. After tion. Antibiotic lock therapy with ciprofloxacin few minutes of infusion, the patient presented was prescribed for 14 days. The patient showed shaking chills. Therefore , peripheral and CVC a complete recovery. Blood cultures from CVC blood cultures were collected and an empirical and collected 15 days after ciprofloxacin dis - antibiotic treatment with piperacillin/tazobac - continuation were negative. tam, teicoplanin and levofloxacin was started, together to fluid therapy via a peripheral venous access . PCT levels were high [Thermo Fischer Scientific (B-R-A-H-M-S PCT-Q) ], beta-D- Discussion glucane was negative (Table II). Transthoracic echocardiography (TTE) did not show signs of Patients admitted to Internal Medicine Inpatient endocarditis. Blood cultures were positive for units are typically affected by multiple comorbidi - Rhizobium radiobacter with a differential time ties, taking poly-pharmacotherapy and often need - to positivity of 20h (19h from CVC vs. 39h from ing chronic treatments. Both of the described pa - peripheral blood ). Rhizobium radiobacter is a tients were immune-compromised, the first one be - Gram -negative bacillus infrequently recognized cause of autoimmune disease, the second one be - in clinical setting but emerging as an opportunis - cause of chemotherapy. Both were needing a long- tic human pathogen 10 . However, it has been re - term CVC, one because of haemodialysis, the oth - ported to create a biofilm on implanted er because of chemotherapy.

2442 Catheter-related bloodstream infections by opportunistic pathogens in immunocompromised hosts

Ralstonia pickettii , generally considered of mi - as an alternative pharmacological approach to nor clinical significance, has been frequently control bacterial biofilm growth in human dis - found in dialysis water treatment systems and has eases 23,24 . been shown to be associated to CRBI 12,13 . Rhizo - Gram-positive bacteria are most frequently re - bium radiobacter , also called agrobacterium ra - sponsible for CVC infections . However , other diobacter , represents a phytopathogenic organ - pathogens, such as Gram-negative bacilli, fungi ism widely distributed in soil, being recognized and corynebacteria can be found, particularly in as rare human pathogen affecting mostly im - neutropenic patients 18 . mune-compromised hosts 14 , in particular onco - Risk factors for catheter-related infection are logic patients 15 . Not much is known about the an - associated with the time of catheterization and tibiotic resistance of these bacteria and about the handling of the venous catheter 1,18 , the variables optimal treatment strategy. The majority of Ral - related to the type of tunneled catheter 25 , catheter stonia pickettii isolates showed susceptibility to insertion technique 26 , hematological malignan - most of tested antibiotics. However, the most ef - cies 25 , patient age 27 , parenteral nutrition adminis - fective were found to be the quinolones and sul - tration 22,25 , immunosuppression 28 , and prolonged famethoxazole-trimethoprim 12 . neutropenia 29,30 . Catheter-related bloodstream infections repre - There are two methods for in situ diagnosis of sent a significant cause of morbidity and CRB: quantitative culture of paired blood sam - mortality 16 , being also a major cause of premature ples, and differential time to positivity of catheter catheter removal and increased treatment costs 3. blood sample compared to peripheral vein blood As well as other infections related to almost every culture. Quantitative culture of paired blood sam - temporary or permanent medical implant, such as ples looks for the correlation between a positive orthopedic prosthesis and cardiac and neurological differential quantitative blood culture threefold devices, venous device infections are associated greater than identical bacterial colony count in with biofilm development on the surface of the peripheral vein blood specimen. In differential foreign body 17 . Most of the infections originate time to positivity of catheter blood sample com - from the skin microbiota surrounding the insertion pared to peripheral vein blood culture, a cut-off site of the catheter. Micro-organisms reach the differential time to positivity value of 120 min - subcutaneous catheter tract at the moment of ve - utes has 91% specificity and 94% sensitivity for nous catheter implantation. This event is often re - the diagnosis of CRB 16,18 . lated to external colonization of short-term When CRBI is diagnosed, the most effective catheters 1,18 , while catheter manipulation is impli - therapeutic strategy is to remove the device, since cated with intraluminal colonization of long-term catheter retention is associated with a high risk of catheters 18,19 . Repeated access to ports often leads bacteremia recurrence . Withdrawing the catheter to formation of intraluminal biofilm 20 . The first removes the source of infection and enables mi - step in biofilm development is the attachment of crobiological analysis of the catheter, but the deci - bacterial cells to an artificial or native surface. Mi - sion to remove a catheter should take into consid - cro-organisms will form a community that will en - eration at least three factors: the type of the case itself in a self-produced polymeric matrix 21,22 . catheter; the micro-organism involved in the infec - This matrix is a hydrated polyanionic complex of tion; and clinical status of the patient. Usually exopolysaccharides of bacterial origin, but can al - adopted criteria for CVC removal are signs of sub - so be formed by proteins and DNA. As the poly - cutaneous tunnel infection, suppurative phlebitis , meric matrix grows and matures, it builds a so - pulmonary embolization, clinical suggestion of in - phisticated system of water channels resembling a fective endocarditis, persistent bacteremia, recur - circulatory system for biofilm support 17,21 . It has rent infection despite medical treatment, or infec - been estimated that are associated with tion sustained by Staphylococcus (S.) aureus , more than 70% of nosocomial infections, and that Candida spp., or Mycobacterium spp 3,18 . When pa - the treatment of these biofilm-related infections tient’s hemodynamic is stable, CVC was surgical - costs more than a billion dollars annually 1,16 . ly implanted and may not be easily replaced, Antibiotic treatment represents the traditional catheter removal as first step of treatment should approach to eradicate biofilm-producing bacteria. be avoided 4,18 . In these cases, CVC should not be However, since the deepest layers of biofilm are used for infusion and ALT can be started in ad - inaccessible to most of antibiotics, N-acetilcys - junction to systemic treatment for a duration of 10 teine and acetylsalylic acid have been proposed to 14 days 4,5,18 . ALT consists of the infusion of a

2443 A. Mirijello, M. Impagnatiello, V. Zaccone, G. Ventura, L. Pompa, G. Addolorato, R. Landolfi

–––––––––––––––– ––––– concentrated antibiotic solution in a small volume Conflict of Interest to fill the lumen catheter. The rationality of this The Authors declare that there are no conflicts of interest. technique is to expose the interior walls of the catheter to high concentrations of an antibiotic that may penetrate the biofilm, and therefore to References achieve bacterial eradication 31,32 . The antimicro - bial lock has also been studied as a prophylactic 1) LEBEAUX D, J OLY D, Z AHAR JR. Long-term central ve - measure to prevent catheter-related bacteremia 33 . nous catheter-related infections. Rev Prat 2014; According to the Infectious Diseases Society of 64: 626-633 America (IDSA ) guidelines, when the involved 2) TALSMA SS. Biofilms on medical devices. Home micro-organism is a coagulase negative Staphylo - Healthc Nurse 2007; 25: 589-594. coccus or a Gram-negative bacillus , with positive 3) LEBEAUX D, F ERNÁNDEZ -H IDALGO N, C HAUHAN A, L EE catheter -blood cultures but negative from peripher - S, G HIGO JM, A LMIRANTE B, B ELOIN C. Management of infections related to totally implantable venous- al vein, ALT has to be considered from 10 to 14 access ports: challenges and perspectives. 34,35 days, without systemic antibiotic treatment . Lancet Infect Dis 2014; 14: 146-159. Systemic treatment has to be administered in case 4) MERMEL LA, A LLON M, B OUZA E, C RAVEN DE, F LYNN P, 34 of criteria of sepsis . O'G RADY NP, R AAD II, R IJNDERS BJ, S HERERTZ RJ, W AR - As discussed, the decision to remove a REN DK. Clinical practice guidelines for the diagno - catheter should consider at least three factors: the sis and management of intravascular catheter-re - type of CVC, the micro-organism involved in the lated infection: 2009 Update by the Infectious Dis - 18,34 eases Society of America. Clin Infect Dis 2009; infection and clinical status of the patient . 49: 1-45. Both of the described patients were affected by 5) RIJNDERS BJ, V AN WIJNGAERDEN E, V ANDECASTEELE SJ, CRBI of a long-term implanted CVC, both STAS M, P EETERMANS WE. Treatment of long-term in - showed a Gram-negative opportunistic pathogen, travascular catheter-related bacteraemia with an - and both were immune-compromised. Consid - tibiotic lock: randomized, placebo-controlled trial. ered the lack of a standardized treatment for both J Antimicrob Chemother 2005; 55: 90-94. pathogens, CVC-removal had been hypothesized 6) MARR KA, S EXTON DJ, C ONLON PJ, C OREY GR, S CHWAB SJ, K IRKLAND KB. Catheter-related bacteremia and in both cases. However, ALT seemed to be the outcome of attempted catheter salvage in pa - most safe decision. Since the presence of clinical tients undergoing hemodialysis. Ann Intern Med and laboratory signs of sepsis, empiric systemic 1997; 127: 275-280. antibiotic treatment was started after collecting 7) FERNANDEZ -H IDALGO N, A LMIRANTE B, C ALLEJA R, R UIZ blood cultures. In both cases, the differential time I, P LANES AM, R ODRIGUEZ D, P IGRAU C, P AHISSA . An - to positivity was > 120 min, allowing us to diag - tibiotic-lock therapy for long-term intravascular nose CRBI 34,36,37 . On this connection, antibiotic catheter-related bacteraemia: results of an open, non-comparative study. A. J Antimicrob Chemoth - lock -therapy was started with the resolution of er 2006; 57: 1172-1180. catheter infection, as documented by the negativ - 8) BOOKSTAVER PB, W ILLIAMSON JC, T UCKER BK, R AAD II, ity of the cultures after ALT discontinuation. The SHERERTZ RJ. Activity of novel antibiotic lock solu - use of ALT strategy led us to save the CVC, tions in a model against isolates of catheter-relat - avoiding surgical procedures (CVC removal and ed bloodstream infections. Ann Pharmacother implant of a new CVC) with the relative infective 2009; 43: 210-219. risks and costs. 9) DOMBROWSKY M, K IRSCHNER A, S OMMER R. PVC-pip - ing promotes growth of Ralstonia pickettii dialysis water treatment facilities. Water Sci Technol 2013; 68: 929-933. Conclusions 10) CHEN CY, H ANSEN KS, H ANSEN LK. Rhizobium ra - diobacter as an opportunistic pathogen in central This report highlights on a common situation venous catheter-associated bloodstream infec - in clinical practice: the need to use the best treat - tion: case report and review. J Hosp Infect 2008; ment strategy with the safest solution and the less 68: 203-207. resource-expending choice. Antibiotic lock-thera - 11) HANADA S, I WAMOTO M, K OBAYASHI N, A NDO R, S ASAKI S. Catheter-related bacteremia caused by py seems to be effective in the treatment of CRBI Agrobacterium radiobacter in a hemodialysis pa - and should be considered in the therapeutic tient. Intern Med 2009; 48: 455-457. workup of Internal Medicine patients affected by 12) RYAN MP, A DLEY CC. Ralstonia spp.: emerging glob - CRBI since it has lowers costs compared to al opportunistic pathogens. Eur J Clin Microbiol CVC-removal and re -implantation . Infect Dis. 2013; 33: 291-304.

2444 Catheter-related bloodstream infections by opportunistic pathogens in immunocompromised hosts

13) RYAN MP, P EMBROKE JT, A DLEY CC. Ralstonia pickettii: 26) KOCK HJ, P IETSCH M, K RAUSE U, W ILKE H, E IGLER a persistent gram-negative nosocomial infectious FW . Implantable vascular access systems: ex - organism. J Hosp Infect 2006; 62: 278-284. perience in 1500 patients with totally implanted 14) SOOD S, N ERURKAR V, M ALVANKAR S. Catheter asso - central venous port systems. World J Surg ciated bloodstream infection caused by R. ra - 1998; 22: 12-16. diobacter. Indian J Med Microbiol 2010; 28: 62- 27) ADLER A, Y ANIV I, S TEINBERG R, S OLTER E, S AMRA Z, 64. STEIN J, L EVY I. Infectious complications of implant- 15) DETRAIT M, D'H ONDT L, A NDRÉ M, L ONCHAY C, able ports and Hickman catheters in paediatric HOLEMANS X, M ATON JP, C ANON JL. Agrobacterium haematology-oncology patients. J Hosp Infect radiobacter bacteremia in oncologic and geri - 2006; 62: 358-365. atric patients: presentation of two cases and re - 28) DOMINGO P, F ONTANET A, S ÁNCHEZ F, A LLENDE L, view of the literature. Int J Infect Dis 2008; 12: VAZQUEZ G. Morbidity associated with long-term e7-10. use of totally implantable ports in patients with 16) MERMEL LA, F ARR BM, S HERERTZ RJ, R AAD II, O'G RADY AIDS. Clin Infect Dis 1999; 29: 346-351. N, H ARRIS JS, C RAVEN DE; Infectious Diseases Soci - 29) WINTENBERGER C, E PAULARD O, H INCKY -V ITRAT V, B RION ety of America, American College of Critical Care JP, R ECULE C, F RANÇOIS P, S TAHL JP, P AVESE P. Out - Medicine, Society for Healthcare Epidemiology of come of central venous catheter-related bacter - America Guidelines for the management of in - aemia according to compliance with guidelines: travascular catheter-related infections. J Intraven experience with 91 episodes. J Hosp Infect 2012; Nurs 2001; 24: 180-205. 80: 245-251 17) COSTERTON JW, S TEWART PS, G REENBERG EP. Bacterial 30) BASSI KK, G IRI AK, P ATTANAYAK M, A BRAHAM SW, biofilms: a common cause of persistent infections. PANDEY KK. Totally implantable venous access Science 1999; 284: 1318-1322. ports: retrospective review of long-term complica - 18) BUSTOS C, A GUINAGA A, C ARMONA -T ORRE F, D EL POZO tions in 81 patients. Indian J Cancer 2012; 49: JL. Long-term catheterization: current approaches 114-118 in the diagnosis and treatment of port-related in - 31) SORIANO A, B REGADA E, M ARQUÉS JM, O RTEGA M, fections. Infect Drug Resist 2014; 7: 25-35. BOVÉ A, M ARTÍNEZ JA, M ENSA J. Decreasing gradient 19) MAKI DG, K LUGER DM, C RNICH CJ. The risk of blood - of antibiotic concentration in the lumen of stream infection in adults with different intravascu - catheters locked with vancomycin. Eur J Clin Mi - lar devices: a systematic review of 200 published crobiol Infect Dis 2007; 26: 659-661. prospective studies. Mayo Clin Proc 2006; 81: 32) MANIERSKI C, B ESARAB A. Antimicrobial locks: putting 1159-1171. the lock on catheter infections. Adv Chronic Kid - 20) LIÑARES J, S ITGES -S ERRA A, G ARAU J, P ÉREZ JL, M ARTÍN ney Dis 2006; 13: 245-258. R. Pathogenesis of catheter sepsis: a prosp ective 33) DEL POZO JL, A LONSO M, S ERRERA A, H ERNAEZ S, study with quantitative and semiquantitative cul - AGUINAGA A, L EIVA J. Effectiveness of the antibiotic tures of catheter hub and segments. J Clin Micro - lock therapy for the treatment of port-related ente - biol 1985; 21: 357-360. rococci, Gram-negative, or Gram-positive bacilli 21) MAH TF, O’T OOLE GA. Mechanisms of biofilm resis - bloodstream infections. Diagn Microbiol Infect Dis tance to antimicrobial agents. Trends Microbiol 2009; 63: 208-212. 2001; 9: 34-39. 34) O’G RADY NP, A LEXANDER M, B URNS LA, D ELLINGER EP, 22) COSTERTON JW, L EWANDOWSKI Z, D EBEER D, C ALDWELL GARLAND J, H EARD SO, L IPSETT PA, M ASUR H, M ERMEL D, K ORBER D, J AMES G. Biofilms, the customized LA, P EARSON ML, R AAD II, R ANDOLPH AG, R UPP ME, microniche. J Bacteriol 1994; 176: 2137-2142. SAINT S; Healthcare Infection Control Practices Ad - visory Committee (HICPAC) (Appendix 1). Sum - 23) F. B ULUT , F. M ERIC , E. Y ORGANCILAR , Y. N ERGIZ , M. A KKUS , mary of recommendations: Guidelines for the Pre - S. N ERGIZ , Y. N ASIR . Effects of N-acetyl-cysteine and vention of Intravascular Catheter-related Infec - acetylsalicylic acid on the tonsil bacterial biofilm tis - tions. Clin Infect Dis 2011; 52: 1087-1099. sues by light and electron microscopy. Eur Rev Med Pharmacol Sci 2014; 18: 3720-3725. 35) BERRINGTON A, G OULD FK. Use of antibiotic locks to treat colonized central venous catheters. J Antimi - 24) S. D INICOLA , S. D E GRAZIA , G. C ARLOMAGNO , J.P. P IN - crob Chemother 2001; 48: 597-603. TUCCI . N-acetylcysteine as powerful molecule to destroy bacterial biofilms. A systematic review. 36) MORALES -M OLINA JA, M ATEU -DE ANTONIO J, G RAU S, Eur Rev Med Pharmacol Sci 2014; 18: 2942- SEGURA M, A COSTA P. Stability: a factor to consider 2948. in antibiotic-lock solutions. Enferm Infecc Microbi - ol Clin 2010; 28: 104-109. 25) GROEGER JS, L UCAS AB, T HALER HT, F RIEDLANDER -K LAR H, B ROWN AE, K IEHN TE, A RMSTRONG D. Infectious 37) POOLE CV, C ARLTON D, B IMBO L, A LLON M. Treatment morbidity associated with long-term use of ve - of catheter-related bacteraemia with an antibiotic nous access devices in patients with cancer. Ann lock protocol: effect of bacterial pathogen. Intern Med 1993; 119: 1168-1174. Nephrol Dial Transplant 2004; 19: 1237-1244.

2445