MANUSCRIPT CITATION

Gilbert R, Brown M, Rainford N, Donohue C, Fraser C, Sinha A, Dorling J, Gray J, McGuire W, Gamble C, Oddie SJ; PREVAIL trial team. Antimicrobial-impregnated central venous catheters for prevention of neonatal bloodstream infection (PREVAIL): an open-label, parallel-group, pragmatic, randomised controlled trial. Lancet Child Adolesc Health. 2019 ;3(6):381-390.  PMID 31040096

REVIEWED BY

Dr Sindhu Sivanandan, DM
Assistant Professor,
Neonatology
JIPMER, Puducherry

TYPE OF INVESTIGATION

Prevention

QUESTION

Among neonates requiring a peripherally inserted central venous catheter (CVC) (P), does the use of miconazole and rifampicin-impregnated central CVC (I) compared with a standard (non-impregnated) CVC (C) reduce bloodstream infection, morbidity, and mortality (O)?

METHODS

• Design: multi-centre, pragmatic, randomised controlled trial
• Allocation: Randomization was done at a central location using a secure web-based program. Permuted blocks of 2 and 4 were used, stratified by site.
• Blinding: Blinding of clinicians could not be done as rifampicin caused brown staining of antimicrobial-impregnated PICC. The outcome assessors were blinded.
• Follow-up period: Neonates were followed until hospital discharge, death, or 6 months after randomisation, whichever occurred first. Follow-up for all deaths was done until 6 months after randomisation.
• Setting: Neonatal intensive care unit (NICU)
• Patients: All neonates requiring narrow-gauge (French gauge 1) peripherally inserted CVC (PICC) for any reason (parenteral nutrition or drug administration) were eligible to participate. Neonates with known allergy or hypersensitivity to rifampicin or miconazole, or if previously enrolled in the study were excluded.
  • The study enrolled 861 neonates of which 430 were randomized to receive miconazole-rifampicin impregnated PICC and 431 to standard PICC.  Most neonates were less than 32 weeks gestation (754 [88%]) and 83% of them were enrolled in the trial before 7 days of age. The median (IQR) gestation ages of the intervention and control group; 27·9 (25·78–29·94) and 28·06 (26·23–30·14) weeks respectively were comparable. The groups were also similar with respect to birth weights; median (IQR) of 962·5 g (729–1220) vs. 960·0g (770–1250), antenatal steroid coverage, receipt of antibiotics or antifungals before randomisation, age of enrollment, type of respiratory support before randomisation, the site of PICC insertion after randomisation and other baseline characteristics.
• Intervention: Neonates in the intervention group received miconazole and rifampicin-impregnated PICC (Premistar; Vygon, Swindon, UK) and control group received a standard non-impregnated PICC (Premicath; Vygon) within 48 hours of randomization. If the allocated PICC was not inserted within this period, the standard PICC was used
• Outcomes
  • Primary outcome: Primary outcome was the time to first microbiologically confirmed bloodstream or cerebrospinal fluid (CSF) infection (termed as blood stream infection) during a time window of 24 h after randomisation and 48 h after PICC removal or death.
  • Secondary outcomes
    • Related to infection: Type of organism isolated from bloodstream infection meeting primary outcome criteria, the rate of bloodstream infection per 1000 days with PICC, number of bloodstream infections, rate of catheter-related bloodstream infection (defined by isolation of the same organism from the PICC tip and blood or CSF) per 1000 days with PICC, rifampicin resistance in any isolate from 1) blood or CSF culture, 2) from PICC tips or both
    • Clinical outcomes: chronic lung disease, necrotizing enterocolitis stage 2 or more, retinopathy of prematurity requiring treatment, abnormalities on cranial ultrasound (periventricular leukomalacia or intracranial hemorrhage of any grade), time from random allocation to full milk feeds (150 mL/kg per day), total duration of parenteral nutrition from random allocation until discharge and death before discharge.
• Sample size and analysis: In an audit from three participating NICUs, the rate of bloodstream infection when using standard PICC was noted to be 14%. Using a log-rank test for equality of survival curves for the primary outcome, with a 5% significance level and 90% power to detect a 50% reduction in blood stream infection (from 14% to 7%) with intervention, 429 neonates were required in each group. This includes an allowance for 5% loss of follow up.
  • SAS software, version 9.4 was used for analysis. Primary and secondary survival outcomes were analysed with the log-rank test. Kaplan-Meier curves were used to present the numbers at risk and Cox regression was used to calculate hazard ratios (HRs). The authors also planned four pre-specified sensitivity analysis to check the robustness of results of primary outcome. The assumptions included; 1) time to serious bloodstream infection, defined as treatment with antimicrobials for 72 h or longer or death during treatment; 2) time from PICC insertion to first bloodstream infection; 3) time to first bloodstream infection, excluding samples obtained via arterial cannulas or CVCs; and 4) time to first bloodstream infection, including clearly pathogenic organisms and excluding skin organisms (e.g., coagulase-negative staphylococci)
• Patient follow-up: Twenty-nine (7%) and 16 (4%) neonates in the intervention and control group respectively did not receive any PICC insertion after randomisation. Twenty-eight (7%) and 8 (2%) in the intervention and control groups, respectively, did not receive the allocated intervention. Four neonates in the intervention group and three in the control group were lost to follow up after primary outcome time window was complete. Analysis was by intention-to-treat technique.
• Safety monitoring: An independent Data Monitoring Committee monitored the trial and an interim analysis of the primary outcome was done after 50% of enrollment. The trial was registered with ISRCTN. The trial protocol and its amendment were published online. The trial was funded by the UK National Institute for Health Research Health Technology Assessment Program who had no involvement in data interpretation or writing of the report.

MAIN RESULTS

The primary and secondary outcomes are listed in Table 1 and discussed below

Efficacy in preventing blood stream infection: The median time from randomisation to the first bloodstream infection (not reported because of the small number of events), as illustrated by the Kaplan Meir curve did not show any difference between antibiotic impregnated and standard PICC groups (HR 1·11, 95% CI 0·73-1·67, p=0·63). The number of neonates having one or more bloodstream infection during the primary outcome window was similar; 46 (11%) in antibiotic impregnated PICC vs. 44 (10%) in standard PICC group; RR (95% CI) 1·05 (0·71–1·55). The results were consistent even after performing sensitivity analyses for various pre-specified assumptions.

Outcomes relating to infection: Rate of bloodstream infection per 1000 days with PICC, rate of catheter-related bloodstream infection per 1000 days with PICC and rate of blood or CSF culture sampling per 1000 days with PICC were similar in two groups.

Clinical outcomes: In-hospital mortality, mortality during follow up, clinical outcomes during NICU stay, and time to PICC removal were similar between the two groups. There was no difference in the duration of antimicrobial treatment and parenteral nutrition or time to achieve full feeds in both the groups.

Organisms and selection of rifampicin resistance:  Among the organisms isolated during the primary outcome time window there was a predominance of coagulase-negative staphylococci in both groups (66/94;70%). Gram negatives constituted 17% and fungi, a mere 3%. Rifampicin resistance in organisms isolated from blood or CSF culture, and rifampicin resistance in either blood or CSF or PICC tip culture rifampicin was similar. However, rifampicin resistance in positive cultures of PICC tips was higher in the antimicrobial-impregnated PICC group (relative risk 3·51, 95% CI 1·16-10·57, p=0·018).

PICC related complications: Adverse events related to PICC insertion like mechanical blockage, thrombophlebitis, extravasation and hematoma were similar in both groups.

Table 1: Primary outcome and relevant secondary outcomes

CONCLUSION

1. Authors conclude that, compared to standard PICCs, miconazole and rifampicin-impregnated PICCs showed no evidence of benefit as indicated by the time to first bloodstream infection in neonates. There was no difference in mortality at 6 months and other morbidity at discharge between the two groups
2. The incidence of rifampicin resistance in positive cultures of PICC tips was higher in the antimicrobial-impregnated PICC group.

COMMENTARY

Neonates admitted to neonatal intensive care units often require central venous catheters (CVC) for parenteral nutrition and drug administration.  Catheter related bloodstream infection (CRBSI) is an important complication related to their use leading to increased mortality and, morbidity.(1,2)  Guidelines recommend the implementation a bundle of interventions during catheter insertion and maintenance to decrease CRBSI.(3) Use of antiseptic/antibiotic impregnated CVCs  had shown promise in reducing the incidence of CRBSI in adults(4) and children.(5) But the evidence is scant to recommend its use in in neonates(6), a high risk population who require catheters for longer duration, and encounter a different profile of micro-organisms.(7)

Gilbert and Colleagues(8) address this gap in a large, multicentric, randomised trial (PREVAIL trial) that compares rifampicin (antibacterial) -miconazole (antifungal)  impregnated PICC to standard PICC in neonates. All neonates requiring a narrow caliber CVC irrespective of indication were randomised. Those with known allergy/hypersensitivity to rifampicin or miconazole were excluded. The study was pragmatic as PICC insertion was done according to unit policy and blood cultures were sent when the treating team suspected infection.

The authors investigated the efficacy (prevention of CRBSI) and safety of antibiotic impregnated PICC, as well the clinical outcomes in neonates. The study showed no difference in the time (from random allocation) to first microbiologically confirmed bloodstream infection (BSI) (including cerebrospinal fluid (CSF) infection) and the proportion of any BSI between the two groups. Rifampicin-miconazole combination is the only antimicrobial-impregnated CVC licensed for use in neonates.

The possible reasons for its lack of efficacy could be; 1) use of an antifungal, when in general, the risk of fungal sepsis is low in NICU. Only 3 out of 94 isolates grew Candida spp. in the PREVAIL study population; two in the antibiotic impregnated group and one in standard PICC group 2) rifampicin is a hydrophobic antibacterial that does not penetrate into gram negative bacteria, thus offering no protection against gram negative sepsis. In this trial coagulase-negative staphylococci was the predominant pathogen isolated in both trial groups. Another concern with rifampicin monotherapy is the rapid emergence of resistance and selective emergence of gram-negative organisms. In the PREVAIL trial, the antibiotic impregnated PICC tips grew more rifampicin resistant organisms than the standard one. However, there was no difference in rifampicin resistance among organisms isolated from blood or CSF; although only a half of these were tested. This study was not powered to address the emergence of rifampicin resistance or the selection of gram-negative organisms.

Other considerations to ponder in this study are; 1) the primary outcome is “time to first microbiologically confirmed bloodstream/CSF infection” rather than its incidence 2) delay of up to 48 hours between the time of randomisation to the insertion of allocated PICC is likely to introduce selection bias. The number of neonates not receiving the allocated PICC or not receiving any PICC at all is higher (57 vs.24) in the antibiotic impregnated arm compared to standard arm,3) while the initial audit suggested a baseline incidence of BSI to be 14%, it was lower ( 10%) in both the arms during the study period. We wonder if this was due to Hawthorne effect due to better implementation of CVC bundle components in the study. 4) the lack of blinding due to brown discoloration of rifampicin coat could have influenced treating team’s decision to send blood cultures, adherence to infection control practices and catheter removal.

The PREVAIL trial has several strengths; it was a large multicenter trial enrolling predominantly preterm neonates below 32 weeks of gestation, an independent data safety monitoring board oversaw the study and several safety outcomes were tracked.   The authors looked at any culture positive blood stream infection as well as CRBSI. Tracking the latter alone might be misleadingly as it may be falsely low due to inhibition of bacteria in antibiotic coated catheter tips.

Although the results do not support the use of rifampicin-miconazole coated CVCs in neonates, it paves the way for future research.  Investigators should explore the efficacy of other antibiotic combinations especially those that cover gram negative organisms, study the intervention in low-middle-income settings where the risk of BSI is high and gram-negative organisms predominate and as an addendum to routine implementation of CRBSI prevention bundle.

REFERENCES

1. Shah DK, Doyle LW, Anderson PJ, et al. Adverse neurodevelopment in preterm infants with postnatal sepsis or necrotizing enterocolitis is mediated by white matter abnormalities on magnetic resonance imaging at term. J Pediatr. 2008;153(2):170-175, 175 e171.
2. Stoll BJ, Hansen N, Fanaroff AA, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics. 2002;110(2 Pt 1):285-291.
3. O’Grady NP, Alexander M, Burns LA, et al. Guidelines for the prevention of intravascular catheter-related infections. American journal of infection control. 2011;39(4 Suppl 1):S1-34.
4. Lai NM, Chaiyakunapruk N, Lai NA, O’Riordan E, Pau WS, Saint S. Catheter impregnation, coating or bonding for reducing central venous catheter-related infections in adults. Cochrane Database Syst Rev. 2016;3:CD007878.
5. Gilbert RE, Mok Q, Dwan K, et al. Impregnated central venous catheters for prevention of bloodstream infection in children (the CATCH trial): a randomised controlled trial. Lancet (London, England). 2016;387(10029):1732-1742.
6. Bertini G, Elia S, Ceciarini F, Dani C. Reduction of catheter-related bloodstream infections in preterm infants by the use of catheters with the AgION antimicrobial system. Early human development. 2013;89(1):21-25.
7. Balain M, Oddie SJ, McGuire W. Antimicrobial-impregnated central venous catheters for prevention of catheter-related bloodstream infection in newborn infants. Cochrane Database Syst Rev. 2015(9):CD011078.
8. Gilbert R, Brown M, Rainford N, et al. Antimicrobial-impregnated central venous catheters for prevention of neonatal bloodstream infection (PREVAIL): an open-label, parallel-group, pragmatic, randomised controlled trial. Lancet Child Adolesc Health. 2019;3(6):381-390.