Background

Neonatal late-onset sepsis

Neonatal late-onset sepsis (LOS) is defined as sepsis occurring 48–72 h after birth. The causative organisms associated with neonatal LOS tend to be those found in the neonatal intensive care unit (NICU) or are normal skin or gut commensals. Coagulase negative staphylococci (CoNS) are the most commonly recovered organisms from the blood cultures of neonates with LOS. Overall mortality rates secondary to CoNS are reported as < 2% in resource-rich settings; however, this increases to ~ 9% in very low birth weight (VLBW) and preterm infants [1,2,3,4]. CoNS-associated LOS has a significant morbidity with episodes of LOS having an adverse effect on neurodevelopmental outcomes as well as increasing the cost of hospitalisation, primarily due to prolonged length of stay [1, 5,6,7,8]. Interventions on the NICU are often invasive, e.g. central venous catheters, and these can increase a neonate’s exposure to hospital-acquired pathogens such as CoNS [1].

Diagnosis of neonatal sepsis

Diagnosing neonatal sepsis can be difficult, particularly as clinical signs of sepsis can frequently overlap with disorders unrelated to sepsis [9]. Blood culture positivity rates in neonates are low [10] and pharmacodynamic (PD) markers, such as C-reactive protein (CRP) can demonstrate variable results in neonates, particularly preterm and VLBW babies [11, 12]. Novel methods for diagnosing neonatal sepsis such as biomarkers based on host RNA are under development. A prospective, case-control study provided clinically validated evidence that host RNA from whole blood can serve as a causally relevant biomarker for detecting neonatal infection [13, 14].

Vancomycin use and emerging antimicrobial resistance

Glycopeptides are the most commonly prescribed antibiotics in children for hospital-acquired infections in Latin America, North America, Australia and Europe, with vancomycin being one of the most widely utilised antibiotics in the treatment of Gram-positive sepsis. Methicillin and gentamicin resistances in CoNS are reported as 85% and in the range of 32%-94%, respectively [15,16,17,18]; vancomycin is, therefore, a first-line therapy in sepsis due to CoNS [19].

It is well-known that high rates of antibiotic use are associated with antimicrobial resistance [20,21,22]. The wide variation in vancomycin dosing and monitoring practices seen across NICUs demonstrates that optimal dosing and monitoring regimens are not well standardised in this population [23] and highly variable between countries and centres [24]. One UK study revealed 17 different combinations of dose, timing of dose and timing of monitoring [25]. Current vancomycin-dosing regimens recommended for preterm neonates, such as those in the European Society for Paediatric Infectious Diseases Manual of Childhood Infection – The Blue Book and the British National Formulary for Children, are based on expert opinion or on studies including only a small number of babies rather than on data generated from clinical trials [26,27,28].

The wider impact of different vancomycin-dosing regimens and course durations on normal, neonatal, gut and skin flora are not known. However, antibiotic use can lead to the disruption of the microbiome leading to colonisation with resistant bacteria and candida [5, 29, 30]. Colonisation with vancomycin-resistant enterococci and candida are known risk factors for invasive infection with associated morbidity and mortality [31, 32].

Trial concept

In view of the lack of validated, neonatal, vancomycin pharmacokinetic (PK), safety and efficacy data to support optimal use, the European Medicines Agency (EMA) included vancomycin on its ‘Revised priority list for studies into off-patent paediatric medicinal products’. The EMA called for more data on the optimal dosing and monitoring regimens for preterm and term neonates with sepsis caused by staphylococci and non-pyogenic streptococci [33]. It was advised that new data should be generated from PK/PD and safety-based efficacy studies. The NeoVanc project has been developed in direct response as a planned drug discovery programme. The discovery and licencing of new antibiotics is a lengthy process; therefore, ascertaining optimal dosing of the older, off-patent vancomycin is essential to ensure the continued efficacy of this important antibiotic in the treatment of neonatal LOS [34].

Methods and trial design

Objectives

The primary objective of the NeoVanc trial is to compare the efficacy of an optimised vancomycin-dosing regimen to a standard vancomycin-dosing regimen in patients with LOS, known or suspected to be caused by Gram-positive microorganisms.

The Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist is provided in Additional file 1.

Study design and selection of optimised dosing regimen

The NeoVanc clinical trial (www.neovanc.org) is an open-label, European, multi-centre, Phase IIb, randomised, active control, parallel group, non-inferiority trial. A total of 300 participants are expected to be enrolled from approximately 20 NICUs in five European Union countries (Estonia, Greece, Italy, Spain and the United Kingdom).

The NeoVanc project forms part of a Paediatric Investigation Plan (PIP) which has received a favourable opinion from the EMA Paediatric Committee (PDCO). The clinical trial, as part of the PIP, aims to provide data on dosing, efficacy and safety of vancomycin in neonates, which will lead to a Paediatric-Use Marketing Authorisation (PUMA) application.

The pre-clinical studies—(1) a hollow-fibre infection model, a rabbit model [35] and a staphylococcal biofilm in vitro model [36]; and (2) a population pharmacokinetic meta-analysis of neonatal, vancomycin studies [37] to define the optimal dosing regimen for the clinical trial—are detailed in Fig. 1.

Fig. 1
figure 1

NeoVanc flow chart demonstrating how the pre-clinical studies, NeoVanc-1 (hollow fibre infection and rabbit models) and NeoVanc-2 (population pharmacokinetics meta-analysis) advised the dosing for the optimised arm of the NeoVanc Clinical Trial (NeoVanc-3)

Full size image

Ethical approval has been gained in all participating European countries as well as approvals from the relevant Regulatory and Ethics Authorities.

Study participants

Infants admitted to the NICU with suspected/confirmed LOS. Participants can be recruited if they fulfil the inclusion criteria and have none of the exclusion criteria.

Inclusion criteria

The inclusion criteria for the clinical trial are adapted from the European Medicines Agency definition for neonatal LOS [38].

  • Postnatal age ≤ 90 days at randomisation

AND

  • Postnatal age ≥ 72 h at onset of sepsis

AND

  • Clinical sepsis as defined by presence of any three clinical or laboratory criteria from the list below, in the 24 h before randomisation

OR

  • Confirmed, significant bacterial sepsis as defined by positive culture with a Gram-positive bacterium from a normally sterile site and at least one clinical or one laboratory criterion from the list below, in the 24 h before randomisation.

Clinical criteria

  • Hyperthermia or hypothermia

  • Hypotension or impaired peripheral perfusion or mottled skin

  • Apnoea or increased oxygen requirement or increased requirement for ventilatory support

  • Bradycardic episodes or tachycardia

  • Worsening feeding intolerance or abdominal distension

  • Lethargy or hypotonia or irritability

Laboratory criteria

  • White blood cell (WBC) count < 4 or > 20 × 109 cells/L

  • Immature to total neutrophil ratio (I/T) > 0.2

  • Platelet count < 100 × 109/L

  • CRP > 10 mg/L

  • Glucose intolerance as defined by a blood glucose value > 180 mg/dL (> 10 mmol/L) when receiving normal glucose amounts (8–15 g/kg/day)

  • Metabolic acidosis as defined by a base excess (BE) < − 10 mmol/L (< − 10 mEq/L) or a blood lactate value > 2 mmol/L

Exclusion criteria

  • Administration of any systemic antibiotic regimen for > 24 h before randomisation, unless the change is driven by the apparent lack of efficacy of the original regimen

  • Treatment with vancomycin for ≥ 24 h at any time within 7 days of randomisation

  • Known toxicity, hypersensitivity or intolerance to vancomycin

  • Known acute renal impairment as defined by urinary output < 0.7 mL/kg/h for 24 h or a creatinine value ≥ 10 μmol/L (1.13 mg/dL)

  • Patient receiving (or planned to receive) haemofiltration, haemodialysis, peritoneal dialysis, extracorporeal membrane oxygenation (ECMO) or cardiopulmonary bypass

  • Severe congenital malformations where the infant is not expected to survive for > 3 months

  • Patient known to have S. aureus (MSSA or MRSA) bacteraemia

  • Patient with osteomyelitis, septic arthritis, urinary tract infection (UTI) or meningitis

  • Patient with high suspicion of/confirmed sepsis caused by Gram-negative organisms or fungi

  • Other situations where the treating physician considers a different empiric antibiotic regimen necessary

  • Current participation in any other clinical study of an investigational medicinal product (IMP)

Post-randomisation exclusions

  • Any participant found to have Gram-negative or fungal sepsis, osteomyelitis, septic arthritis, urinary tract infection, meningitis or S. aureus (MSSA or MRSA) bacteraemia after randomisation will be excluded from the efficacy analysis but followed up for safety.

Dosing schedule by arm

Standard regimen: 15 mg/kg q24h (PMA < 29 weeks), q12h (PMA 29–35 weeks) or q8h PMA > 35 weeks. Duration is 10 ± 2 days. Optimised regimen: all participants receive a single loading dose of 25 mg/kg followed by a maintenance dose of 15 mg/kg q12h (PMA ≤ 35 weeks) or q8h (PMA > 35 weeks). Duration is 5 ± 1 days. Dose adjustments should only occur within the context of routine therapeutic drug monitoring. The only exception is if renal impairment develops, when study vancomycin dose may be adjusted dependent on vancomycin levels/local practice.

Concomitant antibiotics

Use of concomitant antibiotics during vancomycin therapy is allowed except for specific anti-staphylococcal agents (flucloxacillin, oxacillin, linezolid, tedizolid, daptomycin and teicoplanin).

Assignment of interventions

Infants will be randomised in a 1:1 ratio to either the standard treatment regimen or the optimised treatment regimen. Randomisation will be managed via a secure web-based randomisation facility. The randomisation program will use a minimisation algorithm to ensure balance between the groups with respect to clinical centre, PMA at randomisation, and presence or absence of an umbilical catheter/central venous line.

Local investigators and participants’ parents/guardians will not be blinded to regimen allocation as this is an open label study. Investigators undertaking the biomarker and microbiological sub-analyses will be blinded to arm allocation. The trial management group and trial data analysts will be blinded to aggregate outcomes from interim analyses; unblinding will, therefore, not occur.

Primary outcome measure

The primary outcome incorporates clinical and microbiological factors. Successful outcome at the Test of Cure (TOC) visit (10 ± 1 days after the end of actual vancomycin therapy [EVT])

  • Participant is alive

and

  • Successful outcome at EVT (defined as the participant is alive, there is a significant improvement in the participant’s overall, clinical status, there is microbiological resolution or presumed eradication of bacteria and no new vancomycin-susceptible pathogens potentially associated with sepsis have been identified)

and

  • Participant has not had a clinically or microbiologically significant relapse or new infection, requiring treatment with vancomycin or other specific anti-staphylococcal antibiotics (flucloxacillin, oxacillin, linezolid, tedizolid, daptomycin and teicoplanin) for > 24 h within 10 days of EVT visit

Secondary outcome measures

  • Evaluation of efficacy

    • Clinically or microbiologically, significant relapse or new infection within 10 days of EVT, requiring treatment with any antibiotic (other than vancomycin or specific anti-staphylococcal antibiotics [flucloxacillin, oxacillin, linezolid, tedizolid, daptomycin and teicoplanin]) for > 24 h

    • Clinically significant relapse or new infection at short-term follow-up (STFU; 30 ± 5 days from initiation of study vancomycin) visit

  • Evaluation of safety by allocation group

    • Abnormal renal function tests at the STFU visit

    • Abnormal hearing screening test

    • Comparative safety of vancomycin (related to all other parameters other than renal and hearing safety) at STFU visit

  • PK/PD evaluation

    • PK parameters of vancomycin using population PK modelling by allocation group

    • Probability of target attainment with different study regimens

  • Microbiological evaluation

    • Relationship between CoNS species and duration of treatment and CRP response

    • Gut, skin and mucosal colonisation by vancomycin resistant organisms at baseline, EVT and STFU visit

    • Bacterial DNA polymerase chain reaction (PCR) analysis in babies ≥ 29 weeks PMA

  • Biomarker evaluation

    • Assessment of changes in host biomarker panel profiles [13, 14] from baseline to EVT and the relationship between host biomarker and duration of treatment

Study procedures

Recruitment and consent

The parent(s)/guardian(s) of potential participants are approached by research staff to assess eligibility, provide information and obtain informed consent. Informed consent is only gained after information has been provided, questions have been answered and time has been given for consideration. If consent is gained, the informed consent form is signed by the parents/guardians before study recruitment according to the national ethical regulations of the participating countries. Parents/guardians can also pre-consent to the trial. In this case, the same procedure applies; however, parents/guardians are approached before their baby becomes unwell. If pre-consent is in place and a baby develops signs of sepsis and meets the eligibility criteria, the parents/guardians can be contacted via telephone or in person to confirm continued consent.

Consent for the trial as well as the PK and microbiological sub-analyses are gained together; information is provided in relation to all sub-analyses in the participant information sheet and informed consent form. Participants’ parents/guardians will also be asked for permission for the research team to share relevant data with people from the organisations taking part in the research or from regulatory authorities, where relevant. A copy of the informed consent form can be provided by the corresponding author on request.

Visits

Eight visits are planned, from the neonate’s screening and randomisation to the audiology follow-up visit and will be performed as outlined in Table 1 – Schedule of Study Assessments for the NeoVanc Clinical Trial.

Table 1 Schedule of study assessments for the NeoVanc clinical trial
Full size table

Monitoring of clinical and laboratory parameters

Participants will be monitored for specific clinical signs and laboratory parameters in line with the EMA’s criteria for the diagnosis of neonatal sepsis, at day 3, day 5 ± 1 and day 10 ± 2 (standard arm only) [38], so as to determine the rate of recovery. At EVT, an assessment will be made according to specified criteria as to whether the baby has had a significant improvement in their overall, clinical status. Participants who have met this criteria will proceed to the TOC visit where an assessment to determine whether the baby has had a clinically significant new infection (three clinical or laboratory signs), microbiological relapse (positive blood culture with phenotypically similar organism + one clinical/laboratory sign) or microbiological new infection (positive blood culture with an organism of interest + one clinical/laboratory sign) which has required > 24 h of anti-staphylococcal antibiotics since the EVT visit. All participants will undergo hearing screening.

Study-specific assessments

Blood isolates

Blood culture will be collected at baseline and subsequently if positive. Isolates will be stored and sent to the central microbiology laboratory for identification and antimicrobial susceptibility testing. Relevant isolates will be tested for biofilm production and biofilm susceptibility.

Bacterial DNA PCR

Bacterial DNA PCR will be utilised in the identification of causative organisms where the blood culture is negative, as well as quantitative analysis as a PD marker. Bacterial DNA analysis will only take place in those infants who are ≥ 29 weeks PMA in view of the restrictions on the blood volumes allowed in clinical trials. Samples are taken at baseline, Visit 3 and the STFU visit.

Disruption of normal colonising flora

Stool (or rectal swab if stool unavailable) as well as axilla and nasal swabs will be collected at baseline, EVT and at the STFU visit to assess the impact of vancomycin on gut, skin and mucosal flora. Babies will be screened for vancomycin-resistant enterococci, candida and a subset for CoNS with reduced susceptibility to vancomycin, which will all be used as surrogate markers of disruption to normal flora. It will be determined whether colonisation with these organisms changes over the treatment course, as well as between specific groups, e.g. preterm versus term babies.

Biomarkers

A set of biomarkers will be performed as a classifier with high accuracy in predicting specifically bacterial infection. Biomarkers samples are taken at baseline, visits 3, 4, 5 (standard arm only) and the STFU visit.

Sample size

This is a non-inferiority study comparing a standard regimen and an optimised regimen. It is hypothesised that 95% of the participants in the standard treatment arm will not require a second course of treatment with anti-staphylococcal antibiotics and that the proportion will be the same in the optimised treatment group. Using the Wilson-score method, a sample size of 150 babies per arm will give at least 90% of power to demonstrate a non-inferiority margin of 10% (nQuery Software, ‘Statsols’; Statistical Solutions Ltd., Cork, Ireland).

Statistical analysis

Analysis of the primary outcome will be performed using binomial regression with an identity link to report the risk difference and its 95% confidence interval (CI) (risk of a successful outcome at TOC visit in the optimised regimen – risk in the standard regimen). The optimised regimen will be considered non-inferior if the lower limit of the 95% CI is ≥ − 10%. Non-inferiority of the optimised regimen will be tested using a one-sided test at the 2.5% level of significance.

An unadjusted analysis will be carried out as well as an adjusted analysis, adjusting for the minimisation factors used during randomisation by including them as covariates – centre, PMA at randomisation (< 29 weeks/29–35 weeks/> 35 weeks), and presence or absence of an umbilical catheter/central venous line. If there are insufficient participants in some centres to use this as a covariate factor, the centre will be used to adjust the variance estimator by including it as a clustering factor. This analysis will be carried out using both per-protocol (PP) and intention-to-treat (ITT) populations for comparison but the primary inference will be based on the adjusted analysis of the PP population.

For the secondary outcomes, risk ratios and their 95% CIs will be estimated using a log binomial regression model or using a log Poisson regression model with a robust variance estimator if the binomial model fails to converge. The analyses will be carried out for both the PP and ITT populations and, in each case, an unadjusted and adjusted analysis will be carried out as described for the primary outcome.

Pre-specified subgroup analysis will be performed for the primary outcome using the statistical test of interaction, based on PMA at randomisation, birth weight category, and presence or absence of an umbilical catheter/central venous line at the onset of sepsis.

Pharmacokinetics analysis

Pharmacokinetic data will be analysed by nonlinear mixed-effect modelling method (NONMEM). The vancomycin plasma concentration data will be imported into R for exploratory analysis and formatted for subsequent modelling using NONMEM version 6.2 (Globomax, USA). An interim PK analysis will be performed when PK and scavenged samples have been obtained from 50 participants.

Adverse event reporting

All adverse events occurring from the administration of the first dose of study vancomycin to the final follow-up visit will be recorded. In addition, serious adverse events, which meet the criteria for expedited reporting, will be reported to the Sponsor within 24 h of the awareness of the event.

Data management

Data will be collected in an electronic case report form which will be managed by Consorzio per Valutazioni Biologiche e Farmacologiche. All data collected will remain strictly confidential and anonymous. Access to the data is mandated by the Sponsor.

End of trial

The end of trial will be the point of database lockdown.

Committees

Network Management Group (NMG) is an executive body comprising of NeoVanc Work Package leaders and the Trial Statistician with decision powers over the whole NeoVanc project, its technical development and work plan updates. The NeoVanc Trial Management Group is a subgroup of the NMG and comprises of the individuals actively involved in this clinical trial. The group is responsible for managing the operational aspects of the trial. The Independent Data Monitoring Committee (IDMC) will monitor progress, safety and PK data with the aim of protecting the safety and interests of the trial participants and scientific integrity of the trial. The IDMC is composed of a neonatologist, a microbiologist and a statistician. The members of the IDMC are independent and have no involvement with the NeoVanc clinical trial.

Monitoring

Monitoring of the study will be performed by Therakind Ltd. in accordance with the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use Good Clinical Practice guidelines, local regulations and Therakind Standard Operating Procedures.

Audit can be performed at any time by independent, properly authorised individuals mandated by the Sponsor. The audit is aimed to ensure the quality of the research, data accuracy, and respect of laws and regulations.

Specific considerations

Study-specific procedures have been kept to as few as possible and the volumes of blood to be obtained fall well within the guidelines set out by the EMA [39]. Bacterial DNA analysis will only be undertaken in neonates ≥ 29 weeks PMA. Study procedures will be carried out at the same time as routine investigations, where possible, to minimise discomfort.

Neonatal sepsis is serious and requires rapid treatment initiation; therefore, it will not always be possible to gain consent for the study within this timeframe. Up to 24 h of antibiotic administration prior to randomisation has been allowed within the protocol to allow for this. However, pre-consent is also allowed where consent can be gained before a baby becoming unwell, allowing for IMP to be initiated as soon as possible.

Trial sponsor

The trial sponsor is Fondazione PENTA Onlus (Email: info@penta-id.org). The sponsor has not participated in the study design and will not participate in the collection, analysis or interpretation of the data.

Discussion

Clinical trials and neonates

The NeoVanc trial opted to adapt the EMA criteria for the diagnosis for neonatal sepsis, by allowing entry to the trial on three clinical or laboratory signs as opposed to requiring two clinical and two laboratory signs. The reason for this was due to the less severe presentation and milder sepsis course seen with CoNS [1,2,3,4].

The NeoVanc trial is an open-label study. The use of a placebo in neonatal trials is controversial particularly if preterm infants are included in recruitment. In studies evaluating intravenous medication, the use of placebo can be a challenge in view of the careful fluid balance required in this population. Blinding would be difficult in view of the differing course lengths unless a placebo was used for the remaining days in the short course arm. It is recognised by the NeoVanc Consortium that this could lead to ascertainment bias particularly in view of the varying prescribing practices for vancomycin throughout Europe; some clinicians may favour a short course over a long course or vice versa [23, 25]. Investigators will be strongly encouraged to adhere to the allocated vancomycin duration and an assessment of the baby’s clinical status will be undertaken on the day vancomycin is actually stopped as well as on the day vancomycin should have been terminated so as to ascertain if the baby fulfilled the protocol’s definition of recovered/significantly improved.

The NeoVanc trial provides the unique opportunity to evaluate multiple aspects of neonatal, vancomycin pharmacokinetics within one study. Different course lengths will be explored but also the use of a loading dose, in the optimised arm, will help to ascertain if vancomycin levels can be safely increased more quickly and thus allow course duration to be reduced. More frequent dosing in babies < 29 weeks PMA will also be assessed in the optimised arm where babies receive dosing every 12 h instead of every 24 h as in the standard arm.

Finding an acceptable balance between antibiotic resistance, safety, duration and efficacy is a major challenge when treating neonatal sepsis; the NeoVanc trial aims to focus on all four components.

Trial status

The trial opened to recruitment on 3 March 2017; the first participant was recruited on 17 March 2017. The current protocol version is 6.0 (dated 7 September 2016). Recruitment is expected to complete by 31 December 2019.

Dissemination policy

There are no current plans for granting public access to participant-level datasets or statistical code. Outcomes of this trial will be published in a peer-reviewed journal.