Objective To evaluate fosfomycin-related adverse events (AEs) and pharmacokinetics and changes in sodium levels in neonates with clinical sepsis.
Between March 2018 and February 2019, 120 neonates aged ≤28 days received standard of care (SOC) antibiotics for sepsis: ampicillin and gentamicin.
Intervention We randomly assigned half of the participants to receive additional intravenous fosfomycin followed by oral fosfomycin at a dose of 100 mg/kg twice daily for 7 days (SOC-F) and followed up for 28 days.
Results 61 and 59 infants aged 0-23 days were assigned to SOC-F and SOC, respectively.There is no evidence that fosfomycin has an effect on serum sodium or gastrointestinal side effects.During the 1560 and 1565 infant-day observation periods, we observed 50 AEs in 25 SOC-F participants and 34 SOC participants, respectively (2.2 vs 3.2 events/100 infant days; rate difference -0.95 events/100 infants) day (95% CI -2.1 to 0.20)).Four SOC-F and three SOC participants died.From 238 pharmacokinetic samples, modeling indicated that most children required a dose of 150 mg/kg intravenously twice daily to achieve pharmacodynamic goals, and for neonates <7 days old or weighing <1500 g daily The dose was reduced to 100 mg/kg twice.
Conclusions and Relevance Fosfomycin has potential as an affordable treatment option for neonatal sepsis with a simple dosing regimen.Its safety needs to be further studied in a larger cohort of hospitalized neonates, including very preterm neonates or critically ill patients.Resistance suppression can only be achieved against the most sensitive organisms, so it is recommended to use fosfomycin in combination with another antibacterial agent.
Data is available upon reasonable request.Trial datasets are deposited at https://dataverse.harvard.edu/dataverse/kwtrp and are available from the KEMRI/Wellcome Trust Research Program Data Governance Committee at dgc@kemri-wellcome.org.
This is an open access article distributed under a Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which allows others to copy, redistribute, remix, transform, and construct this work for any purpose, provided that it is properly cited The original work is given, a link to the license is given, and an indication of whether changes have been made.See: https://creativecommons.org/licenses/by/4.0/.
Antimicrobial resistance poses a threat to the survival of newborns and there is an urgent need for affordable new treatment options.
There is a significant sodium burden with intravenous fosfomycin, and oral fosfomycin preparations contain large amounts of fructose, but there are limited safety data in neonates.
Pediatric and neonatal dosing recommendations for intravenous fosfomycin differ, and there are no published oral dosing regimens.
Intravenous and oral fosfomycin at 100 mg/kg twice daily, respectively, had no effect on serum sodium or gastrointestinal side effects.
Most children may require intravenous fosfomycin 150 mg/kg twice daily to achieve efficacy goals, and for neonates <7 days old or weighing <1500 g, intravenous fosfomycin 100 mg/kg twice daily.
Fosfomycin has the potential to be combined with other antimicrobials to treat neonatal sepsis without the use of carbapenems in the setting of increased antimicrobial resistance.
Antimicrobial resistance (AMR) disproportionately affects populations in low- and middle-income countries (LMICs).The reduction in neonatal mortality was lower than in older children, with at least a quarter of neonatal deaths attributable to infection.1 AMR exacerbates this burden, with multidrug-resistant (MDR) pathogens accounting for approximately 30% of neonatal sepsis deaths globally.2
WHO recommends ampicillin, penicillin, or cloxacillin (if S. aureus infection is suspected) plus gentamicin (first-line) and third-generation cephalosporins (second-line) for the empirical treatment of neonatal sepsis.3 Along with extended-spectrum beta-lactamase (ESBL) and carbapenemase, 4 clinical isolates are often reported to be insensitive to this regimen.5 Retention of carbapenems is important for MDR control, 6 and reintroduction of traditional antibiotics is advocated to address the lack of new affordable antibiotics.7
Fosfomycin is a non-proprietary phosphonic acid derivative that has been deemed “essential” by the WHO.8 Fosfomycin is bactericidal9 and exhibits activity against Gram-positive and Gram-negative bacteria, including methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, ESBL producers and may penetrate biofilm.10 Fosfomycin has shown in vitro synergy with aminoglycosides and carbapenems 11 12 and is commonly used in adults with MDR urinary tract infections.13
There are currently conflicting recommendations for the dosing of intravenous fosfomycin in pediatrics, ranging from 100 to 400 mg/kg/day, with no published oral dosing regimen.Four neonatal studies estimated an elimination half-life of 2.4-7 hours following intravenous administration of 25-50 mg/kg.14 15 Protein binding was minimal, and maximal concentrations were consistent with adult data.16 17 Bactericidal effects were considered to be associated with either time above the minimum inhibitory concentration (MIC) 16 or the area under the curve (AUC):MIC ratio.18 19
A total of 84 case reports of neonates receiving intravenous fosfomycin at 120-200 mg/kg/day indicated that it was well tolerated.20-24 Toxicity appears to be lower in adults and older children.25 However, parenteral fosfomycin contains 14.4 mmol/330 mg sodium per gram—a potential safety concern for neonates whose sodium reabsorption is inversely proportional to gestational age (GA).26 In addition, oral fosfomycin contains a high fructose load (~1600 mg/kg/day), which may cause gastrointestinal side effects and affect fluid balances.27 28
We aimed to assess pharmacokinetics (PK) and sodium level changes in clinically sepsis neonates, as well as adverse events (AEs) associated with intravenous following oral fosfomycin.
We conducted an open-label randomized controlled trial comparing standard of care (SOC) antibiotics alone with SOC plus IV followed by oral fosfomycin in neonates with clinical sepsis at Kilifi County Hospital (KCH), Kenya.
All newborns admitted to KCH were screened.Inclusion criteria were: age ≤28 days, body weight >1500 g, gestation >34 weeks, and criteria for intravenous antibiotics in WHO3 and Kenya29 guidelines.If CPR required, Grade 3 hypoxic-ischemic encephalopathy, 30 sodium ≥150 mmol/L, creatinine ≥150 µmol/L, jaundice requiring exchange transfusion, allergy or contraindication to fosfomycin, specific indication of another class of antibiotics disease, the neonate was excluded from another hospital or not in Kilifi County (Figure 1).
Try out the flowchart.This original figure was created by CWO for this manuscript.CPR, cardiopulmonary resuscitation; HIE, hypoxic-ischemic encephalopathy; IV, intravenous; SOC, standard of care; SOC-F, standard of care plus fosfomycin.*Causes include mother (46) or severe illness (6) after caesarean section, discharge from hospital (3), discharge against recommendation (3), abandonment by mother (1) and participation in another study (1).†One SOC-F participant died after completing follow-up (Day 106).
Participants were enrolled within 4 hours of the first dose of SOC antibiotics until September 2018, when protocol amendments extended this to within 24 hours to include overnight admissions.
Participants were assigned (1:1) to continue on SOC antibiotics alone or to receive SOC plus (up to) 7 days of fosfomycin (SOC-F) using a randomization schedule with random block size (Supplementary Figure S1 online).Concealed by sequentially numbered opaque sealed envelopes.
According to WHO and Kenyan pediatric guidelines, SOCs include ampicillin or cloxacillin (if staphylococcal infection is suspected) plus gentamicin as first-line antibiotics, or third-generation cephalosporins (eg, ceftriaxone) as second-line antibiotics.3 29 Participants randomized to SOC-F also received intravenous fosfomycin for at least 48 hours, switching to oral when adequate feed was tolerated to assume adequate absorption of the oral drug.Fosfomycin (intravenous or oral) was administered for 7 days or until discharge, whichever occurred first.Fomicyt 40 mg/mL fosfomycin sodium solution for intravenous infusion (Infectopharm, Germany) and Fosfocin 250 mg/5 mL fosfomycin calcium suspension for oral administration (Laboratorios ERN, Spain) twice daily with 100 mg/kg/dose administered.
Participants were followed for 28 days.All participants were cared for in the same highly dependent unit to regulate AE monitoring.Complete blood counts and biochemistry (including sodium) were performed on admission, days 2, and 7, and repeated if clinically indicated.AEs are coded according to MedDRA V.22.0.Severity was classified according to DAIDS V.2.1.AEs were followed until clinical resolution or judged chronic and stable at the time of treatment.”Expected” AEs were pre-defined as those expected to be common in this population, including possible deterioration at birth (protocol in Supplementary file 1 online).
After the first IV and first oral fosfomycin, patients assigned to SOC-F were randomized to one early (5, 30, or 60 minutes) and one late (2, 4, or 8 hour) PK sample.An unsystematic fifth sample was collected for participants who were still hospitalized on day 7.Opportunistic cerebrospinal fluid (CSF) samples were collected from a clinically indicated lumbar puncture (LP).Sample processing and fosfomycin measurements are described in Supplementary file 2 online.
We reviewed admission data between 2015 and 2016 and calculated that the mean sodium content of 1785 neonates weighing >1500 g was 139 mmol/L (SD 7.6, range 106-198).Excluding 132 neonates with serum sodium >150 mmol/L (our exclusion criteria), the remaining 1653 neonates had a mean sodium content of 137 mmol/L (SD 5.2).A sample size of 45 per group was then calculated to ensure that the 5 mmol/L difference in plasma sodium on day 2 could be determined with >85% power based on local prior sodium distribution data.
For PK, a sample size of 45 provided >85% power to estimate PK parameters for clearance, volume of distribution, and bioavailability, with 95% CIs estimated using simulations with an accuracy of ≥20%.To this end, an adult disposition model was used, scaling age and size to neonates, adding first-order absorption and presumed bioavailability.31 To allow for missing samples, we aimed to recruit 60 neonates per group.
Differences in baseline parameters were tested using the χ2 test, Student’s t-test, or Wilcoxon’s rank-sum test.Differences in day 2 and day 7 sodium, potassium, creatinine, and alanine aminotransferase were tested using analysis of covariance adjusted for baseline values.For AEs, serious adverse events (SAEs), and adverse drug reactions, we used STATA V.15.1 (StataCorp, College Station, Texas, USA).
Model-based estimates of PK parameters were performed in NONMEM V.7.4.32 using first-order conditional estimates with interactions, full details of PK model development and simulations are provided elsewhere.32
On-site monitoring was performed by DNDi/GARDP, with oversight provided by an independent data security and monitoring committee.
Between March 19, 2018, and February 6, 2019, 120 neonates (61 SOC-F, 59 SOC) were enrolled (Figure 1), of whom 42 (35%) were enrolled before the protocol revision. Group.Median (IQR) age, weight and GA were 1 day (IQR 0-3), 2750 g (2370-3215) and 39 weeks (38-40), respectively.Baseline characteristics and laboratory parameters are presented in Table 1 and online Supplementary Table S1.
Bacteremia was detected in two neonates (Supplementary Table S2 online).2 of 55 neonates who received LP had laboratory-confirmed meningitis (Streptococcus agalactiae bacteremia with CSF leukocytes ≥20 cells/µL (SOC-F); positive Streptococcus pneumoniae cerebrospinal fluid antigen test and CSF leukocytes ≥ 20 cells/µL (SOC)).
One SOC-F neonate incorrectly received only SOC antimicrobials and was excluded from the PK analysis.Two SOC-Fs and one SOC Neonatal withdrew consent – including pre-withdrawal data.All but two SOC participants (cloxacillin plus gentamicin (n=1) and ceftriaxone (n=1)) received ampicillin plus gentamicin on admission.Online Supplementary Table S3 shows the antibiotic combinations used in participants who received antibiotics other than ampicillin plus gentamicin at admission or after a treatment change.Ten SOC-F participants were converted to second-line therapy due to clinical worsening or meningitis, five of whom were before the fourth PK sample (Supplementary Table S3 online).Overall, 60 participants received at least one intravenous dose of fosfomycin and 58 received at least one oral dose.
Six (four SOC-F, two SOC) participants died in hospital (Figure 1).One SOC participant died 3 days after discharge (day 22).One SOC-F participant missed follow-up and was later found to have died on day 106 (outside of study follow-up); data were included through day 28.Three SOC-F infants were lost to follow-up.Total infants/days of observation for SOC-F and SOC were 1560 and 1565, respectively, of which 422 and 314 were hospitalized.
On Day 2, the mean (SD) plasma sodium value for SOC-F participants was 137 mmol/L (4.6) versus 136 mmol/L (3.7) for SOC participants; mean difference +0.7 mmol/L (95% CI) -1.0 to +2.4).On day 7, the mean (SD) sodium values were 136 mmol/L (4.2) and 139 mmol/L (3.3); mean difference -2.9 mmol/L (95% CI -7.5 to +1.8) (Table 2).
On day 2, mean (SD) potassium concentrations in SOC-F were slightly lower than in SOC-F infants: 3.5 mmol/L (0.7) vs 3.9 mmol/L (0.7), difference -0.4 mmol/L ( 95% CI -0.7 to -0.1).There was no evidence that other laboratory parameters differed between the two groups (Table 2).
We observed 35 AEs in 25 SOC-F participants and 50 AEs in 34 SOC participants; 2.2 events/100 infant days and 3.2 events/100 infant days, respectively: IRR 0.7 (95% CI 0.4 to 1.1), IRD -0.9 events/100 infant days (95% CI -2.1 to +0.2, p=0.11).
Twelve SAEs occurred in 11 SOC-F participants and 14 SAEs in 12 SOC participants (SOC 0.8 events/100 infant days vs 1.0 events/100 infant days; IRR 0.8 (95% CI 0.4 to 1.8) , IRD -0.2 events/100 infant days (95% CI -0.9 to +0.5, p=0.59). Hypoglycemia was the most common AE (5 SOC-F and 6 SOC); 3 of 4 in each group 3 SOC-F and 4 SOC participants had moderate or severe thrombocytopenia and were doing well without platelet transfusions on day 28. 13 SOC-F and 13 SOC participants had an AE classified as “expected” (Supplementary Table S5 online). 3 SOC participants were readmitted (pneumonia (n=2) and febrile illness of unknown origin (n=1)) All were discharged home alive. One SOC-F participant had a mild perineal rash and another SOC-F participant had moderate diarrhea 13 days after discharge; both resolved without sequelae. After exclusion of mortality, Fifty AEs resolved and 27 resolved with no change or sequelae resolved (online Supplementary Table S6). No AEs were related to study drug.
At least one intravenous PK sample was collected from 60 participants.Fifty-five participants provided the full four sample sets, and 5 participants provided partial samples.Six participants had samples collected on day 7.A total of 238 plasma samples (119 for IV and 119 for oral fosfomycin) and 15 CSF samples were analyzed.No samples had fosfomycin levels below the limit of quantitation.32
Population PK model development and simulation results are described in detail elsewhere.32 Briefly, a two-compartment PK disposition model with an additional CSF compartment provided a good fit to the data, with clearance and volume at steady state for typical participants (body weight (WT) 2805 g, postnatal age ( PNA) 1 day, postmenstrual age (PMA) 40 weeks) were 0.14 L/hour (0.05 L/hour/kg) and 1.07 L (0.38 L/kg), respectively.In addition to fixed allometric growth and expected PMA maturation based on renal function31, PNA is associated with increased clearance during the first postnatal week.The model-based estimate of oral bioavailability was 0.48 (95% CI 0.35 to 0.78) and the cerebrospinal fluid/plasma ratio was 0.32 (95% CI 0.27 to 0.41).
Online Supplementary Figure S2 illustrates the simulated steady-state plasma concentration-time profiles.Figures 2 and 3 present the AUC Probability of Target Attainment (PTA) for the study population (body weight >1500 g): MIC thresholds for bacteriostasis, 1-log kill, and resistance inhibition, using MIC thresholds from smaller neonates. data to infer.Given the rapid increase in clearance during the first week of life, the simulations were further stratified by PNA (Supplementary Table S7 online).
Probability goals achieved with intravenous fosfomycin.Neonatal subpopulations.Group 1: WT >1.5 kg +PNA ≤7 days (n=4391), Group 2: WT >1.5 kg +PNA >7 days (n=2798), Group 3: WT ≤1.5 kg +PNA ≤7 Days (n=1534), Group 4: WT ≤1.5 kg + PNA >7 days (n=1277).Groups 1 and 2 represented patients similar to those who met our inclusion criteria.Groups 3 and 4 represent extrapolations to unstudied preterm neonates in our population.This original figure was created by ZK for this manuscript.BID, twice daily; IV, intravenous injection; MIC, minimum inhibitory concentration; PNA, postnatal age; WT, weight.
Probabilistic target achieved with oral fosfomycin doses.Neonatal subpopulations.Group 1: WT >1.5 kg +PNA ≤7 days (n=4391), Group 2: WT >1.5 kg +PNA >7 days (n=2798), Group 3: WT ≤1.5 kg +PNA ≤7 Days (n=1534), Group 4: WT ≤1.5 kg + PNA >7 days (n=1277).Groups 1 and 2 represented patients similar to those who met our inclusion criteria.Groups 3 and 4 represent extrapolation of preterm neonates using external data not studied in our population.This original figure was created by ZK for this manuscript.BID, twice daily; MIC, minimum inhibitory concentration; PNA, postnatal age; PO, oral; WT, weight.
For organisms with MIC > 0.5 mg/L, resistance suppression was not consistently achieved with any of the mock dosing regimens (Figures 2 and 3).For 100 mg/kg i.v. twice daily, bacteriostasis was achieved with an MIC of 32 mg/L of 100% PTA in all four mock layers (Figure 2).Regarding 1-log kill, for groups 1 and 3 with PNA ≤7 days, the PTA was 0.84 and 0.96 with 100 mg/kg i.v. twice daily and the MIC was 32 mg/L, but the group had a lower PTA , 0.19 and 0.60 for 2 and 4 PNA > 7 days, respectively.At 150 and 200 mg/kg twice daily intravenously, the 1-log kill PTA was 0.64 and 0.90 for group 2 and 0.91 and 0.98 for group 4, respectively.
The PTA values for groups 2 and 4 at 100 mg/kg orally twice daily were 0.85 and 0.96, respectively (Figure 3), and the PTA values for groups 1-4 were 0.15, 0.004, 0.41, and 0.05 at 32 mg/L, respectively. Kill 1-log under MIC.
We provided evidence of fosfomycin at 100 mg/kg/dose twice daily in infants with no evidence of plasma sodium disturbance (intravenous) or osmotic diarrhea (oral) compared with SOC.Our primary safety objective, detecting the difference in plasma sodium levels between the two treatment groups at day 2, was sufficiently powered.Although our sample size was too small to determine between-group differences in other safety events, all neonates were closely monitored and the reported events help provide evidence to support the potential use of fosfomycin in this susceptible population with sepsis alternative empiric therapy.However, confirmation of these results in larger and more severe cohorts will be important.
We aimed to recruit neonates ≤28 days of age and did not selectively include suspected early-onset sepsis.However, 86% of neonates were hospitalized within the first week of life, confirming the high burden of early neonatal morbidity reported in similar LMICs.33-36 Pathogens that cause early-onset and late-onset sepsis (including ESBL E. coli and Klebsiella pneumoniae have been observed) to empirical antimicrobials,37-39 may be acquired in obstetrics.In such settings, broad-spectrum antimicrobial coverage including fosfomycin as first-line therapy may improve outcomes and avoid carbapenem use.
As with many antimicrobials, 40 PNA is a key covariate describing fosfomycin clearance.This effect, distinct from GA and body weight, represents a rapid maturation of glomerular filtration after birth.Locally, 90% of invasive Enterobacteriaceae have a fosfomycin MIC of ≤32 µg/mL15, and bactericidal activity may require >100 mg/kg/dose intravenously in neonates >7 days (Figure 2).For a target of 32 µg/mL, if PNA >7 days, 150 mg/kg twice daily is recommended for intravenous therapy.Once stabilized, if a switch to oral fosfomycin is required, the dose can be selected based on neonatal WT, PMA, PNA, and likely pathogen MIC, but the bioavailability reported here should be considered.Studies are needed to further evaluate the safety and efficacy of this higher dose recommended by our PK model.
Post time: Mar-16-2022