Antibiotic Analgesic-Sedative and Antiseizure Medications Frequently Used in Critically Ill Neonates: A Narrative Review

1. Introduction

Preterm and sick neonates admitted to neonatal intensive care units (NICUs) are at high risk of developing hospital-acquired infections which are associated with severe complications, extended need for intensive care, and increased mortality [1,2]. Factors contributing to the high incidence of infections in NICU neonates include their immature defense mechanisms, the multiple-resistant microorganisms colonizing neonates in NICUs, and the invasive procedures they are subjected to [1]. The latter problem exposes the NICU neonates to numerous painful and stressful procedures which may have serious short- and long-term consequences [3,4]. In addition, high risk neonates may suffer central nervous system (CNS) complications, such as peri-intraventricular hemorrhage and hypoxic-ischemic encephalopathy, which may present with seizures. All these severe and life-threatening complications should be prevented or promptly treated. To this aim, evidence-based management protocols should be applied by all neonatologists. However, previous studies and reviews revealed a wide variability in whether and which medications are used in NICUs worldwide [3,5,6]. A major factor contributing to this variability is the lack of adequate studies regarding the effectiveness and safety of medications administered to neonates, especially the preterm ones. Reported differences primarily concern the indications and dosage of the first-line medications [7].

Several studies have explored the pattern of medications used in NICUs in different countries. It was found that antimicrobials – antifungals and analgesics - sedatives are among the most commonly used medications in NICUs [8]. In this review, we address the antimicrobials – antifungals, analgesics – sedatives, and antiseizure medications used in neonatology, focusing on existing evidence or knowledge gaps regarding their indications, dosage, and potential side effects. We believe that this review could be useful in daily clinical practice to all of those involved in neonatal care. Future perspectives for the safe use and development of drugs, specifically for the neonatal population, are also discussed.

2. Method of Literature Review

Electronic databases (PubMed, Scopus, and the Cochrane Library) were searched up to March 2024 for articles on medications for the treatment of neonatal infections, pain/stress, and seizures. Additionally, a manual search of the reference lists of the included studies was conducted to find additional relevant articles. The MeSH terms used included “acetaminophen”, “alpha-2 agonists”, “analgesics”, “antibacterial”, “antibiotics”, “antifungal”, “antimicrobial”, “antiseizure”, “benzodiazepines”, “dexmedetomidine”, “drug therapy”, “fentanyl”, “ketamine”, “levetiracetam”, “midazolam”, “morphine”, “necrotizing enterocolitis”, “neonatal infections”, “neonatal pharmacotherapy”, “neonatal sepsis”, “neonate”, “pain”, “phenobarbital”, “phenytoin”, “phosphophenytoin”, “placebo”, “preterm infants”, “propofol”, “randomized controlled trials”, “remifentanil”, “review”, “sedatives”, “systematic review”, “topiramate”. Studies included full reports in the English language.

3. Antibiotics

3.1. Antimicrobial Agents

Neonatal sepsis, classified as either early- or late-onset, remains an undeniably significant cause of substantial morbidity and mortality in both high and low - middle income countries, although the precise estimates of its burden vary by setting [1,2,9]. Prompt empiric antimicrobial therapy is the “gold standard” regarding the management of neonates with suspected sepsis. Not surprisingly, in 305 NICUs (USA, 2005–2010), 26 of the 100 most commonly prescribed medicines were antibiotics with ampicillin, gentamicin, and vancomycin being at the top of the list; at the first, second, and fifth position, respectively [9]. Both ampicillin and gentamicin have retained their position in the list until recently [8,10]. From a global perspective, according to the NO-MAS-R (no-more-antibiotics and resistance) study, one out of five high-risk neonates receives at least one antimicrobial agent (92%, antibacterial; 19%, antifungal; 4%, antiviral). “Rule-out” sepsis (32%) and “culture-negative” sepsis (16%) were the most common indications for ampicillin (40%), gentamicin (35%), amikacin (19%), vancomycin (15%), and meropenem (9%) administration, while for definitive treatment of presumed/confirmed infection, vancomycin (26%), amikacin (20%), and meropenem (16%) were the most prescribed [11].

Wide inter-NICU variability and off-label use of antibiotic agents continue to characterize the treatment strategies in neonatal sepsis and antibiotic prescribing practices [12,13,14,15,16,17]. Management of neonatal sepsis varies among NICUs in Europe with less homogeneity in the combination of antibiotic regimens for late-onset sepsis (LOS) when compared to early-onset sepsis (EOS). For EOS, neonatologists prefer to prescribe ampicillin, penicillin, gentamicin, and amikacin with the most frequent combination being ampicillin plus gentamicin (54.6%). In cases of LOS, vancomycin (52.4%), gentamicin (33.9%), cefotaxime (28%) and meropenem (15.5%) are frequently chosen [6,18]. High variability in dosage is common in clinical practice, and this variability has been documented in European NICUs. For instance in one study, a total of 444 dosage regimens were used, yet only 41 different antibiotics were administered [19,20]. Limited knowledge exists for either pharmacokinetics (PKs), safety, dosage regimen (dose, dosing interval, route of administration, formulation), or evidence-based guidelines for the optimal antibiotic use in neonatal sepsis. The lack of robust evidence from high quality randomized controlled trials in neonates, especially in those born preterm, has led to the off-label/unlicensed use of anti-infective agents [10,17,21]. It is noteworthy, only 6 antibiotics have been approved by the Food and Drug Administration (FDA) since 1998, with the majority for term infants; ceftolozane/tazobactam, clindamycin, dalbavancin, ceftaroline-fosamil, ampicillin, meropenem, and linezolid [22].

The most recent Cochrane meta-analysis, aiming to assess the effect of different antibiotic regimens for EOS, failed to find superiority of any antibiotic combination due to the low quality of evidence in the included studies (5 trials with 865 participants) [23]. Similar conclusions were drawn regarding the effect of antibiotic regimens for LOS (5 trials) [24].

The general principle is that empirical therapy should be guided by the epidemiology of EOS and LOS, as well as local antimicrobial resistance patterns settings [1]. Thus, active infection surveillance programs should be implemented. EOS is most often caused by group B streptococcus (GBS) followed by Escherichia coli (E. coli), with the latter pathogen predominating in very low birth weight infants (VLBWI). LOS is mainly caused by Gram-positive bacteria, most often coagulase-negative Staphylococcus (CoNS) [18,25]. However, in low-middle income countries Klebsiella pneumonia was found to be the leading cause of neonatal sepsis which stresses the importance of developing local infection surveillance programs [26]. Ampicillin and gentamicin are recommended by the American Academy of Pediatrics (AAP) as the first choice for empirical therapy of EOS. Another important issue neonatologists should be aware of is the increasing ampicillin resistance of E. coli and other Gram-negative pathogens, while the majority of them remains susceptible to gentamicin [27,28,29]. The National Institute for Health and Care Excellence (NICE) recommends the use of benzylpenicillin instead of ampicillin, while a combination of narrow-spectrum antibiotics (such as intravenous flucloxacillin plus gentamicin) is recommended for LOS, as well [30]. Third-generation cephalosporins, such as cefotaxime, or fourth-generation cephalosporins should be preserved for suspected Gram-negative meningitis. Nevertheless, their overuse should be avoided, as their administration has been associated with the emergence of multidrug resistant organisms and candidiasis.

Infections due to extended-spectrum beta-lactamase-producing Gram-negative bacteria require treatment with carbapenems, such as meropenem [31]. Treatment with piperacillin–tazobactam and ampicillin–sulbactam is being used increasingly in NICUs. However, penetration of tazobactam into the CNS is questionable and should not be used for treatment of meningitis. On the other hand, the β-lactamase inhibitor sulbactam, when combined with ampicillin, seems to achieve high concentrations in cerebrospinal fluid [1,32]. The use of vancomycin in the empiric therapy of LOS is based on the predominance of coagulase-negative Staphylococcus, and concerns for methicillin-resistant Staphylococcus aureus (MRSA) infection. Nevertheless, there are several arguments against its empiric use in empiric regimen, due to the risk of ototoxicity and nephrotoxicity. Active surveillance of local pathogen distribution, screening policies for methicillin-resistant Staphylococcus aureus colonization, and the use of oxacillin in empiric antibiotic therapy can reduce vancomycin overuse in VLBWI [25,33].

Written policies for antimicrobial use and auditing are essential for guiding clinical practice. Antibiotic stewardship programs in NICUs can help develop strategies to identify neonates at risk of EOS, optimize antibiotic choice - including dosage and route of administration both for empiric treatment and targeted therapy for definitive LOS, and standardize the decisions regarding the duration of therapy [33,34,35]. The value of such strategies has been recently highlighted in the NO-MAS-R study, since NICUs with specific Antimicrobial Stewardship Programs have lower antibiotic utilization rates regardless of the country’s income level [11]. Such strategies should be implemented in every NICU to prevent the adverse associated with antibiotic overuse and misuse, including the emergence of multi-drug resistant organisms and increased health care costs.

In the era of antimicrobial resistance, neonatal LOS due to multidrug-resistant bacteria has become a significant issue in many NICUs worldwide [36]. In the Neonatal Antimicrobial Resistance Research Network, the resistance rates of Gram-negative isolates to cephalosporins ranged from 26% to 84% and to carbapenem from 0% to 81%, while glycopeptide resistance rates among Gram-positive isolates ranged from 0% to 45% [37]. Higher mortality and morbidity are attributed to multidrug-resistant organisms when compared to non-multidrug-resistant organisms causing neonatal sepsis, with case fatality rates for carbapenem resistant organisms reaching 36% [38,39,40,41]. The Infectious Diseases Society of America updates guidelines for the treatment of infections caused by extended-spectrum beta-lactamase and AmpC beta-lactamase producing Enterobacterales, carbapenem-resistant Enterobacterales and Acinetobacter baumannii, P. aeruginosa with difficult-to-treat resistance, and Stenotrophomonas maltophilia. These guidelines apply to both adult and pediatric patients, although data for the optimal treatment of such infections in children is limited [40,42]. Neonates are not included in these guidance reports. A recent systematic review aimed to identify and appraise the current antimicrobial options for multidrug-resistant bacteria and extensively drug-resistant Gram-negative bacterial infections in neonates. Colistin in combination with other antimicrobials such as meropenem, amikacin, ciprofloxacin, or tigecycline were used for carbapenem-resistant Enterobacterales infections, whereas colistin plus ciprofloxacin were prescribed for difficult-to-treat resistant and extensively drug-resistant P. aeruginosa. Authors concluded the following: (a) in the last two decades, colistin is the most frequently studied and used antimicrobial, especially in low - middle income countries, with variable evidence of efficacy; (b) carbapenems are still the leading antibiotics for extended-spectrum beta-lactamase Enterobacterales; and (c) newer antibiotics, such as beta-lactam agents / beta-lactamase inhibitor combination (e.g., ceftazidime–avibactam infrequently used as salvage therapy), are promising, but data are few and limited to high income countries [43,44]. Indications, dosing regimens, and side effects of antibacterial agents are summarized in Table 1, while more details are presented in the Supplementary Table S1 [25,31,36,44,45,46,47,48,49,50].

3.2. Antifungal Agents

Although fungi are a less frequent cause of neonatal LOS, the EUROCANDY study revealed that neonates are the second most vulnerable age-specific group to invasive candidiasis among pediatric patients ≤18 years old [51,52]. The incidence of neonatal invasive candidiasis has been reduced in the last decade in high-income countries, but the burden of fungal disease is still significant in high-risk neonates, with severe neurodevelopmental sequelae and a high case-fatality rate of 20% up to 50% in extremely low birth weight infants [53]. Meanwhile, the majority of neonates with invasive candidiasis included in the NeoOBS invasive candidiasis sub-study (data from low-middle income countries) had gestational age > 28 weeks (81%) and birth weight > 1000 g (73%) [53,54,55,56]. Differences between NICU level of care, underlying comorbidities, feeding strategies, antifungal prophylaxis use, antibiotic stewardship, and infection prevention and control bundle measures may explain the great variance of the epidemiology of invasive candidiasis in different geographical regions [53,56,57,58,59,60]. While Candida (C.) albicans remains the leading cause, followed by C. parapsilosis, C. tropicalis, C. glabrata and C. krusei, whereas C. auris has emerged as the third most commonly encountered species causing neonatal invasive candidiasis in low - middle income countries [51,56,59,61]. In a recent European 12-week modified point prevalence study (mPPS) which included 26 NICUs, 17 hospitals, and eight countries, the median percentage of neonates receiving antifungal agents per mPPS week across all NICU Level III was 9.6% (range 7.5-11.4%). According to GARPEC-PPS study (Global Antimicrobial Resistance, Prescribing and Efficacy in Neonates and Children -PPS), 43.5% of all antifungal prescriptions were in the extreme preterm neonates (gestational age <28 weeks), whereas antifungals represented 17.3% and 13.4% of all antimicrobial drugs administered in this specific group in high and low - middle income countries, respectively (P=0.26) [62].

Antifungal drugs used against neonatal invasive fungal infections belong to the following 4 classes: polyenes (amphotericin Β deoxycholate [AmB-D], liposomal AmB), triazoles (fluconazole, voriconazole, posaconazole), echinocandins (micafungin, caspofungin, anidoulafungin), and nucleoside analogues (flucytosine) [60]. Prophylaxis and empirical treatment are the main reasons for antifungal administration to neonates, with fluconazole being the most commonly prescribed drug followed by AmB-D. The most common reasons for prophylaxis are prematurity, birth weight <1000g, and central venous catheters. Variability in antifungal usage appears to be an important problem in neonates, with antifungal prophylaxis being frequently prescribed outside of common risk-factor based recommendations, in addition to reports of underdosing of fluconazole [62,63].

3.2.1. Polyenes and Triazoles

Existing guidelines for the treatment of neonatal invasive candidiasis are endorsed by the European Society for Clinical Microbiology and Infectious Diseases (ESCMID, 2012) and more recently (last updated in 2016) by the Infectious Diseases Society of America [64,65]. AmB-D, the oldest of the polyenes, is strongly recommended at a dose of 1 mg/kg, IV, as a first-line therapy, including CNS infections [64]. Due to its effective penetration into the cerebrospinal fluid and urinary tract, fluconazole (12 mg/kg, q24h, IV or orally) is considered a reasonable alternative option, provided the neonate has not been on fluconazole prophylaxis. In cases of meningoencephalitis, it is suggested as a step-down therapy, after determination of in vitro susceptibility and proven clinical response to the initial therapy with AmB-D or liposomal AmB (5 mg/kg/day) [64,66]. It is active against most Candida spp. except for C. krusei and C. glabrata. Due to the reduced renal excretion of the lipid formulation of AmB B, the Infectious Diseases Society of America warns that it should be used with caution particularly in the presence of urinary tract infection and renal fungal balls [64,65]. Robust evidence for the preferred first-line empiric antifungal drug for neonatal invasive candidiasis is lacking, which underlies the existing variability in antifungal prescription practices among NICUs [61]. A higher mortality rate in infants treated with AmB lipid formulations compared to AmB-D or fluconazole was reported by Ascher et al. [67]. This could be attributed to inadequate penetration of the AmB lipid products into the kidneys, understudied dosage in preterm infants, or other unknown factors. In another study the case fatality rate was lower in the fluconazole treated infants (33%) compared to the AmB-D group (45%) and was found to have fewer side effects, despite the use of fluconazole at doses lower than current recommendations suggest for neonates [67,68]. Although nephrotoxicity is a common serious adverse effect of AmB-D therapy in adults, it seems that this is not the case in neonates. It is generally considered that its use in neonates is safer than in older children and adults, albeit clinicians should still monitor renal function [69,70]. An unfavorable impact of AMB-D use in pediatric patients from 13 months of age onwards suggests this age group, and not the neonatal period, as a turning point for increased adverse events [71]. Fluconazole seems to be safe in neonates with rare adverse effects, the most common of which comprises a mild elevation of liver enzymes that reverses after drug discontinuation.

Growing evidence from population PKs has shed more light on the optimal dosing of fluconazole in term and preterm neonates which has influenced expert opinion, and favors the use of higher doses than those labeled in term neonates, namely a loading dose of 25 mg/kg followed by 12 mg/kg/day [61,65,72,73,74]. A physiology-based PK model investigating fluconazole CNS exposure supported the above dosing regimen, as it resulted in a more rapid attainment of the target exposure in plasma and cerebrospinal fluid of preterm infants with CNS infection [75]. Furthermore, as part of the DINO (Drug dosage Improvement in Neonates, NCT02421068) study, a more recent PK study revealed that the actual body weight and serum creatinine are better predictors of fluconazole clearance, and an optimized dosing regimen has been proposed for VLBWI.

Based on a recent meta-analysis of nine randomized clinical trials on the prophylactic use of fluconazole in VLBWI, fluconazole is effective in reducing the Candida colonization rate, incidence of invasive candidiasis, and the in-hospital and infection-attributed mortality, which is similar to what was reported in previous systematic reviews and meta-analyses [66,76,77]. Several prophylactic dosage regimens have been studied, with different dose ranges (from 3 to 6 mg/kg), different dosing intervals (every 24 - 72 hours), and varying duration (4 to 6 weeks) [61]. Given that the higher dose (6 mg/kg) does not increase efficacy, while potentially increasing risk of toxicity and cost, the use of the lowest dose (3 mg/kg) seems to be preferred [78]. Both Infectious Diseases Society of America and European Society for Clinical Microbiology and Infectious Diseases recommend fluconazole prophylaxis in high-risk neonates in NICUs with a high frequency of invasive candidiasis (>10%). When prophylaxis is necessary, an interesting approach is that NICUs may choose to use either the 3 mg/kg or the 6 mg/kg dosing strategy based on local minimum inhibitory concentration (MIC) data, provided that surveillance and in vitro susceptibility testing are implemented (higher dose of 6 mg/kg for Candida spp. with MIC > 2-4 mg/L) (Table 2, Supplementary Table S2) [61].

3.2.2. Echinocandins

The role of echinocandins in neonatal invasive candidiasis is limited to salvage therapy or to situations in which resistance or toxicity preclude the use of AmB-D or fluconazole, and they are not recommended for the treatment of CNS infections in neonates [64]. Micafungin has been approved by the FDA (2019) andEuropean Medicines Agency(EMA) (2016) for use in patients younger than 4 months without meningoencephalitis, with a dosing label of 4 mg/kg/day [79]. Target populations for this dose regimen include stable full-term infants younger than 4 months with line-related candidemia or those with significant toxicities of other antifungal drugs [79]. Micafungin at a dose of 4 to 10 mg/kg/day is recommended for treatment of neonatal candidiasis by European Society for Clinical Microbiology and Infectious Diseases (ESCMID) guidelines. Given the difficulty of safely ruling out Candida meningoencephalitis in premature and critically-ill infants <4 months of age with candidemia and the risk of underdosing, experts appreciate that a substantially higher micafungin dose of at least 10 mg/kg once daily is likely needed for the treatment of candidemia with meningoencephalitis [79]. Micafungin seems to have a safe profile in term and preterm infants. Anidulafungin and caspofungin are not currently approved in neonates. Anidulafungin has approval for infants >1 month of age and caspofungin for infants >3 months of age, due to limited data in neonates (Table 2) [80].

The indications, dosing regimens and side effects of antifungal agents are summarized in Table 2, while more details are presented in the Supplementary Table S2.

4. Analgesics and Sedatives

During the last decades, it has become evident that fetuses and preterm infants, not only feel pain, but are more sensitive and show cardio-respiratory, hormonal, and metabolic stress responses similar to or even more intense than that seen in adults [81]. This is of great importance as neonates receiving intensive care are exposed to numerous painful and/or stressful procedures [82]. Moreover, cumulative evidence suggests that prolonged exposure to painful events in the neonatal period is associated with significant long-term consequences [83,84]. In this context, international scientific societies have provided guidelines for the prevention and management of pain and stress in neonates [85,86,87].

4.1. Analgesic Medications

4.1.1. Opioids

The opioids morphine, fentanyl, and remifentanil are the most frequently employed analgesic agents in neonates [4,8]. Typically, they are used for procedural or post-operative pain, either as monotherapy or in combination with other drugs [88,89]. Overall, the use of specific pain scales in neonates revealed that opioids reduce procedural pain. However, there is a considerable uncertainty regarding the relationship of opioids with episodes of bradycardia, hypotension, or severe apneas [90]. Therefore, the use of standardized protocols for pain management have been suggested to minimize the exposure to opioids [87,91].

4.1.1.1. Morphine

Morphine is a commonly used medication for analgesia in neonates. It has a slow onset of action (mean onset in 5 min.), reaches peak effect in 15 minutes, and is metabolized in the liver via glucuronidation, oxidation, and sulfidation [89]. The glucuronide biproducts of morphine include morphine-3-glucuronide and morphine-6- glucuronide. The latter metabolite has strong affinity with the morphine receptor and thereby possesses analgesic properties and augments the analgesic effect of morphine [92,93]. The main indication for morphine analgesia is invasive mechanical ventilation in preterm neonates, while it is not recommended for procedural pain in non-ventilated neonates, such as examination for retinopathy of prematurity [94,95]. Variable dosing regimens have been recommended by various authors and scientific societies. Early studies by Quinn et al. and Chay et al., based on available PK studies, recommended morphine dosing regimens with of loading doses of 100 and 150 mcg/kg/h, respectively, for about 2 hours followed by a continuous IV infusion of 25 or 22.5 mcg/kg/h, respectively [96,97]. Based on these studies, Saarenmaa et al. administered morphine in ventilated neonates at a loading dose of 140 mcg/kg over one hour followed by a continuous IV infusion of 20 mcg/kg/h for at least 24 hours [93,97]. Five years later, Anand et al., in a study of 898 ventilated neonates from 16 centers, administered morphine at a dosing regimen based on PK studies available at the study-time [94]. The treatment protocol included a morphine loading dose of 100 mcg/kg in IV infusion over 1 hour, followed by continuous infusions ranging from 10 to 30 mcg/kg/h depending on GA. No increase in early neurological adverse effects was associated with morphine use, except for the hypotensive infants and the extremely low birth weight infants who received high doses (> 10 mcg/kg/h) [94]. Reported acute adverse effects include respiratory depression, miosis, hypotension, constipation, increased biliary pressure, urinary retention, tolerance and withdrawal [89,94]. Results of the NOPAIN pilot trial suggested that morphine administered prophylactically in ventilated preterm infants may improve neurologic outcomes [98]. However, as documented in a following randomized clinical trial (NEOPAIN), pre-emptive analgesia with morphine did not decrease the composite outcome of death, severe intraventricular hemorrhage (IVH), or periventricular leukomalacia in preterm neonates but intermittent boluses of morphine actually increased the incidence of the composite outcome [94]. Overall, the results of these studies in preterm infants warrant extreme caution with the use of this morphine during key stages of brain development, especially in extremely low birth weight infants. It should be noted, though, that it is difficult to solely attribute the adverse outcomes to morphine (and other opioids), as low GA, and pre-existing arterial hypotension are also associated with IVH [84,89,99,100,101]. Interestingly, the analgesic effectiveness of morphine in preterm neonates was questioned, as well. Results of the Procedural Pain in Premature Infants (POPPI) study showed no beneficial effect of morphine on procedural pain in preterm neonates, whereas a higher number of morphine treated neonates required non-invasive ventilation due to apneas compared to the placebo group [95].

4.1.1.2. Fentanyl

Fentanyl is a synthetic opioid characterized by a rapid onset of action, within 1-2 minutes when administered intravenously, and an intermediate duration of action (30 minutes) [102]. These properties have made this medication a suitable opioid for acute, short lasting procedural pain [89,93]. A randomized control trial comparing fentanyl to morphine as analgesia for ventilated neonates showed equal efficacy, with fentanyl having fewer side effects [93]. As a result, fentanyl has become the most commonly used analgesic-sedative medication in many NICUs [4,89,93]. PK studies for fentanyl use in neonates and children are scarce. A comprehensive review by Ziesenitz et al. showed significant age-related changes, and a great variability in fentanyl kinetics, in preterm infants following bolus or continuous intravenous infusion, with a clearance ranging widely between 3.4 to 58.7 mL/min/kg [103]. The dosing regimens used in studies in neonates included a loading dose ranging from 5 to 12.5 mcg/kg followed by a continuous infusion of 0.5 to 2.0 mcg/kg/h [104].^.A recent PK study suggests a quickly increasing clearance threshold within the first three postnatal weeks in preterm infants allowing for a reduced infusion dose by 50% and 25% on postnatal days 0-4 and 5-9, respectively. These results support a decrease in fentanyl dose regimen potentially mitigating some of the adverse effect of fentanyl [105]. Schofer et al. suggested a dose of 3 mcg/kg of fentanyl IV three minutes prior to intubation. Doses greater than 5 mcg/kg have been associated with increased incidence of hypotension [106]. Serious adverse effects of fentanyl include dose-dependent respiratory depression, chest wall rigidity, and arterial hypotension [106]. Of note, in a retrospective study involving very low gestational age infants, fentanyl was independently associated with the need for inotropes [99]. Nevertheless, the effect of fentanyl on blood pressure in neonates is still unclear. A recent Cochrane review including 13 independent studies (enrolling 823 newborn infants) concluded that opioids probably are more effective in reducing pain scores than placebo. However, existing evidence cannot clarify the effect of opioids on episodes of bradycardia, hypotension, or apnea [90,103]. Thus no definite conclusion could be reached concerning the potential effect of opioids, in general, or fentanyl per se, on blood pressure in preterm and term infants [90,103]. A long-term study of infants treated with fentanyl early after birth did not find any significant correlation between cumulative fentanyl exposure and neurodevelopmental outcomes at five years of age [107]. These results were consisted with previous reports in VLBWI at the age of two years [108].

4.1.1.3. Remifentanil

Remifentanil is another synthetic opioid and a selective morphine-receptor agonist, with rapid onset and ultra-short duration of action due to a quick degradation by nonspecific plasma and tissue esterases [109]. In neonatology, it has been used for brief procedural analgesia, i.e., prior to intubation, and prolonged sedation/analgesia with relative safety [89,110,111]. When using remifentanil (1 to 3 mcg/kg) as a single premedication for INSURE, a faster infusion over 30 seconds (compared to over 60 seconds) was associated with higher incidence of chest rigidity (43%) and shorter duration of sedation [112]. Due to the risk of chest wall rigidity, it should only be used in intensive care units with strict monitoring capabilities. Long-term consequences of remifentanil administration in neonates are unknown and, therefore, its optimal use in the neonatal population warrants further studies [89,109,111,112,113,114].

4.1.2. Non-Opioid Analgesics

In the context of the existing controversy on the effectiveness and safety of opioids, alternative analgesics, mainly paracetamol (acetaminophen), have been used for treatment of postoperative and procedural pain in neonates. It exerts its central analgesic effect via activation of descending serotonergic pathways, and inhibition of prostaglandin synthesis [115]. Pooled analyses of PK results suggest an oral dose of 25 mg/kg/day in preterm neonates at 30 weeks, 45 mg/kg/day at 34 weeks, 60 mg/kg/day in term neonates, and 90 mg/kg/day at 6 months of age [116,117]. Other authors recommend a loading dose of 20 mg/kg, followed by 10 mg/kg every 6 hours for 32-44 weeks’ neonates. For neonates < 32 weeks, a loading dose of 12 mg/kg and a maintenance dose of 6 mg/kg every 6 hours is recommended [118]. The safety profile of paracetamol might have contributed to its increasing use in term and preterm neonates, despite the off-label use in this population [116,119]. A recent meta-analysis and a review showed that existing data are not sufficient to support a role for paracetamol in reducing procedural pain in neonates, but may reduce the need for morphine following major surgery [115,120].

4.2. Sedatives

4.2.1. Benzodiazepines

Benzodiazepines effectively relieve patients’ stress, but they exert no analgesic action. Therefore, they are mainly used as an adjunct to analgesics and may rarely be used for in minor procedures [121]. Midazolam is the most common sedative utilized by the NICU physicians for sedation [4,8]. Due to its pharmacological advantages (lack of active metabolites), it has replaced diazepam [122]. Midazolam is a short-acting benzodiazepine that possesses sedative and anticonvulsant properties [121]. The advantages of midazolam over other sedatives are the rapid onset of action and fast termination of effects [123]. Various dosing regimens have been recommend, but one report suggested that when midazolam is used as the only sedative agent, the optimal dose was of 209 mcg/kg/hour (range: 100 to 500 mcg/kg/hour) [121]. However, the limited existing data raises significant concerns about its safety when given as continuous infusion for sedation in infants receiving intensive care [121].

Adverse effects reported in neonates include respiratory depression, hypotension, and a decrease in cerebral blood flow. Moreover, paradoxical agitation, such as hyperexcitability and myoclonus have been reported, which be attributed to the low number of GABA-A receptors in the neonate [124]. As documented in the most recent Cochrane review (2017) which included only three trials and 148 neonates, the duration of NICU stay was significantly longer in the midazolam group than in the placebo group. Moreover in one of the included studies, the incidence of a combined adverse outcome of death, grade 3 or 4 peri- intraventricular hemorrhage, or periventricular leukomalacia at 28 days’ postnatal age was significantly higher in the midazolam group compared with the morphine group [98,125]. Consequently, the limited existing data raises significant concerns about its safety when given as continuous infusion for sedation in infants receiving intensive care.

4.2.2. Ketamine

Ketamine is an analgesic and anesthetic (analgo-sedative) agent that has been used more in children than in neonates. It is an antagonist of n-methyl D-aspartate (NMDA) receptor, gamma-aminobutyric acid (GABA), and other brain receptors, and has a rapid action [126]. It is indicated as analgesia for minor procedures in neonates, such as cannulation for extracorporeal membrane oxygenation [127]. It has also been administered for endotracheal suctioning at a dose of 2 mg/kg [127]. Several administration routes have been used (intravenous, intramuscular, or intranasal). Recommended analgesic doses are 0.15–0.25 mg/kg, IV or 0.5–1 mg/kg intramuscularly [126]. Overall, it is believed to offer cardiorespiratory stability, as ketamine causes mild increases in BP and heart rate while having minimal effects on cerebral blood flow. Moreover, it decreases respiratory drive and induces bronchodilation [127]. Few perioperative complications and satisfactory operative conditions were reported with ketamine analgesia compared to general anesthesia during laser treatment for retinopathy of prematurity in a small group of infants [128]. In the larger “NOPAIN-ROP” randomized trial, IV fentanyl versus IV ketamine for pain relief during laser photocoagulation for retinopathy of prematurity in preterm infants was tested. With both drug regimens, adequate analgesia was provided only in a minority of infants [129]. Concerns regarding potential ketamine-mediated neurotoxicity in the immature brain give pause to the use of ketamine infusion as a therapeutic option for refractory neonatal seizures [130].

4.2.3. Propofol

Propofol is a highly lipophilic compound that is rapidly distributed from blood to subcutaneous fat and the CNS with subsequent redistribution and metabolic clearance. It positively modulates the inhibitory function of the neurotransmitter GABA through the ligand-gated GABAA receptors. It is a short acting sedative (without analgesic properties) that is rapid in its onset and short in duration. These properties make propofol attractive for short-duration interventions. On the other hand, due to the reduced clearance capacity, both preterm and term neonates in the first week of postnatal life are at risk for accumulation following propofol administration [131]. In a recent prospective trial investigating the optimal propofol dose in neonates requiring non-emergency endotracheal intubation, effective sedation without side effects was reported as “difficult to achieve”; optimal result was observed more often with the high starting dose of 2.0 mg/kg compared to the lower doses of 1.0 and 1.5 mg/kg. However, propofol-induced hypotension occurred in 59% of patients, and this risk was found even with low initial doses [132]. In a following study, the profound and prolonged decrease in blood pressure following propofol administration as premedication for intubation in neonates was mainly attributed to its starting dose rather than the cumulative dose [133]. There is only one Cochrane review on propofol for procedural sedation/anesthesia in neonates and it was published more than a decade ago. No practice recommendation could be made at that time [134]. As well, there has been no updated review which reflects the lack of available evidence regarding the use of propofol in neonates.

4.2.4. Alpha-2 Agonists

Centrally acting alpha-2 agonists, such as clonidine and dexmedetomidine, possess sedative, analgesic, and anxiolytic properties. The two main adverse effects of alpha-2 agonists are bradycardia and hypotension. However, contrary to opioids, they do not cause significant respiratory depression. Due to this advantage, alpha-2 agonists have been used in critically ill children as adjunctive sedative agents alongside opioids benzodiazepines and help to minimize the use of benzodiazepines and opioids in children and prevent withdrawal syndrome [135]. There are few data on the “off-label” use of clonidine and dexmedetomidine in neonates. The most recent (2017) relevant Cochrane review included only one small trial (112 neonates) comparing clonidine with placebo. Although sedation-pain scale values were lower among treated infants, clonidine was not associated with reduce death, duration of mechanical ventilation, or duration of stay in the NICU [136]. Interestingly, there are no studies regarding the use of clonidine for the prevention or treatment of procedural and postoperative pain, or pain associated with clinical conditions in non-ventilated neonates [137]. Clonidine has been used for neonatal abstinence syndrome, as well. In a systematic review of randomized clinical trials, clonidine was found to be more efficacious than morphine with respect to the duration of treatment, and better than phenobarbital in reducing morphine treatment days [138].

Like clonidine, dexmedetomidine was reported to be effective in achieving sedation/analgesia in neonates and reduce the need for adjunctive sedative or analgesic agents. Furthermore, it was found to decrease the time to extubation as well as the duration of mechanical ventilation [139]. It is worth noting that during the last several years, dexmedetomidine has attracted the interest of neonatologists as an analgesic/sedative agent in neonates undergoing therapeutic hypothermia due to its possible neuroprotective action [140,141] However, so far, there are insufficient data from randomized clinical trials evaluating the use of any analgesic-sedative agent during therapeutic hypothermia, including clonidine and dexmedetomidine [142]. Dexmedetomidine has been used in neonates at a starting dose of 0.2 to 0.3 mcg/kg/h, escalating in 0.1 mcg/kg/h increments depending on pain / sedation assessment scores, up to a median maximum dose of 0.5 mcg/kg/h (Table 3) [143].

The suggested dosing regimens of analgesics and sedatives which have been most frequently used in neonates are summarized in Table 3.

5.1. Phenobarbital

Phenobarbital (or phenobarbitone) is a barbiturate with antiseizure and hypnotic/sedative properties. Phenobarbital sodium (powder for injection) is the first and only medication approved by the FDA for the treatment of neonatal seizures [150]. Phenobarbital acts by increasing GABAA mediated inhibition of GABA [150]. The drug is metabolized in the liver and excreted in urine. Regardless of the seizure etiology, phenobarbital is the most frequently used ASM in the NICU setting [145].

According to the recent recommendations by The Neonatal Task Force of the International League Against Epilepsy (ILAE), phenobarbital should be the first-line ASM (evidence-based recommendation) regardless of etiology (expert agreement), unless channelopathy is likely the cause for seizures (e.g, due to family history), in which case phenytoin or carbamazepine should be used. Phenobarbital should be given at a loading dose of 20 mg/kg IV, followed by a maintenance dose of 5 mg/kg/day IV or orally. A second loading dose at 10-20 mg/kg could be administered IV if required [149]. A 40% response was reported after the initial loading dose, but sequential administration of IV phenobarbital lead to improved control of seizures in term and preterm neonates [150]. Based on the results of a retrospective population PK study across a pediatric population including neonates, an initial phenobarbital dose of 30 mg/kg has the highest probability of attaining a therapeutic concentration at seven days. Moreover, postmenstrual age and drug-drug interactions should be incorporated into dosing regimen [160]. Monitoring of circulating levels (target concentration: 20-40 mcg/mL) should be considered if on maintenance therapy [149,160]. In asphyxiated neonates with seizures undergoing therapeutic hypothermia, a second loading dose of 20 mg/kg has been suggested for better seizure control. Nevertheless, prophylactic administration of phenobarbital before the onset of hypothermia is not recommended [151].

The most common adverse effects of phenobarbital include respiratory depression, hypotension, depressed consciousness, somnolence, and poor feeding [145,150]. Moreover, the fact that phenobarbital may cause apoptotic neurodegeneration in the developing rat brain at plasma concentrations relevant for seizure control in humans is alarming [161].

5.2. Phenytoin/Fosphenytoin

Phenytoin and fosphenytoin (the phosphorylated prodrug of phenytoin) [152] are still considered as second-line ASMs for most seizure etiologies in neonates not responding to phenobarbital [162]. However, its use declined during the last two decades and has been surpassed by levetiracetam as the second most widely used antiseizure medication in NICUs [163].

With respect to seizure control, two studies evaluating phenytoin and fosphenytoin in neonates, demonstrated similar efficacy to phenobarbital (45% and 56% became seizure free after initial treatment with phenytoin and fosphenytoin, respectively) [164,165]. Ιn a recent meta-analysis, high uncertainty was expressed about the effect of phenobarbital compared to phenytoin on achieving seizure control after a maximal loading dose of the ASM [166].

A loading dose of phenytoin/fosphenytoin (20 mg/kg phenytoin equivalent IV over 30 min) is initially administered followed by maintenance (5 mg/kg/day IV or orally in two divided doses). Adjustments are made according to response (max. per dose 7.5 mg/kg) and plasma concentration (target level: 10–20 mcg/mL). Phenytoin has poor oral bioavailability and limited hepatic metabolism capacity. Low albumin levels and bilirubin displacement by phenytoin from its protein-binding sites may result in increased serum free levels of bilirubin. Levels may also increase in infants receiving therapeutic hypothermia [162].

Careful cardiac monitoring is needed during and after administering IV phenytoin/fosphenytoin because of the risk of severe hypotension and cardiac arrhythmias. Other common adverse reactions are infection, site irritability/necrosis, hypotonia, and respiratory depression/arrest [162]. Fosphenytoin is preferred over phenytoin due to lower risk of adverse effects [152]. Similar to phenobarbital, there are experimental data indicative of apoptotic neurodegeneration with phenytoin [161].

5.3. Levetiracetam

Levetiracetam is an FDA-approved ASM for adults and infants older than one month postnatal age since 2012, mainly used in combination with other anticonvulsive agents. According to a prospective cohort study, it is used off-label in neonates, being the next most commonly used ASM for neonatal seizures following phenobarbital [145]. A reduction by 50-88% in the frequency of seizures without serious side effects has been reported using levetiracetam [167,168]. Moreover, when used as a first-line antiepileptic drug for neonatal seizures, levetiracetam achieved better control than phenobarbital [153]. However, in another recent phase IIb study, phenobarbital was found to be more effective than levetiracetam for the treatment of neonatal seizures (80% versus 26%, respectively), although more adverse effects were seen in subjects assigned to phenobarbital [169]. This finding is also supported by a recent (2023) Cochrane meta-analysis; it was concluded that phenobarbital as a first-line ASM is probably more effective than levetiracetam in achieving seizure control after the first loading dose and after the maximal loading dose (moderate-certainty of evidence) [166]. It should be noted that levetiracetam might have a neuroprotective effect on the neonatal brain by reducing neuronal apoptosis, as shown in animal models [170].

The ILAE, suggested that levetiracetam should be administered as follows: a loading dose of 40 mg/kg IV, followed by a second loading dose of 20 mg/kg IV if required, and maintenance dose of 40–60 mg/kg/day IV or orally in three divided doses [162]. However, higher loading doses (60 mg/kg) may be more effective and substantially decreased the seizure burden [169]. Moreover, high doses seem to be well tolerated. Specifically, in a neonatal study, the initial cumulative dose of levetiracetam ranged from 50 to 100 mg/kg [171], while in another study involving infants suffering hypoxic-ischemic encephalopathy, the mean total and maintenance doses of levetiracetam were 63 and 65 mg/kg/day, respectively [172]. In clinical practice, transition from IV to oral administration should be done, when feasible, and at equivalent doses. Levetiracetam is mainly excreted from the kidneys and to a lesser degree from the liver resulting in fewer interactions with other drugs [146,169].

5.4. Midazolam

The use of midazolam as sedative in ventilated infants is discussed elsewhere in this article. Midazolam was approved in 2022 by the FDA for rescue treatment in adults with status epilepticus, but not in infants and neonates. However, this non-FDA approved benzodiazepine has been utilized for the management of refractory neonatal seizures as well [123,155]. Castro-Conde et al. reported that midazolam effectively controlled EEG-confirmed seizures in all non-responders (n=13) to phenobarbital/phenytoin, significantly improving their long-term neurodevelopment [155]. The authors proposed an antiseizure dosing regimen comprising 0.15 mcg/kg IV bolus, followed by a continuous infusion of 1 mcg/kg/min, increasing by 0.5 to 1 mcg/kg every 2 min, up to a maximum of 18 mcg/kg/min. In other studies, a lower efficacy was reported [173,174].

In the suggested by the ILAE treatments for the management of neonatal seizures, midazolam is included in the second-line ASM options. Its administration includes a loading dose of 0.05–0.15 mg/kg, followed by continuous infusion of a maintenance dose (1 mcg/kg), which may be titrated up in steps of 1 mcg/kg/min to a max. of 5 mcg/kg/min) [162]. Midazolam may cause respiratory depression, hypotension, poor feeding as well as dyskinetic movements and myoclonus [125,162]. Nevertheless, most worrisome is its potential effect on brain development and ultimately neurodevelopment. Animal data showed that benzodiazepines may induce apoptotic neurodegeneration of the developing brain [161], although these concerns have not been validated in human studies [175].

5.5. Topiramate

Topiramate, an anticonvulsant agent widely used in adults and children, is characterized by good absorption, high bioavailability, and good tolerability [176]. Due to the encouraging experimental evidence as neuroprotective agent (through several mechanisms including glutamate-receptor inhibition), topiramate has received additional interest as an adjunct treatment in neonates undergoing therapeutic hypothermia [156]. In the “NeoNATI” feasibility study, 21 neonates were allocated to hypothermia plus topiramate at a dose of 10mg/kg once daily for the first three days of life, and 23 to hypothermia alone. Co-administration of topiramate was safe and associated with a reduction in the prevalence of epilepsy while did not reduce the frequency of the combined outcome of mortality or severe neurological disability [177].

Investigators also evaluated the influence of hypothermia on topiramate PKs. In an early study by Filippi et al. (2009), 13 neonates undergoing whole body hypothermia were given oral topiramate (5 mg/kg once a day for the first three days of life) while seven had concomitant phenobarbital treatment. In most neonates (11/13), topiramate concentrations were within the reference range for the entire treatment duration [157]. A therapeutic range of 5-20 mg/L has been proposed for epilepsy therapy [158,159]. A more recent study evaluated topiramate PKs in hypothermic neonates. Aiming to decrease seizure events, topiramate was given at 5 mg/kg on day 1, and at 3 mg/kg on days 2-5. Based on PKs, the authors suggested an optimized alternative dosage regimen consisting of a loading dose of 15 mg/kg for cycle one and maintenance doses of 5 mg/kg for the following four cycles, to allow topiramate concentrations to be within the therapeutic range after the first dose in more than 90% of cooled neonates [159].

6. Future Perspective and Conclusions

In every-day clinical practice, neonatologists use several drug categories, often with very little proof of evidence as for their safety and efficacy. This fact is indicative of important gaps in knowledge in neonatal pharmacology and the need for new treatments. Therefore, there is an urgent need for clinical trials, in which the diversity of neonates and their medical conditions will be taken into consideration. Only then, will we be able to identify the most effective (old or new) medication, and at the same time decrease the risk of adverse effects, primarily concerning the developing brain, thus improving neurodevelopmental outcomes. Additionally, scientific information derived from the application of new technologies (e.g., genetics, -omics) along with the analysis of pharmacometric parameters and clinical-imaging data, may allow for a more precise use of medications in the neonatal population. Although improvement have been made in legislation, infrastructure, and clinical trial methodology during the last decades ensuring a safer framework surrounding neonatal drug development, more work is still needed [178].