**1. Introduction**

Neonates are not small adults, and neither can they be classified as small children when it comes to medicinal products and their formulation development.

Neonates include term, post-term and preterm babies. The neonatal period for term and post-term newborn infants is defined as the day of birth plus 27 days. The neonatal period for preterm newborn infants is defined as the day of birth through to the expected date of delivery plus 27 days [1].

Each year, some 15 million babies arrive in the world, more than one in 10 babies are born prematurely, according to the report 'Born Too Soon: The Global Action Report on Preterm Birth' (2012) [2]. Even if born at term and ready to grow outside of the mother's womb, most organs and their functions are still immature. This immaturity of organ and function is more profound and impactful in preterm infants. For example, neonates have reduced gastric emptying, intestinal transit time and surface area, and transporter immaturity, which have relevance for oral drug delivery. Additionally,

the skin barrier may not be fully formed, and respiratory function may be immature. A host of other physiological factors such as gastro intestinal (GI) pH, body surface to volume ratio, body fat to lean tissue ratio are also di fferent and are known to change rapidly with time [3].

The underlying complexity of pharmacology and biopharmaceutics in neonates, especially in preterm babies, can lead to altered and variable pharmacokinetics and pharmacodynamics even compared to that in young babies [4–6]. In turn, this can lead to potential lack of e fficacy or reduced safety of a medicinal product leading to additional special requirements for development of products for these age groups. This includes (amongst others) formulation aspects, dosage form choice, drug administration considerations and storage and handling advice.

The recent European Medicines Agency (EMA) report to the European Commission on the experience acquired as a result of the first 10 years application of the Paediatric Regulation, acknowledged that neonates still represent a particularly neglected paediatric subpopulation in the development of medicines despite the regulatory push [7].

In paediatric patients, there is still significant o ff-label use of medicines due to a lack of medicines developed and authorised for the specific needs of the very young [8,9]. This remains an even more significant problem in the neonatal population, due to the di fficulty in conducting the necessary clinical trials in vulnerable subsets with lower patient numbers. Considering the extra challenges they present, pharmaceutical companies are not incentivised to formulate medicines for the neonatal population. In fact, trials open for recruiting neonates were included in only a quarter of all agreed paediatric investigation plans (PIPs), often at the request of the EMA Paediatric Committee (PDCO), due to an array of reasons: Lack of neonate specific indications, recruitment/enrollment challenges, lack of incentives [10]. There is a 'Catch 22 situation because to protect neonates, trials have been deferred so that safety and e fficacy data are obtained in older age groups meanwhile the unmet needs gap widens, necessitating o ff-label use [11]. An international consortium of experts has produced a white paper to facilitate successful neonatal clinical trials of medicines and includes useful information on neonatal dosage forms and formulations [12].

The situation is not helped by a relative lack of relevant guidelines on the development of medicines for neonates (let alone those born prematurely). In terms of international guidance, it is only in the latest revision of the ICH E11 Guideline on Clinical Investigation of Medicinal Products in the Paediatric Population (2018) [1] that neonates are specifically mentioned as an age classification and paediatric subgroup. In these guidelines, a rather general and brief set of considerations are made in terms of formulation and around polypharmacy via parenteral routes of administration in the hospital setting. The capability to administer small volumes in relation to dosing error is also mentioned. This is insu fficient considering the much wider and complex needs of neonates [13] and lack of clarity about how this limited guidance is to be translated into patient-centric product development, which meets with regulatory approval.

The common ground in current guidelines is that there are pointers to the lack of safety data on excipients and that available data generally does not apply to neonates requiring further justification including provision of non-clinical safety data [14,15].

The aim of this review is, therefore, to provide insights and factors to consider in order to assist those developing products for neonates, but with little or no neonatal medicine knowledge or paediatric formulation development background, to overcome the range of challenges posed by this patient group and so enhance the provision of clinically relevant products. This includes factors that di fferentiate the needs of preterm and term neonates from those of the older paediatric subsets, in terms of environment of care, ability to measure and administer the dose (from the patient, the routes of administration, the device and the product perspectives), formulation development, neonatal biopharmaceutics and regulatory challenges.

### **2. Formulation Considerations**

### *2.1. Environment of Care*

Neonates who receive treatment in hospital, will most often be located in the Neonatal Intensive Care Unit (NICU). The NICU admits high-risk premature and full-term neonates with serious medical or surgical conditions.

Factors which may necessitate neonatal admission include (modified from [16])


Often multiple conditions exist concomitantly leading to a need for polypharmacy.

Neonates admitted to the NICU often require periods in specialised incubators to maintain optimum environmental conditions and may also be attached to electrocardiogram (ECG), oxygen saturation and blood pressure monitors (Figure 1). In addition, they may require respiratory support through oxygen supplementation or require phototherapy for jaundice [17]. The majority of pharmaceutical medicinal products administered to neonates are liquid oral or parenteral formulations [18]. In this population, the environment of care is critical with the vast majority of IV infusions delivered in an intensive care setting where the additional environmental requirements of light, temperature and oxygen may impact the physicochemical stability of medication being delivered [19]. The effect of the environment of care should be a critical consideration in the formulation design of a parenteral product intended for this patient population and in use stability studies will be discussed in more detail below.

Neonates in incubators also provide some challenges for the provision of suitable nutrition and hydration requiring enteral or parenteral feeding. Given the limited number of access points and the requirement for polypharmacy that is common in this patient group, there are more opportunities for stability challenges, interactions and incompatibilities.

**Figure 1.** Schematic of NICU support for neonate (used with permission of Mayo Foundation for Medical Education and Research, all rights reserved).

### *2.2. Ability to Dose: Patient (Developmental Age)*/*Physiological*/*Administration Routes Factors to Consider*

Some drugs can be administered to these patient groups orally, topically, via inhalation or indeed by any of the usual administration routes and the specific issues associated with the ability to dose via these routes is discussed under each of these sections below. However, the main route of administration in the neonatal patient is parenteral and in particular the IV route for seriously ill preterm and term neonates [20].

### 2.2.1. Parenteral Delivery

The neonatal IV infusion is a demanding process, involving vulnerable patients and complex IV administration apparatus. Venous access in neonates can be via a catheter with tip placed in the vena cava, a central venous catheter (CVC) (threaded through from a peripheral vein and known as a 'PICC' or peripherally-inserted central catheter, inserted through the umbilical vein (UVC) or placed surgically) or via a peripheral cannula or catheter (accessing smaller veins in the hands, arms and feet). A UVC remains in situ for up to two weeks after birth and PICC may last for several weeks whilst cannulas and catheters accessing the small peripheral veins may only be patent for hours or days [21].

All cannulas and catheters require scrupulous care to avoid blockage and infection. Since blood flow in central versus peripheral veins is greater, an administered drug is rapidly diluted [22]. Some irritant (chemotherapy, amiodarone, vasopressors [22] and hyperosmolar (glucose >12.5%, total parenteral nutrition (TPN)) [23] medications, may be indicated for central administration only. Many medications are compatible with a peripheral route of administration. However, some may cause phlebitis (e.g., dopamine, dobutamine, sodium bicarbonate, calcium gluconate) [24].

The choice of available catheters is summarised in Table 1 alongside characteristics and issues associated with each.

Tubing of small internal diameter is often used to reduce the dead space in apparatus for IV drug administration. Even so, drugs may be exposed to adverse temperatures and light from the point of administration until they reach the vascular system. This should be taken into account when assessing in-use stability as well as pharmacokinetics and clinical outcomes.


**Table 1.** Characteristic of available types of the vascular catheters (adapted from [21,22,24–26].

An important formulation consideration associated with IV administration in preterm and term neonates is the volume of fluid that can be tolerated. Neonates, especially those delivered preterm, will have critical fluid and electrolyte requirements. At full term, a neonate has a fluid allowance of 100–140 mL/kg/day resulting in administration of 10–20 mL/hr, which includes all routes of administration [27]. In the case of severely fluid restricted neonates, IV fluids and drugs must be concentrated and so may need to be infused at flow rates as low as 0.02 mL/hr. Formulations for IV drugs and infusions should be designed with such requirements in mind and should take account of the sickest babies requiring multiple therapies, each with their own method of administration to be considered.

Generally, drugs administered intravenously should not be required to be administered in a fixed volume. It is preferable to investigate and report the minimum and maximum concentration at which the drug is sufficiently stable and to note any restrictions that this may pose for vascular access. For example, the need to use the central venous routes and slow rates of administration if higher concentrations might irritate or damage veins. The consequences of inadvertent extravascular administration should also be considered.

When drugs require dilution for administration the carrier fluid studied should include dextrose (5% and 10% *w*/*v*) as well as sodium chloride 0.9% w/v to maintain isotonicity. If dextrose solutions are suitable, this will help with sodium restriction and provide additional energy. Water for Injection should not usually be considered a suitable infusion fluid because of the potential risk of infusing hypotonic solutions, but information on stability and osmolarity may be useful if dextrose is inappropriate and sodium balance a problem.

Many neonatal patients in a critical care setting receive between 15 and 20 IV medications daily, the majority of these are unlicensed or used off label [4]. Lack of knowledge around the physicochemical incompatibilities of IV drugs in NICU and PICU settings often necessitates the use of a dedicated IV catheter in neonates and infants who have limited IV access [28,29]. Drug incompatibilities are often an underestimated aspect of clinical practice and are concerning in the neonatal population where a lower capability to compensate for adverse drug reactions may lead to higher morbidity and death [30,31]. This concern is exacerbated in neonates by the frequent requirements for polypharmacy, multiple infusions delivered through a single catheter due to limited vascular access, low infusion rates exposing drugs to longer interaction and the possibility of incomplete dissolution or precipitation of drug due to low volumes of drug solutions [32]. Realistically, limited venous access can result in little choice but to co-administer drugs.

A recent report highlighted the extent of the problem and reported that among medicines tested as a co-infusion of two drugs listed as a common NICU medication, only 4% of the combinations were fully compatible [28]. The compatibility of IV injections with other commonly administered drugs and infusion fluids should be studied when mixing at a 'Y'-site when the drug is likely to be used in the intensive care situation.

### 2.2.2. Oral Delivery

Medicine administration via the oral route to children less than two years of age can be di fficult for both parents and children [13,33]. The main issue in administering oral medicines to neonates lies in their ability to e ffectively swallow the medicine. Typically, most oral processes are present from birth (rooting, lip, lateral tongue, mouth opening, biting, and emerging chewing behaviours). However, even oral syrups are not always fully swallowed when administered to neonates and infants [34]. This is further exacerbated in preterm neonates, where issues around swallowing and ADME (absorption, distribution, metabolism, and excretion) factors may be less well understood. Clinical evidence suggests, that the oral absorption process of a drug undergoes substantial changes after birth [35,36]. The impact of these physiological changes on oral drug therapy is currently unclear.

Historically oral liquid formulations have been used in neonates. Examples of commonly used oral medicines for neonates include vitamin D drops, analgesic suspensions (ibuprofen, paracetamol), antibiotics, glucose gel for treatment of hypoglycaemia and anti-reflux medicines. The first approved medicine to treat neonatal diabetes (Amglidia ®) is a suspension formulation of glibenclamide [37].

EMA guidance [20] suggests, that the following oral formulation options are suitable from birth: Powders and granule (administered as a liquid) and oral liquid preparations (solution, suspension and oral drops).

It is important that appropriate strength medications are available for neonates to ensure that appropriate doses/volumes can be accurately measured and administered. The maximum recommended single dosing volume is 5 mL for children aged below five years and 10 mL for children aged below 10 years [38]. For neonates, dose volumes as low as 0.1 mL may be required.

The importance of accurate dosing to the younges<sup>t</sup> children has previously been highlighted [39]. Dosing devices for this age group mainly involves the use of an oral syringe or dropper though both have significant issues in terms of accuracy [40,41]. Therapeutic nipple shield [42,43] and medicine dispensing pacifiers [44,45] have been suggested for home use, but these are as ye<sup>t</sup> experimental. In hospitalised patients, enteral tubes can be used for the administration of oral medicines where liquids (ideally solutions rather than suspensions) are the preferred dosage forms. Emulsion formulations can also theoretically be delivered via this route. One example is enteral nutrition or milk. There is a risk that the PK of a highly lipophilic drug, delivered via this route, may be altered if co-administered with an emulsion formulation or milk. Although during development compatibility of drug-enteral feed will be evaluated for new medicines, it is impossible to consider every feed available to review the likelihood of an interaction. Thus, there are circumstances where a drug–feed interaction may lead to PK variability as a result of a novel feed. Many medicines currently used in neonates are unlicensed for use in this population, and there is little or no information on their compatibility with milk or enteral feeds which may lead to a change in the anticipated PK. This is addressed in more detail in section (biopharmaceutical consideration).

Palatability considerations for neonates are more commonly associated with volume and texture of medicines in addition to taste [46,47]. However, some taste preferences seem to be innate (e.g., sweetness), and in extreme cases, bitter taste can lead to vomiting, so taste cannot be ignored even in neonates. New products that are designed for children are likely to be those that o ffer flexible dosing to target not only neonates but also older infants and children. Therefore, a product that meets the needs of all age groups will be developed which will address issues of palatability for infants and children.

### 2.2.3. Rectal Delivery

Rectal administration is more commonly used from infancy onwards [48]. This route su ffers in neonates from unpredictable absorption which seems to be the limiting factor [36]. This is highly correlated with faecal incontinence/retention of the drug dose, which is inversely related to age/maturation. Solid dosage forms are usually better retained in the rectum than liquids. However, flexible and accurate dosing, which is an important product feature, cannot be delivered when suppositories are split at the point of administration [49]. This does not completely prevent the use of the rectal route to reduce the need for parenteral administration, or overcome oral administration restrictions (e.g., physiologically reduced gastrointestinal absorption, nausea, vomiting, seizures, nil by mouth), but safety and e fficacy with appropriate bioavailability studies and patient size-appropriate product are paramount. In fact, recent studies (pain, patent ductus arteriosus closure) [50–52] demonstrated similar e ffectiveness for rectal delivery (mainly small enemas) as compared with IV or oral delivery even in very low birth weight preterm infants as well as being cheap and safe. Therefore, in resource-limited settings, rectally formulated drugs for pre-referral use could have grea<sup>t</sup> potential, e.g., neonatal septicaemia, pneumonia or malaria [53]. Recently, rectal antibiotic (Ceftriaxone) to reduce treatment delays in neonatal sepsis presented as rectodispersible capsules have been proposed [54].

### 2.2.4. Pulmonary Delivery

The respiratory route is not much used therapeutically in preterm and term neonates with one major exception.

Respiratory distress syndrome (RDS) is a life-threatening condition, which occurs almost exclusively in preterm neonates with a deficiency [55,56], dysfunction or inactivation of pulmonary surfactant. The physiological role of surfactant is to allow the lungs to expand and avoid collapse (atelectasis) during the expiratory phases. Lack of surfactant results in di fficulty in breathing, with low oxygenation, increased breathing e ffort and the need for respiratory support.

The administration of exogenous surfactant can alleviate the symptoms of RDS by supplementing the endogenous pool of surfactant, thereby enabling the biofilm to be replenished, dramatically reducing mortality and morbidity. This is often administered via invasive methods, such as endotracheal intubation and mechanical ventilation (MV). The gold standard version of this approach [57,58] is called INSURE (intubation, surfactant administration, extubation). Less invasive approaches, using thinner catheters (such as LISA technique—less invasive surfactant administration [59,60]), have been designed to supply exogenous surfactant to spontaneously breathing neonates, but these are still partially invasive. Therefore, there is a grea<sup>t</sup> interest by neonatologists in the development of a truly non-invasive procedure for surfactant administration, such as nebulisation. Some interesting pilot studies have been published on this issue [61–63].

These various intubation techniques also allow administration of other therapeutic interventions to treat a wide range of breathing di fficulties should these be required. It is important that any formulations administered in this way, are of the correct concentration to be able to deliver the required dose in a small volume, and that the rheological profile is adequate to permit flow through the narrow tubes involved.

For neonates who can breathe spontaneously the major route of drug delivery is via nebulisation. A wide range of breathing di fficulties can be treated, including respiratory syntical virus (RSV) bronchiolitis or asthma-like reactive airways disease. Typical drugs administered include beta agonists, steroids, ribavirin and sometimes adrenalin.

An advantage of using the pulmonary route is that e fficacy can be achieved with reduced systemic drug levels and hence, decreased side e ffects [64]. For example, a recent review [63] has shown that pulmonary administration of corticosteroids can e ffectively prevent bronchopulmonary dysplasia in preterm infants with RSD, without the adverse e ffects on growth and neurodevelopmental outcome associated with systemic delivery. An emerging practice consists of adding budenoside to surfactant [65].

The formulation for nebulisation is usually a sterile aqueous solution or suspension, mainly based on the solubility and stability proprieties of the active drug substance. Many times, in order to achieve the suitable physico-chemical characteristics (i.e., osmolality, viscosity, pH etc.) and stability, it may be necessary to include excipients in the formulation. However, only a few excipients are approved for the inhalation route, and usually, there is little if any specific safety data available for neonates or the wider paediatric population [15], let alone after pulmonary exposure. Nebuliser solutions should be presented in an age-related unit dose of appropriate concentration if possible. Where this is not possible, the device used to measure the dose must not be a syringe designed for injection, thus reducing the risk of the nebuliser solution being inadvertently injected. The most appropriate nebuliser system to achieve good inhalation performance in terms of output rate, dose delivery to the lung, aerodynamic particle/droplet size distribution (i.e., fine particle or respirable fraction below 5μm or 3.3μm) and respirable drug delivery rate must be chosen. It is worth highlighting that the lung deposition of a nebulised product can be influenced by the breathing pattern (i.e., tidal volume, breath frequency and inhalation/exhalation ratio). The breathing pattern depends on the patient's age, and so, neonates (preterm and term) and children have significantly di fferent and variable breathing profiles, making reliable dosing di fficult when the e fficiency of drug delivery systems are breathing pattern dependant [66,67]. This is further complicated if the infant is crying as this also changes breathing patterns.

It is possible to model the influence of various anatomical, physical, and physiological factors on aerosol delivery in preterm neonates on the e fficiency of the delivery of an aerosolized drug to the bronchial tree using 3D models such as the (PrINT model) [68].

Nebulised delivery may also be important in older paediatric patients during times when the child's inspiratory force is low such as during an acute exacerbation of their condition. However, for both babies and older paediatric patients, some studies have shown that drug delivery from a Pressurised Metered Dose Inhaler (pMDI) and spacer system with suitably sized facemask can be at least as e ffective as nebulisation [69,70]. Delivery via pMDI and a spacer system is recommended in NICE guidance [71] for routine treatment of older paediatric patients.

### 2.2.5. Nasal Delivery

Although the nasal aperture is small in diameter in neonates and the nasal mucosa is often delicate and coated with mucus, it nevertheless o ffers a potential route for employing local drug administration to e ffect systemic drug delivery, thus reducing the need for injectable administration. In emergency care, this could also decrease the need for additional painful procedures such as insertion of IV cannulas for medication administration. For example, intranasal fentanyl has been used in the palliative care of term neonates and infants [72]. As such, it has attracted significant research interest in the last few years for use in this age group. Even if intubation is required, nasal administration could provide a less impactful means of aiding that procedure. Nasal midazolam was used and was more e fficient than nasal ketamine, to adequately sedate neonates requiring intubation in the delivery room [73]. A typical dose volume was 0.1 mL/kg in each nostril.

Potential advantages of this route of administration, are that it is less invasive than IV whilst maintaining rapid onset since medications are directly absorbed (parenterally) through the nasal mucosa into systemic circulation, it also results in higher bioavailability compared to oral medications, as first-pass hepatic metabolism is bypassed. There is also some evidence that brain penetration can be enhanced as a consequence of nasal delivery [74–76].

Compared to buccal administration, (another method of avoiding first-pass metabolism), where salivation and swallowing issues after administration in neonates could have a substantial impact on dosage accuracy, more consistent absorption and bioavailability have been obtained following intranasal administration [77].

Bypassing the gastrointestinal tract with nasal drug delivery allows for the delivery of some macromolecules with low permeability across the gastrointestinal tract or susceptibility to

chemical/proteolytic degradation. Recent studies in neonatal mice highlighted the potential of intranasal immunization of neonates with live vaccines [78]. As vaccination at birth would provide early protection for neonates and infants, expanding and improving the available means of neonatal vaccination is a global health priority.

There is a well-established link between nasal dosing and enhanced blood–brain barrier (BBB) permeability. This may or may not be of therapeutic benefit. If central activity is not desired, then nasal dosing may lead to enhanced toxicity. The BBB is a dynamic physiological barrier which regulates the passage of hydrophilic molecules into the central nervous system (CNS) via a physical barrier formed by tight junctions (TJ) between the endothelial cells and also a system of influx and e fflux transporters and enzymes. Recent reviews and papers on the development of the BBB in preterm and term neonates sugges<sup>t</sup> that although the physical barrier is formed very early, many other factors that a ffect brain penetration favour increased drug levels in the brain and cerebrospinal fluid (CSF) and that these factors change rapidly with maturation [79–81]. As yet, there is little or no information on the e ffect of actives and excipients on the development of the BBB in these age groups. It is known, however, that some penetration enhancers used in intranasal formulations for adults work via breaking tight junctions which is clearly not safe in neonates [82]. Conversely, cyclodextrins (e.g., hydroxypropyl-β-cyclodextrin), which are increasingly being used as vehicles to transport lipophilic drug through BBB [83], may have a neuroprotective e ffect and are used therapeutically in the treatment of neonatal hypoxia-ischemia [84].

Alterations in BBB development and in TJ expression could lead to anomalies later in life as well as to increased predisposition to some diseases. Neonatal hypoxia-ischemia (HI) causes severe brain damage and remains a major cause of neonatal morbidity and mortality [85]. At present, treatment options for neonatal HI brain damage are very limited and have only modest e fficacy [86]. A study conducted in several rodent models of ischemic brain injury demonstrated the therapeutic potential of mesenchymal stem cells (MSCs) transplantation that improves functional outcome and also restores brain structure [87]. These findings in rodents indicate that the nasal route was an e fficient route for stem cell transplantation after brain injury in the neonates [88].

During the development of a new formulation for nasal administration to the neonate, as well as all the usual factors (e.g., pH, osmolality, chemical irritation and overall acceptability) the formulator needs to keep in mind:


### 2.2.6. Dermal and Transdermal Delivery

In neonatal practice, many of the topical treatment decisions are made by specialist nurses. It is therefore very important for formulators to provide a high level of support to enable safe and e ffective use of topical formulations.

Formulators should be aware that the skin barrier may not be fully formed in preterm infants both at birth and for up to four weeks afterwards [90]. There are also changes in the type and proportion of lipids present and the development of ancillary skin structures such as sweat ducts and hair follicles [91]. This has implications for potential toxicity of active pharmaceutical ingredients (API's) and formulations usually delivered for a localised e ffect via the topical route. In the absence of a fully competent skin barrier, higher systemic exposure of both API and common topical formulation excipients could be expected. The skin barrier may also be breached in term neonates and infants due to conditions such as cradle cap, nappy rash and eczema/dermatitis. Even if the skin is fully formed and functional, the lipid composition changes rapidly and the stratum corneum tends to be thinner in young babies and more hydrated. Additionally, occlusive dressings including nappies/diapers with impermeable plastic coverings can increase absorption. Besides permeability, the simple BSA/kg ratio puts neonates at a higher risk of increased absorption. Therefore, any adverse e ffects tend to be exacerbated by the relatively higher surface to volume ratio in children leading to a risk of unwanted high systemic exposure if large areas of skin are treated.

Formulators, therefore, need to ensure that APIs delivered topically in neonates have excellent systemic safety (limitation to the skin alone cannot be assumed). It is insu fficient to only undertake topical safety studies as systemic absorption should be assumed and therefore systemic safety studies are required. Toxicity of ingredients in subpopulations such as neonates requires a careful risk assessment to avoid the use of excipients with unclear safety in target population. It is also necessary to select formulation excipients that are intrinsically safe and have a low potential to cause irritation or other skin reactions, e.g., sodium lauryl sulfate [92].

Finally, it is important to ensure that the pH and tonicity of the formulation is well matched to the specific requirements of the skin at the various stages of development and an understanding that this can change rapidly [93,94].

Coupled with the state of development of the barrier function of the skin, it is important to remember that skin in the neonate is often fragile and that it can be damaged by mechanical abrasion. Thus, the rheological profile and other cosmetic attributes of the formulation need to be taken into account. It may well be desirable to use a specific neonatal formulation that is more fluid/easier to spread than is common for creams and ointments used in older children. Similar considerations apply to the potential skin damage that can be caused by adhesives used in transdermal patches.

In some cases, the topical route is deliberately used to achieve systemic delivery [91]. In this case, it is still important to remember that the barrier properties of the skin may be rapidly changing during early development as this might a ffect the systemic blood levels achieved from a particular dose with implications for the topical dose that needs to be given.

Indeed, dose flexibility is a challenge in the neonate due to the above-mentioned developmental factors specific to the skin and due to the rapidly changing weight of the neonate. For standard semi-solid formulations, this can be accommodated to some extent by the area that is treated, but a wide range of strengths is still likely to be required. This is often achieved by diluting a formulation extemporaneously with a base. If this is done the stability implications (chemical, physical and microbiological) need to be considered along with the issues of ensuring homogeneity of the diluted formulation. For 'unit dose' topical formulations such as patches, dose control is achieved either by cutting patches or masking them (usually o ff-label) to reduce the contact area. Both practices are subject to significant risks of poor dose control unless the developer fully validates this practice.

Once a fully competent skin barrier is formed, it is possible to use the full range of topical and transdermal delivery [91] including creams, ointments, sprays, lotions, baths and, less frequently, transdermal patches with suitable adjustment to the dose. Some patches have paediatric labelling supported by clinical trials, whereas others are used o ff-label. There is little literature on the utility of innovative delivery methods such as needleless injectors, iontophoresis, sonophoresis and microneedle patches in the neonate. However, there is evidence of them being tested for use in older paediatric patients in order to attain more reliable drug levels, either locally in the skin or systemically, that are bioavailable from creams and ointments, especially where dermal permeation of the API is low or slow.

### *2.3. Ability to Administer: Product Factors to Consider*

The lack of appropriate dosage forms frequently results in o ff-label and/or unlicensed use of modified adult formulations for administration in neonate and paediatric patients [95]. It has been reported that between 71% and 100% of patients in the NICU receive at least one o ff-label or unlicensed medicine [96].

The dose of a medicine can vary 100-fold between that for a preterm baby and an adult. Whilst a liquid medicine may be acceptable for all ages, using just one concentration for all ages would mean that an appropriate dose volume for an adult (say 10 mL) would be impossible to measure accurately for a preterm neonate (say 0.1 mL). Thus, more than one concentration would be required in this example.

Fortunately, many medicines can be dosed according to weight bands (e.g., 5–10 kg, 11–20 kg) or by age bands making it easier to produce an acceptable standard concentration of a drug. Formulations available from manufacturers should follow this rule, making it more feasible to manipulate with drug doses for neonatal application. However, doses of potent medicines for neonates necessitates individualised dose calculations based on body weight (e.g., mg/kg) or body surface area (e.g., mg/m2). This could lead to the preparation of very low drug concentrations, several manipulations in preparation of the end formulation and the managemen<sup>t</sup> of multi-infusion IV administration [97]. This has been reported to lead to a significant risk of medication errors and adverse drug events, especially in the NICU [98].

To overcome these issues associated with individualised concentration calculations based on low body weight, formulations of standard concentrations [97,99], specifically of high alert medication, have been implemented into NICU settings [100–102]. These can then be used with intelligent infusion pumps programmed with the standard concentrations allowing automated calculation of the individual flow rate based on body weight. Standardised concentrations also make it easier for pharmacies to prepare or source drugs which are 'ready to infuse' rather than them being prepared at the bedside with the attendant risks. They also reduce the burden of complex patient-individualised calculations on the prescriber and nurse administering the medication [97,99,103] and standard concentrations of di fferent strengths allow for flexibility to compensate for a patient's di ffering fluid requirements, though the existence of more than one concentration introduces the potential for product selection error.

As the flow rates of standard concentrations are linked to patient weight, the flow rate decreases with decreasing patient's weight. This can result in very low flow rates in extremely low birth weight neonates where rates may be as low as 0.02 mL/h. Low flow rates have been associated with long delays in drug delivery times and hence the onset of e ffect, delays in transitioning to new infusion rates and delays in reducing e ffects when the infusion is stopped [27]. Such unpredictable delivery is unacceptable in the administration of life-sustaining medicines to critically ill patients [104–106]. Lack of awareness of these issues among clinical sta ff may lead to inappropriate clinical decision-making [27,104]. Observed lack of response to treatment due to delivery delay may cause precocious dose increase leading to overdosing and toxicity. When examined together with low flow rates, several other factors may contribute to prolonged and unpredictable delivery such as, the dead space of the administration system [27,104], the adsorption of the drug to tubing, backflow of the infusion [105], change of the flow rate of other infusions or a carrier fluid [27], the type of pump used, the placement of the infusion tubing and the line architecture [27].

When designing new formulations or reformulations of existing drugs, clinicians and drug formulators should work together to maximise an optimal drug delivery under special neonatal drug administration conditions (Table 2). Information and advice provided to the end user should facilitate a practical dosing regimen taking into account the potency of the drug and potential adverse e ffects and the setting in which the drug is to be measured and administered (NICU, hospital ward, home).

Using standard concentrations of medicines can help avoid medication errors when patients move between care settings [107]. Due to the implementation of standard concentrations to the neonatal care, the number of individually prepared drug doses reduced dramatically. Formulation of medication strengths that could represent standard concentration used in the NICU would eliminate the step of drug manipulation to prepare standard concentrations.

> **Table 2.** Factors to consider in neonatal parenteral drug formulation and administration.

Chemical and physical compatibility of drug formulation used in multi-drug administration [28] including generic brands

Chemical and physical compatibility of drug formulation used in combination with neonatal TPN [108,109]

Compatibility of drug with diluents typically used in the NICU and stability after dilution

Compatibility of drug formulation while mixing at Y-site junction at different mixing ratios [108,109]

Stability of drug formulation over extended period of time (e.g., over 24 h infusion)

Stability of drug formulation exposed to different environmental conditions (high temperature, strong light, high oxygen levels) [110]

Stability and compatibility of excipients used in drug formulation

Stability and compatibility of excipients used in drug formulation with IV administration set and container CompatibilityofdrugformulationwithIVadministrationsetandcontainer

Strength(s)/concentration of drug that can cover neonatal weight- or age-bands as well as fluid restricted patients

Performance of medical equipment delivering drug—volumetric and smart pumps, syringe drivers

Design of IV administration set minimising drug delivery delays

### 2.3.1. In use Stability Issues

As well as all the administration factors discussed above, it is clear that formulations need to be stable in order to deliver the expected dose to the patient.

Although all the usual ICH stability requirements will apply, the therapeutic environment will often present some additional 'in use' stability challenges that should be considered by the formulator when developing products for this age group. This is particularly true for preterm infants and those being treated in high dependency incubators or intensive care environments and when those infants are treated almost exclusively via the parenteral route. They are complex to anticipate and therefore to reproduce in vitro despite regulatory expectations. Yet, conditions in NICU may be more controlled than in other environments. Clinical colleagues can provide information on likely treatment scenarios that will allow the most common potential interactions to be mimicked. Drug concentrations, diluents, formulations, mixing ratios and environmental conditions (light and temperature) are some factors pertinent to the design of compatibility studies. Some of the major issues are true pharmacy issues that need to be considered within the clinical context:

a) Photostability—useful reviews have been published [111,112]. Light intensity is often higher in the neonatal environment. Phototherapy is often used to treat postnatal jaundice either with lamps having the required spectral power distribution to break down the bilirubin in the skin or even exposure to direct sunlight. The implications of this treatment to the stability of any drug that partitions to the skin or eye need to be considered along with the implications for photosafety particularly if the metabolic immaturity of the neonate might be expected to affect clearance of any photoproducts that are generated. When bilirubin is present in the skin, protein binding will also be high. This can either exacerbate or mitigate photostability issues, depending on the mechanism of photodegradation. It is extremely difficult to predict what effect this protein binding might have *a priori.*

If the infant is being treated parenterally, it should be remembered that solution concentration tends to be low (though they can also be high if the neonate is fluid restricted) and infusion rates can also be low leading to long residence times in the administration lines. There is often a transit time of several hours between the infusion pump and the patient [27,106]. Added to this is the potentially very large surface to volume ratio due to the solution flowing through a long narrow bore tube. All of these

factors can lead to significant levels of photodegradation that may not be fully predicted by some ICH Q1B photostability tests. It is therefore important to undertake a well-designed in use photostability test to allow for suitable advice to be provided to the healthcare practitioner.

Another factor that needs to be considered here is that the material of construction of any transfer line may be quite variable. Thus, some plastics will contain quite high levels of UV stabilisers whilst others may have much lower levels or even none at all. Clearly, this can lead to di fferent levels of photoprotection for the content of the line.

If the product is known to be photolabile, then it may be wise to advise that any holding vessel (such as a syringe or infusion bag) and particularly any transfer lines are shielded from light exposure either with a suitable coloured cover or with a light impermeable barrier such as aluminium foil. One downside of doing this is that it will render visual inspection more di fficult.

b) Other environmental factors—due to potential di fficulties in breathing and less developed thermoregulation preterm infants may be in an oxygen-rich environment that is maintained at a higher temperature than would usually be the case in the hospital environment. This has obvious potential implications for formulations that are subject to oxidation or thermally driven hydrolysis if they are stored in such an environment for extended periods. Again, this may be further exacerbated by low concentration. The formulator will need to consider whether or not these conditions might have an effect on their product and provide suitable handling advice accordingly.

c) Potential Interactions—due to the limited number of access points for parenteral delivery to the neonate, it is often necessary to provide several drugs via the same entry point. In addition, neonates may be provided with nutrition via the same route. Thus, it is important to consider potential incompatibilities (both chemical and physical) between the various medications and the potential for API to adsorb to components of the parenteral or enteral feeding formulation used. In some cases, high concentrations of dextrose may be administered in this age group, both of which could have effects on product stability. On top of that, there is a risk that the API may interact with the material of the transfer tube leading to adsorption of the API. If the concentration is low, there may be a significant loss of therapeutic activity.

d) It may be a challenge to provide a su fficient range of formulation strengths to cover all needs of the paediatric population. Dilution may be required to allow accurate measurement using routinely available syringes. Large dilution factors may be required if only a single product concentration (often designed for adults or older children) is available. If dilution is to be avoided it may be necessary to develop and provide an appropriate dose measurement device and/or advice on providing an accurate dosage. Another method for controlling dose (volume) delivery is to change the infusion rate as discussed in the parenteral section.

If the formulation contains a stabiliser of some sort (e.g., antioxidant, photostabiliser) then dilution may render that stabiliser ine ffective with obvious potential consequences for the formulation stability. The extent of the impact will be determined by a knowledge of the stability of the unprotected API and the level of stabiliser that remains. If the parenteral product is an emulsion or suspension, then dilution may lead to physical instability potentially leading to blockage of cannulas or even phlebitis. On the positive side, dilution may also reduce intake of an excipient that might be harmful to the neonate, e.g., sodium metabisulfite in parenteral inotropes.

The diluent used is also of importance. In some instances, photodegradation of the diluent can lead to degradation of an otherwise photo stable API (e.g., [113]). Formulators should be aware of this possibility when considering which diluent to recommend.

Photostability (and possibly oxidative stability) is also of concern for formulations that are applied topically in such a high-intensity light environment. Clearly, there is a risk that stability could be poor, leading to suboptimal dosing and/or the generation of relatively high levels of photo products in situ. The fact that the API concentration could be low should be considered when assessing the likely impact.
