Prodrug approaches for the development of a long-acting drug delivery systems

Abstract

Long-acting formulations are designed to reduce dosing frequency and simplify dosing schedules by providing an extended duration of action. One approach to obtain long-acting formulations is to combine long-acting prodrugs (LA-prodrug) with existing or emerging drug delivery technologies (DDS). The design criteria for long-acting prodrugs are distinct from conventional prodrug strategies that alter absorption, distribution, metabolism, and excretion (ADME) parameters. Our review focuses on long-acting prodrug delivery systems (LA-prodrug DDS), which is a subcategory of long-acting formulations where prodrug design enables DDS formulation to achieve an extended duration of action that is greater than the parent drug. Here, we define LA-prodrugs as the conjugation of an active pharmaceutical ingredient (API) to a promoiety group via a cleavable covalent linker, where both the promoiety and linker are selected to enable formulation and administration from a drug delivery system (DDS) to achieve an extended duration of action. These LA-prodrug DDS results in an extended interval where the API is within a therapeutic range without necessarily altering ADME as is typical of conventional prodrugs. The conversion of the LA-prodrug to the API is dependent on linker cleavage, which can occur before or after release from the DDS. The requirement for linker cleavage provides an additional tool to prolong release from these LA-prodrug DDS. In addition, the physicochemical properties of drugs can be tuned by promoiety selection for a particular DDS. Conjugation with promoieties that are carriers or amenable to assembly into carriers can also provide access to formulations designed for extending duration of action. LA-prodrugs have been applied to a wide variety of drug delivery strategies and are categorized in this review by promoiety size and complexity. Small molecule promoieties (typically MW<1000 Da) have been used to improve encapsulation or partitioning as well as broaden APIs for use with traditional long-acting formulations such as solid drug dispersions. Macromolecular promoieties (typically MW>1000 Da) have been applied to hydrogels, nanoparticles, micelles, dendrimers, and polymerized prodrug monomers. The resulting LA-prodrug DDS enable extended duration of action for active pharmaceuticals across a wide range of applications, with target release timescales spanning days to years.

Graphical Abstract

1. Introduction

The effectiveness of medications requiring strict dosing schedules are often limited by inadequate patient adherence, resulting in poor outcomes and increased cost of care[1]. Adherence is critical in a variety of clinical applications where dosing intervals are prescribed to maintain an active pharmaceutical between its minimum effective and maximum tolerated concentrations (Fig. 1). Adherence to treatment regimens can be drastically improved with formulations that extend the duration within this therapeutic window on timescales spanning days to years from a single dose [2]. Long-acting formulations (LAF) include any strategy that extends the duration of drug action such as solid drug dispersions as well as drug encapsulation approaches into biodegradable or non-degradable implants, polymers, or hydrogels [2–6]. Long-acting strategies sustain drug release at a level that balances the clearance rate to maintain therapeutic concentrations in blood or tissue [2]. LAFs typically alter pharmacokinetic parameters such as T_max and C_max, but ultimately provide significantly longer duration of action and decrease the frequency of dosing. The time interval for long-acting may refer to multiple days or years depending on the context of the application and treatment. For example, the target dosing frequency of long-acting injectables for HIV treatment is one month or more, whereas dosing frequencies for long-acting antipsychotics is 2–4 weeks[4,7]. Even shorter durations of action may be considered long-acting in the case of opioid delivery, where the target dosing frequency is 72 hours[8]. Long-acting formulations are also designed for specific routes of administration. While oral administration is advantageous with respect to patient acceptability, absorption and distribution limits its applicability in long-acting formulations[9]. Instead, LAFs are typically designed for parenteral administration [5][10]. The interplay between application requirements and user acceptability can also lead to different long-acting target specifications. For example, in the field of HIV treatment and pre-exposure prophylaxis, acceptable dosing frequencies depend on whether the route of administration is oral (≥ 1 week), injectable (≥ 1 month) or implantable (≥ 6 months)[4]. Given the wide range of timescales when referring to a long-acting formulation, the criteria for inclusion in this review is whether drug concentrations are prolonged relative to standard delivery of the parent API unless there is a consensus in the field on the dosing frequency required for long-acting delivery.

Fig 1. Summary of Long-Acting Prodrugs Effect on Extending the Duration of Action (EDA) of an API.

Long-acting prodrugs enable formulation into a variety of drug delivery systems that extend the duration of action of a parent API by a wide range of time-scales [19,20,29–31,21–28]. Here, LA-prodrug DDS results in an extended interval where the active pharmaceutical is within a therapeutic range between its minimum effect and maximum tolerated concentrations. In this schematic, the LA-prodrug DDS extends the duration of action of a parent API without necessarily altering its absorption, distribution or elimination rates.

Our review focuses on long-acting prodrug delivery systems (LA-prodrug DDS), which is a subcategory of LAFs where prodrug design facilitates DDS formulation to achieve an extended duration of action that is greater than the parent API. Many APIs are incompatible with LAFs such as nanocrystalline solid drug dispersions, and LA-prodrug strategies are one approach to modify APIs for LAF platforms that may differ in form and dosing frequency. For example, a wide range of chronic diseases require routine management involving highly hydrophilic drugs, in which patients suffering from such diseases can benefit from the availability of long-acting formulations. For example, essential thrombocythemia (ET) is a clonal stem cell disorder that results in elevated levels of platelets and persistent thrombocytosis[11]. First-line treatment can involve daily aspirin and hydroxycarbamides (hydroxyurea, HU) to prevent thrombosis. HU is a highly hydrophilic antiproliferative drug that can potentially be formulated as a long-acting prodrug to facilitate LAF, which can benefit patients by minimizing adherence issues[12]. Although such chronic conditions and disorders are rare, FDA grants potential seven years of market exclusivity, tax credits for clinical trials, and exemption from user fees to incentivize development of orphan drugs[13]. Long-acting prodrug strategies discussed in this review can facilitate long-acting formulations for many diseases and benefit patients suffering from chronic conditions.

The design criteria for LA-prodrugs are distinct from conventional prodrug strategies. Conventional prodrugs typically describe a single molecular entity composed of the API, a promoiety of variable complexity, and a cleavable covalent linkage between the API and promoiety. Following uptake and distribution, the covalent linkage is cleaved enzymatically or hydrolytically to release the API with the timing and location of cleavage largely determined by the structure of the linkage[14]. Conventional prodrugs focus on optimizing ADME properties of APIs and are not necessarily long-acting[14]. For example, conversion of the reverse transcriptase inhibitor tenofovir into tenofovir alafenamide reversibly masks a phosphate group, resulting in significantly enhanced cell permeation that leads to higher intracellular drug concentrations [15]. Conventional prodrugs can offer prolonged duration of action since ADME properties directly impact pharmacokinetics. The key distinction between conventional prodrugs versus LA-prodrugs is the criteria used to select promoieties and linkers. In a conventional prodrug, the promoiety and linker is selected to optimize some aspect ADME for the prodrug itself. In contrast, for LA-prodrugs, the promoiety and linker is typically selected to enable a drug delivery strategies that promote extended duration of action.

Several long-acting formulations can extend the duration of drug action without requiring the use of a LA-prodrug. For example, drug encapsulation into polylactide-co-glycolide (PLGA)/polylactic acid (PLA) based microspheres are used in formulations for octreotide acetate (Sandostatin^® LAR, Novartis) and leuprolide acetate (Lupron^®, Takeda) to achieve dosing frequencies of 4–12 weeks[16]. However, these approaches are most successful for hydrophobic and insoluble drugs where high drug loading and slow dissolution from the hydrophobic matrix drives extended duration of action[17]. In contrast, hydrophilic drugs typically have low drug loading and rapid dissolution from these same DDS [16,18]. Alternatively, conjugating the drug to the carrier via a cleavable linkage is possible for drugs with appropriate functional groups, where drug release is dependent on linker cleavage[17]. In this review, we focus on strategies to broaden the APIs that are amenable to LAFs by using prodrugs. These LA-prodrugs can improve upon conventional prodrug development focused on ADME optimization by further implementing designs that will facilitate a carrier affect or integration with a carrier for the purpose of sustaining drug bioavailability. This review focuses on strategies applied to drugs that are already clinically approved where long-acting functionality is desirable to improve adherence and clinical outcomes. In this review, we describe LA-prodrugs where the selection of the promoiety and linker is driven by the design of the drug delivery formulation or device (Table 1). First, we present long-acting prodrugs employing small molecule promoieties (typically MW<1000 Da), where the promoiety structure is a well-defined chemical entity. Subsequently, we discuss prodrugs employing macromolecular promoieties (typically MW>1000 Da), where the promoiety structure is typically polydisperse, leading to distributions of prodrug structures. The criteria used to evaluate literature for inclusion in the review is summarized in Table 1. Drug release mechanisms are also considered in the various formulations and strategies. Lastly, we summarize the challenges and outlook associated with LA-prodrug DDS.

Table 1.

Inclusion criteria for prodrug designs that enable extended duration of action (EDA)

Prodrug Example	Prodrug Design Contributes to EDA*	Criteria 1: Promoiety Selection Contributes to EDA?	Criteria 2: Clinically Relevant Need for API to Have EDA?	Inclusion in Review? Criteria 1 + 2 = Yes
Tenofovir Alafenamide (TAF) [15]
	No	No	No	No
TAF Implant (subdermal) [32]	No	No (Implant enables EDA, not prodrug)	Yes	No
Pluoronic 407-Coumarin-Tetracaine [33]
	No (Design promotes rapid release not EDA)	No	No	No
Myristol-Cabotegravir [34]
	Yes (Prodrug design enables nanocrystals)	Yes	Yes	Yes
TAF Drugamer [29]
	Yes (Prodrug design enables depot)	Yes	Yes	Yes
ProGel-Dexamethasone Hydrogel [35]
	Yes (Prodrug design enables depot)	Yes	Yes	Yes

^*EDA = Extended Duration of Action; Chemical Structure: Red = API, Blue = Polymer backbone, Black = Linker

2. Long-Acting Prodrugs with Low Molecular Weight (MW<1000 Da) Promoieties

In this section, we review LA-prodrugs employing small molecule promoieties (typically MW<1000 Da), where the promoiety structure is a well-defined chemical entity that leads to LA-prodrugs with defined chemical structure and molecular weight (Fig. 2). Strategies of this type are further organized by how the LA-prodrug is used to achieve extended duration of action. The first subsection focuses on tuning the physiochemical properties of LA-prodrugs using different promoieties in order to promote encapsulation and slow dissolution from various drug delivery systems. The second subsection focuses on lipophilic prodrugs and how their high-order assembly yields slow dissolution and sustained bioavailability.

Fig 2. Cleavable and Irreversible Bonds Used in Prodrugs Designs for Extended Duration of Action.

Schematic shows cleavable bonds discussed in this review. The innermost concentric circle (white) depicts functional groups on the drug that are amenable to chemical modification to generate a cleavable linker. The middle concentric circle (yellow) depicts cleavable bonds that can be obtained from different functional groups. The outermost concentric circle (green) shows various promoieties that can be used for long-acting prodrugs.

2.1. Promoieties to Improve Encapsulation in Drug Carriers

Various drug carriers can be used to improve the delivery of drugs to achieve extended duration of action. One of the main challenges for passive encapsulation via noncovalent interactions is poor drug loading[36]. Small hydrophilic drugs with molecular weight <1000 Da and logP or logD <3.0 are particularly difficult to encapsulate in hydrophobic carriers[18]. The weak interaction between drug and carrier in these systems leads to rapid partitioning into the aqueous phase under physiological conditions, limiting the duration of action[37,38]. In certain cases, drug encapsulation of hydrophilic drugs in carriers such as PLGA microspheres is poor due to the rapid partitioning into the external aqueous phase during the formulation step[39]. Likewise, insufficient drug loading is one of the major hurdles preventing widespread clinical usage of polymeric micelles, despite their potential to treat tumors in a long-acting manner due to the improved solubility and bioavailability of chemotherapeutic drugs[40]. Long-acting prodrugs designed to enhance non-covalent interactions with the carrier can improve encapsulation efficiency and loading while sustaining release by minimizing partitioning to plasma or tissue.

Several research groups have applied LA-prodrug strategies to improve the loading and release of hydrophilic drugs from polymeric carriers, often focusing on the effect of promoiety lipophilicity. Creighton et al. (2019) investigated various acyl ester prodrugs of raltegravir (RAL) that were designed to prevent ionization and increase lipophilicity. LA-prodrug modifications of RAL increased encapsulation in poly(lactic-glycolic acid) (PLGA) nanoparticles by as much as 20–25 fold when compared to free RAL[41]. Creighton et al. (2019) attributed the greater encapsulation efficiency of RAL prodrugs to the enhanced non-covalent interactions with the polymer/solvent oil phase of the emulsion. While increased octanol:water partition coefficients (P_ow) were necessary for high loading a direct correlation with P_ow values was not observed, likely due to prodrug specific interactions in 1-octanol compared to the polymer/solvent oil phase of the emulsion. In vitro results demonstrated release was sustained with a heptanoate RAL-prodrug (approximately 100% cumulative release just under 80 h) and a benzoate RAL-prodrug (approximately 100% cumulative release just under 100 h). Although the duration of action is shorter compared with other formulations for HIV treatment, this study demonstrates a strategy to minimize burst release from PLGA nanoparticles. Conversion of drugs into LA-prodrugs with acyl esters and other lipophilic promoieties have also demonstrated sustained in vitro and in vivo effects. Forrest et al. (2008) conjugated acyl esters of variable length to the hydroxyl group of paclitaxel (PTX)[42]. The Flory-Huggins interaction parameter for each prodrug and poly(ethylene glycol)-co-poly(caprolactone) (PEG-b-PCL) was calculated to gauge the polymer-solvent interaction and predict encapsulation. Forrest et al. (2008) determined the encapsulation of 7’hexanoate-PTX (PAX7’C₆) was improved by a factor greater than 36.5 times in PEG-b-PCL, due to the Flory-Huggins interaction parameter being less than 1. Intravenous administration of PAX7’C₆ in PEG-b-PCL micelles in rats improved t_{1/2, serum} from 1.55 h to 6.95 h when compared with PTX as the control. Wang et al. (2014) conjugated vitamin E and various amino acids to SN-38 in order to shield the phenolate moiety (10-OH), which had been posited to contribute to poor compatibility with polymeric NPs[43]. The high polarity of the phenolate moiety prevented facile encapsulation in PEG-PLA nanoparticles during nanoprecipitation. Wang et al. (2014) identified five promoieties that demonstrated the optimal lipophilicity to associate with the hydrophobic core of PLA, while being sufficiently soluble in water for encapsulation at a sufficiently high mass. Furthermore, encapsulated SN-38 LA-prodrugs in PEG-PLA micelles successfully inhibited tumor growth by 3.29 times over 20 days when compared to the free drug. Lipophilic prodrugs have also been applied to dendrimers. Ward et al. (2013) showed that both oligoethyl glycol or tert-butyl additions to morphine resulted in LA-prodrugs with improved encapsulation within poly(amidoamine) (PAMAM) dendrimers [44]. Dissociation of free morphine from the dendrimer carrier was nearly complete in just 5 min after incubation in guinea pig plasma. In vivo analgesic tests with rats indicate tert-butyl morphine LA-prodrug provided sustained analgesic effect up to 6 h, while the oligoethyl glycol morphine LA-prodrug was unable to provide similar analgesic effect. This phenomenon may be explained by the slower conversion of the oligoethyl glycol morphine prodrug to active morphine.

Similar strategies can also be used to improve encapsulation in liposomal carriers for sustained delivery. Pignatello et al. (2016) demonstrated that attaching an C-12 alkyl chain via an amide linkage to a luteinizing hormone-release hormone (LHRH) peptide improves its encapsulation within liposomes[45]. Although release data comparing free LHRH and LHRH LA-prodrug were not include, this study demonstrates the possibility of chemically modifying peptides as well as small molecules for improved encapsulation efficiency, and points to potential extended duration of action. Zhang et al. (2016) synthesized moeixitecan, a LA-prodrug comprised of SN-38 conjugated to a succinic acid via an ester linkage, which was further modified with a Trolox component followed by a hexadecanol chain[46]. The increased lipophilicity of moeixitecan resulted in much greater encapsulation within liposomes (90% efficiency) compared to CPT11 (irinotecan), a water-soluble prodrug of SN-38 (34–60% efficiency). The improved encapsulation efficiency was attributed in part to the partitioning of moeixitecan in the hydrophobic region of lipid bilayers, which was supported by MD simulations. When delivered intravenously in Sprague-Dawley rats in vivo, moeixitecan loaded in liposomes exhibited a two-fold increase in t_{1/2, plasma} and almost three-fold increase in mean residence time relative to CPT11 loaded liposomes[47]. Xing et al. (2019) demonstrated the moeixitican-loaded liposomes inhibited tumor growth for up to 15 days with an inhibition rate of 66.86%, whereas the corresponding inhibition for CPT11 loaded liposomes was 46.06%. The authors attributed the increased in vivo antitumor activity to the sustained drug release of moeixitican-loaded liposomes, improved systemic circulation, and tumor accumulation via the enhanced permeability and retention.

Long-acting prodrug strategies have also improved encapsulation in polymeric drug carriers, often by conjugating fatty acyl promoieties to optimize lipophilicity. Gemcitabine is a deoxycytidine analogue used to treat pancreatic cancer. The short half-life of gemcitabine requires frequent administration may lead to significant dose-related side effects[48,49]. To address this, Daman et al. (2014) synthesized a stearoyl-gemcitabine prodrug encapsulated within poly(ethylene glycol)-b-poly(D,L-lactic acid) (PEG-PLA) micelles for sustained antitumor effect. Stearoyl-gemcitabine demonstrated high encapsulation efficiency (>90% encapsulation) due to the association of the stearyl chain with the hydrophobic PLA interior[50]. Stearoyl-gemcitabine loaded PEG-PLA micelles provided sustained release with only 30% of drug released over 72 h in vitro. Another example of a drug that is poorly encapsulated and thus difficult to formulate into LAFs is dexamethasone, a glucocorticoid used in the treatment of inflammatory and autoimmune diseases that demonstrates unfavorable pharmacokinetics and side effects[51,52]. Simón-Vázquez et al. (2022) conjugated a palmitate group to dexamethasone to obtain dexamethasone palmitate (DXP) for improved encapsulation in poly(DL-lactide-co-glycolide)-polyethylene glycol (PLGA-PEG) NPs at high encapsulation efficiency (75–85%)[53]. Intravenous injections of DXP NPs exhibited significant reduction of arthritis score (6 to 1 and 6 to 4 for DXP NPs and dexamethasone sodium phosphate, respectively) over 6 days after treatment. The enhanced therapeutic efficacy can be attributed to sustained release of DXP from PLGA-PEG NPs, leading to high concentrations of dexamethasone for up to 18 h.

In some cases, a second drug can also serve as the lipophilic promoiety for LA-prodrug strategies. Liu et al. (2019) exploited this strategy by conjugating PTX to suberoylanilide hydroxamic acid (SAHA) and subsequently loaded the co-prodrug into PEG-PLA micelles[54]. The PTX-SAHA co-prodrug was designed to mask polar functional groups in both PTX and SAHA, leading to increased lipophilicity and improved encapsulation. This design also enabled potentially synergistic effects between two anticancer drugs that would otherwise be difficult to deliver concurrently. Interestingly, despite the stoichiometric conjugation of PTX and SAHA, the drugs showed differential release of 12% PTX and 40% SAHA over 48 h in vitro that was not explained by the authors. Minimal PTX-SAHA prodrug was detected, which indicates rapid conversion into the active drugs after release from the nanomicelle. This strategy can be applied to other diseases that require chronic treatment with combination therapy.

In summary, in this section we discussed how drugs with poor compatibility to the carrier can result in poor encapsulation and lead to rapid partitioning out of the carrier, limiting opportunities for slow dissolution. Conversion of free drugs into more lipophilic prodrugs improves encapsulation for a variety of carriers, but additional promoiety properties could result in improved loading and release for specific drugs and carriers. For example, promoieties that introduce specific hydrogen bonding functional groups or hydrophobic interactions may be useful in improving encapsulation. However, carriers used for LAF of drugs are often hydrophobic and LA-prodrugs derived from lipophilic promoieties are often employed[2].

2.2. Fatty Acyl Promoieties for Oil Depots

The first long-acting depots were developed using drugs solubilized in an oil depot, in which the rate of drug release is controlled by the partitioning from the oil phase into body fluids[55,56]. This strategy typically requires a dosing frequency of 2–3 weeks[57]. To further extend release from these depots, drugs were modified with fatty acyl promoieties to increase solubility in the oil phase. Esterification to fatty acyl chains is a common method to increase prodrug solubility in an oil depot[56,58]. Decanoate, enanthate, and caproate are common promoieties employed to increase the solubility in the oil vehicle and enhance partitioning in the fatty tissues[59]. By adjusting the length of the fatty ester promoiety the partitioning of the prodrug between the oil and aqueous medium can be tuned[60]. Oil depots loaded with fatty acyl prodrugs are distinct from the strategies discussed in Section 2.1 since the oil depot does not act as a carrier to control drug release. Instead, drug release is controlled by the rate of partitioning between the oil and aqueous medium and the subsequent cleavage of hydrolysis of the ester linker.

Fatty acyl prodrugs for oil depots have been used in the treatment for various conditions and diseases over the last several decades. Examples of long-acting antipsychotics that utilize this strategy include fluphenazine decanoate and haloperidol decanoate, where the parent drugs are conjugated to decanoate by an ester linkage to increase solubility in oil[19,20]. This strategy is also used in extending the duration of anabolic steroids, with examples such as nandrolone decanoate and testosterone undecanoate[61,62]. Other examples of treatment with fatty acyl prodrugs in oil depots are as contraceptives, treatment of menopause symptoms, and prostate cancer[63–65]. Research groups have continued to study other areas of treatment that can be addressed with prodrugs in oil depot. Gaekens et al. (2016) investigated this strategy in the long-acting treatment of alcohol disorder with a library of nalmefene prodrugs employing C8–16 fatty acid promoieties with cleavable ester or carbamate linkages[66]. Intramuscular injections of carbamate prodrugs in beagle dogs produced an initial burst release, possibly due to unconjugated drug, but was too stable to release drugs to the target plasma level (0.5–1 ng/mL). Gaekens et al. (2016) also developed an octadecyl glutarate diester prodrug, but it failed to confer additional advantage over the monoester prodrugs (octanoate, decanoate, dodecanoate, and palmitate). Gaekens et al. (2016) found that longer alkyl chain length of monoester promoieties led to more sustained nalmafene plasma concentrations. In particular, dodecanoate and palmitate nalmafene prodrugs provided sustained 4-week long sustained therapeutic levels at a dose of 5 mg nalmefene/kg. Thing et al. (2013) synthesized a lipophilic naproxen prodrug conjugated to an N,N-diethyl glycolamide ester promoiety to achieve sustained release via intraarticular (IA) injection[67]. The IA injection of prodrug solution maintained total joint concentration of naproxen at 350 μg/mL at 720 min, whereas IA injection of free naproxen solution yielded 1.6 μg/mL at 720 min, which is lower than the therapeutic plasma concentration of 42–54 μg/mL.

In general, LAFs based on oil depots using molecularly soluble species are cost-effective and can be manufactured simply. The multitude of examples of prodrugs in oil depots demonstrate how this strategy can be used to achieve extended duration of drug action to treat multiple diseases or conditions. However, they are ultimately limited the low drug loading and low injection volume for parenteral injections, and thus this strategy must be used for highly potent drugs[55]. Tuning drug release kinetics from oil depots is also highly limited[10]. In Section 2.3, we discuss fatty acyl and other monodisperse promoieties used in prodrugs for drug delivery systems that employ an assembled carrier with different mechanisms of extending the duration of action. Conjugation with fatty acyl promoieties can facilitate nanocrystal formation and formulation as nanosuspensions, in which duration of action is extended by slow dissolution, bioconversion to active API, and secondary depots[14,34,68]. Synthesizing prodrugs with improved encapsulation in other carriers is also a viable strategy to extend duration of action.

2.3. Fatty Acyl Promoieties for Solid Drug Dispersions and Micelles

Solid drug dispersions (SDD) are an important class of LAF for poorly water-soluble drugs that naturally self-assemble into amorphous or semicrystalline solids and are used for treatment in several diseases[10]. Most pharmaceutical nano/micron-sized SDDs result from top-down approaches for bulk drug comminution by milling or homogenization. Drug release from SDD is mediated by the slow dissolution of drug from the surface of these nanosized solids (Fig. 3). By reducing the particle size of solids, solubility and dissolution rate of poorly water-soluble drugs are greatly increased due to the increased surface area and decreased diffusion layer thickness[69]. As defined by the Nernst-Brunner equation, the diffusion coefficient is directly proportional to the dissolution rate[70,71]. The desired release rate of a drug from a SDD can be obtained by tuning the dissolution rate according to desired particle size and drug aqueous solubility. A lipophilic promoiety conjugated to a drug can decrease its aqueous solubility and dissolution rate from an SDD. SDD are almost entirely drug with minimal amounts of excipients to offer a “carrier-free” LAF [72]. This is especially advantageous compared with oil depots of fatty acyl prodrugs where the drug partitioning limits dissolution. SDD are formulated as suspensions stabilized with surfactants and injected subcutaneously or intramuscularly to form depots at the injection site. Drugs such as cabotegravir (CAB, logP=0.16[30]) and rilpivirine (RPV, logP=5.47[73]), have poor aqueous solubility, which makes it amenable for formulation as an SDD that has been successfully translated to clinical products[74,75]. For APIs with high aqueous solubility, conjugation to a fatty acyl promoiety generates a LA-prodrug that may facilitate formulation as SDD with tunable dissolution rates for extended duration of drug action. SDD formulated from LA-prodrugs have demonstrated clinical efficacy and are advantageous for long-term treatment of diseases[58,76]. For example, paliperidone palmitate (PP) is a prodrug of paliperidone conjugated with palmitic acid that results allows formulation as an SDD and results in slower dissolution in a LAF (Invega Sustena, Invega Trinza)[77]. Aripiprazole lauroxil (AL) is an N-acyloxymethyl prodrug of aripiprazole with decrease aqueous solubility due to its conjugation with lauric acid, which facilitates formulation as an LAF to treat schizophrenia (Aristada)[78]. Oral formulations of paliperidone and aripiprazole are dosed daily, whereas the SDD formulations are dosed once every 1–3 months[79–82]. The success of long-acting antipsychotic prodrug formulated as SDD has been a paradigm for other APIs such as antiretroviral (ARV) drugs. In addition, fatty acyl prodrug modifications of hydrophilic drugs have been successfully formulated as SDD.

Fig 3. Comparison of Oil Solutions and Nanocrystals of Fatty Acyl Prodrugs.

Conjugation of an example drug with a fatty acyl promoiety to generate a long-acting prodrug that can be used in formulating an oil depot or nanocrystalline solid drug dispersion. Dissolution from oil solutions is controlled by the partitioning of the LA-prodrug from the oil into tissue or body fluid where the active drug is released. Dissolution from SDDs is controlled by diffusion of drug at the surface of the solid.

A library of ARV drugs has been chemically modified to achieve various physiochemical properties into LAFs, which has been coined as LASER-ART (long-acting slow effective release antiretroviral therapy). LASER-ART entails dissolution of the LA-prodrug ARV from the SDD followed by hydrolysis of the prodrug from the promoiety to release the active pharmaceutical[83]. Conversion of LA-prodrug into the active species can be mediated by carboxyesterases enzymes that hydrolyze ester, carbamate, and amide linkages[84]. Long-acting prodrug strategies enable successful formulation of many available hydrophilic ARV drugs as nanocrysalline SDD to extend duration of drug action. For instance, abacavir (ABC), a relatively hydrophilic ARV drug (logP=1.20) [85], was successfully formulated as a LA-prodrug by conjugating it to myristic acid via an ester linkage. The poorly water soluble LA-prodrug (MABC) could be formulated into a nanocrystal SDD that prolonged therapeutic levels of ABC in plasma and reduced dosing frequency[86]. Myristic acid, a 14-carbon fatty acid, is a recurring promoiety used in LASER-ART to enable nanocrystal formulation as well as inhibiting enzymes necessary for HIV replication[86]. Tuning the fatty acid chain length offers a design lever to control dissolution rate by tuning aqueous solubility and hydrolysis rate of the LA-prodrug and release of the active drug. Deodhar et al. (2022) investigated the effect of fatty acyl promoiety carbon lengths (C14, C18, C22) on duration of action of dolutegravir (DTG). They found that C18-DTG prodrug SDDs afforded the longest duration of action as measured by plasma concentration levels remaining above PA-IC90 (in vitro protein associated-90% inhibition concentration) for 367 days[31]. The C14-DTG prodrug SDD also provided a long duration above PA-IC90 for at least 112 days, whereas C22-DTG prodrug SDDs maintained PA-IC90 levels for only about 7 days. The authors noted that the dissolution rate is driven primarily by the LA-prodrug aqueous solubility rather than the carbon length of the fatty acyl chain. To confirm the role of solubility, Deodhar et al. (2022) synthesized a C18-DTG prodrug and a C18 dimer DTG prodrug comprised of two DTG conjugated together via a C18 diester. The C18 dimer DTG prodrug exhibited a shorter duration of action as indicated by plasma concentrations dropping below PA-IC90 after <140 days. This is due to the significantly greater octanol solubility of the C18 DTG prodrug, indicating change in drug release rate was driven by solubility as opposed to solely the alkyl carbon length. LASER-ART have also been successfully applied to synthesize LA-prodrugs of lamivudine, tenofovir, emtricitabine, darunavir, and other ARV drugs for formulation as SDD with extended duration of action, thus demonstrating the wide variety of drugs suitable to this LAF strategy[87–89].

LASER-ART has also been applied to APIs that have been formulated as conventional SDDs and shown to result in promising attributes such as reducing injection-site reactions, large injection volumes, and frequent dosing [90–92]. For example, CAB is already clinically approved as a long-acting injectable SDD, and was used for LASER-ART by Kulkarni et al (2020). They showed that a single intramuscular injection of C18-CAB prodrug SDD in mice yielded concentrations above the PA-IC90 for at least 364 days[30]. These data along with favorable administration and manufacturing properties suggest that a C18-CAB prodrug SDD may be promising as a once-yearly injectable. These data also suggest that different acyl chain lengths may be suitable to tune the desired dissolution rate of drug depending on the inherent aqueous solubility of the parent API. Hilaire et al. (2019) also applied LASER-ART to RPV where they developed a library of N-acyloxyakyl RPV prodrugs with varying akyl carbon lengths (6–18 carbons)[93]. N-acyloxyalykl RPV prodrugs undergo bioconversion into N-hydroxymethyl RPV in the presence of esterases in vivo and is further converted into the parent drug after hydrolysis. Following intramuscular injections in BALB/cJ mice (45 mg/kg RPV), a single intramuscular injections of C14-RPV (NM3RPV) SDD maintained RPV concentrations above PA-IC90 (12 ng/mL) over 4, 12, and 25 weeks depending on the injection dose of 45, 75, 100 mg/kg, respectively. In contrast, nanoformulation of the parent drug (NRPV) displayed faster decay in plasma concentrations and maintained RPV plasma concentration above PA-IC90 for only 7 weeks at the highest dose (100 mg/kg). When comparing the duration of action of the two formulations, NM3RPV exhibited 13- and 26-fold increases in plasma RPV apparent half-life and mean residence time compared to NRPV. This study demonstrates that, even for highly hydrophobic and insoluble APIs that are readily formulated as SDD, synthesizing LA-produgs with altered solubity can further tune dissolution rates promote longer durations of drug action in tissue and plasma. In addition to promoiety architecture (alkyl carbon length, linker selection, etc.), other parameters have shown to influence the duration of action of LASER-ART. For example, particle size and choice of surfactant can influence cellular uptake of nanocrystals and generate intracellular drug depots[94]. Intracellular depots of nanocrystals provide a secondary depot within cells, and have been observed to provide greater antiretroviral activity for a longer duration compared to a depot formed at the injection site[86,95]. Poloxamers used as coatings to stabilize solid dispersions have also been functionalized with folic acid (FA) to facilitate folate-receptor targeted endocytosis and increase intracellular depots within macrophages[96]. In contrast to simple oil vehicles, LASER-ART and nanocrystal formulations offer multiple levers to tune to achieve the desired release rate depending on the inherent aqueous solubilty of the API.

While LASER-ART focuses primarily on ester linkages for the promoiety conjugation and high-pressure homogenization to obtain solid drug particles, other approaches recapitulate this general LA-prodrug design and extended drug duration outcomes. Xu et al. (2022) successfully modify paliperidone with a stearyl chain joined with a ketal linker (SKP-MC) and formulated this LA-prodrug into micron-sized SDD[97]. SKP-MC demonstrated longer duration of action than Invega Sustenna, a commercially available long-acting paliperidone that used ester linkages and a smaller particle size. Compared to Invega Sustenna, SKP-MC demonstrated a greater t_1/2 of up to 494.4 h and sustained paliperidone concentration out to 43 days. Xu et al. (2021) also used the stearoxyl ketal promoiety on dexamethasone and formulated an intraarticular LAF injectable to treat arthritis[98]. Available intraarticular glucosteroid injections such as micelles released drugs rapidly in the order of days[99]. By tuning the dose delivered and particle size, the desired drug release rate can be achieved to treat chronic arthritis. In addition to micron/nano-sized SDD, other carriers can be formulated from LA-prodrugs based on lipophilic monodisperse promoieties. Cheng et al. (2015) synthesized a LA-prodrug of cidofovir (CDV) by a covalent phosphate linkage to a hexadecyl promoiety[100]. The hydrophobic hexadecyl promoiety and the hydrophilic CDV self-assembles into a micelle, which demonstrated sustained release of CDV above its half maximal effective concentration (EC50, 0.9 nM) by at least 238 times (214 nM) for over 8 days in ex vivo models. Using a similar approach, Daman et al. (2014) synthesized stearoyl-gemcitabine (GemC18), which self-assembles into LA-prodrug micelles that release less than 20% of gemcitabine over 72 hours in vitro[50]. Compared to GemC18 loaded in PEG-PLA polymer nanoparticles, the micellar GemC18 exhibited slower release, which was attributed to the greater affinity of the stearic acid domains to form a more stable core. Oligomers also serve as promoieties that can be conjugated to obtain micellar structures for LAFs. Hydrophobic drugs conjugated to both termini of low molecular weight PEG results in micellar self-assembly, where the prodrug segment makes up the hydrophobic core and the hydrophilic segments extend outwards. Diclofenac conjugated to the end of PEG by cleavable ester linkages self-assembled into micelles for sustained anti-inflammatory effect[33]. In the presence of esterases, diclofenac is hydrolyzed from the PEG chains and inhibits acute inflammation in BALB/c mice out to more than 48 h.

Here we summarize LA-prodrug modifications formed by addition of a lipophilic promoiety to various APIs that will enable formulation of nano/micron-sized solid drug dispersions and micelles for extended duration of action. Depending on the physiochemical properties of the prodrug and the DDS formulation employed, a wide range of release behavior can be achieved. For example, ABC LA-prodrug SDD and HPD-CVD micelles resulted in extended duration of action on the order of days and weeks, whereas C18-DTG and C18-CAB LA-prodrug SDDs extends duration of drug action up to one year. Use of fatty acyl prodrugs is an effective and facile to enable LAF for many drugs, and it is amenable to different types of linkers. In addition, release and linker cleavage kinetics is highly tunable to achieve the desired release rate.

3. Long-Acting Prodrugs with Macromolecular (Mw>1000Da) Promoieties

In this section, we discuss long-acting prodrugs employing macromolecular promoieties (typically MW>1000Da) (Fig. 2). These strategies typically involve drugs directly conjugated to a reactive macromolecular entity (post-modification) and the direct polymerization of polymerizable prodrug monomers[101]. Specifically, we focus on specific polymer-drug conjugates that facilitate formation of higher-order structures or enable other LAFs for extending duration of drug action. In the following subsections, we delineate various types of carriers obtained from macromolecular promoieties. The first four subsections discuss macromolecular prodrugs obtained by postmodification of macromolecular promoieties. The last subsection discusses polymer prodrugs obtained by direct polymerization of polymerizable prodrug monomers.

3.1. Macromolecular Prodrugs for Long-Acting Hydrogels

Hydrogels are a type of drug delivery carrier made up of a large amount of water and a polymer network that exhibit excellent biocompatibility and a wide range of mechanical properties to mimic physiological environments[17]. Hydrogels are able to formulate drugs within the polymer network but suffer from rapid partitioning of drug into plasma or tissue. This leads to low encapsulation and burst release, particularly for small molecule drugs that can rapidly diffuse within the polymer network[102,103]. Covalent drug attachment to the hydrogel polymer is a strategy to immobilize the drug and prevent its rapid diffusion or release until linkage cleavage or in response to a stimulus[17]. Below, we examine examples of LA-prodrugs comprised of drugs conjugated to the terminal or pendant groups of polymers for the purpose of sustaining drug release from hydrogels (Fig. 4).

Fig 4. Comparison of End Group and Pendant Group Polymeric Prodrugs for LAF.

Polymeric prodrugs can be obtained by conjugating drug onto a polymer chain. How the drug is conjugated to the polymer backbone can result in the polymeric prodrug assembling in different ways. A) depicts a polymeric prodrug where the drug is conjugated to an end group of a polymer. B) depicts a polymeric prodrug where drugs are conjugated to pendant functional groups of the polymer. The left depicts the general scheme, and the right shows examples of such polymeric prodrugs.

PEG hydrogels have been studied extensively for drug delivery given its biocompatibility to mimic extracellular matrices and bioerodibility [17,104,105]. PEG contains end-terminal hydroxyl groups that are easily conjugated to hydrolytically cleavable or other cleavable linkers to obtain polymer-drug conjugates that can be used as LA-prodrugs when assembled into hydrogels. Benoit et al. (2006) developed a hydrogel of PEG conjugated to fluvastatin by a lactic acid spacer and a hydrolysable ester bond[24]. By tuning the length of the spacer, the accessibility to the hydrolysable ester linker tuned drug release from 7 days (6 theoretical lactic acid groups) out to ~17 days (2 theoretical lactic acid groups in linker). Ma and Zhang (2011) successfully conjugated indomethacin to the terminal hydroxyl group of PEG to obtain a PEGylated indomethacin prodrug (MPEG-indo)[106]. MPEG-indo was induced to form a hydrogel using cyclodextrin, which achieved in vitro release of indomethacin over 20 days. The authors concluded that the release of indomethacin from MPEG-indo can be increased by lowering the weight percent of MPEG-indo or cyclodextrin, which increases the dissolution properties by increasing the dissociation of the hydrogel.

Li et al. (2020) conjugated alendronate (ALN) to PEG by an amide rather than an ester linkage and then formulated the prodrug within a tetra-PEG hydrogel (tetra-PEG@PEG-ALN)[107]. PEG-ALN cannot move as freely within the hydrogel compared to ALN, thus preventing any initial burst release. The relative stability of an amide compared to an ester linkage also likely contributed to a slower release. The release of ALN was driven by the degradation of the tetra-PEG hydrogel as well as the hydrolysis of the PEG-ALN prodrug. The authors verified that osteroporosis induced by ovariectomy could be effectively cured over 85 days with an injectable tetra-PEG@PEG-ALN using an in vivo rabbit model. For certain drugs, conventional ester linkages are not suitable due to generation of unwanted byproducts, such as the generation of diclofenac lactam in diclofenac-PEG prodrugs[25]. Peptides have also been conjugated to PEG for formation of hydrogels that provide sustained release. Schneider et al. (2016) by conjugated exentatide to tetra-PEG, which released the peptide following cleavage of the β-eliminative linker[108]. In addition to tuning the release rate by changing hydrogel properties, Schneider et al. (2016) tuned the release by using electron withdrawing groups to adjust the cleavage rate of the β-eliminative linker[109]. A subcutaneous injection of the exenatide conjugated tetra-PEG hydrogel microspheres resulted in a t_1/2=160 h[108]. In contrast, subcutaneous injections of free exenatide has a t_1/2 of 2.4 h and is administered twice daily[110].

In addition to PEG, polysaccharides have also been used as promoieties for drug conjugation to facilitate sustained drug delivery from hydrogels [111]. For instance, chitosan (CS) has been conjugated with cefuroxime (CEF) via an ester linkage by physically mixing CEF with CS under acidic conditions[112]. Pawar et al. (2019) used the CS-CEF prodrug hydrogel for treatment of chronic wound infections and showed 43% release over 25 days under in vitro conditions. The CS-CEF prodrug platform further exhibited 5–6 log reduction of gram-positive bacteria following incubation, demonstrating strong antibacterial activity after conversion to free drug. Another polysaccharide that can be used as a promoiety is carboxymethyl cellulose (CMC), which has good compatibility with skin and mucous membranes and is also clinically approved for parenteral delivery. Capanema et al. (2018) conjugated doxorubicin (Dox) to the pendant carboxylic acid groups of CMC with amide bonds, which formed hydrogels when crosslinked with citric acid[113]. The release kinetic of Dox could be tuned by altering the degree of substitution with carboxymethylcellulose, which resulted in different degrees of swelling within the hydrogel and 18% and 25% cumulative release for up to 24 h under in vitro conditions. Fu et al. (2015) achieved sustained release of Dox from polysaccharide hydrogels based on hyaluronic acid (HA)[114]. In this example, Dox was thiolated to HA via an acid-labile hydrazone linkage and this conjugate prodrug formed a gel when exposed to air at room temperature due to pendant thiols forming disulfide crosslinks. In acidic conditions (pH=5.0), cumulative release of Dox was 29.8% over 72 h, while cumulative release was 11.2% in neutral conditions (pH=7.4). Acid sensitivity can provide stimuli-controlled release in certain active sites such as intracellular endosomes or within tumor microenvironments[115].

Poly(2-oxazoline) (POZ), which has been found to hydrolyze slower than PEG, can also be used for drug conjugations for sustained delivery from hydrogels[116]. The underlying hypothesis is that the flexible POZ chains fold around the hydrophobic drugs conjugated to the pendant functional groups, resulting in a “core-shell” that prevents esterases from easily accessing the cleavable ester linkers[116]. Various drugs and linkers can be coupled to the POZ backbone by click chemistry resulting in a drug-linker-azide compound. Although drug release rate can be controlled using linkers with different hydrolysis rates, Harris et al. (2019) found that hydrolysis rate also depended on the drug attached to the linker. For example, when using 3-proprionate as the linker and a similar drug loading, rotigotine (logP=4.9) exhibited a hydrolysis half-life of 76 h, while Δ⁹-tetrahydrocannabinol (Δ⁹-THC, logP=7.0) exhibited a much greater half-life of >7 days. The half-life of POZ-drug conjugates were greater than transdermal rotigotine patches (t_1/2=5.3–5.7 h) and Δ⁹-THC (t_1/2=1.3 d for infrequent users)[117,118]. POZ-rotigotine was used as a once weekly subcutaneous injections and showed potential to reducing episodes of motor fluctuations and dyskinesia in Parkinson’s disease[116]. For POZ-drug conjugates, hydrolysis rate is slowed with a more hydrophobic drug due to a more tightly folded conjugate. Inclusion of inert, hydrophilic molecules in pendant positions allowed the conjugate to swell and increase the hydrolysis rate.

Drug depots formed from thermoresponsive polymers that transition from a solution to a gel when their lower critical solution temperature is close to the body temperature have been investigated as LAFs. Zhao et al. (2021) conjugated dexamethasone (Dex) to N-2-(hydroxypropyl)methacrylamide (HPMA) by a hydrazone linkage[35]. Gelation of Dex-HPMA results from increased hydrophobic interactions at elevated temperatures, where hydrogen bonds between water and HPMA are disrupted, resulting in the formation of a ProGel-Dex hydrogel. In vitro release demonstrated slow release of 2–10% Dex over 28 days, which can be explained by the hydrophobic aggregation and limited exposure to the release media. Following intraarticular injection in arthritic rat models, ProGel-Dex presence within arthritic joints were maitained for more than a month, and resulted in significant and progressive reduction in joint swelling when compared with free Dex. Thermoresponsive hydrogels have also been used for the sustained delivery of cisplatin by covalent conjugate of Pt(IV) to the hydrophobic end of two methoxyl PEG-b-PLA copolymer chains[119]. In aqueous conditions, the resulting Bi(mPEG-b-PLA)-Pt(IV) conjugate formed micelles with Pt(IV) conjugated to PLA within the core, which can subsequently undergo thermogelation to obtain a semisolid gel. This property is advantageous because it is easily injectable and forms a depot in situ due to micellar aggregation in physiological conditions. In addition, while free cisplatin encapsulated in mPEG-PLA demonstrated almost instantaneous release in vitro, the covalently linked cisplatin exhibited up to two months of release. The LA-prodrug not only extended the release, but also enhanced cytotoxicity against tumor cells. The enhanced potency of the Bi(mPEG-b-PLA)-Pt(IV) was attributed to the greater cellular uptake, which was confirmed via in vitro CCK-8 assays and cellular uptake assays. Upon intracellularization within cancer cells, Bi(mPEG-b-PLA)-Pt(IV) can be reduced to release Pt(II) to bind to the DNA[120].

3.2. Macromolecular Polymer Prodrugs for Nanoparticles

Long-acting formulations based on polymeric nanoparticles can provide tunable and sustained drug release as well as improve drug safety and stability[121]. Long-acting prodrugs can broaden the APIs that are amenable to formulation in polymeric nanoparticles by improving drug loading and preventing premature drug release[122][123]. Selection of covalent linkers that vary in hydrolysis rate as well as incorporating steric effects to hinder water or enzyme accessibility to the linkers are strategies to mediate API release from LA-prodrug nanoparticles[123,124]. Long-acting prodrug NPs can be fabricated with a variety of polymers such as polyacetal [125], chitosan [126], hyaluronic acid [127], PEG [128,129], and PCL [126].

LA-prodrugs can be physically encapsulated as well as conjugated to the polymers comprising the nanoparticle (NP). For instance, arginine-conjugated chitosan polymer (CHITARG) were prepared by linking the end-groups of chitosan to arginine and then nanoparticles were precipitated in water at pH=3.5[130]. Arginine was further loaded into the NPs by mixing CHITARG with an arginine solution prior to NP precipitation. CHITARG did not exhibit high encapsulation efficiency due to the repulsive positive charges between arginine and chitosan. In vitro arginine release from CHITARG was 20% in the initial 2 h followed by a slow and sustained release out up to 12 h. The authors associated the phase of slow arginine dissolution to be due to arginine clusters having a greater affinity to CHITARG as opposed to the aqueous medium.

LA-prodrug designs can also enable combination delivery of physicochemically diverse APIs co-encapsulated into the same long-acting nanoparticle formulation. For example, cisplatin (CDDP) and curcumin (CUR) were co-formulated in nanoparticles by first synthesizing a PLGA-CDDP conjugate and using a layer-by-layer (LBL) technique with glyceryl monostearate to encapsulate CUR[23]. The resulting CDDP-PLGA/CUR LBL NPs near complete release of both drugs at 72 h. In contrast, unconjugated CDDP/CUR LBL NPs reached near complete release of both drugs at 48 h, demonstrating that the CDDP-PLGA prodrugs produced more sustained drug release. By conjugating CDDP to PLGA, release of CDDP first required hydrolysis of the ester linkage and dampened the rate of CDDP dissolution. From the observation of tumor inhibition in NSCLC bearing mice over 21 days. CDDP-PLGA/CUR exhibited significantly greater tumor inhibition efficacy (76.7%) than CDDP/CUR LBL NPs (56.4%) or free CDDP/CUR (23.1%).

3.3. Polymer Prodrugs for Micelles

Micelles fabricated from polymer-drug conjugate prodrugs have been studied extensively and is an effective strategy to sustain release of drug to an active site[131]. Here, we describe micelles formed from LA-prodrugs designed with higher molecular weight (HMW) polymer promoities rather than the fatty acyl or low molecular weight PEG species described in Section 2.3. Micelles can be obtained by amphiphilic block copolymer self-assembling in aqueous conditions, where cleavable linkers connect the drug to the polymer[132]. This strategy is an effective way to extend the release, as there are now two steps to drug release: 1) the drug must be cleaved from the polymer, and 2) the drug must diffuse from the micelle. LA-prodrugs with HMW polymers also improve the encapsulation of drug within the micelle and the stability of the drug within the micelle.

One early example of polymeric prodrug micelles is a Dox-PLGA-PEG micelle developed by Yoo and Park (2001). Dox covalently conjugated to PLGA-PEG was loaded in micelles at a much higher efficiency (2.18%, w/w) compared with physical encapsulation of the parent Dox (0.51%, w/w)[133]. Physically encapsulated Dox micelles only sustained release for 3 days, whereas Dox-PLGA-PEG micelle exhibited 50% cumulative release over 2 weeks. Using a similar strategy, cabazitaxel (CTX) was covalenty conjugated to PEG-PLA to form micelles for sustained tumor inhibition[21]. Micellar CTX-polymer conjugates (mPEG-PLA-CTX) exhibited greater drug loading (10.57%) than physically encapsulated CTX micelles (CTX-NPs, 8.04%). mPEG-PLA-CTX also sustained CTX release in vitro and showed tumor inhibition in vivo while reducing systemic toxicity by preventing premature rapid drug release above the tolerated dose. Polymer prodrug micelles can also be obtained by conjugating drugs to pendant rather than end-groups of polymers to also facilitate micellar self-assembly. Conjugation to functional groups attached to the polymer backbone may be difficult due to steric hindrances, but Pang et al. (2014) used an ester linkage to conjugate quercetin (QT) to hyaluronic acid (HA) using an adipic acid dihydrazide spacer attached to HA by carbodiimide coupling [134]. This polymeric prodrug self-assembled into micelles (HA-QT) in aqueous conditions where the hydrophobic QT was buried within the inner domain, and the hyaluronic blocks were at the surface interacting with the aqueous phase. Without esterases, HA-QT micelles released only 21% of QT over 96 h in vitro. HA-QT micelles increased plasma half-life by 23.3 times and the AUC by 4.9 times in rats compared to free QT injection. Tumor growth in mice was also inhibited by 62.9% over 10 days when injected with HA-QT, versus just 25.1% with free QT. Other chemical strategies allow for specific drug conjugation to polymer pendant groups that generate unique micellar properties to facilitate extended duration of action.

Indomethacin was conjugated to PEG via carbodiimide-mediated coupling to form amide linkages. In aqueous conditions, the amphiphilic IND-PEG-IND conjugate self-assembled into nanoparticles where water-insoluble IND was buried within the interior while the hydrophilic PEG formed the exterior. The prodrug undergoes conversion when the amide linker can be hydrolyzed in the presence of cathepsin B, which is a protease overexpressed in metabolically active bone microenvironment induced after inflammation[135]. Following cleavage of the amide linker, IND is released through diffusion of the IND-PEG-IND NP. In vitro release experiments demonstrated rapid release of free IND in PBS buffer at pH=7.4 and 37°C, wherein 95% cumulative release was observed within 24 h. The prodrug was observed to have 20% cumulative release within 24 h, which is a slower release compared to the free IND. In the presence of cathepsin B, the release of IND is greatly accelerated, thus enabling stimulus-triggered release of IND at the active site. In vitro anti-inflammatory experiments demonstrated that IND-PEG-IND inhibited TNF-α and other pro-inflammatory markers (MCP-1 and IL-6) to a greater degree than free IND, indicating an improved therapeutic efficacy with the usage of polymeric prodrug micelles.

Synergistic co-delivery of drug combinations using micelles can promote greater therapeutic efficacy in certain diseases such as cancers[136]. Use of polymer-drug LA-prodrugs can facilitate certain drug combinations that are difficult to co-formulate in micelles. Coumarin and 5-FU were conjugated to PEG chains by click chemistry to form amphiphilic conjugates (CouPEG and 5-FUPEG, respectively)[137]. CouPEG and 5-FUPEG amphiphiles self-assembled into micelles that physically encapsulated GEM, PTX or curcumin (CUR), where the weight ratio of drug and LA-prodrug was used to tune drug loading. Although GEM was released rapidly due to micelle instability, CouPEG and 5-FUPEG released CUR and PTX at a steady rate. As some drugs cause micelles to become too unstable, one strategy to further stabilize micelles is by using cross-linking groups in the micelle core to prevent premature micelle dissociation. For example, a PEG-Dox micelle encapsulating curcumin was synthesized by conjugating an aldehyde-terminated PEG to the amine group of Dox to obtain an oxime linkage[138]. In vitro release demonstrated minimal release (>20%) of Dox and CUR under neutral (pH=7.4) conditions, while release of both drugs was rapid under acidic conditions (pH=5). In a similar approach, a pH-responsive polymer prodrug micelle (SN38@P(Asp-(Hyd-Dox))-Se₂) was prepared by conjugating mPEG to aspartic acid, which is then conjugated to Dox via a hydrazine linkage[139]. SN38 was physically encapsulated into the micelle during its self-assembly and then further crosslinked with diselenide bonds by reacting 3,3’-diselanediyldipropanionic acid. The crosslinked micelles exhibited a more prolonged release of Dox and SN-38 (cumulative release of ~10% in PBS) compared to non-crosslinked micelles (cumulative release of ~20% in PBS).

3.4. Macromolecular Prodrugs for Use with Dendrimers

Dendrimers are a class of highly branched macromolecules with well-defined architecture with surface functional groups that offer sites for covalent attachment of drugs to obtain dendrimer prodrugs[140]. Conjugation of drugs to dendrimers can increase the bioavailability and decrease dosing frequency[141]. Among dendrimers, polyamidoamine (PAMAM) has been studied extensively as a drug carrier, where drugs can be directly conjugated to the surface functional groups or indirectly conjugated using a linker[142,143]. The surface functional groups of PAMAM can be substituted with hydroxyl groups, which have been used to form ester-cleavable drug for slow release of ibuprofen[144] and erythromycin[145]. Kurtoglu et al. (2010) noted the amide linkers are highly stable and resulted in modest release of only 3% in acidic conditions (pH=1.2)[144]. The surface of PAMAM dendrimers can also be functionalized with a hydrazine to facilitate drug conjugation via a hydrazone bond, which has greater cleavability than amide bonds. Chang et al. (2012) functionalized PAMAM dendrimers with Dox linked by cleavable hydrazone bonds, which were highly stable under neutral pH but easily cleavable under acidic conditions (pH=5.03) and released 75% of Dox over 15 h[146]. Other cleavable linkers that can be used for sustained release of drugs include enzyme cleavable linkers[147] and disulfide reducible linkers[26]. Satsangi et al. (2014) conjugated paclitaxel to PAMAM via a GFLG tetrapeptide linker that can be cleaved in the presence of cathepsin B[147]. Xu et al. (2016) conjugated doxorubicin to PAMAM dendrimers via disulfide bonds, which can be reduced to release Dox in the presence of intracellular glutathione[26]. PEG spacers have also been used to conjugate camptothecin to PAMAM dendrimers by click chemistry, resulting in slow release with zero order kinetics over 7 days (10.6%) via ester hydrolysis[148].

Dendrimers can also be covalently crosslinked to obtain greater control for sustained delivery. Wang et al. (2019) developed a PEG-diacrylate that crosslinked surface amine groups by Michael addition to form a gel network[149]. Camptothecin was conjugated to dendrimer prior to crosslinking and released from the gel by cleavage of the ester linkages. Self-cleavage of ester linkages driven by free amine groups drives the generation of free camptothecin within the hydrogel, and diffusion of camptothecin is controlled by the hydrogel properties. Camptothecin-dendrimer hydrogels (DH-G3-CPT) exhibited prolonged tumor inhibition in a mouse model, and more effective tumor inhibition compared to physically encapsulated camptothecin loaded into a blank crosslinked hydrogel dendrimer (CPT/DH-G3).

Dendrimers synthesized from other polymers have also demonstrated sustained release. England et al. (2017) modified the surfaces of poly-L-lysine dendrimers with polyoxazoline, which was used to conjugate to SN38 by click chemistry[150]. Release rates were controlled using various linkers based on in vitro studies: primary amine carbamate (fast), ester (medium), and secondary amine carbamate (slow). These conjugated dendrimer prodrugs were compared with free iriniotecan (a prodrug that is converted to SN-38) delivered at 25 mg/kg. After weekly intravenous dosages, ester conjugated SN38-dendrimers (DEND-38) sustained release of SN38 and caused tumor regression by 69%, while free irinotecan did not reduce tumor size. The authors also studied how dosage affects tumor growth regression. Weekly injections up to 21 days of conjugated dendrimers caused tumor regression of 98% at 55 days and 96% at 48 days with 4 mg/kg and 8 mg/kg dosages of DEND-38. In contrast, weekly intravenous injections of irinotecan at 50 mg/kg mg slowed tumor growth relative to saline injections, but failed to decrease tumor volume. In a separate study England et al. (2020) also functionalized the surface of poly-L-lysine dendrimers with poly(sarcosine) for “stealth” properties[151]. Three weekly doses of intravenously delivered poly(sarcosine)-SN38-dendrimers in mice caused near-complete tumor regression, whereas free irinotecan, a prodrug that is converted to SN38, was unable to inhibit tumor growth. The increased efficacy was reflected in the greater SN38 concentration in relevant tissues such as the spleen and tumor compared to injection of free irinotecan, a prodrug that is converted to SN38, where SN38 concentration was minimal.

3.5. Macromolecular Polymers Comprised of Prodrug Monomers

In this section, we introduce the strategy of synthesizing polymer prodrugs by conjugating a drug to a polymerizable promoiety to obtain a prodrug monomer, and then subsequently polymerizing the prodrug monomers to yield a well-defined polymeric prodrug. Polymerization can be either through polycondensation, RAFT polymerization, etc. The strategy of synthesizing macromolecular prodrugs from polymerizable prodrug monomers results in a larger fraction of active agent in the final product compared to other polymeric drug delivery formulations[152,153]. The previous sections described polymeric prodrugs obtained by methods that postmodify fully synthesized polymers, which may sometimes leave significant portions of the reactive groups unreacted with drug resulting in low drug loading. Strategies to incorporate drugs during polymerization of the polymer have been termed “polyprodrugs” and “drugamers”[154,155]. These strategies focus on using the drug itself as one of the building blocks to obtain a well-defined polymeric prodrug.

3.5.1. Polyanhydrides, Polyanhydride esters, and Other Polymer Prodrugs with Drugs within the Backbone

Polyanhydrides (PA) or polyanhydride esters (PAE) can be used as polymeric prodrugs with the drug incorporated as monomers in the polymer backbone, which is released as the backbone degrades[156]. Use of prodrugs within the polymer backbone allows for high drug loading (50–80%) in a reproducible manner[27]. PA/PAE prodrugs release drugs from the polymer backbone by hydrolysis, and thus are inherently biodegradable and can be used as drug eluting implant that does not require surgical removal[157]. PA can be obtained by ring-opening polymerization, transesterification, and melt condensation. Rosario-Meléndez et al. (2012) synthesized a collection of PAE prodrugs for extended release. For example, PolyMorphine is a PAE prodrug synthesized by melt-condensation of morphine monomers at 170°C and resulted in polymers with high weight average molecular weight (Mw) (26000 Da), low PDI (1.14), and high yields (70%)[27]. Morphine monomers were synthesized by a ring-opening reaction of glutaric anhydride to obtain morphine-based diacid, which is followed by acetylation with acetic anhydride to obtain two terminal anhydride groups. In the presence of water, the anhydride bonds undergo hydrolysis to release the morphine diacid monomer prodrug, which is further hydrolyzed to release glutaric acid and free morphine. In vivo experiments of intraperitoneally injected PolyMorphine in mice demonstrated pain relief for up to 3 days, which was more than 20 times the analgesic window of free morphine. In addition, Gulrajani et al. (2022) developed long-acting salicylic acid (SA)-based PAEs (SAPAEs) that can release SA in a controlled manner[158]. SAPAEs were also polymerized via melt condensation at 180°C to obtained polymers with Mw=17800 Da, PDI=1.3, and a SA loading of ~75%. SA monomers were synthesized by reacting the hydroxyl group with the two carboxylic acid groups of sebacic acid, which were then polymerized together by condensation polymerization to obtain SAPAEs. SAPAEs were then formulated as disks and demonstrated extended delivery of SA with near zero-order release rate of 0.11–1.7% SA/day depending on the pH (pH=2–8). SAPAEs exhibit a two-step hydrolytic degradation to release SA starting with the cleavage of the anhydride to obtain oligomers that are subsequently broken down to a SA diacid that is further hydrolyzed to obtain two molar equivalents of SA and one molar equivalent of adipic acid. Other drugs that have been successfully incorporated in PA/PAEs include antimicrobials (carvacrol, thymol, and eugenol)[159] and antioxidants[160]. Heyder et al. (2021) synthesized poly(anhydride-ester) gemcitabine (poly(GMT)) by polymerizing a modified dianhydride GMT monomer with Mw=19740 Da and PDI=1.6[161]. Poly(GMT) exhibited much greater drug loading (58%w/w) compared to physically encapsulating GMT in nanoparticles and liposomes (~5%w/w), demonstrating the high drug loading that can be achieved by polymerizing drugs into PAEs. Poly(GMT) was further formulated into a nanoparticle and demonstrated long-acting delivery with zero-order release of 33% GMT after 45 days, making the platform a potential treatment for tumors. Shakil et al. (2022) used step growth polymerization techniques to obtain poly(carbamate-carbonate) prodrugs that releases emtricitabine (FTC) upon hydrolysis of the polymer backbone[162]. Similar to the PAE prodrugs, Shakil et al. (2022) achieved high drug loading (>57 wt%) by incorporating GMT within the polymer backbone. The authors used tris(chloroformate) of trimethyloyl propane (TMP) in the step-growth polymerization, which resulted in a crosslinked polymeric prodrug. The crosslinked polymeric prodrug over 3 days. Poly(carbamate-carbonate) prodrugs could also potentially be used as long-acting drug delivery devices without surgical removal due to their biodegradability.

3.5.2. RAFT, ROMP, and ROP Polymerization to Obtain Polymers Comprised of Prodrug Monomers

Polymerizable prodrug monomers can be used to obtain well-defined block copolymers to formulate micellar polymeric prodrug with improved encapsulation and drug stability. Polymerizable prodrug monomers are obtained by modifying the drug with moieties such as acrylamide and methacrylate that promote drug polymerization to the pendant groups of the polymer backbone[163,164]. RAFT polymerization is especially useful in obtaining polymer prodrugs with low polydispersity and precise control of molecular mass relative to other polymerization techniques (Fig. 5) [165]. In this subsection, we specifically discuss examples of polymer prodrugs obtained via RAFT polymerization and exclude other RAFT polymer prodrugs if the drug was conjugated to the polymer by post-modification techniques which has been discussed in other reviews[165]. In contrast to strategies discussed in subsection 3.5.1, RAFT results in polymer prodrugs conjugated to pendant groups as opposed to integrated in the polymer backbone. For example, Hasegawa et al. (2013) used RAFT to polymerize ibuprofen-acrylamide (IBUAAm) prodrugs for micelles[163]. IBUAAm was obtained by conversion of the carboxylic acid of IBU into an acyl chloride group, and linked to acrylamide by an ester bond. Block copolymerization of IBUAAm using RAFT was facilitated with a PEG macromolecular chain-transfer agent (PEG-CTA, MW=10k Da). Hasegawa et al. (2013) successfully polymerized IBU polymers with the number average molecular weight (Mn) ranging from 9706–26410 Da and PDI ranging from 1.15–1.20. IBU polymers improved the stability of ibuprofen within the micelle compared to physically encapsulated drug and resulted in release of days to months depending on the length of the hydrophobic drug segments of the copolymer (≤ 20 and 53 units). Interestingly, the use of longer hydrophobic segments resulted in the formation of worm-like micelles rather than spherical micelles, which may also affect drug release kinetics.

Figure 5. General Scheme of Drugamer or Polymers Comprised from Polymerizable Prodrug Monomers.

The scheme (top) describes conversion of a drug into a prodrug monomer that is subsequently polymerized to obtain a “drugamer”. An example drugamer (bottom) synthesized by RAFT using a TAF monomer[29].

Extensive research has been conducted on engineering micelles for improved delivery of drugs from different polymer architectures generated by RAFT polymerization. Guo et al. (2017) presented a novel poly(triethylene glycol methacrylate)-b-poly(podophyllotoxinmethacrylate) copolymer (PTP) that self-assemble into prodrug nanoparticles[166]. Here, the authors conjugated podophyllotoxin to a methacrylate group to obtain the polymerizable prodrug monomer (PODMA). PODMA was then polymerized by block copolymerization with triethylene glycol methacrylate, yielding hydrophilic and hydrophobic segment lengths with Mn of 8500–27200 Da and PDI 1.16–1.49. The resulting copolymer drug loading could be varied from 0–67% depending on the length of the poly(podophyllotoxinmethacrylate) block. Guo et al. (2017) selected PT₄₅P₁₄ (drug loading=40%) as the model prodrug given that it cannot be dissolved in water and could be dispersed, and successfully achieved sustained release over 72 h in vitro. An apomycin (APO)-loaded micelle (cPAM) was fabricated from the diblock copolymerization of allyloxy-PEG with APO-2-hydroxyethyl methacrylate by RAFT polymerization[28]. Furthermore, copolymers were functionalized with a scar-homing peptide (CAQK) and allowed to self-assemble in water to obtain cPAM. Fluorescent imaging verified greater localization of the micelles in the spinal cord after intravenous administration of cPAM compared to micelles without targeting peptides. The cPAM also demonstrated 35% of APO release by degradation with esterases over 24 h, demonstrating controlled drug release. Mice that received cPAM intravenously after thoracic spinal cord contusion exhibited significant motor function recovery 30 days after injury compared to free drug and non-functionalized micelle (Basso Mouse Scale score system of 2.2 ± 0.3 versus 1.3 ± 0.3 and 1.3 ± 0.3, respectively).

Codelivery is also possible by synthesizing copolymer prodrugs micelles that physically encapsulate a second drug. Xu et al. (2018) synthesized a diblock copolymer prodrug micelle where the hydrophobic segment is made up of suberoylanilide hydroxamic acid (SAHA) prodrug[167]. Diblock copolymerization was achieved by using a poly(oligo(ethylene glycol)) (POEG) macro chain transfer agent to polymerize SAHA-monomers, wherein POEG made up the hydrophilic segment. Dox was physically encapsulated within the micelle upon self-assembly of the diblock copolymer (POEG-b-PSAHA/Dox) in aqueous conditions. Compared to free Dox HCl, the co-delivery of Dox and SAHA via POEG-b-PSAHA/Dox inhibited 4T1.2 breast tumor growth in female BALB/c mice over 23 days. Free Dox HCl and Doxil, a form of Dox delivered in liposomes, decreased tumor volume to 500–900 mm³, whereas POEG-b-PSAHA/Dox decreased the tumor volume to 200–300 mm³. The increased antitumor effect of POEG-b-PSAHA/Dox was attributed to the synergistic effect of SAHA and Dox and improved accumulation at the tumor site.

Auriemma et al. (2021) also synthesized an APO copolymer prodrug by RAFT polymerization of the hydrophilic zwitterionic phosphorylcholine methacrylate (MPC) with variable amounts of HEMA-caprolactone (HEMA-CL5) and diapocynin cross-linker monomer[168]. The diapocynin cross-linker monomer is synthesized by dimerizing APO, and then conjugating mono-2-methacryloyloxy succinate on the two free hydroxyl groups of the dimer by ester bonds. This conjugation made the dimer amenable for radical chemistry at both ends, which can be used to crosslink the hydrophobic polymer core of the micelle. The prodrug self-assembles into a NP in aqueous environments, and releases covalently linked diapocynin upon hydrolysis of the ester linkages. The prodrug NP extended release of diapocynin for up to 9 days in in vitro conditions compared to the non-crosslinked NP with physically encapsulated diapocynin. However, Auriemma et al. (2021) conceded that diapocynin drug loading was less in the polymer prodrug compared to physically encapsulated diapocynin (1.1% and 3.9%, respectively), but the polymer prodrugs still offered significant advantages in terms of release rate. The authors noted the mechanism of release is dependent on NP hydrolytic degradation, where ester bonds within the core are hydrolyzed in water, which causes NP swelling and further dissolution[169]. When the degradation of the oligoesters from caprolactone and diapocynin is complete, the water-soluble poly(MPC)-b-poly(HEMA) backbone can be easily excreted, ensuring no polymer accumulation.

After cleavage of the pendant drugs, the backbone of many conventional polymeric prodrugs remains due to it being nonbiodegradable. Thus, Joubert and Pasparakis (2020) developed a biodegradable polymeric prodrug with an ester bond within the polymer backbone by using RAFT polymerization of a GEM-monomer with 2-methylene-1,3-dioxepane (MDO)[170]. Joubert and Pasparakis (2020) synthesized a library of copolymers with Mn ranging 6840–25693 Da with PDI of 1.17–1.71. GEM-monomer content in the copolymer ranges from 30–75%. P(GEM-co-MDO) demonstrated backbone biodegradability with 41–81% decrease in molecular weight in acidic or basic conditions, depending on the content of MDO within the backbone. The degradability of p(GEM-co-MDO) was corroborated with the shifting of size-exclusion chromatography (SEC) traces to an increased retention time after an incubation period in pH=11 over 24 h. In acidic environment (pH=5.2) and in the presence of Cathepsin B, GEM was released at a pseudo zero-order rate (47–50% cumulative drug release after 25 days). MDO also modulated the release kinetics of GEM from the polymer due to the increase in hydrophobicity of the polymer. Joubert et al. (2020) noted that the slow release of GEM is due to the two-step hydrolysis: hydrolysis of the ester releasing the GEM-prodrug, and then the rate-limiting hydrolysis of the amide linking GEM to the succinate group[164].

In addition to RAFT, techniques such as ring-opening metathesis polymerization (ROMP) use prodrug monomers to produce high molecular weight polymer prodrugs [171]. Venu et al. (2021) used ROMP to polymerize norbornene monomers covalently linked with chlorambucil (CL), folate (FOL), and methyl orange (METH) as a triblock copolymer of with controlled molecular weight (Mn=34100 Da) and narrow PDI (1.06)[172]. In aqueous conditions, the triblock copolymer self-assembled into micelles with FOL as a surface targeting ligand, and CL and METH in the hydrophobic core. CL and METH exhibited a pH-responsive release in vitro with minimal release at neutral pH of <10% but higher release of >55% at pH=5.5. Liu et al. (2015) polymerized chlorambucil (CL) and camptothecin (CPT) prodrug monomers with a mPEG-CTA using ROMP to make diblock copolymers with Mn ranging from 10600 to 17400 Da with PDI ranging from 1.17–1.29[173]. Prodrug monomers were synthesized by disulfide conjugation with a 5-methyl-2-oxo-1,3-dioxane group. The resulting diblock copolymer self-assembled into micelles with mPEG chains at the surface to enhance stealth properties for targeted delivery to tumors. The redox-responsive polymer prodrug micelle released <20% in oxidizing conditions (absence of glutathione) but rapidly released ~75% of drug in reducing conditions (presence of glutathione). The authors suggested that polymer prodrug micelles have longer plasma circulation compared to free drug due to the stealth-like properties from PEG based on an observed nine-fold increase in plasma AUC compared to free CPT injection observed from intravenous injection.

Das et al. (2016) used RAFT polymerization to synthesize various polymeric prodrugs of ciprofloxacin (cipro) with different polymeric architecture and linkers[174]. Cipro prodrug monomers were copolymerized with poly(ethylenegylcol methacrylate) (950 Da) (O950) as diblock copolymers and random copolymers to observe the effect on micelle self-assembly. The diblock copolymer self-assembled into micelles in aqueous conditions where the hydrophilic corona of O950 encapsulated the hydrophobic polymer segments of the prodrug. The diblock polymer architecture was observed to have reduced ester hydrolysis rates relative to the random copolymer prodrug, which was claimed to be a result of hindered water penetration into the hydrophobic core and demonstrated the importance of polymer architecture on sustained release. Das et al. (2016) also used a fast-hydrolyzing phenyl ester (CPM) linker and slow-hydrolyzing aliphatic ester (HBC) linker to further modulate cipro release. Copolymers containing CPM and HBC linkers show that 50% of drug was released over 120 h and 21 d, respectively. An inhalable version of the polymer prodrug was also developed where the cipro methacrylate monomer was conjugated to a protease-cleavable valine-citrulline (VC) linker with either a self-immolative spacer or phenyl ester (CTM) linker. Copolymerization via RAFT was carried out with the prodrug monomers with different linkers and a hydrophilic mannose comonomer, which acts as both the hydrophilic component to improve solubility and a targeting ligand for alveolar macrophages[155]. Mannose receptors on the drugamer induced receptor mediated endocytosis by alveolar macrophages, which is followed by cleavage of the VC linker by intracellular cathepsins to release the active cipro compound[175]. After intratracheal aeroslization of drugamers in mice, both the VC and CTM-drugamer exhibited sustained concentration of cipro within the lung at concentrations greater than the minimum inhibitory concentration (0.065 μg/g), whereas free cipro was quickly cleared from the lungs within 2 h. VC-drugamer also exhibited superior AUC over 48 h compared to CTM-drugamers and free cipro (25.6x and 74.8x, respectively), and sustained intracellular delivery to alveolar macrophages at >3 magnitudes greater than the minimum inhibitory concentration. In a challenge study with Burkholderia thilandensis, 100% of the mice administered the VC-drugamer survived 14 days after bacteria innoculation compared with only 25% survival of mice administered free cipro[175]. The sustained delivery of cipro up to 7 days improved biodistribution, and resulted in a dose-sparing regimen that illustrates the drugamer platform as promising candidate for further clinical development.

Ho et al. (2021) also used polymer prodrug comprised of polymerizable prodrug monomers to formulate injectable depots to achieve long-acting delivery of ARV drugs[29]. The authors synthesized a library of tenofovir alefanamide (TAF) drugamers with drug wt% of 34–73%. TAF-drugamers were subcutaneously injected to form a depot capable of sustaining release over an order of months. A single subcutaneous injection of the TAF drugamer sustained release of TAF, which accumulated intracellularly as the active metabolites (tenofovir diphosphate) for up to 50 days. The release rate of TAF was tuned by using a faster-releasing carbamate linker (p(benzyl-TAFMA)) or slower-releasing carbamate linker (p(alkyl-TAFMA). In addition to controlling release rates with different linker chemistry, local environments in the depot may result in differential water and enzyme activity that will affect hydrolysis of the carbamate linkers. The authors describe that TAF is released from the polymer backbone upon hydrolysis of the carbamate linkers and results in hydration of the pendant functionalized hydrophilic chains that are shed from the depot. Diblock designs of p(glycerolmonomethacrylate)-b-p(alkyl-TAF-methacrylate) and p(glycerolmonomethacrylate)-b-p(benzyl-TAF-methacrylate) incorporate glycerol segments to increase hydrophilicity of the copolymer, which may effect depot structure and stability in a way that promotes faster dissolution. The drugamers developed by Ho et al. (2021) were also amenable to further encapsulation in other drug delivery carriers. Drugamers have also been integrated with delivery systems other than micelles and depots. For example, Dart et al. (2022) formulated these same TAF drugamers in PCL electrospun fibers via blend electrospinning[176]. Sustained release of the active metabolite from these electorspun fiber patches was observed over a 2-month period under in vitro conditions.

4. Summary and Outlook

Long-acting drug delivery strategies address the need to improve the drug dosing needed to maintain long-term drug concentration within the therapeutic index to improve patient adherence for treatment of different diseases. However, many approved drugs have physicochemical properties not amenable to traditional long-acting formulations. Here, we reviewed various long-acting prodrugs that expand the inclusion of these APIs for long-acting formulations. A key requirement for the use of the strategies presented in this review is that functional groups on the parent drug be functionalized with cleavable linkers. For example, fourteen of the fifteen FDA-approved small molecule drugs in 2022 have functional groups that can potentially be chemically modified to obtain long-acting prodrugs discussed in this review[177]. LA-prodrugs can be designed to enable a variety of long-acting formulations that provide different timescales for drug delivery, and thus the challenge is to select the ideal LA-prodrug strategy for a given application. For example, although nanocrystals offer the greatest increase in duration of action, it is not suitable for some anticancer drugs as it can have off-target effects when delivered systemically and is not reversible in the case of adverse reactions[178]. LA-prodrugs involving high molecular weight macromolecular promoieties show a great diversity of possible modifications to tune drug release rate with different cleavable linkers and polymer architectures. However, for these systems, careful consideration must be taken to assess factors alone and in combination to tune for optimal drug release. Lastly, consideration of drug properties should be incorporated early during drug development in order to make a drug amenable for long-acting formulation. Many drugs in development are optimized such that they can be bioavailable following oral administration, but these optimizations may result in drugs not suitable for LAFs or LA-prodrugs[179]. The pharmaceutical industry can consider designing novel APIs functional groups that can be chemically modified to achieve long-acting formulations by LA-prodrug strategies.

Highlights.

• Long-acting prodrugs can improve the compatibility of physicochemically diverse active pharmaceutical ingredients for a variety of drug delivery strategies.

• Extended duration of action for long-acting prodrug systems depends on both prodrug release and conversion rates to the active metabolite.

• Release of active metabolite from macromolecular prodrugs depends primarily on the rate of linker cleavage.