A Holistic Review of the Preclinical Landscape for Long-Acting Anti-infective Drugs Using HIV as a Paradigm

Abstract

Lack of predictive preclinical models is a key contributor to the steep attrition rate in drug development. Successful clinical translation may be higher for new chemical entities or existing approved drugs reformulated for long-acting (LA) administration if preclinical studies designed to identify any new uncertainties are predictive of human exposure and response. In this review, we present an overview of standard preclinical assessments deployed for LA formulations and delivery systems, using human immunodeficiency virus LA therapeutics preclinical development as a paradigm. Key progress in the preclinical development of novel LA antiretrovirals formulations and delivery systems are summarized, including bispecific broadly neutralizing monoclonal antibody and small molecule technologies for codelivery of multiple drugs with disparate solubility properties. There are new opportunities to take advantage of recent developments in tissue engineering and 3-dimensional in vitro modeling to advance preclinical modeling of anti-infective activity, developmental and reproductive toxicity assessment, and to apply quantitative modeling and simulation strategies. These developments are likely to drive the progression of more LA anti-infective drugs and multipurpose technologies into clinical development in the coming years.

The progression of more long-acting antiretrovirals and multipurpose technologies into clinical development in the coming years will partly be driven by recent progress in the preclinical development of novel formulations and delivery systems that are reviewed in this article.

Decisions about progressing new drug candidates to clinical development are guided by evidence from preclinical studies. These are the cornerstone of new drug development and clinical translation, providing the critical data that inform first-in-human trials (see Figure 1). The landscape for long-acting (LA) anti-infective development over the past 12 months was marked by some significant developments, particularly for human immunodeficiency virus (HIV), which will be used as a paradigm to discuss the preclinical development of LA anti-infective drugs. LA therapeutics offer a transformational approach to HIV treatment and preexposure prophylaxis, largely circumventing established issues around treatment fatigue and associated nonadherence by offering regimens with extended dosing intervals.

Figure 1.

An overview of the stages in preclinical development of long-acting drug formulations, the studies conducted within each stage and opportunities to integrate new approaches. Abbreviations: ADME, absorption, distribution, metabolism, and excretion; PBPK, physiologically based pharmacokinetic model.

Multiple LA candidates for treatment of HIV are in development and are at different points in the drug development pipeline. Multicountry approval of LA-injectable cabotegravir plus rilpivirine administered intramuscularly every two months was a major milestone for LA HIV treatment, despite virological failure in patients with baseline nonnucleoside reverse transcriptase inhibitor or integrase strand transfer inhibitor resistance [1–3]. The US Food and Drug Administration (FDA) issued a Complete Response Letter with easily resolvable concerns about incompatibility with borosilicate glass vials in response to a New Drug Application for the LA lenacapavir (capsid inhibitor) administered every 6 months for treatment in heavily treatment-experienced patients with multidrug resistant HIV-1 infection [4]. The clinical hold has since been lifted by the FDA for injectable lenacapavir [5]. Investigational New Drug Applications for LA islatravir (nucleoside reverse transcriptase translocation inhibitor) monthly oral formulation and annual implant formulation have been placed on clinical hold due to observations of reductions in total lymphocyte and CD4+ T-cell count.

From the 12% of novel drug candidates that pass preclinical testing, only 11% successfully make it to the market, mostly because of failures resulting from unanticipated toxicity (about 50%) or suboptimal efficacy. About 50% of postapproval withdrawals are also from toxicity. Lack of predictive preclinical models is a key driver of this steep attrition rate, and it has been suggested that clinical trial success rate can be theoretically increased to > 55% with better preclinical models [6]. If preclinical studies designed to identify any new uncertainties are predictive of human exposure and response, translation of LA medicines should have a much higher success rate. However, additional considerations are evident for LA medicines and key knowledge gaps relating to drug absorption remain.

Here, we present a holistic review of the current preclinical development landscape for LA antiretrovirals, to provide a basis for expectations about novel products and technologies that may progress to clinical development in the coming years. First, we present an overview of standard preclinical assessment deployed for LA formulations and delivery systems. We also identify some opportunities to advance preclinical modeling of antiviral activity, developmental and reproductive toxicity assessment (DART), and the application of quantitative modeling and simulation.

PRECLINICAL ASSESSMENT OF LA DELIVERY TECHNOLOGY AND FORMULATION PROPERTIES

As summarized in Table 1, the strategies used in the preclinical characterization of LA formulations and delivery systems depend on their properties, the route and site of administration and the intended dosing interval. Adequate characterization of LA delivery systems and formulation allows assessment of how their properties affect quality, safety, and efficacy. Considering the increasing diversity of these systems, a risk-based framework that involves evaluation of potential risk factors during development and reduction of residual uncertainties throughout the product lifecycle is recommended by the FDA [7]. Based on the likelihood of clinically significant changes in exposure or response, there are 3 risk categories for LA formulations repurposed from an approved oral formulation: (1) low (eg, those with nanomaterials that revert to molecular constituents shortly after administration); (2) medium (eg, nontargeted nanomaterials with predictable active ingredient release characteristics); or (3) high (eg, targeted nanomaterials with complex and unpredictable active ingredient release characteristics).

Table 1.

Common Preclinical Studies Conducted During Long-Acting Formulation Development

Preclinical Studies	Application	Model/Conditions
Drug dissolution assay	Measurement of release rate from drug-eluting implants.	Samples incubated in media at 37°C on an orbital shaker. Can be dynamic and incorporate a continuous flow of media within the system at a set flow rate.
Drug release assay	Measurement of release rate from microarray needle patches.	Completed through insertion of the microarray needle patch onto porcine skin for 24 h or less.
Viral neutralization assay	Measurement of efficacy and potency of drugs.	Completed in pseudo-virus–infected 2D cell lines, most commonly TZM-bl or PBMCs.
Dose ranging study	Toxicology assessment: Completed to inform dosing, frequency, and route of administration of novel or reformulated therapeutics or modalities. As well as to assess toxicology and tolerability of dose and formulation.	Mice, rats, rabbits, NHPs
Toxicology bridging study	To assess the effect of certain changes on toxicity profile, including formulation, manufacturing process, test species selection, impurities, active metabolites found in significant amounts (usually ≥10% of total AUC) in humans, or impurities.	Mice, rats, rabbits, NHPs
Pharmacokinetics study	DMPK evaluation.	Mice, rats, rabbits, NHPs
PKPD study	Measurement of efficacy and potency.	Humanized mice, NHPs
Qualitative biodistribution study	Visualizing the nanoparticles by electron microscopy, and in particular transmission electron microscopy, or environmental scanning electron microscopy which allows direct imaging of samples from dosed animals.	Mice, rats
Quantitative biodistribution study	To quantify tissue nanoparticle distribution using radiolabeled experiments or inductively coupled plasma mass spectrometry, X-ray fluorescence, or neutron activation analysis.	Mice, rats

Abbreviations: 2D, 2 dimensional; AUC, area under the curve; DMPK, drug metabolism/pharmacokinetics; NHP, nonhuman primates; PBMC, peripheral blood mononuclear cell; PKPD, pharmacokinetics/pharmacodynamics.

Of particular importance are biodistribution and toxicology studies to bridge existing gaps in knowledge compared with already approved immediate release formulations, including where changes to excipients may alter their tissue distribution. The formulation properties and molecular mechanisms that influence in vivo release, biodistribution, and elimination after intramuscular and subcutaneous administration of LA therapeutics are not well understood [8]. Growing evidence indicates influence from a variety of distinct and interrelated formulation, delivery system and biological factors, including recruitment of immune cells, angiogenesis, and granuloma formation [9]. Furthermore, passive and active uptake of drug into surrounding vasculature or the lymphatics can alter in vivo drug release and subsequent pharmacokinetics [10]. Administration into interfascial spaces instead of intramuscular in animal models leads to leeching into the subcutaneous space along the needle track and can potentially result in misunderstanding of the pharmacokinetics [9, 11]. Several methods are currently used for preclinical qualitative and quantitative biodistribution studies (Table 1). Importantly, physiological differences (including drug transporter tissue expression) between human and preclinical animal models used in such studies can result in interspecies differences in drug distribution and pharmacokinetics [12]. The use of pharmacokinetic data from preclinical animal models to inform first-in-human studies is further complicated by differences in depot volume, shape, and size, which are known to impact drug release [13].

LA DRUG DELIVERY SYSTEMS AND ANTIRETROVIRALS IN PRECLINICAL DEVELOPMENT

An increasing number of LA drug delivery systems and formulations are currently in preclinical development as highlighted in our previous review [14]. Those published on the Long-acting Technologies Patents and Licenses Database of Medicines Patent Pool (https://lapal.medicinespatentpool.org/) in the past 12 months, are outlined in Table 2. Most of these technologies are being evaluated for combination antiretroviral therapy delivery, but they are disease agnostic, and many may be readily adaptable to multipurpose delivery systems. A selection of different antivirals in preclinical development and some essential features are summarized in Table 2. In this section, some of the key lessons from preclinical assessment of drug release are highlighted.

Table 2.

Examples of Novel Long-Acting Antiretroviral Formulations Currently in Preclinical Development

Drug Name (Class)	Application and Administration Route	Model/Species	Overview
BilA-SG (bNab)	Therapeutic antibody by intramuscular or intravenous injection	Chinese origin rhesus macaques	A bi-specific bNab made from PGT128, Hu5A8, and HuIG. Superior in vivo potency in comparison to 12 other bNabs in NHPs infected with SHIVSF162P3CN. Preexposure BiIA-SG injection prevents productive viral infection. Day 1 or 3 postviral challenge, single injection significantly reduces peak viremia, achieves undetectable setpoint viremia, and delays disease progression for years in treated animals [15].
Islatravir and etonogestrel (NRTTI/progestin)	Multipurpose technology subcutaneous biodegradable Implant	In vitro drug dissolution assay	Sustained release of islatravir and etonogestrel co-formulated within a single biodegradable implant for up to 12 mo [16].
	Multipurpose technology subcutaneous biodegradable implants	Wistar rats	Extruded polymer device enables long term islatravir delivery over 6 mo and etonogestrel for 12 mo. Islatravir-triphosphate detectable in PBMCs for 6 months [17]. Islatravir development is now on hold because of lower total lymphocyte and CD4+ T-cell counts being observed as a side effect of treatment in clinical trials.
Cabotegravir (INSTI)	Refillable nanofluidic delivery subcutaneous implant	Sprague-Dawley rats	2-Hydroxypropyl-β-cyclodextrin cabotegravir (β-CAB) loaded implant had 100 × higher drug solubility than cabotegravir alone, over 5-fold increase in release rate versus CAB-only implants. Plasma concentration was 4 × PBA IC90 for 45 days, and 2 × PBA IC90 for 91 days. Well tolerated with minimal inflammation [18].
Tenofovir alafenamide (NRTI)	Subcutaneous nanofluidic implant using electrostatic gating to modulate drug release	Indian rhesus macaques	Sustained subcutaneous delivery maintained tenofovir diphosphate concentration at 9 times above clinically protective levels over 4-month period, conferring >60% reduction in risk of infection per exposure through low-dose rectal SHIVSF162P3 challenge compared with the control [19].
Tenofovir disoproxil fumarate/Lamivudine/Dolutegravir (DcNP)	Subcutaneous injection	NHPs	Drug combination nanoparticle platform allows coformulation of lipophilic and hydrophilic molecules using multi-drug domain matrices. Allows greater distribution of components into PBMC and lymph nodes versus plasma (unpublished), demonstrated for multiple drug combinations in the GLAD project.
Dolutegravir, darunavir, ritonavir, MK-2048, atazanavir, rilpivirine (DcNP cART)	Ultra-long-acting biodegradable, removal in-situ forming subcutaneous or intramuscular implant	Humanized mice	Single injection of single or coformulated drugs (darunavir/ritonavir, atazanavir/ritonavir or dolutegravir/darunavir/ritonavir) demonstrated sustained plasma concentration over 90 d [20].
Cabotegravir/ rilpivirine (DcNP)	Microarray intradermal patches	Sprague-Dawley rats	Bilayer microarray patch with high drug loading (3 mg/0.5 cm2). Single application maintained plasma concentration similar to effective concentration in humans over 30 d. Well tolerated, no adverse events [21].

Abbreviations: BNabs, broadly neutralizing antibodies; cART, combination antiretroviral therapy; DcNP: drug combination nanoparticle; IC₉₀, 90% of the concentration that inhibits viral replication; INSTI; integrase strand transfer inhibitor; NHP, nonhuman primates; NNRTI, nonnucleoside reverse transcriptase inhibitor; NRTI; nucleoside reverse transcriptase inhibitor; NRTTI, nucleoside reverse transcriptase translocation inhibitor; PBMCs, peripheral blood mononuclear cells; SHIV, simian human immunodeficiency virus.

Adaptations of current in vitro methods for assessing release rate may be needed for application to novel delivery systems. For example, a refillable nanochannel delivery implant containing cabotegravir formulated with 2-hydroxypropyl-β-cyclodextrin, described further in Table 2, demonstrated cabotegravir plasma concentrations twice the protein adjusted 90% of the concentration that inhibits viral replication of 166 ng/mL for up to 91 days in Sprague-Dawley rats [18]. The 2-hydroxypropyl-β-cyclodextrin had 100 times higher dissolution than that seen for cabotegravir alone. In vitro, this translated to a 5.44-fold increase in release rate from 2-hydroxypropyl-β-cyclodextrin implants compared with implants containing cabotegravir alone. This lack of scalability from drug loading to release may in part be attributable to saturable sink conditions within the in vitro assay or within the nanochannels of the implant. As higher drug loading of implants becomes the norm, selection of conditions within dissolution experiments will need to be optimized to ensure maintenance of sink conditions and better mimicry of in vivo release. The absence of biology in most in vitro systems remains a major limitation. Ultimately, more complex in vitro models will be required to mimic the in vivo microenvironment with use case-relevant cellular and molecular processes.

Evidence from preclinical studies indicates potential differential release kinetics when drugs are coformulated compared with when they are administered as separate injections. For example, studies of multipurpose technology implants containing the antiretrovirals islatravir, or tenofovir alafenamide, in combination with the contraceptives etonogestrel or levonorgestrel showed differences in release rates of tenofovir alafenamide and etonogestrel in vitro when coformulated with either contraceptive or either antiretroviral, respectively [16], in comparison to their single drug formulations (Table 2). This highlights the importance of using the target drug combinations as early as possible in preclinical studies.

BROADLY NEUTRALIZING ANTI-HIV MONOCLONAL ANTIBODIES AS A PARADIGM FOR PRECLINICAL DEVELOPMENT

Combinations of anti-HIV broadly neutralizing monoclonal antibodies (bNabs) are a promising HIV prevention and treatment strategy that offer an alternative to small molecule LA antiretrovirals. They have the potential for dosing every six months or less frequent dosing, through the introduction of mutations, which extend antibody half-life [22, 23]. A wide range of singular, multispecific, and combinations of bNabs have entered clinical trials, and the available results have been robustly reviewed elsewhere [24, 25]. Additional information about the preclinical and clinical development of LA bNabs can be found in an accompanying article in this supplement of CID [26].

Preclinical investigations of bNabs have to date been conducted primarily using virus panels generated in cell lines, followed by in vivo studies in humanized mice and nonhuman primates (NHPs) [22, 27]. Existing data suggest that bNabs may not be as broad or as potent in humans as predicted in in vitro assays [24, 25]. This may be due to their activity being assessed most commonly in panels of HIV-1 Env-pseudotyped viruses generated in 293 T cells. Env-pseudotype viruses generated in 293 T cells are more sensitive to neutralization than other infectious molecular clones generated in 293 T cells or Env-pseudotype viruses generated in peripheral blood mononuclear cells, and this is consistent across antibodies targeting the same or different epitopes [27]. Furthermore, neutralizing assays are conducted in TZM-bl cells, which although highly reproducible, have been demonstrated to overestimate the breadth and potency of bNabs against diverse HIV strains, compared with studies in peripheral blood mononuclear cells [27–29]. This highlights the need for testing bNabs in multiple panels as part of in vitro screening, to ensure the most accurate assessment of neutralization activity of novel bNabs before candidates are selected for further investigations.

The role of antibody Fcγ receptor (FcγR) binding and activation of Fc effector functions in HIV neutralization has previously been completed in TZM-bl cells stably expressing FcγR [30]. However, in vitro neutralization assays do not provide a complete picture of antiviral activity because they do not include the full library of immune cells expressing FcγR and so are not able to fully recapitulate the complexity of FcγR-mediated pathways. Another difficulty relates to the species-specific nature of the Fc region binding in bNabs. Also, although humanized mice have human CD4 and CD8 T cells, their murine NK cells and macrophages limit the its translatability to human Fc pathways [31, 32]. NHPs have comparable distribution and expression of FcγRs to humans [33], but the FcγR in NHPs has been demonstrated to be genetically heterogenous [34]. Humanized mouse models exclusively expressing human FcγR allow direct assessment of antibody binding to human FcγR [35–38], and chimeric anti-HIV bNabs with Fc regions specific to mice have been generated, which enables the study of FcγR binding in bNab antiviral activity against HIV [35]. Human–mouse chimeric forms of 8 bNabs have been demonstrated to produce differential mouse FcγR engagement compared with their respective human counterpart, without impairing fragment antigen binding region-mediated recognition, antigen affinity, or specificity [35]. These models have shown their utility for studying bNab receptor interactions and the role of Fc effector functions in bNabs antiviral effect, but their use as part of bNab candidate screening has yet to be investigated. Nevertheless, when used in combination with other methodologies, they have the potential to provide key information about bNab pharmacological activity.

ADVANCING PRECLINICAL MODELS OF ANTIVIRAL ACTIVITY

Key preclinical animal models of HIV infection such as humanized mice and NHPs have different receptor and immune cell expression profiles compared with humans [31, 34]. Overall, there is a need to refine existing preclinical models and develop novel methods that better mimic conditions for HIV transmission and the tissues relevant to drug pharmacokinetics and pharmacodynamics. These methods will need to be tailored to the specific requirements of novel LA antiretrovirals formulations and delivery systems. Preclinical in vitro studies routinely use cell lines in static 2-dimensional (2D) culture systems for high-throughput screening of drug candidates [39]. However, static 2D culture systems are poor representations of the complex and dynamic in vivo microenvironments in humans. Also, genome-wide differences in gene expression profiles and transcription regulation of cell lines and their tissues of origin have been reported [40].

Generation of perfused 3-dimensional (3D) cocultures, which account for dynamic movement across epithelial and endothelial cell layers, would enable enhanced mimicry of the tissue microenvironment and therefore more accurate modeling of preclinical safety, absorption, distribution, metabolism, and excretion. A number of well-established methods for 3D cell culture have been adapted to enable vascularization including microfluidic devices [41], 3D organoids [42], and bioprinting [43]. Further work is required to overcome the established difficulties [44] of producing and maintaining complex 3D cell cocultures over extended study periods, which would be required for investigations focusing on LA therapeutics. Replication of complex biological systems within in vitro studies is challenging, especially when considering the need to recapitulate immune responses and the lymphatic system in the study of LA HIV therapeutics. The 3D cell models enable the study of cell–virus interactions in a way that is not achievable within 2D models, as demonstrated through the application of live cell imaging of macrophage and T-cell interactions within a 3D collagen matrix [45].

This method was used to study how HIV infection altered macrophage T-cell interactions and drug efficacy during dynamic cell–cell transmission of HIV [45]. This model was of utility because it enabled cell migration within the collagen matrix, better capturing the high motility of T cells that is seen in vivo within lymphoid tissues, compared with previous methods of layering cells on top of macrophages within static 2D models. This model has the potential to be adapted to study immune cell recruitment to LA administration sites and macrophage uptake of antiretrovirals from LA formulations, both of which are dynamic processes that require further study.

ROLE OF QUANTITATIVE MODELING AND SIMULATION FROM FORMULATION DEVELOPMENT TO FIRST-IN-HUMAN STUDY

The review by Shen et al provides a general overview of the main goals of preclinical testing to inform first-in-human trials, and the factors to consider in selecting an effective strategy [46]. For LA therapeutics, one of the key guiding principles in deciding what additional preclinical testing may be required is the understanding that equivalent active ingredient exposure in plasma does not always translate to equivalent response from potential differences in tissue distribution of nanoformulated drug (eg, by endocytic routes or via enhanced permeation retention effects) versus free drug. Hence, the overarching goal is to reduce residual uncertainties especially in medium to high-risk situations, which often requires exploration of the exposure-response profile. A review by Li et al provides an overview of how quantitative modeling and simulation can be used to streamline LA products development [47]. Physiologically based pharmacokinetic (PBPK) modeling now plays an increasingly important role in projecting first-in-human dose using input parameters from preclinical models [48]. Mayer et al published a pharmacokinetic/pharmacodynamic modeling approach to optimize combination ratios of bNAbs that integrates data on in vitro potency, in vivo pharmacokinetics, a correlate that can translate in vitro potency to in vivo efficacy and in vivo interactions of components [49]. The cross-cutting LA Therapeutics Development Program at the University of Liverpool involves extensive application of quantitative modeling and simulation across our entire preclinical activities, including through the LEAP Modelling and Simulation Core service [50]. During the lead generation stage, best-in-class candidates for LA formulation are selected based on predicted pharmacokinetic profiles using essential parameters from in vitro experiments and published literature. PBPK model estimates of doses required to maintain systemic drug concentration above the effective level and projected achievable dosing interval are used in acute pharmacokinetic/pharmacodynamic studies in preclinical species to generate in vivo data in the lead optimization stage. Details of these models are provided in an accompanying article in this supplement of CID [50]. These studies provide essential input parameters for PBPK/PD model predictions of optimal doses in higher species to further characterize the exposure-response relationship and obtain reliable dosage predictions for first-in-human clinical trials. Importantly, the reliability of these predictions depends not only on the quality of the input data, but also on the closeness of the preclinical model to the relevant aspects of human physiology. Hence, the use of available and relevant 3D in vitro models may lead to further improvements, reducing the reliance on animals in development and minimizing potential difficulties due to interspecies scaling.

MINDING THE GAPS IN DART ASSESSMENT OF LAI CANDIDATES

Another key consideration during preclinical development of LA formulations is identification of existing gaps in DART knowledge for new drug candidates and repurposed drugs with limited or no prior information. In preclinical drug testing, DART liability is assessed in 3 different types of studies: fertility and early embryonic development, embryo foetal development, and pre- and postnatal development. Currently used preclinical models, typically 1 rodent and 1 nonrodent such as rabbit [51], have historically proved inadequate in their representation of human physiology. The unresolved uncertainties often lead to significant time lag in the availability of pregnancy-specific safety information and widespread off-label drug use. Human relevant preclinical models of DART are needed to inform decisions about the inclusion of pregnant women in the early stages of drug development trials. In recent years, there is a growing regulatory interest in novel technologies to improve predictivity of preclinical studies and replace, reduce, and refine reliance on animal testing [52]. For example, a recent review highlights key developments in organ-on-a-chip technology for the study of different components of the female reproductive system [53]. Organ-on-a-chip models have been shown to recapitulate human physiology and disease states with higher fidelity than other in vitro models or animal studies [54]. As this and similar technologies progress, it is crucial to qualify them for use cases that are relevant to DART to facilitate an early insight into potential toxicity liability of LA antiretrovirals, and to support go/no-go decision-making about inclusion of pregnant women when these products progress into clinical trials.

ADDITIONAL CONSIDERATIONS

The novel LA antiretrovirals currently in preclinical development that are highlighted in this review have the potential to provide huge benefits for patients if they successfully progress into clinical development and market authorization. Hence, it is crucial to adapt and apply novel preclinical in vitro models to increase the human relevance of findings and resolve key uncertainties about the clinical significance of differences in exposure and tissue distribution between LA and oral formulations of already approved drugs. A better understanding of the mechanisms that alter the release kinetics and pharmacodynamics of LA drug candidates for different formulations and administration routes remains a major gap for current preclinical models.

Consideration must be given to how best to coformulate and study combination LA therapies to minimize the risk of drug resistance with LA monotherapy particularly where 1 molecule has a longer pharmacokinetic tail than the other(s). Importantly, the emergence of resistance in the LA cabotegravir and rilpivirine preexposure prophylaxis clinical trials was mainly from low or unquantifiable drug levels in certain participants, leading to infection, exposure to subtherapeutic drug concentrations, and subsequent viral resistance [55, 56]. Hence, factors that may result in significant interindividual variability in tail-phase pharmacokinetics of LA formulations need to be studied during development.

To ensure that new LA candidates have supporting information to guide their use across different populations, it is recommended that existing knowledge gaps are identified during preclinical development. For instance, resolving key uncertainties about DART liabilities early in LA formulation development using novel human preclinical models, or, where there is substantial human pregnancy exposure data from off-label use or pregnancy registry, will ensure pregnant women are not left behind in clinical development. Furthermore, practical limitations that may constitute barriers to implementation should be considered early in development. For example, LA antiretrovirals that can be self-administered would be hugely beneficial in widening access to LA HIV therapies. Last, novel methods that will facilitate experiments of extended dosing intervals and novel technologies for enhanced drug loading in LA formulations will be important new developments that will advance this field in the coming years.

Notes

Acknowledgments. This supplement was sponsored by the Long-Acting/Extended Release Antiretroviral Research Resource Program (LEAP).

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health grant for the Long-Acting/Extended Release Antiretroviral Resource Program (LEAP), www.longactinghiv.org—(R24 AI 118397). A. Owen also acknowledges funding from Unitaid for project LONGEVITY, Wellcome Trust (222489/Z/21/Z), EPSRC (EP/R024804/1; EP/S012265/1), and NIH (R01AI134091).

Supplement sponsorship. This article appears as part of the supplement “Long-Acting and Extended-Release Formulations for the Treatment and Prevention of Infectious Diseases,” sponsored by the Long-Acting/Extended Release Antiretroviral Research Resource Program (LEAP).