Alternative Routes of Administration for Therapeutic Antibodies—State of the Art

Background: For the past two decades, there has been a huge expansion in the development of therapeutic antibodies, with 6 to 10 novel entities approved each year. Around 70% of these Abs are delivered through IV injection, a mode of administration allowing rapid and systemic delivery of the drug. However, according to the evidence presented in the literature, beyond the reduction of invasiveness, a better efficacy can be achieved with local delivery. Consequently, efforts have been made toward the development of innovative methods of administration, and in the formulation and engineering of novel Abs to improve their therapeutic index. Objective: This review presents an overview of the routes of administration used to deliver Abs, different from the IV route, whether approved or in the clinical evaluation stage. We provide a description of the physical and biological fundamentals for each route of administration, highlighting their relevance with examples of clinically-relevant Abs, and discussing their strengths and limitations. Methods: We reviewed and analyzed the current literature, published as of the 1 April 2022 using MEDLINE and EMBASE databases, as well as the FDA and EMA websites. Ongoing trials were identified using clinicaltrials.gov. Publications and data were identified using a list of general keywords. Conclusions: Apart from the most commonly used IV route, topical delivery of Abs has shown clinical successes, improving drug bioavailability and efficacy while reducing side-effects. However, additional research is necessary to understand the consequences of biological barriers associated with local delivery for Ab partitioning, in order to optimize delivery methods and devices, and to adapt Ab formulation to local delivery. Novel modes of administration for Abs might in fine allow a better support to patients, especially in the context of chronic diseases, as well as a reduction of the treatment cost.


Introduction
Over the past 30 years, therapeutic antibodies (Abs) have been found to be valuable therapeutics [1]. A total of 6 to 12 new Abs are approved by the U.S. FDA and/or the EMA each year, and new molecules are reaching clinical trials every month [2]. Therapeutic antibodies are used in the treatment of numerous diseases, including infection, cancer, and autoimmune disorders, in which they have already demonstrated their efficacy [3,4].
The success of Abs is due to (I) a high level of specificity and affinity to their target antigen, (II) a favorable safety profile, and (III) a unique pharmacokinetic profile, supporting a longer half-life as compared to other drugs [5]. These characteristics have allowed Abs to move rapidly from pre-clinical studies to clinical trials, as observed during the COVID-19 pandemic [6].
From the historical full-length antibody, molecular engineering has enabled the development of multiple and diverse Ab formats, including multi-specific Abs, fragments, and conjugated Abs that are now extensively evaluated in clinical trials [7]. The routes of administration used to deliver an Ab affects its pharmacokinetics and pharmacodynamics, as well as exposing patients to potential harmful side effects. The systemic intravenous and subcutaneous routes are conventional routes for Abs, but novel local modes of administration are currently being investigated to improve the Ab therapeutic index. Optimal delivery of Abs may improve support for patients in numerous clinical situations. The figure was created with Biorender.com.  After IV injection, the second most popular route for the delivery of antibodies is the subcutaneous (SC) route. It consists in the injection of Abs using a syringe and needle under the skin of patients at an angle of 90 • C, thus bypassing the barrier formed by the epidermis and dermis layers [18]. The choice of the anatomical site is important due to differences in dermal thickness which may reduce the absorption of the injected Abs. Nowadays, around 30% of the approved Abs are delivered by SC injection (Table 1).

Routes of Administration for Therapeutic Antibodies
If the delivery of drugs, mainly opioids, by SC administration, has been in practice since the middle of the 19th century, the administration of Abs by this route is recent. The first subcutaneous injected Ab was Adalimumab, used in the treatment of rheumatoid arthritis and approved by the FDA in 2002, and by the EMA in 2003 [19]. After this first success, and particularly since 2009, the number of marketed Abs delivered by the SC route has significantly increased. It is noteworthy that SC administration is already the standard route in the treatment of chronic diseases such as rheumatoid arthritis. Indeed, it allows self-administration and improves patients' compliance. The SC route is mainly used for the delivery of Abs targeting interleukins such as TNF-α, critically involved in the development of rheumatoid arthritis (Adalimumab, Golimumab, Certolizumab pegol) or cytokine receptors such as the IL-17a receptor, involved in the progression of psoriasis (Brodalumab, Secukinumab, Ixekizumab) [20].
The development of Abs intended for a subcutaneous injection necessitates understanding the physiology of the skin. After injection, the drug reaches the hypodermis interstitial space between the dermis and the deep fascia covering the muscle tissue. This layer is composed of adipose tissue, blood, lymph vessels, and resident immune cells such as fibroblasts and macrophages. All components are enmeshed in an extracellular matrix (ECM) network, rich in collagen, elastin, and glycosaminoglycans [21]. To pass into the systemic compartment (via either the blood capillaries or lymphatic vessels), and thus reach their target, Abs have to diffuse through the ECM, which constitutes both a physical and chemical barrier. The fate of the Ab is dictated by its size, charge, and affinity with transporters. Despite the presence of the positively charged collagen fibrils, the hypodermis interstitial space displayed an overall negative charge due to important concentrations of hyaluronic acid and chondroitin sulfate, two major glycosaminoglycans of the hypodermis ECM, which are negatively charged. The global negative charge of ECM favors the transport of negatively charged drugs thanks to electrostatic repulsion [22]. However, the majority of therapeutic Abs are positively charged. Once the ECM is traversed, drugs may enter the systemic circulation by two different mechanisms. Molecules smaller than 16 kDa diffuse directly into the bloodstream, taking advantage of the permeability of the vascular endothelium [22]. However, Abs, along with drugs with a higher molecular weight, are absorbed by convection into lymphatic vessels. Thus, the subcutaneous route is particularly interesting to target lymphoid cells and the molecules they secrete. Abs in the lymphatic vessels pass to larger lymphatics and then reach the blood vascular system, from where they diffuse throughout the body.
If the development of Abs for subcutaneous injection is quite challenging, multiple factors explain the attractiveness of this route as compared to other parenteral ones. In the hypodermis, the walls in the fat lobule are thinner than those in the dermis, which facilitates the diffusion of drugs into blood capillaries [23]. Moreover, the absence of antigenpresenting cells in the hypodermis, usually present in the top layers of the skin (Langerhans cells and/or macrophages), may decrease the immunogenicity of the antibody. Thus, an increasing number of Abs delivered by SC are being developed, allowing a quicker delivery time of administration as compared to IV injection, enabling longer dosing intervals and, in fine, reducing the frequency of administration. In addition, SC administration is less invasive and painful than IV injection [22] and allows self-delivery at home [24]. Thus, the subcutaneous route may improve patient comfort and compliance, which is critical for the treatment of chronic diseases, and may be associated with a reduction in treatment costs, consuming fewer healthcare resources.

Abs Approved for Subcutaneous Delivery
Abs approved for subcutaneous administration must be formulated at a high concentration, thus necessitating a careful control of their stability and formulation viscosity. Different strategies have been considered to ensure the efficient absorption and bioavailability of Abs after hypodermis injection. They include, but are not limited to, the increase in delivered Abs concentration (e.g., SC administration limiting the injection volume to 1-2 mL [25]), the development of specific formulations to reduce physical and chemical destabilization (e.g., the use of polysorbate preventing aggregation and particle formation [26]) and the development of novel administration devices (e.g., the autoinjectors enabling a faster delivery for larger concentrations of Abs [27]). Those strategies have led to the approval of around 40 different Abs (Table 1).  2 The approval for Daclizumab was revoked, due to important side effects, by both the FDA and the EMA in 2018. 3 The first dose of Caplacizumab is delivered through IV while the subsequent ones are delivered through SC. Data were obtained from the list of approved medicines and drugs by both agencies, and from the prescribing information provided by the manufacturers [28,29]. A major concern for SC injection is the isoelectric point (pI) of Abs, found between 7 and 9, making Abs positively charged at the physiological pH. A study by Bumbaca Yadav et al., showed that positively charged Abs present a reduced bioavailability by 31%, while their negatively charged counterparts demonstrate enhanced bioavailability up to 70% after SC administration [30]. Another study found that the reduced bioavailability of Abs delivered subcutaneously is due to their interaction with ECM components, thus limiting the amount of Ab reaching the vascular compartment [31]. Moreover, the overall negative charge of the hypodermis interstitial space increases the interaction of ECM components with water molecules resulting in a low hydraulic conductivity and limiting the subcutaneous injection volume [32]. To circumvent this serious issue, one strategy consists in combining Abs with hyaluronidase. Hyaluronidase degrades hyaluronic acid, lowering the amount of negatively charged molecules and enhancing the bioavailability of Ab after SC injection [33]. Moreover, combining Abs with hyaluronidase may facilitate bulk fluid flow and improve the pharmacokinetic profile after SC injection [34], as demonstrated in multiple clinical studies [35][36][37]. Based on these results, the regulatory agencies approved Rituximab, Trastuzumab, and Daratumumab in combination with recombinant human hyaluronidase (rHuPH20), in 2017 (Rituxan Hycela/mAbThera s.c), 2019 (Herceptin Hylecta), and 2020 (Darzalex Faspro), respectively.
These encouraging results have fueled the repurposing of Abs approved for delivery by IV injection to this novel modality of administration. Notably, Tocilizumab ((Ro)-Actemra), an antibody used in the treatment of rheumatoid polyarthritis, was formulated for the SC route, in response to patient demand, and to allow a less invasive route for a treatment usually delivered monthly [38]. It is noteworthy that multiple studies have demonstrated the absence of significant differences between the IV and SC routes of administration for Abs, thus making SC a legitimate option for patients [37,38].

Abs in Clinical Development for the SC Route
The clinical development of subcutaneously-delivered Ab concerns either de novo development, expansion of the disease target, and/or new formulation for already approved Abs. A high number of those Abs are currently found in clinical trials. Here, we listed subcutaneously delivered Abs either in active phase 3 trials or under review by regulatory agencies (Table 2). 1 Those antibodies are already on the market, either delivered by another route, or for another application. Data were obtained from clinicaltrials.gov and EMA/FDA's website. Abbreviations: IL-Interleukin; CD-Cluster of Differentiation; COPD-Chronic obstructive pulmonary disease; PSCK-Proprotein convertase subtilisin/kexin; RANK-L-receptor activator of nuclear factor kappa-B ligand; TNF-Tumor Necrosis factor; BCMA-B-cell maturation antigen; TSLP-Thymic stromal lymphopoietin; HNGF-Human nerve growth factor; CCR-C-C chemokine receptor; FIX-Factor IX; FX-Factor X.
Novel developments include Fasinumab, a recombinant fully human IgG4, targeting the nerve growth factor (NGF) and evaluated for pain relief in patients suffering from osteoarthritis (OA). A phase 2b/3 trial showed that Fasinumab provides improvement in OA pain and motor function, even in patients that are non-responsive to analgesics [39]. The drug approval is pending an evaluation by the FDA (NCT03161093; NCT02683239). In the meantime, studies are also investigating lower doses of Fasinumab in patients with knee or hip OA.
The repurposing of IV delivery approved Abs for SC application in a disease context different than the original approval is also investigated. For example, Ofatumumab (Arzerra ® , Novartis) is a monoclonal antibody targeting CD20 and causing cytotoxicity in cells expressing CD20. It was first approved in 2010 for the treatment of certain chronic lymphocytic leukaemia by IV injection, and has been reformulated (Kesimpta) for SC administration and evaluated in patients with relapsed multiple sclerosis. Two ongoing phase 3 trials, OLIKOS (NCT04486716) and ARTIOS (NCT04353492) are evaluating the efficacy, safety, and tolerability of the SC drug in patients with relapsing multiple sclerosis, all of whom are transitioning from a CD20 Ab therapy (Rituximab or Ocrelizumab), or dimethyl fumarate therapy [40].
Many Abs have been developed or repurposed as emergency treatments since the beginning of the SARS-CoV2 pandemic, and target either the virus or the host inflammatory response. Among them, REGEN-COV2 comprising Casirivimab and Imdevimab has been approved for emergency use by IV infusion and is now undergoing regulatory review, for its use by SC administration to treat and prevent SARS-CoV-2 infection in non-hospitalized individuals [41]. The first phases of its clinical investigation showed a significant efficacy with improved survival. A phase 3 trial is also in progress to evaluate its potency for the prevention of COVID-19 in immunocompromised patients (NCT05074433).

Conclusion and Perspectives Regarding the Subcutaneous Route
The relevance of the SC route for the administration of Abs has been illustrated by several clinical successes. However, the bioavailability of Abs delivered subcutaneously remains difficult to predict. Advances in preclinical models would be necessary to investigate the fate of Abs at the SC injection site and their diffusion into the blood/lymphatic compartment. Interestingly, novel in vitro tools have been developed to predict the in vivo absorption of biopharmaceuticals after SC injection, by modeling the environmental changes an Ab will experience after its injection. Among those, Scissor (Subcutaneous Injection Site Simulator) device provides a tractable method to study the fate and the pharmacokinetics of biopharmaceuticals once in the hypodermis [42,43]. Nevertheless, as no in vitro model is yet accurate enough, the pharmacokinetics of Abs still relies on in vivo studies.
Formulating Abs for the SC route remains challenging, as formulations need to afford high concentration with low viscosity, aggregation and immunogenicity [22]. Biotechnological platforms have been developed to support the switch from intravenous infusion to SC delivery, using proprietary excipients and proteins, allowing the reduction of ionic strength and hydrophobicity areas of the molecule, thus limiting aggregation when the Ab is highly concentrated. If the intramuscular route (IM) is often used for the administration of vaccines, this modality is rarely considered for the delivery of Ab. The IM injection consists in administering a drug deep into the muscle mass, where the blood supply allows a rapid and better absorption than the subcutaneous route. As for the SC route, the drug is administered via a syringe and needle into the skin at an angle of 90 • . However, where the SC route uses small length-and diameter-needles, IM injection requires longer needles-2.5 to 4 cm-to bypass the different layers of the skin. The volume of administration is relatively low~2-5 mL, depending on the muscle chosen for the injection, necessitating a more concentrated product than for IV administration. The preferred injection sites are the deltoid muscle in the upper arm, the vastus lateralis found at the front of the thigh toward the outside of the leg, especially for IM injection in young children [44], the ventrogluteal muscle of the hip, the safest site for adults and children older than 7 months, and finally the dorsogluteal muscles of the buttocks, albeit less used nowadays due to potential injury to the sciatic nerve.

The Intramuscular
The pharmacokinetics of a drug injected by IM follows a specific sequence: dissolution rate, solvent supply, phase transfer, and diffusion to the vascular system. Therefore, injection depth is an important parameter which influences the absorption rate of the drug after IM injection. Consequently, a too superficial injection will deliver the drug either SC, or into the fat layer, retarding the action of the drug. The absorption will also depend on the muscle mass and its vascularization. Patients with muscular atrophy see a delay in drug absorption, as well as an increase in the risk of neurovascular complications. These anatomical parameters also influence the rapidity of action of the drug [44].

Ab Approved for Delivery by the IM Route
One of the most threatening pathogens responsible for lower respiratory tract infection in children is the respiratory syncytial virus (RSV). Nearly every infant develops an RSV infection during their childhood. A total of 60% of them have an immune system mature enough to control the infection [45]. However, between 15 and 40% of infants, especially preterm infants, develop a more serious airway infection, which may eventually lead to bronchiolitis and pneumonia. Although no vaccine or specific RSV treatment exist to treat the infection (apart from general anti-viral therapy), a passive immunization may limit the infection in young infants with high risk factors. In 1998, the FDA approved Palivizumab (Synagis) [46], a humanized monoclonal IgG1 antibody targeting the glycoprotein F on the RSV virus, which is responsible for membrane fusion and infection of the host cell. Intramuscular injection of palivizumab is recommended for premature infants and young children with heart or lung comorbidities, and injections usually start before the expected RSV epidemic season [47,48]. Intramuscular delivery was chosen to facilitate delivery and limit invasive risks to the young recipients of the treatment. Aggregated clinical data have established the half-life of Palivizumab administered via intramuscular injection to be around 17-27 days, with a mean of 20 days, in infants younger than 24 months. As such, a monthly administration during the epidemic season is necessary to protect effectively against RSV. Palivizumab prophylaxis has shown a significant decrease in the hospitalization, ICU stay, and mortality [49,50]. Unfortunately, its high cost limits the number of children who benefit from this prophylaxis protection, especially in low-income countries [51]. Although Palivizumab is not licensed for the treatment of RSV disease, it has paved the way to the clinical use of intramuscular injection, and has shown some potential to facilitate everyday care, promoting a less invasive route for patients, in particular infants, in whom IV injection is more complicated [52].

Abs in Clinical Development Delivered by the IM Route
The IM injection allows a more rapid absorption and onset of action compared to SC delivery: mathematical in silico models associate the IM route with higher drug concentration on target and shorter time to reach the peak of concentration as compared to SC delivery. This route benefits from the long experience of vaccine administration. As shown below, several anti-infective Abs delivered by IM administration are close to reaching patients (Table 3). In 2021, Sotrovimab (Vir Biotechnology-GlaxoSmithKline) obtained an emergency use authorization (EUA) from the FDA for the treatment of mild-to-moderate COVID-19. However, since May 2022, and due to its inefficacy against the BA.2 variant of SARS-CoV-2, its authorization has been revoked [53]. A phase 3 trial (COMET-TAIL, NCT04913675) which has established that the intramuscular injection of Sotrovimab had a similar clinical effectiveness as compared to an IV injection is still active. SYN023, a therapeutic cocktail comprising two monoclonal Abs, CTB011 and CTB012, is under development for post-exposure prophylaxis to rabies. Nowadays, the standard of care treatment after rabies infection is an on-site (wound injection) and IM administration of anti-rabies immunoglobulin (RIG), called post-exposure prophylaxis, followed by four doses of the vaccine, delivered by IM or intradermal routes. The full dose of RIG facilitates rapid protection until the immune system can produce its own immunoglobulins thanks to the vaccination. However, the current RIG products are limited and expensive, thus their use is limited in high-burden countries [54]. SYN023, administered by IM injection, targets protein residues found on human rabies virus, and displays a neutralization capacity equivalent or superior to RIGs, even at a lower dose [55]. The ongoing phase 3 trial will provide additional information regarding the efficacy of SYN023 in populations at risk of rabies infection.
Ibalizumab-uiyk (Trograzo) is an anti-HIV monoclonal antibody, approved in 2018, for the treatment of patients with resistant forms of AIDS by IV injection. It targets CD4 protein expressed by T cells, inhibiting the entry of HIV. As a "first class medication" (defined by the FDA), the use of Ibalizumab-uiyk needs to be expanded across the world and for patients reluctant to receive IV injection or where the access to medical facilities and personnel qualified to administer IV injections is limited. A phase 3 is currently evaluating the IM route in comparison to IV injection, notably in terms of safety, pharmacokinetics, and limiting the spread of infection (NCT03913195).

Conclusions and Perspectives on the Intramuscular Route
The IM injection is usually considered as a rescue route of administration when other ways to deliver drugs are not appropriate. For example, drugs inducing vein irritation or sensitive to oral digestion have been moved from IV and per os injection, respectively, to IM injection. Among other advantages, IM administration is associated with a quicker and more uniform absorption of the drug as compared to other routes, and it is considered to be as efficient and potent as IV injection, with less invasive characteristics. Even with these potential advantages over other routes of administration, the IM route is still underestimated for Abs. The development of depot injections may represent an opportunity for the expansion of the IM route. Depot injections are a slow-release form of medication usually implanted in the muscle. They are already used for neuroleptic drugs (fluphenazine), prolonging their pharmacological effect. Moreover, they allow a sustained delivery over time [56] which may be of particular interest regarding the treatment of chronic diseases. With the population aging, the prevalence of ophthalmic disorders has increased over the last 20 years and is estimated to reach several hundred million in the next decades [57]. The most prevalent age-related eye disease is macular degeneration (AMD), accounting for 196 million patients, in 2020. This continuously evolving disease affecting the retina leads to near blindness, due to the overexpression of vascular endothelial growth factor (VEGF), and has no curative treatment. During AMD, VEGF expression is abnormally increased promoting the development of microvessels, leading to micro-hemorrhage and the progression of the disease.
Drug delivery to the retina is challenging, and various routes of administration have been considered to circumvent the complexity of the eye anatomy and improve the therapeutic index of drugs. The eye comprises an anterior and a posterior part. The anterior part, closer to the external environment, composed by the cornea, the conjunctiva, the ciliary body, the aqueous humor, lens, and the lachrymal system, forms a static barrier which prevents the access of foreign particles to the eye. Topical administration on this side results in low bioavailability of the drug (less than 3%) inside the eye, and an even lower in the posterior part [58,59]. Penetration enhancers may be required to improve topical administration [60]. The posterior part encompasses the vitreous humor, retina, sclera, and choroid. The main barrier on this side is due to the cornea, a negatively charged tissue repulsing negatively charged drugs [61]. The intra-ocular space is also protected by two additional obstacles, the blood-aqueous, and the blood-retinal barriers [62], both composed of tight junctions which limit the transfer of drugs from the blood compartment to the eye, especially considering high molecular weight biotherapeutics. In this context, the local delivery of Ab, directly in the internal parts of the eye by intravitreal injection has been developed to provide a high amount of drug in the retina and the vitreous compartments.
The intravitreal administration is an invasive procedure consisting in the injection of the Ab, posterior to the limbus, thanks to an incision through the sclera at a specific angle of 30 • . The only approved syringe systems for intravitreal injection are the ones already prefilled with Ab. Once the needle is in position, the Ab solution (with a limited volume, 100-200 µL) is slowly applied into the vitreous cavity, avoiding damage to the retinal surface [63].

Approved Abs by Intravitreal Administration
Up to now, three Abs delivered by intravitreal injection have been approved for the treatment of AMD and equivalent ophthalmic diseases (Table 4). Data were obtained from EMA/FDA's website and from the prescribing information provided by the manufacturers [28,29]. Abbreviations: VEGF-A-Vascular Endothelial Growth Factor; AMD-Age-related Macular degeneration; ANG-2-Angiopoietin 2.
Ranibizumab (Lucentis), a humanized Fab fragment, [64] and Brolucizumab (Beovu), a humanized single-chain antibody fragment, both targeting all isoforms of VEGF-A [65] have been approved in 2006/2007 and 2019/2020, respectively. Recently, Faricimab (Vabysmo), a bispecific antibody targeting both VEGF-A and ANG-2 [66], has been approved by the FDA, for the treatment of AMD and diabetic macular edema. All these Abs aim at blocking the neo-vascularization, which is the main pathological process associated to retinal diseases [67]. It is noteworthy that Bevacizumab (Avastin), the first anti-VEGF-A antibody approved for cancer treatment, was also considered as an off-label indication for AMD by intravitreal injection, due to its more affordable price [68]. Unfortunately, it did not attain the required professional or political consensus to obtain final approval [69].
The pharmacokinetics of these Abs has been extensively studied. In aqueous humor, Bevacizumab and Ranibizumab were shown to have a half-life of around 10 days, and 7 days, respectively [70,71]. Abs concentration peaked on the first day of injection, and then rapidly declined. The short intravitreal half-lives implied frequent injections to ensure optimal effect, increasing the potential risks of side-effects. Interestingly, Faricimab demonstrated longer duration of action in a phase 3 clinical trial, with a sustained efficacy over 16 weeks post-injection [72].
To better understand the PK-PD of Abs after intravitreal administration, Mazer et al. developed a mechanistic model of intravitreal pharmacodynamics of anti-VEGF in the eye and demonstrated the interrelationship between the half-life of the Ranibizumab and the in vivo VEGF kD [73]. They also showed that the ocular Ab t1/2 was proportional to the hydrodynamic radius of the antibody, and the radius of the vitreous globe. Thus, the biodistribution and absorption of Ab depends on the physiological state of the eye, with a potential decrease in Abs efficacy for older eyes, or eyes impaired with either globe abnormalities or underlying pathologies (e.g., myopia, or hypermetropia) [74].

Abs in Clinical Development for Intravitreal/Intraocular Delivery
The global aging of the human population accounting for an increased prevalence of AMD, or retinal vein occlusion along with the effectiveness of intravitreal administration, has paved the way for further developments of intravitreally/intraocularly-delivered Abs products (Table 5). It is noteworthy that several ongoing trials are further evaluating currently approved Abs delivered by the intravitreal route (Table 4) to either identify their long-term effects in patients with AMD (NCT04777201) or evaluate their efficacy in other ophthalmic diseases (NCT04740905). Notably, Faricimab, after being recently approved for AMD is currently being investigated in macular edema and branch retinal occlusion.
KSI-301 is a humanized antibody against VEGF-A conjugated to a biopolymer, under investigation for the treatment of different ophthalmic disorders [75,76]. Apart from its excellent safety and better efficacy than the current treatment using the fusion protein Aflibercept, the phase 1 trial showed longer durability, allowing patients to achieve treatment-free intervals of 4 to 6 months. Few novel drugs delivered by intravitreal injection are currently being developed. However, biosimilars are currently under investigation. HLX04-O, a biosimilar of Bevacizumab, is a monoclonal antibody under development by Shanghai Henlius Biotech, targeting VEGF to treat AMD. Contrary to the original molecule, this biosimilar has been developed de novo for intraocular administration. Early-stage clinical trials, analyzing the safety and toxicity of the molecule delivered intravitreally, showed promising results, similar to the results obtained with Bevacizumab. As of April 2022, a Phase 3 trial was launched with the administration of HLX04-O to a first patient [77]. MW02, a recombinant anti-VEGF humanized monoclonal antibody, is under phase 2 and 3 evaluation. The trials compare its efficacy and safety to Ranibizumab (Lucentis) (NCT05297292).

Conclusion and Perspectives on the Intravitreal Route
While anatomic barriers prevent Abs access to the eye from the blood compartment, intravitreal injection has demonstrated efficacy and has become the standard-of-care in the treatment of eye disease. Although it is invasive, it is associated with limited sideeffects such as uveitis and vitreitis [78], and a higher therapeutic index. To support the development of intravitreally-delivered Abs, advances are necessary in formulation and medical devices, to allow prolonged action, longer periods between two injections, and better tolerability.
For example, the Port Delivery System (PDS) has been proposed for the delivery of Ranibizumab. It consists of a device surgically implanted into the vitreous cavity, allowing continuous delivery of the Ab. It dispenses the need for frequent intravitreal injections, thereby reducing invasiveness for patients. Its clinical evaluation is on-going and the first results for patients suffering from neovascular age-related macular degeneration are promising, showing equivalent control of the disease as compared to the standard care treatment, a monthly intravitreal injection of Ranibizumab (NCT04657289).
The repertoire of excipients recognized as safe by the intravitreal route is limited due to ocular toxicity. For instance, only polysorbate (80 and 20) and sugar residues have been included in the formulation of Brolucizumab, Faricimab, and Ranibizumab [79,80]. Similar to what is observed in SC administration, Abs diffuse poorly into the vitreous cavity due to their positive charge [81]. Multiple parameters may influence the diffusion of Abs inside the eye, including: (i) the age of the eye, which can be associated with a change in the viscosity of the vitreous humor; and (ii) the quality of the lens. Patients with cataract surgery might experience a more rapid clearance of Abs from the eye, as observed with intravitreal delivery of antibiotics [82].
Overall, intravitreal injection is quite challenging, and inter-individual variability will dictate the fate and the activity of the antibody inside the eye. Like other local routes of administration, intravitreal injection of Ab provides effective on-target treatment, while limiting systemic deleterious exposure, and thereby improving Ab therapeutic index.

Rationale of Delivering Abs through the Airways
Respiratory diseases are a major worldwide public health issue. They represent the fourth most common cause of death worldwide, which is mainly attributable to lung cancer, lung inflammatory diseases (e.g., chronic obstructive pulmonary disease (COPD)), and lower respiratory tract infections. The increase in antimicrobial resistance, seasonal virus outbreaks, newly emerging pathogens, or atmospheric pollution have severely complicated the management of airway diseases. Owing the success of Obiltoxaximab/Raxibacumab against pulmonary anthrax and Benralizumab against asthma, Abs have emerged as powerful therapeutics to tackle multiple respiratory diseases [83]. However, Abs administration through systemic routes led to a low bioavailability in the airway compartment [84,85]. This may limit the efficacy of Abs to treat respiratory diseases, if their target antigen primarily acts within the respiratory organ, while exposing the rest of the body to potential side effects [86].
Oral inhalation is the gold standard route of administration for small molecules commonly used for the treatment of inflammatory diseases (asthma, COPD). It allows direct drug access-as an aerosol-to both the upper and lower respiratory tract. It is associated with a rapid onset of action and a better therapeutic index as compared to parenteral injections [87,88]. Abs aerosolization is generally difficult since they are highly sensitive to mechanical, thermal, and physical stresses occurring during aerosol generation. Two types of devices have been evaluated for the oral inhalation of proteins: dry-powder inhalers, which deliver solid aerosol, and nebulizers, delivering liquid-formed aerosols. So far, metered-dose inhalers are not recommended for proteins due to the use of propellants [89].

Challenges Associated with the Clinical Development of Inhaled Abs
Researchers and pharmaceutical companies have been interested by local delivery of Abs through the airways for the last 20 years [90]. However, the first attempts were unsuccessful [91] and may be explained by multiple factors: − Abs instability during aerosolization/spraying, resulting in aggregation and thereby potential alteration of activity and immunogenicity concerns [86,92,93]. − Relevant target antigen, which must operate within the respiratory tract and be critical in the pathophysiology of the respiratory disease − Selection of the appropriate population, which may benefit from inhaled Abs. − Biological barriers, which may impair Abs PK and activity [94].
Several inhaled Abs are in early clinical trials, after preclinical studies showing promising efficacy in animal models (for example anti-IL-13 Fab for the treatment of asthma [95]). A few inhaled Abs are in phase 2 and 3 clinical evaluation, notably for the treatment of SARS-CoV-2 infection (Table 6). CSJ117 (Ecleralimab) is an inhaled TSLP inhibitory antibody fragment developed to treat moderate to severe asthma [96] and COPD. The antibody is provided as a powder in hard capsules to be delivered daily in the lungs using a dry powder inhaler. First results showed promising results, with CSJ117 having the ability to attenuate airways inflammation in asthma patients.
The recent SARS-CoV-2 pandemic has brought attention to the inhalation route, and several inhaled anti-SARS-CoV-2 Ab are being tested in the clinic by oral inhalation. Among the most advanced is a combination therapy developed to tackle the emerging SARS-CoV2 mutants, showing promising results. CT-P63 and CT-P59 (Regdanvimab) are monoclonal Abs, both targeting the receptor binding domain of the spike protein of SARS-CoV-2. More than just eliciting a neutralizing response, these Abs have the property to trap viral particles within the mucus, promoting their elimination by mucociliary clearance [97].
It is noteworthy that one molecule is currently under development by nasal inhalation, to address the Ab into the nasal cavity rather than in the lungs (COVI-DROPS).

Conclusion and Perspectives on the Inhalation Route
Although airway delivery of Abs has been demonstrated as feasible and promising in preclinical studies, the benefits of Abs inhalation have not materialized yet in the clinical setting. A major advantage of the inhalation route is on the possibility of self-administration of Abs treatment for non-hospitalized patients, thereby reducing healthcare costs, and increasing patient comfort [98]. However, groundwork is still necessary to optimize Abs design and formulation, inhalation devices, as well as improving our knowledge of the physical and biological barriers that may impair the therapeutic response to inhaled Abs. Abs used for the treatment of cancer are systematically delivered through IV injection. Their targets cover a wide range of proteins. They will promote the blocking of signals needed for cancer cell survival/growth, induce cancer cell destruction when coupled with toxins, or mark the tumor so that the immune system will better recognize and destroy it. Recently, the development of Abs targeting immune checkpoints has changed the treatment paradigm in many cancers. Abs targeting immune checkpoints inhibitors (ICI) block the immune "off" signal induced by the tumor cell in order to prevent the immune system from destroying the cancer, thus allowing a natural anti-cancerous immune response, and a higher survival for patients [99]. Systemic anti-cytotoxic T-lymphocyteassociated protein 4 (CTLA4) Ab was the first licensed ICI against advanced metastatic melanoma [100], and was then quickly followed by anti-programmed cell death ligand and protein 1 (PD-L1/PD-1) Abs. However, if ICI antibodies have shown efficacy in patients with melanoma or non-small cell lung cancers, many cancers remain refractory to them. In addition, if the systemic ICI treatment has undeniable advantages, such as predictable pharmacokinetics, it is associated with hazardous on-target/off-tumor side effects, including widespread inflammation. Thus, improving the tumor targeting of ICI Abs appears to be an evident goal.
In fact, most solid tumors are poorly reachable from the systemic circulation, due to multiple surrounding barriers [94]: − Densified tumor-associated ECM with the overexpression of collagens, and fibronectin, preventing the diffusion of Abs [101]. − Abnormal growth of blood vessels in the vicinity of the tumor resulting in nonvascularized, inaccessible areas for Abs coming from the blood circulation [102]. − Disorganized vessel structure associated with blood flow resistance, impeding the transport of the drug [103].
Consequently, several strategies have been proposed to limit the side effects of Abs targeting ICI, as well as optimizing their bioavailability within the tumor environment. A novel accurate method of injection consists in the delivery of Abs intratumorally (IT), allowing access to the tumor vicinity, circumventing extracellular barriers, and in theory improving on-target efficacy. Intra-tumoral injection is performed in a clinical environment by a physician, under the guidance of imagery (e.g., ultrasound, computed tomography (CT), or endoscopy) [104].

Abs in Clinical Development by Intra-Tumoral Injection
Preclinical studies have established that local administration of anti-CTLA-4 Abs was able to restore an anti-tumor response, eradicating both local and distant tumors, and used a significant lower amount of Abs as compared to IV Ab [105,106]. Thanks to the success of those pre-clinical studies, novel intratumorally-delivered Abs are under clinical evaluation (Table 7). Data were obtained from clinicaltrials.gov and the Embase online platform. Abbreviations: CD40-Cluster of differentiation 40; OX40-Tumor necrosis factor receptor superfamily, member 4; CTLA-4-cytotoxic T-lymphocyteassociated protein 4; PD/PDL1-programmed cell death ligand and protein 1.
In 2019, an anti-CD40 agonist antibody (ADC-1013) was developed and clinically evaluated in solid tumors after intra-tumoral administration (NCT02379741). The results from the phase 1 study showed a good safety profile and were associated with positive pharmacodynamic responses [107].
Ipilimumab is a human anti-CTLA4 antibody already approved for the treatment of metastatic melanoma and given by IV injection. A phase 1 study showed that Ipilimumab in combination with IL-2, given intratumorally, was well tolerated and was able to generate anti-tumoral responses in the majority of patients [108]. A phase 2 study is currently evaluating the tolerance of Ipilimumab IT and Nivolumab IV for the treatment of stage III/IV melanoma patients (NCT02857569). Moreover, a phase 2/3 trial is comparing Pembrolizumab and/or Ipilimumab in patients with advanced solid tumors after intratumoral injection and IV infusion. (NCT03755739).

Conclusion and Perspectives on the Intra-Tumoral Route
Intra-tumoral administration of anti-cancerous Abs has multiple advantages, especially in reducing the harmful side effects associated with on-target/off-tumor of Abs after IV injection. In addition, local administration, within the tumor environment, may promote a superior priming of T-cells already on-site, and a multi-clonal response [106].
For now, clinical validation of intra-tumoral delivery of Abs is still in the preliminary stages. Further work is needed to optimize the dosage and the formulation according to the type and the stage of the tumor lesion. In addition, this route of administration is operator-dependent, which means that a subtle change in needle position may change the outcome of the administration. The development of new live imaging tools may improve injection reproducibility. Combinatorial approaches with intra-tumoral and intravenous treatment, or tumor ablation, are also being investigated and may represent a promising therapeutic alternative [109]. Joints degeneration is one of the leading causes of permanent motor disability worldwide requiring long-term therapeutic treatment. Among them, osteoarthritis (caused by trauma to the joint cartilage), rheumatoid arthritis (autoimmune disease where immune cells attack the joints), and gout (joint inflammation due to excess of uric acid), are the most common forms [110]. Standard care includes systemic administration of analgesic and anti-inflammatory agents, including anti-TNF-α (Infliximab, Etanercept, Adalimumab) molecules and anti-IL1β Ab (Canakinumab) [111]. However, although symptoms show substantial improvements thanks to these treatments, non-responsive patients and side effects are increasing with time [112].
Efforts have been made in recent years toward the development of intra-articular (IA)delivered treatments, with multiple benefits such as a better bioavailability and a reduced systemic exposure [113]. Intra-articular injection is performed using syringe/needle into a joint. To help the guidance of the needle, ultrasound or fluoroscopy techniques can be used. The development of drugs dedicated to IA injection necessitates understanding the anatomy and physiology of the joints, to identify the key parameters influencing the pharmacokinetics and pharmacodynamics of the drug.
Synovial joints, or diarthrosis, joins bone endings and hyaline cartilage with a fibrous capsule delimitating the synovial cavity filled with synovial fluid. Synovial joints are particularly affected during the degeneration processes [114]. Cartilage is an avascular tissue composed of chondrocytes embedded in a negatively charged ECM which make this tissue impermeable to molecules bigger than 50 kDa (such as antibodies) depending upon their charge and conformation. Consequently, the cartilage is inefficiently targeted by drugs administered systemically, which first need to reach the synovial fluid, before diffusing through cartilaginous ECM. To enter the joints, the drug needs to pass through the capillary endothelium of the synovium, the ECM of the synovial intima, and the synoviocytes composing the synovial membrane. Both cellular layers are highly fenestrated, allowing high diffusion of molecules smaller than 10 kDA. For larger molecules, the fenestration will allow a size-dependent diffusion, slowing the passage of large molecules such as antibodies [115]. During inflammation, there is an increased permeability of both capillary and synovial membranes, allowing macromolecules to reach the synovial compartment. However, joint inflammation will also accelerate synovial clearance [116]; studies have revealed a mean clinical half-life of around 3 h for an anti-inflammatory antibody, not leaving enough time for an optimal action of the antibody.

Abs in Clinical Development for IA Administration
Despite some promising features, intra-articular delivery of therapeutic antibodies remains rare. Studies performed 20 years ago have shown conflicting results regarding IA delivery of Infliximab depending on the disease, with positive outcomes in patients suffering from ankylosing spondylitis [117], while there was no positive effect in patients with acute joint inflammation [118]. In 2007, the clinical evaluation of intra-articularlydelivered Infliximab in patients suffering from intractable knee monoarthritis, reached a phase 3 clinical trial but the development was stopped in 2015 due to the insufficient recruitment of patients [117] (NCT00521963). More recently, two novel intra-articularlydelivered Abs have entered clinical evaluation (Table 8). AMB-05X is a fully human antibody targeting colony-stimulating factor I (c-FMS), a protein overexpressed in many cancers, and on tumor-associated macrophages. This antibody entered a clinical trial phase 2 (NCT04731675), to treat patients suffering from tenosynovial giant cell tumor, and pigmented villonodular synovitis, two afflictions of the knee, after IA delivery. The ongoing study is investigating the safety, tolerability, and efficacy of the treatment delivered by IA injection.
Canakinumab, a human anti-IL1β antibody, is currently in phase 2 of clinical evaluation, in combination with LNA043 (a protein inducing chondrogenesis and cartilage repair) for the treatment of patients with knee osteoarthritis (NCT04814368) after IA delivery.

Conclusion and Perspectives Regarding the Intra-Articular Route
Intra-articular delivery may be of particular importance for the treatment of joint diseases (and notably rheumatoid arthritis) as around 30% of patients are resistant to biotherapies, and few novel molecules are under development. The limited half-life of antibodies reaching the inflamed joint is one of the major limitations associated with IA delivery, requiring frequent injections and thus increasing side effects, discomfort, and morbidity. Specific formulations including hydrogels [119], micro/nano particles [120], or in situ implants [121] are under consideration to improve antibody concentration on-target. A long-lasting Ab formulation through the use of biopolymers was recently developed for IA administration, showing a sustained high concentration of Abs in the synovial fluid, with minimal inflammatory side effects [122]. However, controlled-release methods are often associated with a loss of bioactivity of the drug, and/or inflammatory side effects that could accelerate the degeneration of the cartilage and the disease [123]. Additional work is needed to develop safe and active intra-articularly-delivered Abs to fight joints auto-immune diseases. The CNS is an essential part of the nervous system consisting of the brain and the spinal cord. The CNS integrates and coordinates essential functions of the body [124]. Because of its importance, the CNS is well protected, notably by the blood-brain barrier (BBB), limiting its exposure to exogenous particles carried by the blood circulatory system. The BBB is a highly selective semipermeable barrier preventing the passage of solutes from the blood into the extracellular fluid of the CNS. The BBB is composed of specialized endothelial cells sealed together with tight junctions reinforcing trans-endothelial electrical resistance, as well as the high expression of energy-dependent efflux transporter, inducing a selective passage of solutes [125]. In particular, FcRn, expressed in the microvascular endothelium and in the choroid plexus epithelium, essential components of the BBB, is involved in the reverse transcytosis of IgG, from the brain to the blood vessels. Therefore, it is estimated that less than 0.1% of systemically injected IgG enter the CNS through nonspecific pathways. Thus, to attain a therapeutic dose in the CNS, Abs have to be administered in high quantitywhich may be associated with toxicity-or specific transporter pathways existing between the circulatory system and the CNS have to be used [126,127].
In order to circumvent these barriers, novel methods have been considered aimed at addressing the drug into the CNS by surgery, either via the intracerebroventricular, intracerebral routes or convection-enhanced delivery [126,128]. The intracerebral injection, the most direct method, consists in intermittent bolus injections, which are administered locally in the brain, after a surgical intervention. The intracerebroventricular technique allows the injection of drugs directly into the cerebrospinal fluid in the cerebral ventricles. Both methods use a syringe and needle system filled with the drug to be administered. Convection-enhanced delivery consists in the generation of a pressure gradient at the tip of an infusion catheter or canula (usually implanted in the cerebral tumor), allowing the delivery of drugs through the interstitial spaces of the CNS.

Abs in Clinical Development for Direct Delivery to the CNS
There are many diseases targeting the CNS, including the neurodegenerative Parkinson or Alzheimer's disease (AD), for which the development of Abs has been considered.
For example, Aducanumab, an anti-Aβ Ab was approved by the FDA for the treatment of AD, after parenteral injection [129,130]. Cerebral delivery of Abs is mainly being evaluated in the preclinical phases with some success [131]; only one Ab has already reached clinical trials.
The antibody 131I-Omburtamab (Y-Abs Therapeutics), is a murine IgG1 recognizing CD276, used for radioimmunotherapy. This antibody is injected via the intracerebroventricular route, and phase 1 and 2 trials in patients with CNS Neuroblastoma, CNS metastases, or leptomeningeal metastases, have shown that the antibody was well tolerated and improved survival. A phase 3 trial is ongoing, evaluating the efficacy and safety of the Ab in children (NCT03275402). In the meantime, the same antibody is entering a phase 1 trial to test its efficacy, once delivered through a convention-enhanced delivery in patients with diffuse pontine gliomas profile (NCT05063357) [132].

Conclusion and Perspectives on CNS Delivery
The CNS is particularly well protected from the environment impeding an efficient access to therapeutic molecules. The BBB remains the major obstacle, when considering systemic infusion of therapeutic Abs, and limits CNS bioavailability. Novel methods bypassing the BBB, described above, are still in infancy and require further work to be standardized and to reduce their invasiveness. In this context, the investigation of the "nose-to-brain" route may be of particular interest. It consists of administering drugs in the olfactory region of the nose, by inhalation, from where they may transfer to the brain through the epithelial layer and via neuronal bundles that project to the olfactory bulb. The "nose-to-brain" route is under preclinical evaluation for Abs [133].

Future Perspectives
If the pharmacokinetic and pharmacodynamic properties of therapeutic antibodies depend on their format and their target, they also depend on their route of administration. In fact, depending on the location of the Ab's target, alternative modes of administration are being explored and developed to optimize Ab deposition in the vicinity of its target. The choice of route of administration is thus a critical factor for efficient Ab-based therapy and needs to be considered in the early stages of the development concomitantly with the appropriate formats, formulations, and devices (Table 9). If among the alternative routes some have progressed quite well to clinical trials, others are still at the very beginning of the research process.

Oral Delivery: Protects Abs from the Harsh Environment of the Intestinal Tract
Some Abs have been approved over the years for the treatment of intestinal diseases, such as inflammatory bowel disease (IBD). However, their delivery through parenteral administration has displayed several limitations, such as non-responding patients, or the increase of anti-drug antibodies limiting the effect of the Abs. Oral delivery has been used for centuries and is the most accessible and the least invasive route for drug administration. Specific issues associated with the gastro-intestinal (GI) tract have to be considered to ensure efficient drug activity. The intestinal mucosa, in contact with external environment, maintains a state of homeostasis that sustains tolerance and detects and eliminates exogenous materials or pathogens. Within each part of the gastrointestinal tract, specific cellular and extracellular barriers exist, such as the harsh pH conditions, which denature Abs, and the presence of multiple proteases that may degrade proteins [134]. Studies have shown that less than 20% of Abs administered orally were still immunologically active after proteolytic digestion by gastric enzymes [135]. Thus, oral administration of Abs necessitates their protection them from the harsh conditions of intestinal transit. Among the most advanced studies, AVX-470 is an orally delivered Ab targeting TNF, used to treat IBD. A phase 1 study concluded that the drug was safe and well-tolerated, and suggested a beneficial effect (NCT01759056) [136]. However, a high amount of AVX-470 was required because the majority of the antibody was degraded. This highlights the importance of protecting the antibody from the environment encountered in the GI tract.
One protection strategy may consist in encapsulating the Ab in either nanoparticles or enteric coated capsules, a strategy evaluated in a phase 1 trial for Foralumab-an anti-CD3 Ab-given to patients with active Crohn's disease [137] (NCT05028946). Another strategy consists in decreasing the gastric acidic pH by combining the Ab with a proton pump inhibitor. An anti-CD3 (OKT3) was combined with omeprazole in patients suffering moderate to severe ulcerative colitis [138]. The six patients who received both OKT3 and omeprazole orally showed a promising reduction in inflammatory genes expression in peripheral blood cells, associated with low side effects.
Although studies have shown consistently promising results, it is important to better understand the fate of Abs once ingested. Strategies to bypass Abs degradation and uphold their activity remain to be investigated, and may also include the rational selection of the best isotypes which could survive to the extreme gastro-intestinal environment [139].

Skin Administration: A Topical Delivery of Ab in Wounds
Abs offer interesting opportunities to treat local skin diseases or wounds. While Abs may be administered directly to open wounds and allow direct healing, the treatment of blisters or scars would require their administration on intact skin [140]. However, when applied topically, antibodies showed low skin penetration, which may be explained by the hydrophobic nature of the keratin-essential and main component of the skin-and the hydrophilic nature of antibodies, and their size, limiting cellular and transcellular diffusion [141]. Thus, a high dose of antibody should be required to achieve a therapeutic dose inside the tissue. Moreover, for topical administration of drugs, gels or creams are recommended, which imply complex formulations of Abs to ensure their stability. Several strategies using cell-penetrating peptides, physical penetration enhancer, or injection with microneedles have been tested, but none of them has entered clinical trials. Funding: This work was supported by a scholarship from the University of Tours and the French National Research Agency as part of the "Investissements d'Avenir" program (LabEx MABImprove, ANR-10-LABX-53-01), as well as the Region Centre-Val de Loire (Novantinh program).

Conflicts of Interest:
A.P. and T.S. have nothing to declare. N.H.-V. is the co-founder and scientific expert for Cynbiose Respiratory. In the past 3 years, she has received consultancy fees from Eli Lilly, Argenx, and Novartis.