Synthetic oxygen carriers as substitutes of red blood cells for blood transfusion

Abstract

The search for an efficacious blood substitute has remained a persistent challenge for more than a century. Despite extensive efforts to develop synthetic oxygen carriers, no broadly approved alternative to human blood has yet been realized. In this review, we first provide an overview of the historical development and repeated failures of artificial blood substitutes, while highlighting the continuing clinical demand for transfusion alternatives. We then discuss recent setbacks in blood substitute development, particularly adverse events observed in human trials. Current research efforts with goals to improve biocompatibility are further examined, with a particular focus on oxygen-carrying micro-/nano-particles, emerging fabrication technologies, and unmet clinical needs. Hemoglobin-based cellular and acellular oxygen carriers, as well as novel non-hemoglobin-based systems, are reviewed in detail with emphasis on technical innovations and biological advances. Finally, future development in the field was briefly envisioned to shed light on the rational design and versatile applications of safe and efficient oxygen carriers.

This review summarizes the design, oxygen-delivery mechanisms, and translational challenges of artificial blood substitutes, including perfluorocarbons, hemoglobin-based carriers, and emerging oxygen nanocarriers.

1. Introduction

The clinical need for transfusion-enabled oxygen delivery has driven the search for artificial blood substitutes for more than a century. The discovery of blood circulation by William Harvey and the later identification of human blood antigen types by Karl Landsteiner laid the physiological and immunological foundations of modern transfusion medicine.^1,2 However, the need for blood type compatibility, together with the limited shelf-life and blood availability, continues to motivate the development of cell-free or cell-mimetic oxygen carriers. As red blood cells (RBCs) mediate the oxygen transport through hemoglobin (Hb), a tetrameric heme protein whose Fe²⁺-porphyrin centers reversibly bind oxygen by means of coordination chemistry, this review focuses primarily on RBC transfusion and on strategies designed to temporarily replace the oxygen-carrying function of RBCs.^3,4

The 2023 National Blood Collection and Utilization Survey (NBCUS) in the United States (U.S.) reported the continued stabilization in the blood supply, consistent with the 2019 and 2021 data, but the gap between blood collection and transfusions keeps widening.⁵ In fact, the stock and consumption of human blood have become both ends difficult to meet. Especially, the negative impact of an emerging pandemic on blood collection and supply was significant, as a systematic review reported 8.3% to 95.4% reduction in blood donation rate amid and immediately after H1N1 influenza or COVID-19 pandemic.⁶ Moreover, given the ageing society in many countries like U.S., which means a declining population of eligible blood donors and an increasing group of potential blood needers, it necessitates a rich storage of transfusable blood.⁷

The availability of transfusable blood after collection has been hampered by several obstacles for further medical applications.⁸ Firstly, the rising concerns towards the possible contamination of harmful pathogens in the collected blood, along with the need to crossmatch the blood type between the donor and the recipient, require stringent and expensive screening steps before clinical uses. Secondly, storage and usage of donated blood are susceptible to the short shelf-life. The aged blood is associated with the decreased deformability and impaired oxygen delivery, lessening the circulatory time, and may also release a variety of cytokines to affect the recipient's immune system.^9,10 Thirdly, blood administration per se is never risk-free. Varying from patient to patient, the possible transfusion reactions could cause allergic response, serious shock or even death, owing to the physical transfer of donor blood containing exogenous cells, proteins and nucleic acids, etc., to the recipient.¹¹

These constraints define the real-world deficiency of human blood and urgent demand for artificial oxygen carriers. Therefore, an effective RBC substitute must solve more than one supply problem: it must reproduce key chemical functions of RBCs, including high but reversible oxygen binding, controlled oxygen release in hypoxic tissues, suppression of Hb oxidation to methemoglobin (metHb), limited nitric oxide scavenging, and compatibility with the vascular endothelium and immune system.^12–14 Compared with previous reviews that primarily summarize clinical development or biomedical applications of blood substitutes, this review focuses on how the physicochemical properties of artificial oxygen carriers, such as molecular structure, ligand coordination, redox stability, interfacial chemistry, polymer architecture, particle size, surface charge and biological corona formation, could determine their oxygen transport, circulation parameter, toxicity profile and translational potential. By linking historical failures to underlying biological and chemical mechanisms, we aim to explore design principles for next-generation oxygen carriers.

2. Failure in the early development of synthetic blood substitutes

Artificial blood substitutes can be of great benefit by overcoming the drawbacks of human blood and providing rapid oxygen delivery in emergency or perioperative settings. Insofar the most extensively studied are perfluorocarbon (PFC) emulsions and hemoglobin-based oxygen carriers (HBOCs).^8,15 Their early development demonstrated that oxygen transport can be achieved through either physical dissolution of oxygen in fluorocarbon phases or reversible coordination of oxygen to Hb.

2.1. Perfluorocarbon (PFC) emulsions

PFCs are chemically inert fluorinated hydrocarbons that dissolve large quantities of oxygen and carbon dioxide through physical solubility rather than covalent or coordination binding (Fig. 1A).¹⁶ At ambient conditions the solubility of O₂ in most PFCs is >40% (v/v). This oxygen content follows Henry's law and is therefore proportional to the oxygen partial pressure (pO₂), in contrast to the sigmoidal oxygen-binding curve of Hb (Fig. 1B).^17,18 Thus, PFCs can be highly oxygenated under hyperoxic conditions. Their oxygen delivery depends on dissolved gas gradients and requires stable emulsification because PFCs are hydrophobic and immiscible with plasma.¹⁹ Differing from the human blood or blood derivatives, PFCs are thermally and chemically stable, economically produced, and easy to sterilize. Simultaneously, they can be storable for years and especially suitable for transfusion recipients on religious grounds that decline human or animal blood derivatives. Nevertheless, the therapeutical performance of PFCs is governed by their volume fraction, droplet size, surfactant composition, interfacial tension, colloidal stability and immune clearance by the mononuclear phagocyte system (MPS).²⁰

Fig. 1. Representative perfluorocarbons (PFCs) and acellular hemoglobin-based oxygen carriers (HBOCs). (A) Commonly used PFC emulsions for oxygen transport. (B) Comparative oxygen dissociation curves of red blood cells (RBCs), representative acellular HBOCs, and PFC emulsions. Hemospan, HemAssist, and Hemolink represent conjugated, cross-linked, and polymerized hemoglobins (Hbs), respectively. (C) Potential adverse effects associated with PFCs and acellular HBOCs, including macrophage activation, cytokine release, and nitric oxide (NO) scavenging-related vasoactivity. (D) Representative structures and size comparison of RBCs, acellular HBOCs, and Hbs. Panels B and D are adapted from ref. 133 with permission from Elsevier, copyright 2014. Other panels are created with Adobe illustrator 2019.

The first-generation PFC product Fluosol-DA, developed by Green Cross Corporation in Japan, was a mixture of perfluorodecalin and perfluorotripropylamine (7 : 3 by weight ratio) emulsified by Pluronic F-68 and egg-yolk phospholipids.²¹ The emulsion contained particles of ∼100–200 nm on average with a broad size distribution, due to the instability of emulsion and possible fusion between particles. Fluosol-DA exhibited no major acute or chronic toxicity in vivo and was approved by US Food and Drug Administration (FDA) for oxygenation of human tissue during angioplasty in 1989 but withdrawn from the market in 1994 due to the minimal effect in transporting oxygen. Fluosol-DA possessed only 7.2% (v/v) oxygen-carrying capacity, less than half of RBCs' (17–20% v/v), and a short intravascular half-life (4–6 h) (Table 1).²² From a chemical viewpoint, this failure revealed the trade-off between emulsification and oxygen loading: increasing hydrophilic stabilization improves dispersion but reduces the fraction of oxygen-dissolved PFC in each droplet.²³

Table 1. Summary of PFC products and acellular HBOC products^a.

Type	Product names	Chemical design	Oxygen-transport mechanism	Major limitations	Clinical status	Ref.
PFC emulsions	Fluosol-DA; Perftoran; Oxyfluor; Oxygent	Fluorocarbon droplets stabilized by surfactants or lipids	Physical dissolution of O2 within PFC droplets, with O2 loading and release determined primarily by the local pO2 gradient	Low O2-carrying capacity at clinically feasible doses, emulsion instability, MPS uptake, and inflammatory reactions	Clinical development has been limited or discontinued for most products	22–28
Cross-linked Hb	HemAssist	Intramolecular covalent cross-linking to stabilize Hb tetramers	Reversible coordination of O2 to heme Fe2+, with cross-linking altering Hb stability and O2 affinity	NO scavenging, vasoconstriction, hypertension, and increased mortality	Clinical development discontinued	44 and 45
Polymerized Hb	PolyHeme; Hemopure/Oxyglobin; OxyVita	Intermolecular polymerization of Hb to increase molecular size and reduce extravasation and renal filtration	Reversible binding of O2 to heme Fe2+ within polymerized Hb molecules, with polymerization modifying circulation and O2-release properties	Product heterogeneity, hypertension, oxidative stress, and potential myocardial injury	PolyHeme development was discontinued; Hemopure has limited human approval in South Africa; Oxyglobin has been approved for veterinary use in Europe and the USA; OxyVita remains investigational	50–53, 57 and 58
Conjugated Hb	Hemospan; Sanguinate	Surface conjugation of Hb with PEG chains to increase hydrodynamic size	Reversible binding of O2 to heme Fe2+, with polymer conjugation modifying Hb circulation, ligand accessibility, and O2 affinity	MetHb accumulation, linker instability, altered O2 affinity, oxidative toxicity, and vasoactivity	Products have shown mixed clinical progress, with no broad clinical approval	61, 62 and 65–69

^aAbbreviations: Fe²⁺, ferrous iron; Hb, hemoglobin; MetHb, methemoglobin; MPS, mononuclear phagocyte system; NO, nitric oxide; O₂, oxygen; PEG, polyethylene glycol; PFC, perfluorocarbon; pO₂, partial pressure of oxygen; P₅₀: partial pressure of oxygen required for 50% saturation of hemoglobin oxygen binding sites.

Being the only authorized product as artificial oxygen carriers in the world, Russia-developed Perftoran™ was approved for clinical use in several countries including Russia and Mexico, although adverse effects were occasionally reported, such as pulmonary complications.²⁴ The emulsion was constituted by two components of PFCs, i.e., perfluoromethylcyclopiperidine and perfluorodecalin (1 : 2 by weight ratio), stabilized by non-ionic Proxanol 268 (analogous to Pluronic F-68), having particles of 30–150 nm in diameter (averaged particle size <100 nm). The cyclic structures of both PFCs were attributed to the stable emulsion with small particle suspensions, leading to the rapid elimination from the body and the reduced toxicity risks. However, it contains only 10% (v/v) PFC with limited oxygen carrying capacity of 6.9 ml dL⁻¹ under PaO₂ of 760 mmHg.²⁵ Currently, this product was rebranded as Vidaphor by an U.S. company for manufacturing and marketing in the North America, with FDA approval pending.²⁶

With the help of advanced emulsion techniques such as high-pressure homogenization and microfluidics, the second-generation PFC products were developed with an enhanced capacity of delivering oxygen and prolonging the circulation.²⁷ Oxyfluor™, invented by HemaGen corporation, and Oxygent™ (also known as perflubron), produced by Alliance Pharmaceutical Corporation, both US-based manufacturers, resemble in their compositions, where they contain 50–60% (w/v) linear PFC with halogen-substituted C₈ chains and egg-yolk phospholipid as emulsifier. With less electronegativity, chloride or bromide replacement of fluoride in PFC molecules decreases O₂ affinity, but substantially improves their lipophilicity when emulsified, further promoting the encapsulation of O₂-enriched PFC in the products. Both products owned uniform particle size within a narrow range (220–250 nm and 160–180 nm for Oxyfluor™ and Oxygent™, respectively), and shared many features in common, such as viscosity, stability and O₂ loading capacity.²⁷ However, due to severe side effects associated with effective dosage, medical developments of Oxyfluor™ and Oxygent™ were ceased after Phase I and Phase III clinical trials, respectively.²⁸

As thermodynamically unstable systems, emulsions are mixtures of two or more immiscible liquids stabilized by surfactants and interfacial energy input for well dispersion.²⁹ They tended to be heterogenized back to their segregated oil and water forms over time as the interfacial tension unavoidably increased. Optimization of PFC emulsion through gradual changes in temperatures and pressures may enhance the stability of storage for better clinical use.³⁰ Once administered via blood circulation, PFC emulsion droplets are immediately recognized by MPS, a group of phagocytes that originate in the bone marrow and linger in the reticular connective tissues (RCTs) of various organs. Largely depending on the size, PFC droplets with > ∼100 nm in dimension could be intercepted by MPS, followed by accumulation of particles in RCTs or direct hepatic or renal clearance. On the contrary, PFC particles with MPS-passable sizes flow through the blood stream to reach deoxygenated sites, releasing O₂, whereas the vaporized PFCs are finally exhaled through the lungs.³¹ In this biological course, the physicochemical properties of PFC emulsions determine their stability, such as the chemical composition, the type of emulsifier, and the size of emulsion droplets, inducing different biological effects on their O₂ delivery mission.^20,32 The main adverse effects, including febrile reactions, thrombocytopenia and leukocytosis, are linked to particle phagocytosis, complement activation and cytokine release (Fig. 1C).³³ Future PFC designs therefore require simultaneous optimization of droplet size, interfacial robustness, oxygen solubility, PFC loading and biological deposition/clearance.

2.2. Acellular hemoglobin-based oxygen carriers (HBOCs)

Hb in native form is a tetrameric metalloprotein that comprises two identical pairs of α and β polypeptide chains. Assembled through hydrogen bonds in a nearly tetrahedral geometry, each subunit contains an Fe²⁺-containing heme group, responsible for reversible coordination with O₂.³⁴ When encountering O₂, Hb undergoes a conformational change that favors further O₂ binding, and this cooperative binding mechanism can be depicted in a characteristic sigmoidal-shaped oxygen-Hb dissociation curve, which allows the steep release of O₂ when pO₂ drops in a narrow physiological range (from 40 to 100 mmHg) (Fig. 1B).³⁵ Hbs transport the inhaled O₂ from the lung to different organs through blood flow and collect the metabolized CO₂ back to the lung for exhalation.

Hbs in HBOCs come from human, bovine, or recombinant sources.³⁶ Cell-free Hb does not express RBC surface antigens and therefore obviates blood-type matching. Its smaller hydrodynamic size than that of RBC may permit oxygen delivery through microcirculation to arterioles, capillaries and venules.³⁷ However, severe toxicity has been associated with the infusion of cell-free Hb. For instance, Hbs can be easily dissociated into dimers or monomers, being eliminated by kidney and inducing toxicity by blocking the renal tubules.³⁸ Moreover, the lack of a chemical effector in stroma-free Hbs increases their O₂ affinity, making Hbs prone to be oxidized and the bound O₂ difficult to be unbound.^39,40 In RBCs, 2,3-diphosphoglycerate (2,3-DPG) as a product of glycolysis in cytosol binds preferentially to deoxygenated Hbs in tissues and therefore lowers the O₂ affinity for oxygenated Hbs.⁴¹ Its absence in free Hbs leads to an abnormally high O₂ affinity, thereby reducing the efficiency of O₂ unloading. Hence, to minimize the toxicity while preserving the O₂-delivery capacity, chemical modification of acellular Hbs became indispensable.⁴²

Three types of acellular HBOCs have been developed; namely, cross-linked Hbs, polymerized Hbs, and surface-modified Hbs (Fig. 1D). Firstly, cross-linked Hbs were prepared by connecting polypeptide chains of Hbs through covalent bonds using a linker molecule, to prevent the breakdown of Hbs into dimers and to avoid the ensuing nephrotoxicity.⁴³ However, HemAssist™ (Baxter Healthcare, USA), a diaspirin cross-linked Hb produced using bis-(3,5-dibromosalicyl) fumarate as an α–α chain cross-linker at lysine residues, was discontinued after increased mortality in Phase III trials.⁴⁴ Mechanistically, small acellular Hb species can extravasate across the endothelial barrier and rapidly scavenge nitric oxide (NO), thereby impairing vasodilation, increasing systemic and pulmonary vascular resistance, and compromising tissue perfusion (Fig. 1C).⁴⁵ As a result, vasoconstriction has been commonly observed in acellular Hb-based O₂ therapeutics, giving rise to a heightened systemic blood pressure but a reduced blood flow in local organs. The resulting shortage in cardiac output outweighs the benefit of O₂ delivery in generating more severe life-threatening problems.

Secondly, polymerized Hbs were synthesized by intermolecular cross-linking of tetrameric Hbs. Molecular glues, such as glutaraldehyde, reacted with amino acids in Hbs and polymerized a different number of Hb tetramers with various sizes.⁴⁶ Human serum albumin can be enveloped to evade the MPS clearance, whereas incorporation of catalase as antioxidant within the core may effectively suppress Hb autoxidation.^47,48 A typical product containing polymerized Hbs is PolyHeme™, developed by Northfield Laboratories Inc., USA, was approved by FDA for Phase III clinical trials as RBC substitutes infused for patients with traumatic blood loss.⁴⁹ Initially, this product showed successful treatment with no further need for regular RBC infusion, where vasoconstriction was not noticed, possibly due to increased size of polymeric Hbs which are unable to penetrate the interstitial space to scavenge NO.⁵⁰ Nevertheless, FDA denied the Biological License Application (BLA) of this product due to several medical complications including the higher rate of myocardial infarction and death after administration compared to conventional RBC infusions.⁵¹

Another glutaraldehyde-polymerized bovine-based Hbs, Hemopure™ (produced by Biopure, Boston) was approved for clinical use in South Africa and Phase III clinical trials in USA. This product may improve local and systemic oxygenation for effective resuscitation in various ischemic and hypoxic clinical scenarios, such as traumatic hemorrhagic shock, cardiopulmonary bypass, and organ transplantation.⁵² It can be stored at room temperature up to 36 months, and a sister product developed by the same company (Oxyglobin™) has been approved for veterinary use in the U.S. and European Union.⁵³ However, vasoconstriction and increased blood pressure were still of great concern and remained the most common incidence of adverse effects for such a product of polymerized Hbs. One strategy to reduce vasoconstriction and hypertension is to increase the molecular size of Hb through polymerization, so limiting extravasation and NO scavenging while maintaining oxygen-delivery capacity.⁵⁴ Polymerized human Hb (PolyhHb) represents a new generation of synthetic oxygen carrier with a larger molecular diameter than the bovine counterpart, preventing extravasation and limiting adverse reactions such as cardiac toxicity in preclinical models.^55,56 Other efforts include the development of OxyVita derived from bovine Hbs via the zero-link polymerization.⁵⁷ The manufacture utilizes chemical activators (i.e., water-soluble carbodiimide) instead of linking agents to form a superpolymer of Hb tetramers with a size of ∼17 MDa, greatly enhancing the oxygen affinity and reducing the Hb oxidation.⁵⁸ Such hyperpolymerized Hbs with abundant Hb content, low colloid osmotic pressure and high viscosity can effectively diminish tissue extravasation, avoiding the subsequent NO scavenging and the resultant rise in mean arterial pressure,⁵⁹ as well as reducing renal tubular necrosis.⁶⁰

Thirdly, free Hbs can be modified on their surfaces with large molecules (e.g., polymers) to alter their physicochemical and biological properties for a favored intravascular O₂ delivery. For instance, Hemospan™ (Sangart Inc., California) was produced by conjugation of human Hbs with six-armed maleimide-polyethylene glycol (PEG, m.w. = 5 kDa), a well-known stealth surface group that significantly prolongs the circulation by reducing immune recognition and organ uptake.⁶¹ With increased molecular size, this PEG-modified Hbs exhibited an increased viscosity, osmotic pressure and oxygen affinity with no adverse vasoconstriction and hypotension reported, and were enrolling Phase III clinical trials.⁶² However, like many other HBOCs that were finally discontinued, PEG-modified Hbs lacked an ability to adjust redox potentials in themselves due to the absence of reductases in acellular milieu, therefore failing in converting Fe³⁺-contained metHb back to Hb.⁵³ The accumulated metHbs cannot bind O₂, so inactivating HBOCs to lose their O₂ delivery capacity. It was reported that 10% accumulation of metHbs would significantly impair O₂ transport while >70% could be fatal.⁶³ Furthermore, under aerobic conditions PEGylation may induce structural changes, typified by tetramer dissociation, so losing O₂ binding cooperativity, while the same PEGylation under anaerobic conditions yields a low-affinity HBOC that obeys cooperativity of O₂ binding.⁶⁴ At the same time, compared to the low-affinity one, the high-affinity HBOC enabled greater tissue oxygenation but also induced higher oxidative stress in blood plasma and tissues like heart and kidney.⁶⁴

Later, Sanguinate was developed to equip carboxylated bovine Hbs with 8–10 PEG-arms of 5 kDa (total m.w. = 109–120 kDa). Sanguinate has an average P₅₀ (defined as the value of PO₂ at which Hb is 50% saturated with O₂) of 7–16 mmHg between those of normal erythrocytes (∼27 mmHg) and ischemic tissues (<5 mmHg).⁶⁵ This unique Hb–O₂ binding affinity allows it to bypass obstructions in the microcirculation and favors O₂ transfer to hypoxic tissues.^66–68 Despite adverse effects like transient hypertension, lethargy, and troponin elevation, Sanguinate that entered clinical Phase I study demonstrated clinical improvements in patient mortality and secondary efficacy measures such as oxygenation levels, respiratory support, and vasopressor requirements.⁶⁹ Recent findings indicated that maleimide-PEG products were prone to deconjugation or cross-conjugation to thiol species, thus causing structural instability and the ensuing vasodilation.⁷⁰ In comparison, the mono-sulfone-PEG Hb products were more stable, suitable as acute oxygen therapeutic with prolonged vascular retention (weeks-months) and enhanced vasoactivity.⁷⁰

Overall, decades of HBOC research demonstrate that the principal obstacles are chemical as much as clinical. Short circulation half-life (usually <12 h), vasoconstriction, hypertension, oxidative stress, metHb accumulation, renal toxicity, and myocardial injury can be traced to physicochemical and physiochemical characters of acellular Hb, such as particle size, ligand accessibility, heme redox chemistry, NO reactivity and interactions with vascular and immune systems.⁷¹ Acellular HBOCs may still be lifesaving when donor RBCs are unavailable, but the high risk-to-benefit ratio has shifted the field toward encapsulated, biomimetic and redox-stabilized systems that better reproduce the protective environment of RBCs.

3. Ongoing search for blood substitutes

3.1. The persistent and increasing needs for an artificial blood substitute

The growing demand for blood transfusion has intensified the need for safe and effective RBC substitutes. Global population aging is expected to substantially increase transfusion requirements, particularly because elderly patients already account for a large proportion of transfusions related to surgical care, while the pool of eligible blood donors continues to decline.^72,73 Blood shortages may become even more critical during public health emergencies, natural disasters, and disease outbreaks. In addition, stored RBCs can contain bioactive plasma components and storage-related products that may trigger adverse inflammatory responses, potentially contributing to cancer progression or acute organ injury.⁷⁴ Therefore, despite recent setbacks, the pursuit of clinically viable blood substitutes remains an urgent and unceasing goal.

Given the previous failures of earlier blood substitutes, current research gradually views artificial oxygen carriers as integrated chemical systems rather than passive oxygen reservoirs. Beyond oxygen loading and release, an ideal blood substitute should be inexpensive and scalable to manufacture, carry minimal risk of pathogen transmission, remain stable during long-term storage, and require no donor-recipient compatibility testing. After administration, it should avoid allergic or hemolytic transfusion reactions, provide effective oxygen delivery with appropriate intravascular retention time, and minimize systemic or local toxicities, including oxidative injury, immunogenicity, nephrotoxicity, pulmonary hypertension, and hematological disturbances.

To meet these requirements, new-generation oxygen carriers are being designed to overcome the limitations of PFC emulsions and acellular HBOCs. Major strategies include encapsulating Hb within organic or inorganic shells to prevent direct vascular exposure; engineering surface chemistry to regulate protein adsorption, complement activation, and MPS clearance; tuning particle size, charge, and morphology to balance circulation time with biodegradation and elimination; co-loading antioxidants, reductases, or nanozymes to maintain Hb in its functional ferrous state; and using coordination chemistry or self-assembly approaches to stabilize Hb at high loading while preserving reversible gas binding. These advances suggest that future blood substitutes will depend not only on oxygen-carrying capacity, but also on rational chemical architecture that integrates biocompatibility, redox control, vascular safety, and manufacturability (Table 2).

Table 2. Comparison of emerging oxygen-carrier platforms with different chemical and structural designs^a.

Platform	Key design principle	Physicochemical advantages	Main limitations	Proposed improvements	Translational stage	Ref.
PEG-LEH/HbV	Encapsulation of concentrated Hb within PEGylated lipid vesicles	Separates Hb from the vascular endothelium; reduces renal filtration and extravasation; permits co-encapsulation of antioxidants and allosteric effectors	MPS uptake, complement or immune responses, lipid oxidation, and membrane/interfacial instability	Optimize PEG density, lipid composition, particle size, membrane stability, and redox cofactors	Early clinical translation: extensive preclinical evaluation; HbV has entered first-in-human safety evaluation	83–86
PEH/polymersome	Encapsulation of Hb within self-assembled amphiphilic polymer vesicles	Greater mechanical and membrane stability than conventional liposomes; tunable shell thickness, permeability, surface chemistry, and degradation	Potentially restricted O2 permeability, incomplete biodegradation, polymer-related toxicity, and limited in vivo validation	Tune polymer block length, shell thickness, crystallinity, gas permeability, and biodegradation	Preclinical proof-of-concept: primarily physicochemical, in vitro, and limited animal evaluation	106, 107 and 114–118
RBC-mimetic core–shell particles	Replication of RBC size, shape, surface properties, and deformability using synthetic or biomimetic materials	Potentially improves rheology, capillary passage, circulation, and microvascular O2 delivery	Complex fabrication, uncertain long-term stability and biodegradation, batch variability, and limited reproducibility at scale	Standardize particle size, elasticity, morphology, Hb loading, shear resistance, and scalable manufacturing	Early preclinical stage: laboratory-scale fabrication and initial in vitro or animal proof-of-concept	119 and 121–123
MOF-Hb carriers	Encapsulation or biomimetic mineralization of Hb within porous coordination frameworks	High Hb-loading capacity; protection against denaturation and oxidation; tunable porosity and surface functionalization	Potential metal ion or linker release, uncertain framework degradation, limited gas transfer characterization, and unclear long-term clearance	Develop biocompatible and biodegradable frameworks; optimize pore structure, surface PEGylation, and clearance profiles	Early preclinical stage: mainly material characterization, in vitro studies, and small-animal evaluation	126–128
Hb-free O2 micro-/nano-bubbles	Encapsulation of gaseous O2 within a transient lipid, polymeric, or protein shell	Avoids Hb-associated oxidation and heme toxicity; enables rapid and potentially image-guided O2 delivery	Short O2-delivery duration, bubble instability, embolic or vascular obstruction risk, and limited control over systemic dosing	Develop rapidly dissolving and size-controlled shells; improve circulation safety, controlled gas release, and CO2-management strategies	Exploratory preclinical stage: predominantly in vitro and localized animal applications; systemic transfusion use remains unestablished	130–132

^aAbbreviations: Hb, hemoglobin; HbV, hemoglobin vesicle; LEH, liposome-encapsulated hemoglobin; MOF, metal–organic framework; MPS, mononuclear phagocyte system; O₂, oxygen; PEG, polyethylene glycol; PEH, polymersome-encapsulated hemoglobin; RBC, red blood cell.

3.2. Liposome-encapsulated HBOCs

A new class of nano-/micro-scale Hb-encapsulated particles has been developed for acute hemorrhage or short-term restoration of blood oxygen-carrying capacity. Unlike acellular HBOCs, these cell-like HBOCs mimic natural RBCs by enclosing oxygenated Hb within protective core–shell structures, typically composed of aqueous Hb cores surrounded by semi-permeable phospholipid or polymer shells.⁷⁵ This design preserves Hb oxygen affinity while allowing modulation of key physicochemical properties, including stability, viscosity, and oncotic pressure. The shell permits oxygen diffusion but prevents direct Hb exposure to tissues, thereby reducing Hb-associated toxicity; it can also be surface-engineered to evade MPS clearance, prolong circulation, or/and enhance tissue targeting.⁷⁵ In addition, cofactors such as 2,3-DPG, antioxidants, reductases, and catalases can be co-encapsulated to better recapitulate RBC function, suppress metHb formation, and improve oxygen unloading.

To this end, liposome-encapsulated Hbs (LEHs) were prepared using bovine Hb, glutathione as an antioxidant, and a lipophilic ^99mTc compound for radiolabeling, followed by infusion into rabbits to assess biodistribution. No Hb leakage was detected after 20 h post-infusion, and LEHs showed a circulation half-life of ∼18 h, with organ uptake ranked as blood > spleen > liver > lungs > muscle > urine and minimal accumulation in the kidney, heart, and brain. These findings indicated that LEHs were cleared mainly by MPS, particularly in the spleen and liver, rather than by renal filtration.⁷⁶ Although modifying liposome membranes, such as using saturated phospholipids or adding cholesterol, can improve particle stability, it does not substantially prolong circulation, with typical half-lives remaining around 16–20 h.⁷⁷ Moreover, unmodified LEHs tend to aggregate and fuse during storage, limiting their functionality and clinical development.

To overcome these limitations, surface functionalization has been used to improve colloidal stability and circulatory persistence of LEHs, with PEGylation being the most widely adopted and effective strategy.⁷⁸ PEG is a biologically inert FDA-approved polymer commonly used to coat therapeutic particles, enhancing circulation time, rheological behavior, hemodynamic stability, and blood compatibility.^79,80 In PEG-LEHs, phospholipid-conjugated PEG is incorporated into the liposome bilayer, increasing surface hydrophilicity, reducing particle aggregation, minimizing nonspecific serum binding, and limiting MPS uptake.⁸¹ In the rabbit model using ^99mTc-labeled PEG-LEHs, particle accumulation after 48 h infusion followed the order blood > liver > bone > muscle > spleen > kidney ≈ lung, with only trace levels in the heart and brain.⁸² This altered biodistribution reduced splenic burden and markedly prolonged the circulation half-life to 65.2 h, more than three times that of unmodified LEHs under similar physiological conditions.

In 2022, cell-like PEGylated LEHs, termed Hb vesicles (HbVs), were approved for a Phase I clinical trial in Japan (Fig. 2).⁸³ With a mean diameter of 225–285 nm, HbVs mimicked erythrocyte-like architecture while avoiding free Hb-associated toxicities, such as platelet activation and neutrophil extracellular trap (NET) formation.⁸⁴ In humans, HbVs show an ∼8 h circulation half-life, with longer retention times observed at higher doses.⁸³ Optimized kneading-based production has yielded thermostable HbV suspensions with minimal particle size change, Hb leakage, or denaturation during two years of storage.⁸⁵ Owing to these properties, HbVs hold potential not only as transfusion alternatives but also as oxygen therapeutics and organ perfusates.⁸⁴ More recently, scalable production using a high-pressure cell disruptor generated PEG-LEH nanoparticles with well-characterized gas (O₂, CO, NO) binding/release kinetics, long-term storage stability, and comparable Hb concentration to that of whole blood (∼12 g dL⁻¹), further supporting their potential as lifesaving oxygen therapeutics.⁸⁶

Fig. 2. Preparation and representative applications of Hb vesicles (HbVs). (A) Schematic illustration of HbV production from nucleic acid amplification test (NAT)-inspected red blood cell (RBC) concentrates. (B) Glass vials containing deoxygenated HbV (deoxy-HbV), carbon monoxide-bound HbV (CO-HbV), and methemoglobin vesicles (met-HbV), showing color differences according to the state of encapsulated Hbs. (C) Representative functional forms of HbVs: deoxy-HbV for O₂ transport, CO-HbV for carbon monoxide (CO) carrying, and met-HbV for detoxification applications. Panel A is reproduced from ref. 83 with permission from Elsevier, copyright 2022. Panels B and C are adapted from ref. 134 with permission from Frontiers, copyright 2022.

The steric shielding effect of PEG arises mainly from a dense and flexible polymer layer on the liposome surface that acts as a “molecular shield”, reducing protein adsorption and opsonization. This effect, and consequently the intravascular retention of PEG-LEHs, can be tuned by adjusting key physicochemical parameters, including particle size and shape, and molecular weight, branching, grafting density, and surface charge of PEG (Fig. 3).⁸¹ Among these factors, particle size is a primary determinant of MPS recognition, as the larger foreign particles are more readily engulfed by phagocytes. However, simply reducing PEG-LEH size does not necessarily prolong circulation, because smaller particles carry less Hb and may show increased interactions with hepatocytes, renal cells, or splenic cells, thereby enhancing tissue uptake and potential toxicity. Empirically, PEG-LEHs of approximately 200 nm have been shown to provide an optimal balance between Hb loading and blood retention.⁸⁷

Fig. 3. Physicochemical determinants of PEGylated liposome-encapsulated Hb (LEH) nanoparticles. The circulation time and intravascular persistence of PEGylated LEH (PEG-LEH) nanoparticles are influenced by particle size, PEG molecular weight (MW), PEG side-chain density, surface charge, and particle shape. The panel illustrating the effect of particle shape is adapted from ref. 135 with permission from Wiley, copyright 2022. Other panels are created with BioRender.

The effect of molecular weight of PEGs on the pharmacokinetics of PEG-modified liposomes has been long evaluated.^88,89 In general, the particle circulation lasts longer as the molecular weight of surface-modified PEGs becomes larger, although the circulatory half-life may always reach a plateau with still growing sizes of PEGs. In mice, the circulation time of a phospholipid-PEG in a typical phospholipid/cholesterol unilamellar liposome was found proportional to the molecular weight of PEG as PEG₁₂₀₀₀ = PEG₅₀₀₀ > PEG₂₀₀₀> PEG₁₀₀₀ (with all sizes = 200 nm).^90,91 Additionally, employment of PEG < 5 kDa or insufficient PEG density for LEH modification may not effectively prevent complement activation in vivo.^92,93 When moving along the blood vessels, larger PEGs with the increasing hydrophilicity in PEG-LEHs may lessen more protein opsonization or particle aggregation, and travel more slowly to extravascular tissues with less tissue clearance from blood, therefore extending their blood residence. With the same molecular weight, the increasing number of PEG side chains may decrease the particle size and enlarge the effective surface of PEG shield, further prolonging the circulation.⁹⁴ Experimental findings derived that liposomes modified by 10% PEG with chain length of 5 kDa have an optimal anti-opsonization effect without formation of curved micelles, resulting in a long circulation half-life.⁹⁵

Surface charge on PEG-LEHs plays a vital role in determining their biological fate, influencing immune recognition, tissue uptake, and systemic clearance. In a pilot study on evaluating this effect, anionic and neutral PEG-LEHs were prepared by incorporating phosphatidylethanolamines (PE)-linked PEG into the Hb-encapsulated liposome bilayer, with or without additional anionic lipids such as dimyristoyl phosphatidylglycerol (DMPG).⁹⁶ Both particles showed comparable size distribution (∼130–200 nm), O₂ affinity and metHb level, with significantly longer circulation time than that of their non-PEGylated counterpart in rabbits. The circulation half-lives were 19.3 and 16.5 h for neutral and negatively charged PEG-LEHs, respectively, suggesting that neutral surface charge may favor intravascular persistence.⁹⁶

After entering the bloodstream, PEG-LEHs interact with phagocytes through a multistep process involving serum protein adsorption, attachment to the phagocyte membrane, endocytic internalization, lysosomal degradation, and eventual exocytosis of residual fragments. Among these steps, phagocyte adsorption is considered rate-limiting.⁹⁷ Particles with high positive or negative zeta potentials tend to bind more plasma proteins, thereby enhancing phagocyte adsorption and uptake, whereas near-neutral particles reduce nonspecific protein binding and phagocytosis, prolonging blood circulation.⁹⁷ This charge-dependent behavior of PEG-LEHs is consistent with observations from other PEGylated nanoparticles.^95,98 Although native RBCs are negatively charged and long-circulating, PEG-LEHs appear to benefit from surface neutrality, likely because their smaller size and exogenous nature make them more readily recognized by MPS.

A novel approach to increase the circulation half-life of PEG-LEHs is to incorporate actin, a cytoskeletal scaffolding protein forming microfilament, into the hydrophilic core of liposome.⁹⁹ Compared to spherical PEG-LEHs, ellipsoidal actin-containing PEG-LEHs owned similar sizes (both in a range of 110–150 nm in diameter) but longer in vivo circulatory half-life (10.5 h versus ∼8 h) after infusion into Sprague-Dawley rats. This improvement occurred without compromising Hb encapsulation capacity, O₂ binding affinity or metHb maintenance.¹⁰⁰ Thus, this disk-like morphology rendered actin-containing PEG-LEHs greater mechanical stability and deformational flexibility when flowing in circulation and microcirculation, resembling crucial biophysical features of natural RBCs and offering advantage over conventional spherical PEG-LEHs.

3.3. Polymer-encapsulated HBOCs

In addition to LEHs, polymer-encapsulated Hbs (PEHs) represent another major class of cell-like HBOCs and have attracted increasing interest because of their low cost, diverse polymer sources, tunable structures, and potential for prolonged circulation. Unlike polymer-conjugated Hbs, PEHs are typically designed as core–shell particles in which amphiphilic polymers encapsulate high Hb payloads within hydrophilic cores through supramolecular self-assembly¹⁰¹ or double emulsion (water/oil/water type).¹⁰² The synthetic process and polymer matrix properties, including micellar geometry, curvature, and molecular weight, strongly influence Hb loading capacity. In an early study, Chang and colleagues used biodegradable polymers such as polylactic acid (PLA) and poly-isobutyl-cyanoacrylate (PIBCA) to encapsulate bovine Hb, producing PEH particles with mean diameters of 80–120 nm, polymeric membrane thicknesses of 5–15 nm, Hb contents of 30–45% by weight, and P₅₀ values comparable to free Hb.¹⁰³ Although these polymers could be biodegraded into carbohydrates, the resulting PEHs were rapidly cleared from rats after 2 h post-infusion.

Replacing PLA with PLA-PEG copolymer membranes increased the circulation time of PEHs, with 80–150 nm particles persisting in blood for up to 12 h.¹⁰⁴ Further optimization within the PLA-PEG system, including the use of higher-molecular-weight PLA, increased PLA-PEG concentration, encapsulation of chemically cross-linked Hb instead of free Hb, and glutaraldehyde cross-linking of PEH particles, extended the circulation half-life to 24.2 h in rats (equivalent to t_1/2 = 41.5 h in humans).¹⁰⁵ Adjusting PEG content in PLA-PEG copolymers also improved performance. PEHs containing 5–20 wt% PEG formed 100–200 nm particles with bovine Hb encapsulation efficiencies of 80–88%, showed longer circulation and less liver accumulation compared with PLA-based PEHs, and achieved an optimal half-life of 34.3 h in mice at 10 wt% PEG.¹⁰⁶ In addition, incorporating ionic surfactants to minimize the absolute surface charge of PLA-PEG PEHs further prolonged blood circulation.¹⁰⁷

Among polymers used for PEH oxygen carriers, amphiphilic diblock copolymers are especially attractive as they can self-assemble into Hb-loaded polymersomes with nano to submicron sizes (50–300 nm).^108,109 Typical formulations combine a hydrophilic, biocompatible polyethylene oxide (PEO) block with a hydrophobic, biodegradable poly(ε-caprolactone) (PCL) or PLA block, and are commonly obtained by hydration of PEO-b-PCL or PEO-b-PLA thin films within human or bovine Hb in plasma-simulated buffer (Fig. 4A).¹¹⁰ Compared with liposomes (Fig. 4B), polymersomes possess thicker membranes, greater mechanical strength, lower water permeability, and improved structural stability.^111,112 PCL-based membranes degrade by ester hydrolysis in physiological conditions (Fig. 4C), offer favorable permeability to small biomolecules, and erode more slowly than PLA or PGA without substantially altering local pH.¹¹³ The PEO block forms a fully PEGylated surface, enhancing colloidal stability, bioavailability, and blood circulation compared with partially PEGylated LEHs. Hb-loaded polymersomes therefore can integrate efficient Hb loading, RBC-like gas binding and release, uniform size, blood-like viscosity and oncotic properties, and scalable production and storage, making them promising next-generation oxygen therapeutics.^114–118

Fig. 4. Structure and degradation mechanism of cell-like hemoglobin-based oxygen carriers (HBOCs). (A) Chemical structures of representative polymers used as hydrophobic or hydrophilic blocks, as well as polyethylene oxide (PEO)-based biodegradable block copolymers. (B) Structural comparison of a polymersome and a liposome, showing representative diameters and membrane thicknesses. (C) Schematic mechanism of esterase-catalyzed polymer degradation, illustrating hydrolysis of a polylactic acid (PLA) ester bond by the catalytic triad in the enzyme active site. All panels are adapted from ref. 120 with permission from Royal Society of Chemistry, copyright 2023.

Bio-inspired material design has emerged as an important strategy for translating natural structures into engineered oxygen carriers. In this context, Hb-containing polymer particles have been developed to mimic the morphology and oxygen-delivery function of RBCs.^119,120 One approach used electrohydrodynamic jetting to generate spherical PLGA templates with RBC-like dimensions (7 ± 2 µm), followed by layer-by-layer deposition of Hb/bovine serum albumin (BSA) or poly(allylamine hydrochloride)/BSA shells and subsequent removal of the PLGA core in an alcohol/tetrahydrofuran solvent mixture. This process produced oxygen-carrying particles with biconcave discoidal morphology and mechanical elasticity resembling RBCs (Fig. 5A).^119,121 However, whether these RBC-like structural and mechanical properties can be preserved under biological conditions remains unclear.

Fig. 5. Synthesis strategies for cellular hemoglobin-based oxygen carriers (HBOCs). (A) Preparation of red blood cell (RBC)-like particles using poly(lactic-co-glycolic acid) (PLGA) templates, followed by layer-by-layer (LbL) coating, protein cross-linking, and template removal to generate cellular HBOCs. (B) Production of hemoglobin nanoparticles (Hb-NPs) by desolvation precipitation, followed by polydopamine (PDA) coating through dopamine self-polymerization. (C) One-pot synthesis of hemoglobin-loaded zeolitic imidazolate framework-8 nanoparticles (Hb@ZIF-8 NPs) using hemoglobin (Hb), zinc ions (Zn²⁺), 2-methylimidazole (HmIm), and polyethylene glycol (PEG) or polyvinylpyrrolidone (PVP) as capping agents. Panel A is adapted from ref. 136 with permission from National Academy of Sciences, copyright 2009. Panel B is adapted from ref. 123 with permission from American Chemical Society, copyright 2021. Panel C is adapted from ref. 137 with permission from Elsevier, copyright 2024.

Recently, maleimide-activated stroma-free Hbs were crosslinked using dithiothreitol, and residual surface maleimide groups were further conjugated to sulfhydryl-terminated poly(2-ethyl-2-oxazoline) (m.w. = 5 kDa).¹²¹ The resulting spherical PEH of 30 nm in diameter showed high oxygen affinity with low cooperativity (P₅₀ = 9 mmHg), good hemocompatibility, and minimal vascular, renal, or hepatic toxicity. Resuscitation with this PEH solution in a murine model of 50% hemorrhagic shock achieved 100% survival, together with rapid restoration of blood pressure and heart rate. Alternatively, Hb nanoparticles have been prepared by desolvation precipitation of stroma-free bovine Hb and coated with polydopamine or metal-phenolic networks composed of tannic acid and Fe³⁺ ions (Fig. 5B).^122,123 These coatings prevent Hb dissociation and oxidation while preserving reversible oxygen binding and release, although the in vivo performance remains to be evaluated.

As innovative materials winning 2025 Nobel Prize in Chemistry, metal organic frameworks (MOFs) are promising candidates for drug delivery applications due to their ordered crystallinity, structural diversity, tunable porosity, and high loading capacity, which enable effective encapsulation of biomolecules within their framework.^124,125 A one-pot synthesis by adding Hb, Zn²⁺, 2-methylimidazole, and PEG encapsulated a large amount of Hbs within zeolitic imidazolate framework (ZIF), forming ZIF-8 nanoparticles (i.e., Hb@ZIF-8 NPs) (Fig. 5C). With a uniform size at 168 nm, Hb@ZIF-8 NPs achieved a remarkable 95% oxyhemoglobin content and Hb entrapment at 34 mg mL⁻¹ with P₅₀ = 5.5 mmHg.¹²⁶ The PEG coating not only confers stealth properties to prolong circulation times in vivo but also stabilizes the particles, preserves the Hb functionality, and protects the entrapped Hbs from oxidation, finally endowing Hb@ZIF-8 NPs with excellent bio- and hemo-compatibility.¹²⁷ An upgraded variant of Hb@ZIF-90 NPs was recently fabricated by optimizing the ratios of Zn²⁺ : imidazole-2-carboxaldehyde (2-ICA) to maximize the resultant Hb capacity, resulting in enhanced O₂ affinity (P₅₀ ≈ 8 mmHg), which makes it suitable for ischemia reperfusion.¹²⁸ Further preclinical and clinical experiments are needed to validate their capable role as an RBC substitute.

3.4. Hb-free oxygen carriers

A novel Hb-free micron-scale lipidic oxygen carrier has been developed to rapidly transfer large amounts of O₂ to deoxygenated Hb and immediately restore blood oxygenation. These non-Hb microparticles, with a mean diameter of 2–4 µm, consisted of a gas core containing up to 90% O₂ by volume, a phospholipid monolayer, and outer block-copolymer layers.¹²⁹ After intravenous injection, the particles transferred encapsulated O₂ to deoxygenated Hb within RBCs, reversing hypoxemia in rabbits within seconds. Continuous low-rate infusion (∼4 mL kg⁻¹ min⁻¹) during 15 min of tracheal occlusion maintained systemic oxygenation without evidence of toxicity.¹²⁹ Notably, this oxygenation strategy does not require an intact airway-lung unit, making lipid-stabilized O₂ emulsions a promising emergency gas delivery platform for refractory hypoxemia or acute respiratory distress syndrome (ARDS).

Such O₂-containing microbubbles/nanobubbles have been developed as oxygen-delivery platforms to enhance oxygen therapeutics. One common method is to suspend a mixture of DSPE-PEG/phospholipids or lecithin/glycyrrhizic acid/citric acid/glycerol in the saline solutions accompanied by repeating sonication with O₂ supply at a low flow rate, so forming lipid-shelled microbubbles (1.56 ± 0.76 µm) or nanobubbles (262.8 ± 78.1 nm) preferentially engulfed by phagocytic cells.¹³⁰ Alternatively, inspired by freshwater algae that produce gas-filled and protein-shelled vesicles, the collected gas vesicles were further coated by a lipid layer and thoroughly sonicated, thus forming nanobubble formulations with a mean diameter of 300–330 nm and a uniform distribution (Fig. 6A and B).¹³¹ These lipid-coated gas vesicles exhibited favorable O₂ release kinetics and effectively improved tumor oxygenation within minutes after tail-vein injection in hypoxic murine models.

Fig. 6. Fabrication and characterization of lipid gas vesicles (GVs) and polymeric microbubble (PMB) gas carriers. (A) Schematic illustration of lipid-GV preparation using dioleoyl phosphatidylcholine (DOPC). (B) Transmission electron microscopy (TEM) image showing the morphology of lipid-GVs. Scale bar = 100 nm. (C) Design and fabrication of low-molecular-weight (LmD) dextran PMBs for O₂ delivery. LmD PMBs are prepared by homogenization with simultaneous acid titration to induce interfacial cross-linking and shell formation. After contact with blood, the PMBs rapidly dissolve to release gas, while the shell components revert to soluble LmD constituents for subsequent clearance. (D) Cryo-scanning electron microscopy image of LmD PMBs, showing a thin shell and smooth surface. Panels A and B are adapted from ref. 138 with permission from Elsevier, copyright 2020. Panels C and D are reproduced from ref. 139 with permission from Springer Nature, copyright 2024.

Nevertheless, the lipid shell may prevent complete bubble dissolution through surface reorganization and gas exchange, potentially causing obstruction via gas embolism even at low doses in healthy animals. This safety concern may limit these lipid-coated O₂ micro-/nano-bubbles from being used as resuscitation therapeutics in emergency settings. To address this issue, a low-MW dextran of 6 kDa (LmD) chemically modified with acetyl and carboxyl groups was prepared as a pH-responsive amphiphile that is soluble at physiological pH but insoluble in an acidic milieu. The dissolved LmD polymer was homogenized at the air–water interface, followed by gradual acidification and acid-induced protonation of carboxylic groups, which resulted in formation of a thin-shelled (<50 nm) O₂ microbubble (mean diameter = ∼5 µm) around the gas core.¹³² The polymeric microbubbles can store a high volume of O₂ with good stability, while the pH-responsive dextran shell may quickly dissolve in blood, enabling a fast delivery of O₂ to maintain critical oxygenation in swine models with acute and severe hypoxemia (Fig. 6C and D).¹³²

Further development of O₂-containing particles or vesicles for clinical use should focus on two improvements. First, the lipid-polymer shells should be rapidly degraded or excreted after oxygen transfer to prevent toxic or potentially fatal intravascular accumulation. Second, for emergency applications, particularly when pulmonary ventilation is impaired, these oxygen-delivery systems should be engineered to facilitate exchange of O₂ with metabolic gases such as CO₂, thereby reducing the risk of hypercapnia and subsequent acidosis.

4. Conclusions

The increasing demand for transfusable oxygen carriers, together with persistent concerns regarding blood shortages, storage limitations, compatibility testing, and transfusion-transmitted infections, continues to drive the development of RBC substitutes. However, the history of synthetic oxygen carriers shows that oxygen transport alone is insufficient for clinical success. Earlier formulations of PFCs and acellular HBOCs were limited by several recurrent problems, including Hb oxidation, NO scavenging-mediated vasoconstriction, complement and macrophage activation, suboptimal oxygen-release profiles, and difficulties in reproducible large-scale manufacturing. These limitations indicate that future RBC substitutes should not be designed merely as oxygen reservoirs, but as integrated systems that combine the advantages of engineered production and long-term storage with the vital physiological functions of natural erythrocytes.

Among the various oxygen-carrier platforms developed to date, cell-like HBOCs appear particularly promising because they compartmentalize Hb within vesicular, polymeric, or hybrid shells, thereby avoiding direct Hb exposure to the vascular endothelium while allowing more controlled oxygen loading, release, circulation, and clearance. Platforms such as PEG-LEH, HbVs, RBC-like particles, and nanoparticle-based Hb carriers provide valuable opportunities to combine Hb stabilization, redox control, immune evasion, and biodegradable shell design within a single system. Compared with earlier acellular HBOCs, these cell-like carriers are better positioned to overcome the principal challenges associated with previous clinical failures, although their long-term safety, immunogenicity, degradation kinetics, and organ accumulation still require systematic evaluation.

Looking forward, future RBC substitutes should therefore prioritize mechanism-driven strategies rather than empirical modifications. Major directions include optimizing Hb encapsulation and stabilization to reduce oxidative injury; tuning particle size, surface chemistry, and mechanical properties to balance circulation time with organ clearance; adopting biodegradable shells to avoid long-term tissue retention; and adjusting oxygen affinity and release kinetics for specific clinical scenarios, such as trauma resuscitation, ischemia-reperfusion injury, or perioperative anemia. Importantly, standardized pharmacological, toxicological, and immunological assays should be established to enable comprehensive assessments of hemocompatibility, pharmacokinetics, and biodistribution and to allow meaningful comparisons across different platforms. Ultimately, successful translation will depend not only on enhanced oxygen-transport capacity, but also on demonstrable biosafety, disease-specific functionality, therapeutic efficacy, manufacturing scalability and reproducibility, and rigorously validated clinical utility.

Author contributions

XS, HQ and ZT proposed the idea and outline, YS, KZ, YL, XZ and ZT drafted the original manuscript. YS and KZ prepared the figures and tables. YS, WX, XS, HQ and ZT revised the manuscript. All authors contributed to the writing and submission of this manuscript.

Conflicts of interest

The authors declare no competing interests.

Abbreviations

BLA

Biological license application

BSA

Bovine serum albumin

DOPC

Dioleoyl phosphatidylcholine phatidylcholine

DSPE

Distearoyl phosphoethanolamine

FDA

Food and Drug Administration

HBOCs

Hemoglobin-based oxygen carriers

Hbs

Hemoglobins

HbVs

Hemoglobin vesicles

LEHs

Liposome-encapsulated Hbs

LmD

Low-molecular-weight dextran

metHb

Methemoglobin

MPS

Mononuclear phagocyte system

MW

Molecular weight

NBCUS

National Blood Collection and Utilization Survey

NET

Neutrophil extracellular trap

NAT

Nucleic acid amplification test

NO

Nitric oxide

PCL

Poly(ε-carprolactone)

PE

Phosphatidylethanolamines

PEG

Polyethylene glycol

PEHs

Polymer-encapsulated Hbs

PEO

Polyethylene oxide

PFC

Perfluorocarbon

PIBCA

Poly-isobutyl-cyanoacrylate

PLA

Polylactic acid

PLGA

Poly(lactic acid-co-glycolide)

pO₂

Oxygen partial pressure

RBCs

Red blood cells

RCTs

Reticular connective tissues

WBCs

White blood cells

Data availability

No primary research results, software or code have been included, and no new data were generated or analyzed in this review.