Serum albumin (SA) is the most abundant plasma protein in blood, and its well-known role in maintaining plasma oncotic pressure and regulating fluid distribution between vascular compartments (1–3) has historically overshadowed its other important functions. These include binding and transport of a diverse array of endogenous molecules and exogenous drugs, combined with its antioxidant and anti-inflammatory properties (1, 4). It is difficult to isolate a single function of SA, as its many roles are interconnected and collectively serve to maintain homeostasis in the body.

The objectives of this review are to expand clinical knowledge regarding the multiple functions of SA, elucidate the potential causes and effects of dysregulated SA concentration, and discuss the clinical relevance of hypoalbuminemia (defined as SA concentration <3.5 g/dL) (5) resulting from various diseases or certain drugs. The detailed mechanisms modulating drug interaction with SA in the presence and absence of endogenous and exogenous ligands have been described extensively elsewhere (6, 7) and will not be reviewed in detail here. Rather, we focus on several interconnected effects of SA on health and the potential clinical impact of SA concentration in patients with known risk factors, patients with chronic inflammatory disease or an autoimmune disorder, and patients with multiple comorbidities (e.g., dyslipidemia) (8). We also examine the potential for certain immunomodulatory neonatal Fc receptor (FcRn) inhibitors to reduce SA concentration and for reduced SA concentration to, in turn, affect the efficacy and toxicity of some medications. A better understanding of the interplay among SA concentration and endogenous and exogenous molecules will help clinicians recognize the potential for SA-related complications, elicit novel therapeutic insights, optimize treatment strategies and regimens, and improve quality of life for patients.

Albumin is exclusively synthesized by hepatocytes at a rate of roughly 150 mg/kg/day (9), meaning a person weighing 70 kg would synthesize approximately 10.5 g/day. The rate of synthesis is dependent on the body’s needs and on alterations in colloid osmotic pressure, as well as osmolality of the hepatic extravascular space (10). When a change in oncotic pressure in the hepatic vascular beds is detected by osmoreceptors in the hepatic interstitial matrix, the rate of albumin synthesis changes accordingly (10). It has been suggested that a healthy liver can boost synthesis of albumin by up to 3-fold in response to increased turnover or catabolism; however, there is scant evidence to support this widely held assumption (11).

Following synthesis in the liver, albumin is released in the extracellular space via exocytosis; ~40% is intravascular, and ~60% is distributed throughout organ/tissue interstitial spaces, predominantly in muscle, adipose and connective tissue, and skin (11). Via lymphatic vessels, there is movement of albumin between interstitial and intravascular spaces (1). Albumin has a high concentration in plasma, comprising roughly half of total protein content, and the high concentration together with a strong negative net charge are major contributors to the maintenance of plasma oncotic pressure (1). Interstitial concentration of albumin is lower than in serum and varies by anatomical location (1); concentrations of ~7 g/L in adipose tissue and ~13 g/L in skeletal muscle have been reported (12).

SA has a dominant role in binding and transport of numerous endogenous molecules, including thyroxine, bilirubin, amino acids, and fatty acids (FAs), throughout the body (1, 11). SA is considered the major plasma carrier of FAs, which play a critical role in generation and storage of energy for various cells and tissues throughout the body (13, 14). SA can accommodate a range of FA chain lengths among the 7 binding sites with differing affinities (FA1 through FA7) located across its 3 domains (15–17). SA binds cholesterol-containing vesicles and mediates transport of molecules between fibroblasts and other peripheral cells (18, 19). It also strongly binds and transports exogenous molecules—and their affinity for SA can influence drug behavior, potentially affecting the rate of delivery and the efficacy or toxicity of a drug (17, 20).

Its multiple binding sites provide an ideal platform for free radical scavenging, giving SA its robust anti-inflammatory and antioxidant properties (4, 21). SA also binds various inflammatory mediators and is involved in regulating the immune response in systemic inflammation (4). Because SA has been suggested to be responsible for >70% of total free radical–trapping activity, it constitutes the dominant antioxidant in the circulatory system (22). The reduced sulfhydryl group of Cys34 in domain I works as a free radical scavenger for multiple reactive oxygen species (ROS) and reactive nitrogen species (21, 23). SA can also bind free metal ions, thereby controlling their reactivity, limiting their availability, and decreasing ROS formation (21).

In healthy humans, the half-life of SA is roughly 21 days; ultimately, ~84% is catabolized, with the remainder excreted via feces (~10%) and urine (~6%) (11). It should be noted that the amount of SA lost through urinary output is minimal, normally <20 mg/day (24). The unusually long half-life of SA is regulated by FcRn, a broadly expressed cellular receptor predominantly found in acidified endosomes (25, 26). This regulation occurs via pH-dependent engagement by the receptor (27), with binding at acidic endosomal pH (6.5-6.0) that becomes negligible at near-neutral pH (7.0-7.4), directing recycling or transcytosis across polarized cell layers and resulting in rescue of albumin from intracellular degradation (28, 29) (Figure 1A). Specifically, following fluid-phase pinocytosis, albumin binds FcRn at acidic pH within endosomes, then the FcRn-albumin complex recycles to the cell surface, where exposure to the neutral pH of the extracellular milieu triggers release of albumin into the circulation (31–33). Similarly to the endosomal pathway of immunoglobulin G (IgG) recycling (34), the albumin recycling process occurs in both hematopoietic and nonhematopoietic cells, particularly macrophages and endothelial cells, ensuring wide distribution of albumin throughout the body (35, 36). As such, in addition to liver synthesis, pH-dependent FcRn binding is critical for maintaining a high circulatory concentration of SA.

Because IgG antibodies also bind FcRn in a comparable pH-dependent manner, IgG has a similarly long half-life as albumin (25, 37, 38); however, IgG and albumin binding sites on FcRn are distinct and do not overlap (29, 38). In analyses of the structure of FcRn-albumin-antibody complexes, it has been shown that although binding of IgG and albumin to FcRn can occur simultaneously (30, 32) (Figure 1B), the fragment antigen-binding (Fab) arms of bound IgG may lead to steric clashes with the cell membrane, resulting in an orientation of FcRn with respect to the membrane that is less favorable for albumin binding (37). It has been proposed that, to overcome these clashes, IgG binds FcRn in a T-shaped conformation, which would be enabled by the highly flexible nature of the Fab arms (39). Studies have shown that the Fab arms of an IgG molecule can affect behavior in FcRn-mediated functions (40, 41), indicating that these effects vary for antibodies with different variable domain sequences and Fab arm flexibility (41, 42). The importance of FcRn for maintaining SA concentration has been demonstrated in multiple in vitro and preclinical studies in which FcRn deficiency decreased both half-life and concentration of SA (28, 35, 43, 44). FcRn knockout mice reportedly have an albumin catabolic rate twice that of normal mice, but synthesis rates only 20% higher than in normal mice and SA concentration that is 40% lower (45). FcRn directs newly synthesized albumin into the vascular space rather than into bile, a process that protects the liver from being damaged by albumin-bound toxins or drugs. Hepatic FcRn deficiency in mice has been shown to increase loss of SA into bile and influence the development of hypoalbuminemia (35). In addition, preclinical studies investigating maintenance of adequate SA concentration showed that FcRn expressed in the kidneys is also involved in preventing loss of SA. Albumin is filtered across the glomerulus and reclaimed by proximal tubule cells (46). FcRn in the apical area of the proximal tubule facilitates reclamation and mediates pH-dependent transcytosis with the albumin-binding cubilin-megalin complex (46, 47). FcRn-deficient mice were shown to excrete more SA into urine than did wild-type mice, and wild-type mice transplanted with a kidney lacking FcRn developed albuminuria (48).

SA is the principal plasma protein for the binding of drugs (49). The 3-dimensional structure of albumin comprises a single nonglycosylated polypeptide chain with 3 domains (I, II, and III), each consisting of 2 subdomains (A and B) and multiple reversible and irreversible binding sites (3, 20, 50). There are 2 main drug-binding sites, one in the IIA subdomain (Sudlow site I) and the other in the IIIA subdomain (Sudlow site II) (51). SA can also be harnessed to extend the half-life of drugs and reduce drug toxicity (23). Used as a carrier, SA joined to a therapeutic compound via covalent binding or noncovalent conjugation or fusion enables extension of the drug half-life and improved pharmacokinetic properties (52, 53).

Alterations to the structure of albumin or changes in its concentration in serum can alter drug-binding capacity, affecting pharmacokinetics and impacting therapeutic efficacy and/or side effects (7). Because only the unbound drug can interact with its target(s), alterations in SA concentration can lead to a drug being metabolized less or more quickly, influencing therapeutic effect (20). In addition, because low SA concentration leads to a reduction in available binding sites, a larger ratio of an administered SA-binding drug can be present as free drug, increasing the potential for side effects (54). The free fraction of a drug can be affected by concomitant administration of other drugs that also bind to SA, resulting in drug-drug interactions at the level of protein binding (17).

The long circulating half-life of SA increases its susceptibility to posttranslational modifications (PTMs), including glycation and oxidation (23, 52, 55). PTMs can alter the structure and/or function of SA (4, 56) and negatively impact its ability to bind FcRn, thereby decreasing its half-life (52, 57). Increased rates of glycation of SA, commonly seen in diabetes, can reduce drug-binding capacity, as can glycation-induced alterations to its molecular structure, independent of rate (3, 56). Increased glycation of SA has also been reported in patients with heart failure (HF), with highest levels in those with the most severe HF (58). Oxidation can similarly impact the binding properties of SA, and because oxidized SA is removed more quickly from circulation, overall availability and capacity for binding and transport are reduced (50). Increased S-thiolation (oxidation) of albumin has also been shown in the plasma of patients with HF, and level of thiolated SA may represent a viable marker for HF prognosis and diagnosis (59).

Many of the drugs in the standard treatment armamentarium for autoimmune disorders, including glucocorticoids such as prednisolone and methylprednisolone, as well as nonsteroidal immunosuppressants such as mycophenolic acid and tacrolimus, are highly protein bound, as are many common cardiovascular (CV) medications (Table 1). Patients with autoimmune conditions are frequently treated for CV comorbidities; in a sample of patients with myasthenia gravis (MG), dyslipidemia was the most common comorbidity (60%; 140 of 234 patients) (63). In addition, certain lipid-lowering drugs can affect SA concentration by increasing its urinary excretion. Statin treatment has been shown to be independently associated with microalbuminuria (64), defined as urinary excretion of 30 mg to 300 mg of albumin in a 24-hour period (65). It has been reported that approximately one-third of patients with MG are receiving statins (66) for the treatment of dyslipidemia, a prominent risk factor for CV disease (67). Accordingly, the potential for consequences from altered SA concentration in specific disease states may have greater clinical significance.

The clinical impact of hypoalbuminemia is drug-specific and most well characterized for narrow therapeutic index drugs, defined by the US Food and Drug Administration (FDA) as drugs that can result in “serious therapeutic failures and/or adverse drug reactions that are life-threatening or result in persistent or significant disability or incapacity” due to minor differences in dose or blood concentration (62). For example, with valproic acid and phenytoin, it has been shown that reduced SA concentration can lead to a higher free fraction of these drugs (68), increasing the probability of (neuro)toxicity. The anticoagulants warfarin and rivaroxaban are both highly protein bound (99% and 92%-95%, respectively) (69) and hypoalbuminemia is associated with higher risk for overanticoagulation in patients taking warfarin (70). Lower SA level has also been associated with higher risk of bleeding events in rivaroxaban-treated patients; each decrease of 1.0 g/dL of SA in one inpatient population was shown to increase bleeding risk 4.4-fold (69). For many other drugs, both decreased efficacy and/or safety are still rather theoretical risks, based on mechanistic insights. To what extent these risks will be clinically relevant is a matter of ongoing debate.

At present, SA concentration is normally assessed and monitored only in acute disease flares and following trauma, e.g., in hospitalized patients, although decreased SA concentration has been linked to worsening disease severity and increased mortality in multiple chronic conditions, including CV, metabolic, and autoimmune diseases and disorders (71, 72). Hypoalbuminemia may also result from primary chronic inflammatory disease, systemic inflammation, and kidney and liver disease, as well as from comorbid conditions (8).

The clinical implications of alterations in SA concentration seen with various disease states are not widely understood by clinicians (8) nor are the potential interactions with pharmacologic treatments, including the potential to enhance or hinder efficacy and increase toxicity of commonly prescribed drugs (7) or trigger or worsen comorbidities such as hyperlipidemia.

Antibody-based drugs targeting the IgG binding site on FcRn are increasingly being explored to treat IgG-mediated autoimmune disorders in humans (88, 89); these drugs block the interaction of FcRn with endogenous IgG. Of the FcRn inhibitors in development, some have demonstrated, in preclinical, phase 1 (Table 2), and phase 2 trials (Table 3), an effect on SA concentration as well (89). It is currently unclear whether differences in effect on SA concentration may relate to differences in FcRn inhibitor design, such as format (full-length IgG vs Fc fragment), subclass, Fc effector functions, mode of binding, affinity, pH dependency, or binding epitope.

Rozanolixizumab/UCB7655 (UCB Biopharma), a humanized IgG4 monoclonal antibody (mAb), showed modest decreases in mean SA concentration in a phase 1 study in healthy adult volunteers treated with a low-dose regimen (91). In a phase 2 study of rozanolixizumab in adult patients with primary immune thrombocytopenia (ITP), a maximum mean decrease in SA concentration of 4.5% was observed with the 20-mg/kg dose at day 8, which returned to baseline levels (98).

A phase 1 study of nipocalimab/M281 (Momenta/Janssen), a fully human IgG1 mAb (aglycosylated to reduce effector functions), showed asymptomatic transient reductions in SA concentration (up to 22%) in patients in the single ascending and multiple dose phases of the study (89, 93). In a phase 2 study in generalized MG (gMG), reductions in SA were observed across 4 intravenous (IV) dosing regimens (5 mg/kg every 4 weeks, 30 mg/kg every 4 weeks, 60 mg/kg every 2 weeks, and 60 mg/kg single dose) over 8 weeks of treatment; the largest reduction was observed in the group receiving 60 mg/kg every 2 weeks (83).

Orilanolimab/ALX1380 [AstraZeneca/Alexion (formerly SYNT001, Syntimmune)], a humanized IgG4 mAb, was reported to have no effect on SA concentration in a phase 1 healthy volunteer study using single ascending doses (94). In a phase 1b/2 study of orilanolimab in patients with PV, no noteworthy changes in SA concentration after 5 weekly IV doses of 10 mg/kg were reported (100).

Administration of batoclimab/IMVT-1401/HBM9161 (Immunovant), a fully human IgG1 mAb with Fc mutations (L234A/L235A) to reduce effector functions, resulted in reversible, asymptomatic, dose-dependent reductions in SA concentration, to 20% below baseline by day 28 after 4 weekly subcutaneous (SC) doses of 340 mg and to 31% below baseline after 4 weekly SC doses of 680 mg, in a phase 1 healthy volunteer study (96). In another phase 1 study of batoclimab in healthy volunteers, a single SC dose of 680 mg resulted in transient reductions of SA up to 10% below baseline at day 11 (95). A phase 2 study in gMG showed reductions in SA concentration of 16% with 6 weekly SC doses of 340 mg and 26% with 6 weekly SC doses of 680 mg (89). A phase 2b clinical trial of batoclimab as treatment for thyroid eye disease was paused when a 12-week assessment showed elevated total and LDL cholesterol levels in active-treatment patients (89, 101).

Efgartigimod alfa (argenx), recently approved by the FDA for treatment of gMG (105), is an engineered Fc fragment derived from human IgG1 that blocks FcRn recycling of IgG by binding naturally via its Fc domain to FcRn (97, 106). SA concentration was not decreased by efgartigimod in a phase 1 trial in healthy volunteers (97). In a phase 2 study of efgartigimod in patients with ITP randomly assigned to receive 4 weekly IV infusions of placebo or efgartigimod at a dose of 5 mg/kg or 10 mg/kg, changes in SA concentration from baseline to 3 days after final infusion were similar for the placebo and treated groups (102). In a phase 2 trial of up to 34 weeks of efgartigimod treatment in PV/pemphigus foliaceus, average SA concentration increased slightly and transiently, within normal limits, and returned to baseline levels after treatment was stopped (103). Total serum cholesterol and LDL levels, as measured in 1 cohort, remained within normal limits across all active study and follow-up assessment points (103). In the phase 3 ADAPT trial of efgartigimod in gMG, no decreases in SA concentration or increases in cholesterol were seen (104).

The clinical importance of SA concentration and the relationship of SA to disease, in combination with therapeutics, is complex. Increasing evidence suggests it may be useful to monitor SA concentration in patients with diseases or disorders that affect its steady-state concentration. Dysregulation of SA in CV disease, diabetes, and certain autoimmune disorders, including MG and GBS, can affect multiple body systems. Oxidation and glycation of SA, resulting from disease processes, as well as from aging, complicate the clinical picture. Awareness of these changes could help clinicians better manage disease.

Although the published literature is limited, hypoalbuminemia in autoimmune disorders has been associated with increased disease severity. Because SA is largely responsible for the antioxidant capacity of serum, the decreased SA concentration commonly seen in patients with autoimmune disorders may contribute to the underlying inflammatory processes. Additionally, many of the primary medications, including prednisolone, methylprednisolone, and the nonsteroidal immunosuppressant mycophenolate mofetil, used to treat various autoimmune diseases are protein bound. Therefore, their efficacy and safety could be affected by reductions in available SA binding sites, leading to increased drug clearance.

Some newer drugs on the horizon for the treatment of autoimmune disorders also have SA-related effects. FcRn inhibitors targeting the IgG binding site on the receptor are not intended to decrease SA, only to accelerate clearance of endogenous IgG. In clinical trials, however, some IgG mAbs that reduce IgG levels by inhibiting FcRn-IgG interactions have had the off-target effect of concomitantly decreasing SA concentration. Important questions remain regarding the link between FcRn inhibitors and SA. Flow cytometry and microscopy data have shown that higher affinity and/or avidity and/or lower pH-dependent binding to FcRn of an anti-FcRn antibody, relative to the engineered Fc fragment efgartigimod, results in greater retention in FcRn-positive compartments, as well as increased lysosomal accumulation (97). Upon investigation of intracellular trafficking of FcRn by immune complexes, it was shown that cross-linking of FcRn diverted the majority of the cross-linked receptors to lysosomes (107). These findings suggest that the different binding properties of FcRn-specific antibodies and an engineered Fc fragment (with increased affinity for FcRn) might modulate the behavior of FcRn within cells, leading to possible effects on SA recycling.

An increased understanding of the nononcotic functions of SA, especially during this period of accelerated drug development in MG (89, 108), may help improve the risk/benefit assessment and selection of optimal therapeutic regimen for a wide array of disease conditions and patient types. Although increased catabolism or excretion of SA is partly compensated for by increased synthesis, resulting in a milder net loss, stimulating liver synthesis of proteins such as apoB-100 and fibrinogen can result in comorbidities, including dyslipidemia. Greater clinical attention to SA concentration is warranted for older patients, patients on multiple drugs that could potentially affect or be affected by SA concentration, and patients with multiple comorbidities.

As FcRn inhibitors become used more widely in clinical trials, the potential for mild to moderate decreases in SA concentration to have subsequent impacts on lipid panels and CV risk may become increasingly apparent as additional patient populations are treated. Because many of these patients will be older and may have comorbidities, the importance of a fuller understanding by healthcare providers of the role of SA in healthy and disease states will grow more urgent. The most obvious clinical indications are the various IgG-mediated auto- and alloimmune- diseases, meaning these patients may be inherently more susceptible to treatment side effects resulting from lowered SA concentration. Fortunately, several candidate FcRn inhibitors have shown little or no effect on SA concentration and might, therefore, be safer in a broader array of disease settings.

Increased awareness is needed regarding the clinical importance of monitoring plasma SA concentration in patients with diseases or disorders known to correlate with decreased SA concentration and/or who are being treated with drugs that may lower SA concentration, independently or in combination with concomitantly administered drugs. As more data become available, the question whether clinicians should routinely evaluate SA concentration to aid in monitoring disease progression and decreasing the potential for side effects and adverse events could gain traction. As such, SA concentration may become an increasingly important consideration in the design and outcomes of clinical trials using FcRn inhibitors, as well as potential future treatment with FcRn inhibitors that are associated with decreases in SA concentration.

ESW – provided substantial contributions to the design of the work and interpretation of current data based on expertise in FcRn-SA-IgG interaction and to figure development; drafted sections of the work and revised the manuscript critically for important intellectual content; shares first authorship; DG – provided substantial contributions to the design of the work and interpretation of current data based on her clinical expertise in the neuromuscular field; drafted sections of the work and revised the manuscript critically for important intellectual content; shares first authorship; ED – provided content on exogenous drug binding and revised the manuscript critically for important intellectual content; JVS – provided original concept, initial research, and content; revised the manuscript critically for important intellectual content; JTA – provided expertise in and content on FcRn-SA-IgG interaction; revised the manuscript critically for important intellectual content; NS – provided clinical and application expertise in neuromuscular disorders and revised the manuscript critically for important intellectual content; JEK – provided clinical expertise on albumin in hematology; revised the manuscript critically for important intellectual content; DS – guided understanding on the interaction of human albumin with FcRn and other proteins; revised the manuscript critically for important intellectual content; DR – provided clinical and research expertise on albumin’s relation to cholesterol metabolism; provided text and content for new direction of the manuscript; revised the manuscript critically for important intellectual content; JJPK – provided expertise on cholesterol metabolism for new direction of the manuscript and reframed the discussion of albumin’s potential role in cholesterol metabolism; revised the manuscript critically for important intellectual content; EL – developed original concept for manuscript and provided nonclinical expertise on FcRn; provided content and critical revision of manuscript for important intellectual content; GV – provided expertise on FcRn-SA-IgG interaction; revised the manuscript critically for important intellectual content; IS – provided content on exogenous drug binding and revised the manuscript critically for important intellectual content. All authors contributed to the article and approved the submitted version.

DG, EL, and JVS are employees of argenx, the manufacturer of efgartigimod (FDA approved in gMG and under investigation in primary immune thrombocytopenia, pemphigus vulgaris and foliaceus, chronic inflammatory demyelinating polyneuropathy, bullous pemphigoid, and idiopathic inflammatory myopathy [myositis]). JTA has received funding from Syntimmune and argenx as part of fee-for-service agreements. ED is a postdoctoral research fellow of the Research Foundation – Flanders (grant number 12X9420N). JEK has participated on the Medical Advisory Committee of Sanofi Genzyme and Plasma Advisory Committee of Haemonetics. DR is the founder of Staten Biotechnology. NS is a member of the scientific advisory board and a speaker for argenx. IS, is supported by the Clinical Research Fund of the University Hospitals Leuven. ESW may receive royalties from patents owned by the UK Medical Research Council, UT Southwestern Medical Center, and Texas A&M University, and has financial interests in argenx, Astero BioPharma LLC, Astero Erado Inc., and Astero Technologies LLC. GV is a paid consultant for argenx.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The development (medical writing and editorial support) of this manuscript received funding from argenx. Authors who are not affiliated with argenx were not paid for their contributions to this work. The funder was involved in conception of design, interpretation of data, writing of this article, and decision to submit it for publication.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.