Physiology and pharmacology of albumin

This chapter is relevant to Section Q4(i) of the 2017 CICM Primary Syllabus, which expects the exam candidates to demonstrate "understanding of the pharmacology of blood and its components, including individual factor replacement". To some extent, it also satisfies Section Q4(iii), which expects them to "understand the process of collection and production of blood and its components". The properties of albumin as a blood product have never been discussed in the CICM Part One exams, but the consequences of infusing it into people have come up occasionally, most notably in:

• Question 9 from the first paper of 2022
• Question 12 from the second paper of 2020
• Question 1 from the second paper of 2015
• Question 1(p.2) from the second paper of 2009

It is hard to figure out where to put albumin-related material. Is it a blood product? A colloidal fluid? A resuscitation strategy? A pharmacological agent? Fortunately, the ethos of Deranged Physiology supports and celebrates disorganization in all its forms. As the result, the physiological role of albumin, the various effects of albumin infusion and the metabolic fate of albumin molecules all ended up being scattered randomly throughout this resource. 

Or, alternatively, you could view it as a drug, and use a completely pharmacological structure, in which case:

In terms of peer-reviewed resources, the Australian supplier is CSL, and they have this handy data sheet here. For a reference without the taint of corporate sponsorship,  Matejtschuk et al (2000) discuss the properties and production of albumin with enough detail, and their paper is available for free. For a discussion of the physiological effects of albumin, one may be referred to the (also free) article by David & Michael Levitt (2016). And, if one absolutely has to pay for something, make it Theodore Jr. Peters All About Albumin (1995), which is basically four hundred pages all about albumin. 

Contents of the 20% concentrated albumin solution

This soup is a concentrated separated product of multiple unremunerated donors, which is heated at 60 degrees for 10 hours. As you can see, not only albumin comes in the bottle. You get some sodium octanoate as well.  The 20g of albumin has its own osmolality, and is dissolved in the electrolyte solution which supplies most of the remaining osmoles, but still in comparison to human plasma this solution is hypo-osmotic. The manufacturer data reports a nominal osmolality of 130mOsm/Kg. Given that osmolality relates to the number of particles, one makes sense of this by recalling that 20 grams equates to only 0.3 mmol of albumin (its molecular mass is 66.438 kDa), so most of that 130mOsm comes from electrolytes.

The reason for the hypotonicity of this solution has nothing to do with any specific pharmaceutical properties of the albumin itself. Rather, it was a conscious decision on the part of the manufacturer. Matejtschuk et al (2000) discusses the production of albumin in some detail, and explains that the low-salt option was developed "to avoid electrolyte disturbances". Realistically, you could make 20% or 4% albumin with an isotonic sodium concentration, and if you did you might find that it no longer has any ICP-increasing effect (Iguchi et al, 2018 did exactly this).

Preparation and pharmaceutics of albumin

Albumin is prepared by a series of steps which start with plasma. This is usually the plasma that remains following the centrifuge separation of whole blood into plasma and packed red blood cells. Historically, apparently expired red blood cells were also used (i.e. this is where whole blood was collected and stored). Albumin is then extracted from the plasma either by using ethanol fractionation or chromatographic fractionation (which is the more modern method).

Back in the days of World War 2, people started adding sodium caprylate (now known as sodium octanoate) to stabilise the albumin for transport and storage. Additionally, octanoate is weakly antimicrobial and antifungal; it also keeps the albumin from denaturing during the ten hours of 60-degree heat treatment. It has the delicious distinction of being a common constituent of mammalian milk, and it is named after goats (caprylate, derived from capris). This poster from Germany warns us that octanoate is not very well metabolised by people who have dysfunctional livers.

Pharmacokinetics of albumin

Administration and endogenous synthesis

Albumin is administered intravenously, as to give it orally would make no sense whatsoever. Internally, it is produced by the liver, at a rate of approximately 10g/day. Most people who have spent some time in the ICU will have noticed how it tends to drop rather rapidly in critically ill patients, and a part of that is the typically leaky capillaries of the critically ill patient which allow a lot of the albumin to escape into the interstitial compartment. However , another part is a cytokine-induced decrease in synthesis (i.e the liver preferentially prioritising the synthesis of acute phase reactants). Overall, there are several possible factors which influence the rate of albumin synthesis. Here is the content of a table from Chien et al (2017) which outlines some of these:

Factors which increase albumin synthesis:

• High protein diet
• High caloric intake
• Decreased colloid oncotic pressure
• Growth hormone
• Corticosteroids
• Insulin

Factors which decrease albumin synthesis:

• Protein malnutrition
• Decreased caloric intake
• Increased plasma oncotic pressure
• Diabetes
• Liver disease
• Sepsis
• Trauma

Distribution

As albumin is infused, its concentration in the vascular compartment falls rapidly. At least 50% of it is already gone by the end of the first day. This albumin is migrating into the extracellular compartment, where (it would appear) much of the albumin in your body resides.

This might sound weird (given how much weight is attributed to the plasma oncotic pressure as a mechanism of keeping fluid in the intravascular compartment).  One might be fooled into thinking that interstitial fluid is some sort of crystal-clear mountain stream, devoid of oncotic pressure-generating solids. This is far from the truth. Remember how much protein ends up in transudative fluids. Remember also how an entire system of lymphatic vessels is required to de-proteinate this supposedly protein-free extracellular fluid.

The transcapillary movement rate is an average of two rates, on fast (12%, into the viscera) and one slow (2%, into the skin and muscle).  Some of this movement is by filtration, and some is (surprisingly) by active transcytosis. The responsible protein ("albondin") has been identified. It is present on selective endothelia; for instance, not in the brain (in keeping with the low albumin content of the CSF).  The diagram below is a crude approximation of the multi-compartment comings and goings of human albumin, with numbers calculated for your standard 70kg Homo vulgaris. Theodore Jr Peters offers a table where the distribution of albumin among the tissues is broken down; the data therein is offered below, because it is excellent even though much of it is derived from rat and rabbit models.

There is at any given time about twice as much albumin outside the circulation as there is inside. An average 70kg person contains about 360g of albumin in total. The average plasma concentration is about 40g per L. The total intravascular albumin content is about 118g, and another 177g is sloshing around in the interstitial fluid. Additionally, there is some 65g of albumin trapped in a non-exchangeable state in various body compartments.   Albumin is weakly acidic, with a pKa of 6.75, and it contributes to the Gibbs-Donnan effect by sitting in the intravascular compartment and repelling all the other anions, which accounts for the slightly higher concentration of chloride in the interstitial fluid.

Metabolism and elimination

Albumin is ultimately degraded into amino acids, which enter the free amino acid pool (of the total amino acids available for protein synthesis or enegry production). Its destruction contributes to about 5% of the total daily protein turnover in the body. At a degradation rate of 3.7% per day, an average albumin molecule lives for 27 days, circulating around the fluid compartments. Theodore Jr Peters mentions that albumin molecules collect various small molecules by covalent binding along their life in the circulation, like "barnacles" collected by "any tramp steamer doing its rounds".

The degradation occurs at a rate of 13.3-13.6g/day in a typical 70kg person. The sites of degradation are ubiquitous. The liver contributes only about 15%. The kidneys degrade 10%, and a further 10% is lost via leakage into the GI lumen. Persisting radiolabels that remain in lysosomes have demonstrated that muscle and skin are the major sites of albumin degradation, accounting for 40-60%. The major cell types involved seem to be the fibroblasts and macrophages.

The degradation rate is increased by catabolism-inducing hormones, such as corticosteroids. Sensibly, a rising plasma albumin level increases the rate of degradation (and conversely falling plasma albumin increases the rate of synthesis). The change in the rate of degradation with albumin excess seems to reflect the change in total albumin store (intravascular + extravascular) rather than plasma concentration alone. When Andersen and Rossing overloaded their healthy volunteer with albumin (and doubled his albumin stores), the rate of degradation increased to 200%, but the plasma concentration only rose from 40g/L to 55g/L. The "barnacle-encrusted" albumin with lots of covalently bound small molecules deforming its structure is more easily degraded. Healthy normal albumin seems to survive and recirculate.

Physiological effects of albumin

It would be tempting to continue the pharmacology metaphor and call this section "pharmacodynamics", but realistically albumin is a human blood product and when it is infused it continues performing the physiological role it was always supposed to perform. This role, one must point out, is relatively trivial. Albumin is not an essential component of human life. Some rare examples of its total congenital absence demonstrate that "the virtual absence of albumin is tolerable despite its multiple functions". People who are congenitally without albumin discover this fact randomly in the course of some routine blood test, i.e. they are generally not aware of it being a problem.

So, if it's so useless, what do we even have it for?

Oncotic role of albumin

The oncotic pressure exerted by the albumin is due to the fact that it is a plasma colloid, trapped forever in the intravascular compartment. Normal albumin values exert about 75% of the total 20-30mmHg of oncotic pressure; it stands to reason that a 5-times concentrated infusion of albumin should exert 75-113mmHg of oncotic pressure.

This pressure sucks water out of the interstitial space, which contributes to the plasma-expanding effects of albumin. An American company’s prescribing information reports that every 100ml of 20% human albumin draws an additional 250ml from the interstitial fluid over 15 minutes. The mileage may vary considerably in critically ill patients whose capillaries are leaky, and who may lose more of the infused volume into the extracirculatory spaces.

Additionally, the higly negative charge of albumin attracts sodium into the intravascular space, and holds it there by the Gibbs-Donnan effect. According to David & Michael Levitt (2016), this sodium effect adds another 50% to the intrinsic osmotic activity of albumin.

Antioxidant activity

The aforementioned negative charge of albumin gives it the ability to bind all kinds of nasty reactive chemicals in the bloodtream. Taverna et al (2014), upon reviewing this, came to the conclusion that divalent cations (mainly copper and iron) were the most important ligands for those negatively charged sites, mainly because they are probably responsible for most of the reactive oxygen species synthesis in vivo. By binding these ions, albumin limits their bioavailability and therefore decreases their potential oxidative damage. Additionally, albumin binds bilirubin and homocysteine (thus preventing them from peroxidising lipids). The list of unpleasant radicals controlled directly or indirectly by albumin is probably massive, and according to the the Levitts it accounts for 50% of the total antioxidant activity of the plasma.

Binding function

Albumin binds numerous endogenous and exogenous chemicals; when drugs are "highly protein-bound" they are usually bound to albumin. It carries 50% of the calcium in the bloodstream, as well as fatty acids, bile acids, zinc, and about a thousand other things. This mechanism can decrease the bioavailability of toxic xenobiotics by decreasing their free soluble fraction.