Non-respiratory functions of the lungs

This chapter is most relevant to Section F10(vi) from the 2017 CICM Primary Syllabus, which expects the exam candidates to be able to "outline the non-ventilatory functions of the lungs". This appeared in the exam just once, in Question 7 from the first paper of 2013. Literally, "outline the non-respiratory functions of the lungs" was the entire text of the question, and the college comments were a three-line sigh about how poorly the question was answered.

Fortunately, as the examiners also pointed out, Nunn's has an entire chapter on this. In the 8th edition, that is Ch. 11 (p.203-214). For those who are either without a Nunn's or for whatever reason unwilling to read it, an excellent alternative is Joseph et al (2013), certainly enough to pass a written question. An older paper by Heinemann & Fishman (1969) has a few additional fragments of information which are slightly different to either of the others, and these are explored in much more detail (over 33 pages). With luck, the content of these resources has been effectively remixed and summarised below.

In summary:

• Trap for airborne particles:  generally, nothing larger than 2.5μm gets to the alveoli
• Reservoir of blood: the lungs contain about 10% of the circulating blood volume
• Route of drug administration (eg. nebulised steroids and bronchodilators)
• Route of drug elimination  (eg. volatile anaesthetics)
• Metabolism (eg. conversion of of angiotensin-I, and degradation of neutrophil elastase by α1-antitrypsin)
• Modulator of acid-base balance by virtue of CO₂ elimination
• Modulator of the clotting cascade: the lungs contain thromboplastin, heparin and tissue plasminogen activator
• Filter for the bloodstream: particles larger than an RBC are trapped (~8  μm size barrier), which includes clots, tumour cells and other emboli
• Antimicrobial and immune functions: Alveolar macrophages and sequestered neutrophils, mast cells in the lung and bronchi, immunoglobulin in the respiratory mucus (IgA)
• Modulation of body temperature: heat loss can occur by respiration
• Organ of speech: the lungs form a part of the system which permits communication by sound and language

In addition, one might list some "para-respiratory" functions which are unrelated to respiration per se, such as the pulmonary arterial reflexes or secretion of surfactant. These are largely omitted from the discussion which follows because they have too much to do with gas exchange or lung mechanics, and therefore enjoy a more detailed discussion elsewhere.

The lungs as a barrier against airborne particles

One often sees the lungs described as some sort of defence against airborne particles. This seems somewhat counterintuitive, as those particles just seem to lodge in the lung anyway and cause trouble there. However, that does tend to prevent them from getting into the systemic circulation, so in a way this can be viewed as a barrier function.

One might be able to calculate that, assuming a constant minute volume of 5L/min, a normal person will inhale 7200L of air every day. Of this air, assuming it is indoor air, each 1 m³ will contain up to 259 μg of particles around 10 μm in diameter and up to 202  μg of 2.5  μm particles (Morawska et al, 2017).  That's 3.3 mg of 2.5-10 μm particles inhaled every day, or 1.2g every year- just in normal household air; obviously the situation will be very different if one spends most of one's day at a poultry or swine confinement building. Even when it is not full of aerosolised pig faeces, inhaled air is not exactly clean. It is usually full of inorganic material, as well as swarming with bacteria and fungal spores (Barberán et al, 2015). Even at high altitude, Valsan et al (2016) found pollen, spores, cocci, mineral dust and salt crystals (seen in the stolen SEM images below). 

To give an idea of the distribution of size, the table below is borrowed from the excellent article by Harada & Repine (1985), where the various possible inhaled materials are listed along with their effective diameter in micrometres:

The fate of inhaled particles depends largely on their size. An excellent article by Theodore F. Hatch (1960) contains a graph which perfectly describes the distribution of deposited particles in the respiratory tract:

In short:

• Particles of 5-10 μm in diameter end up largely deposited in the upper airway by what Nunn's calls "impaction", i.e. their velocity in the wair stream causes them to hit walls when the direction of the airflow changes (for example, in the turbulent upper airway).
• Particles of 2-6 μm have a high likelihood of ending up in the lower respiratory tract and alveoli
• Particles around 0.5-2 μm are deposited almost exclusively in the alveoli, where they drop onto the alveolar floor as respiratory gases decelerate. Nunn's calls this "deposition"
• Particles smaller than 0.2-0.5 μm tend to wash in and out of alveoli without interacting very much with the walls, and are not very well deposited, as the influence of gravity is outweighed by the viscosity of the gas at that scale.
• Particles smaller than 0.2μm are sufficiently enslaved by the forces of Brownian diffusion that their deposition onto the walls actually increases because of chance interactions. Hatch (1960) reports that deposition for particles smaller than 0.05 μm is close to 100%.

Whatever these inhaled particles are, clearly it would be counterproductive to allow them to simply land on a surface which is exposed to the entire blood volume. The respiratory system deploys a series of defence mechanisms which assist in protecting the delicate alveolar membrane from inhaled material. These mechanisms are listed by Harada & Repine (1985) in a particularly clear structure which is reproduced here with minimal modification:

• Physical mechanisms
  • Upper airway filtering systems (nasal turbinates)
  • Mucociliary escalator
  • Reflexes
    • Sneezing
    • Coughing
• Cellular mechanisms
  • Alveolar macrophages
  • Lung neutrophils
• Immunologic mechanisms
  • Lymphatic system of the lung
  • Immunoglobulin in mucus and cell surfaces
  • Direct antibacterial action of surfactant (Wu et al, 2003)

The lungs as a route of drug administration

Following logically from the above, the lung can be used as a means of administering therapeutic substances, whether to the alveoli locally or to achieve a systemic effect. Patton & Byron (2007) give an excellent overview. In brief, in order for a drug to be well absorbed by the lung, the drug delivery vehicle has to have certain characteristics:

• Small size: the droplets should be smaller 5 μm in diameter if you intend for them to get out into the systemic circulation (conversely, if your target is the mucosa, feel free to inundate the upper airway with comically huge droplets)
• High lipid solubility and small molecule size: large molecules, eg. proteins, will tend to be mopped up by alveolar macrophages.
• It helps to be recognised as something useful: for instance, the relatively huge IgG molecules (150 kDa) are absorbed relatively rapidly by active transcytosis. This allows the noninvasive delivery of large molecules
• Technique matters: for example, inhaling rapidly may increase or decrease the alveolar delivery of drug particles, depending on particle size
• The excipients of the pharmaceutic preparation must be relatively benign, i.e. one would not want to go to market with an inhaler which causes some sort of pneumoconiosis with prolonged use.
• Unless you specifically want this, it would be important to make sure the drug does not have some sort of weird preference for binding to lung tissue. Particularly, lipophilic drugs with a positive charge tend to do this. Examples of intentional retention in lung tissue are steroids like formoterol and salmeterol and antibiotics like tobramycin and pentamidine. In contrast, the binding of verapamil to lung tissue is completely pointless.

The lungs as a route of drug elimination

Technically speaking, any substance which has a saturated vapour pressure higher than 0 mmHg at 37 °C can be eliminated via the lungs, though obviously, the more volatile the substance is the better.  The lungs may routinely eliminate a number of volatile substances:

• Volatile anaesthetic gases
• Acetone
• Methyl mercaptan (foetor hepaticus)
• Ammonia
• Alcohols (eg. ethanol and methanol) 
• Allicin (garlic)
• Paraldehyde
• Marker substances used to measure circulatory and respiratory physiology, eg. ether, krypton, nitrous oxide, etc

The lungs as a reservoir of blood

Yes, you can store your blood in there. At any given time, the total pulmonary blood volume is somewhere between 200-300ml/m² (indexed to body surface area). These are Evans Green dye dilution measurements from Tarazi (1985), which are presumably still accurate (the exact range 204-314 ml/m² was borrowed from an old 1969 book by Yu & Resnekov, who gave a mean value of 271ml/m²). Thus, a normal-shaped adult male with a BSA of 1.9 m² would have approximately 514ml of blood in their pulmonary circulation at any given time, or about 10% of their total blood volume. Approximately 20-25% of this is present in the pulmonary capillaries at any given time, i.e. about 100-125ml, though this may increase to 50% (250ml) with heavy exercise. These numbers are obviously going to be virtually always inaccurate, as the pulmonary circulation changes its capacity dynamically to frustrate people who are trying to measure its volume as a static parameter. Aggrawal et al (2015) give a table where all the human studies are arrayed, demonstrating a range between 211ml to 550ml. 

What are you supposed to do with this information? It does not appear to be a valuable piece of knowledge. For instance, it is not as if one can draw on this reserve of blood by completely emptying it into the systemic circulation in times of haemorrhage, as some mammals may be able to do with their spleen. Sure, you can flush about 50% of the pulmonary blood volume into the systemic circulation with a particularly vigorous Valsalva, but this is really not a satisfactory solution to acute blood loss. 

One might be able to find significance in the fact that though the right and left ventricle are usually beheld together as an uninterrupted system, in fact there might be differences in their performance from beat to beat, and the pulmonary blood reservoir will therefore act as a temporary store of blood to supply left ventricular preload, even where right ventricular output may is variable. This was demonstrated brutally in a conscious dog model by Guz et al (1961). As it is hard to get a copy of Federation Proceedings from that era, one is only able to rely on citing articles as witnesses to these experiments.

In short, it appears there's enough blood in the pulmonary circulation to support a couple of normal left ventricular stroke volumes, before filling pressure declines and cardiac output suffers.

Metabolic function of the lung

The lung does various metabolic things. These can be described broadly as "biotransformation". Chitkara & Khan (1983) were able to produce 53 pages on this topic. In the briefest form:

• Inactivation of deadly toxins, eg. inactivation of neutrophil proteases by α1-antitrypsin
• Activation of deadly toxins, eg. chlorine gas turns into hydrochloric acid, and harmless chemicals in tobacco smoke turn into potent carcinogens
• Activation of circulating hormones, eg. angiotensin-I and arachidonic acid
• Metabolism of circulating molecules, eg. noradrenaline adenosine and bradykinin

The lung as a modulator of electrolyte homeostasis

One might not think of it as a major contributor to the blood biochemistry, but the respiratory system is in fact involved in modifying the concentration of serum electrolytes, most notably of blood pH and bicarbonate. By exhaling more CO₂, the bicarbonate levels can drop as pH rises.  The lungs are therefore able to act as modulators of acid-base balance; they are the other system (other than the kidney) which is able to modify the pH of the body fluids. This would be important to mention as a non-respiratory function.

The lung as a modulator of fluid balance

Depending on which textbook you read and which ambient conditions they used for their calculations, values given for a normal rate of water loss through the lung will vary. Heinemann & Fishman (1969)  give 250ml as the normal 24-hourly water loss. Obviously that's not going to be the case for the tachypnoeic febrile patient with a tracheostomy. 

The lung as a modulator of the clotting cascade

The lung is one of the richest sources of endogenous procoagulants, anticoagulants and fibrinolytic agents. 

• Thromboplastin is a potent procoagulant which catalyses the conversion of prothrombin to thrombin. The mammalian lung is a rich source of thromboplastin. To compare it to other organs, Østerud et al (1986) put rabbit arteries into a blender and tested the resulting goop for thromboplastin content. The lung had the most - six times more than the kidney and 25 times more than the spleen.
• Heparin occurs naturally in the lung. It is mainly restricted to the lung mast cells, and it resembles the "unfractionated" variety of injectable heparin, as its molecular weight is usually around 20,000 Da (Metcalfe et al, 1979). It is hard to say how much heparin is currently present in any given lung (your lung, reader), but it may be as high as 44.7 μg/g of lung tissue, or approximately 225 Howell units per gram. 
• Plasminogen activator, a fibrinolytic product, is abundant in the lung.  Microvascular cells secrete vast amounts of it (up to 15 times more than other plasminogen activator-secreting cells, according to Takahashi et al , 1998). This works well, because the pulmonary microcirculation is a favourite destination for microthrombi. 

The lungs as a filter for blood

Given that the entire cardiac output comes through the lungs, they act as a convenient filter which prevents fragments of solid material from being flushed into the arterial circulation where they would wreak embolic havoc. That conventionally consist of various thrombotic debris (as mentioned above the lung is particularly well set up to take care of those) but also lots of other things - Jorens et al (2009) list sickle cells, fat globules, foetal material, septic emboli, hydatic cysts, gas and injected particulate material from IV drugs, among others.

As for any sieve, there is of course a size barrier, but for the lung it is unclear what exactly that is. Theoretically, in order to get filtered out by the lung, you'd have to be a particle larger than a red blood cell (which are about 8 μm in diameter). To test the behaviour of larger emboli, Knisely & Mahaley (1958) pureed a tumour and injected its fragments into the venous circulation of eighty-four rabbits. The microembolic phenomena were observed in vivo by means of quartz rod transillumination and a stereoscopic dissecting microscope. The investigators described their observations in poetic detail:

"Opaque, gray masses tumbled along the stream of blood and lodged abruptly at a bifurcation or a point where the vessel narrowed to a diameter approaching that of the particle. Blood continued to flow around emboli for 5-10minutes and up against emboli and out the adjacent proximal capillaries. Soon after lodging, the emboli "grew" by the formation of a white, opaque deposit, particularly on portions of tumor exposed to flowing blood... The blood proximal to the impacted mass continued to pulsate, and sometimes red cells became packed up against the embolus."

These medium-sized emboli (75-100 μm) got stuck at the level of small pulmonary arterioles which taper rapidly down to a diameter of less than 50 μm before joining the pulmonary capillaries.  

Particles in the 50-10 μm range can still get through. The width of capillaries is not uniform (some might be wider) and in any case the diameter changes with respiration, which means emboli could potentially be "milked" through the capillaries by the effects of respiration. Beyond that, there are arteriovenous communications which bypass the capillary bed and allow the embolisation of much larger particles. Niden & Aviado (1956) detected occasional glass bead particles as large as 400 μm in the systemic circulation of dogs.

Some substantial emboli may bypass the size barrier by being particularly deformable. Famously, fat and amniotic fluid are among the embolised substances which are well known to easily gain transpulmonary access to the systemic circulation, particularly where the pulmonary arterial pressure is high (which in both circumstances it usually is). Experimental models certainly have confirmed this. For instance,  Byrick et al (1994) were able to demonstrate that though only 1% of their 15  μm radiolabelled spheres made it through into the arterial circulation of their mongrel dogs, "intravascular fat was found in all brain, heart, and kidney specimens" after bilateral cemented arthroplasty.

Immunological function of the lungs

There are multiple immune functions performed by the lung, which can be broadly separated into "barrier" and "antimicrobial". Bienenstock (1984) did a good overview, which has aged gracefully.

In summary:

• Barrier functions consist of
  • Thick mucus
  • Mucociliary escalator
• Antimicrobial features consist of
  • Alveolar macrophages and sequestered neutrophils
  • Mast cells in the lung and bronchi
  • Immunoglobulin in the respiratory mucus (IgA)
  • Antigen presentation functions of the lung lymphatics
  • Lymphoid tissue in the lung (eg. BALT, bronchus-associated lymphoid tissue) 

The lungs also act as a fairly significant storage reservoir for neutrophils. In fact PEEP can influence this:  Loick et al (1993) showed that this transpulmonary difference in leukocyte count can occur with  10 cm H₂O (the transpulmonary leukocyte gradient was increased four-fold from baseline).

The lung as a modulator of heat balance

According to Heinemann & Fishman (1969), about 350 kcal of heat are lost through respiratory evaporative loss per 24 hr period, though this is obviously going to be higher in the Siberian tundra. It would only be a little higher because the upper respiratory tract is so very awesome at reclaiming heat and water from the expired gas mixture.  Cain et al (1990) subjected some healthy volunteers to a range of temperatures (including a truly Siberian environment of −40°C ) and determined that heat loss due to respiration only varied from 25% to 30% of the resting metabolic rate, even though 60-degree swings of ambient temperature. 

The lungs as an organ of speech

As one of the readers (Dr Emerson) has pointed out, the lungs form a part of the human communications apparatus, as they are the gas reservoir used to generate the noisily turbulent flow we refer to as "speech". This is a really interesting proposition, which appears to be "a valid and possibly even "mark worthy" non-respiratory function of the lung". It is not mentioned in textbooks. Would it score marks? That is impossible to say. In attempting to guess the minds of the examiners, one would have to conclude that it would depend on how it was phrased.

On one hand, it must be acknowledged that the lungs themselves do not contribute, as lung tissue plays no role in the generation of speech except by acting as a storage volume for gas. Sure, the lung can aid in expiration by recoiling passively and expelling gas, but this sort of passive airflow is not how you talk - it is a voluntary and controlled expiration, which is actually the function of thoracic and abdominal respiratory muscles.  Moreover, lungs are not essential for speech (not like the vocal cords), as any source of gas flow will suffice. On the other hand, totally passive things done by the lung (eg. filtering the blood) and totally muscle-related activities (eg. CO₂ elimination for acid-base homeostasis) are listed here, so why not speech?

What is the real answer? There is nothing to guide us here, except textbook publications like the chapter by Docio-Fernandez & Garcia-Mateo from the Encyclopedia of Biometrics (2009), which define speech as a byproduct of the respiratory cycle. In short, speech is just expiration, which still has the primary role of clearing CO₂; it just also so happens to communicate complex thoughts. Ergo, if management of acid-base balance is considered a "non-respiratory function", then so should be speech.  Though this is not mentioned as one of the canonical non-respiratory functions of the lung in Nunn's, trainees are encouraged to try their luck with this point, but probably only after they have listed all the other functions.