Lung liquid and protein exchange: the four inhomogeneities

William of Ockham, 14th-century scholastic philosopher at Oxford and Munich, emphasized the principle of economy, "pleurality is not to be supposed without necessity" (Ockham's razor). Necessity is the key word. In the modeling of steady-state lung liquid and protein exchange, the desire for simplicity has sometimes outweighed good judgment. In fact, we and others have shown that simple models do not work. It is necessary to include several forms of inhomogeneity. The air-filled lung shows regional (top to bottom) variations of mass, microvascular pressure, and perimicrovascular protein concentration. Normally, the small longitudinal (arterioles to venules) gradient of microvascular and perimicrovascular pressures is not a major concern, but in nonuniform disease processes, such as microembolism, longitudinal inhomogeneity, and parallel inhomogeneity are dominant. Multiple pores should also be considered a form of inhomogeneity. The effect on liquid and protein exchange, when plasma protein concentration or microvascular pressure change, can be readily explained using pore heterogeneity. The model I am currently using consists of a large number of discrete compartments (18), rather than a continuous distribution. We have recently identified a fifth inhomogeneity, which is that lung lymph flow might not always represent steady-state transvascular filtration because interstitial liquid may leak through the pleura or along the bronchovascular liquid cuffs into the mediastinum.