J. clin. Path., 23, Suppl. (Roy. Coll. Path.), 4, 132-142 Genesis of fat emboli J Clin Pathol: first published as 10.1136/jcp.s3-4.1.132 on 1 January 1970. Downloaded from A. J. WATSON University ofNewcastle upon Tyne In pathological terms fat embolism may be Some may be stuck fast, but there is evidence defined as the blockage of blood vessels by liquid that others continue to flow slowly through the fat globules. As Szabo (1970) emphasizes, on small vessels and recirculate, returning eventually page 123 of this issue, a clear distinction must be to the lungs (Scriba, 1880; Scuderi, 1953; Moser drawn between the histopathological findings and Wurnig, 1954). When the lungs are heavily and the much less common clinical syndromes of embolized, the globules passing through the fat embolism. Fat emboli in the lungs have been lung vessels into the systemic circulation may reported in a great variety of associations, but by become very numerous. Possibly this is the mostcopyright. far the most common and the most important serious consequence because of the multifocal association is with major fractures and accom- brain damage which results from cerebral panying soft tissue damage due to severe trauma. embolization (Scriba, 1880; Sevitt, 1962) and Controversy exists regarding the clinical signi- may in turn lead to secondary lung changes. But ficance of the emboli and there is even a sugges- some would give pride of place to the pulmonary tion that they are not essential for the changes changes and regard the cerebral damage as underlying the 'fat-embolism' syndrome. This is secondary to hypoxaemia (Peltier, 1967; Szabo,http://jcp.bmj.com/ one of the many anomalies with which this 1970). intriguing condition is beset. A wide divergence of opinion exists regarding the genesis of fat emboli. According to the clas- sical 'mechanical' explanation fat droplets are Aetiology of Fat Embolism set free from disrupted fat cells, enter torn venules at sites of injury or fracture, and are carried to INJURY TO ADIPOSE TISSUE AND TO BONE on September 29, 2021 by guest. Protected the lungs. Globules 10 ,t or more in diameter MARROW being unable to pass immediately through the Leaving aside trivial or low-grade pulmonary fat lung capillaries are retained as emboli. The embolism, a common incidental finding at alternative explanation, which also has many necropsy, much the most important association adherents, envisages fat emboli as derived partly is with fractures and soft tissue damage. This or entirely from plasma lipids, possibly after was recognized early and has been amply corro- increased mobilization from depot fat. If prophy- borated. Certain fractures, such as those of long lactic and therapeutic measures are to be well bones, pelvis, or spine, are more prone than founded, it is important to ascertain which others to be followed by severe pulmonary fat explanation is correct. embolism (Fig. 1). Further, the intensity of fat Fat emboli are initially arrested in the lungs embolization tends to be directly related to the which act as a filter to protect the systemic severity of injury, and especially to multiplicity circulation. They have certain peculiarities qua of fractures (Emson, 1958; Sevitt, 1962). Histo- emboli. Often they are very numerous: in the logical fat embolism can occur with remarkable most heavily embolized lungs thousands of rapidity and even very severe (grade 4) lung emboli per cubic millimetre of lung tissue may be involvement may be found occasionally in those seen. They are relatively small: their diameter in who die 'immediately' after injury. lung vessels ranges from about 10 to 100 ,u, and Jarring of the skeleton has been said to cause therefore they lodge in capillaries and arterioles. significant fat embolism but this is doubtful, Genesis offat emboli 133 J Clin Pathol: first published as 10.1136/jcp.s3-4.1.132 on 1 January 1970. Downloaded from r --m 1 Injury to the soft tissues seldom causes severe femur - 0 0 ? 9?0 0 fat embolism though cases have been reported shaft (Scully, 1956). Severe pulmonary and systemic embolism were seen in a 64-year-old man who other - long 00 0 00o 0 died 10 days after the second of two operations bones for aneurysm of the abdominal aorta. Personal observations confirm Grondahl's (1911) opinion - 10 To that severe burns cause only insignificant fat pelvis embolism. Acute pancreatic necrosis has been reported as a cause (Edmondson and Fields, 1942; Lynch, 1954), but a personal study of eight t 00 spine fatal cases failed to confirm this. In a fatal case of Weber-Christian disease the occurrence of fat embolism was attributed to the necrotizing ri bs 6 0 process in adipose tissue (Miller and Kritzler, 1943). A fatty liver may be regarded in the present context as a form of adipose tissue from skull - 0000 0 0 0000 00 which embolic fat might be liberated by trauma (Grondahl, 1911; Killian, 1931; Hallgren, Kers- tell, Rudenstam, and Svanborg, 1966), by necrosis femur 0 0 0 0 ooooooocD (MacMahon and Weiss, 1929; Tonge, Hurley, neck and Ferguson, 1969), or even spontaneously grade 4 grade 3 grade2 gradesl and O | (Cammermeyer and Gjessing, 1951; Hartroft and Ridout, 1951; Kent, 1955; Lynch, Raphael, and Dixon, 1959). Nevertheless, the liver is Fig. 1 Severity ofhistological pulmonaryv fat seldom if ever a significant source of embolic fat. embolism at necropsy related to fractures. Each point, or vertically joined series offpoints, represents one subject in a necropsy series of 43 INGRESS OF FREE LIQUID FAT INTO THE fracture cases. Certain fractures, particularly of BLOODSTREAM copyright. the shaft offemur, other long bones, andp7elvis, and Observations at operation or at necropsy confirm multiple fractures in general, tend to resul n that fat is at severe (grades 3 and 4) fat embolism. Mos (21132) liquid freed fracture sites where the of the isolatedfractures are associated witth lesser marrow is fatty. Some of the freed fat may be grades (0, 1, and 2) ofembolism. Some of*the derived from adjacent adipose tissue (vide infra). apparent anomalies are attributable either to very As originally envisaged, the embolic process short or to prolonged survival. required a pool of liquid fat, venules with gaping ends held open by attachment to bony canals http://jcp.bmj.com/ and although cases are reported frc )m time to (Gauss, 1916), and an impelling force provided time (Beitzke, 1912; Silverstein and E Ionzelman, by local increase in extravascular pressure due to 1940), it is well known that fractures can be over- haemorrhage at the fracture site (Flournoy, 1878). looked even at necropsy. Such an ovevrsight may Later it was found that the venules in bone also account for instances of fat embolism marrow are not incapable of collapsing (Urist attributed to convulsions (Meyer cand Teare, and Johnson, 1943; Whitson, 1951) and that 1945). fracture leads to a fall in intramedullary pressure on September 29, 2021 by guest. Protected It is surprising how rarely clinical fat embolism (Rehm, 1956 and 1957; Szab6, Jankovics, and results from surgical injury to bone marrow such Magyar, 1967). as intramedullary nailing (Scuderi, 1953). The process of embolism was illuminated by Embolization of fat and marrow ltissue may the work of Young and Griffith (1950) who follow external cardiac massage witih injury to studied its dynamics using a model representing the ribs and sternum, and in some insstances this a thin-walled blood vessel suspended in a fluid is an agonal if not actually a postmorrtem occur- medium. As the pressure within the 'vessel' fell rence (Sack and Wegener, 1968). Ain unusual below the external pressure there was a rapid cause of fat embolization is infarction,of the bone alternation of collapse and re-expansion; During marrow in cases of sickle-cell anaennia (Vance each re-expansion phase small plastic spherules and Fisher, 1941; Wyatt and OrrahcDod, 1952). within the supporting medium were drawn into In dogs, decompression sickness following the lumen of the 'vessel' through a slightly larger exposure to hyperbaric conditions may be hole previously made in its wall. Fuchsig, Brucke, accompanied by pulmonary fat anid marrow Blilmel, and Gottlob (1967) found that liquid fat embolism secondary to disruption otf the bone injected subcutaneously or intramuscularly was marrow by gas bubbles, but it is uncertain rapidly mobilized into the bloodstream following whether this is an important feature of decom- the induction of haemorrhagic shock. Presum- pression sickness in man (Bennison, C_atton, and ably this is determined by a critical reduction of Fryer, 1965). intravascular relative to extravascular pressure. A. J. Watson 134 J Clin Pathol: first published as 10.1136/jcp.s3-4.1.132 on 1 January 1970. Downloaded from An extreme fall in intravascular pressure follow- amount of oil which actually entered the venous ing cardiac arrest may also account for the high system. incidence of fat and bone-marrow embolism Estimates of the quantity of embolic fat in the after external cardiac massage (SackandWegener, lungs of patients who died after developing the 1968). fat-embolism syndrome have been attempted According to Fuchsig et al (1967) the main although normal lungs contain adipose tissue route of ingress for the fat was through the and other sources of lipid, and the amount varies lymphatic system into the innominate vein, but widely. The values range from about 20 ml the evidence for this is tenuous. If any fat reaches (Killian, 1931), through 36 ml (Armin and the bloodstream via the lymphatics the amount is Grant, 1951) to the probable overestimates of likely to be unimportant (Szabo, Serenyi, and Elting and Martin (1925). Assuming that 20% Kocsafr, 1963). The idea is of historical interest of the embolic fat may have passed into the in that thoracic duct drainage was once advocated systemic circulation, then the total quantity as a prophylactic measure.