Dysbaric Osteonecrosis: Aetiology and Pathogenesis

In terms of clinical pathology, DON is simply one in a long list of causes of aseptic necrosis, but perhaps one of the most fascinating. The most common cause of aseptic necrosis of the femoral head is fracture of the neck of the femur. The necrotic lesions of high-dose steroid therapy, even though multiple and bilateral, often involve the articular surface of the knee and ankle joints, virtually never seen with DON. This variation in distribution suggests that the pathogenesis may be different, even if the pathological features are identical. Osteonecrosis is also frequently reported in association with those diseases in which there is some disturbance of fat metabolism, e.g. diabetes mellitus, pancreatitis, alcoholism and cirrhosis, Gaucher’s disease and hyperlipidaemia. Trauma and steroid administration are the most common associations. Aseptic osteonecrosis may occur without any known risk factors (idiopathic aseptic necrosis). Certain specific isolated-site bone necrosis disorders, such as Legg-Calvé-Perthes disease, may be associated with specific systemic or anatomical abnormalities (Table 14.3).


Table 14.3 Some causes of aseptic necrosis

  • Decompression sickness or dysbaric exposure.
  • Trauma (e.g. fractured neck of femur, dislocated hip and unrelated fractures).
    Steroids (Cushing’s syndrome and steroid therapy).
  • Collagen diseases (e.g. lupus erythematosus, rheumatoid arthritis, polyarteritis nodosa).
  • Occlusive vascular disease.
  • Diabetes mellitus.
  • Hyperlipidaemia.
  • Liver disease (fatty liver, hepatitis, carbon tetrachloride poisoning).
  • Alcoholism.
  • Pancreatitis.
  • Gaucher’s disease.
  • Gout.
  • Haemophilia.
  • Polycythaemia/marrow hyperplasia.
  • Haemoglobinopathies (especially sickle cell).
  • Sarcoidosis.
  • Charcot joint.
  • Specific bone necrosis disorders (Legg-Calvé-Perthes, Kienbock’s, Freiberg’s and Kohler’s diseases).
  • Radiotherapy.

It has been postulated that many of these conditions may be associated with fat emboli and these emboli obstruct end arteries in rigid haversian canals of bone, leading to avascular osteonecrosis. These fat emboli may arise from a fatty liver, coalescence of plasma lipoproteins, disruption of bone marrow or other fat tissue or a combination of the foregoing mechanisms. Enhanced coagulability may add to blood vessel obstruction.

The exact mechanism leading to bone necrosis in association with hyperbaric exposure has not been fully elucidated. The most widely held belief is that it results from the decompression phase and represents a delayed or long-term manifestation of DCS (see Chapter 12). There is a definite relationship between DON and exposure to inadequate decompression, experimental diving and clinical DCS.

There are, however, numerous variations on this basic concept. One theory is that the infarction is caused by arterial gas emboli produced during decompression. Certainly, ‘silent’ bubbles can be detected by Doppler techniques during clinically apparently safe decompression schedules. However, several series indicate a relationship with musculoskeletal DCS or total DCS, rather than specifically neurological or serious DCS, and it is the latter that are more likely to be associated with intra-arterial bubbles.

Others propose that the fat in bone marrow takes up large amounts of nitrogen during longer pressure exposures. During or after decompression, gas is liberated from the fat, and expansion with decompression increases intramedullary pressure, thus compromising blood flow within non-compliant bone cavities5. Prompt recompression may prevent later deterioration because there is probably a critical period of bone ischaemia after which pathological changes become irreversible. Osteocytes are known to die after about 4 hours of anoxia. Some affected areas may spontaneously recover, whereas others progress to the typical necrotic lesions.

Bubbles have been found post mortem in the large venous sinusoids in animal experiments with DCS, and they may well have obstructed venous outflow from marrow, leading to areas of infarction. Bubble formation within bony lacunae and subsequent destruction of osteocytes are also possible following decompression.

Changes secondary to intravascular bubbles, whether arterial or venous, such as platelet aggregation and intravascular coagulation, may cause further vascular obstruction (Figure 14.1). Release of fat, thromboplastin and vaso-active substances could also trigger disseminated intravascular coagulation and exacerbate DON2. This model is supported by the post-dive observation of increased platelet adhesiveness and decreased platelet count in volunteers who display higher intravascular bubble counts on Doppler imaging6.

Suggested pathogenesis of dysbaric osteonecrosis.
Figure 14.1 Suggested pathogenesis of dysbaric osteonecrosis. (From Kawashima M, Tamura H, Noro Y, et al. Pathogenesis and prevention of dysbaric osteonecrosis. In: Proceedings of the 12th Meeting of the United States-Japan Cooperative Program in Natural Resources (UJNR) Panel on Diving Physiology, Washington DC, July 13–14. Silver Spring, Maryland: National Undersea Research Program; 1993.)

It is possible that a number of factors may combine to produce necrosis in a given situation and that the aetiology is complex and multifactorial. Experimental evidence is available to suggest that both intravascular and extravascular aetiologies are consistent with the bone pathology, but a direct cause-and-effect relationship has not been proven. Asymptomatic or ‘silent’ bubbles during or after decompression are incriminated in those divers who have had neither DCS nor exposure to hazardous diving practices.

All embolism theories (gas, fat or other) do not adequately explain why other tissues do not appear to be embolized and why the femur and upper end of the humerus are particularly affected.

Oxygen toxicity is another possible cause of DON. Several mechanisms have been postulated. One suggests that the local vasospastic reaction to high oxygen pressures leads to ischaemia. High oxygen pressures have been shown to cause swelling of fat cells, which may produce increased intramedullary pressure and ischaemia or, if insufficient to obstruct blood flow completely, could inhibit the clearance of gas from the marrow during decompression. Given the low rates of DON in those who practice oxygen decompression techniques, this seems an unlikely cause of DON.

An osmotic aetiology has also been suggested, incriminating the movement of water into or out of the bone. Rapid pressure changes during compression are associated with large gas gradients because the intravascular partial pressure of all inspired gases is transiently much higher than in the tissues. Thus, a gradient exists across the capillary wall, and water would then move into the vascular compartment. Expansion of the intravascular space within the rigid bone structure may to lead to local bone ischaemia. It has even been suggested that the absolute pressure within the medulla of bone may be transiently lower than that outside the cortex during a rapid compression, and that this alone could promote venous stasis and bone necrosis. It is unclear how this apparent violation of Pascal’s Principle could occur, but such transient differences in pressure have been reported, with no clear explanation.

Dysbaric osteonecrosis is thought to be a long-term effect of inadequate decompression.

Various animal models have been developed to study the aetiology of DON because of the obvious difficulties in early detection and monitoring of such a capricious and chronic disease. Much research thus involves the experimental induction of bone necrosis in animals such as guinea pigs and mice, but it is difficult to be convinced that these lesions are strictly comparable to those of divers and caisson workers. Studies in larger animals such as sheep that have a large fatty marrow compartment in long bones similar to humans have been more successful. These studies in sheep and human post-mortem studies tend to support raised intramedullary pressure combined with hypercoagulability mechanisms.

Any theory must account for the following observations:

  • Dysbaric osteonecrosis may follow a single exposure to pressure.
  • Although there appears to be a relationship between DCS and DON, not all divers with DON have a history of DCS.
  • Not all divers who have DCS develop DON.
  • Not all divers at high risk develop DON.

The development of effective strategies for prevention and treatment depends on further research elucidating the precise pathophysiological mechanisms involved.