Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet.

Local compression is known to damage peripheral nerves. Severe compression may actually crush the fibres and lead to Wallerian degeneration; the excitability of the nerve distal to the lesion is lost and recovery may take many months (Seddon, 1943; Denny-Brown & Brenner, 1944 a). Mild compression produces a physiological block which is reversed as soon as the pressure is released. On the basis of their studies with pneumatic cuffs or clamps in man, Lewis, Pickering & Rothschild (1931) concluded that this physiological block was due to local asphyxia. There is a lesion, distinct from either of the two described above, which follows compression of intermediate severity. In such a lesion there is a local conduction block with preservation of distal excitability; this may take several weeks to recover (Erb, 1876; Dejerine & Bernheim, 1899; Seddon, 1943). Using a tourniquet or a spring clip to produce compression, Denny-Brown & Brenner (1944a, b), found in cats that the anatomical basis for this was demyelination with preservation of axonal continuity at the site of compression. Denny-Brown & Brenner thought that this lesion, like that produced by Lewis et al. (1931), was due to local ischaemia of the nerve and that direct mechanical pressure was an unlikely explanation. In support of this they cited the experiments of Grundfest (1936) who had shown that enormous pressures were necessary to abolish conduction in excised frog nerve enclosed in an oxygenated pressure chamber. During the course of recent experiments in which a tourniquet was used to produce long-standingconduction block in the peripheral nerves of baboons (Fowler, Danta & Gilliatt, 1972), we have found a characteristic lesion which precedes paranodal demyelination. The anatomical features and distribution of this lesion suggest that the damage to the nerve fibres is a direct result of the applied pressure, and not a consequence of secondary ischaemia. A brief preliminary account of our findings has been reported (Ochoa, Danta, Fowler & Gilliatt, 1971).