To the Editor:--Amoiridis et al. are to be congratulated on redirecting attention to the potentially catastrophic dangers of an exaggerated lordotic posture in patients receiving epidural blockade. [1]Surgeons and anesthesiologists have been unwilling to acknowledge the harmful effects of prolonged hyperextension of the spine, though the evidence has been available for nearly half a century, [2,3]and cited repeatedly for more than 20 yr. [4–6]The cardiovascular mechanism of spinal venous infarction from hyperlordosis is quite clear. Hyperlordosis causes rotation of the liver and obstruction of the inferior vena cava in its intrahepatic course. Pressure increases within the alternative venous pathway (the azygos system) and is transmitted via the intercosto-vertebral veins and epidural venous plexuses to the intraspinal veins. [2,3].

In their opening paragraph, Amoiridis et al. list a total of four published reports of spinal cord infarction, including their own case, attributable to an imposed hyperlordotic posture, and one of these cases holds the key to appreciating the underlying pathophysiology. It is not appreciated generally that two of those cases ([4,8]in Amoiridis' reference list) were operated by the same surgeon, in the same year (1969), in the same surgical posture of imposed hyperlordosis, and under the same general anesthetic technique, except for the method of controlled arterial hypotension. In one case, high epidural blockade was induced with 36 ml 1.5% lidocaine and 1/200,000 epinephrine, and the intraoperative blood pressure was maintained at 80–90 mmHg; this case was published in the journal Anaesthesia under the title “Paraplegia following epidural analgesia.”[7]In the second case, arterial pressure was reduced with intravenous pentolinium, and no epidural was used. This second case is critical to resolving the etiologic issue of arterial versus venous infarction of the cord, because it is the only one to receive thorough macroscopic and microscopic examination of the spinal contents and to show all the signs of venous infarction. Both cases revealed a patent anterior spinal artery. After being rejected for publication by a widely read anesthesia journal in 1969, the second case was eventually published 5 yr later, by the New Zealand Medical Journal, a journal not widely read by Western anesthesiologists. [8].

The educational influence of these two cases and their publication history is reflected in subsequent listings in Science Citation Index, with the epidural case receiving 24 citations, most imputing causality to the epidural block. The second has received only four citations, two of which were by this writer, inveighing against the dangers of spinal hyperextension and the folly of ascribing them to a coincidental epidural blockade.

One more case must be added to the list of parapareses after spinal hyperextension. [9]In this recent report, hyperlordosis was maintained for 2 h 45 min, but nowhere in the discussion was this considered as a possible etiologic factor, because epidural analgesia was used as an adjunct to general anesthesia, and accidental subarachnoid cannulation was suspected. Once again, the publication process failed to acknowledge the unphysiologic posture of hyperlordosis as a possible cause of injury.

Finally, epidural analgesia will continue to attract obloquy unless appropriate monitoring standards are maintained to avoid associated but unrelated complications. Any analgesic medication of the spinal cord is a neurologic intervention that requires informed and appropriate neurologic surveillance on the part of the anesthesiologist. Any hope of rescuing a jeopardized spinal cord depends on prompt diagnosis and treatment within 8 h of injury. [10]In my opinion, a prudent course of management for postoperative epidural analgesia should assume that rare, unrelated neurologic dangers are always lurking, and that steps must be taken to provide early warning of their possible intrusion and their timely containment. This implies that early and regular neurologic examination of the lower limbs must take precedence over continuous intense epidural analgesia, and epidural infusions should be interrupted periodically while alternative pain relief is administered, to allow neurologic examination and confirmation of uncompromised neural function. [11].

Philip R. Bromage, M.B.B.S., F.F.A.R.C.S., F.R.C.P.(C.); Professor (post-retirement), McGill University Department of Anaesthesia, Montreal, Canada.

(Accepted for publication April 29, 1996.)

Amoiridis G, Wohrle JC, Langkafel M, Maiwurm D, Przuntek H: Spinal cord infarction after surgery in a patient in the hyperlordotic position. ANESTHESIOLOGY 1996; 84:228-30.
Bull GM: Postural proteinuria. Clin Sci 1949; 7:77-108.
Brigden W, Howarth S, Sharpey-Schafer EP: Postural changes in the peripheral blood flow of normal subjects with observations on vaso-vagal fainting reactions as a result of tilting, the lordotic posture, pregnancy and spinal anaesthesia. Clin Sci 1950; 9:79-90.
Bromage PR: Spinal thrombophlebitis after hypotensive anaesthesia (letter). N Z Med J 1974; 80:519-20.
Bromage PR: Epidural Analgesia. W.B. Saunders, Philadelphia, 1978, pp. 373-5, 676-9.
Bromage PR: Neurological Complications of Epidural and Spinal Techniques, Clinical Anaesthesiology. New Developments in Epidural and Spinal Drugs Administration. Edited by Van Aken H. Bailliere Tindall, Philadelphia, 1993, pp. 793-815.
Urquhart-Hay D: Paraplegia following epidural analgesia. Anaesthesia 1969; 24:461-70.
Urquhart-Hay D, Teague CA: Spinal thrombophlebitis after prostatectomy with hypotensive anaesthesia: Case report. N Z Med J 1974; 79:654-6.
Cheng ACK: Intended epidural anesthesia as possible cause of cauda equina syndrome. Anesth Analg 1994; 78:157-9.
Bracken MB, Holford TR: Effects of timing of methylprednisolone or naloxone administration on recovery of segmental and long-tract neurological function in NASCIS 2. J Neurosurg 1993; 79:500-7.
Bromage PR: Masked Mischief (editorial). Reg Anesth 1993; 18:143-4.