An unusual complication of emergency tracheal intubation

Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

C A S E R E P O RT

Serious spinal cord injury due to haematomyelia caused by spinal anaesthesia in a patient treated with low-dose heparin J. D. Greaves Newcastle School of Anaesthesia, Royal Victoria Infirmary, Queen Victoria Road, Newcastle Upon Tyne NE1 4LP, UK Summary

Failed spinal anaesthesia for left total hip arthroplasty was followed postoperatively by dense motor paralysis and sensory deficit in the right leg. The patient had received a dose of subcutaneous heparin 1 h before the spinal anaesthetic was attempted. She died of pulmonary embolism on the ninth postoperative day. At autopsy extensive haematomyelia was found in relation to the needle track. Keywords Anaesthetic techniques, regional; spinal. Complications; haematoma, spinal. Blood; coagulation, heparin. Embolism; thrombo-embolism. ...................................................................................... Correspondence to: Dr J. David Greaves Accepted: 15 October 1996

Case report

An overweight woman of 66 years (height 163 cm weight 87 kg) presented for a left Charnley total hip replacement. She had a history of deep vein thrombosis some 10 years previously, apart from which and notwithstanding her weight, she had remained in good general health. Her only medication was fenbufen for arthritis, which she had not taken for 1 week prior to the operation and co-codamol as required for pain. The patient was frankly advised of the danger of hip surgery but was keen to proceed because her arthritis was severely disabling and her pain constant. She had been waiting for surgery for some months and had been able to lose 8 kg in weight. Pre-operative laboratory investigations showed a haemoglobin of 12.9 g.dlÿ1 and platelets of 170 000 × 109.lÿ1. The patient’s clotting function was not investigated. Despite the practical difficulties caused by her weight, regional analgesia was chosen. Premedication was with midazolam 5 mg intramuscularly and low-dose heparin was prescribed (5000 i.u. calcium heparin subcutaneously twice daily) the first dose being given 1 h pre-operatively. Spinal anaesthesia was attempted in the left lateral position with a 26G needle. It was extremely difficult to palpate the bony landmarks of the back but the

150

subarachnoid space was located at the first attempt in what was believed to be the L3–4 interspace and free flow of cerebrospinal fluid was observed. On injection of the local anaesthetic the patient complained of intense pain, described as sharp and stabbing in nature, down the left leg. The subarachnoid injection was abandoned and the needle withdrawn. A total of 0.3 ml of heavy bupivacaine 0.5% in 12% dextrose had been injected. The situation was discussed with the patient who was now pain free and it was decided to proceed with general anaesthesia. This was induced with thiopentone 250 mg and controlled ventilation with nitrous oxide, oxygen 30% and isoflurane 0.5% was facilitated with vecuronium. A total of 12 mg of morphine was administered during surgery. The operation was uneventful and the systolic blood pressure remained in the range 110–140 mmHg throughout. The duration of surgery was 90 min. The patient complained of moderate pain at the operative site and was given a further 5 mg of morphine intravenously in the recovery unit. Her immediate postoperative progress was otherwise unremarkable. The following morning when the patient awoke she complained that her right leg was numb and weak. Neurological examination confirmed this and neurological and neurosurgical opinions were sought.

Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

On physical examination a right-sided unilateral sensory deficit to all modalities was found extending from T12 to S3. This was accompanied by a dense flaccid monoplegia. There was complete absence of flexion and extension at hip and knee and the only remaining motor power in the right leg was internal rotation at the hip and extension of the big toe. The lesion was exclusively on the right side. The specialist opinion of a neurologist and a neurosurgeon was that the unilateral nature of the lesion precluded a cord compression aetiology and exploration of the cord was not advised. The patient’s clotting function was investigated and was found to be normal. (platelets 190 000 × 109.lÿ1 prothrombin time 13 s, APTT 36 s). Subcutaneous heparin was continued in view of the risk of deep vein thrombosis. Over the following 9 days there was some improvement in motor function but the sensory deficit was unchanged. Mobilisation was begun with difficulty. On the tenth post-operative day the patient was found dead in bed. Post-mortem findings

The cause of death was massive pulmonary embolus. A careful examination was made of the tissues of the back in the region of the lumbar puncture. A needle track was identified entering the spinal canal in the T12 –L1 interspace. This penetrated the conus medullaris of the spinal cord and was associated with haemorrhage within the substance of the cord which extended in a proximal direction for 4.5 cm on the right-hand side of the midline. The haemorrhage was confined to the substance of the cord and no subdural or extradural clot was seen. Discussion

The most dreaded complication of subarachnoid anaesthesia is spinal cord injury. Its incidence is very low but the catastrophic effect upon the patient has ensured that it has had a major effect upon the practice of anaesthesia. There have been reports of serious complications with spinal anaesthesia since shortly after it was first described [1, 2]. The Woolley and Roe case and its aftermath led a generation of British anaesthetists to eschew spinal anaesthesia [3] and worldwide there was a reluctance to use the technique until the publications of Dripps & Vandam [4] and Phillips et al. [5] each of which reported careful follow up of 10 000 cases. Those individuals here and abroad who continued its use stressed its safety when used with meticulous attention to detail and they emphasised that most symptoms initially attributed to the use of the technique were on careful evaluation found to be unrelated [6–8]. Nevertheless, sporadic cases of serious neurological Q 1997 Blackwell Science Ltd

injuries have continued to be reported [9–25]. The incidence of such problems is not known and the low estimated incidence of one serious neurological sequel per 11 000 administrations makes it difficult to obtain epidemiological data [4, 5, 8, 26–38]. Some cases are probably not recorded in the literature. A recent questionnairebased survey revealed two cases of apparently unreported paraplegia [39]. For an analysis of the issues surrounding spinal cord damage by needles the reader is referred to the comprehensive reviews by Usubiaga [40] and Thorsen [8]. The incidence of spinal haematoma following epidural or spinal anaesthesia has recently been estimated to be less than 1:100 000 administrations [41]. The present case is exceptional in that the patient’s death from pulmonary embolism allowed a full evaluation of the spinal cord lesion in the post-mortem room whilst the acute appearances were still evident. Evaluation of the extent of the lesion before death by use of nuclear magnetic resonance imaging was not possible at the time that this case occurred. At post-mortem examination the needle track could be clearly identified and was seen to be associated with the haematomyelia. The space chosen was thought to be L2–3 and was recorded as such. The spinal cord was of normal length and in order to have entered it at the position described the space chosen must have been in the region of T12 –L1. The difficulty of correctly identifying the surface marking of an intervertebral space, particularly in obese patients, is well recognised [42]. In a study on cadavers it was shown that the usual method of locating the L3–4 interspace was inaccurate and could lead to cephalad misplacement of needles [43]. It is probable that the needle was misplaced at least two spaces cephalad in this patient. The location of the subarachnoid space was uncomplicated. A 26G needle with a diamond point tip was used. Cerebrospinal fluid appeared slowly at the hub of the needle as is usual with fine-gauge needles. Entering the subarachnoid space was painless but the pain on injection was immediate and localised to the area where the neurological deficit was subsequently evident. The experience of such pain is rare and should lead to immediate removal of the needle. Intramedullary injections are always followed by sharp pain [42]. The association of pain on injection and subsequent haematomyelia has been noted previously [44, 8]. The anaesthetist is faced with a dilemma. Extradural and subarachnoid anaesthesia have been shown to be associated with lower mortality and also a decreased incidence of deep vein thrombosis when used for major hip surgery [45–50]. Many surgeons prescribe prophylactic subcutaneous heparin which also reduces the incidence of postoperative thrombo-embolism [51]. It is not known 151

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

whether low-dose heparin thromboprophylaxis is associated with an increased risk of spinal haematoma, or whether heparin or regional anaesthesia is the more effective in reducing postoperative venous thrombosis. In the absence of such information many anaesthetists are willing to proceed with spinal or epidural anaesthesia in patients receiving ‘low-dose’ heparin [52]. This course of action has received guarded support [53–59] but some writers have called attention to the possible role of low-dose heparin in the aetiology of spinal haematoma and advised against the use of regional techniques when heparin thromboprophylaxis is being employed [17, 60–63]. Monitoring of heparin levels following a single dose of 5000 units subcutaneously has revealed that they may be in the anticoagulant range in some patients for as long as 4 h and advice has been given that either the dose be administered more than 4 h prior to surgery or the activated partial thromboplastin time be checked prior to the regional anaesthetic [64, 65]. There is no doubt in my mind that this patient’s ultimately fatal complications resulted from a spinal cord injury caused by the needle intended for the subarachnoid injection. It is unusual to be able to report conclusively on the spinal cord injury in such a case. The decision to use spinal anaesthesia for this operation was reinforced by the high risk of deep vein thrombosis and pulmonary embolism following surgery. It is ironic that such was indeed the cause of the patient’s death. What is the anaesthetist to do? The risks of the use of regional spinal or epidural anaesthesia in patients receiving heparin thromboprophylaxis have recently been reviewed [41]. It is clear that the current state of knowledge of the relative risks is inadequate for clear guidance to be provided for anaesthetists. The epidemiological data relating to the incidence of spinal haematoma suggest that with [66] or without heparin thromboprophylaxis the incidence of spinal haematoma is very small. The current case may be the first report of such an association. In the absence of absolute guidelines it is reasonable for the anaesthetist to follow rational practice. Where heparin thromboprophylaxis has been started the regional anaesthesia should be delayed for 4 h after the heparin injection (12 h in the case of low-molecular-weight preparations [67]). Where possible, if regional anaesthesia is judged to be of benefit to the patient, the first dose of heparin should be postponed until after the block has been instituted. This is now my practice. References 1 Dandois R. Accidents cerebro-spineaux, tardifs et prolonge´s apre`s cocainisation de la moe`lle. Journale de

152

2

3

4 5

6

7

8

9

10 11 12 13

14

15

16 17

18 19 20

Chirurgerie et Annales de Societe´ Belge de Chirurgerie 1901; 1: 282. Ferguson FR, Watkins KH. Paralysis of the bladder and assosciated neurological sequelae of spinal anaesthesia (cauda equina syndrome). British Journal of Surgery 1938; 25: 735–52. Cope RW. The Woolley and Roe Case. Woolley and Roe versus Ministry of Health and others. Anaesthesia 1954; 9: 249–70. Dripps RD, Vandam LD. Hazards of lumbar puncture. Journal of the American Medical Association 1951; 147: 1118. Phillips OC, Ebner H, Nelson AT, Black M. Neurologic complications following spinal anesthesia with lidocaine: a prospective review of 10 440 cases. Anesthesiology 1969; 30: 284–9. Kennedy F, Effron AS, Perry G. The grave spinal cord paralyses caused by spinal anesthesia. Surgery, Gynecology and Obstetrics 1950; 91: 385–8. Nicholson MJ, Eversole UH. Neurological complications of spinal anesthesia. Journal of the American Medical Association 1946; 132: 679–85. Thorsen G. Neurological complications after spinal anesthesia and results for 2493 follow up cases. Acta Chirugica Scandinavica (supplement) 1947; 121: 1. Bergner RP, Roseman E, Johnson H, Smith WR. Severe neurological complications following spinal anesthesia: a report of six cases. Anesthesiology 1951; 12: 717–27. Michelsen JJ. Neurological manifestations following spinal anaesthesia. Neurology 1952; 2: 255. Payne JP, Bergentz SE. Paraplegia following spinal analgesia. Lancet 1956; 1: 666–7. Davies A, Soloman B, Levene A. Paraplegia following epidural anaesthesia. British Medical Journal 1958; 2: 654–8. King OJ, Glas WW. Spinal subarachnoid haemorrhage following lumbar puncture Archives of Surgery 1960; 80: 574–7. Roscoe MJW, Barington TW. Acute spinal subdural haematoma- a case report and review of the literature. Spine 1984; 9: 672–5. Skouen JS, Wainapel SF, Willock MM. Paraplegia following epidural anesthesia. A case report and a review of the literature. Acta Neurologica Scandinavica 1985; 72: 437–43. Adriani J, Naragi M. Paraplegia associated with epidural anesthesia. Southern Medical Journal 1986; 79: 1350–5. Owens EL, Kasten GW, Hessel EA. Spinal subarachnoid hematomata after lumbar puncture and heparinisation: a case report, review of the literature, and discussion of anesthetic implications. Anesthesia and Analgesia 1986; 65: 1201–7. Schou H, Hole P. Neurologic deficit following spinal anaesthesia. Acta Anaesthesiologica Belgica 1987; 38: 241–3. Barker GL. Spinal subdural haematoma following spinal anaesthesia. Anaesthesia 1988; 43: 664–5. Gustafsson H, Rutberg H, Bengtsson M. Spinal hematomata following epidural analgesia. Anaesthesia 1988; 43: 220–2.

Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

21 Bredtmann RD, Wright J, Weisflog M. A space occupying hemorrhage after spinal anaesthesia in the presence of an undiagnosed small spinal tumour. Regional Anaesthesie 1989; 12: 38–40. 22 Bachman MB, Michaelis G, Biscoping J, Kleinstein J. Neurologic complication following spinal anesthesia for manual detatchment of placenta. GeburtshilfeFrauenheilkunde 1990; 50: 231–3. 23 Parnass SM, Schmidt KJ. Adverse effects of spinal and epidural anesthesia. Drug Safety 1990; 5: 179–94. 24 Pryle BJ, Carter JA, Cadoux-Hudson T. Delayed paraplegia following spinal anaesthesia. Anaesthesia 1996; 51: 263–5. 25 Sage DJ. Epidurals, spinals and bleeding disorders in pregnancy: A review. Anaesthesia and Intensive Care 1990; 18: 319–26. 26 Vandam LD, Dripps RD. Long term follow up of patients who received 10,098 spinal anesthetics. III. Neurological disease incident to traumatic lumbar puncture during spinal anesthesia. Journal of the American Medical Association 1960; 172: 1483–7. 27 Brown S. Fractional segmental spinal anesthesia in poor risk surgical patients: report of 600 cases. Anesthesiology 1951; 12: 16–28. 28 Kane RE. Neurologic deficits following epidural or spinal anesthesia. Anesthesia and Analgesia 1981; 60: 150–61. 29 Sadove MS, Levin MJ. Neurological complications of spinal anaesthesia. A statistical study of 10,000 consecutive cases. Illinois Medical Journal 1954; 105: 169–74. 30 Sadove MS, Levin MJ, Rant-Sejdinaj I. Neurologic complications of spinal anaesthesia. Canadian Anaesthetists’ Society Journal 1961; 38: 405–16. 31 Moore DC, Bridenbaugh LD. Spinalsubarachnoid block. A review of 11,574 cases. Journal of the American Medical Association 1966; 195: 907–12. 32 Arner O. Complications following spinal anaesthesia Acta Chirurgica Scandinavica (Supplement) 1952; 167: 1. 33 Almgard LE. Complications of epidural and spinal anaesthesia. A comparative study. Acta Chirurgica Scandinavica 1959; 117: 433. 34 Marinacci AA. Neurological aspects of complications of spinal anesthesia. Bulletin of Los Angeles Neurological Society 1960; 25: 170–92. 35 Greene NM. Neurological sequelae of spinal anesthesia. Anesthesiology 1961; 22: 682–95. 36 Vandam L. Neurological sequelae of spinal and epidural anesthesia. International Anesthesiology Clinics 1986; 24: 231–55. 37 Wildsmith J, Lee JA. Neurological sequelae of spinal anaesthesia British Journal Anaesthesia 1989; 63: 505–7. 38 Keating V. Anesthetic Accidents (2nd edn). Chicago: Year Book, 1961. 39 Owens EL, Kasten GW, Hessel EA. Spinal subarachnoid hematomata after lumbar puncture and heparinisation: a case report, review of the literature, and discussion of anesthetic implications. Anesthesia and Analgesia 1986; 65: 1201–7.

Q 1997 Blackwell Science Ltd

40 Usubiaga J. Neurological complications following epidural anesthesia. International Anesthesiology Clinics; 13. Boston: Little Brown and Co, 1975, 1–153. 41 Bullingham A, Strunin L. Prevention of postoperative venous thromboembolism. British Journal of Anaesthesia 1995; 75: 622–30. 42 Lee JA, Atkinson RS Sir Robert Macintosh’s lumbar puncture and spinal analgesia: intradural and extradural Edinburgh: Churchill Livingstone, 1978. 43 Ievins F. Accuracy of placement of extradural needles in the L3–L4 interspace. British Journal of Anaesthesia 1991; 66: 381–2. ¨ eber dauerschadigungen nach 44 Franke M. U lumbalanaethesie mit Novocain suprerenin-losung. Deutsche Zeitschrift fur Chirurgerie 1927; 202: 262. 45 Modig J, Borg T, Karlstrom G, Maripuu E, Sahlstedt B. Thromboembolism after total hip replacement: role of epidural and general anesthesia. Anesthesia and Analgesia 1983; 62: 174–80. 46 Thorburn J, Loudon JR, Vallance R. Spinal and general anaesthesia in total hip replacement: frequency of deep vein thrombosis. British Journal of Anaesthesia 1980; 52: 1117–21. 47 Scott NB, Kehlet H. Regional anaesthesia and surgical morbidity. British Journal of Surgery 1988; 75: 299–304. 48 Katz J, Aidinis SJ. Complications of spinal and epidural anesthesia. Journal of Bone and Joint Surgery 1980; 62: 1219–22. 49 Editorial. Prevention of post-operative deep vein thrombosis and pulmonary embolism. Annales de Chirugerie 1992; 46: 290–6. 50 Prins M, Hirsh J. A comparison of general anaesthesia and regional anaesthesia as a risk factor for deep venous thrombosis following hip surgery: a critical review. Thrombosis and Haemostasis 1990; 64: 497–500. 51 Collins R, Scrimgeour A, Yusuf S, Petro R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. New England Journal of Medicine 1988; 318: 1162–73. 52 Schwander D, Bachman F. Heparin and spinal or epidural anaesthesia: decision analysis. Annales Francais d’ Anaesthesie et de Reanimation 1991; 10: 284–96. 53 Lowson SM, Goodchild CS, Bentgsson B. Low dose heparin therapy and spinal anaesthesia. Anaesthesia 1989; 44: 67–8. 54 Parnass SM, Rothenberg DM, Fischer RL, Ivanovitch AD. Spinal anesthesia and mini dose heparin. Journal of the American Medical Association 1990; 263: 1496. 55 Matzsch T, Bergquist D, Johansson A. An inquiry shows minimal risk of hemorrhage resulting from thrombosis prevention in regional anaesthesia. Lakartidningen 89: 4028–30. 56 Mattingly SB, Stanton-Hicks M. Low dose heparin therapy and spinal anesthesia. Journal of the American Medical Association 1981; 246: 886. 57 Odoom JA, Sih IL. Epidural anesthesia and anticoagulant therapy. Experience with one thousand cases of continuous epidurals. Anaesthesia 1983; 38: 254–9.

153

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

58 Tryba M. Haemostatic requirements for the performance of regional anaesthesia Regional Anaesthesia 1989; 12: 127–31. 59 Wille-Jorgensen P, Jorgensen NL, Rasmussen LS. Lumbar regional anaesthesia and prophylactic anticoagulant therapy. Anaesthesia 1991; 46: 623–7. 60 Dupeyrat A, Dequire PM, Merouani A, Moullier P, Eid G. Subarachnoid Hematomata and spinal anesthesia. Annales Francaises d’ Anesthesie et de Reanimation 1990; 9: 560–2. 61 Onishchuk JL, Carlsson C. Epidural hematoma associated with epidural anesthesia: complications of anticoagulant therapy. Anesthesiology 1992; 77: 1221–3. 62 Sternlo JE, Hybinette CH. Spinal subdural bleeding after attempted epidural and subsequent spinal anaesthesia in a patient on thromboprophylaxis with low molecular weight heparin. Acta Anaesthesiologica Scandinavica 1995; 39: 557–9.

63 Bergquist D, Lindblad B, Matz CHT. Risks of combining low molecular weight heparin for thromboprophylaxis and epidural or spinal anaesthesia. Seminars in Thrombosis and Hemostasis 19 (Supplement) 1993; 1: 147–51. 64 Brozovic M, Stirling Y, Abbosh J. Plasma heparin levels after low dose subcutaneous heparin in patients undergoing hip replacement. British Journal of Haematology 1975; 31: 461–6. 65 Cooke ED, Lloyd MJ, Bowcock SA, Pilcher MF. Monitoring during low dose heparin prophylaxis. New England Journal of Medicine 1976; 294: 1066–7. 66 Berquist D, Lindblad B, Matzsch T. Low molecular weight heparin for thromboprophylaxis and epidural/spinal anaesthesia- is there a risk? Acta Anaessthesiologica Scandinavica 1992; 36: 605–9.

C A S E R E P O RT

An unusual complication of emergency tracheal intubation G. C. Ooi,1 M. G. Irwin,2 L. K. Lam3 and S. W. K. Cheng3 1 Department of Diagnostic Radiology, 2 Department of Anaesthesiology, and 3 Department of Surgery, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong Summary

A 68-year-old woman presented with dyspnoea having undergone a short period of tracheal intubation during an episode of congestive cardiac failure 10 days earlier. On the fourth day of admission, she suddenly developed acute respiratory distress followed by a cardiac arrest and was found to have an enlarging neck mass. Following intubation and resuscitation, computerised tomography was carried out and surgical exploration revealed a retropharyngeal abscess and a pseudo-aneurysm of the common carotid artery resulting from pharyngeal rupture which most probably occurred as a complication of the original tracheal intubation. Keywords Complications; intubation, tracheal, retropharyngeal abscess, pseudo-aneurysm. Imaging; angiography, computerised tomography. ...................................................................................... Correspondence to: Dr M. G. Irwin Accepted: 12 September 1996

Perforation of the pharynx or oesophagus during tracheal intubation is an uncommon complication which may occur when intubation is difficult and conditions are unfavourable, such as during cardiopulmonary resuscitation [1–5]. The consequences of pharyngo-oesophageal perforations include retropharyngeal abscess, mediastinitis, respiratory distress, septic shock, subcutaneous emphysema and death [1, 5]. We are not aware of a previous report of septic erosion of the common carotid artery 154

by a retropharyngeal abscess associated with tracheal intubation. Case report

A 68-year-old Chinese woman with a history of ischaemic heart disease was admitted to hospital with shortness of breath 2 days after discharge from another institution where her trachea had been intubated 10 days previously Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

Figure 1 Chest X-ray showing a paratracheal mass on the right side of the neck displacing the trachea (arrows) to the left and extending into the superior mediastinum.

during treatment for congestive cardiac failure. On the fourth day of the current admission, she suddenly developed acute respiratory distress followed by a cardiac arrest. She was resuscitated but remained unconscious and was transferred to our hospital. On admission, 500 ml of blood was aspirated from the tracheal tube but direct laryngoscopy was normal. Extensive right cerebral hemisphere infarction was seen on computerised tomography of the brain. It was also noted that the patient had an enlarging neck mass and right arm swelling. The chest X-ray showed that the trachea was displaced to the left by a soft tissue mass in the superior mediastinum (Fig. 1). Review of chest X-rays taken before the cardiac arrest showed that this abnormality, although less obvious, was already present. Computerised tomography of her neck showed a large retropharyngeal mass of heterogeneous density measured as 65 Hounsfield Units (circled area 1, Fig. 2), occluding the hypopharynx and extending distally to the right parapharyngeal space to displace the larynx and trachea anteriorly and to the left. The mass diminished in size at the aortic arch. After intravenous administration of iodine-based radiographic contrast, extravasation was seen near the origin of the right carotid sheath (Fig. 3). An anomalous right Q 1997 Blackwell Science Ltd

subclavian artery was also noted, traversing posterior to the oesophagus at the level of the manubrium (Fig. 3). This was confirmed by an arch aortogram which also demonstrated a pseudo-aneurysm of the right common carotid artery as it crossed the anomalous right subclavian artery. The pseudo-aneurysm was sited too superiorly for accidental puncture from subclavian vein cannulation. At surgical exploration, a tense pseudo-aneurysm was found in the caudal two-thirds of the neck extending over the superior mediastinum. Five hundred millilitres of foul-smelling clots were evacuated and a necrotic hole was seen at the anteromedial wall of the proximal right common carotid artery. There was also a 1-cm transverse laceration of the right posterior pharyngeal wall at the level of the lower thyroid cartilage. The pharyngeal laceration was incorporated into a drain pharyngostomy and the common carotid artery was divided and ligated. After surgery the patient never regained consciousness, requiring enteral feeding to this day. Discussion

It is likely that the patient suffered a pharyngeal perforation during attempts at tracheal intubation 10 days before 155

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

Figure 2 Plain computerised tomography scan through the region of the lower larynx showing a heterogeneous retropharyngeal mass

(circle labelled 1) causing anterolateral displacement of the larynx (arrowheads) and oesophagus (arrows).

admission. This gave rise to a retropharyngeal abscess which enlarged and tracked down the neck to erode into the proximal right common carotid artery. The enlarging pseudo-aneurysm, in turn, caused respiratory embarrassment which resulted in cardiac arrest. In the pre-antibiotic era, vascular complications occurring as a result of deep neck infections were not uncommon [6]. Since then, the incidence of such complications has fallen considerably although there have been a few reports of carotid pseudo-aneurysms secondary to septic erosion by oropharyngeal infections [7–14]. Owing to the close proximity of cranial nerves IX to XII and the sympathetic plexus to the parapharyngeal space, Horner’s syndrome and cranial nerve neuropathies have been reported in association with deep neck infections [8–10]. Repeated minor nasopharyngeal bleeds before the final rupture of the pseudo-aneurysm with resulting fatal haemorrhage are also described [9]. The internal carotid artery is the most common vessel involved (in up to 62% of cases) as it lies closest to the pharyngeal wall [6–9, 10]. The location of carotid rupture in our patient is 156

considerably more caudal in the neck, at the proximal third of the common carotid artery. This may be explained by the site of pharyngeal rupture at the lower hypopharynx and the time lapse before presentation with the pseudoaneurysm, allowing the abscess to track distally to involve the common carotid artery. Most patients with retropharyngeal abscesses associated with tracheal intubation present with pyrexia, pain, swelling or dysphagia within 48 h, although later presentation at 2 weeks has been reported [1–5]. Cerebral complications following surgical ligation of the internal or common carotid artery are seen in approximately 30% of patients [9], with a mortality rate of 10– 14% [8, 9]. The cerebral infarction in our case was present before operation and may have been caused by a combination of vascular insufficiency secondary to compression by the abscess and to hypoxia sustained during the cardiac arrest. Embolisation can be attempted to occlude the bleeding site during angiography. If this is unsuccessful, a balloon catheter can be left at the bleeding site to occlude the erosion prior to surgical ligation [7, 8]. Our case Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

Figure 3 Post-contrast computerised tomography scan at the level of the manubrium showing extravasation of contrast (arrowheads) and an anomalous right subclavian artery (arrows) crossing posterior to the oesophagus (E).

illustrates the usefulness of computerised tomography in the investigation of an enlarging neck mass in an unconscious patient in whom a history cannot be obtained. Computerised tomography detected the presence of a retropharyngeal mass, the extent of its involvement, the presence and extent of airway compromise, the presence of haemorrhage, the integrity of neighbouring vascular structures and the presence of an anomalous vessel. Our computerised tomography findings provided clinical grounds for further investigation with angiography which identified the location of the pseudo-aneurysm. Before surgery, the mass was thought to be due wholly to haematoma derived from the pseudo-aneurysm and a pharyngeal perforation was not suspected. Pharyngeal rupture complicated by vascular erosion therefore should be added to the list of differential diagnoses in an unconscious patient with an enlarging neck mass whose trachea has recently been intubated. This case also highlights that catastrophic complications can occur when tracheal intubation is attempted under suboptimal conditions and that tracheal intubation should normally be performed by an experienced and properly trained operator. Q 1997 Blackwell Science Ltd

References 1 Padmanabhan K, Gadde H, Vora S. Acute mediastinal widening following endotracheal intubation and gastric lavage. Western Journal of Medicine 1991; 155: 419– 20. 2 De Clercq LD, Chole RA. Retropharyngeal abscess in the adult. Otolaryngology – Head and Neck Surgery 1980; 88: 684–9. 3 Wolff AP, Kuhn FA, Ogura JH. Pharyngeal-esophageal perforations associated with rapid oral endotracheal intubation. Annals of Otology, Rhinology and Laryngology 1972; 81: 258–61. 4 Wong YK, Novotny GM. Retropharyngeal space – A review of anatomy, pathology and clinical presentation. The Journal of Otolaryngology 1978; 7: 528–36. 5 Wengen DFA. Piriform fossa perforation during attempted tracheal intubation. Anaesthesia 1987; 42: 519–21. 6 Salinger S, Pearlman JJ. Hemorrhage from pharyngeal and peritonsillar abscesses. Archives of Otolaryngology 1933; 18: 464–509. 7 Stevens HE. Vascular complication of neck space infection: Case report and literature review. The Journal of Otolaryngology 1990; 19: 206–10. 8 Gonda RL, Gutierrez OH, Hengerer AS, Hengerer AS, 157

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

De Weese JA. Pharyngeal abscess with external carotid artery erosion and pseudoaneurysm. Pediatric Neurosurgery 1990–91; 16: 21–4. 9 Eneroth AM, Tham R. Pseudoaneurysm of the internal carotid artery. Acta Otolaryngologica 1971; 72: 445–50. 10 Alexander DW, Leonard JR, Trail ML. Vascular complications of deep neck abscesses: a report of four cases. Laryngoscope 1968; 78: 361–70. 11 Langenbrunner DJ, Dajani S. Pharyngomaxillary space abscess with carotid artery erosion. Archives of Otolaryngology 1971; 94: 447–57.

12 Shapiro SL. Carotid artery erosion secondary to pharyngeal infection. Ear, Nose Throat Journal 1974; 53: 513–8. 13 Blum DJ, McCaffrey TV. Septic necrosis of the internal carotid artery: a complication of peritonsillar abscess. Otolaryngology – Head and Neck Surgery 1983; 91: 114–8. 14 Garino JP, Ryan TJ. Carotid hemorrhage: a complication of peritonsillar abscess. American Journal of Emergency Medicine 1987; 5: 220–3.

C A S E R E P O RT

Central airway compression

Anaesthetic and intensive care consequences

C. C. McMahon,1 L. Rainey,2* B. Fulton2 and I. D. Conacher1 1 Department of Cardiothoracic Anaesthesia, Freeman Group of Hospitals Trust, High Heaton, Newcastle upon Tyne NE7 7DN, UK 2 Department of Anaesthesia and Intensive Care, Newcastle General Hospital, Newcastle upon Tyne NE4 6BE, UK Summary

We present the case of a 32-year-old woman with compression of the central airway, in whom airway control was lost on two separate occasions, regained using a rigid bronchoscope and whose subsequent intensive care management was complicated by dynamic hyperinflation. Keywords Airway; obstruction. Lung; hyperinflation, trachea, compression. Complications; hypoxaemia. Surgery; mediastinoscopy, bronchoscopy. ...................................................................................... Correspondence to: Dr C. C. McMahon * Present address: The Women’s and Children’s Hospital, 72 King William Road, North Adelaide, South Australia 5006 Accepted: 13 October 1996

Patients with anterior mediastinal masses may present for surgery to obtain a tissue diagnosis, for staging procedures or for excision of the mass. It is recognised that airway management can be difficult and control lost at any stage of anaesthesia. We report a patient in whom the ability to ventilate was lost on two separate occasions and in whom the additional complication of dynamic hyperinflation developed during subsequent positive pressure ventilation, secondary to central airway distortion. Case report

A 32-year-old female was referred for bronchoscopy and mediastinoscopy. She gave a 2-month history of weight loss, increasing breathlessness and an irritable cough, 158

provoked by lying down. She was a non-smoker. Stridor was present when she lay flat and there was clinical evidence of superior vena caval obstruction. Chest X-ray showed a large superior mediastinal mass and a limited CT scan of thorax confirmed a homogeneous mass, 8 cm in diameter, causing compression of the lower trachea, carina and right main bronchus (Fig. 1). Pulmonary function tests demonstrated an FEV1 of 1.85 l and an FVC of 2.5 l (54% and 62% of predicted, respectively) while her oxygen saturation breathing air was 99%. She was 15 weeks pregnant at the time of referral. She was premedicated with morphine 10 mg and hyoscine 0.4 mg intramuscularly and pre-induction preparation included radial artery cannulation for blood pressure monitoring. After pre-oxygenation and with the patient Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

Figure 1 Pre-operative CT scan showing the anterior mediastinal mass with compression of the lower trachea (the lowest of three

available images).

semirecumbent, anaesthesia was induced with fentanyl and thiopentone. Vecuronium was given after an ability to manually ventilate the lungs was confirmed and a rigid bronchoscope was inserted. There was severe compression of the lower third of the trachea and during an attempt to define the anatomy of the central airway the patient became totally obstructed. Pulmonary ventilation could only be re-established after the rigid bronchoscope was forced through the area of compression into the left main bronchus. The minimum oxygen saturation at this time was 83%. In view of the need to establish a tissue diagnosis, it was decided to proceed to mediastinoscopy, with the rigid bronchoscope in place. The patient’s lungs were ventilated with a jet ventilator [1] and anaesthesia was maintained with a propofol and alfentanil infusion, with intermittent vecuronium. As the rigid bronchoscope was withdrawn, it was seen that the right main bronchus was functionally obstructed by extrinsic compression and by rotation of the carina 458 from the vertical. An uncut armoured tracheal tube (size 6.5), from which the cuff had been removed and an eye cut for the right main bronchus, was inserted. The provisional Q 1997 Blackwell Science Ltd

histological diagnosis was non-Hodgkin’s lymphoma and she required transfer to another (nonthoracic) centre for emergency radiotherapy. Ventilation of her lungs suddenly became very difficult shortly after arrival in the radiotherapy suite. Despite high inflation pressures with manual ventilation and adjustment of the tracheal tube, there was little chest wall movement and oxygen saturation decreased to 72%. Ventilation with a Sander’s injector attached to a rigid bronchoscope, which was passed beyond the obstruction into the left main bronchus, improved oxygenation. A more permanent airway was established using a left-sided double lumen tube (Portex 40 FG). She was transferred to the ICU, manual ventilation being applied via the bronchial limb of the double lumen tube. The patient’s lungs were artificially ventilated (Dra¨ger Evita 2), with settings to deliver intermittent positive pressure ventilation (IPPV) with a tidal volume of 700 ml, a respiratory rate of 12 breath.minÿ1 and an I:E ratio of 1:2. She became progressively hypotensive and the amplitude of the pulse oximeter trace decreased. Despite an increase in the inspired oxygen fraction (F IO2 ) to 1.0, 159

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

Figure 2 Chest X-ray taken at time of deterioration in gas exchange, with an infiltrative pattern in keeping with ARDS.

her oxygen saturation continued to decrease, from 95% to 68%. The right side of her chest visibly became overinflated. Finally, she was disconnected from the ventilator to allow deflation of the right lung, at which stage blood pressure and oxygen saturation improved rapidly. Alternative modes of ventilation were tried, including high-frequency jet ventilation (Penlon Bromsgrove Humidified Jet Ventilator), but all were associated with similar periods of hypotension and desaturation. The most satisfactory proved to be IPPV via the Evita 2, with a tidal volume of 500 ml, delivered at 6–8 breath.minÿ1 with a prolonged expiratory phase (an I:E ratio of 1:6) and an F IO2 of between 0.9 and 1.0. Even with this pattern of ventilation, her oxygen saturations slowly decreased and intermittent periods of disconnection from the ventilator by the nursing attendants were required for the next 36 h. By the fourth day and after three episodes of (unplanned) mediastinal radiotherapy, stability was achievable with IPPV at a normal rate, with an I:E ratio of 1:3 and an F IO2 of 0.4. The double-lumen tube was replaced by a size 160

8.0 low-pressure cuffed orotracheal tube and weaning was attempted. Her progress was very slow. Dynamic collapse of the lower third of the trachea on expiration, observed at fibreoptic bronchoscopy and continuing problems with expiratory ability of the right lung were considered to be the main contributing factors to her poor progress and fatiguability during periods of assisted spontaneous breathing. There was little evidence of tumour regression on completion of mediastinal radiotherapy and chemotherapy. Despite this, additional planned radiotherapy was regarded as the only therapeutic option since both the insertion of a silastic tracheal stent and tumour debulking were considered to be too hazardous. However, oxygen requirements began to increase and radiological evidence of ARDS became apparent (Fig. 2). Despite a variety of ventilatory manoeuvres, including prone positioning, her oxygenation deteriorated until, on the twenty-third day after mediastinoscopy, she suffered a hypoxic cardiac arrest and died. Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

Figure 3 Chest X-ray with the Portex double lumen tube in place, showing hyperinflation of the upper lobe of the right lung with compression of the lower lobe and mediastinal shift.

The patient had undergone two courses of mediastinal radiotherapy and a single course of chemotherapy, without apparent effect on the size of the tumour. Post-mortem examination confirmed ARDS with bronchopneumonia as the cause of death. The 230 g mediastinal mass was necrotic and there was no evidence of residual lymphoma. Discussion

Investigative options to define the nature of the mediastinal mass were limited in this patient. Computeriesd tomography (CT) guided fine needle aspiration was felt to be inappropriate because of the patient’s persistent cough, orthopnoea and the increased risk of haemorrhage in the presence of superior vena cava obstruction. Blind radiotherapy, relatively contra-indicated because of the potential to worsen critical airway narrowing, was not considered as an option because of her pregnancy. The presence of a pregnancy also placed a limit on the number Q 1997 Blackwell Science Ltd

and extent of radiographic investigations. Mediastinoscopy was thought to be the only way of obtaining an accurate histological diagnosis so that optimal therapy could be planned. It is possible to make an assessment of the airway with CT scans and expiratory spirograms [2], but these are not adequate to gauge three-dimensional anatomy, functional patency and optimal size and position for a tracheal tube to splint dynamically compressible areas. Rigid bronchoscopy is invaluable because it provides this additional information and at the same time ensures ventilation, which in this case was critical in enabling surgery to proceed. There is debate about the method of induction of anaesthesia in patients with large airways obstruction. Most solutions, including the use of local anaesthetic techniques, relate to upper airway obstruction and are of little relevance to the situation of central airways obstruction, for which tracheostomy does not provide relief. Loss of control of the airway is reported to have occurred 161

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

during both inhalational [3, 4] and intravenous [5, 6] inductions for surgery involving mediastinal masses. Since narrowed central airways can result in delays in equilibrium between inspired and alveolar concentrations of volatile agents, intravenous agents generally result in a more rapid and smoother induction. Coughing and raised intrathoracic pressure, which exacerbate airway narrowing and the effects of vena caval obstruction, are less likely to be induced. During the management of this patient it became clear that the off-the-shelf aids to replace the rigid bronchoscope for more prolonged airway support do not exist outside specialist units. Uncuffed, low-profile tracheal tubes obstructed the right main bronchus and did not splint beyond the area of critical airway narrowing and it proved impossible to align a customised hole in an armoured tracheal tube with the opening of the right main bronchus. The blind passage of a plastic double lumen tube (Portex) proved best in the emergency event. The narrow anterior–posterior diameter and the long length peculiar to this tube ensured that the bronchial lumen reached beyond the area of compression and secured ventilation of the left lung. With this tube it proved possible to secure some air-flow via the bronchial limb to the right lung, but in turn the arrangement precluded independent lung ventilation at an early stage and contributed to the development of right lung dynamic hyperinflation. When the dynamic hyperinflation became critical, the patient’s blood pressure decreased and oxygen desaturation occurred. A concurrent chest X-ray (Fig. 3) showed hyperinflation of the right upper lobe, with lower lobe compression and mediastinal shift. Such changes are associated with a rise in intrathoracic pressure, reduced systemic venous return and increased right ventricular afterload, with a subsequent fall in cardiac output. Unrelieved, barotrauma results and cardiac arrest can occur [7, 8]. The paradoxical effect of improvement in blood pressure and oxygenation when ventilation was disconnected was reflected in the pulse oximetry [9]. These warning signs were used by the nursing attendants to determine the

162

pattern of the apnoeic phases that resulted in the most effective ventilation. There is little doubt that the longterm effects of the dynamic hyperinflation contributed to the acute lung injury that ultimately proved fatal. Patients with central airways obstruction can prove particularly awkward when undergoing advanced airway management and positive pressure ventilation. In such cases, we recommend that units have available a variety of aids, including rigid and fibreoptic bronchoscopes. In addition, it is important to remember that anatomical abnormalities of the central airway may predispose to the development of air-trapping and dynamic hyperinflation. References 1 Paes ML, Conacher ID, Snellgrove TR. A ventilator for carbon dioxide laser bronchoscopy. British Journal of Anaesthesia 1986; 58: 663–9. 2 Gascoigne AD, Corris PA, Dark JH, Gibson GJ. The biphasic spirogram – a clue to unilateral mainstem bronchial narrowing. Thorax 1990; 45: 637–8. 3 Mackie AM, Watson CB. Anaesthesia and mediastinal masses. A case report and review of the literature. Anaesthesia 1984; 39: 899–903. 4 John RE, Narang VPS. A boy with an anterior mediastinal mass. Anaesthesia 1988; 43: 864–6. 5 Newman GG, Weingarten AE, Abramowitz RM, Kushins LG, Abramson AL, Ladner W. The anaesthetic management of the patient with an anterior mediastinal mass. Anesthesiology 1984; 60: 144–7. 6 O’Leary HT, Tracey JA. Mediastinal tumours causing airway obstruction: a case in an adult. Anaesthesia 1983; 38: 67. 7 Myles PS, Madder H, Morgan EB. Intraoperative cardiac arrest after unrecognised dynamic hyperinflation. British Journal of Anaesthesia 1995; 74: 340–2. 8 Rosengarten PL, Tuxen DV, Dziukas L, Scheinkestel C, Merret K, Bowes G. Circulatory arrest induced by intermittent positive pressure ventilation in a patient with severe asthma. Anaesthesia and Intensive Care 1991; 19: 118–21. 9 Conacher ID, McMahon CC. Pathognomonic pulse oximetry paradox. Lancet 1995; 346: 448.

Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

C A S E R E P O RT

An epidural abscess due to resistant Staphylococcus aureus following epidural catheterisation M. Yuste, J. Canet, M. Garcia, M. A. Gil and F. Vidal Department of Anaesthesiology, Hospital Universitari Germans Trias i Pujol, Carretera del Canyet s/n, 08916-Badalona, Barcelona, Spain Summary

We report a case of abscess formation after epidural analgesia, a rare complication that developed in our patient 13 days after placement of a thoracic epidural catheter for patient controlled analgesia. Culture of the pus grew methicillin-resistant Staphylococcus aureus. Although early diagnosis and rapid management have been reported to yield a satisfactory outcome, the case we describe ended in severe sequelae. Keywords Anaesthetic techniques, regional; epidural. Complications; epidural abscess. ...................................................................................... Correspondence to: Dr M. Yuste Accepted: 28 September 1996

Epidural abscess is a rare complication of epidural cannulation for anaesthesia or analgesia. Although a favourable outcome is dependent upon rapid surgical management, the great variability in presentation makes diagnosis difficult. Delay in diagnosis and treatment, however, can result in permanent neurological damage. Case report

A 44-year-old man was admitted to the emergency room after a traffic accident; he was awake after transient loss of consciousness. A chest X-ray showed some fractures of the left third and fourth ribs and the left clavicle but no obvious pneumothorax. A CT scan of the chest confirmed that no pneumothorax was present and showed additional fractions of the lowest right ribs. Physical examination was otherwise normal, as were blood count, coagulation screening and biochemistry. Arterial blood gases breathing room air showed moderate hypoxaemia (PaO2 7.46 kPa), and haemodynamic parameters were stable. Pulmonary contusion was suspected and the patient was prepared for transfer to the Intensive Care Unit (ICU). Prior to transfer, a thoracic epidural catheter for patientcontrolled analgesia (PCA) was inserted with the patient in the sitting position, via a paramedian puncture at the level of T5 using a Portex 18G Tuohy needle. Full asepsis was Q 1997 Blackwell Science Ltd

maintained throughout the procedure. After a test dose, one bolus of 0.25% bupivacaine 4 ml was administered with adrenaline 1 : 200 000 and fentanyl 50 mg to provide adequate analgesia. When the patient had been admitted to the ICU, an epidural infusion of 0.125% bupivacaine with adrenaline 1 : 400 000 and fentanyl 5 mg.mlÿ1 was started at 2 ml.hÿ1 with optional boluses of 0.5 ml every 20 min, via the PCA system, a programmed PharmaciaDeltec pump which had been prepared in the pain clinic under aseptic conditions. The pump cassette was changed on the second day by the pain clinic staff using sterile gloves for manipulation. The epidural catheter was removed intact on the fifth day when the patient was discharged from the ICU. The injection site was inspected and appeared clean. No incidents had been reported by the nurses during the 96 h the catheter was in place and the patient’s progress in the ICU was favourable, with blood gases returning to normal on the first day and radiological improvement on the second day. No fever developed and white blood cell counts were normal. Treatment in the ICU consisted of the usual procedures for hydration, gastric protection and antithrombotic prophylaxis. No antibiotics were administered apart from a single dose of cefazoline 1 g in the emergency room. After 3 days with no adverse events on the ward the patient was discharged and given an appointment for 163

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

follow up at the outpatient clinic 5 days later. At that visit he mentioned general discomfort in the upper back, which was judged to be related to contusions from the accident. The next day the patient was readmitted to another hospital complaining of asthenia, fever of 38.7 8C and thoracic back pain in a band at T4–T5 which did not improve with nonsteroidal anti-inflammatory drugs; hyperreflexia of both legs was observed. Cervicodorsal spine injury was suspected. Eight hours after admission, paraplegia with total lack of sensation below T4–T5 and loss of sphincter control appeared. Treatment was begun with methylprednisolone 2 g and ampicillin/gentamycin. Magnetic resonance imaging (MRI) showed a lesion in the posterior epidural space with compression of the spinal cord from T1 to T6 suggesting an epidural abscess or haematoma and a lipomatoid mass at T6–T9. Emergency decompressive laminectomy was performed under general anaesthesia within 12 h of the first sign of neurological symptoms; the diffuse abscess containing pus and granulation tissue was drained and a sample taken for culture while antibiotic treatment was continued with vancomycin. The bacterial culture grew methicillin-resistant Staphylococcus aureus (MRSA) sensitive only to vancomycin and phosphomycin. The latter antibiotic was added immediately. Sensory deficit recovered partially, but 6 months later the patient was still paraplegic. Discussion

Epidural abscess is very rare, with an estimated incidence of 0.2–1.2 per 10 000 hospital admissions. Although the number that is actually caused by anaesthetic and analgesic procedures is unknown [1, 2], an incidence of 15 abscesses per one million cases of extradural anaesthesia has been estimated [3]. Asepsis is extremely important in manipulation of epidural catheters. The operator should use a sterile grown and gloves and wear hat and mask when siting the catheter and also when preparing the pump cassettes. Single use particulate and antimicrobial filters must be used. Because we are certain that we followed the usual recommendations for asepsis and reduced manipulations to a minimum by using a fully watertight continuous infusion in a PCA system, we believe that other explanations should be sought to account for the abscess formation in the present case. The most common causative organism in epidural abscess is Staphylococcus aureus, although recent reviews suggest that the spectrum is broadening to Gram negative bacilli (13%) and anaerobic bacteria (8%) [1, 4–7]. Our case was unusual in that bacterial culture of pus from the epidural abscess grew a resistant strain of Staphylococcus aureus and to our knowledge this is the first such case 164

reported in the literature. As MRSA in our hospital has been found to be epidemic only in the ICU to date, we must assume that contamination of the epidural space occurred after the catheter was inserted in the emergency room. Haematogenous spread reaching a zone of low resistance to colonisation is the most frequent cause of epidural abscess in general [4]. A ‘weak zone’ could have been produced either by injury sustained in the traffic accident or by the existence of an epidural lipomatoid mass at the T6–T9 level just caudal to the puncture site (T5), unrelated to the type of analgesic technique used. Prior trauma at a site of puncture, present in 15–35% of cases, is considered to be a predisposing factor and diabetes, alcoholism and vascular catheters are others [4]. A ‘weak zone’, however, could also have been directly caused by the epidural technique itself, as a result of needle trauma, presence of a foreign body, or the formation of a small haematoma during catheter insertion or removal associated with heparin prophylaxis. Among pain treatments, long-term placement of epidural catheters and corticosteroid therapy have also been considered risk factors [8]; in fact, although the thorax was once the most common location in the earliest reported cases, abscess in the lumbar region is now seen more often, probably because of the greater use of this area for therapeutic procedures [4]. The location of our patient’s abscess in the most typical place, according to Dandy [9], does not in itself provide sufficient evidence to support one hypothesis over another. Early diagnosis of our patient’s condition was achieved but is, in general, difficult because of the rarity of this complication and the insidiousness of its symptoms. It is first suspected in only about 40% of cases [10]. If, in addition to the symptoms, however, there is a history of recent epidural analgesia, epidural abscess should be looked for and Gadolinium-MRI is the diagnostic procedure of choice [7]. After diagnosis, the most accepted treatment is immediate drainage of the abscess and antibiotic coverage [1, 4, 5, 7, 10, 11] although some authors have defended a wait-and-see attitude [12–17]. Prognosis has improved greatly since the earliest descriptions of this complication and mortality has decreased from 86% in 1926 to 7% at present [5]. Outcome, however, depends not only on rapid diagnosis and early treatment but also on the severity of neurological deficit before treatment [10]. In our case, surprisingly rapid degeneration of the patient’s condition to paraplegia worsened the prognosis, as paralysis is permanent in up to 50% of cases [10]. Still more important for prognosis is the duration of neurological deficit, as paralysis lasting more than 48 h is usually irreversible [18]. In our patient, the duration of paralysis before surgical decompression was in fact less than 12 h but a favourable outcome nevertheless did not ensue. Q 1997 Blackwell Science Ltd

Anaesthesia, 1997, 52, pages 150–168 Case reports ................................................................................................................................................................................................................................................

In summary, we must emphasise the importance of aseptic technique, which provides the only known way of preventing epidural space infection that falls fully under the anaesthetist’s control. We also believe that a wait-and-see attitude is inappropriate when symptoms which may be related to abscess develop, given that the sequelae associated with this rare complication are severe. References 1 Baker AS, Ojeman RG, Swartz MN, Richardson EP. Spinal epidural abscess. New England Journal of Medicine 1975; 293: 463–7. 2 Ngan Kee WD, Jones MR, Thomas P, Worth RJ. Extradural abscess complicating extradural anaesthesia for caesarean section. British Journal of Anaesthesia 1992; 69: 647–53. 3 Chadwick HS, Bonica JJ. Complications of regional anaesthesia. In: Bonica JJ, McDonald JS, eds. Principles and Practice of Obstetric Analgesia and Anaesthesia, 2nd edn. The Williams and Wilkins Co., 1995: 538–72. 4 Daroiche RO, Hamill RJ, Greenberg SB, Weathers SW, Musher DM. Bacterial spinal epidural abscess. Review of 43 cases and literature survey. Medicine 1992; 71: 369–85. 5 Del Curling O, Gower DJ, McWhorter JM. Changing concepts in spinal epidural abscess: a report of 29 cases. Neurosurgery 1990; 27: 185–92. 6 Kaufman DM, Kapaln JG, Litman N. Infectious agents in spinal epidural abscesses. Neurology 1980; 30: 844–9. 7 Maslen DR, Jones SR, Crislip MA, Bracis R, Dworkin RJ, Flemming JE. Spinal epidural abscess. Optimizing patient care. Archives of Internal Medicine 1993; 153: 1713–21.

8 Fine PG, Hare BD, Zahniser JC. Epidural abscess following epidural catheterization in a chronic pain patient: a diagnostic dilemma. Anesthesiology 1988; 69: 422–4. 9 Dandy WE. Abscesses and inflammatory tumors in the spinal epidural space (so-called pachymeningitis externa). Archives of Surgery 1926; 13: 477–94. 10 Danner RL, Hartman BJ. Update of spinal epidural abscesses: 35 cases and review of the literature. Reviews of Infectious Diseases 1987; 9: 265–74. 11 Hlavin ML, Kaminski HJ, Ross JS, Ganz E. Spinal epidural abscess: a ten-year perspective. Neurosurgery 1990; 27: 177–84. 12 Bouchez B, Arnott G, Delfosse JM. Acute spinal epidural abscess. Journal of Neurology 1985; 231: 343–4. 13 Hanigan WC, Asner NG, Elwood PW. Magnetic resonance imaging and the non operative treatment of spinal epidural abscess. Surgical Neurology 1990; 34: 408–13. 14 Leys D, Lesoin F, Viaud C, et al. Decreased morbidity from acute bacterial spinal epidural abscess using computed tomography and nonsurgical treatment in selected patients. Annals of Neurology 1985; 17: 350–5. 15 Mampalam TJ, Rosegay H. Andrews BT, Rosenblum ML, Pitts LH. Nonoperative treatment of spinal epidural infections. Journal of Neurosurgery 1989; 71: 208–10. 16 Messer HD, Lenchner GS, Brust JCM, Resor S. Lumbar spinal abscess managed conservatively: case report. Journal of Neurosurgery 1977; 46: 825–9. 17 Wheeler D, Keiser P, Rigamonti D, Keay S. Medical management of spinal epidural abscesses: case report and reviews. Clinics of Infectious Diseases 1992; 15: 22–7. 18 Heusner AP. Nontuberculous spinal epidural infections. New England Journal of Medicine 1948; 239: 845–54.

C A S E R E P O RT

Plasmapheresis in neuroleptic malignant syndrome L. Gaitini,1 M. Fradis,2 S. Vaida,1 S. Krimerman3 and A. Beny 4 1 Department of Anesthesiology, 2 Department of Otolaryngology – Head and Neck Surgery, 3 Intensive Care Unit, and 4 Regional Psychiatric Services, Bnai-Zion Medical Center, PO Box 4940 31048 Haifa, Israel Summary

Neuroleptic malignant syndrome is a severe and potentially fatal reaction to neuroleptic drugs. Treatment requires withdrawal of the neuroleptic agent, metabolic and cardiovascular support, physical cooling and treatment with dantrolene sodium and bromcriptine mesylate. We report a therapeutic success of plasmapheresis in a case of neuroleptic malignant syndrome in which conventional therapy had failed. We postulate that plasmapheresis may prove to be a useful tool in treating this frequently fatal disease. Q 1997 Blackwell Science Ltd

165

Case reports Anaesthesia, 1997, 52, pages 150–168 ................................................................................................................................................................................................................................................

Keywords Anaesthetics, intravenous; neuroleptics. Complications; neuroleptic malignant syndrome. ...................................................................................... Correspondence to: Dr L. Gaitini Accepted: 21 August 1996

Neuroleptic malignant syndrome (NMS) is a severe and potentially fatal reaction to neuroleptic drugs [1–3]. The syndrome was first described in 1968 by Delay & Deniker [4] and its incidence is estimated at 0.5–1.4% of all patients exposed to neuroleptic drugs [5, 6]. The drugs associated with NMS include phenothiazines, butyrophenones and thioxanthenes. Of these, haloperidol and long-acting preparations like fluphenazine decanoate are often involved. Events which may exacerbate NMS include hyponatremia [7], pre-existing fever damage, stress and exhaustion [8]. It is thought that NMS is due to a primary dopamine receptor blockade in the basal ganglia and hypothalamus [1]. Treatment requires withdrawal of the neuroleptic agent, metabolic and cardiovascular support, physical cooling and treatment with dantrolene sodium and bromcriptine mesylate [9–11]. Case history

A 36-year-old male suffering from mental retardation and a long history of schizophrenia was admitted to the hospital with high fever. He was suspected to be suffering from leptospirosis due to a rat bite 5 days before his admission. He had been given fluphenazine decanonate 3 days before admission for an exacerbation of his psychosis. He was treated with ampicillin 4 g intravenously daily and tetanus toxoid, and his temperature returned to normal after 2 days. The serology for leptospirosis was negative. Because of the patient’s psychotic symptoms, haloperidol 25 mg daily was started. After 3 days there was a rise in body temperature, with sweating, tachycardia and a fall in blood pressure (80/50 mmHg). He was confused and

catatonic, with extrapyramidal signs of tremor and muscle rigidity. Creatine phosphokinase (CPK) was elevated at 1600 IU.lÿ1 (normal: