Secondary Logo

Journal Logo

A massive intrathecal cefazoline overdose

Lang, E. W.*; Weinert, D.*; Behnke, A.*; Pflug-Rolfes, K.; Rautenberg, M.*

European Journal of Anaesthesiology: March 1999 - Volume 16 - Issue 3 - p 204-205
Case Report

This case report describes the clinical course of a patient who inadvertently received a massive intrathecal cefazoline overdose through a lumbar drain, which had been placed after trans-sphenoidal surgery for a pituitary adenoma. She received high-dose barbiturate therapy and was monitored with electroencephalography (EEG), somatosensory evoked potentials (SSEP), brainstem auditory evoked potentials (BAEP) and transcranial Doppler ultrasound (TCD). No cerebrospinal fluid (CSF) exchange was performed, but CSF was drained continuously with daily CSF-cefazoline level monitoring. Despite the extremely toxic dose, the patient soon recovered completely.

Departments of *Neurological Surgery †Microbiology, Christian Albrechts Universität, Weimarer Strasse 8, D-24106 Kiel, Germany

Accepted September 1998

Correspondence: E. W. Lang.

Back to Top | Article Outline

Case Report

A 53-year-old woman presented with acromegaly that had been progressive for 5 years. Notably, she had suffered from poliomyelitis during childhood, which had left her with a mild 5-/5 right-sided hip flexion paresis. A trans-sphenoidal approach was used to resect a somatotropin-producing pituitary adenoma. A lumbar drain was placed to reduce post-operative cerebrospinal fluid (CSF) rhinorrhoea. Intravenous prophylactic antibiotic therapy was given. Five days after surgery, cefazoline-sodium ≈1.5 g was administered inadvertently into her lumbar drain. Ten minutes after this, she complained of bilateral sciatica radiating over her thighs into her knees. She soon became diaphoretic and nauseated, her thigh pains increased, she reported a feeling of paraplegia and, 5 min later, she developed a series of generalized tonic-clonic seizures and bilateral-dilated pupils. After intubation, her grand mal seizures continued; they were eventually controlled with phenobarbitone (100 mg), diazepam (30 mg) and cisatracurium besilate (10 mg). Her pupils returned to normal size. Further drug therapy included sedation (midazolam) and analgesia (fentanyl), phenytoin and high-dose barbiturate therapy (methohexitone and phenobarbitone), which lead to a burst-suppression EEG. Accompanying systemic hypotension was treated aggressively with volume expansion, dopamine and dobutamine infusions. CSF (600 mL) was collected from the lumbar drain over the next 24 h, followed by about 250 mL over the subsequent 2 days. Her cefazoline CSF levels were approximately 3700 mg L−1 (=3.7 mg mL−1) shortly after the inadvertent dose, 69 mg L−1 27 h later, 55 mg L−1 28 h later and < 1 mg L−1 48 h later. Her PT fell from 100% to 53% and was treated with vitamin K; it returned to within the normal range after a further 2 days, while PTT remained within normal limits. Follow-up EEGs showed severe general changes with delta and theta activity and intermingled high-amplitude activity of 15-18 Hz without focal changes or seizure patterns. Somatosensory and auditory-evoked potentials and intracranial blood flow velocities measured by transcranial Doppler ultrasound were normal throughout her treatment in the ICU. Sedation and analgesia were reduced on day 3, she followed commands on day 4 and was extubated shortly thereafter. An MRI obtained 7 days later showed no signs of focal or generalized ischaemia. She was discharged 16 days after the incident with a subjective 50% worsening of her right-sided leg weakness; formal motor strength demonstrated a 4-/5 hip flexion paresis. Her neurological examination was otherwise unremarkable. The patient was fully informed of this complication.

On follow-up examination 15 months later, her paresis had returned to a 5-/5 strength, a follow-up MRI scan was unchanged, and her endocrine function had been well controlled. An EEG showed a slightly irregular and slow alpha activity of 7-8 Hz, with a slowing of activity over the fronto-central areas on both sides. There was no delta activity, and the amplitudes of the superimposed beta activity had returned to normal values. An EMG of the right quadriceps femoris muscle showed a rarefaction of the activity pattern after maximum innervation with increased mean amplitudes of the single potentials. Pathological spontaneous activity was observed on both sides, predominantly on the right side.

Back to Top | Article Outline


Intravenous antibiotic therapy is sometimes indicated prophylactically for patients with CSF drains. This is the first reported case of cefazoline, a widely used i.v. cephalosporine, administered directly into the human CSF space. Several important lessons can be learned from this case.

The effects of a lumbar intrathecal cephalosporine overdose appear to occur with a short latency, and they require immediate intensive management. The clinical sequelae observed confirm previously observed signs and symptoms of excessive intrathecal antibiotic administration in humans, i.e. paraplegia and convulsions [1]. Safe intrathecal concentrations for similar cephalosporines have been suggested between 50 and 100 μg mL−1 for cephalothin [2], between 100 and 250 μg mL−1 for cephaloridine [2-4] and ≈100 μg mL−1 for ceftazidime [5]. For cefotaxime, a level of 500 μg mL−1 CSF is reported to have caused ophistotonus and nystagmus in humans [3]. Our patient's condition at discharge and 15 months later suggests that, although the presumably safe cephalosporine dose was exceeded by highly toxic levels (3700 μg mL−1 CSF), a quick and complete recovery is possible. On day 2, cephalosporine CSF levels had returned to therapeutic levels (≈70 μg mL−1) without any CSF exchange, simply by drainage. There were no detectable signs of cerebral ischaemia, at least insofar as was amenable to neuroimaging, intracranial blood flow velocity and electrophysiological monitoring.

The most toxic effect apparently affected nervous structures at the spinal level at or close to the drug entry site, leading to a transient worsening of a preexisting mild leg paresis which followed poliomyelitis during childhood. Our electromyographic findings show typical signs of a neurogenic lesion, although it is not possible to distinguish whether this was caused by poliomyelitis or by cefazoline toxicity.

In summary, based on the experience gained here, we would propose that a therapeutic regimen for treating an intrathecal cephalosporine overdose should consist of seizure control with deep sedation, relaxation, artificial ventilation and close electrophysiological and coagulation monitoring until all monitored parameters appear sufficiently normal to start weaning from the ventilator. Systemic hypotension should be treated aggressively, and CSF drainage is recommended with close monitoring of cefazoline CSF levels. Finally, this case illustrates the importance of labelling any fluid-filled line properly for the prevention of such complications.

Back to Top | Article Outline


The authors wish to thank Professor Hisashi Onoue, Department of Neurological Surgery, The Jikei University School of Medicine, Tokyo, Japan, for translation from the Japanese language [3].

Back to Top | Article Outline


1 Wolman L. The neuropathological effects resulting from the intrathecal injection of chemical substances. Paraplegia 1966; 4: 97-115.
2 Quintiliani R. The intrathecal and intraventricular administration of antibiotics in central nervous system infections. Conn Med 1970; 34: 321-326.
3 Kanda R. Clinical study on intraventricular and intrathecal administration of antibiotics. Neurol Med Chir (Tokyo) 1978; 18: 733-741.
4 Murdoch JM, Speirs CF. Clinical trial of cephaloridin (ceporin), a new broad-spectrum antibiotics derived from cephalorine C. Br Med J 1964; 2: 1238-1240.
5 Thilmann AF, Mobius E, Podoll K. Intraventricular antibiotic therapy. Nervenarzt 1992; 63: 108-112.

cephalosporine, intrathecal administration, overdose, neurotoxicity

© 1999 European Academy of Anaesthesiology