Low-dose Propofol Infusion for Controlling Acute Hyperspasticity after Withdrawal of Intrathecal Baclofen Therapy

A 65-yr-old man was admitted to the hospital with suspected baclofen pump failure. The pump had been sited 6 yr previously for severe spasms after a complete C5 injury, necessitating ambulatory noninvasive ventilation. The patient had felt increasingly unwell over the previous 24 h, with increasing spasticity and pyrexia (38.1°C). Laboratory investigations did not indicate an infective component. Five hours after admission, the hyperspasticity became more marked, with bilateral clonus and increased abdominal and respiratory muscle tone despite oral loading with baclofen. At admission to the critical care unit, arterial blood gas analysis revealed a metabolic acidosis (base deficit = 3.2). Plasma creatinine kinase was and remained within normal limits. Propofol was infused intravenously at 5–15 mg/h. Within 2 h, the base deficit had normalized, and both the spasms and clonus were markedly reduced. Abdominal and respiratory muscle tone decreased, with the patient feeling settled. Low-dose propofol had no effect on cardiorespiratory parameters or sedation status. The propofol infusion (target-controlled plasma concentration 0.6 μg/ml) was then continued, preventing the return of hyperspasticity, and stopped 2 days later, during which time a new baclofen pump was resited surgically.

Three separate episodes of acute hyperspasticity were treated in a 34-yr-old man who required intrathecal baclofen therapy after a complete C4 spinal cord lesion.

Admission 1  After an elective change of intrathecal baclofen pump, severe hyperspasticity developed within 24 h, accompanied by hyperpyrexia (40.1°C), tachycardia (160 beats/min), hypertension, and a respiratory rate of 33 breaths/min. Propofol was infused intravenously at 20–150 mg/h, in addition to 20 mg sublingual nifedipine and 100 mg dantrolene. Within 1.5 h after commencing the propofol, the spasms ceased. Within 5 h, heart rate and respiratory rate had decreased to 110 beats/min and 22 breaths/min, respectively, with normal core temperature and conscious level throughout the admission. The plasma creatinine kinase concentration was also normal throughout. During the first 5 h of admission, the patient also received 100 μg intrathecal baclofen via  a newly introduced intrathecal catheter. Surgical relocation of the catheter was undertaken the next day. Propofol was infused for optimal control of spasms for a further 24 h while intrathecal baclofen was reintroduced via  a newly inserted pump.

Admission 2  Two years after admission 1, the patient was admitted with new onset spasms, pyrexia (38.6°C), and respiratory discomfort (respiratory rate 24 breaths/min, no metabolic acidosis). Propofol was administered solely, with two boluses of 50 mg plus a 20-mg/h background infusion resulting in complete resolution of hyperspasticity, pyrexia, and respiratory rate within 2 h. The spasms returned after 7 h, during which time the propofol infusion had been reduced. The spasms resolved after increasing the propofol background infusion (50–150 mg/h). No changes in gas exchange or sedation levels were observed throughout the admission.

Admission 3  Three yr after admission 2, pyrexia (37.6°C) and increasingly severe spasms were noted 4 days after the scheduled reimplantation of a new baclofen pump. Propofol was commenced solely (10 mg/h), with resolution of tachycardia, pyrexia, and spasms within 1.5 h. Propofol was continued for 20 h, over which time pump-delivered baclofen administration was established. Again, no changes in gas exchange or sedation levels were observed throughout the admission.

Oral baclofen administration is usually inadequate to treat or prevent intrathecal baclofen withdrawal.20The clinical data presented support the hypothesis that propofol is the agent of choice for acute control of acute hyperspasticity. Although the neuropharmacologic actions of propofol and baclofen are incompletely understood, they share many similarities (fig. 1). Propofol enhances GABA-mediated synaptic inhibition in a number of neuronal systems.23Propofol enhances presynaptic inhibition at primary afferent terminals in human spinal cord24and dose-dependently decreases the firing rate and burst activity of nigral dopaminergic neurons, an effect reversed by a selective GABA type B antagonist.25Interestingly, propofol administered via  both intrathecal and intraperitoneal routes produces antinociception in rats by acting on spinal cord Δ-opioid receptors, whereas only intrathecally administered midazolam produces the same effect.26This may mirror our clinical experience of low-dose propofol infusions being effective in contrast to several previous case reports requiring high-doses of benzodiazepines for control of hyperspasticity. Furthermore, the antiinflammatory27and antinociceptive28,29effects of propofol may also confer some protective effect during baclofen withdrawal.

These clinical data lend support to the idea that propofol is the agent of choice for treatment of this and perhaps other hyperspasticity syndromes. Propofol possesses important neurophysiologic, neuropharmacologic, and pharmacokinetic properties to ameliorate the symptoms and prevent the pathophysiologic progression of acute baclofen withdrawal. In clinical practice, titration of propofol therapy using a target-controlled intravenous infusion against an objective score of spasticity (such as the Ashworth scale score30or Penn Spasm Scale score31) offers a new therapeutic approach to this potentially life-threatening syndrome.