The Efficacy of Intrathecal Baclofen in Severe Tetanus

A 35-yr-old man (weight, 65 kg) was admitted to our ICU with a 2-day history of severe tetanus (case #13, see Table 2). He had never been immunized with tetanus toxin and had sustained a knee wound 11 days previously. Trismus and dysphagia occurred 8 days after the wound, and generalized spasms occurred 1 day later. On admission, the patient presented with complete lockjaw, neck and general contractures with frequent, severe tonic paroxysms resulting in opisthotonos. He was fully conscious, temperature was 39.3 [degree sign]C, heart rate (HR) was 84 beats/min, and blood pressure was 120/80 mmHg. Dakar score [6]was 3/6 (Table 1). Plasma CPK level was 1,170 U/l (normal value < 100 U/l).

The knee wound was debrided. The patient was treated with tetanus antitoxin (heterologous serum 1,500 U; Serum antitetanique, Pasteur vaccins, Marnes la Coquette, France), repeated on hospital days 2 and 6, intravenous ampicillin, and vaccinated with tetanus toxin (Tetavax, Merieux, Lyon, France) on days 1, 30, and 60.

Sedation was immediately achieved with intravenous diazepam. An intrathecal catheter (Epidural Minipack, Portex Ltd, UK) was inserted via the L3-L4 vertebral interspace and advanced about 3 cm cephalad. Diazepam administration was ceased, and a bolus of 750 [micro sign]g baclofen (Lioresal, Ciba Geigy, Rueil Malmaison, France; 500 [micro sign]g/ml into saline solution) was administered intrathecally. Paroxysmal contractures disappeared within 2 h of the injection, and the patient was free of rigidity during the next 24 h. A second bolus was administered 24 h after the first one. Because of mild drowsiness during the few hours after injections, a continuous infusion (1,750 [micro sign]g/day) was instituted on the third day. Baclofen was continuously infused from day 3 to day 17 (mean daily dose, 1,690 [micro sign]g), allowing control of muscle rigidity, although trismus persisted. Additional sedation or tracheal intubation was not required. There were no adverse effects except for well-tolerated sinus bradycardia periods (HR, 45 beats/min) from day 12 to day 16.

After 15 days of intrathecal baclofen administration, rigidity was well controlled, and the CPK normalized. The baclofen daily dosage was decreased to 1,500 [micro sign]g and further to 1,000 [micro sign]g on day 17. On day 18, some spasms and moderate neck rigidity reappeared, and temperature increased to 40.5 [degree sign]C and CPK level to 560 U/l. Cerebrospinal fluid (CSF) and blood cultures were positive for Klebsiella pneumoniae. The intrathecal catheter was removed, and ampicillin was switched to ciprofloxacin. Residual muscle rigidity was controlled with diazepam. On day 20, the CSF was sterile, and the CPK level decreased to 143 U/l. Soon afterward the patient rapidly recovered; he was able to walk on day 22 and was discharged from the hospital on day 24.

The ICU is located in Bobo-Dioulasso, Burkina Faso. Mechanical ventilation and neuromuscular blockade are usable, but ventilators and monitors are scarce. Additionally, paramedical staff are less experienced than in developed countries.

From May 1995 to August 1996, 14 patients with severe tetanus (i.e., Dakar score [6]>or= to 2/6 and Mollaret stage [7]II or III) were treated with intrathecal baclofen. Demographic data, main therapeutic modalities, and outcome are summarized in Table 2. All patients received standard tetanus therapy with antibiotics (penicillin, ampicillin), antitoxin, and tetanus immunization. Baclofen administration was started as soon as possible. The dose of the first intrathecal injection was determined according to body weight (500 [micro sign]g if weight < 55 kg; 1,000 [micro sign]g if weight >or= to 55 kg). After administration of the first bolus, continuous intrathecal infusion of baclofen was the preferred method of administration if appropriate equipment was available (baclofen concentration in saline solution ranging from 20 to 80 [micro sign]g/ml). Supplementary intravenous diazepam was used only for extra sedation in those with pain. The dose of baclofen was adjusted every 12 h to allow control of muscle rigidity without adverse effects. If consciousness decreased (Glasgow Coma Score < 13) or bradypnea (respiratory rate < 10 breaths/min) occurred, the infusion was stopped until recovery and then started again at a decreased dose. Intrathecally administered baclofen was continued until persistent resolution of muscular rigidity and then progressively decreased by steps of 500 [micro sign]g every 2 days. The mean baclofen dosage was 1,225 +/- 309 [micro sign]g/day. Mean duration of treatment was 19 +/- 8 days for survivors. When respiratory depression required tracheal intubation, a tracheotomy was performed within 24 h, although there was always an attempt to keep intrathecal baclofen as the only sedation treatment. If intrathecal baclofen failed (lack of resolution of muscle rigidity while severe side effects prohibited dosage increase or respiratory distress required mechanical ventilation), treatment was withdrawn, and general anesthesia was undertaken with mechanical ventilation.

Cerebrospinal fluid cultures were systematically performed at admission and twice a week and additionally when temperature increased to more than 38.5 [degree sign]C. The catheter was removed when a meningeal infection was diagnosed, and baclofen was administered via daily lumbar punctures.

In 12 patients, intrathecal baclofen was effective as antispastic treatment with resolution of generalized muscle rigidity and spasm, although dysphagia and trismus most often persisted. However, two patients (14.3%) had to be intubated because of tracheal obstruction and laryngeal spasm (Table 2). Overall mortality rate was 5 of 14 (35.7%).

These results can be compared with those of the previous 11 patients admitted in our ICU for severe tetanus before the use of intrathecal baclofen. Despite demographic data and severity scores similar to the baclofen-treated patients, 45.5% of these previous patients were intubated, requiring deep sedation in each case. Overall mortality rate was 9 of 11 (81.8%), reaching 100% for intubated patients. These mortality rates are higher than in developed countries (now between 20% [8,9]and 50%, [8,10]depending on age and severity of illness). However mortality is commonly more than 50% in developing countries (up to 74% in Gabon [11]). In such countries, facilities for specialized cares with respect to equipment, medications, and qualified staff are more limited. Although comparison with retrospective controls is difficult, it is our impression that patient outcome is better since the introduction of intrathecal baclofen.

To our knowledge, 26 adults with severe tetanus treated by intrathecal baclofen have been reported to date, [12-22]including 11 cases in Africa. Our series is the largest reported in the literature. Our results agree with the literature, showing efficacy of intrathecal baclofen in controlling muscle rigidity. They also suggest a decreased need for tracheal intubation or deep anesthesia. These findings are important for developing countries wherein intensive care facilities are scarce.

Because of slight liposolubility baclofen does not easily cross the blood brain barrier. Therefore local administration yields to higher concentration on effect site, with fewer adverse systemic effects. After intrathecal injection, the effect of baclofen begins within 1 or 2 h and persists 12-48 h. [5,12]Baclofen elimination half-life in CSF has been found to range from 0.9 to 5 h. [23]After lumbar intrathecal administration, cervical-to-lumbar concentration ratio is 1:4. [24]This distribution probably explains the cephalic progression of hypotonia and the higher posology necessary to relax trismus than legs, which is reported by every author.

The major adverse effect of baclofen is depressed level of consciousness (LOC) and respiratory compromise. [5,25]These potentially deleterious effects can develop at any dose and can develop before resolution of tetanospasm. [22,26] 

Seven of our patients developed decreased LOC or respiratory depression that was readily reversed by decreasing the baclofen dose. This side effect seems more closely related to overdose than to accumulation because it is more often reported after first injections [5]or during bolus administration. [27] 

Three cases of sinus bradycardia were noted, but two of them did not seem to be related to dosage nor to timing of baclofen injection.

Infection of the intrathecal catheter is another serious adverse effects of intrathecal baclofen administration. Meningitis occurred in three patients in our series (patients #2, 13, and 14 on days 5, 8, and 16). However, recovery without sequelae was obtained after withdrawal of the catheter and antibiotic therapy. To avoid these risks, administration by repeated lumbar punctures has been proposed. [15]However, this technique is uncomfortable and also creates the risks of overdose and [28]inability to titrate effect, resulting in consciousness disorders.

Consequently these inherent adverse effects preclude use of this treatment for moderate tetanus that can be managed with benzodiazepine alone without need of tracheal intubation. Nor is intrathecal baclofen appropriate for the most severe forms that need deep sedation and even muscular blockade.

In our 14 patients with severe tetanus, intrathecal baclofen was an efficient treatment allowing effective control of muscle rigidity and spasms in 12 patients and reduced the need for deep sedation in the remaining 2 who required tracheal intubation. These results, which appear to represent an improvement compared with our previous experience and which agree with those previously reported, argue in favor of use of intrathecal baclofen for severe tetanus, especially in developing countries where respiratory resuscitation techniques are seldom available. Further prospective studies comparing intrathecal baclofen with conventional treatment in regard to overall mortality rate, cost, and length of hospital stay are needed.