BRAIN INJURY, VOL. 18, NO.5(MAY 2004), 409–417 Pharmacological management of Dysautonomia following traumatic brain injury IAN J. BAGULEYyz, IAN D. CAMERON}, ALISA M. GREENy, SHAMERAN SLEWA- YOUNANy, JENO E. MAROSSZEKYzô and JOSEPH A. GURKAy y Brain Injury Rehabilitation Service, Westmead Hospital, Wentworthville, NSW, Australia z Department of Rehabilitation Medicine, University of Sydney, Sydney, NSW, Australia } Rehabilitation Studies Unit, University of Sydney, Sydney, NSW, Australia ô Department of Rehabilitation Medicine, Westmead Hospital, Wentworthville, NSW, Australia (Received 21 November 2002; accepted 31 March 2003) Primary objective: To document and critically evaluate the likely effectiveness of pharmacological treatments used in a sample of patients with Dysautonomia and to link these findings to previously published literature. Research design: Retrospective case control chart review. Methods and procedures: Data were collected on age, sex and GCS matched subjects with and without Dysautonomia (35 cases and 35 controls). Data included demographic and injury details, physiological parameters, medication usage, clinical progress and rehabilitation outcome. Descriptive analyses were For personal use only. undertaken to characterize the timing and frequency of CNS active medications. Main outcomes and results: Dysautonomic patients were significantly more likely to receive neuro- logically active medications. A wide variety of drugs were utilised with the most frequent being morphine/midazolam and chlorpromazine. Cessation of morphine/midazolam produced significant increases in heart rate and respiratory rate but not temperature. Chlorpromazine may have modified respiratory rate responses, but not temperature or heart rate. Conclusions: The features of Dysautonomia are similar to a number of conditions treated as medical emergencies. Despite this, no definitive treatment paradigm exists. The best available evidence is for morphine (especially intravenously), benzodiazepines, propanolol, bromocriptine and possibly intrathecal baclofen. Barriers to improving management include the lack of a standardized nomen- clature, formal definition or accepted diagnostic test. Future research needs to be conducted to improve understanding of Dysautonomia with a view to minimizing disability. Brain Inj Downloaded from informahealthcare.com by George Washington University on 03/23/11 Introduction Dysautonomia is a syndrome of episodic autonomic nervous system dysfunction and increased muscle tone affecting a small but significant sub-group of survivors Correspondence to: Dr Ian J. Baguley, Research Team Leader, Brain Injury Rehabilitation Service, Westmead Hospital, PO Box 533, Wentworthville, NSW 2145, Australia. e-mail: ianb@ biru.wsahs.nsw.gov.au Brain Injury ISSN 0269–9052 print/ISSN 1362–301X online # 2004 Taylor & Francis Ltd http://www.tandf.co.uk/journals DOI: 10.1080/02699050310001645775 410 I. J. Baguley et al. of severe traumatic brain injury (TBI). Dysautonomia has a dramatic presentation characterized by extreme, paroxysmal changes in blood pressure, heart rate, respira- tory rate, temperature and/or sweating. To date, it has been poorly reported in the medical literature where it is variously referred to as Dysautonomia, Autonomic Dysfunction Syndrome, autonomic or sympathetic ‘storming’, hyperpyrexia associ- ated with muscle contraction, hypothalamic-midbrain dysregulation syndrome, acute midbrain syndrome or diencephalic epilepsy. Despite this, research has shown statistically significant physiological differences between survivors of severe TBI with and without the condition [1]. Dysautonomia-like syndromes have been reported in a range of other central nervous system (CNS) disorders including cerebral tumours [2], hydrocephalus [3–6], intracerebral and subarachnoid haemorrhage [3, 7], hypoxic encephalopathy [8] and on discontinuation of dopaminergic [9] and GABA-ergic drugs [10]. One of the more concerning syndromes that Dysautonomia mimics is Neuroleptic Malignant Syndrome (NMS) [3, 11]. NMS is characterized by hyper- thermia and muscle rigidity along with other features such as sweating, dysphagia, tachycardia, blood pressure changes, leucocytosis, elevated Creatinine Kinase (CK) levels and altered consciousness (DSM-IV). NMS is thought to result from CNS dopamine blockade and has a 15–30% mortality rate [12]. Dysautonomia is also similar to another potentially fatal syndrome resulting from abrupt intrathecal baclofen withdrawal [13, 14]. This syndrome is characterized by fever, tachycardia, variable blood pressure, exaggerated rebound spasticity, muscle rigidity, altered mental state and increased CK levels. There are further similarities evident between Dysautonomia and Autonomic Dysreflexia (AD). In AD, non- specific noxious stimuli in people with a spinal cord injury above T6 can result in severe hypertension, increased spasticity, profuse sweating and vasodilatation above For personal use only. the spinal lesion. This association between noxious stimuli and increased symptoms has also been reported in Dysautonomia [15]. Unlike Dysautonomia, AD is usually associated with bradycardia due to intact baroreceptor reflexes mediated by the vagal nerve [16]. Given the clinical similarities between Dysautonomia and a range of potentially fatal syndromes, it could be argued that Dysautonomia also warrants the aggressive clinical management seen with NMS and AD. Taken alone, the increased core tem- peratures following acute TBI represent a potentially preventable cause of secondary brain injury [17]. The extent of these temperature increases are significantly higher and more prolonged in patients with Dysautonomia [1]. Brain Inj Downloaded from informahealthcare.com by George Washington University on 03/23/11 Despite the above, there is little or no systematic evidence available to optimize clinical management. Most of the published literature has concentrated on anecdote, individual case studies or small case series reporting medication effects. Some of these latter publications are difficult to interpret given that some patients were simultaneously trialled on multiple drugs and few studies published their criteria for determining efficacy. The aims of this study are to document medications used in a large published sample of patients with Dysautonomia [1], to critically evaluate the likely effective- ness of some of these medications based on observed physiological changes and to link these findings with a review of the published literature to suggest a tentative treatment protocol for this syndrome. Pharmacological management of dysautonomia 411 Method Subjects A retrospective case control methodology was used. For the purposes of this study, Dysautonomia was defined as simultaneous, paroxysmal increases in at least five out of the seven reported features of Dysautonomia (heart rate (HR), respiratory rate (RR), blood pressure, temperature, posturing, dystonia and sweating). File reviews of 35 age, sex and GCS matched subjects with and without Dysautonomia were performed. Control cases were taken from a database of 121 consecutive rehabilitation inpatients with a severe TBI on GCS criteria (maximum GCS 8 in the first 24 hours) and without evidence of Dysautonomia on file review. Data recorded were demographic and injury details, CT scan findings, physiological parameters (HR, RR, temperature), evidence of infections over the first 28 days, clinical progress and rehabilitation outcome. The dates that medica- tions, particularly antibiotics, sedatives and centrally acting drugs, were commenced or ceased were also recorded. Approval was obtained from the relevant human research ethics committee. Statistical analyses Descriptive analyses were undertaken to characterize the study participants. The timing and frequency of use of CNS active medications were reported descriptively. The data for the most frequently used medications (morphine/ midazolam infusions and chlorpromazine) were further evaluated. Dependent vari- ables included the mean of the maximum daily temperature, HR and RR during the pre-, on or off phase of Dysautonomia for each subject. Morphine/midazolam infusions were commenced on arrival in ICU and, thus, the effect of morphine/ For personal use only. midazolam infusions were investigated via a repeated measures t test. Chlorpromazine was utilized later during the ICU admission, allowing pre-commencement variables to be analysed. The effect of chlorpromazine status (pre-commencement, on chlor- promazine, off chlorpromazine) was analysed via a one-way repeated measures ANOVA. In the latter analysis, only cases with seven consecutive days of data were analysed. Results Brain Inj Downloaded from informahealthcare.com by George Washington University on 03/23/11 Demographic and injury related variables for the Dysautonomia and control groups appear in table 1. Records of medications used in the early intensive care management were available for 32 and 31 subjects in the Dysautonomia and control groups, respectively. Significantly more Dysautonomic patients received neurologically active medi- cations compared to controls (27/32 (84%) vs 9/31 (29%), (w2 ¼ 19.69, p < 0.001)). The only exceptions to this pattern were that there were no significant differences in the number of people managed with morphine/midazolam infusions ( p ¼ 0.773) or phenobarbital ( p ¼ 0.753) between the two groups (table 2). The average interval before morphine/midazolam infusions were replaced with PRN bolus doses was 4.1 days