Behavioral dysregulation exacerbated by intrathecal baclofen in an adolescent with severe traumatic brain injury

1 Introduction

Dystonia, an abnormal, involuntary sustained or intermittent muscle contraction that can cause repetitive movements, twisting, and/or abnormal posturing [1], can be a sequela of traumatic brain injury (TBI) that can result in significant functional impairment. Intrathecal baclofen (ITB) pumps can treat dystonia following TBI [2]. ITB is often reserved for severe dystonia, which is generally considered dystonia refractory to enteral pharmacologic agents and botulinum toxin injections. Adverse drug reactions to both oral and ITB are rare but can include confusion, agitation, depression, seizures, or behavioral disinhibition [3–6]. While changes in emotional and behavioral responses are common after TBI, agitation typically occurs within the first few weeks of recovery and does not typically present suddenly years after injury [7]. Few reports have shown worsening psychosis with oral baclofen in patients with [5, 9] and without [10, 11] history of mania. Currently, no reports have demonstrated behavioral dysregulation in patients using ITB pumps with a history of TBI.

2 Case

In April 2016, a 15-year-old male sustained a severe TBI after an 80-foot fall (Table 1). He was previously healthy, independent in mobility and activities of daily living (ADLs) without any cognitive or behavioral dysfunction. He lost consciousness from the fall and initially scored a 3 on the Glasgow Coma Scale. MRI brain scans showed diffuse axonal injury, most notable in the right superior frontal gyrus, left inferior frontal lobe, right temporal lobe, right cerebellum, left pons, and corpus callosum, with a large parenchymal hemorrhage of the right thalamus. His course was complicated by generalized tonic-clonic seizure. He was in a coma for >24 hours, in post-traumatic amnesia for >7 days, and treated for autonomic storming but did not have any significant behavioral issues.

The TBI caused left greater than right-sided dystonia, torticollis, and significant cognitive and communication deficits requiring admission to inpatient rehabilitation (IPR). He was first started on oral baclofen for dystonia without any behavioral issues, and then received botulinum toxin injections with some benefit. Carbidopa-levodopa was added next but his dystonia remained significant and functionally limiting. Trihexylphenidyl was also considered but due to limited pharmacologic benefit thus far, the decision was made with the family to pursue ITB pump placement June 2016. Catheter placement was confirmed to be at T8 after initial placement. He was nonverbal throughout his IPR stay, confused but not agitated, and followed simple commands. He was discharged from IPR three months after injury, requiring maximum assistance for transfers, grooming, dressing, and feeding while also using a wheelchair for mobility. After initial ITB pump placement, his baclofen dose was 437.9 mcg/day, and he continued carbidopa-levodopa with good control of his dystonia. His ITB dose was gradually increased over the next couple of years to 745.2 mcg/day for his dystonia.

In late 2018, he was diagnosed with bipolar II due to episodes consistent with hypomania. He has no family history of psychiatric disorders and slept approximately two hours a night for weeks at a time. In addition, he had paranoid beliefs, and some auditory and visual hallucinations. When sleep improved, the hypomania improved for a few weeks, but the cycle would begin again. He was taken off of carbidopa-levodopa as it was thought to be contributing to low mood, and was trialed on lamotrigine, lithium, and quetiapine for bipolar disorder. Quetiapine and lithium were not tolerated due to side effects. ITB dose at that time was increased to 819.72 mcg/day since carbidopa-levodopa was discontinued. Eventually, he was only on lamotrigine and tolerated it without any recurrent mood episodes.

By late 2019, he was ambulating with a 4-wheel walker, climbing stairs, independent with ADLs, communicating with sign language, and on a modified diet. The patient was very active but in early 2020, he developed increasing difficulties with dystonia again. Despite up-titration of ITB dose to 1,100 mcg/day, his dystonia did not improve. Given concern for lack of efficacy, the ITB catheter was evaluated with a catheter dye study and found to be in the epidural space rather than intrathecal. It is uncertain how long the catheter was in the incorrect place. His ITB pump was down-titrated to 100 mcg/day to avoid baclofen withdrawal. He underwent catheter revision in March 2020 with a catheter tip placed at T5. The post-revision ITB dose was 100 mcg/day. It was complicated by cerebral spinal fluid leakage, requiring additional weeks of immobility. Over the following weeks, his dose was increased to 275 mcg/day.

By April 2020, the family noted paranoid behaviors such as accusing his father of stealing his possessions. At that time, his ITB dose was around 375.4 mcg/day. He was re-evaluated by his psychiatrist in late April 2020 and started on cariprazine and lithium in addition to his home lamotrigine for presumptive bipolar disorder. After starting these medications, he developed severe dysphagia, increased falls, decreased participation with therapies, and functional decline. From May through June 2020 his behaviors escalated with paroxysms of head-banging, yelling, throwing objects, punching hard surfaces, slamming doors, and choking himself with his hands.

Over these same few months, ITB dosing was also being up-titrated for dystonia. Despite psychiatric medication changes, he continued to exhibit mood changes, dystonia, and developed a new tremor. He was admitted to the hospital in June 2020 for his behavioral changes. Medications prior to admission included valproic acid 375 mg BID, lamotrigine 150 mg BID, diazepam 1 mg BID, and citalopram 10 mg QHS. Due to the side effects, lithium had been discontinued prior to admission. During hospitalization, diazepam was tapered off to obtain a more accurate cEEG. No epileptiform activity was seen after 5 days of EEG monitoring. Valproic acid was consolidated to an extended-release form because it was thought to be contributing to his tremor [12]. Due to concerns for polypharmacy, his citalopram was discontinued. Neuroimaging demonstrated remote TBI and no acute intracranial findings.

Rehabilitation medicine was consulted for ITB pump management. During the initial consult, the baclofen pump dose was 230.2 mcg/day with simple continuous dosing. Physical exam showed good strength but severe dystonia. The family reported he was functioning below his functional baseline prior to ITB pump revision and unable to ambulate safely or navigate stairs, activities he was able to do previously. His fall risk increased, and he required more caregiver assistance for simple self-care tasks, such as feeding, so he was admitted to IPR. ITB was increased to 253.2 mcg/day to assist with dystonia but resulted in a significant increase in behavioral outbursts a couple of days later. He was started on pindolol around the same time due to concern for TBI-related agitation and reached maximum dosing, but no significant improvement in his behavior was noted. ITB continued to be up-titrated for dystonia (maximum of 278.1 mcg/day), but the patient’s family noted his mood instability worsened with ITB increase. Thus, his pump was down-titrated 5–10% every couple of days, which appeared to decrease the frequency of behavioral outbursts. His psychiatric medications were discontinued aside from lamotrigine, and a behavior plan was created with psychology to assist with prevention of behavioral outbursts and de-escalation strategies. Once his outbursts were controlled, he was discharged home on pindolol, lamotrigine, and a reduced dose of ITB of 178.2 mcg/day in simple continuous mode with stable dystonia and overall improved functional status.

Since discharge from IPR, he continued to be down-titrated on his ITB dose and remained on pindolol and lamotrigine. He is administered 1 mg lorazepam for breakthrough outbursts when he displays violent or self-injurious behaviors. However, he has rarely needed to use this. For dystonia, he received physical therapy, occupational therapy, an ITB dose of 150 mcg/day with a plan to continue down-titration, as well as receiving botulinum toxin injections. The patient and family reported improvement in overall behavior.

This case suggests a possible correlation between ITB and behavioral dysregulation in a patient with severe TBI. While it is difficult to determine whether the patient’s behavioral changes can be explained by the increased ITB after catheter revision or an exacerbation of an underlying mood disorder from his TBI, the timing of events suggests baclofen may play a role. Bipolar disorder is a rare sequela of TBI [13] and difficult to diagnose in the setting of TBI due to overlapping symptoms common to both [14]. In retrospect, the consulting psychiatric team that evaluated the patient after baclofen pump revision felt it was unclear if he truly was bipolar because his TBI was a major confounder of the initial bipolar diagnosis. At the time of bipolar diagnosis, his presentation was potentially a consequence of TBI-related insomnia which has been seen in up to 50% of TBI patients [15]. Sleep deprivation may have caused hallucinations rather than a primary psychiatric disorder. Additionally, the episodes of disinhibition and agitation after catheter revision were not consistent with bipolar but more consistent with behavioral dysregulation (agitation) related to his former TBI and concomitant medication changes.

While it is unclear if the patient truly met diagnostic criteria for bipolar disorder, bipolar could be a risk factor for increased sensitivity to baclofen-related behavioral dysregulation. Reports have suggested baclofen provokes mania in patients with known bipolar disorder [5, 8] and those without psychiatric history [10, 11]. In one case, baclofen was used off-label for treatment of alcohol use disorder in a patient without a history of mood disorders and presented with mania (inflated self-esteem, decreased need for sleep, flight of ideas, increased loquacity, and distractibility) after taking 180 mg/day of oral baclofen. The symptoms decreased when the dose was decreased. In all case reports, the action taken was to decrease or discontinue baclofen.

In this case, it is thought that minimal baclofen was reaching the patient’s central nervous system (CNS) due to the positioning of the ITB catheter in the epidural space. Once the catheter was revised and the baclofen dose up-titrated, he was noted to have worsening behaviors, possibly due to increased CNS exposure to baclofen. Additionally, his behavioral outbursts became less severe and less frequent as his ITB dose was decreased overall by 36% during IPR. While pindolol may have slightly improved his behaviors, he was kept on a stable dose of 10 mg TID throughout his stay. The most significant behavioral changes were seen soon after baclofen dose adjustments. Of the medications he was taking, the primary interactions between them included altered seizure threshold, CNS and respiratory depression, with no significant psychiatric effects. Furthermore, no clear alternative causes of his symptoms were evident after thorough metabolic, neurologic, infectious, and psychiatric work-up.

The mechanism that may induce these behavioral changes is not well understood but may be indirectly related to the dopaminergic and serotonergic systems. Baclofen is a gamma-aminobutyric acid type B (GABA-B) agonist that acts on the CNS. It can increase the rate of norepinephrine turnover and up-regulate serotonin receptors [16]. Additionally, GABA-B receptor stimulation can affect dopaminergic neuron firing rate [17]. Mania has been hypothesized to be associated with variations in dopaminergic transmission in different regions of the brain [18]. Thus, baclofen may play a role increasing susceptibility to mood disorders. The patient was also on several medications for bipolar disorder that affect the dopaminergic and serotonergic systems. He did not tolerate quetiapine or lithium, which influence serotonin and dopamine. He was able to tolerate lamotrigine, which primarily acts on glutamate and sodium channels and has a weak effect on serotonin. It is not clear why oral baclofen did not exacerbate his symptoms as much as ITB initially. The most significant difference is that when he was on oral baclofen, he was still in the acute phase of his TBI recovery, whereas his catheter revision occurred a couple of years later. More research is needed to determine the pathophysiology of baclofen in TBI.

Only a few case reports have demonstrated episodes of baclofen-induced mania, but none in patients with TBI. This is the first case that indicates that intrathecal baclofen may cause behavioral dysregulation in patients with severe TBI. Patients and providers should be aware of the risks and benefits of ITB, including behavioral dysregulation.

4 Conclusion

Intrathecal baclofen may trigger behavioral dysregulation in patients with severe TBI. Clinicians should be cautious when initiating intrathecal baclofen and monitor closely for any behavioral changes associated with increased ITB doses.

Author contributions

Each person listed as an author has participated in the study to a significant extent.

Conflict of interest

The authors have no conflict of interest to report.

Ethical considerations

Informed consent from the patient was obtained and approved by the authors' Institutional Review Board.