Shaking spells can be difficult to diagnose perioperatively. Even though phenomena of excitability such as sustained clonus during reflex testing are frequent in the immediate postoperative period, overt myoclonus is a rare complication of general anesthesia. We present a patient who underwent 6 procedures over 5 years at our institution, with vastly different outcomes, because postoperative myoclonus after general anesthesia was initially misdiagnosed and treated as refractory epileptic activity. The episodic nature of perioperative anesthesia care presents a challenge to the continuity of care required to solve uncommon medical problems. The patient consented to the publication of this case report.
Our patient was a 38-year-old woman with a medical history of mild cerebral palsy, seizures, gastroesophageal reflux disease, gastroparesis, and scoliosis. She had no history of seizures as a child. She spoke at 9 months. She did not walk until the age of 22 months, leading to a diagnosis of cerebral palsy. As an adult, the manifestations of her cerebral palsy comprised mild loss of motor speech fluency and mild spastic left hemiparesis, but she was physically active, with a full-time job and a master’s degree in a basic science discipline.
One of us (C.N.S.) first met the patient during an informal consultation referred by faculty from another college because of concerns about “severe reactions to anesthesia,” which led the patient to forego any procedures requiring either sedation or anesthesia. Her first shaking spell occurred in the immediate postoperative period after a knee surgery at an outside hospital. An electroencephalogram (EEG) and magnetic resonance imaging at the outside hospital were reportedly normal. She had similar shaking spells after 2 further knee surgeries. After the third, she shook for 11 hours and jerked for 2–3 days. For this procedure, the patient reported a traumatic endotracheal reintubation and prolonged distress, while intubated and restrained, in the intensive care unit. After a fourth procedure to excise sebaceous cysts in her scalp, she was started on phenytoin, but she complained of difficulty walking, as though her legs did not want to move. Phenytoin was discontinued. She also experienced 11 days of postoperative stuttering, which did not improve after phenytoin, but did improve after clonazepam. Magnetic resonance imaging and EEG were persistently normal. A movement disorders specialist treated her briefly with carbidopa/levodopa for possible early-onset isolated dystonia (DYT1), but no change in her gait occurred. She was followed by an outside neurologist and started on levetiracetam, a second-generation antiepileptic drug. Baclofen improved her gait. An Epilepsy Monitoring Unit stay at another outside hospital recorded leg jerking without EEG changes.
The patient first presented to our hospital in 2012 for an esophagogastroduodenoscopy because of symptoms of gastroparesis, acid reflux, and weight loss. Attempts to gather outside records and EEG reports were still ongoing when she underwent the esophagogastroduodenoscopy as a routine procedure under general endotracheal anesthesia, premedicated with midazolam and maintained with a propofol infusion. Details of her anesthetic experiences for this and subsequent procedures are described in the Table. On emergence from general anesthesia, she was tachycardic, tachypneic, and unresponsive to commands but able to move her extremities. She had an intermittently fixed gaze and “tremors” of all her extremities that were felt to be epileptic in nature. Her blood glucose concentration was 65 mg/dL. The tremors did not improve after the administration of midazolam, lorazepam, and diazepam. She required reintubation and admission to the neurological intensive care unit. The consultant neurologist described nonstereotyped recurrent episodes with her eyes open and head back, upper extremity flexion with rhythmic jerking, and adduction of the lower extremities. This occurred every 10 minutes for several hours. A 48-hour video EEG recorded the movements but showed only myogenic artifact superimposed over a normal EEG background (Supplemental Digital Content 1, Supplemental Video 1, http://links.lww.com/AACR/A145, showing movements at a time of normal EEG background). Valproic acid was added to the levetiracetam but was discontinued at discharge. She continued on extended-release levetiracetam monotherapy.
Four years later, the patient presented with a complex comminuted tibial pilon fracture from a water slide accident. She underwent external fixation under combined regional and spinal anesthesia, supplemented with fentanyl during the placement of the regional anesthetic. She did not develop shaking spells and recovered well, but found the experience of being awake in the operating room profoundly terrifying. Because of her reaction to general anesthetics, the plan for the definitive orthopedic repair also included combined regional and spinal anesthesia, supplemented with dexmedetomidine to address the patient’s anxiety. Unfortunately, the combination of gastroparesis, opiates for block placement, and anxiety resulted in nausea and vomiting. The movement associated with ongoing dry heaves, despite multimodal antiemetic therapy, prohibited the conduct of surgery. Therefore, general anesthesia was induced with sevoflurane, and her airway was secured with an endotracheal tube. Sevoflurane was chosen to limit the number of medications and to avoid the use of neuromuscular blocking agents and their reversal. Because of the patient’s history, 24-channel video EEG was applied by a registered EEG technician and interpreted in real time at bedside by a neurologist with specialization in epilepsy (J.E.C.) for the remainder of the case and the patient’s initial stay in the postanesthesia care unit (PACU). The patient had taken her usual dose of levetiracetam preoperatively, but because she had significant vomiting and the intraoperative EEG demonstrated some possible low-amplitude generalized sharp waves despite fairly deep anesthesia (end-tidal concentrations of sevoflurane of 1.7%–2.0% in the setting of adequate regional anesthesia and normal end-tidal carbon dioxide of 35–40 mm Hg) with intermittent EEG attenuation (Figure 1), an additional dose of levetiracetam was administered. She emerged from general anesthesia uneventfully. EEG monitoring continued into the PACU. Shortly after her transfer to the PACU, she developed self-limited myoclonic jerks in her upper extremities without EEG changes (Figure 2; Supplemental Digital Content 2, Supplemental Video 2, http://links.lww.com/AACR/A146, showing movements limited to the hands again with normal EEG background). These myoclonic jerks declined in frequency and duration over the first 45 minutes of her PACU stay. During the myoclonic jerks, the patient was responsive and amenable to reassurance, emphasizing the benign nature of these movements. She recovered from her surgery without any further issues or myoclonic events. At a subsequent neurology follow-up visit, the levetiracetam was weaned and no further events have occurred. We have since anesthetized her 3 times, for removal of hardware, for irrigation and debridement of a tibial wound, and a cholecystectomy. She underwent general anesthesia each time, emerging after a loading dose of levetiracetam to brief, self-limited shaking episodes in the recovery room that did not adversely affect recovery or discharge.
Myoclonus is an abrupt, rapid movement with involuntary contractions of part or all of a muscle or muscle group. It can be single or repetitive, and involve the trunk, extremities, or face; it is often widespread (but not necessarily synchronous). Negative myoclonus is a brief and involuntary loss of movement of a muscle or a group of muscles.1 Nonepileptic myoclonus can be observed in generalized encephalopathy or degenerative disorders such as Creutzfeldt-Jakob disease, spinal cord and peripheral nerve disorders, but can also be entirely benign, such as with hypnic jerks, the involuntary muscle contractions preceding sleep.2
Myoclonus is seen frequently in the perioperative period, especially at induction with etomidate,3 and to a lesser extent, propofol.4 Myoclonus may originate in the cortex, subcortex, or spinal cord.1 Epileptic seizures, on the other hand, are a sudden change in behavior from electrical hypersynchronization of neuronal networks in the cerebral cortex.5 The behavior seen during a seizure is dependent on the function of the cortex underlying the seizure onset. Myoclonic epilepsy is a type of primary generalized epilepsy characterized by a generalized spike-wave pattern on EEG. Some have known genetic associations, whereas others are sporadic, but onset and diagnosis are typically in childhood or young adulthood, and neuroimaging studies are normal. In this group of disorders, the characteristic seizure manifestation is myoclonus, and can include with multiple seizure types such as generalized convulsions as well as staring spells.
Perioperative myoclonus is generally self-limited, resolving in most cases without further treatment. If myoclonus persists or is distressing to the patient, many cases will resolve with the administration of benzodiazepines, which facilitate inhibitory neuronal input because the myoclonus can be a product of changes in the balance between excitatory and inhibitory input to muscles caused by administration or withdrawal of anesthetic agents and adjuncts. Previously, our patient did not respond to treatment with benzodiazepines. Because of her history of cerebral palsy, an ischemic/hypoxic insult to the central nervous system, as well as postoperative myoclonus refractory to benzodiazepines, we tried levetiracetam, which may treat myoclonus caused by hypoxic events. Levetiracetam is efficacious in several forms of posthypoxic myoclonus6 and is commonly used as a broad-spectrum anticonvulsant, effective for both focal onset and primary generalized epilepsy. Although we believe our patient’s myoclonus was cortical, levetiracetam has been shown to be efficacious in patients with spinal myoclonus in 1 small case study.7
Our patient avoided invasive procedures because of her consistently complicated postoperative course despite the best intentions of those caring for her and attempts to identify the underlying cause. In the episodic and fragmented setting of perioperative care, it can be difficult to record acute events such as the postoperative shaking in a way that communicates across specialties. Although the shaking was seen in 2012 on the video EEG recorded postoperatively in the neurological intensive care unit, we failed to recognize the shaking as the event of concern and to positively identify it as myoclonus. Instead, the report emphasized the absence of epileptic activity. Furthermore, unexpected events such as this patient’s myoclonic movements bias toward a course of treatment that deals with the worst plausible cause, that is, a seizure in the setting of a patient with cerebral palsy, first, before considering a more benign cause such as myoclonus or simply seeking out additional data. Our patient’s EEG was persistently normal, and careful review during the events showed only myogenic artifact.
This patient’s postoperative myoclonus was likely caused by her cerebral palsy, which predominantly affected motor control as evident by her relatively delayed acquisition of walking and the spasticity interfering with her ability to walk well. General anesthesia profoundly and dynamically affects the balance of excitatory and inhibitory input of the nervous system in general and motor control in particular, probably best known to anesthesiologists in the concept of Guedel stages of inhalation anesthesia. We hypothesize that withdrawal of anesthesia affected inhibitory control of motor function in this patient in a way that set up myoclonus. Levetiracetam, by suppressing excessive excitation, suppressed this myoclonic disinhibition of motor function.
Postoperative myoclonus can resemble a seizure, making a clear diagnosis in a critical time of care difficult, even knowing that not all that shakes is epilepsy. Assessment is further complicated by the broad differential diagnosis of myoclonic movements.8 Given this patient’s medical history and her past reactions to general anesthesia, we are encouraged that her most recent general anesthetics were relatively uneventful after pretreatment with levetiracetam, a broad-spectrum anticonvulsant that has shown significant efficacy in the treatment of posthypoxic as well as spinal myoclonus.
Name: Todd E. Jones, MD.
Contribution: This author helped review the medical record and draft the manuscript.
Name: Jean E. Cibula, MD.
Contribution: This author helped with neurological consultation, electroencephalography (EEG) analysis, selection of figures and videos, and draft of the manuscript.
Name: Don Bohannon, MD.
Contribution: This author helped with anesthetic care, review the medical record, and draft the manuscript.
Name: Christoph N. Seubert, MD, PhD, DABNM.
Contribution: This author helped with coordination of care throughout the described anesthetics, review the medical record, and draft the manuscript.
This manuscript was handled by: Hans-Joachim Priebe, MD, FRCA, FCAI.
2. Haerer AF; Abnormal Movements. DeJong’s The Neurological Examination. 1992:5th ed. Philadelphia, PA: Lippincott, 408–410.
3. Reddy RV, Moorthy SS, Dierdorf SF, Deitch RD Jr, Link L. Excitatory effects and electroencephalographic correlation of etomidate, thiopental, methohexital, and propofol. Anesth Analg. 1993;77:1008–1011.
4. Jeon HW, Kang JH, Kim HS, Jo HY, Kim SH. A case of propofol-induced delayed-onset refractory myoclonic seizures. J Clin Neurol. 2007;3:154–157.
6. Genton P, Gelisse P. Suppression of post-hypoxic and post-encephalitic myoclonus with levetiracetam. Neurology. 2001;57:1144–1145.
7. Keswani SC, Kossoff EH, Krauss GL, Hagerty C. Amelioration of spinal myoclonus with levetiracetam. J Neurol Neurosurg Psychiatry. 2002;73:457–458.
8. Caviness JN. Treatment of myoclonus. Neurotherapeutics. 2014;11:188–200.
Supplemental Digital Content
© 2018 International Anesthesia Research Society