Focal cooling may provide a safe, nondestructive alternative to resective and disconnective strategies that have been proposed or used to control refractory epilepsy. Observations of the effects of direct application of iced saline on the cortical surface during cortical mapping surgery and induced seizures have led to interest in developing implantable cooling therapy devices for refractory localizable epilepsies. In this article, the authors provide an overview of the historical background, physiology, and animal and human data leading to the development of implantable cooling devices for the treatment of medically refractory epilepsy.
The clinical problem of epilepsy
Humans have been affected by epilepsy for thousands of years. By the end of the twentieth century, there were several remarkable advances toward understanding and treating many of the epilepsies. Scientific speculation that some genetic epilepsies were caused by mutations in ion and voltage-gated channels was validated by a series of landmark genetic discoveries. New anticonvulsants, several with highly favorable therapeutic indices, were introduced. Sensitive magnetic resonance– and positron emission–based imaging tests emerged and rapidly evolved, revealing the focal cause of many complicated epilepsies. The once radical therapeutic option of surgical resection entered the epilepsy treatment mainstream. Yet, despite these advances, many patients are still affected by epilepsy as investigators search for more effective treatments.
Particularly problematic are the focal extratemporal epilepsies, occurring in up to half of the patients with poorly controlled seizures. Focal and multifocal seizures arising from the neocortex have proven refractory to conventional anticonvulsant therapy and newer surgical approaches. Even with guidance from modern imaging techniques that allow functional anatomic correlation of seizure onset with surface rendering of the neocortex, surgical treatment of extratemporal epilepsy results in only 36% to 76% of patients being seizure free.
Alternatives to permanent resection for neocortical epilepsy
A variety of invasive but nondestructive strategies have been proposed or used to reduce the frequency and severity of neocortical seizures. These strategies include vagal nerve stimulation (VNS), which for the past 15 years has been used for a subset of children and adults with refractory epilepsy. Although clinical studies have validated its efficacy, the overall reduction in seizure frequency with VNS was 28% in one randomized controlled trial. Although this reduction represents improved seizure control for some patients, for many others there is insignificant improvement in quality of life.
There is limited experience with several variations of intermittent electrical stimulation for human epilepsy. Recent publications have identified subsets of patients that benefited from intermittent hippocampal or thalamic stimulation. A 2010 prospective trial of anterior thalamic nuclear stimulation reported a seizure reduction of approximately 50%, which was substantial but still insufficient to dramatically improve the quality of their lives.
The most sophisticated intervention currently in clinical trials is a totally implantable closed-loop feedback system for seizure detection and cortical stimulation to terminate focal seizures. This system uses the same electrodes for recording and stimulation with a custom-designed seizure detection algorithm and a device that abruptly terminates seizures by delivering a burst stimulation lasting for about a tenth of a second. This method is being tested at several North American epilepsy centers, yet, to date, its utility seems similar to thalamic stimulation.
An alternative to these strategies is focal cooling, which is a safe nondestructive modality that may provide better seizure control if delivered at the ictal onset. Reports of the effects of direct application of iced saline on the cortical surface during cortical mapping surgery and induced seizures have led to interest in developing implantable cooling therapy devices for refractory localizable epilepsies. In the following sections we provide an overview of the historical background, physiology, and animal and human data leading to the development of implantable cooling devices for the treatment of medically refractory epilepsy.
Alternatives to permanent resection for neocortical epilepsy
A variety of invasive but nondestructive strategies have been proposed or used to reduce the frequency and severity of neocortical seizures. These strategies include vagal nerve stimulation (VNS), which for the past 15 years has been used for a subset of children and adults with refractory epilepsy. Although clinical studies have validated its efficacy, the overall reduction in seizure frequency with VNS was 28% in one randomized controlled trial. Although this reduction represents improved seizure control for some patients, for many others there is insignificant improvement in quality of life.
There is limited experience with several variations of intermittent electrical stimulation for human epilepsy. Recent publications have identified subsets of patients that benefited from intermittent hippocampal or thalamic stimulation. A 2010 prospective trial of anterior thalamic nuclear stimulation reported a seizure reduction of approximately 50%, which was substantial but still insufficient to dramatically improve the quality of their lives.
The most sophisticated intervention currently in clinical trials is a totally implantable closed-loop feedback system for seizure detection and cortical stimulation to terminate focal seizures. This system uses the same electrodes for recording and stimulation with a custom-designed seizure detection algorithm and a device that abruptly terminates seizures by delivering a burst stimulation lasting for about a tenth of a second. This method is being tested at several North American epilepsy centers, yet, to date, its utility seems similar to thalamic stimulation.
An alternative to these strategies is focal cooling, which is a safe nondestructive modality that may provide better seizure control if delivered at the ictal onset. Reports of the effects of direct application of iced saline on the cortical surface during cortical mapping surgery and induced seizures have led to interest in developing implantable cooling therapy devices for refractory localizable epilepsies. In the following sections we provide an overview of the historical background, physiology, and animal and human data leading to the development of implantable cooling devices for the treatment of medically refractory epilepsy.
Cooling and the brain
Focal brain cooling is an attractive nondestructive strategy to terminate and possibly prevent focal seizures. For several decades, published neurophysiologic studies have illustrated that cooling reduces synaptic transmission in the mammalian brain and that it should be possible to use new engineering technology to deliver focal cooling to superficial and deep brain structures. An 1895 report from Stefani and a 1900 report from Deganello provide the earliest descriptions of the central neurologic effects of focal cooling. A decade later, the German physiologist Trendelenburg began a systematic study of local hypothermia, investigating its effects on brainstem autonomic reflexes and the neocortex. These early investigators concluded that cooling reduced neurologic function at the system level, although none of them had any cellular insights. Throughout the rest of the twentieth century, physiologists continued to use local cooling to investigate cortical and subcortical localization of specific brain functions.
Over the last 50 years, information about the neurobiological effects of cooling has markedly increased. In 1965, Katz and Miledi used intracellular recording at the frog neuromuscular junction to show that cooling prolonged the latency of end plate potentials. The investigators stated that this prolongation likely occurred by desynchronizing acetylcholine release from presynaptic terminals. More recent experiments in mammalian tissue culture and brain slices showed that cooling affects neurotransmission through presynaptic and postsynaptic mechanisms. Two studies published in 1992 and 2001 show that cooling can augment the magnitude of neuronal action potentials and inhibit the sodium/potassium adenosine triphosphatase (ATPase). Although the initial ATPase inhibition and concomitant cell depolarization elicited transient hyperexcitability, cooling eventually reduced neuronal excitability. Our own published observations indicate that hypothermia rapidly reduces transmitter release from presynaptic vesicles and suggest that this may be a dominant effect of rapid cooling on central neuronal excitability.
The potential clinical utility of cooling for neurologic disease has been discussed for half a century. Fay began an extensive investigation of brain cooling in 1938. He suggested that systemic hypothermia or local cooling of the brain using chilled fluid circulating through a sealed metal capsule might be an efficacious treatment of head trauma and intractable pain. The investigator also believed that reducing brain temperature could inhibit tumor growth and therefore tried intracranial cooling for inoperable gliomas ( Fig. 1 ).
More recent controlled studies have reported the benefits of cooling in patients after acute head trauma or asphyxia. The temperature reduction in these recent clinical reports is generally no more than 4°C, which would not be expected to have a very large effect on synaptic transmission. This decrease is also much smaller than the cooling required to terminate severe experimental seizures in rodents.
Physicians have been aware of a causal relationship between elevated temperature and seizures since Hippocrates in 400 bc . Moreover, numerous in vitro and in vivo experimental epilepsy studies have consistently demonstrated that cooling reduces paroxysmal bursting and can diminish or stop seizure activity. Some clinical reports suggest that there is merit in using brain cooling for human seizures as well. At least 5 separate clinical investigations have documented the efficacy of cooling in the therapy for patients with intractable status epilepticus or chronic recurrent seizures. The first study in 6 patients with refractory status epilepticus documented that external cooling (31°C–36°C) stopped seizures in all but one. The second study described 25 patients, aged 8 to 46 years, who had frequent major motor seizures that were poorly controlled by chronic anticonvulsant therapy ( Fig. 2 ). While under general anesthesia, these patients were placed in an enclosed chamber and chilled to 29°C with cold air. Bilateral frontal burr holes were opened, and the subarachnoid space or ventricles were irrigated with iced saline, which reduced the cortical surface temperature below 24°C in 21 of the patients. A total volume of 500 to 20,000 mL of saline was infused into the patients who were slowly recovered from anesthesia. Of the 15 patients followed up for 1 year, 11 showed a marked reduction in seizure frequency, including 4 who had been seizure free.
The third study documented that vascular cooling to 31°C to 35°C resolved intractable status epilepticus in 4 patients whose seizures persisted despite aggressive intravenous drug therapy that included continuous midazolam or pentobarbital infusion. All 4 patients had a dramatic reduction in seizures despite the severity of their precooling seizures and the modest degree of cooling ( Fig. 3 ). Furthermore, there was clinical and electrographic documentation that cooling influenced the seizures ( Fig. 4 ). Two more published studies of human brain cooling during operative neurosurgical mapping confirmed that acutely lowering cortical temperature terminates paroxysmal discharges. In these cases, focal spikes abruptly stopped when iced saline was applied to the neocortex in the operating room. Both sets of studies have encouraged the development of more convenient methods of brain cooling for epilepsy.
Cooling methodology
Although it would be impractical to design an implanted cooling device for epilepsy therapy based simply on circulating cold water or conventional compressive refrigeration methodology, improvements in thermoelectric devices and their necessary supporting technologies have allowed the reevaluation of cooling as a therapy for some forms of focal epilepsy. Thermoelectric devices use Peltier’s 1834 observation that a temperature gradient develops at the junction between 2 dissimilar conductors when an electric current is applied across them. The discovery in the 1920s that synthetic semiconductors were superior to metals as thermoelectric elements hastened progress in this field. Modern semiconductors, typically alloys of bismuth, tellurium, selenium, and antimony, are now used to fabricate small light thermoelectric modules or Peltier devices, which are only a few millimeters in length and width and about 1.5 mm thick ( Fig. 5 A). In these modules, pairs of N- and P-type semiconductors are connected between 2 ceramic plates so that they are electrically in series and thermally in parallel. The latest thermoelectric devices are fabricated with thin film, which was developed for the microelectronics industry, and can be less than 200 μm thick. These devices have up to 10 times the heat pumping capability of conventional devices, making them ideal for medical devices (see Fig. 5 B).
Thermoelectric devices can generate temperature differentials of 70°C, but, for this to result in cooling, heat must be efficiently removed from the warm side. For most of our work, we attach the warm side of the device to a copper rod, which removes heat and acts as a convenient holder for a manipulator. When our work advances to the point of requiring internal heat dissipation, we use heat pipes that have been designed to efficiently transfer heat to highly vascularized areas of the skull or dura.
Results using epilepsy animal models
In 1999, we learned about the attractive features of commercially available thermoelectric devices and thus began experiments to determine whether they could terminate acute seizures. After we found that we could use cooling to stop seizurelike events in rat brain slices, we looked for an in vivo effect of cooling on focal cortical seizures. We discovered that we could reliably induce focal seizures in halothane-anesthetized rats by injecting a tiny volume of 4-aminopyridine (4-AP) solution 0.5 mm below the pial surface of the motor cortex using a micropipette. Within 20 minutes, the animals developed recurrent focal seizures and continued to have electrographic seizures for approximately 2 hours. Untreated seizures lasted between 60 and 80 seconds. We then maneuvered a thermoelectric device to directly contact the neocortex immediately above the injection site and at seizure onset and activated cooling to 20°C. This process reduced seizure duration to approximately 7 seconds ( Figs. 6 and 7 ). We believed that this rapid effect was because of the local cooling because the thermoelectric device did not influence seizure duration if it was not in direct contact with the cortical surface (see Figs. 6 B and 7 ). In this case, there was a progressive decrease in seizure duration when we cooled between 26°C and 22°C.
After completing our initial cooling experiments, we developed a frequency-based seizure detection algorithm for the 4-AP seizures, which allowed us to use a closed-loop system to abort seizures. To minimize false-positives, we set a relatively high threshold before cooling was activated. In these circumstances, our closed-loop system was as effective in terminating seizures as manual activation. Had we decided to accept more false-positives, we could have initiated more rapid cooling and further reduced seizure duration.
Since that time, we investigated other important aspects of cooling for focal seizures. In one case, we used a small thermocouple inserted into a 30-gauge needle to map the cortex below our thermoelectric device, which showed that cooling extends only about 4 mm below the surface. This finding demonstrated that the cooling effect should be localized to just a small region of neocortex below the pia.
A series of experiments replicated and extended our observation that cooling attenuates experimental epilepsy. In kindled rat hippocampal seizures, seizure severity and afterdischarge duration were reduced with cooling between 23°C and 26°C and with saline flowing through thin copper tubing next to the dorsal hippocampus ( Fig. 8 ). Another laboratory using Peltier devices almost identical to those used in our work showed that cortical excitability triggered by local injection of kainic acid was reduced by focal cooling. This group then showed that they could cool deep into the rat hippocampus by attaching an insulated needle to a Peltier device held away from the brain surface.
Ongoing experiments are evaluating the efficacy of focal cooling on antiepileptogenesis and anti-ictogenesis in a rodent model of pharmacoresistant posttraumatic epilepsy: the fluid percussion injury model (CURE/DOD USAMRMC 05154001.5 and 08149006) ( Fig. 9 ). We are currently preparing a comprehensive manuscript detailing substantial data confirming the efficacy of a double-blinded randomized trial of focal cortical cooling using this model.
Related posts:
Electrical Stimulation for Epilepsy: Experimental Approaches
Thalamic Stimulation for Epilepsy
Responsive Neurostimulation for the Treatment of Epilepsy
Hippocampal Stimulation in the Treatment of Epilepsy
Seizure Prediction and its Applications
Implanted Subdural Electrodes: Safety Issues and Complication Avoidance
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
