Introduction
Febrile infection-related epilepsy syndrome (FIRES) is clinically defined as a subset of new-onset refractory status epilepticus that presents with fever 2 weeks to 24 hours prior to onset of refractory status epilepticus (RSE).1,2 FIRES is a rare presentation that was first defined in children, although it is considered to be the same disease process when seen in adults.1,2 The acute phase, characterized by RSE, is devastating with a mortality rate of 12% in children and ranging from 16–27% in adults.1 The majority of survivors experience medically refractory epilepsy and long-term neurological sequela (including a vegetative state in severe cases) in the chronic course of the disease.1,3 The etiology is unknown, but it is believed to be secondary to a pro-inflammatory state.1–3
Multiple studies have explored the process by which central nervous system (CNS) inflammation contributes to the development of epilepsy. The inflammatory cascade triggered by interleukin-1 (IL-1) has been specifically implicated,2,3 with inhibition of this pathway demonstrating a disease modifying effect in multiple animal models.2,4 Anakinra is the recombinant IL-1 receptor antagonist (IL1-RA) used in these studies. Early reports show anakinra use is safe2,5,6 and believed to show good therapeutic response in many FIRES patients.1,2,7–10 Reviews on the topic have recommended starting anakinra within 2 weeks of onset, in the setting of continued super-refractory status epilepticus after corticosteroid and/or intravenous immunoglobulin use.1,2 Case reports in children have shown efficacy in prolonged dosage periods (7 months to 1 year).7–9 One case report features an adult with FIRES successfully treated with anakinra.10 In this case, the patient was seizure free within 24 hours of starting anakinra and remained seizure free at 6 months follow up. Here, we report an adult with FIRES successfully treated with short-term anakinra.
Case Report
A 47-year-old physics teacher initially presented to an emergency department (ED) after falling down 16 stairs in his house, found to be having new onset seizures. His only medical history prior to this event was generalized anxiety disorder. One week prior to this event, the patient experienced a sensation of overheating while doing yard work. Over the course of the next week, he experienced multiple episodes of bilious emesis, headache, and self-reported fever. On the day of presentation, his wife heard an ictal cry prior to finding him at the bottom of their basement stairs. After an unremarkable workup including normal magnetic resonance imaging (MRI) of the brain with and without contrast at a local ED, he returned home where he had two more seizures with an additional seizure en route to the hospital. Vitals were significant for a temperature of 101.7°F without any source of infection or other abnormalities on chest X-ray, urinalysis, or complete metabolic panel. He received lorazepam 1 mg and fosphenytoin 15 mg/kg, then was started on maintenance levetiracetam and transferred from a local ED to a neurological intensive care unit (ICU) at a tertiary care facility. He arrived alert, following commands, and oriented only to self, complaining of a headache and fatigue.
Continuous video electroencephalogram (EEG) monitoring captured focal status epilepticus with bilateral independent temporal lobe seizures (left greater than right) (Fig. 1A). Initial electroclinical seizures were typified by agitation sometimes accompanied by oral automatisms and right greter than left facial twitching. However, the majority of seizures were subclinical. He required intubation after receiving high doses of multiple anti-seizure medications (ASMs) in an attempt to treat status epilepticus. Further, his clinical course was characterized by escalating use of multiple ASMs including anesthetics (Fig. 2). Following resolution of status epilepticus, the EEG was characterized by inter-ictal sharp waves and brief potentially ictal rhythmic discharges occurring independently in bilateral temporal lobes (Figs. 1B, C, 3). As anesthetic medications were added, his EEG transitioned to a burst-suppression pattern which returned briefly to focal status epilepticus once these medications were weaned (Figs. 1D, 2, 3).
ASMs used, not all concurrently, included levetiracetam, fosphenytoin, phenytoin, lacosamide, clobazam, topiramate, phenobarbital, propofol infusion, ketamine infusion and midazolam infusion prior to initiation of immunomodulatory medications (Fig. 2). On hospital day 7, methylprednisolone 1 g daily was started for a 5-day course. Ketogenic diet was initiated through a nasogastric tube formula at a 4:1 ratio, though achieving this ratio was limited by carbohydrate containing medications and ketosis was not achieved. On hospital day 9, anakinra 100 mg TID was started, then increased to 5 mg/kg (200 mg BID) on hospital day 12 and maintained at this dose for 10 days before discontinuation. The patient was fully weaned off sedation by hospital day 17. EEG continued to show occasional bilateral independent temporal epileptiform discharges until discontinuation on hospital day 20; however, electrographic seizures did not recur after day 18. He was evaluated by neuropsychology starting on day 20, showing quick recovery of mental capabilities. He was discharged to inpatient rehabilitation on hospital day 35 on an ASM regimen of levetiracetam, lacosamide, and phenobarbital, as well as a prednisone taper.
Concerning the workup for etiology of his presentation, cerebrospinal fluid (CSF) studies on arrival to the ICU were as follows: RBC 2,000, WBC 7 (lymph 47%, Neu 43%, and Mono 10%), and protein 54, glucose 84 (serum glucose 124). Herpes simplex virus (HSV) and meningoencephalitis panel could not be obtained due to a traumatic tap. Follow up CSF studies obtained on hospital day 2 revealed; RBC 350, WBC 22 (lymph 53%, Neu 38%, and Mono 9%), protein 137, glucose 87 (serum glucose 175), and studies obtained 3 days later showed; RBC 515, WBC 2 (lymph 87%, Neu 0%, and Mono 13%), protein 51, glucose 53 with all additional tests negative including West Nile, CMV, cryptococcus, HSV-1/-2, HHV-6, enterovirus, E.Coli, parechovirus, listeria, neisseria, VZV, VDRL, lyme, coccidioides, cysticercosis, and autoimmune panel (NMDA, AMPA, GABA-B, LGI1, CASPR2, DPPX, mGluR1, GFAP, ANNA-1, ANNA-2, ANNA-3, CRMP-5, GAD65, PCA-2, PCA-1, PCA-Tr, AGNA-1, amphyphysin, IgLON5, and NIF IFA). Serum testing was likewise unrevealing with negative results including Q-fever, bartonella, HIV, ACE, thyroglobulin antibodies, and serum autoimmune panel (same as CSF).
MRI was unrevealing on admission, however repeat MRI on day 29 showed T2 hyperintensity within bilateral mesial temporal lobes (Fig. 4A). Subsequent repeat MRI 2.5 months later showed resolution of these findings (Fig. 4B).
On 3-month neuropsychology follow up assessment, he was not yet planning to return to teaching, as he was continuing to report issues with short-term memory and concentration. Testing revealed impairments in attention, timed visual scanning, cognitive flexibility, and verbal memory. Epilepsy clinic follow up 3 months and 2 weeks from initial hospitalization noted one episode of visual aura and one episode of word finding difficulty with no further events concerning for seizures. Two weeks later, tracheostomy decannulation was complicated by a post-operative seizure. He remained seizure free since this episode on 6 months follow up; however, 5 days after the 6-month anniversary of his hospital admission, he had a break-through seizure lasting 60–90 seconds described as patient becoming stiff and leaning on his wife; patient was amnestic of event. Further brief breakthrough seizures described as behavioral arrest with motor manifestations occurred at 10 months and 24 months post hospitalization with no further seizures at 28 months. By about 13 months follow up, he returned to teaching.
Discussion
This case describes a 47-year-old male who presented with new onset refractory status epilepticus without a clear etiology and associated with prodromal fever, headache, and gastrointestinal distress. This presentation meets the clinical definition of FIRES, though it is unique in that most adults who present with FIRES are younger and female.1 Aside from meeting the clinical definition, other aspects of this case that are more typical for an adult FIRES presentation include the lack of etiology, seizure semiology (focal aware facial twitching), and imaging findings implicating mesial temporal/limbic structures.1
His remarkable clinical recovery should be viewed in light of the well documented devastating course of this disease.1,3 Studies show that over 90% of those who survive into the chronic phase of the disease develop refractory epilepsy.1,3 The need for induction of a barbiturate coma and exposure to high number of anesthetics are both associated with a worse cognitive outcome.1 In this case report, three different anesthetic continuous infusions and multiple ASMs failed to establish adequate seizure control. Although our patient continues to deal with refractory epilepsy, he maintained seizure freedom for over a year and has regained cognitive ability to a significant extent in that he has returned to teaching. These results provide support for a therapeutic response to anakinra.
Only one prior report describes an anakinra-responsive FIRES case in an adult,10 in which a 21-year-old first received the medication on hospital day 32 and achieved seizure cessation in 24 hours. We were able to initiate anakinra more quickly, on hospital day 9. This was within the recommended guidelines of initiation of therapy 1–2 weeks after initial seizure presentation.2 However, our patient’s acute response was not as robust. It is unclear what factors could account for this differing response. The dose in the prior adult study was initiated at 300 mg daily for over 2 weeks, similar to our initiation of 300 mg daily for 3 days, then increased to 5 mg/kg/day (~400 mg). Perhaps there is some effect of prolonged IL-1 activation that renders the inflammatory pathway more susceptible to blockade after continuous cytokine release.
Another significant consideration is dose duration. In comparison to our patient’s anakinra therapy duration of 2 weeks, the prior adult report maintained the patient on therapy through 6 month follow up. Although it is unclear if there is any direct attribution of this difference to therapeutic benefit, the prior patient experienced a faster chronic phase recovery, and reportedly achieving baseline cognitive function as well as seizure freedom at 6 months follow up.
Little is currently known regarding ideal initiation dose or therapy duration with anakinra in FIRES. No level 4 or higher evidence has been able to provide standardized treatment guidelines for FIRES due to the rare nature of the disease. The incidence is believed to be higher in pediatric patients, where it is estimated to be about 1:1,000,000.1,2 Multiple reviews studying the epileptogenic quality of CNS inflammation and its presumed pathogenesis in FIRES have provided recommendations for treatment that includes anakinra initiation within 1–2 weeks of RSE onset.1,2 Anakinra dosing has been studied in adults in two separate phase II trials for stroke and traumatiic brain injury with patients tolerating doses as high as 2 mg/kg/hour over 72 hours without any noted adverse reaction.5,6 Doses as low as 100 mg daily continued to show neuroinflammatory response in the CNS.5
A few pediatric studies have also reported on anakinra efficacy in FIRES, though response has been less robust than in our case. The first pediatric case to show efficacy reported on a 32-month-old who received the medication with good response in the acute phase and was continued on 10 mg/kg/day.7 This patient was unable to wean anakinra at 6 months due to refractory seizures. In another report anakinra was utilized in two separate patients, before or after centromedian nucleus deep brain stimulation to treat FIRES. A better outcome was seen in the patient who received anakinra after deep brain stimulation (DBS) (43 days after initial seizure presentation) with continued short focal seizures in the chronic phase, while the patient who received anakinra prior to DBS (22 days after initial seizure presentation) remained in a vegetative state.8 In the final case report, anakinra was dosed similarly to our case presentation, though for a longer duration, with initial 3-day 2.5 mg/kg/day dosing followed by 5 mg/kg/day for 7 months.9 During this period the patient suffered two clinical seizures and the clinical course was further complicated by hospital admission for status epilepticus 11 months after IL-1 inhibitor cessation.
Looking specifically at the efficacy of anakinra in the chronic phase of the disease, a retrospective cohort study included five pediatric patients who received anakinra during the chronic phase, a median interval of 7 years from the acute phase.11 All patients initially received a dose of 100 mg daily with further titration offering no observable benefit. Therapy was continued for a median duration of 9 months. Ultimately, most patients showed 20–50% seizure reduction, improvement in seizure intensity, reduced need for rescue medications, and improved behavior/communication. There were no significant adverse effects to motivate discontinuation of anakinra in any of the study patients.
In the limited number of reports described above, there is evidence for a robust therapeutic effect of anakinra in both the acute and chronic phases of FIRES despite wide variations in dose, timing of therapy initiation, and duration of therapy. The best response seen was in the young adult, where seizures abated at anakinra initiation, and seizure freedom was noted at 6 months follow up as was complete return to prior cognitive function. Our patient is the oldest discussed in the literature and also returned fully to his cognitive baseline, though after a more prolonged recovery period and continues to have few breakthrough seizures. Pediatric cases developed refractory epilepsy with some evidence for cognitive improvement. It is yet to be determined if there is a clearly demarcated response to anakinra based on age given the few cases discussed. A larger scale investigation may be able to determine alternate variables responsible for individual long-term outcomes to anakinra, and ultimately a clearer understanding of the pathogenesis of FIRES.
This disease process has been shown to involve a pathophysiological connection with CNS inflammation.1–4 Among acute phase reactants, IL-1 is a powerful cytokine with local and systemic downstream effects that have been implicated in the process of epileptogenesis in animal models.4 CSF and serum analysis in drug resistant epilepsy patients and in animal models showed reduced expression of intracellular IL1-RA isoforms in comparison to expression of the molecule it inhibits, IL-1β.3,4 This indicates that the ratio IL1-RA:IL-1β has an impact on the seizure threshold. Another area of variability to consider is that there is not only a quantitative deficiency in IL1-RA, but also a functional deficiency in IL1-RA inhibitory activity shown in these studies.3 One potential source for the development of this susceptibility lies in variability of the RN2 allele of IL1-RN.2
An alternative cytokine pathway has been targeted with tocilizumab, an IL-6 antagonist. Evidence for this pathway has been provided with CSF findings of massively increased IL-6 levels in FIRES patients, 3 and while tocilizumab has been found effective in lowering these levels, anakinra has also been shown to normalize these values.3,7 This is not unexpected as anakinra acts antagonistically at the IL-1 receptor and IL-6 is one of the downstream cytokines activated by IL-1. Tocilizumab has also been effective in attaining seizure freedom.1,2 In some cases, this has been when administered following anakinra use.11 This would seem to indicate that tocilizumab has treatment effect in an alternative stream distinct from anakinra. In consensus review of the data, it was ultimately recommended that tocilizumab be considered only when anakinra is ineffective due to less evidence backing its use, decreased CSF penetration, and potentially greater side effects related to infection.2 More thorough analysis of FIRES patients’ serum and CSF markers, as well as genetic variation in cytokine-related genes may lead to a clearer understanding of this disease process and the mechanisms underpinning the variable response to therapeutic intervention in different individuals.
In conclusion, this report focuses on a unique presentation of FIRES that is responsive to IL-1 antagonist, anakinra. Notably, this patient is the oldest reported FIRES patient treated with anakinra. This is also the first case report discussing short-term anakinra use. Our patient experienced a remarkable therapeutic response to an IL-1 antagonist after failure of multiple conventional anti-seizure medications and anesthetic medications. His chronic phase recovery has been characterized by return to his cognitive baseline and few breakthrough seizures separated by a year of seizure freedom. These outcomes provide additional support for therapeutic recommendations offered in reviews for the treatment of FIRES, including a therapeutic response to anakinra. Further studies with a focus on determining the underlying mechanism that accounts for variability in patient response to anakinra is essential. Such studies may aid in the development of ideal dosing and therapy duration of anakinra in FIRES.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
