The Scope of the Unmet Medical Need

PNES occur worldwide, and in the United States up to 20% of civilians and up to 25% of Veterans diagnosed as having epilepsy actually have PNES,Reference Salinsky, Evrard, Storzbach and Pugh ⁵ making PNES as common as multiple sclerosis and Parkinson disease.Reference Hirtz, Thurman, Gwinn-Hardy, Mohamed, Chaudhuri and Zalutsky ⁶ Comprehensive epidemiological investigations are lacking in PNES; however, the estimated prevalence of PNES is up to 33/100,000 individuals.Reference Asadi-Pooya and Sperling ⁷ The incidence rate from a prospective study of 367,566 individuals followed over 3 years found that new-onset PNES was observed in 4.9/100,000/year.Reference Duncan, Razvi and Mulhern ⁸ As is typical for PNES, individuals were 80% women with a mean age of onset of 31 (±15) years. The majority of individuals with PNES were unemployed, and 46% were previously diagnosed with anxiety or depression; 57% of patients also had at least one additional medically unexplained symptom (MUS), highlighting the overlap between PNES and other FNDs. The coincidence of PNES with ES is approximately 10%.Reference Duncan and Oto ⁹

The economic and healthcare utilization costs incurred in caring for PNES are substantial. As one illustrative example, 15 of 28 subjects with PNES accounted for 34 emergency room visits, and 12 of 28 individuals accounted for 66 inpatient hospital days pre-diagnosis.Reference Razvi, Mulhern and Duncan ¹⁰ In a retrospective study from Ireland, the annual estimated cost per patient with undiagnosed PNES was 20,995 euros.Reference Magee, Burke, Delanty, Pender and Fortune ¹¹ While healthcare expenditures decrease substantially post-diagnosis,Reference Razvi, Mulhern and Duncan ^{10 ,} Reference Magee, Burke, Delanty, Pender and Fortune ¹¹ multiyear delays in diagnosis following symptom onset, continued PNES events, unemployment, and medical disability perpetuate healthcare and economic costs.

Further compounding the problem, many healthcare providers report frustration when caring for patients with PNES. In a semistructured interview of 75 healthcare clinicians and workers in the Veterans Health Administration system, while some providers endorsed sentiments of hope, many expressed frustration over the complexity of PNES, which included difficulties with patient acceptance of the diagnosis, uncertainties in treatment interventions, and a failure of cross-disciplinary collaborations.Reference McMillan, Pugh and Hamid ¹² While neurologists increasingly recognize the scope of the unmet medical need, many remain uncomfortable discussing somatoform symptoms with patientsReference Monzoni, Duncan, Grunewald and Reuber ¹³ ; psychiatrists and other mental healthcare providers may also potentially feel ill-equipped to treat individuals with PNES.

Semiologic Differentiation of PNES from ES and Diagnostic Criteria

Differentiating PNES from ES can be clinically challenging, and several studies have investigated clinical signs that aid the diagnosis of PNES.Reference Devinsky, Gazzola and LaFrance ^{1 ,} Reference Avbersek and Sisodiya ¹⁴ A retrospective review detailing signs that reliably distinguished PNES from ES suggested that a diagnosis of PNES was favored for events showing a long duration, a fluctuating course, asynchronous or side-to-side movements, ictal eye closure at onset, ictal crying, and post-ictal recall of information when presented ictally (see Table 1). In addition, urinary incontinence and tongue biting do not reliably distinguish between ES and PNES.Reference Brigo, Nardone and Ausserer ¹⁵ A prospective study of 120 seizures in 35 consecutive subjects showed that video-documented preserved awareness, eye fluttering, and the modulation of event intensity by bystanders reliably predicted PNES; abrupt onset, ictal eye-opening, and post-ictal confusion/sleep reliably predicted ES.Reference Syed, LaFrance and Kahriman ¹⁶ It is also worth noting that apart from differentiating PNES from ES, additional diagnostic considerations should also be evaluated and ruled out, including paroxysmal movement disorders, panic attacks, and physiologic forms of nonepileptic events including cardiac arrhythmias among other conditions.

Table 1

Semiologic features that support the diagnosis of psychogenic nonepileptic seizures (PNES) vs. epileptic seizures (ES).

* Indicates that sign may not reliably differentiate between PNES and frontal lobe partial epileptic seizures.

** Indicates that sign may only be helpful in distinguishing convulsive PNES and ES.

Adopted from Avbersek and SisodiyaReference Avbersek and Sisodiya ¹⁴ with permission.

While semiological features may suggest a diagnosis of PNES over ES, the gold standard for diagnosis is the recording of typical event(s) on video electroencephalography (vEEG) and noting a lack of epileptiform activity in the peri-ictal period, with semiologic and historical consistency with PNES. While this is the diagnostic work-up of choice, access to vEEG may not be available, and some patients may have sufficiently low event frequencies that hospital admission for long-term monitoring is impractical. Recently, the International League Against Epilepsy (ILAE) Commission on Neuropsychobiology Nonepileptic Seizures Task Force published consensus guidelines on the minimum requirements to diagnosis PNES (see Table 2).Reference LaFrance, Baker, Duncan, Goldstein and Reuber ¹⁷ Levels of diagnostic certainty are based on the clinical history, ictal semiology of an event, and EEG findings. To meet criteria for documented PNES, the clinical history should be consistent with PNES and habitual events must be recorded on vEEG demonstrating the absence of epileptiform activity. Lower levels of certainty (clinically established, probable, possible) are based on availability of diagnostic components.

Table 2

Diagnostic levels of certainty for psychogenic nonepileptic seizures

+, history characteristics consistent with PNES; PNES, psychogenic nonepileptic seizures; ES, epileptic seizures; EEG, electroencephalogram.

* Captured ictus should not resemble types of ES which thatmay not show ictal epileptiform correlate on EEG (e.g., simple partial epileptic seizures).

Modified from Ref. Reference LaFrance, Baker, Duncan, Goldstein and Reuber17 [LaFrance WC Jr, Baker GA, Duncan R, et al. Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force. Epilepsia 2013;54(11):2005—2018]. With permission.

Given the clinical need to clarify a diagnosis of PNES vs. ES, whenever possible, seizure induction protocols have drawn increasing interest. While ethical concerns have been raised regarding provocation techniques, given that they may be perceived as misleading, studies suggest that use of conventional EEG activation procedures, including hyperventilation and photic stimulation, may be effective adjunctive techniques in the diagnosis of PNES,Reference Hoepner, Labudda, Schoendienst, May, Bien and Brandt ¹⁸ making provocation unnecessary. Providing correct and explicit information about the potential utility of standard induction protocols may alleviate ethical concerns and aid diagnostic evaluations. Surface electromyography recordings may also potentially aid differentiation of convulsive ES and PNES events.Reference Beniczky, Conradsen and Moldovan ¹⁹ Approximately twice normal serum prolactin levels drawn 10–20 minutes following an ictal event can also help differentiate convulsive ES from PNES.Reference LaFrance, Baker, Duncan, Goldstein and Reuber ¹⁷

The Emerging Neurobiology of PNES

Systems-level functional and structural neuroimaging studies have started to elucidate the neurobiology of PNES.Reference Perez, Dworetzky and Dickerson ³ . Several functional magnetic resonance imaging (fMRI)Reference van der Kruijs, Bodde and Vaessen ^{43 –} Reference Li, Liu and Ma ⁴⁷ and one positron emission tomography (PET) study used resting state techniques to investigate neural circuit disturbances in PNES. In a region-of-interest “seed”-based analysis, increased functional connectivity was observed between motor regions (precentral sulcus) and regions involved in emotional processing [anterior cingulate cortex (ACC), insula] and executive functions (inferior frontal gyrus, parietal cortex); among other findings, trait dissociation scores positively correlated with the functional connectivity strength between the precentral sulcus and the posterior insula.Reference van der Kruijs, Bodde and Vaessen ⁴³ Several studies used data-driven, multivariate functional connectivity analyses to demonstrate widespread functional connectivity alterations in emotion control, executive, fronto-parietal (attentional), sensorimotor, and default mode networks.Reference van der Kruijs, Jagannathan and Bodde ^{44 –} Reference Ding, An and Liao ⁴⁶ Another study specifically evaluated functional connectivity patterns across insular subregions, observing that PNES compared to healthy subjects showed increased functional connectivity between the left ventral anterior insula and the left post-central gyrus and bilateral supplementary motor area (SMA); functional connectivity strength within the bilateral SMA in relation to the left ventral anterior insula positively correlated with PNES event frequency.Reference Li, Liu and Ma ⁴⁷ Functional connectivity strength between the SMA and ACC has also been shown to positively correlate with PNES frequency.Reference Li, Li, An, Gong, Zhou and Chen ⁴⁸ In comparison to healthy subjects, a 2-deoxy-2-[fluorine-18]fluoro-D-glucose PET study showed that PNES patients exhibited bilateral ACC and right inferior parietal lobule hypometabolism.Reference Arthuis, Micoulaud-Franchi, Bartolomei, McGonigal and Guedj ⁴⁹

Quantitative structural MRI comparisons have also been conducted in PNES.Reference Labate, Cerasa and Mula ^{50 –} Reference Ristić, Daković, Kerr, Kovačević, Parojčić and Sokić ⁵² Compared to matched controls, patients with PNES showed ACC, SMA, and pre- and post-central gyrus atrophy on cortical thickness and voxel-based morphometry analysesReference Labate, Cerasa and Mula ⁵⁰ ; bilateral precentral gyrus cortical thickness reductions were independently replicated, along with observed increases in insular and orbitofrontal cortical thickness in PNES.Reference Ristić, Daković, Kerr, Kovačević, Parojčić and Sokić ⁵² A diffusion tensor tractography study showed a right-ward asymmetry of the uncinate fasciculus (a tract connecting medial prefrontal and medial temporal regions) in PNES compared to healthy controls.Reference Hernando, Szaflarski, Ver Hoef, Lee and Allendorfer ⁵¹ While functional and structural neuroimaging studies require further replication and control of potential confounding variables, these early results suggest that PNES may manifest in the context of alterations within and abnormal interactions across sensorimotor, emotion regulation/processing, cognitive control, and multimodal integration neural circuits (see Figure 1 for plots of peak coordinates in the above discussed functional and structural neuroimaging studies). Research is needed to evaluate the specificity of PNES-related neural circuit alternations in relation to neurologic (ie, epilepsy, TBI) and psychiatric (ie, PTSD, depression, other functional neurological disorders) control groups. Additional research efforts are also needed to investigate if semiologic differences in PNES (ie, convulsive vs. atonic events) are linked to distinct patterns of neural circuit alterations.

Figure 1

Display of volumetric and resting-state functional connectivity (FC) findings in psychogenic nonepileptic seizures (PNES) neuroimaging studies. The left panel displays peak coordinates of cortical thickness and voxel-based morphometry findings in PNES vs. healthy controls.Reference Labate, Cerasa and Mula ^{50 ,} Reference Ristić, Daković, Kerr, Kovačević, Parojčić and Sokić ⁵² The right panel shows peak coordinates of abnormal resting-state FC in individuals with PNES vs. healthy controls.Reference van der Kruijs, Bodde and Vaessen ^{43 ,} Reference Ding, An and Liao ^{45 ,} Reference Li, Liu and Ma ^{47 ,} Reference Li, Li, An, Gong, Zhou and Chen ⁴⁸ Blue circles indicate decrease and red circles indicate increase. Evidence from neuroimaging studies, while requiring replication and further investigations, suggests that PNES may potentially develop in the context of alterations within and across brain networks implicated in sensori-motor (eg, pre- and post-central gyrus, premotor regions), emotion processing/regulation [eg, anterior cingulate cortex (ACC), middle cingulate cortex (MCC), orbitofrontal cortex, insula], cognitive control (eg, lateral prefrontal cortex, dorsal ACC, MCC), and multimodal integration functions (eg, cingulate gyrus, posterior parietal cortex, precuneus). Only main findings are displayed using Caret 5. Cerebellar and basal ganglia foci, and seed region coordinates are not shown.

Electrophysiology and autonomic nervous system studies have also contributed to the understanding of PNES pathophysiology. In a cohort of 18 PNES and 18 healthy subjects, resting state high-density source EEG analyses showed that patients with PNES exhibited decreased functional connectivity between the basal ganglia and cortical regions, along with reduced interhemispheric connectivity across paralimbic regions.Reference Barzegaran, Carmeli, Rossetti, Frackowiak and Knyazeva ⁵³ In a unique case of intracranial recordings during PNES, decreased power in the theta band was observed over the posterior parietal cortex.Reference Arzy, Halje, Schechter, Spinelli, Seeck and Blanke ⁵⁴ Several investigations have probed autonomic profiles peri-ictally in PNESReference Reinsberger, Sarkis and Papadelis ⁵⁵ ; some noted increased sympathetic tone ictally in ES vs. PNES,Reference Ponnusamy, Marques and Reuber ⁵⁶ while others characterized pre-ictal and post-ictal autonomic changes as distinguishing between groups.Reference Reinsberger, Perez, Murphy and Dworetzky ⁵⁷

A few studies have investigated serologic biomarkers in PNES. Positive correlations were observed between baseline salivary cortisol levels and negative attentional bias during performance of a masked emotional Stroop taskReference Bakvis, Spinhoven and Roelofs ⁵⁸ ; in this same cohort, negative attentional bias also correlated with past sexual trauma.Reference Bakvis, Roelofs, Kuyk, Edelbroek, Swinkels and Spinhoven ⁵⁹ Increases in basal diurnal cortisol levels in PNES compared to healthy controls were mediated mainly by patients reporting past sexual trauma.Reference Bakvis, Spinhoven and Giltay ⁶⁰ Reductions in serum brain derived neurotrophic factor, implicated in synaptic reorganization and neurogenesis, have also been reduced in samples with PNES and with ES, compared to healthy controls.Reference LaFrance, Leaver, Stopa, Papandonatos and Blum ⁶¹

Treatment and Outcome Studies

In the past decade, significant progress has been made in the evidence-based management of PNESReference LaFrance, Reuber and Goldstein ⁶² (see Table 3). In a pilot, randomized, controlled trial of 12-weeks of cognitive-behavioral therapy (CBT) plus standard medical care (SMC) vs. SMC alone, conducted in 66 patients with PNES, intention to treat analyses showed a statistically significant reduction in seizure frequency at the end of the treatment period associated with CBT.Reference Goldstein, Chalder and Chigwedere ⁶³ At a 3-month follow-up, however, mean seizure frequency between the treatment arms did not remain statistically significant. A pilot, double-blind, randomized, placebo-controlled trial of flexible-dose sertraline, conducted over 12 weeks, showed that PNES subjects assigned to the sertraline treatment displayed a 45% reduction in seizure frequency as compared to an 8% seizure frequency increase to placebo.Reference LaFrance, Keitner and Papandonatos ⁶⁴ Recently a multicenter, randomized clinical trial was conducted in 38 patients with PNES comparing 4 treatment arms: flexible-dose sertraline only, CBT-informed psychotherapy (CBT-ip) only, CBT-ip with flexible-dose sertraline, and SMC.Reference LaFrance, Baird and Barry ⁶⁵ Within-group analyses for each group showed that the 2 psychotherapy treatment arms exhibited a significant reduction in seizures (51% in the CBT-ip only arm and 59% in the CBT-ip plus sertraline arm), while the sertraline and the SMC arms did not show a significant reduction in seizures. The CBT-ip only group also showed significant improvements in depression, anxiety, quality of life, and global functioning. These findings support the use of a manualized CBT-ip PNES workbook.Reference Reiter, Andrews, Reiter and LaFrance ⁶⁶ The intervention is being disseminated for treatment by a range of clinicians, from epileptologists/neurologists, psychiatrists, psychologists, and social workers.Reference LaFrance and Wincze ⁶⁷

Table 3

Summary of controlled clinical trials in the treatment of psychogenic nonepileptic seizures (PNES).

CBT, cognitive behavioral therapy; CBT-ip, cognitive behavioral therapy–informed psychotherapy; SMC, standard medical care; AED, anti-epileptic drug; Sz, seizures; f/up, follow-up; ↓ - reduction in.

A randomized, controlled trial of a post-PNES diagnosis group psychoeducation program (3 successive monthly, 1.5-hour group sessions) also showed that patients in the psychoeducation intervention reported improvements in perceived psychosocial functioning and a trend toward decreased seizure-related emergency room visits or hospitalizations; however, seizure frequency reduction was not observed.Reference Chen, Maheshwari, Franks, Trolley, Robinson and Hrachovy ⁶⁸ Last, a randomized controlled trial of immediate vs delayed withdrawal of antiepileptic drugs (AED) in PNES showed that immediate AED withdrawal was associated with a reduction in seizures, decreased use of rescue medications, and greater perceived internal locus of control.Reference Oto, Espie and Duncan ⁶⁹ The long-term benefits of therapy and other treatments can be examined in randomized, controlled, clinical trials along with other potential pharmacologic and psychotherapy treatment interventions.Reference Baslet, Dworetzky, Perez and Oser ^{70 ,} Reference Bullock, Mirza, Forte and Trockel ⁷¹

Outcome and predictor of outcome studies have been performed in PNES. Given that the initial stages of treatment include accurate diagnosis and effective communication of the diagnosis, it is noteworthy that diagnostic acceptance by patients improves outcomes.Reference Duncan, Graham and Oto ⁷² Effective communication alone, however, is insufficient to improve outcomes in most patients.Reference Mayor, Brown and Cock ⁷³ In a retrospective study of 260 consecutively evaluated PNES patients measuring outcome at 6 and 12 months post diagnosis, 18% showed an increase in seizures, 38% were event free, but the majority of patients continued to have seizures.Reference McKenzie, Oto, Russell, Pelosi and Duncan ⁷⁴ Women, individuals with anxiety/depression, and patients receiving social security benefits were less likely to become event-free. Post-diagnosis, many also developed other MUS functional neurological symptoms.Reference Jones, TJ and Adams ⁷⁵ In addition to psychosocial and psychiatric variables impacting outcome, patients with PNES with a history of TBI exhibit worse outcomes, as exemplified by increased likelihood of having major depression, behavioral impulsivity, PTSD, receiving disability, and lower global functioning.Reference LaFrance, Deluca, Machan and Fava ²³

Disclosures

David Perez has the following disclosure: National Institute of Neurological Disorders and Stroke, Researcher, Research grant (D.L.P., R25NS065743-05S1). W. Curt LaFrance has the following disclosures: Cambridge University Press, book co-editor, editor’s royalties; Oxford University, book co-author, author’s royalties; Matty Fund, researcher, research support; Veterans Administration, researcher, research support; Brown University, researcher, research support.