We retrospectively assessed all patients in a large cohort of patients with epilepsy surgery at the National Hospital for Neurology and Neurosurgery (NHNN) over 12 years, to identify patients with postoperative psychogenic nonepileptic attacks (PNEA). Twenty-nine patients (23 women) were identified of a total of 790 patients, a frequency of 3.7%. Female gender and presurgical psychiatric diagnosis, other than psychosis, were significant risk factors for PNEA development. In female patients with a preoperative psychiatric diagnosis the chance of developing PNEA after epilepsy surgery was 8.5%. PNEA developed between 2 weeks and 10 years after epilepsy surgery, independently of outcome of epileptic seizures. In most cases, PNEA differed from the present or past epileptic seizures, and motor symptoms were the most common manifestations. Seizures after epilepsy surgery should be carefully evaluated. Physicians should consider the possibility of PNEA, especially in female patients with preoperative psychiatric comorbidity developing “atypical” seizures with motor manifestations postoperatively, even many years after epilepsy surgery.
Psychiatric symptoms, especially emotional liability and depression, anxiety, and/or psychotic symptoms, may occur after epilepsy surgery (Foong & Flugel, 2007). Psychogenic nonepileptic attacks (PNEAs) may develop as part of this psychiatric instability that may follow any resective surgery, such as intracranial neurosurgery for indications other than refractory epileptic seizures (Reuber et al., 2002); of interest are the reports of nonneurosurgical surgery cases developing postanesthetic PNEA (Parry & Hirsch, 1992; Ng & Chambers, 2003; Allen & Farling, 2004; Lichter et al., 2004) as a result of operation anxiety and the altered state induced by anesthetic agents (Lichter et al., 2004).
PNEAs, developing de novo after epilepsy surgery, have not been well studied. Two case reports and four retrospective series have been published with a total of 44 patients with postsurgical PNEA (Ferguson & Rayport, 1965; Krahn et al., 1995; Ney et al., 1998; Parra et al., 1998; Glosser et al., 1999; Montenegro et al., 2000). In one more recent postsurgical series, patients after anterotemporal lobectomy were screened for postoperative PNEA (Kanner et al., 2009).
The present study is an evaluation based on chart review of a surgical cohort that was carefully followed. We retrospectively identified the patients who developed postoperative PNEA, estimated the risk factors, and identified PNEA features.
Patients and Methods
In this retrospective chart review, we used the long-term outcome data of 805 patients, with a period of follow-up of 6,000 patients-years, as described previously (de Tisi et al., 2011). Patients had epilepsy surgery at the National Hospital for Neurology and Neurosurgery (NHNN), London, United Kingdom, from February 15, 1990, until March 15, 2012.
This study was approved by the Joint Ethics Committee of the National Hospital for Neurology and Neurosurgery and UCL Institute of Neurology. The Research Ethics Committee classified this work as a service evaluation; therefore, individual consent from patients was not required.
All clinical and investigational data were reviewed. Comprehensive neuropsychological assessments, including cognitive testing, were carried out in all patients preoperatively, and also 3 and 12 months postoperatively. Psychiatric diagnoses preoperatively and postoperatively were made by an experienced neuropsychiatrist, and written documentation was recorded, according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) Axis I diagnoses (American Psychological Association, 2000).
Patients had regular follow-up, and patients and witnesses were questioned in depth about any seizures occurrence, their description, and whether these differed from preoperative seizures. Postoperative PNEAs were diagnosed by experienced clinicians based on interviews, records review, and video–electroencephalography (EEG) telemetry.
Patients with preoperative and postoperative PNEAs were excluded. Patients with two surgical procedures were included in the study at the first surgical operation and the second operation was censored.
Statistical analysis was undertaken using SPSS 20.0 for (IBM Corporation, Armonk, NY, U.S.A.) Windows.
Seven hundred ninety patients were evaluated. One patient with preoperative and postoperative PNEA was excluded, and only the first operation was considered in 14 patients having epilepsy surgery on two occasions.
Twenty-nine patients (3.7%) developed postsurgical PNEA; 23 of the patients were female (79%). Twenty-six patients had PNEA diagnosed on video-EEG, and in three patients PNEA were diagnosed by neurologist' s review of home video recording of episodes or by witnessing episodes directly and comparing with preoperative video-EEG.
PNEA and non-PNEA patients' data are provided in Table 1.
Table 1. PNEA and non-PNEA data
CI, confidence interval; NS, not significant; OR, odds ratio.
Duration of epilepsy (years, range)
Epilepsy onset age (years, range)
Age at surgery (years, range)
Preoperative psychiatric diagnosis (%)
Preoperative depression (%)
Preoperative psychosis (%)
Preoperative psychiatric disorder other than depression and psychosis (%)
A preoperative psychiatric diagnosis was associated with increased risk of PNEA onset (Table 1). Further analysis showed that preoperative depression and the presence of other, than psychosis and depression, disorders was associated with increased risk of postoperative PNEA. These psychiatric disorders included anxiety (six patients), history of abuse (two patients), personality disorders (two patients), and anorexia (one patient), as recorded in history. Further analysis for each of these disorders was not conducted because of the small numbers of cases in the PNEA group, the absence of systematic screening tools to assess these disorders in both groups, and the comorbidity with depression (seven cases) and psychosis (one case). Preoperative psychosis, either postictal or interictal, was not associated with an increased risk of PNEA.
There was a trend toward a higher proportion of surgery complications in the PNEA group, including transient postoperative issues, such as severe headache and infections, and permanent complications, such as hemiparesis and quadrantanopia.
Further multivariate regression analysis showed that the female gender (p=0.015) and a preoperative psychiatric disorder (p=0.01) are associated with increased risk of PNEA. Consequently, a female patient with a preoperative psychiatric disorder had a 8.5% chance of developing PNEA (p=0.0002).
In the postoperative period, 19 PNEA patients had comorbid psychiatric disorders. No statistical difference was found for any postoperative psychiatric disorder, other than PNEA, between the two groups (Table 1).
The postoperative epileptic seizure outcome was not associated with PNEA onset. At the time of PNEA onset, 26% of PNEA patients were epileptic seizure free, 17.4% had simple partial seizures, and 39% patients were classified as Epileptic Seizures Outcome Classification 4, during the year preceding PNEA onset (Table 2), according to International League Against Epilepsy (ILAE) surgery outcome scale (Wieser et al., 2001).
Table 2. Epileptic seizure outcome classification in PNEA patients and PNEA features
Epileptic seizure outcome
Classification (SOC) at time of PNEA onset
SOC1 (seizure free): 26%
SOC2 (SPS only – no other seizures): 17.4%
SOC3 (Seizures on <4 days per year, with or without SPS): 8.8%
SOC4 (>50% reduction in numbers of days affected by seizures): 39.1%
SOC5 (No significant change in seizures with up to 50% reduction ranging to 100% increase in days affected by seizures): 8.7%
SOC6 (>100% increase in days affected by seizures): none
PNEA development (months)
Daily – per 4 months
Motor symptoms (16)
Falling to the floor still (4)
Staring and unresponsiveness (2)
Anxiety and panic attacks (2)
Psychic symptoms (2)
Similarity to epileptic seizures
2 lost to follow-up
PNEA developed in 2 weeks to 10 years after epilepsy surgery. The frequency ranged from daily attacks to one attack every 4 months. The most common clinical manifestations were motor manifestations, followed by episodes of falling to the floor and being still. In 24 patients (83%), PNEA had a character different from that of the habitual epileptic seizures (Table 2).
In the present series, 78% of PNEA patients accepted psychiatric input. In six patients (21%) PNEA stopped permanently (five of these received psychiatric therapy). Seventeen patients continued having PNEA; two patients were lost to follow up, and in four patients it was uncertain whether PNEA had stopped (Table 2).
In our cohort, postoperative PNEA developed at a frequency of 3.7%. We acknowledge the possibility of some patients having gone undiagnosed, particularly if PNEAs were self-limiting. Krahn et al. (1995) identified 2 patients of 700, whereas in the other series the frequency ranged from 1.8% to 8.8% (Ney et al., 1998; Parra et al., 1998; Glosser et al., 1999; Table S1). The highest frequency is reported by Glosser (8.8%), but preoperative PNEA was suspected in some cases, raising the possibility of a lower true postoperative PNEA frequency (Glosser et al., 1999).
As in a nonsurgical context, female patients appear more vulnerable to developing PNEA (McKenzie et al., 2010). Female gender is the most prominent risk factor for postoperative PNEA, when all cases reported in the literature are amalgamated, followed by the presence of preoperative psychiatric disorder (Table S1).
Krahn et al. (1995) and Ney et al. (1998) found longer epilepsy duration in PNEA patients, whereas Glosser et al. (1999) reported a later epilepsy onset age. In our study, no difference in epilepsy duration and epilepsy onset age was found. Noteworthy, all PNEA patients had epilepsy for >6 years.
PNEA development was not related to postoperative psychiatric disorders. Under the light of these findings we could potentially interpret the development of PNEA as a manifestation of psychosocial distress in vulnerable individuals. In addition, the fact that physical complications from surgery were almost significantly more prevalent in the PNEA group may have implications in the etiology of the disorder for at least a subgroup of patients.
In the nonneurosurgical surgery PNEA literature, PNEA are reported to ensue in response to anxiety associated with having an unexpected and alarming complication of the operation (Ng & Chambers, 2003; Lichter et al., 2004).
Health related trauma has been identified in PNEA patients (Duncan & Oto, 2008) which, coupled with poor seizure outcome, may contribute to helplessness and reinforce illness beliefs that increase vulnerability toward PNEA.
It has been reported that most PNEA cases emerge within 6 months (Glosser et al., 1999), but this is not the case in our cohort, where two thirds of patients developed PNEA after the first year following epilepsy surgery and one patient developed PNEA 10 years after surgery.
Postoperative PNEA does not often resemble the habitual epileptic seizures (Parra et al., 1998). In our study, PNEA were dissimilar to the past or present epileptic seizures in most cases, a dissimilarity further shown by previous literature (Table S1). In accordance with postanesthetic “shaking seizures” in nonneurosurgical surgery PNEA cases (Parry & Hirsch, 1992; Ng & Chambers, 2003; Allen & Farling, 2004), PNEAs were manifested, in most cases of the present study, as motor “seizures” (Table 2), whereas panic attacks were the most common “seizures” in other series (Table S2).
There are certain limitations to the present study. Axis II diagnoses, although assessed, were not consistently recorded. No formal scales were used to assess personality. No screening tools were used for trauma or abuse, so this was difficult to quantify or analyze. Although we infer PNEA to be a manifestation of psychosocial distress in vulnerable individuals, we acknowledge that we did not have comprehensive psychosocial measures. In addition, the influence of new stressors in this period of 2 weeks to 10 years, which could potentially be related to PNEA development, was not assessed and not reported.
Conclusively, de novo PNEA after epilepsy surgery is underdiagnosed and may contribute to the reported recurrence rate of seizures after epilepsy surgery. Seizures after epilepsy surgery should be carefully evaluated. The development of “atypical seizures” with motor manifestations, especially in female patients with presurgical psychiatric disorders, independently of epileptic seizures outcome, may represent PNEA.
This work was carried out at University College London Hospitals/UCL Biomedical Research Centre, which receives partial funding from the Department of Health's NIHR Biomedical Research Centre, and at Epilepsy Society Chalfont St Peter, Buckinghamshire, United Kingdom. Thanks to all the members of the multidisciplinary Epilepsy surgery team at Queen Square.
The authors have no conflicts of interest. We confirm that we have read the Journals position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Sofia Markoula Consultant Neurologist at Ioannina Hospital, Greece. Honorary Clinical Assistant at Clinical and Experimental Epilepsy Department, NHNN, United Kingdom.