To present new information on the semiology and short-term evolution of seizures associated with primary brain tumors (PBTs) in a prospective study.
To present new information on the semiology and short-term evolution of seizures associated with primary brain tumors (PBTs) in a prospective study.
This study is a section of the PERNO study – Project of Emilia Romagna Region on Neuro-Oncology, the main aim of which is to collect prospectively all cases of PBTs occurring in the Emilia-Romagna region, northeast Italy (3,983,346 population) from January 2009 to December 2011, to allow epidemiologic, clinical, and biomolecular studies.
The epilepsy section of the PERNO study included all the patients who experienced seizures, either as first symptom of the tumor or appearing during the course of the disease. Each patient was interviewed by the referring neurologist with a specific interest in epilepsy. The patients who entered the study were followed up with visits on a quarterly basis.
We collected 100 cases with full clinical, neuroradiologic, and pathologic data. The majority (79%) had high grade PBTs (glioblastoma in 50 cases), whereas the remaining patients had low-grade gliomas, mostly localized in the frontal (60%), temporal (38%), and parietal (28%) lobes. Seizures were the first symptom of the tumor in 72 cases. Overall, the initial seizures were tonic–clonic (48%) (without clear initial focal signs in more than half of the patients), focal motor (26%), complex partial (10%), and somatosensitive (8%). The majority of cases (60%) had isolated seizures or a low seizure frequency at the onset of the disease, whereas a high seizure frequency or status epilepticus was observed in 18% and 12% of cases, respectively. Ninety-two patients underwent surgical removal of the tumor, which was either radical (38%) or partial (53%). Seven patients underwent only cerebral biopsy. In the 72 patients in whom seizures were the first symptom, the mean time to the surgical treatment was 174 days, with a significant difference between high grade (95 days) and low grade (481 days) gliomas.
At the time of our first observation, the majority of patients (69%) had already undergone surgical removal, with a mean follow-up of 3 months after the procedure. Overall, 39 patients (56%) were seizure free after tumor removal. The good outcome did not depend on presurgical seizure frequency or tumor type, although there was a trend for better results with low-grade PBTs.
These data provide evidence that seizures are strictly linked to the tumoral lesion: They are the initial symptom of the tumor, reflect the tumor location and type, are usually resistant to antiepileptic treatment, and may disappear after the treatment of the lesion.
Epilepsy is a common manifestation of primary brain tumors (PBTs). The frequency of epilepsy in PBTs varies between 30% and 100% and depends on tumor location and type, with frontal, central, and temporal areas having the highest propensity to generate seizures and slow growing tumors being the most epileptogenic (van Breemen et al., 2007; Rossetti & Stupp, 2010). The impact of epilepsy on the total disease burden is high and, in addition, both epilepsy and the use of antiepileptic drugs predispose to deterioration of cognitive functions (Klein et al., 2003), still a major problem in patients with a brain tumor (Douw et al., 2009).
Seizure semiology and evolution of epilepsy associated with PBTs (tumor epilepsy) have been described extensively in the past, but most studies were retrospective and performed at a time in which neuroimaging and pathology were inadequate (Lund, 1952). Moreover, in recent years, there has been a major advance in therapeutic procedures including new antiepileptic drugs, surgery, chemotherapy, and radiotherapy (van Breemen et al., 2007; Rudà et al., 2010; Prakash et al., 2012).
In a prospective registry of PBTs in Emilia Romagna Region (Italy) spanning a 3-year period (2009–2011), we studied the clinical features of tumor epilepsy, with the aim to give new information on the semiology and evolution of seizures in this special group of patients. In this article we present data concerning the epilepsy at the time of the tumor diagnosis and during a short-term follow-up.
This study is a section of the PERNO registry, the main aim of which is to collect prospectively all cases of PBTs occurring in the Emilia-Romagna region, northeast Italy (3,983,346 population) from January 2009 to December 2011, to allow epidemiologic, clinical, and biomolecular studies (de Biase et al., 2012). The PERNO registry is based on the participation of specialists in neurology, neuroradiology, neurosurgery, oncology, radiotherapy, biomolecular, and pathology who are operating in Emilia Romagna region, who were invited to use the project website (www.perno.it) to notify all new adult cases of suspected PBT observed from 1st of January 2009 and living in the Emilia Romagna region.
For each incident case notified, a research assistant collected demographic data and clinical information concerning date and symptoms at onset, neuroradiologic findings, date and type of surgical treatment, histologic classification, and medical therapies.
To complete data on incidence of PBT, all records of patients discharged with ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification) codes from regional and neighboring hospitals from January 1, 2009 to March 31, 2010 were searched. PBT diagnosis was based on histologic confirmation or typical radiologic findings with no evidence of cancer elsewhere. The study did not include cerebral metastases or tumors originating from extracerebral tissues, like meningiomas.
The epilepsy section of the PERNO study included all the patients who experienced seizures, either as first symptom of the tumor or appearing during the course of the disease. Each patient was interviewed by the referring neurologist with a specific interest in epilepsy who used an ad hoc questionnaire to collect information about seizure semiology, neurologic status, tumor location and histology, and treatments (including antiepileptic regimen, chemotherapy, radiotherapy, and surgery). The patients who entered the study were followed up with repeat visits on a quarterly basis. A consent form was obtained by each patient or responsible guardian.
Statistical analysis was performed by Student's t-test and Fisher's exact test.
Of 945 patients included in the registry within a 3-year period (2009–2011), 749 had complete forms with details on seizures and 510 had a histologically confirmed diagnosis of PBTs. Seizures were present in 230 cases (30% of all cases with complete forms). Of these patients, only 114 agreed to participate in the epilepsy study of the PERNO project.
Of the patients entering the study, 14 were excluded for a number of reasons, (including insufficient clinical information, inability to attend visits, informed consent withdrawal) and adequate data for analysis were available for 100 patients (62 male and 38 female, mean age 47 years, range 19–77). Ninety-nine patients had a histologically verified diagnosis.
A family history of seizures/epilepsy was reported in nine patients, and one additional 46-year-old patient with a frontotemporal low-grade glioma had a personal history of childhood absence epilepsy successfully treated with valproate until age 12.
Seventy-seven patients (mean age at onset, 56 years) had high grade (III–IV) gliomas (glioblastoma in 50 cases), whereas 21 cases (mean age at onset, 38 years) had low grade (I–II) gliomas. Two patients had malignant PBTs of different types (Table 1). The most common locations were the frontal (60%), temporal (38%), and parietal (28%) lobes, with multiple areas being involved in 37% of the cases.
|Tumor grade||Tumor type||N|
|Anaplastic xanthoastrocytoma anaplastic||1|
|Other types||Primary CNS lymphoma||1|
|Without histologic confirmation (MRI suggestive of high-grade PBT)||1|
Seizures were the first symptom of the tumor in 72 cases; seizures occurred after other symptoms (but before any treatment) in 7, and after surgery for the neoplasm in 21 (within the first 24 h in 6 cases, the first 2 weeks in 3 cases, and after the first 2 weeks in 12, of which 4 were during radiochemotherapy).
The semiology of the first seizures is shown in Table 2. Overall, the predominant seizure types were tonic–clonic (48%) (without clear initial focal signs in more than half of the patients), followed by focal motor (26%), complex partial (10%), somatosensitive (8%), aphasic (4%), and other types (4%).
|Seizure types||Seizures as first symptom||First seizures: all cases|
|With clear focal signs||20||27||22||22|
|Without focal signs||16||22||26||26|
Subsequent seizure types (occurring after the first seizure) were reported by 36 cases and were tonic–clonic/secondarily generalized (13), focal motor (8), somatosensitive (6), aphasic (3), complex partial (4), and sensory (2). The most frequently reported seizure sequence (16 cases, 43%) was a tonic–clonic/secondarily generalized seizure as first symptom, followed after a few days or weeks by focal (usually motor or somatosensitive) seizures. Less frequently (12 cases, 33%) the reverse sequence was reported (focal seizures as initial symptom and secondarily generalized seizures as subsequent seizures).
The majority of cases had isolated seizures (42%) or a low seizure frequency (<1 per month) (18%) at the onset of the disease. A high seizure frequency (>4 seizures per month) at the onset was observed in 18% of the cases, and status epilepticus in an additional 12%. Status epilepticus was tonic–clonic in six cases, focal motor in five, and aphasic in one. The PBTs associated with status epilepticus were glioblastomas in eight cases, low-grade gliomas in three, and gliomatosis in one, with a frontal or frontotemporal location in the majority of cases.
All the patients but three were on antiepileptic treatment, which was commenced at the time of the first seizure. The five most commonly used drugs were levetiracetam (63%), oxcarbazepine (33%), phenytoin (16%), carbamazepine (10%), and clonazepam (8%). Overall, 64 patients were on monotherapy, 23 on duotherapy, and 10 were taking three or more drugs. Additional medical therapy included steroids in 69 patients, particularly before and after surgical therapy, and radiotherapy.
Ninety-two patients underwent surgical removal of the tumor, which was either radical (38%) or partial (53%). Seven patients underwent only cerebral biopsy. In the 79 patients in which seizures were the first symptom, the mean time to the surgical treatment was 174 days, with a significant difference between high grade (95 days) and low grade (481 days) gliomas (p < 0.001).
In the majority (70%) of high grade tumors, surgery was followed by chemotherapy (usually with temozolomide) and radiotherapy. Surgery plus chemotherapy alone or surgery with radiotherapy alone was performed in seven and eight cases, respectively. Radiotherapy plus chemotherapy was performed in three not operated cases.
At the time of our first observation, the majority of patients (69%) had already undergone surgical removal, and the mean follow-up after the procedure lasted 5 months (range 1–6). Fourteen patients had 1-month follow-up, 14 2-month follow-up, 6 3-month follow-up, 5 4-month follow-up, 6 5-month follow-up, and 24 had 6-month follow-up. Overall, 39 patients were seizure free after tumor removal, 4 had a marked improvement of seizure frequency, 14 were unchanged, and 12 had a worsening of seizure frequency. Of interest, seizure control did not depend on presurgical seizure frequency (seizure freedom or marked improvement in 61% and 65% of patients with low and high presurgical seizure frequency, respectively; Table 3) or tumor type, although there was a trend in favor of better results with low-grade tumors (since seizure freedom or marked improvement was observed in 58% of high grade tumors compared to 76% of low-grade tumors—a difference not statistically significant, p = 0.25; Table 4).
|Pretreatment seizure frequency||N||Postsurgical seizure frequency|
|Low (isolated or <1/month)||46||28||61||7||15||11||24|
|PBT type||N||Postsurgical seizure frequency|
Because the 114 cases who agreed to enter the epilepsy section of the PERNO study were only a fraction of the total number of patients with PBT and seizures (230), we compared the two groups on the basis of general information obtained from the PERNO registry, that is, sex, age, percentage of seizures as first symptom, and tumor type. The two groups did not differ significantly in any of these parameters (Table 5).
|Group 1 (included) 114||Group 2 (excluded) 116||Significance|
|Age (mean, range)||47 (19–77)||62 (27–92)||N.S.|
|Sex (male > female)||71/44||71/43||N.S.|
|Seizure as first symptoms, n (%)||81 (71)||82 (70)||N.S.|
|Low grade, n (%)||22 (21)||15 (20)||N.S.|
|High grade, n (%)||79 (77)||61 (80)||N.S.|
|Other, n (%)||2 (2)||0 (0)||N.S.|
We have reported a series of 100 patients with tumor epilepsy collected from a prospective registry of PBTs in the Emilia Romagna Region spanning a 3-year period. There are two main limitations associated with this study. The first is that our series includes only a part of the total number of patients with seizures associated with PBTs, because only half of the registered patients agreed to enter the epilepsy section of the study. However, when patients entering the epilepsy study and those left out were compared, we did not find significant differences in terms of age, sex, and tumor types. The second limitation is that the series is not representative of the whole population of tumor epilepsy living in the study area, since there is a marked prevalence of high – grade gliomas (Ohgaki & Kleihues, 2005), this bias reflecting mainly the predominant type of patients seen at the neurooncologic centers of our region.
As reported in other studies (van Breemen et al., 2007), the most common locations of PBTs responsible for seizures were the frontal and the temporal lobe. Seizures were the first symptom of PBTs in about 72% of cases and prompted neuroradiologic investigations. There was a different time period, however, to diagnosis between the patients with high grade and low grade gliomas, as inferred from the latency between the first seizure and the time of the surgical procedure. This interval was significantly longer in the group with low-grade gliomas, a finding probably explaining a longer history of seizures in these patients. Of interest, this diagnostic delay is shorter here than in previous series, mostly because of an improvement in the brain imaging techniques (Michelucci & Tassinari, 1985).
Family history of epilepsy was observed in almost 10% of our sample, a rate significantly higher than that observed in the general population. Although the genetic background of tumor-associated epilepsy is unknown and no data are available, a genetic contribution to epileptogenesis in this condition has been suggested. The tumor suppressor gene LGI1, which causes a rare form of genetic epilepsy (autosomal dominant lateral temporal epilepsy; Michelucci et al., 2003), has been suggested to play a role in epileptogenesis in patients with brain tumors (Gu et al., 2005).
In our series, the most common initial seizure type was a tonic–clonic seizure (either secondarily generalized or unknown whether focal or generalized), this semiology mainly reflecting a predominant frontal localization of the tumors. Other common seizure types, somewhat typical of tumoral etiology, were focal motor, complex partial, and somatosensitive seizures. In many cases, an initial tonic–clonic seizure was followed by clear-cut focal seizures, mainly motor or somatosensitive.
The frequency of seizures at the onset was variable, the majority of cases having isolated or very few seizures. A high seizure frequency and/or status epilepticus was observed in about one third of cases. Despite the bias of a short follow-up at the time of our first visit, the resection of the tumor was followed by the disappearance of seizures in the majority of cases (62%), and this was the case also in the subgroup of 23 patients with seizures that were resistant to medical treatment. On the other hand, tumor type seemed to have a higher influence on the response to surgery, since low-grade tumors were more likely to have a better seizure outcome, although the difference did not reach statistical significance.
Available reports seem to indicate that lesionectomy of low-grade gliomas is associated with good postoperative results, seizure freedom being achieved in 65–82% of patients (Duffau et al., 2002; Luyken et al., 2003; Chang et al., 2008; Phi et al., 2009; Babini et al., 2013). The results are improved when lesionectomy is associated with tailored epilepsy surgery (Kurzwelly et al., 2010). In contrast, the prognosis of seizures in glioblastomas after surgery has not been studied extensively, but the outcome is poorer than in low-grade gliomas (Lund, 1952; Rosati et al., 2009).
The efficacy of antiepileptic drugs was difficult to assess in this uncontrolled setting. Indeed, almost all patients were given medical therapy at the time of the first seizure, and many of them underwent surgery after a short interval from the epilepsy onset, resulting in the absence of seizures in the months following the procedure. On the other hand, the number of patients with a long history of seizures was too low to draw conclusions on the effects of AEDs. It is well known, however, that seizures in brain tumors are often resistant to antiepileptic treatment. In a large cohort study, complete seizure control was achieved in only 20 (12.6%) of 158 patients with a brain tumor (Hildebrand et al., 2005). Similar rates of pharmacoresistance are reported in a number of large studies (Duffau et al., 2002), although more promising results have been reported with the use of new antiepileptic drugs, especially in terms of better tolerability and fewer interactions with antineoplastic agents (Michelucci, 2006; Vecht & van Breemen, 2006; de Groot et al., 2012).
In conclusion these data provide evidence that seizures are strictly linked to the tumoral lesion: They may be the initial symptom of the tumor, reflect tumor location and type, are usually resistant to antiepileptic treatment, and may disappear after the treatment of the lesion. In this sense, they represent the prototype of lesional epilepsy and need specific expertise for diagnosis and treatment.
This project was funded by the Research Program Regione-Università 2007–2009 Area 1a – Innovative Research Progetto Emiliano-Romagnolo di Neuro-Oncologia (PERNO; Emilia-Romagna region project).
None of the authors has any conflict of interest to disclose. We confirm that we have read the journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Albani F, Calbucci F, D'Alessandro R, Brandes A, Eusebi V, Pession A, Ceruti S, Fainardi E, Tamarozzi R, Emiliani E, Cavallo M, Franceschi E, Tosoni A, Cavallo M, Fiorica F, Valentini A, Depenni R, Mucciarini C, Crisi G, Sasso E, Biasini C, Cavanna L, Guidetti D, Marcello N, Pisanello A, Guiducci G, de Pasqua S, Testoni S, Agati R, Ambrosetto G, Bacci A, Baldin E, Baldrati A, Barbieri E, Bartolini S, Bellavista E, Bisulli F, Bonora E, Bunkhelia F, Carelli V, Crisci M, Dall'Occa P, de Biase D, Ferro S, Franceschi C, Frezza G, Grasso V, Leonardi M, Marucci G, Morandi L, Mostacci B, Calandri G, Pastore Trossello M, Poggi R, Riguzzi P, Rinaldi R, Rizzi S, Romeo G, Spagnolli F, Tinuper P, Trocino C, Visani M, Cerasoli S, Dall'Agata M, Faedi M, Frattarelli M, Gentili G, Giovannini A, Iorio P, Pasquini U, Galletti G, Guidi C, Neri W, Patuelli A, Strumia S, Casmiro M, Gamboni A, Rasi F, Cruciali G, Cenni P, Dazzi C, Guidi AR, Zumaglini F, Amadori A, Pasini G, Pasquinelli M, Pasquini E, Polselli A, Ravasio A, Viti B, Santini M, Ariatti A, Bertolini F, Bigliardi G, Carpeggiani P, Cavalleri F, Nichelli P, Pettorelli E, Pinna G, Zunarelli E, Artioli F, Bernardini I, Costa M, Greco G, Guerzoni R, Stucchi C, Iaccarino C, Ragazzi M, Zuccoli G, Api P, Cartei F, Granieri E, Latini F, Lelli G. Saletti A, Schivalocchi R, Saraceni S, Tola MR, Urbini B, Giorgi C, Montanari E, Cerasti D, Crafa P, Dascola I, Giombelli E, Mazza S, Ramponi V, Servadei F, Silini EM, Torelli P, Immovilli P, Morelli N, Vanzo C, and Nobile C.
Full affiliations and postal addresses of PERNO participants are available at the study Website: www.perno.it.