Current understanding of risk factors in TLE

There is paucity of literature, especially from Canada, around the etiological risk factors of TLE, apart from surgical case series.^{Reference Shukla and Prasad4,Reference Uijl, Leijten, Arends, Parra, Van Huffelen and Moons5,Reference Junna, Buechler and Cohen-Gadol23,Reference Tassi, Meroni and Deleo24} Between 1993 and 2022, 11 studies reported findings on one or a few risk factors of TLE, with sample sizes ranging between 30 and 333 patients.^{Reference Soeder, Gleissner and Urbach1,Reference Klein, Dingledine and Aronica6–Reference Rai, Shukla and Afsar16} Four larger studies (N = 188–243) primarily investigated febrile seizures.^{Reference Heuser, Cvancarova, Gjerstad and Taubøll10–Reference Ferruzzi, Ferruzzi and Fernandes12,Reference Özkara, Uzan and Benbir15} Other studies investigated additional risk factors, including head trauma, early seizures following acute neurological insult, infections of the central nervous system (CNS) and family history.^{Reference Soeder, Gleissner and Urbach1,Reference Klein, Dingledine and Aronica6–Reference Sztriha, Gururaj, Bener and Nork9,Reference Tubi, Lutkenhoff and Blanco13,Reference Berg14}

Most of the studies were retrospective and focused on post-surgical population. One prospective study evaluated epilepsy development in patients with significant head trauma,^{Reference Özkara, Uzan and Benbir15} along with two comprehensive review papers.^{Reference French, Williamson and Thadani7,Reference Tubi, Lutkenhoff and Blanco13} Significant head trauma emerged as the most common risk factor associated with TLE.^{Reference Klein, Dingledine and Aronica6–Reference Sztriha, Gururaj, Bener and Nork9,Reference Tubi, Lutkenhoff and Blanco13,Reference Berg14} Other factors identified in those studies included early onset seizures following an acute injury, febrile seizures and neuroinfections such as meningitis or encephalitis.^{Reference Soeder, Gleissner and Urbach1,Reference Shukla and Prasad4,Reference French, Williamson and Thadani7–Reference Berg, Langfitt and Shinnar11}

The current study captures data collected through a structured template, applied to consecutive patients attending a referral epilepsy clinic, irrespective of patient’s candidacy for surgery, and hence is more representative of the general Canadian TLE population.

Head trauma

History of minor or major head trauma was reported in 2%–38% of patients with TLE in various studies,^{Reference Klein, Dingledine and Aronica6,Reference Sztriha, Gururaj, Bener and Nork9–Reference Berg, Langfitt and Shinnar11,Reference Tubi, Lutkenhoff and Blanco13–Reference Berg14} while in the current study, history of head trauma was recorded in 50% of patients with TLE. This was significantly higher proportion than in GGE (15%). This difference between our findings and those previously reported may be related to the study design with documentation limitations given varied research objectives of those studies, and may also reflect some socio-cultural variations, especially engagement in outdoor activity, as elaborated upon below.

Although participating in outdoor sports did not show a statistically significant difference between TLE and GGE groups, the frequency was notably higher in the TLE group (25%). On average, there is a higher percentage of Canadian population regularly engaging in outdoor activities (78%), with 15% reporting playing fast-paced sports such as football, soccer, hockey and others,^{Reference Berg14} compared to 44% of the European population²⁵ and 54% of the US population,²⁶ reporting some outdoor physical activity at least once a week.

Being active in fast-paced physical sports is frequently associated with repetitive TBI, which is usually mild and often underreported.²⁷ Interestingly, we also noted a high percentage of patients with EEG abnormality originating from middle/posterior temporal regions (N = 28, accounting for 48% of all temporal EEG abnormalities), suggestive of lateral temporal lobe involvement.

This temporal neocortical involvement is commonly seen in both major head trauma and repetitive concussions among individuals engaged in high-impact sports.^{Reference Mez, Daneshvar and Kiernan28} Our EEG findings align with those of Gupta et al who reported lateral (neocortical) localization in 56% of 115 patients with focal post-traumatic epilepsy.^{Reference Diaz-Arrastia, Agostini and Frol29} Moreover, only 12% of TLE patients in the our study had mesial temporal sclerosis on brain MRI studies, with 40% showing no apparent abnormality, 12% had non-specific findings and only 3% demonstrated clear encephalomalacia or gliosis (indicative of prior major trauma or stroke); these results possibly indicate some association with prior minor (but likely repetitive) head trauma.

Neuroinfections

In our study, 5.6% of TLE patients had a documented history of neuroinfections, compared to none in the GGE group. This is higher than the cumulative range of 3% reported in three studies on TLE patients.^{Reference Rathore, George, Kesavadas, Sankara Sarma and Radhakrishnan3,Reference Heuser, Cvancarova, Gjerstad and Taubøll10,Reference Berg14} The differences between our findings and other studies could be attributed to the difference in study population and study design. For instance, Harvey et al reported only new onset TLE in a pediatric population,^{Reference Heuser, Cvancarova, Gjerstad and Taubøll10} while Ozkara et al specifically reported patients with mesial TLE (mTLE) and hippocampal sclerosis,^{Reference Rathore, George, Kesavadas, Sankara Sarma and Radhakrishnan3} and Feruzzi et al reported a broader epilepsy population from Brazil, 19 of which had TLE.^{Reference Berg14} In a comprehensive review by Klein et al, data from multiple studies indicated that CNS infections, among other acute brain injuries, share many commonalities, particularly inflammatory processes that can eventually lead to epilepsy.^{Reference Rai, Shukla and Afsar16}

Early seizure(s) following neurological insult

While the definition of “early seizures” varies across the literature,^{Reference Harvey, Berkovic, Wrennall and Hopkins8,Reference Berg14} with timeframes ranging between one week and six months following an acute neurological insult, in the current study, we defined early-onset seizures as those occurring within 3 months of an acute brain injury.

In our study, approximately 16% of the TLE group had a history of early seizures following a neurological insult. In comparison, a recent study reported a higher rate of 32%.^{Reference Özkara, Uzan and Benbir15} It is important to note that in 53% of those patients in that study, early seizures occurred within one week of the insult, while the remainder developed seizures within six months. Several factors may explain the differences between our findings and those of Tubi et al^{Reference Tubi, Lutkenhoff and Blanco13}:

1. 1) Although the majority of patients who developed post-traumatic epilepsy in Tubi et al. study had TLE, as evidenced by temporal lobe abnormalities on neuroimaging, recruited patients had various epilepsy syndromes, not exclusively TLE.

2. 2) The Tubi et al. study included only patients with moderate to severe TBI, a population known to have a higher risk for early onset seizures.

3. 3) In Tubi et al.’s study, all patients were subjected to a continuous EEG in the acute phase during the ICU admission, which increases the likelihood of seizure detection. In our study, we primarily relied on patient or family reported seizure history, which could be affected by memory inaccuracies or the misidentification of atypical seizures.

Our findings are similar to those found in a large systematic review, in which Klein et al reported data on a total of 10,211 patients across five studies conducted between 1935 and 2010 in Europe and the USA.^{Reference Harvey, Berkovic, Wrennall and Hopkins8} These studies investigated epilepsy development following acute brain injury, specifically TBI, stroke, intracerebral hemorrhage and CNS infection. Early-onset seizures were found to increase the risk of later developing epilepsy, with incidence rates ranging from 13% to 27%. These authors propose that the acute activation of the innate immune system in the brain following an acute injury and the subsequent changes in the blood-brain barrier and the local structures trigger an inflammatory cascade affecting areas of the brain that are prone to seizures. It is possible that the inflammation itself promotes brain damage, leading to hyperexcitability and lowered seizures threshold, and resulting in early seizures that can further progress to epilepsy.^{Reference Harvey, Berkovic, Wrennall and Hopkins8}

Febrile seizures and febrile status epilepticus

Febrile seizures are among the most frequently reported antecedents and risk factors for TLE, particularly mTLE, with prevalence estimates ranging from 3.3% to as high as 70–78%.^{Reference Rathore, George, Kesavadas, Sankara Sarma and Radhakrishnan3,Reference Tubi, Lutkenhoff and Blanco13}

Interestingly, Heuser, in a 2011 study, considered TLE with childhood febrile seizure (TLE-FS) to be a distinct category of TLE, potentially pointing toward a common genetic ground.^{Reference Ferruzzi, Ferruzzi and Fernandes12} In the current study, we found that only 5.6% of TLE patients had a history of febrile seizures, with 2% having a history of confirmed febrile status epilepticus. Our study included a wider TLE population and not just those with mTLE with hippocampal sclerosis (mTLE-HS). A large population-based study from Denmark found an increased risk of epilepsy (not necessarily TLE) in children with recurrent febrile seizures, with the risk rising as the number of hospital admissions for febrile seizures increased. It is plausible that many of these patients had prolonged febrile seizures/febrile status epilepticus (not elaborated upon in the study), rather than simple febrile seizures alone.^{Reference Gupta, Sayed and Ding30} Independent risk of epilepsy in general, associated with febrile seizures (simple, complex or status epilepticus), remains unclear.^{Reference Dreier, Li, Sun and Christensen31} However, the FEBSTAT study did not identify a clear association between febrile status epilepticus and hippocampal injury on imaging.^{Reference Lewis, Shinnar and Hesdorffer32}

Strengths and limitations

Our study provides systematically collected data with the objective of identifying common antecedents and potential risk factors for TLE in a Canadian population. Having an appropriate control population of GGE is an important strength of this study as most previous studies have reported frequency of different risk factors in TLE populations, without including a comparison group. Another strength of this study is the use of structured data collection template by a single epileptologist, ensuring consistency and reliability across cases.

Limitations include the relatively small sample size, along with some missing data, which could have been more complete through prospective data collection, although the retrospective design minimizes risk of bias. Additionally, we were unable to further stratify TLE, as categorizing subtypes was challenging given the high percentage of patients with normal MRI (40%) and the small proportion with specific TLE syndromes such as mTLE-HS (13.5%).

Nonetheless, this research lays the groundwork for broader prevention and early identification strategies and serves as a steppingstone toward larger, multicenter studies incorporating prospective assessments with advanced structured interviews, neurophysiologic and neuroimaging evaluation, which could yield deeper insights into the optimal management of TLE.