Impact of injury location and severity on posttraumatic epilepsy in the rat: role of frontal neocortex

Abstract

Human posttraumatic epilepsy (PTE) is highly heterogeneous, ranging from mild remitting to progressive disabling forms. PTE results in simple partial, complex partial, and secondarily generalized seizures with a wide spectrum of durations and semiologies. PTE variability is thought to depend on the heterogeneity of head injury and patient's age, gender, and genetic background. To better understand the role of these factors, we investigated the seizures resulting from calibrated fluid percussion injury (FPI) to adolescent male Sprague-Dawley rats with video electrocorticography. We show that PTE incidence and the frequency and severity of chronic seizures depend on the location and severity of FPI. The frontal neocortex was more prone to epileptogenesis than the parietal and occipital, generating earlier, longer, and more frequent partial seizures. A prominent limbic focus developed in most animals, regardless of parameters of injury. Remarkably, even with carefully controlled injury parameters, including type, severity, and location, the duration of posttraumatic apnea and the age and gender of outbred rats, there was great subject-to-subject variability in frequency, duration, and rate of progression of seizures, indicating that other factors, likely the subjects' genetic background and physiological states, have critical roles in determining the characteristics of PTE.

Figure 1.

Basic properties of frontal, parietal, and occipital FPI, its mechanical reproducibility, acute mortality, and ensuing epilepsy. (A) Location of craniotomy and recording (#1–4) and reference (#5) electrode positions for rpFPI, mpFPI, and cpFPI. (B) Graph shows 23 overlapped digitized pressure traces resulting from FPI at 3.4 atm. Amplitude of pressure transducer output is measured in volts. Traces demonstrate the high reproducibility of the mechanical injury. (C) The plot shows mean righting times after 0 atm (sham), 2.0 atm, and 3.4 atm FPI (all locations combined). Righting time increases linearly with severity of injury. (D) Histograms show the acute (≤1 week) mortality for rats injured at different locations (left panel) and for rats receiving moderate and severe trauma (right panel). n.s. indicates no statistical significance. Note that FPI at all locations results in comparable acute mortality when animals are mechanically ventilated after injury. (E) The incidence of PTE increases over time after injury in all experimental groups and depends on the severity of injury (2.0 vs. 3.4 atm) but not on the location of FPI. On the left, the cumulative probability of detecting seizures (G1, G2, and G3) in FPI rats and in sham-injured rats is plotted versus time after injury. On the right, the cumulative probability of detecting unprovoked spreading seizures (G2 and G3) in FPI and sham-injured rats is plotted versus time after injury. Asterisks indicate significant differences in the cumulative incidence of seizures between rats injured at 2.0 atm and rats injured at 3.4 atm. (F) Histograms show percentage of remitters after injury at different locations (left panel) and after injury at 2.0 and 3.4 atm (right panel). n.s. indicates no significance. (G) Remitters (Rem) had low preremission seizure frequency, had few G1 seizures, and virtually no G2 and G3 seizures. Nonremitters (Non Rem) presented with varying and often high frequency of seizures of all types over the period of study.

Figure 2.

Duration of clinical seizures induced by rpFPI, mpFPI, and cpFPI. Overall frequency of all seizure types (A, C, E) or spreading G2 and G3 seizures (B, D, F) early (weeks 2–3; hollow symbols) and late (17–19 weeks; filled symbols) after 3.4 atm rpFPI (A, B), mpFPI (C, D), and cpFPI (E, F). Insets show the same data on a logarithmic scale. Note that seizures are frequent 2–3 weeks after rpFPI (A) but much rarer after mpFPI (C) and cpFPI (E), indicating more rapid epileptogenesis in the frontal neocortex. However, a wide range of seizure durations is a reliable feature of the PTE syndrome regardless of injury location.

Figure 3.

Temporal progression of seizure frequencies and their dependence on location and severity of FPI. (A–G) Frequency of distinct seizure types determined 2–3, 8–9, and 17–19 weeks after rpFPI, mpFPI, and cpFPI at 2.0 and 3.4 atm. RpFPI at 3.4 atm induces a higher frequency of neocortical seizures at weeks 2–3 and 8–9 post-FPI, compared with all other groups. Statistically significant differences (2-tailed unpaired t-test after logarithmic transformation) are indicated by # (rpFPI 3.4 atm vs. mpFPI 3.4 atm), $ (rpFPI 3.4 atm vs. cpFPI 3.4 atm), * (rpFPI 2.0 atm vs. 3.4 atm), o (mpFPI 2.0 vs. 3.4 atm),   (cpFPI 2.0 atm vs. 3.4 atm). (H–I) For each experimental group, mean seizure frequency at 2–3 weeks postinjury is plotted against the intensity of mechanical (H) and functional (righting time; I) injury. RpFPI rats injured at 3.4 atm present higher seizure frequency than mpFPI and cpFPI rats despite comparable mechanical injury and righting time.

Figure 4.

Rat-to-rat variability in progression of posttraumatic epilepsy after FPI. The frequency of neocortical seizures (A) and of limbic seizures (B) is plotted for rpFPI, mpFPI, and cpFPI rats injured at 2.0 atm and 3.4 atm. Note the variable progression of epilepsy in different subjects and the overall lower propensity to epileptogenesis after mpFPI. (C) Rat-to-rat variability in temporal progression of seizure types. Proportions of G1, G2, and G3 seizures are represented with thin, intermediate, and thick line, respectively, at all time points. No line indicates no seizures detected. Note the overall tendency to worsening of seizure type with the time postinjury but with significant rat-to-rat variability. (D) Temporal progression in the proportion of G1, G2, and G3 seizures for rpFPI, mpFPI, and cpFPI animals (2.0 and 3.4 atm combined). Note the faster increase in the frequency of G2 seizures after rpFPI. (E) Relationship between log-transformed frequencies of neocortical seizures at weeks 2–3 postinjury and limbic seizures at weeks 17–19, after rpFPI, mpFPI, and cpFPI at 2.0 atm (hollow symbols) and 3.4 atm (filled symbols). A significant correlation (Pearson; P = 0.03) between the frequencies of early neocortical and later limbic seizures was found only for the rpFPI group. Line represents linear fitting of the data.

Figure 5.

Seizure frequency, duration, and behavioral changes associated with neocortical seizures detected at different neocortical sites. (A) Pie charts showing time-dependent changes in the proportion of G1 events detected by each of the 4 active epidural electrodes (1–4 in the inset) after rpFPI, mpFPI, and cpFPI (2.0 and 3.4 atm combined) at 2–3, 8–9, and 17–19 weeks posttrauma. At 2–3 weeks post-FPI, G1 seizures mostly originate from frontal cortex even when trauma is applied to caudal cortex. For rpFPI rats, the frontal lobe dominance of focal seizures does not change over time after injury. Following mpFPI and cpFPI, substantial fractions of G1 seizures are detected by contralateral or nonperilesional electrodes, indicating multifocal neocortical epilepsy. Electrodes' positions in respect to the skull are shown at right. (B) Inverse correlation between the duration of focal seizures and the distance between the site of onset and the injury site. Frontal cortex G1 seizures detected by electrode 4 (filled symbols) are longer when induced by rpFPI 3.4 atm than when induced by mpFPI or cpFPI at same severity of injury, suggesting that mechanisms of seizure maintenance and termination are recruited to a greater extent when injury is delivered closer to the frontal epileptic focus. Perilesional G1 seizures induced by cpFPI at 3.4 atm (hollow symbol; detected by electrode 3) are shorter than those detected by electrode 4 after rpFPI at same severity, indicating that the frontal cortex sustains seizures better than the occipital cortex. Asterisks indicate P < 0.05 (Mann–Whitney U test); n.s. indicates no significant difference. (C) Frequency of frontal cortex G1 seizures detected by electrode 4 (filled symbols) at weeks 2–3 and 8–9 is higher after 3.4 atm rpFPI than after mpFPI and cpFPI at the same severity. The frequency of these seizures decreases rapidly with the distance from the injury site. Perilesional G1 seizures induced by mpFPI and cpFPI at 3.4 atm (hollow symbol; detected by electrode 3) are much rarer than those induced by rpFPI, indicating that the frontal cortex is more prone to epileptogenesis after injury. (D–E) G1 seizures detected in the frontal cortex by electrode 4 after rostral or caudal injury are most commonly coincident with behavioral changes (D), typically motor arrest with or without facial automatisms, while G1 seizures originating from the occipital cortex (detected by electrode 3) were never associated with changes in behavior (E), regardless of the location of injury. n indicates number of events.

Figure 6.

ECoG traces and video frames of chronic seizures induced by FPI. (A) A ∼42s-long G1 seizure with behavioral arrest following 3.4 atm rpFPI. The rat was engaged in grooming and exploratory behavior before the onset of the seizure, exhibited a freeze-like arrest concomitantly with the ECoG event and then resumed its exploratory behavior immediately upon termination of the epileptiform discharge. (B) A cpFPI rat injured at 2.0 atm having a 8.5 s-long cortical spreading seizure (G2) associated with behavioral arrest without loss of posture. (C) A cpFPI rat injured at 3.4 atm having a ∼32s-long G3 limbic seizure associated with loss of posture. The clinical manifestation of this event begins with the rat crouching down to the floor of the cage prior to detection of the neocortical discharge. The rat then remained motionless until electrographic discharge ended. In all panels, dotted rectangles highlight ECoG traces expanded in insets at right. Numbers left of all ECoG traces represent the electrode and its reference.

Figure 7.

Hippocampal and temporal neocortex atrophy in posttraumatic epileptic rats after frontal, parietal, and occipital FPI. GFAP-stained coronal sections at Bregma −3/−4 mm obtained from rpFPI rats (A, B) and cpFPI rats (C, D) injured at 3.4 atm. Examples illustrate negligible hippocampal and temporal neocortical asymmetry (A and C), pronounced shrinkage of the ipsilateral temporal neocortex (B, arrow), and moderate shrinkage of the hippocampus (D, arrow). Insets in (A–D) represent the position of the coronal section (dotted line) with respect to the FPI craniotomy site (gray circle) and the rat skull. Scale bar is 1 mm in all panels. Hippocampal (E) and temporal neocortex (F) asymmetry in rpFPI, mpFPI, and cpFPI rats injured at 2.0 atm and 3.4 atm. Varying degrees of asymmetry are evident in each experimental group. The temporal cortex asymmetry increases significantly with the severity of FPI (F, right panel). Pron, pronounced; Mod, moderate; Negl, negligible.