The connectivity-based architecture of the human piriform cortex

Piriform cortex is an ictogenic trigger zone in the primate brain

OBJECTIVE

Area tempestas, a functionally defined region in the anterior piriform cortex, was identified as a crucial ictogenic trigger zone in the rat brain in the 1980s. However, whether the primate piriform cortex can trigger seizures remains unknown. Here, in a nonhuman primate model, we aimed to localize a similar trigger zone in the piriform cortex and, subsequently, evaluated the ability of focal inhibition of the substantia nigra pars reticulata (SNpr) to suppress the evoked seizures.

METHODS

Focal microinjection of the γ-aminobutyric acid type A (GABAA) antagonist bicuculline methiodide into the piriform cortex was performed, in macaque monkeys, on a within-subject basis to map the ictogenic regions within this area. Glutamate antagonists were used to characterize the local circuit pharmacology. Focal inhibition of the substantia nigra by infusion of the GABAA agonist muscimol suppressed seizures evoked from piriform cortex.

RESULTS

We documented a well-defined region highly susceptible to bicuculline-induced seizures in the piriform cortex, just posterior to the junction of the frontal and temporal lobes, indicating that a functional homolog to the rodent area tempestas is present in the primate brain. Focal infusion of glutamate receptor antagonists into the area tempestas revealed that α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor-mediated, but not N-methyl-D-aspartate-mediated, neurotransmission was necessary for the expression of seizures. Pharmacological inhibition of the SNpr robustly suppressed area tempestas-evoked seizures.

SIGNIFICANCE

Together, these data point to the area tempestas as a potent ictogenic zone in the primate brain and underscore the antiseizure effects of inhibition of the SNpr. Building on decades of studies in rodents, our present findings emphasize the relevance of these targets to the primate brain and provide further rationale for exploring these targets for clinical use.

Relationship of left piriform cortex network centrality with temporal lobe epilepsy duration and drug resistance

BACKGROUND

We investigated the relationship of piriform cortex (PC) structural network centrality with drug resistance and epilepsy duration as markers of sustained epileptic activity.

METHODS

PCs were manually delineated on retrospectively collected 3D-T1-MRI images of patients with temporal lobe epilepsy (TLE). Connectomes were computed from diffusion MRI scans, including the PC as network nodes. Betweenness centrality (BC) and node degree were computed and compared across drug-resistant versus drug-sensitive patients. Correlations of centrality metrics with the duration of epilepsy were calculated.

RESULTS

Sixty-two patients (36 females, 43/62 drug-resistant) were included in the main analysis. Greater centrality of the left PC was associated with drug resistance (degree: p = 0.00696, d = 0.85; BC: p = 0.00859, d = 0.59; alpha = 0.0125). Furthermore, left PC centrality was correlated with epilepsy duration (degree: rho = 0.39, p = 0.00181; BC: rho = 0.35, p = 0.0047; alpha = 0.0125). Results were robust to analysis of different parcellation schemes. Exploratory whole-network analysis yielded the largest effects in the left PC. Finer parcellations showed stronger effects for both analyses in the left olfactory cortex rostral to PC. In 28 subjects who had received epilepsy surgery, a trend of smaller centrality in patients with ILAE I outcome was observed in this area.

CONCLUSIONS

We demonstrated an increased centrality of the left PC in patients with drug-resistant TLE, which was also associated with the epilepsy duration. Recurring seizures over long periods may lead to changes of network properties of the PC. Large effects immediately rostral to our delineated PC region suggest a role of olfactory cortex anterior to the limen insulae in epileptogenic networks.