Computed tomography of the head with and without contrast in imaging focal and unknown epilepsy – A prospective observational study

Abstract

Purpose

Brain imaging is needed when investigating epilepsy. Imaging options available include MRI and CT scan which may be non-contrast (NCCT) or contrast-enhanced (CECT). The specific clinical question and probable epilepsy substrate in the epidemiological context and socioeconomic milieu are important in determining the choice of imaging. In patients with well-controlled focal or unknown epilepsy who are unlikely to be surgical candidates, is CECT essential or can NCCT be an acceptable choice?

Methods

A prospective observational study was conducted at a tertiary care centre in India. Consecutive patients with focal or unknown epilepsy who were relatively well-controlled on medical treatment underwent NCCT followed by CECT brain. Three neuroradiologists independently reported the images. Proportion of abnormalities missed on NCCT and picked only on CECT were determined. How often abnormalities picked on CECT changed patient management was also analysed.

Results

Two hundred and nineteen patients with focal (87 %) or unknown (13 %) epilepsy underwent NCCT followed by CECT brain. Most had epilepsy for >3 months and an annual seizure frequency of 2–10 seizures. There was a nearly perfect inter-observer agreement between 3 neuroradiologists in reporting the NCCT and CECT as 'normal' or 'abnormal' with kappa (κ) values of 0.9 and 1.0 respectively. The sensitivity of NCCT compared to CECT in detecting an abnormality was 97 % (CI 92.6 - 99.5 %) and the specificity was 99 % (CI 94.9 - 99.9 %). There was no significant difference in the proportion of NCCTs and CECTs found abnormal (50.22 % vs 51.14 %, p = 0.91). A solitary calcified granuloma was the most common abnormality reported on NCCT as well as CECT, 21.0 % and 19.1 % respectively. New findings picked on CECT alone, did not change management in any patient.

Conclusion

When imaging focal or unknown epilepsy, an NCCT performs as well as a CECT, especially in regions where calcified lesions contribute a significant etiological burden. The role of imaging in epilepsy varies between patients and a universal recommendation of an MRI or a CECT in all patients is neither cost-efficient nor evidence-based. In drug responsive focal or unknown epilepsy of longstanding duration, CT scans are either normal or have calcified lesions that are easily picked on NCCT.