Reduction in temporal N-acetylaspartate and creatine (or choline) ratio in temporal lobe epilepsy: does this 1H-magnetic resonance spectroscopy finding mean poor seizure control?