This pilot, open label investigation evaluates the safety and efficacy of a new form of electroconvulsive therapy (ECT). Both the efficacy and adverse cognitive effects of ECT are highly contingent on the intracerebral current paths and current density of the ECT stimulus. However, the impedance of the skull and individual differences in skull anatomy severely limit the spatial targeting of stimulation, and create marked individual differences in intracerebral current density. To address these problems, the investigators are exploring various means of overcoming this limitation.

An approach is to modify the electrical stimulus to induce focal seizures. The most common methods of ECT administration in the US use a bidirectional, constant current, brief pulse, with large (approximately 3 sq. in. surface area) and identically sized and shaped electrodes. In contrast, in this protocol the investigators have coupled unidirectional current flow with an electrode geometry involving a small and large electrode that differ by more than 3:1 in surface area.

Unidirectional currents were widely used in ECT during the, 1940's and continue to be used in European and American devices today. Transcranial electrical stimulation can be made focal by stimulating with an anode-cathode arrangement, with the electrodes differing in surface area. The investigators have shown in nonhuman primates the capacity to produce focal frontal seizure induction under conditions when a unidirectional current flows from a small anterior anode (placed on the forehead over the nasion) to a large posterior cathode just anterior to the motor strip. Furthermore, the investigators expect that some, if not all, of these seizures do not result in motor convulsions.

Thirty outpatients referred for ECT will participate. Relative to concurrent reference data from our ongoing ECT protocols, the investigators hypothesize that acute and subacute adverse cognitive effects of FEAST will be substantially less than those in patients receiving state-of-the art ECT, but with a traditional bidirectional, nonfocal stimulus. The investigators also hypothesize that the majority of patients will remit with FEAST. Thus, by improving the efficiency of the ECT stimulus with the switch to unidirectional current and the use of a new electrode geometry, the investigators expect to be able to induce focal seizures. The investigators hypothesize that this pilot study will provide evidence that this treatment is superior to traditional ECT in having lower dosing requirements and a superior side effect profile.

This study will provide preliminary evaluation of the following:

1. Determination of whether focal seizures can be induced with the FEAST methodology (unidirectional stimulation, small anterior and large posterior electrode).

   1. Focality will be assessed by the occurrence of non-motor seizures.
   2. Electroencephalographic evidence of pronounced asymmetry in frontal leads.

2. Determination of whether the FEAST methodology results in reduced seizure threshold.

   a. Seizure threshold will be quantified at the start of the treatment course using the standard method of limits titration procedure and compared to threshold determinations in matched patients who were treated with conventional ECT methods.

3. Characterization of dynamic impedance using the FEAST methodology.

   a. Dynamic impedance during the passage of the electrical stimulus will be quantified during each administration and compared to the values obtained in matched patients who were treated with conventional ECT methods.

4. Characterization of the efficacy of the FEAST methods and the safety of the treatment.

   1. The primary efficacy measure will be the 24-item Hamilton Rating Scale for Depression. The changes in these scores from before to immediately following the treatment course will be compared in patients treated with the FEAST methodology and matched patients who were treated with conventional ECT methods.
   2. Acute, subacute, and long-term cognitive side effects following FEAST will be assessed with comprehensive neuropsychological batteries. The primary acute measures will include the time to return of orientation following seizure induction and retrograde amnesia for words and shapes. The primary subacute measures will include assessments of anterograde amnesia (forgetting over a delay) for a verbal list and for reproduction of a complex figure, as well as retrograde amnesia for autobiographical information. The primary long-term measure will be retrograde amnesia for autobiographical information, assessed 6-months following the FEAST course. The neuropsychological measures will be compared in the patients treated with the FEAST methodology and matched patients who were treated with conventional ECT methods, as well as in healthy participants who receive were administered the neuropsychological battery at the same intervals as the FEAST patients, but without Intervention.
   3. Safety will also be determined by examining the number and frequency of serious adverse advents and adverse events, as well as scores on the Columbia University ECT Side Effect Scale.

Inclusion Criteria:

• Age between 18 and 90 years (inclusive)
• Diagnosis of major depressive disorder
• Pretreatment HRSD score ≥ 18
• ECT indicated
• Willing and capable of providing informed consent

Exclusion Criteria:

• History of schizophrenia,
• schizoaffective disorder,
• other functional psychosis, or
• rapid cycling bipolar disorder
• History of neurological illness or insult other than conditions associated with psychotropic exposure (e.g., tardive dyskinesia)
• Alcohol or substance abuse or dependence in the past year (RDC)
• Secondary diagnosis of a delirium, dementia, or amnestic disorder (DSM-IV), pregnancy, or epilepsy
• Requires especially rapid antidepressant response due to suicidality, psychosis, inanition, psychosocial obligations, etc.
• Unable to tolerate psychotropic washout and no psychotropic medication during the ECT trial, other than lorazepam (up to 3 mg/d PRN)
• ECT in the past six months