Right Unilateral Ultrabrief Pulse ECT in Geriatric Depression: Phase 1 of the PRIDE Study

The PRIDE study was a multicenter randomized study of the relative efficacy, functional outcome, and tolerability of a novel strategy to enhance long-term outcomes in geriatric depression. The study design involved two phases: an acute ECT phase (phase 1) and a 6-month randomized maintenance phase (phase 2). In phase 1, patients received acute ECT three times per week as well as treatment with venlafaxine. In phase 2, patients who remitted in phase 1 were randomly assigned to receive either pharmacotherapy alone (venlafaxine plus lithium) or combined treatment with pharmacotherapy and continuation ECT enhanced with individualization of the ECT schedule.

The initial eight sites participating in PRIDE were the Icahn School of Medicine at Mount Sinai, New York (clinical coordinating center); Columbia University/New York State Psychiatric Institute, New York (clinical coordinating center); Medical University of South Carolina, Charleston (data management and statistical coordinating center); Zucker Hillside Hospital/Northwell Health System; University of Texas Southwestern Medical Center, Dallas; Mayo Clinic, Rochester, Minn.; Wake Forest University Medical Center, Winston-Salem, N.C.; and New York Presbyterian/Weill Cornell Medical Center, New York and White Plains. Columbia University/New York State Psychiatric Institute discontinued recruitment and study procedures in October 2010 and was replaced by Duke University School of Medicine, Durham, N.C., as a performance site and clinical coordinating center; Mayo Clinic discontinued recruitment in May 2012; Wake Forest University Medical Center discontinued recruitment in August 2012 and was replaced by Augusta University/Medical College of Georgia, Augusta. The trial was approved by the institutional review boards of the participating study sites.

Phase 1, reported here, was completed August 2014; phase 2 was completed in March 2015 and is reported separately (11).

Participants were recruited from February 2009 through August 2014 from among patients who were referred for ECT at participating study sites. All patients provided written informed consent for research procedures and clinical ECT. Participants could be inpatients or outpatients and had to be at least 60 years old, have a DSM-IV diagnosis of unipolar major depressive episode, and have a score ≥21 on the 24-item Hamilton Depression Rating Scale (HAM-D) (12). Exclusion criteria were diagnoses of bipolar disorder, schizoaffective disorder, dementia, or an intellectual disability; a diagnosis of substance abuse or dependence in the past 6 months; any active general medical or neurological condition that would have affected cognition or treatment response; contraindications to lithium or venlafaxine; and history of a failure to respond, in the current episode, to an adequate trial of venlafaxine plus lithium or to ECT.

Baseline assessments were obtained for eligible patients who provided informed consent. During study years 1 and 2, diagnosis was established using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) (14). During study years 3–6, to minimize patient burden, the Mini International Neuropsychiatric Interview (15) was used instead.

Depressive symptoms were assessed at baseline and three times per week thereafter using the HAM-D and the Clinical Global Impressions severity scale (CGI-S) (16). Global cognitive function was measured using the Mini-Mental State Examination (MMSE) (17) at baseline and at each treatment session (three times per week, prior to administration of ECT). Suicidal ideation was assessed using the Beck Scale for Suicide Ideation (18). Baseline medical burden was assessed using the Cumulative Illness Rating Scale for Geriatrics (19).

Ongoing consistency of HAM-D ratings across sites was addressed by having each rater score a new set of videotapes annually for comparison with the consensus ratings of the study co–principal investigator and the project coordinator. A more detailed description of the standardization of ratings is provided in the online data supplement.

Time to reorientation following ECT, measured after ECT treatments 1–3, was determined by assessing the patient’s ability to correctly answer five orientation questions (open eyes on command, name, age, birthday, day of the week). Patients were asked these items at 3, 5, 10, 15, and 20 minutes and were considered reoriented at the time point at which all five items were correctly answered. This method is modeled after that used in other studies (7). Additional specific domains of neuropsychological functioning were assessed with a comprehensive neuropsychological test battery, the results of which will be reported elsewhere.

Safety measures assessed were serious adverse events and adverse events related to either ECT or venlafaxine, recorded according to Office for Human Research Protections guidelines. Safety measures were monitored externally by an NIMH data and safety monitoring board.

The primary outcomes in phase 1 were remission status and the longitudinal trajectory of HAM-D scores. Patients were classified as remitters if 1) they had a HAM-D score ≤10 on two consecutive ratings and 2) the HAM-D score did not increase >3 points on the second consecutive rating, or it remained ≤6. The minimum number of ECT treatments required for remission status was two, but there was no maximum. Patients were classified as nonremitters if they did not meet remission criteria, had at least 12 treatments, and reached a plateau, defined as no clinical improvement (<3-point decrease in HAM-D score after last two consecutive treatments). Patients were categorized as study noncompleters, or dropouts, if they did not meet remission or nonremission criteria and discontinued ECT before having received 12 treatments. There was no fixed time period required for completion of phase 1; patients could be declared remitters (and hence study completers) with as few as two ECT treatments, and nonremitters required at least 12 treatments to be considered study completers but could continue treatment if no plateau of improvement had been reached. Patients were considered study dropouts if they had received fewer than 12 ECT treatments before study exit and had not achieved remission.

Secondary efficacy outcomes in phase 1 were response status and speed of remission. Response was defined as a decrease of at least 50% from baseline in HAM-D score at the time of exit from phase 1. Speed of remission was determined by the number of ECT treatments required to achieve remission.

Frequency distributions for categorical variables and means with standard deviations for continuous variables were used to describe the phase 1 study sample. The proportions of remitters, nonremitters, and dropouts, with corresponding 95% confidence intervals, were determined at the point of exit from phase 1. There were no missing data for the primary three-level categorization (remitter, nonremitter, dropout), and no special methods for imputation were required. For some analyses, a dichotomous remission categorization was created by combining the nonremitter and dropout categories, an approach in which all study dropouts were considered to have had a negative (nonremitter) outcome. Similar analyses were carried out for response status, the secondary efficacy outcome. Speed of remission was estimated as the mean number of ECT treatments (with 95% confidence intervals) required to reach remission. In addition, we determined the percentage of remitters after specific numbers of ECT treatments—for example, percent who achieved remission after two treatments (the minimum required), within 1 week of treatment (three treatments), and within 2 weeks of treatment (six treatments).

The observed means for HAM-D score at each ECT session were used to describe the trajectory of symptom severity through the course of phase 1 for the total sample and within outcome categories (remitters, nonremitters, dropouts). The magnitude of the mean change from baseline was estimated for the total sample and within outcome categories using 95% confidence intervals for difference in means; the significance of the change from baseline was assessed using the paired t test.

Analyses for MMSE scores were as described above for HAM-D scores. For time to reorientation, the proportions of patients falling into each time category (reoriented within 3 minutes, 5 minutes, and so on) were determined.

Of 1,326 patients who were prescreened, 540 were ineligible; of the remaining 786 patients, 240 began phase 1 (for the study CONSORT chart, see Figure S1 in the online data supplement). The most frequent reasons for ineligibility at prescreening (see Table S3 in the data supplement) were diagnosis of bipolar depression (24%) or dementia (11%). A total of 172 patients completed phase 1; of these, 148 were remitters and 24 were nonremitters.

The participants’ mean age was 69.9 years (SD=7.6); 57.5% of the sample was female; 11.7% had the psychotic subtype of depression and 59% the melancholic subtype. The mean baseline HAM-D score was in the severe range (mean=31.2; SD=7.3), the mean MMSE score was within normal limits (mean=27.5; SD=2.4), and the mean CGI-S score indicated “markedly ill” (mean=5.3; SD=0.9). A family history of mood disorder was present in 61.9% of the sample (Table 1).

As noted, 148 patients (61.7%; 95% CI=55.2, 67.9) met remission criteria and 24 (10.0%; 95% CI=6.5, 14.5) were nonremitters; 68 patients (28.3%; 95% CI=22.7, 34.5) dropped out. A total of 169 patients (70.4%; 95% CI=64.6, 76.2) met response criteria (Figure 1). Among remitters, the mean decrease in HAM-D score was 24.7 points (95% CI=23.4, 25.9; p<0.001) and the final HAM-D score was 6.2 (SD=2.5); the average percent change from baseline was 79% (95% CI=77.4, 80.5). Among nonremitters, the mean decrease in HAM-D score was 11.6 points (95% CI=8.5, 14.6; p<0.001) and the final HAM-D score was 19.8 (SD=5.5). Among dropouts, the mean decrease in HAM-D score was 9.6 points (95% CI=7.4, 11.8; p<0.001) and the final HAM-D score was 22.4 (SD=9.4). Among all participants, the mean change from baseline in HAM-D score was 19.1 (95% CI=17.7, 20.5, p<0.001), and the mean percent change from baseline was 60.5% (95% CI=56.8, 64.1) (Figure 1). The trajectory of mean HAM-D scores (Figure 2) at each treatment session reflects the initial rapid decrease in depression symptom severity for the total sample and among all phase 1 outcome groups, followed by a continued general downward trend through ECT treatment 14 for the total sample and the remitters group.

The mean number of ECT treatments to achieve remission was 7.3 (SD=3.1); 19.6% of remitters (N=29) required four or fewer treatments; 45.3% (N=67) of remitters remitted within 2 weeks (six or fewer treatments), and 74.3% (N=110) remitted within 3 weeks (nine or fewer treatments). Only 25.7% of remitters (N=38) required 10 or more ECT treatments for remission (Figure 3).

Global cognitive function remained relatively stable during the ECT treatment course, with no significant difference between baseline and post-ECT mean MMSE scores (27.5 [SD=2.4] and 27.6 [SD=2.6], respectively). The other neuropsychological test results will be reported elsewhere.

Given the similarity in characteristics between nonremitters and dropouts for most baseline and clinical variables (Table 1), the dropout and nonremitter categories were combined to explore predictors of remission (Table 2). Age was significantly associated with remission when treated as a dichotomous variable (ages 60–69 and ≥70). The odds of remitting were 1.89 times greater for those in the older group compared with the younger group (95% CI=1.11, 3.22, p=0.020, logistic regression). The proportion remitting among those age 70 and over was 69.7% (76/109), compared with 55.0% (72/131) for those ages 60–69. The odds of remitting were twice as high for those whose Beck Scale for Suicide Ideation score was 0 compared with those who had scores >0 (odds ratio=2.00, 95% CI=1.11, 3.57; p=0.021, logistic regression); 71.4% (75/105) of the patients with a score of 0 remitted, compared with 55.7% (54/97) of those with scores >0. The relationship remained significant after adjustment for age (adjusted odds ratio=1.9, p=0.041). The early change in HAM-D score after ECT session 1 also predicted outcome, with each one-point increment in early change being associated with a 5% increase in odds of remission (odds ratio=1.05, 95% CI=1.001, 1.09; p=0.045); each 10-point increment in early change was associated with an increase of approximately 60% in the odds of remitting (e.g., the odds of remitting for a patient with an early change of 10 HAM-D points were 1.57 times greater than those for a patient with early change equal to 0; odds ratio=1.57, 95% CI=1.01, 2.43). All observed relationships, significant and nonsignificant, held in sensitivity analyses in which dropouts who had HAM-D scores ≤10 (N=3) were excluded.

In the total sample, there were 16 serious adverse events that affected 13 (5.4%) patients. Serious adverse events that were probably or possibly related to ECT were confusion or disorientation, tardive seizure, atrial fibrillation, and urinary retention, which occurred in one patient each. Serious adverse events that were probably or possibly related to venlafaxine were syncope and hyponatremia, which occurred in one patient each. One patient died from septic shock from a ruptured gall bladder, unrelated to study procedures.