Research Findings on Memory and Cognitive Impairment in ECT
Dr. Peter Como presented to the U.S. Food and Drug Administration Neurological Devices Panel examining the reclassification of electroconvulsive therapy (ECT) devices on January 27, 2011. This presentation comes from the public record of the meeting.
The FDA systematic review of the literature of cognitive adverse events [of ECT] included only randomized controlled trials as I mentioned. However, we did examine data from crossover designs if analyzable pre-crossover data were available. In addition, studies had to use standard psychometrically validated neuropsychologic tests.
The statistical comparisons that were examined included comparisons among various ECT treatment conditions, such as electrode placement, energy dose, frequency of treatment, waveform, and pulse. The comparisons also included ECT versus sham, ECT versus other treatments such as drug and medication placebo, and comparison of pre- and post-ECT changes in baseline cognitive test performance, although the pre- to post- ECT comparisons in themselves were non-randomized.
From this literature search, a total of 68 studies were identified which met these criteria.
This slide summarizes the findings of the published systematic reviews, meta-analyses, and practice guidelines. Overall, these sources indicate that there is evidence for impairment in orientation, anterograde and retrograde memory, and global cognitive function immediately after ECT that may last up to six months. Autobiographical memory is the most commonly reported memory impairment in these reviews. There is limited evidence to suggest that the effects of ECT on memory and cognitive function may not last more than six months.
A greater risk of memory or cognitive impairment is associated with sine wave compared to brief pulse ECT, bilateral and dominant hemisphere electrode placement, and the use of high energy dose ECT. This literature also suggests that raising the electrical stimulus above the individual seizure threshold increases the efficacy of ECT but at the expense of greater memory and cognitive impairment.
To continue, these summaries report that patient self- reported memory loss tends to be more persistent than the deficits that can be measured on formal neuropsychological testing. However, for those patients who do experience memory or cognitive impairment, they consider this to be a considerable source of distress for themselves and their families. The effects of ECT on cognitive function do not appear to differ among various psychiatric diagnoses such as schizophrenia and mania.
These summaries also suggest that factors other than the ECT treatment may impact cognitive function. These include individual variability, degree of improvement in depression, and the use of psychotropic medications at the same time as ECT.
I will not present the findings from the FDA systematic review of the cognitive adverse events literature. As noted earlier, FDA identified 68 studies which met the search criteria.
Cognitive Abilities Impacted by ECT
The specific cognitive domains for which data was available are listed on the next two slides. Bear in mind that the classification of cognitive domains is not mutually exclusive as there is considerable overlap among various cognitive functions.
Orientation includes person, place, and time, is most often measured by the number of seconds to minutes needed for a patient to become reoriented following ECT. Executive function includes aspects of attention, mental tracking and planning, problem solving, response inhibition, set-shifting, and working memory. Global cognitive function is typically a composite score on tasks of multiple cognitive domains. In the ECT literature, the most commonly used measure is the Mini Mental State Examination. Global memory typically is a composite score on a standardized memory battery. The most commonly used measure in the ECT literature is the Wechsler Memory Scale, although there are numerous other batteries that have been studied.
Anterograde memory, also commonly referred to as short- term memory, is the capacity to encode, store, and retrieve novel verbal and non-verbal information. Retrograde memory, also commonly referred to as long-term memory, is the capacity to retrieve information encoded prior to the initiation of ECT and is typically reported in the literature as personal or autobiographical memory, which is the ability to recall past personal information and events, such as birthdays, anniversaries, et cetera. Impersonal retrograde memory is the ability to recall historical or factual information such as the colors of the American flag or past presidents. Subjective memory is typically a patient’s self-report scale of perceived memory problems.
Other cognitive abilities, including language, visual, spatial, and motor function, among others, are typically part of a formal neuropsychological test battery. However, there are relatively few studies in the ECT literature examining these cognitive functions and therefore are not included in this presentation.
In reviewing the cognitive adverse events literature, there is a lack of consistent methodology regarding the time points of when cognitive assessment takes place. In reviewing the literature, the cognitive assessment time points generally fell into these categories. Acute effects are those occurring within the first 24 hours of ECT seizure termination. Subacute effects are those occurring from 24 hours to less than 2 weeks. Medium-term effects are those occurring from 2 weeks to less than 3 months. Longer-term effects are those occurring from 3 months to less than 6 months, and long-term effects are those occurring at 6 or more months.
There’s also some lack of consistency in the literature with respect to energy dose utilized. The FDA review of the cognitive adverse events literature generally categorized energy dose as follows. Low dose is considered to be 1 to 1.5 times the seizure threshold, moderate dose is 1.5 to 3 times the seizure threshold, and high dose is defined as more than 3 times the seizure threshold.
The cognitive and memory adverse events literature also looked at the effects of electrode placement. Electrode placement is generally categorized in the literature as bilateral, which for many studies consists of bitemporal placement, bifrontal placement, unilateral which consists of unilateral nondominant hemisphere and/or right unilateral hemisphere, and finally left unilateral or unilateral dominant hemisphere. All of these terms are in the literature.
Cognitive and Memory Impairment of ECT
The next set of slides summarizes the FDA systematic review of the cognitive and memory adverse events literature. For each of these specific cognitive domains, more specific detailed information is available to the Panel if needed.
For time to reorientation, the literature suggests that there’s a longer period of disorientation with bilateral electrode placement and with high dose ECT, although disorientation was generally quite brief. There does not appear to be any evidence of persistent disorientation over the long term. These data will be discussed in more detail by Dr. Krulewitch in her discussion of the meta-analyses conducted by FDA.
For executive function, there is no evidence of significant differences among the various ECT treatment parameters, although there was a single study which suggested greater executive dysfunction with left unilateral ECT compared to right unilateral ECT. Overall, the literature suggests that there is improvement or no statistically significant change from baseline at up to six months after ECT.
For global cognitive function, bilateral electrode placement was associated with greater impairment than right unilateral ECT. There’s no consensus in the literature on change in test performance on the Mini Mental State Examination from baseline up to two weeks following ECT. However, there was an apparent improvement or no change from baseline by three to less than six months. There are no reported effects of energy dose. Again, these meta-analyses of the MMSE conducted by FDA will follow my presentation.
For global memory function, there were no significant differences by energy dose or waveform or with ECT compared to sham, in the medium term, up to three months. There is limited evidence that bilateral ECT typically performed three times a week may be associated with greater global memory impairment. There’s no change from baseline test performance up to six months identified in the literature.
For anterograde memory, I’m going to break it down into verbal and non-verbal memory. For verbal anterograde memory, overall there are inconsistent results in the literature comparing ECT to sham.
However, there does appear to be a greater risk of verbal memory impairment with sine wave compared to brief pulse ECT, bilateral and dominant hemisphere electrode placement, and high energy dose ECT.
With respect to change from baseline, after about one week of treatment, verbal memory function may return to baseline and might improve following right unilateral electrode placement or low moderate energy dose ECT. After about two weeks of treatment, verbal memory functioning following bilateral electrode placement may return to baseline and may actually improve. From three to six months and beyond, there is limited data to determine if verbal memory impairment persists beyond this time period.
I will now turn to anterograde non-verbal memory data. For non-verbal memory function, the literature review yielded the following. ECT is associated with greater impairment compared to sham ECT immediately after treatment. There do not appear to be any differences in non-verbal memory function with respect to electrode placement. Brief pulse ECT may be associated with greater impairment compared to ultrabrief pulse ECT. After about two weeks of ECT treatment, there is no conclusive evidence in the literature to support any differences among the various ECT treatment parameters. However, there is relatively conclusive evidence of no significant changes in non-verbal memory test performance compared to baseline in the short term, which is the two week to three month period. There is limited data to suggest that in the longer term, non-verbal memory deficits may return to baseline levels.
For impersonal retrograde memory impairment, the literature review yielded the following. Immediately post-ECT, bilateral electrode placement may be associated with greater impairment. There are inconsistent findings with respect to electrode placement, pulse, or energy dose from about 24 hours to 3 months post-ECT. There are no differences between sham ECT and ECT, electrode placement, or pulse wave at six months. Detectable changes from baseline are inconsistent up to six months post-ECT. However, again the literature suggests no significant change from baseline appear to be present at six months.
There was a significant amount of information in the literature regarding the effects of ECT on autobiographical or retrograde personal memory. The majority of the studies in the literature tend to focus on the subacute affects, which is the 24-hour to 2-week time period after ECT. Immediately after ECT, there’s limited evidence to suggest that bilateral electrode placement is associated with greater impairment. In the subacute time period, 24 hours to 2 weeks, there is conclusive evidence to suggest that bilateral ECT is associated with greater impairment of autobiographical memory compared to unilateral, right unilateral or unilateral nondominant ECT. However, there’s limited evidence with respect to the effects of sine wave or high energy dose. There’s also evidence to suggest a decline from baseline test performance during this time period.
For the medium term, which is the two week to less than three month time period, there are limited data regarding the effects of electrode placement, pulse, or energy dose. The data are also limited with respect to change from baseline, although there are some studies that suggest no change or improvement with the use of ultrabrief pulse.
There was a single study comparing maintenance ECT with drug therapy. The results suggested that pharmacologic treatment demonstrated improvement relative to post-ECT performance whereas maintenance ECT demonstrated no change from the post-ECT baseline. At six months, there was only one study which suggested a return to baseline test performance with unilateral ECT. However, there was continued decline with bilateral placement in sine wave pulse.
The meta-analysis of the most commonly used instrument, the Autobiographical Memory Interview Scale, conducted by FDA, will be discussed following my presentation.
Assessment of subjective memory is problematic according to the literature due to the use of self-report scales, are dependent upon the timeframe which these scales are completed by patients, and may be related to the degree of improvement in depressive symptoms. In general, patients are more likely to report memory impairment immediately following ECT treatment. Bilateral was associated with greater impairment than unilateral ECT in the subacute time period, but by six months, there was no difference with respect to electrode placement, waveform, or sham versus ECT.
Improvement or no change from baseline appears evident at six months post-ECT.
To summarize, the systematic review of the cognitive adverse events literature indicates that ECT is associated with cognitive and memory impairment. The degree and duration of the impairment appears to be domain-specific and related to certain ECT treatment parameters. Specifically, there appears to be a greater risk of cognitive and memory impairment associated with bilateral and dominant hemisphere electrode placement and high energy dose ECT.
The key impaired cognitive domains from the review of the literature include disorientation, which is common but typically transient and resolves within seconds to minutes after seizure termination. Of the major cognitive domains, memory dysfunction is apparent for both anterograde and retrograde memory over the short term but may return to either baseline level or possibly improve. Specifically, the literature suggests that bilateral ECT is associated with greater autobiographical memory impairment compared to unilateral ECT, and there’s limited evidence to suggest these deficits may return to pre-ECT baseline test performance at six months.
The next slide, this next table, is a little busy but attempts to summarize the evidence in the literature regarding the pre- to post-ECT changes from baseline cognitive test performance. To try and orient you to the slide, the upward arrows indicate consensus in the literature of improvement relative to baseline test score. The downward arrows indicate consensus in the literature suggesting a decline from baseline test performance. The dashes indicate that the consensus in the literature suggest no change from baseline, and the one +/- sign that you see at six month under retrograde memory indicates that the results are different depending upon the ECT treatment parameter. Finally, the color coding is an attempt to indicate that the red arrow indicates that there is relatively conclusive evidence in the literature to support the finding. The blue arrow indicates that there is limited or equivocal evidence in the literature to support the finding.
Summary of ECT’s Effects on Brain and Memory Functioning
So to try and summarize this table for you, immediately post- ECT, there are deficits across nearly all cognitive domains. The cognitive deficits tend to persist up to two weeks with perhaps the exception of non- verbal memory function. From about two weeks to less than six months, cognitive test performance appears to either return to baseline or possibly improve. The data are limited at six months or greater. However, the available studies reviewed suggest there’s no evidence for persistent cognitive deficits except for perhaps autobiographical memory in which bilateral electrode placement in sine wave ECT appear to be associated with decline from baseline while unilateral brief pulse ECT appears to approach or return to baseline.
U.S. Food and Drug Administration. (2016). Research Findings on Memory and Cognitive Impairment in ECT. Psych Central. Retrieved on August 31, 2016, from http://psychcentral.com/lib/research-findings-on-memory-and-cognitive-impairment-in-ect/