|
Effective Use of Electroconvulsive Therapy in Late-Life Depression
Determining Seizure Threshold
There is good evidence that the dose of electricity relative to a patient’s
seizure threshold affects not only the efficacy of unilateral ECT but also
cognitive side effects, regardless of electrode position (14,15,17). Stimulus
dosing, whereby electrical stimulations are administered at subconvulsive
levels until a generalized tonic-clonic seizure of at least 20 seconds’
duration is induced, is the most precise method for determining a patient’s
seizure threshold. (See Frukacz and Mitchell [22] for examples of stimulus
dosing protocols.) Stimulus dosing has been criticized as potentially increasing
the risk of bradyarrhythmias, on the grounds that repeated subconvulsive
electrical stimulations may result in a sustained parasympathetic response
(23). These concerns, however, have not been borne out in studies that
have routinely used stimulus dosing (24). To attenuate the parasympathetic
response, some authors have recommended the use of atropine or glycopyrrolate
(which does not cross the blood–brain barrier) prior to stimulus dosing,
but it remains to be determined whether this strategy results in a clinically
significant benefit.
Critics of stimulus dosing have proposed the "half-age method" for determining
the dose of electricity for bilateral ECT (23). According to this method,
the amount of energy (joules) is calculated at one-half the patient’s age.
This method, however, is limited by the fact that there is tremendous interindividual
variability in seizure threshold, with studies reporting anywhere from
fourfold to thirty-fivefold variability within samples (15,18,25,26). Further,
in a group of mixed-age patients, Boylan and others found that age accounted
for only 13% of variance in seizure threshold (18), meaning that many other
factors also contribute to the variability of this measure. Thus, the limitation
of the half-age method is that it may result in an unnecessarily high dose
of electricity in patients undergoing bilateral ECT and an underestimation
of electrical dose in patients receiving unilateral ECT (26,27).
Seizure Duration
Length of seizure—beyond the widely quoted minimum criterion of 20 seconds
of motor or 25 seconds of electroencephalographic manifestation—is not
related to ECT efficacy (28). Even this minimum criterion is of uncertain
scientific standing, since it is not experimentally derived but instead
represents a consensus among ECT researchers (22,28). In many patients,
especially older ones, seizure duration decreases and seizure threshold
increases during the course of ECT (28,29). However, there is no correlation
between change in seizure length and change in seizure threshold (30).
Further, substantially suprathreshold stimulus intensities are associated
with shorter, not longer, seizures (28). Thus, seizure duration alone should
not serve as a marker of treatment adequacy (28). In the case of an older
patient receiving low-dose bilateral ECT, it is reasonable to increase
the dose of electricity (initially, by an increment of 50%) if there is
a substantial reduction in seizure length between 2 consecutive treatments
(for example, a decline from 50 seconds to 20 seconds) or if seizure duration
falls below 15 seconds, assuming that the short seizure is not owing to
medication effects or inadequate ventilation (28). Conversely, in the case
of a patient receiving high-dose ECT, retitration of the electrical dose
is the only accurate way of determining whether shortening of seizure length
indicates the need for a higher stimulus intensity.
Managing Medications With Anticonvulsant Effects
Because of their anticonvulsant properties, barbiturates, benzodiazepines,
and antiepileptic drugs can potentially interfere with seizure elicitation
and expression. The dose of barbiturate anesthetic should be based on the
patient’s weight, also taking into account body mass, age, and previous
anesthetic experience. Boylan and others found that methohexital did not
affect seizure threshold or seizure duration when limited to a dose of
0.75 to 1.0 mg/kg (18). Frukacz and Mitchell reported that when thiopentone
was given at a dose of 2.5 to 3.0 mg/kg, a significant proportion of patients
had either no or very brief seizures (22). However, when a dose of 2.0
mg/kg was used, all patients had adequate seizures. Boylan and others did
not find that the lorazepam dosage in the 48 hours prior to the first ECT
session affected seizure threshold, but the dosage was limited to a relatively
small range (mean dose 0.85 mg daily, SD 1.06) (18). Further, the mean
age of patients in this study was under 60 years, and it is possible that
an older group of patients would be more sensitive to the benzodiazepines’
anticonvulsant effects. Many clinicians recommend avoiding the use of benzodiazepines
in elderly patients during ECT, if possible. If a benzodiazepine is required,
lorazepam at a dose of 0.5 to 1.0 mg daily is the most appropriate choice
(3). Antipsychotic medications, which can lower seizure threshold, can
be an effective alternative to benzodiazepines in the acute management
of agitation or severe anxiety associated with late-life depression (3).
In patients with seizure disorder, antiepileptic medication should initially
be maintained at a therapeutic dosage, because dosage reduction or discontinuation
increases the patient’s risk of experiencing seizures between ECT treatments
(31). The dosage should be cautiously reduced only if an adequate seizure
cannot be elicited (31). In the case of patients taking antiepileptic medications
as mood stabilizers, it is preferable to withdraw the medications prior
to ECT (28).
Frequency of ECT
Controlled studies have established that bitemporal ECT administered 3
times weekly results in more rapid improvement than treatment twice weekly,
but there is no difference between the 2 schedules in the total number
of treatments required to achieve response or in the percentage rate of
response (32). Conversely, the more frequent schedule is associated with
more retrograde amnesia, both immediately after finishing the course of
ECT and at 1-month follow-up (32). Thus, twice-weekly administration may
be the optimal schedule for bitemporal ECT in the elderly, unless clinical
indications or other considerations (for example, length of hospitalization)
require the more rapid antidepressant effect of thrice-weekly treatment.
Comparable data on the frequency of RUL ECT are not available.
Number of Treatments
There is considerable variability in the number of ECT treatments required
for response. As a result, the number of treatments in a course of ECT
should be decided on a case-by-case basis. ECT is typically discontinued
once symptoms remit or when symptoms reach a plateau of improvement after
2 consecutive treatments. Among elderly patients with major depression,
6 to 12 treatments are often required to achieve maximal benefit, but some
patients may need more than 12 treatments. In the case of no response or
minimal response, many experts recommend at least 10 to 12 bitemporal treatments
before the depressive episode is labelled nonresponsive (11,15).
Safety and Medical Issues
The mortality rate associated with ECT is only 0.2 to 0.4 per 10 000 treatments,
no higher than that expected with general anesthesia alone (33). Cardiovascular
complications constitute the principal cause of ECT-related morbidity (1).
Delivery of the ECT stimulus induces a brief parasympathetic response that
can result in sinus bradycardia and hypotension (31). Not infrequently,
transient asystole occurs (34). As the patient starts to seize, a discharge
in catecholamines from the adrenal medulla results in increased heart rate
and blood pressure (31). In patients with ischemic heart disease, this
period of increased myocardial oxygen demand may increase the risk of cardiac
ischemia. In turn, ischemia is the main cause of arrhythmias. The primary
preventive measure for these complications is adequate ventilation and
oxygenation (31). In addition, in selected high-risk patients, beta blockers,
calcium channel antagonists, or nitrates are often used to manage the effects
of the sympathetic response to ECT (35,36). However, there has not yet
been a controlled trial to determine whether this strategy has a clinically
significant impact on ECT-related morbidity. Not surprisingly, cardiac
complications of ECT are more likely to occur in elderly patients, particularly
those with preexisting cardiovascular conditions. Nevertheless, several
studies have shown that most complications are transitory and usually do
not prevent the completion of the course of ECT (37–39).
ECT use in medically ill patients has been extensively reviewed elsewhere,
and a detailed discussion is beyond the scope of this article. (For recent
reviews, see Rabheru [40] and Tew and others [31].) There are no absolute
contraindications to ECT, just relative ones (41). ECT has been safely
and effectively performed in the presence of a wide range of serious medical
conditions, including severe ischemic heart disease, aortic stenosis, chronic
airways disease, and osteoporosis; aortic and cerebral aneurysms; brain
tumours; epilepsy; and recent stroke (31,40). Patients with pacemakers
and patients taking anticoagulants can safely undergo ECT (31,40). As with
any treatment, the risks of ECT must be balanced against its potential
benefit and the risks and benefits of alternative treatments or no treatment.
It is worth emphasizing that untreated depression can have severe medical
consequences in older people, including dehydration, malnutrition, skin
breakdown or deep venous thrombosis secondary to prolonged immobility,
and an increased risk of mortality (42). Further, untreated depression
can adversely affect recovery from various medical and neurological conditions
(42).
Cognitive Effects
Impaired cognitive function is the most frequent adverse effect of ECT
and arguably the main factor that has limited use of this treatment (1).
As already noted, several technical factors, including dose of electricity
relative to seizure threshold, electrode position, and frequency of treatment,
can affect cognitive function and need to be kept in mind when balancing
efficacy (rate and speed of response) against potential side effects.
It is unclear whether aging per se increases the severity and persistence
of ECT-related cognitive effects. In a group of patients aged 20 to 65
years, moderately suprathreshold bitemporal ECT resulted in more severe
impairment of verbal and visuospatial anterograde memory in older individuals
tested 24 to 72 hours after the last ECT treatment (43). However, these
deficits were marginal at 1-month follow-up and had disappeared by 6-month
follow-up.Conversely, Tew and others (12) and Wilkinson and others (10)
did not find that older patients fared any worse than their younger counterparts
in terms of change in Mini-Mental State Examination (MMSE) scores between
baseline and the week after finishing ECT. Indeed, Wilkinson and others
found that, when the MMSE was administered between 72 hours and 1 week
after the last ECT treatment, MMSE scores improved above baseline in all
age groups, but the magnitude of improvement was greatest in older patients.
This paradoxical finding of improved cognition following ECT underscores
the complicated interaction between cognitive impairment due to depression
and cognitive impairment induced by ECT. Thus, while ECT can cause circumscribed
anterograde and retrograde amnesia during and following the course of treatment,
it can also result in improved global cognitive function as a result of
improvement in depression.
Reports on ECT in patients with dementia, cerebrovascular disease, or Parkinson’s
disease suggest that these patients have increased risk of interictal confusion,
compared with elderly patients without neurologic impairment (39,44–46).
However, the confusion is usually transient. In fact, in several of these
reports, patients’ performance on the MMSE was again, on average, better
at the end of ECT than before ECT (39,44,45). In these studies, ECT was
usually administered 3 times weekly—the incidence of acute confusion may
possibly have been less if patients had received ECT twice weekly.
Relapse Prevention
Relapse of depression following response to ECT is a significant problem.
Several studies have found that the relapse rate during the 6 to 12 months
following acute ECT exceeds 50%, despite continued antidepressant medication
(14,15,47–49). Relapse prevention remains a major challenge for the field.
Inadequate response to antidepressant pharmacotherapy is a primary indication
for ECT, yet antidepressant medication is usually prescribed as continuation
treatment following ECT. Thus, after responding to ECT, patients are typically
switched back to a treatment modality that had previously proven ineffective.
It is not surprising, therefore, that resistance to antidepressant medication
has been found to predict post-ECT relapse (15,50). One logical solution
to this problem is to continue ECT after response. Currently available,
albeit limited, data suggest that continuation ECT (C-ECT) is a safe, efficacious,
and cost-effective way to prevent relapse (51). In a prospective study
that assigned treatment based on clinical grounds, elderly patients assigned
to C-ECT had a 6-month relapse rate for major depression of 11% (n = 1/9).
This rate was accounted for by 1 patient who stopped ECT and then relapsed.
In comparison, the relapse rate in the continuation medication group was
67% (n = 4/6) (52). A retrospective case-controlled study of 58 patients
with a mean age of 65 years (SD 19) found that the cumulative probability
of relapse or recurrence within 2 years of an acute course of ECT was 7%
for C-ECT and 52% for continuation pharmacotherapy (53). Nevertheless,
despite the apparent benefits of C-ECT, there are barriers to its routine
use. These include issues of patient acceptance and adherence, the need
for sufficient resources to operate a sizable outpatient ECT program, intercurrent
events that interfere with treatment continuity (for example, medical illness,
transportation problems, or inability of caregivers to accompany the patient
to treatment), and, in some countries, financial cost to the patient. Further,
some departments of psychiatry do not provide ECT on an outpatient basis,
meaning that some patients may have to travel a long distance to an outpatient
ECT program. Thus, geography may also hinder patients receiving C-ECT.
Because of the barriers to routinely using C-ECT, investigators have examined
whether other treatment approaches can lessen the risk of relapse. In a
double-blind placebo-controlled study, Sackeim and others compared combined
nortriptyline and lithium with nortriptyline monotherapy in preventing
post-ECT relapse (50). Over the 24-week trial, relapse rates for placebo,
nortriptyline monotherapy, and nortriptyline plus lithium were 84%, 60%,
and 39%, respectively. Thus, even though the combination of nortriptyline
and lithium was more effective than antidepressant monotherapy, the relapse
rate associated with this treatment was still high.
Other strategies that have been proposed to reduce relapse rates include
tapering ECT over a few weeks, rather than stopping it abruptly at the
time of response, and starting the continuation antidepressant medication
during ECT, rather than waiting until ECT is finished (50). Studies undertaken
in the 1960s, in which patients took antidepressant medication in conjunction
with ECT, found that continuation of the antidepressant after ECT was associated
with a 6-month relapse rate of approximately 20%, compared with 50% to
70% in the control groups (54–56). However, at that time ECT was frequently
used as a first-line treatment, and it is possible that some patients in
these studies would have responded to the antidepressant medication and
remained well on it, even if they had not received ECT. The relevance of
these findings to current ECT practice is questionable, because many patients
now referred to ECT have medication resistance. Indeed, the only study
to have examined this issue in recent times reported a relapse rate that
was considerably higher than that of earlier studies (57). In that study,
85% of the patients had received treatment for the index episode of depression
prior to receiving ECT. The 6-month relapse rate among patients randomized
to paroxetine from the start of ECT was 47%, compared with a rate of 65%
among patients randomized to placebo. Mayer and others compared patients
who started antidepressant medication in conjunction with ECT with patients
who started antidepressant medication after the last ECT treatment (58).
Although these authors did not specifically examine relapse rates, they
found no difference between the groups in mean Hamilton Depression Rating
Scale (HDRS) scores 6 weeks post-ECT. Further, consistent with previous
studies (54–56), starting antidepressant medication during ECT conveyed
no benefit in terms of speed or eventual rate of response (58). Thus, the
benefit of simultaneously treating patients with antidepressant medication
and ECT to prevent relapse has yet to be proven.
Studies of patients with psychotic depression also reveal a high post-ECT
relapse rate. Flint and Rifat reported that 53% of elderly patients with
psychotic depression suffered a relapse or recurrence within 2 years of
response to ECT, despite continuation treatment with an adequate dose of
nortriptyline (49). In this study, ECT was given as a first-line treatment,
and the high rate of relapse and recurrence could not be attributed to
prior treatment resistance. Prompted by the observation that psychotic
depression has a better rate of response to combined antidepressant and
antipsychotic medications than to antidepressant monotherapy, Meyers and
others investigated whether combination therapy improved post-ECT outcome
in older patients with psychotic depression (59). These investigators,
however, found that a combination of nortriptyline and perphenazine was
no more effective than nortriptyline alone in preventing relapse of psychotic
depression following response to ECT.
1 | 2| 3 | 4
|
