Download as a PDF

CLIN. CHEM. 34/5, 856-858 (1988)
CardiovascularEffectsof the Standard TncyclicAntidepressants
AleXander H. Glaasman, Randy Pardell, and Sally Woodring
Cardiovascular
effects of tncyclic antidepressants are re-
viewed.
In the early 1970s, as methods gradually developed for
measuring tricyclic antidepressants (TCAs), it became clear
that there were large interindividual
differences in metabolism, and that these resulted in large differences in plasma
steady-state concentrations of various TCAs from one pa-
tient to another. Initially,
the interest in this metabolic
variability
was focused around the question of whether
rapid metabolism resulted in patients with low concentrations in plasma and poor response to these drugs. Only
gradually
did interest shift to include patients who were
slow metabolizers
and had high concentrations in plasma,
and to the question of these high concentrations producing
toxicity.
Since the TCAs were first discovered in the 1950s, it has
been apparent that overdoses of these drugs can result in
death, most frequently cardiac in origin (1). As the interest
in metabolic variability
began to include slow metabolizers
with high concentrations of the drug in plasma, the most
important
question seemed to be whether these serious
cardiac consequences could follow from this metabolic variability. However, the issues are more complex than were
really appreciated, and the cardiac pharmacology
of these
drugs, as we understood
them in the mid-1970s, has
changed enormously over the last decade. In many ways, the
ability to measure plasma concentrations of these drugs has
contributed
heavily to our changing understanding
of the
cardiovascular toxicity of the TCAs.
Early reports on imipramune and other tricyclic drugs
indicated that otherwise-normal patients occasionally developed an abnormal electrocardiogram
(EKG) when they took
these medications. Only gradually has it become clear that
all of the standard tricyclic drugs will delay cardiac conduction. The degree of this delay depends both on the concentration of the drug and the pre-existing cardiac disease of the
patient. In a depressed but otherwise healthy patient,
therapeutic doses of the various antidepressant drugs may
or may not influence conduction times. These times are
characteristically measured by the PR, QRS, or QT intervals in the standard EKG. Thus patients on normal dosesof
antidepressant drugs may show no change or only minor
prolongations of these conduction measures, or occasionally
they even develop what would appear to be EKG abnormalities such as first-degree heart block (slowed atrial to ventricular conduction) or bundle-branch
(intraventricular conduction) patterns. In general, these tricycic-induced conduction
delays are of no clinical significance. Although they may or
may not be present in patients who are receiving the usual
therapeutic oral dosages, as the concentration of the drug
increases in the plasma these conduction changes become
consistent phenomena, and in overdose it is the prolonged
conduction delays that are characteristic of the syndrome
and likely to produce serious sequelae.
ColumbiaUniversity College of Physicians & Surgeons, 722 West
168th St., New York, NY 10032.
856
CLINICAL CHEMISTRY, Vol. 34, No. 5, 1988
Vobra et al. (2) were among the first to identify precisely
where TCAs exert their influence on conduction. They
measured cardiac conduction using intracardiac
His bundle
electrocardiography, and they showed that the All interval
(atrial or nodal conduction) was little affected by these
drugs. They found that the major effect was on HV (intraventricular)
conduction, and this has subsequently been
supported by electrophysiological studies of others (3, 4).
Although the prolongation of conduction induced by TCAs
is seldom significant in people who are free of heart disease,
in those patients with pre-existing intraventricular conduction disease a significant risk develops with these drugs. We
have recently completed a prospective study of 41 depressed
patients with first degree AV block and (or) bundle-branch
block compared with 151 patients with normal pretreatment EKGs (5). Both groups were treated with therapeutic
dosages of imipramine or nortriptyline. The rate of 2:1
atrioventricular block was significantly (P <0.05) higher
(9%) in patients with pre-existing bundle-branch block as
compared with the rate (0.7%) in patients with normal
pretreatment EKGs. In fact, the one patient with a normal
pretreatment EKG who developed 2:1 block on tricyclics had
an abnormal His Purkirje
conduction that was apparent
only on catheterization. This type of block is a symptomatic
extension of the conduction delays seen on the EKG. In this
condition some of the impulses beginning in the atrium fail
to reach the ventricles and actual ventricular beats are
skipped. This can progress to complete heart block, cardiac
standstill, and death.
Because the TCAs can cause severe arrhythmia
in overdose, it has been widely assumed that their use would be
contra-indicated
in patients with pre-existing arrhythmias.
Although this logic is attractive, it has not proved to be the
case. During the past 10 years it has been established that
at antidepressant concentrations in plasma approximating
therapeutic levels, these drugs have powerful anti-arnhythmic activity. The anti-arrhythmic effect of imipramine was
first reported in 1977 (6). Subsequent studies have shown
that imipramme is similar to quinidine in its anti-arrhyth.
mic activity. The available data indicate that nortriptyline
is similar to imipranune with respect to its anti-arrhythmic
activity,
and that other TCAs also show similar
antiarrhythmic
activity (7). These drugs share electrophysiological properties characteristic of Type 1 anti-arrhytbmic
compounds and currently are being used in cardiac-disease
patients for the purpose of controlling
mias.
ventricular
arrhyth.
The confusion in the literature
has arisen primarily
because, in very severe overdoses, patients can developand can die of-ventricular
arrhythmias.
However, this is
not a primary arrhythmogenic activity of these drugs, but is
due to the previously described extensive conduction delays
that develop in these patients. As the QRS interval begins to
exceed 160 ma these patients become particularly
vulnerable to re-entry arrhythmias
and “circus” rhythms (8). To
treat these arrhythmias
with standard
anti-arrhythmic
drugs would only make the condition worse. Arrhythmias
in
the presence of tricyclic
overdose represent a very serious
sign, and they are not seen in mild overdoses. An overdose of
manifesting ventricular
arrhythmias
should be
considered critical.
Until the early 1980s it had been widely assumed that the
tricyclic antidepressants adversely affect left ventricular
function (LYF). A number of earlier studies on nortriptyline, ainitriptyline,
and imipramine produced evidence of
TCA-induced impairment
of LVF (9, 10). However, the
criterion used to assess LVF was usually the systolic time
interval, a measurement that depends in part on the duration of the QRS interval. Because QRS prolongation is
characteristic of the effects of TCAs, it is possible that the
effect of TCAs on the systolic time interval in these early
studies was due to their effect on conduction rather than on
left ventricular
function. The introduction of radionuclide
angiography provided a more reliable means of assessing
TCA effect on left ventricular function. Using this angiography technique, Veith et al. (11) showed that both mupramine and doxepin had no effect on LVF, either in the resting
state or after exercise. However, they used only relatively
low doses of the drug, and they included few patients with
seriously impaired ejection fractions. Subsequently, our
group reported radionuclide data on a series of 15 depressed
patients with moderate to severe impairment of their ejection fraction (12). These patients were all treated to produce
therapeutic concentrations of imiprarnine in their plasma,
and, as Veith et al. had observed, the drug had no measurable effect on left ventricular function. However, it did
produce intolerable orthostatic hypotension in about half of
the patients (imupramine is measured as the parent compound, and its desmethyl metabolite). Subsequently, we
replicated this finding in a second series of 10 patients with
heart failure, where again there was no further impairment
in LVF, but half of the 10 patients developed orthostatic
hypotension
so severe that they were unable to stand.
Thus, in examining
the direct effects on the heart of
standard tricyclic antidepressants, the only serious adverse
sequelae are conduction delays. These effects occur irregularly at the usual therapeutic doses, but their occurrence
increases predictably with increasing doses and is characteristic of overdose. No adverse effect on the pump function
of the heart is seen in TCA-treated
patients even with
seriously impaired
ventricular function. And, finally, in
patients with arrhythmias, the TCAs will usually prove
beneficial.
The most frequent serious cardiovascular
side effect of the
tricyclic antidepressants is not cardiac at all, but vascular.
tricyclics
This is orthostatic hypotension, which can result in fractures, lacerations, and sudden death (13). By far the most
frequently studied drug in this context is imipramine. The
data available on the orthostatic effects of the other tricyclic
drugs are not as complete, but work by Nelson et al. (14),
Hayes et al. (15), Kopera (16), and our own group, strongly
implies that there is an orthostatic effect for amitriptyline,
desipramine, and doxepin as well.
Nortriptyline is the only TCA currently marketed for
which a decreased risk of orthostatic hypotension has been
clearly documented. Freyschuss et al. (17) reported that the
orthostatic effect of nortriptyline in 40 healthy and relatively middle-aged depressed patients was negligible. Thayssen
et al. (18) also reported that there were no posture-related
systolic pressure drops with nortriptyline.
In the United
States, our group has also clearly documented this lesser
risk of orthostatic hypotension with nortriptyline (19).
A recent study of over 100 000 elderly subjects in the state
of Michigan (20) showed that during a three-year period this
population suffered more than 1000 hip fractures. There was
a strong association between the use of psychotropic medication and the risk of occurrence of these fractures. The odds
that a patient taking a tricyclic antidepressant would suffer
a hip fracture were three times greater with even modest
doses of these drugs than for comparably aged patients who
were free of such medication. Given the morbidity and
mortality,
not to mention the costs involved in these frac.
tures, this is not a trivial issue.
The mechanisms by which tricyclics induce orthostatic
hypotension are not entirely clear. Interestingly, a significant risk factor for the development of orthostatic hypotension with imipramine may be the diagnosis of depression
itself. Giardina et al. (21) reported that, of 22 non-depressed
cardiac patients with some degree of congestive failure
(mean ejection fraction of 33%) who were being treated with
imipramme
for arrhythmia,
only one discontinued medication because of orthostatic hypotension. This compares with
a rate of approximately 50% of the patients developing
intolerable orthostatic hypotension in our depressed patients with a similar degree of left ventricular impairment
who were treated to produce comparable concentrations of
imipramine in plasma. Thus it would seem that, when
treated with imipramine, patients with heart disease and
melancholia
are at significantly
greater risk of developing
orthostatic hypotension than are patients with heart disease
alone.
What is the incidence of orthostatic
hypotension
in patients who are being treated with imipramine? It can vary
from 1% for cardiovascular patients free of depression to 5%
or 6% for melancholic patients free of heart disease, to 25%
for cardiac patients with hypertension
but who are free of
cardiac failure, to almost 50% for patients with both melancholia and heart failure. Thus, the incidence of orthostatic
hypotension
is not simply related to the antidepressant
drug; it involves both the illness of the patient and other
drugs that the patient may be ingesting. Although nortriptyline is not free of orthostatic hypotension, it does have a
well-documented clinical advantage over the other standard
tricyclic antidepressants in those patients at particular risk
for this adverse event. Many of the newer antidepressants
would appear to be free of cardiac effects in depressed
patients without heart disease. However, none of these
compounds has been carefully tested in patients with both
depression and cardiovascular difficulties. Thus, comparison
with standard antidepressant drugs remains difficult.
References
1. Williams Jr RB, Sherter C. Cardiac complications of ti-icyclic
antidepressant therapy. Ann Intern Med 1971;74:395-8.
2. Vobra J, Burrows GD, Hunt D, Sloman G. The effect oftoxic and
therapeutic doses of tricyclic
antidepressant
drugs on intracardiac
conduction. Eur J Cardiol 1975;3:219-27.
3. Weld FM, Bigger Jr JT. Electrophysiological effects of imipramine on ovine cardiac Purkinje and ventricular muscle fibers. Cu-c
Res 1980;46:167-75.
4. Rawling DA, Fozzard HA. Effects of unipraimne on cellular
electrophysiological properties of cardiac Purkinje fibers. J Pharmacol Exp Ther 1979;209:371-5.
5. ReeseSP, Glassman AN, Giardina EGV, Walsh BT, Woodring 5,
Bigger Jr JT. Tricyclic antidepressants in depressed patients with
cardiac conduction disease. Arch Gen Psychiatry 1987;44:273-5.
6. Bigger Jr JT, Giardina EGV, Perel JM, Kantor SI, Glassman
AH. Cardiac antiarrhythmic effect of imipramine hydrochloride. N
Engi J Med 1977;296:206-8.
CLINICAL CHEMISTRY, Vol. 34, No. 5, 1988 857
7. Giardina EGV, Johnson IL, Vita J, Bigger Jr JT, Brem RF.
Effect of imipramine and nortriptyline on left ventricular function
and blood pressure in patientstreatedfor arrhythmias. Am Heart J
1985:109:992-8.
8. Boehnert MT. Lovejoy Jr PH. Value ofthe QRS duration versus
the serumdruglevel in predicting seizures and ventricular arrhythmiss after an acute overdose of tricyclic antidepressants. N EngI J
Med 1985;313:474-9.
9. Burckhardt D, Raeder E, Muller V, Imhof P, Neubauer H.
Cardiovascular effects of tricydic and tetracycic antidepressant&J
Am Med Assoc 1978239:213-6.
10. Taylor DJ, Braithwaite HA. Cardiac effects of tricycic antidepressant medication: a preliminary study of nortriptyline. Br Heart
J 1978;40:1005-9.
11. Veith RC, Raskind MA, Caldwell JH, Barnes RF, Gumbrecht
G, Ritchie JL. Cardiovascular effects of tricydic antidepressants in
depressed patients with chronic heart disease. N Engl J Med
1982;306:954-9.
12. Glassman All, Johnson LL, Giardina EGV, et al. The use of
imipramine in depressed patients with congestive heart failure. J
Am Med Assoc 1983;250:1997-2001.
13. Glassman AN, Bigger Jr JT, Giardina EV, Kantor 5!, Perel
JM, Davies M. Clinical characteristics of imipramine-induced orthoatatic hypotension. Lancet 1979i:468-72.
14. Nelson JC, Jatlow P1, Bock J, Quinlan DM, Bowers Jr MB.
Major adverse reactions during desipramine treatment: relation-
858 CLINICAL CHEMISTRY, Vol. 34, No. 5, 1988
ship to plasma drug concentrations, concomitant antipsychotic
treatment, and patient characteristics. Arch Gen Psychiatry
1982;39:1055-61.
15. Hayes JR. Born GF, Rosenbaum AN. Incidence of orthostatic
hypotension in patients with primary affective disorders treated
with tricycic antidepressanta. Mayo Clin Proc 1977;52:509-12.
16. Kopera H. Anticholinergic and blood pressure effects of mianserin, amitriptyline and placebo. Br J Clin Pharmacol 1978;5:29s34s.
17. Freyschuss U, Sjoqvist F, Thck D, Asberg M. Circulatory effects
in man of nortriptyline, a tricyclicantidepressantdrug. Pharmacol
Clin 1970,2:68-71.
18. Thayssen P, Bjerre M, Kragh-Sorensen P. et al. Cardiovascular
effects of imipramine and nortriptyline in elderly patients. Psychopharmacology 1981;74:360-4.
19. Roose SP, Glassman MI, Sins SG, Walsh BT, Bruno RL,
Wright LB. Comparison of imipramine- and nortniptyline-induced
orthostatic hypotension: a meaningful difference. J Clin Psychopharmacol 1981;1:316-9.
20. Ray WA, Griffin MB, Schaffner W, Baugh DK, Melton ifi U.
Psychotropic drug use and the risk of hip fracture. N Engl J Med
1987;316:363-9.
21. Giardina EGV, Johnson IL, Vita J, Bigger Ji- JT, Brem RF.
Effect of imipramine and nortriptyline on left ventricular function
and blood pressure in patientstreatedfor arrhythmias.Am Heart J
1985;109:992-8.