Toxicology of Antidepressant Drugs: Tricyclic Antidepressants


Animal Toxicity

General Toxicology

The LD 50 values for a number of tricyclic antidepressants, when administered to mice and rats in single oral or parenteral doses, are listed in Table Acute LD50 valuesa of some tricyclic antidepressants. Acute poisoning by tricyclic antidepressants usually leads to symptoms of central excitation followed at the higher and lethal dose levels by central inhibition. The symptomatology includes muscular weakness, twitching, stupor, respiratory disorders, ataxia, and tonic-clonic convulsions.

Table Acute LD50 valuesa of some tricyclic antidepressants

Imipramine Doxepine Nortriptyline Viloxazine Maprotiline
Mouse i.v.




15- 20







660- 900

Rat i.v. p.o. 22


13- 19






38- 52


a The values given are for LD50, single administration, in mg/kg body weight

It is evident from Table Acute LD50 valuesa of some tricyclic antidepressants or from the reports of Pluviage () and of Ueki et al. () that no major differences in the acute toxicity of tricyclic antidepressants are apparent. Information on animal studies relating to the tolerance of tricyclic antidepressants upon repetitive administration is relatively scarce. A large amount of animal data on individual compounds are contained in the documentation that is submitted to the Health Authorities. Published information on the results of long term tolerance in animals covering the whole range of toxicity assessment is, however, scarce and generally lacking in detail. Upon reviewing the data available in the literature the following conclusions may be drawn:

a) In response to comparatively small doses of imipramine in the range of 10 mg/kg, rats show a decrease in appetite and fail to gain weight at the normal rate.

b) In guinea-pigs treated with i.p. doses of 60-90 mg/kg imipramine daily, some renal tubular dilatation was observed after 30-60 days; there was no sign of cardiotoxicity however. Intramuscular injections of 30 and 40 mg/kg daily for 30 days resulted in cerebral, hepatic, and renal changes.

c) In rabbits, the striking feature of the response to repeated doses of imipramine is the discrepancy between the tolerance to the drug given by the oral and by the subcutaneous routes. Whereas oral treatment did not affect the animals’ body weight until a dosage level of 100 mg/kg daily was reached, subcutaneous injections of 10 mg/kg daily were sufficient to cause a decrease in body weight gain ().

d) Oral treatment with imipramine in daily doses of 20 mg/kg was relatively well tolerated by rats for 1 year, and by dogs for 6 months, the internal organs showing no histologic changes at autopsy. However, daily oral doses of 60 and 160 mg/kg administered for 360 days, resulted in some hepatotropic effects. In another group of rats treated with oral doses of 60 mg/kg daily the same histologic picture was observed after 168 days.

e) In dogs which had received imipramine in daily oral doses of 60 mg/kg for 6 months, examination at autopsy disclosed no treatment-related changes of organs or tissues ().

f) Rats treated orally with 10 mg maprotiline/kg daily for up to 78 weeks showed minimal effects; 30 mg/kg was tolerated by about two-thirds of the animals and 60 mg/kg/day resulted in the death of more than half of the animals within a 1 year period. Apart from some fatty change in the liver of a number of rats at the higher dose levels that was reversible on withdrawal of medication, no pathology was produced by the treatment.

g) On repetitive oral administration of maprotiline to dogs, 20 mg/kg daily was tolerated by most animals over a 1 year period. A dose of 10 or 1 mg/kg daily given for the same amount of time did not give rise to any adverse effects. At all dose levels, no drug-related changes were seen in any organs or tissues ().

h) Rats tolerated daily s.c. doses of 20 mg nortriptyline hydrochloride per kg body weight for 246 days with no drug-induced pathological changes except a local irritant effect at the site of injection. There was no significant alteration in food intake. Dogs given a daily dose of 40 mg/kg succumbed and priors to death showed emesis, mydriasis, ataxia, and occasional seizures. 20 mg/kg daily for 1 year was tolerated with no deleterious drug effects. Rats maintained for 1 year on a diet containing 1,500 ppm of nortriptyline showed no hematologic or histologic changes. A dose-dependent retardation in growth rate was seen in rats given 750 and 1,500 ppm in the diet ().

i) Dogs treated with doxepin at 25 and 50 mg/kg daily for up to 12 months exhibited slight emesis, ptosis, sedation, and twitching but showed no histopathological changes of their tissues attributable to the drug treatment.

k) When doxepin hydrochloride was fed to rats at dose levels of up to 100 mg/kg daily, fatty metamorphosis of the liver was observed primarly in the male rats at 100 mg/kg. At the same dose level there was an inhibition of body weight gain in the females. Doses of 50 mg/kg daily may be considered as a threshold level in that these dose levels produced no adverse effects in an 18 months study while in another study of 12 months duration, slight hepatic fatty metamorphosis was observed. Lower doses were tolerated without adverse effects ().

Interaction Experiments

The toxicity of a number of tricyclic compounds has been shown to depend on a variety of conditioning factors such as stress, hormonal influences, and the effects of other pharmacologically active agents.

In mice, the acute LD 50 of imipramine depends on the number of animals per cage (). Crowding also increased the toxicity of desmethylimipra-mine in mice. The regression line from desmethylimipramine was steeper under crowded conditions than under uncrowded conditions (). In hyperthyroid animals, tricyclic antidepressants proved more toxic than in euthyroid condition (). In rats, orally administered activated charcoal was found to have only limited value as an antidotal adsorbent in imipramine or desipramine poisoning ().

Pretreatment with α- or β-receptor blockers had no measurable effect on the toxicity of imipramine, but the drug’s toxicity was increased when it was injected simultaneously with ethanol. The combined administration of ethanol together with amitriptyline, trimipramine, imipramine, and nortriptyline to mice increased the duration of coma and loss of righting reflex due to alcohol, mostly two to fourfold, while desipramine protected the mice from ethanol-induced loss of righting reflexes ().

A normal corticosteroid level is the most favourable prerequisite for survival during the first 6 h after an animal has received a toxic dose of imipramine. In the subsequent recovery phase, treatment with adrenocortical hormones improves the animal’s chances of survival ().

A higher death rate was observed when imipramine was given together with a MAO inhibitor than when it was administered alone (). Maxwell () showed that the administration of imipramine and amitriptyline to rabbits pre-medicated with high doses of tranylcypromine, nialamide, or phenelzine was followed by hyperexcitement and fatal hyperpyrexia in a large proportion of animals. The author also found that imipramine, amitriptyline, and trimipramine differed in their ability to produce side effects in rabbits premedicated with MAO inhibitors and that the particular MAO inhibitors used also played a role. In a study on the effects of a number of drugs on the ECG changes induced by amitriptyline, Nymark and Rasmussen () reported that dichloroisoprenaline and propranolol slowed down heart rates but otherwise did not affect the electrocardiogram. Attree et al., () reported on the effects of the administration of tricyclic antidepressants on the cardiotoxicity of digoxin in the rat. In some experiments, rats were stressed by repeated transfer from cage to cage. After acute administration only desmethylimipramine caused a significant increase in the toxicity of digoxin. While stress itself did not significantly increase the lethality of digoxin, stress in the presence of either amitriptyline or imipramine did increase digoxin lethality. When administered chronically, both imipramine and desmethylimipramine significantly increased the cardiotoxicity of digoxin.

A series of experiments was run on the effects of cholinesterase inhibitors, propranolol, and mecamylamine on the toxicity of amitriptyline or protriptyline. Amitriptyline induced tachycardia and neurological signs in all animals at subcutaneous doses of 50 or 70 mg/kg. The chronotropic actions were reversed by all drugs used and the neurological effects favorably influenced only by physostigmine. Protriptyline at 50 mg/kg produced less tachycardia, no neurological signs, and its effect on the heart were less sensitive to physostigmine antagonism. The results suggested that the toxicity of amitriptyline or protriptyline may be due to the anticholinergic activity of the drugs; the effects on the heart resulting from sympathetic activation ().

The interaction of various tricyclic drugs with d-amphetamine toxicity has been studied under standard conditions in nonaggregated mice. Pre- or posttreatment with promazine protected the animals in all cases. The tertiary amines, imipramine and amitriptyline, had either no effect or exerted a protective action; the secondary amines, desmethylimipramine and nortriptyline, either potentiated or reduced the toxicity of d-amphetamine, depending on the dosage schedule with respect to time ().

In mice treated with lethal doses of amitriptyline, the number of fatalities was significantly reduced by physostigmine, pilocarpine and pyridostigmine (). In another study, Nymark and Rasmussen () reported that drugs which interfered with cholinergic transmission (e.g. prostigmine, pyridostigmine) markedly increased the survival of animals poisoned by amitriptyline and exercised a beneficial effect on the ECG changes.

Vance et al. () reported that subtoxic doses of physostigmine potentiated the convulsive toxicity and lethality of amitriptyline and imipramine in mice. The seeming discrepancy with the results reported by Schaerer () is explained by the 100- to 1,000-fold increased doses of physostigmine used in this study. Schaerer () reported that in his experiments higher doses of cholinergic agents increased the lethality since the gut peristalsis which is reduced in tricyclic intoxication is increased by the cholinergic drugs, suggesting the effect of increased resorption.

Based on their experiments, Lee and Spencer () showed that morphine would better be combined with maprotiline than with clomipramine. Tofanetti et al. () reported that doses of doxepin, in themselves lacking any analgesic effect, remarkably enhanced the analgesic activity of propoxyphene. On the other hand, the data proved that doxepin did not significantly alter the acute toxicity of propoxyphene.

In amitriptyline-treated mice, luminal and chlordiazepoxide halved fatalities (); imipramine convulsions could be prevented by diazepam (). In a study on the effects of prolonged treatment of rats with nortriptyline and amitriptyline on the acute toxicity of 26 neuropharmacologic agents, Meyers et al. (1966a) showed that the toxicity of Na-pentobarbital, Na-secobarbital, methapyrilene-HCl, chlordiazepoxide and physostigmine was increased but the toxicity of isoproterenol, atropine sulfate, chlorpromazine, and ephedrine was reduced.

Cardiovascular Effects

Pharmacologic studies indicate that overdosage with tricyclic antidepressants has a complex action on the heart, affecting α- and β-adrenergic receptors, cholinergic receptors, catecholamine metabolism, and atrioventricular and intraventricular conduction (). In response to relatively high i.v. doses of tricyclic antidepressants, evidence of disturbances affecting cardiac conduction and repolarization was observed in the ECG ().

Experiments on anesthetized guinea pigs which were infused until death and additional in vitro studies on guinea pig atria revealed that i.v. infusion of imipramine, amitriptyline, nortriptyline, and doxepin produced sinus tachycardia, repolarization disturbances, prolongation of the corrected QT-interval and disturbances of atrioventricular and intraventricular conduction with atrioventricular and cardiac arrest in asystole. No significant difference between the tricyclic antidepressants was observed. The in vivo model was, therefore, considered to relate to the situation observed in humans suffering from acute intoxication of a tricyclic antidepressant. Guinea pigs infused with doxepin at a constant infusion rate survived significantly longer than those infused with amitriptyline, imipramine, or nortriptyline. In these experiments, sodium bicarbonate had no effects on the arrhythmias induced by tricyclic antidepressants; propranolol, apart from counteracting the tachycardia induced, was without further effect on the ECG.

In the isolated spontaneously beating guinea-pig atria, a slight but significant decrease in the rate and force of contraction was found with several tricyclic drugs at 10-2mM. It was concluded that an interference in impulse conduction could be responsible for the arrhythmogenic effect of tricyclic antidepressants (). Further experiments performed on perfused isolated guinea-pig hearts and on intact guinea pigs, rats, and rabbits showed that amitriptyline had a negative inotropic effect but increased the coronary flow and influenced cardiac frequency, modified the electric axis of the heart, and caused disturbances of ventricular repolarization including block of right branch, supraventricular tachycardia and auricular -ventricular dissociation (). When the effect of imipramine on the rat and guinea-pig isolated myocardial fiber was studied by means of intracellular electrodes, it was shown that imipramine decreased the membrane and action potentials, increased the conduction time and excitation threshold and slowed the depolarization velocity. The effect of imipramine on the action potentials of the isolated myocardial fiber was transiently abolished by sodium lactate, sodium bicarbonate, and calcium chloride. Whereas the calcium salt did not modify the increase in conduction time, the sodium salts almost completely abolished the action of imipramine (). In rabbits, large doses of imipramine also gave rise to changes affecting cardiac activity (). In the rabbit isolated atrium and aortic strip, the cardiac effects of four tricyclic antidepressants correlated better to their “receptor blocking effect” than to their effect on the noradrenaline uptake ().

Experiments on dogs showed that, depending on the dose, the force and rate of heart contractions as well as the aortic and coronary flow were increased. This was followed by a marked depression of contactility with decreased aortic flow and blood pressure followed by death in ventricular arrhythmia (). In dogs with an artificial atrioventricular block, increasing doses of tricyclic antidepressants provoked changes in ventricular automaticity consisting of markedly lengthened ventricular intervals. In dogs treated with 200 mg amitriptyline/ kg, administration of 25 meg mestinon per kg, i.m., brought about rapid normalization of the severely raised heart rate and the altered ECG (). In studies in young anesthetized dogs, hyperventilation, sodium bicarbonate, or trishydroxyaminomethane abolished the arrhythmia produced by tricyclic antidepressants when the blood pH was over 7.4. The best treatment of the intoxication was either physostigmine or sodium bicarbonate (). In cats, cardiac changes involving ST-T wave and myocardial conduction were demonstrated during chronic amitriptyline treatment. Further investigations in the same experimental model led to the conclusion that stress increases the prominence of tricyclic-induced electrocardiographic abnormalities ().

In rats given oral doses of tricyclic antidepressants twice daily, uncoupling of oxidative phosphorylation in heart mitochondria with concomitant changes in oxygen consumption was seen. With maprotiline, the effect appeared after 2 weeks while with amitriptyline, protriptyline, and imipramine it was seen after only two doses. It was concluded that the treatment had caused a mild to moderate impairment of the heart mitochondrial function suggesting that the effect is due to nonspecific binding of the drugs to lipid mitochondrial membranes ().

Induction of Myeloid Body Formation

A variety of drugs and other compounds may, upon single or repeated administration of high enough doses, induce the formation of lamellated cytoplasmic particles with distinct morphology resembling so-called myeloid bodies () in cells of the parenchyma of the liver () and in other tissues of laboratory animals. This morphological alteration, usually associated with intracellular accumulation of polar lipids, is elicited by compounds consisting of a hydrophobic ring system linked to a hydrophilic aliphatic side chain bearing a terminal cationic amine group (). Agents of that sort, also termed amphiphilic may interact with complex membranous structures forming complexes with phospholipids by virtue of electrostatic and hydrophobic forces (). A number of psychotropic drugs, including tricyclic antidepressants, belong to this category and it may be speculated that distinct biologic effects (such as the interference with catecholamine uptake through the cell membrane of sympathetic neurons) may depend on the particular type of membrane interaction ().

Based on ultrastructural studies, the antidepressants iprindole, imipramine, clomipramine, noxiptiline, maprotiline, amitryptiline and the experimental drug 1-chloroamitryptiline have been reported to induce myeloid body-like changes in animals (). The latter compound, similar to chlorphentermine and to a number of tricyclic compounds (), induced, apart from lamellated inclusion bodies, a remarkable increase in alveolar macrophages together with a change in their phospholipid pattern (). Although induced formation of myeloid bodies is usually reported in the liver and, with certain compounds, in steroid-producing endocrine tissues, myeloid bodies may form in a variety of other cell types. The degree of formation of these particles appears to depend both on the amphiphilic compound used and on the local tissue response (); the latter being dependent, at least in part, on the pharmacokinetics and tissue distribution. Myeloid bodies have been produced in in vitro systems (). They have been described in cells of normal animals ().

It has been conclusively demonstrated that myeloid bodies can be assigned to the lysosomal class of cytoplasmic organelles () and that their formation is a reversible process (). The reversibility of the alterations induced by amphiphilic drugs has been documented for a number of compounds. It is thought to depend on the dissociation rate constant of the complex formed between the amphiphilic compound and the phospholipoprotein, and on the actual concentration of drug present ().

Among the antidepressants, maprotiline has been extensively studied with regard tothemorphogenetic,cytochemical and biochemical aspects of myeloid bodies formed in rat liver (). The behavior of subcellular fractions obtained from livers of rats receiving five consecutive daily doses of 200 mg/kg and subjected to velocity-centrifugation in shallow sucrose gradients supported the assumption, on cytochemical evidence, that the induced inclusion bodies are lysosomes. The appearance of these particles, morphologically distinct from normal lysosomes, did not substantially alter the distribution of constitutive enzymes (and hydrolases) which were, however, reversibly displaced to lower sucrose densities, possibly due to the accumulation of phospholipid-rich material within the myeloid bodies. Since the enzyme distribution among the various cell compartments and their fractional activity was virtually unaltered, there is every reason to assume that myeloid body formation per se does not indicate a toxic effect but is compatible with the normal physiology of the cell.

The observations made with maprotiline caution against the indiscriminate use of the term “drug-induced phospholipidosis” coined by Shikata et al., () to describe the tissue changes induced in man and in rats by 4,4′-dimethylaminoethoxyhexoestrol, an amphiphilic coronary vasodilator. The rationale of this term, used in analogy with inborn lipid storage diseases, consists mainly of the structural resemblance of the phospholipid-rich myeloid bodies to the inclusion bodies found in some human gangliosidoses (). Biochemically however, the various inborn lysosomal storage diseases are characterized by a primary defect of a particular hydrolase or biosynthetic enzyme and the progressive accumulation of storage material in a hypertrophied lysosomal compartment, i.e., features that are quite different from the effects produced by maprotiline ().

The available evidence on maprotiline is compatible with the conception that myeloid bodies induced in the animal by this drug, as well as presumably by other tri-cyclic antidepressants in common use, represent transient structures, reversible after discontinuation of treatment (). The limited toxicologic relevance of this feature is borne out further by the fact that long-term treatment of rats with this drug did not produce any tissue changes of an irreversible nature ().

Tricyclic Antidepressants: Teratology

Tricyclic Antidepressants: Intoxication in Man


Incoming search terms:

  • is nortriptyline a controlled substance