Medical toxicology

Chest pain is a manifestation of cocaine toxicity, but whether it is of an ischemic origin is unclear. A reexamination of the role of acetylcysteine in treating acetaminophen overdose has led to substantial changes in the indications for its use.

Pharmacodynamic effects of cotinine in abstinent cigarette smokers

The nicotine metabolite cotinine was administered to abstinent cigarette smokers to determine whether it has pharmacologic activity as assessed by various physiologic and subjective measurements. By means of a randomized, double-blind, placebo-controlled counterbalanced-order design, subjects received cotinine base (30 mg) intravenously after 48 hours of abstinence from cigarette smoking. Serum cotinine concentrations increased to levels commonly achieved during daily cigarette smoking, whereas no change in serum nicotine concentration was observed. Cotinine compared with placebo produced subjective differences in self-reported ratings of restlessness, anxiety and tension, insomnia, sedation, and pleasantness. Cotinine had minimal effects on cardiovascular measurements. These findings indicate that cotinine is behaviorally active in the setting of cigarette abstinence at blood concentrations similar to those commonly achieved through daily cigarette smoking.

12-lead and continuous ECG recordings of subjects during inpatient administration of smoked cocaine

Cocaine can cause myocardial ischemia or infarction. The incidence of these events, and the influence of specific dosing routes or regimens on their occurrence is not established. In the current study, we obtained frequent 12-lead electrocardiograms (ECGs) and continuous 2 or 3 channel ECGs from 20 subjects participating in a behavioral study of smoked cocaine. Subjects received 10 or 11 doses of cocaine 0.4 mg/kg per dose, or 10 doses of 35 mg per dose at 30 min intervals (range 233-408 mg total dose per session). ECGs were also recorded on control days on which subjects received no cocaine. The mean peak plasma cocaine concentration on cocaine days was 640 +/- 262 ng/ml. There were no changes in digitized ST segment amplitude on 12-lead ECGs obtained during cocaine administration (P = 0.098). Of 17 subjects who had technically satisfactory continuous ECGs, four had significant ST segment depression (> 1 mm below the PR segment); two on cocaine days and two on control days (P > 0.5). One subject had frequent premature beats on both cocaine and control days. One subject had an asymptomatic run of 4 ventricular beats 30 s after cocaine administration that could have been due to cocaine. All episodes of ST depression or premature beats were asymptomatic. No evidence of either symptomatic or subclinical cardiac ischemia related to cocaine administration was found. Thus no clinically important adverse events were found as a result of smoked cocaine administered by this dosing regimen to healthy males with a history of heavy cocaine use. Additional study with larger numbers of subjects will be helpful in further assessing the safety of administering smoked cocaine to research subjects.

Reversal of desipramine toxicity in rats with polyclonal drug-specific antibody Fab fragments

Drug-specific antibodies, or fragments containing their binding site(s), are a potential means of treating drug overdose. Affinity purified polyclonal ovine Fab (TFab) with a high average affinity constant (Ka = 1.4 x 10(10) M-1) for the common tricyclic antidepressants was evaluated as a possible treatment for tricyclic antidepressant toxicity. Groups of eight anesthetized rats received 30 mg/kg body weight of desipramine (DMI) intraperitoneally, followed after 15 minutes by a 10-minute intravenous injection of 3 ml normal saline solution, 2 gm/kg nonspecific Fab as a control (CFab), or 1 or 2 gm/kg TFab (representing a molar ratio of TFab to DMI of 0.11 and 0.22, respectively). The animals were observed for 3 hours. During the initial 15 minutes, serum DMI levels in the four groups reached 2.3 to 2.9 micrograms/ml, the QRS duration increased by 67.5% to 77.9%, and the systolic pressure fell to between 65% and 85% of its initial value. The group given saline solution showed a gradual return of all these parameters toward normal, whereas CFab caused a transient further QRS prolongation. CFab also caused an initial rise in blood pressure, which then fell progressively, and two of the rats died 2 to 3 hours later with hypotension and bradycardia. Serum DMI concentration did not change significantly in either the saline or CFab groups.(ABSTRACT TRUNCATED AT 250 WORDS)

Methodological issues in the administration of multiple doses of smoked cocaine-base in humans

Many methodological issues exist in human laboratory research with smoked cocaine-base that include safety, precision of dose delivery of smoked cocaine, and the lack of an adequate placebo. All of these issues are particularly apparent with studies involving multiple doses of cocaine. Addressing these concerns is important in conducting parametric studies that require examining dose-response effects. The purposes of this study were to determine: 1) the safest interval between doses to deliver smoked cocaine; 2) the accuracy or reproducibility of administering precise and multiple doses of cocaine; 3) the potential for using a control dose of cocaine; and 4) the influence of multiple doses on these parameters. Six black males were given 10 doses of either 5 or 35 mg of cocaine-base at 15-, 30-, and 45-min intervals. The dependent measures included physiological, subjective, and performance responses. These measures were taken prior to dosing and at specific time intervals after each dose of smoked cocaine. The results showed: 1) dosing at 30-min intervals allowed sufficient time for recovery of blood pressure and heart rate to permit up to 10 doses to be safely administered; 2) reproducible blood cocaine levels were obtained with repeated dosing using a heated wire-coil device; 3) significant differences were observed between the 5- and 35-mg dose with 5 mg being a low enough dose to produce minimal effects; 4) acute tolerance was evidenced with multiple doses of cocaine for most of the measures; and 5) considerable between- and within-subject variability was observed in the pattern of responses to cocaine.

Reversal of desipramine toxicity in rats using drug-specific antibody Fab' fragment: effects on hypotension and interaction with sodium bicarbonate

The effect of drug-specific antibody Fab' fragment on desipramine (DMI) toxicity was studied in anesthetized rats to determine 1) whether DMI-induced hypotension can be reversed, and 2) whether the effect of this Fab' fragment can be enhanced by the concurrent administration of hypertonic NaHCO3. DMI (60 mg/kg) was administered i.p. to produce marked hypotension. Antitricyclic antidepressant (TCA) Fab' (molar Fab'/DMI ratio = 0.11) or control Fab' was administered 15 min later as a 10 min i.v. infusion. The mean arterial pressure was higher at the end of anti-TCA Fab' infusion than after control Fab' (58 +/- 8 vs. 17 +/- 7 mm Hg, P less than .001). In a second protocol, DMI (30 mg/kg) was administered to prolong QRS duration. Anti-TCA Fab' alone (molar Fab'/DMI ratio = 0.09) and NaHCO3 alone both reduced QRS prolongation compared to control treatment, and combined therapy was more effective than either one alone. In both protocols, anti-TCA Fab' markedly increased the total DMI concentration and the bound fraction of DMI in serum, but did not alter the unbound DMI concentration. In the low DMI dose protocol, anti-TCA Fab' also reduced the cardiac DMI concentration. Concurrent treatment with anti-TCA Fab' and NaHCO3 substantially increased urinary DMI and anti-TCA Fab' excretion compared to treatment with anti-TCA Fab' alone.(ABSTRACT TRUNCATED AT 250 WORDS)

Adaptation of the Müller method to allow quantitative characterization of the affinity and cross-reactivity of antibodies by competitive radioimmunoassay

A quantitative expression is derived for the evaluation of antigen-antibody affinity constants from radioimmunoassays for the completely general situation in which antigen and antibody are both multivalent. The theoretical analysis is then extended to encompass quantitative characterization of the competitive inhibition observed in screening tests for cross-reactivity of antibody with structural analogs of the eliciting antigen. These procedures are illustrated with a radioimmunological study of the cross-reactivity of a desipramine-elicited monoclonal antibody with other tricyclic antidepressants. An unexpected finding to emerge from this immunochemical study is the demonstration that a single affinity constant suffices to describe the interaction of desipramine with a polyclonal antibody elicited by this univalent antigen.

Rapid administration of high-dose human antibody Fab fragments to dogs: pharmacokinetics and toxicity

The treatment of drug overdose with drug-specific antibody fragments may require very high antibody doses. To address the feasibility of this therapy, we studied the pharmacokinetics and toxicity of high-dose human nonspecific Fab fragments in beagles. Three dogs received 5.3 g/kg Fab iv over 1 hr. Because nephrotoxicity was observed, three subsequent dogs received 3.2 g/kg. The fraction of the Fab dose excreted in urine (10 +/- 6%) was lower than reported values for either high or low doses of Fab in other species. The terminal serum elimination half-life (42 hr for the higher and 48 hr for the lower dose) was also longer than reported values for other species, due to lower renal and nonrenal Fab clearance. Fab administration was tolerated without adverse hemodynamic effects. One of three dogs at each dose developed transient oliguria. All dogs developed a transient but marked increase in the serum creatinine concentration. At 2 weeks creatinine clearance had returned to normal. Urinary protein and albumin excretion at 2 weeks were within the normal range for dogs but were increased over their baseline values. The histology of all organs was normal at 3 weeks by light microscopy, and renal histology by electron microscopy was also normal. The mechanism of Fab nephrotoxicity, not observed previously with high-dose Fab in rats or lower doses of Fab in other species including dogs, is not clear. These data suggest that further study of the potential toxicity of high-dose Fab, and its reversibility, is needed to assess the feasibility of treating drug overdose with this antibody fragment.(ABSTRACT TRUNCATED AT 250 WORDS)

Determination of the in vivo antigen-antibody affinity constant from the redistribution of desipramine in rats following administration of a desipramine-specific monoclonal antibody

Quantitative expressions have been derived to determine the affinity constant for the in vivo interaction of an antigen with its elicited monoclonal antibody by analysing the redistribution of antigen following antibody administration. Using this method, the intrinsic binding constant for the interaction of subtoxic doses of DMI in rats with anti-TCA was found to be about two orders of magnitude less than the value obtained in vitro. The disparity is probably due to the presence of endogenous ligands for the antibody.

Pretreatment with drug-specific antibody reduces desipramine cardiotoxicity in rats

The effect of a drug-specific antibody on desipramine (DMI) cardiotoxicity was studied in rats. Animals were pretreated i.v. with 4.2 g/kg of a monoclonal antibody (anti-TCA) followed by DMI HCl 30 mg/kg i.p. (molar ratio of anti-TCA binding sites to DMI = 0.56). Peak QRS complex prolongation was substantially lower after pretreatment with anti-TCA than after control antibody (70 +/- 14 v. 21 +/- 4%, p less than 0.001). Time to peak toxicity was the same in both groups. Binding of DMI by anti-TCA was demonstrated by a higher serum total DMI concentration and increased DMI binding in serum after anti-TCA compared to controls. The DMI concentration in anti-TCA treated animals was lower in some organs (brain, lung, liver, spleen), but not in others (heart, muscle, kidney, fat). The calculated fraction of the DMI dose bound by anti-TCA was 19.9%. The steepness of the DMI dose-response curve was examined by administering DMI alone (without antibody) at various doses to rats. Compared to 30 mg/kg DMI, a dose reduction of 30-50% was needed to reduce QRS duration to the same extent as anti-TCA pretreatment. We conclude that DMI cardiotoxicity was markedly reduced by the binding of a relative small fraction of the DMI body burden to anti-TCA. This disproportionate effect of DMI binding was not due to the steepness of the DMI dose-response curve, nor to slowing of the rate of DMI distribution to tissues.

Drug-specific F(ab')2 fragment reduces desipramine cardiotoxicity in rats

Current therapy for cardiotoxicity due to tricyclic antidepressant (TCA) overdose is often ineffective in seriously poisoned patients. We studied the effect of a drug-specific antibody fragment on TCA cardiotoxicity in rats. Animals received anti-TCA F(ab')2 2.0 g/kg i.v. over 10 min starting 15 min after administration of a toxic dose of desipramine (DMI). This anti-TCA F(ab')2 dose was 36.9% of the molar DMI dose in terms of binding sites. Anti-TCA F(ab')2 infusion had no adverse effects and rapidly reduced DMI induced QRS prolongation compared with control F(ab')2 (23 +/- 14 vs 71 +/- 11% QRS prolongation at the end of infusion, P less than 0.001). This beneficial effect lasted for the 45 min duration of the study. Markedly enhanced DMI binding in serum after anti-TCA F(ab')2 was demonstrated by a 48-fold increase in the total DMI concentration over controls and a reduction in the fraction of unbound DMI (44.5 +/- 19.4 vs 0.7 +/- 0.2%). Anti-TCA F(ab')2 reduced the DMI concentration in brain but not in other organs. We conclude that anti-TCA F(ab')2 substantially reduces DMI cardiotoxicity in rats, and does so rapidly enough to be of potential clinical benefit for patients with DMI overdose. Because only a small fraction of the DMI dose was bound by antibody, these data suggest that antibody fragment doses considerably less than equimolar to the DMI dose may be effective in treating DMI cardiotoxicity.

Redistribution of tricyclic antidepressants in rats using a drug-specific monoclonal antibody: dose-response relationship

A monoclonal antibody was used to study the dose-response relationship for antibody-mediated redistribution of tricyclic antidepressants (TCA) in rats. The antibody (anti-TCA) was an IgG1 with Ka = 3.0 x 10(8) M-1 for desipramine (DMI) and 2.2 x 10(8) M-1 for imipramine (IMI). Anesthetized rats received 1 mg DMI (10% of the toxic dose), followed in 15 min by anti-TCA iv at doses representing anti-TCA/DMI molar ratios of 0.003, 0.013, and 0.07. Anti-TCA produced prompt, dose-related increases in the serum DMI concentration of 33, 164, and 776%. Similar results were obtained in rats treated with IMI. The highest dose of anti-TCA reduced the concentration of IMI in the heart. In a second protocol, the anti-TCA dose was kept constant and the DMI or IMI dose varied. The increase in serum drug concentration was 1750, 1260, and 150% (DMI) and 1460, 1200, and 170% (IMI) at drug doses of 0.1, 10, and 1000 micrograms. Thus, the percentage increase in serum drug concentration was diminished only 12-fold (DMI) or 9-fold (IMI) by a 10,000-fold increase in drug dose. At the highest anti-TCA/DMI ratio (lowest DMI dose), tissue DMI concentrations were significantly reduced. We conclude that 1) anti-TCA can effect substantial redistribution of a subtoxic dose of DMI or IMI, even when the antibody dose is less than equimolar to the TCA dose, and 2) the extent of TCA redistribution depends upon the doses of both antibody and drug; anti-TCA is most effective when the body burden of TCA is high. These data support the potential therapeutic use of anti-TCA for DMI or IMI toxicity, and should be useful in anticipating the dose and affinity of anti-TCA required.

Effects of carbamazepine on acute responses to smoked cocaine-base in human cocaine users

A double-blind, placebo-controlled cross-over study was conducted to determine the effects of carbamazepine on the acute physiological and subjective responses to a single dose of smoked cocaine-base. Male cocaine users (N = 6) were given 400 mg carbamazepine or placebo, each for a period of 5 days. At the end of the 5-day period, a 40 mg dose of smoked cocaine was administered. The results showed a significantly higher heart rate, diastolic blood pressure elevation, and blood pressure-heart rate product under the carbamazepine compared to the placebo condition. There were no effects of carbamazepine on the subjective responses from cocaine. The increase in cardiovascular functions indicates a need to be cautious in the use of carbamazepine in the treatment of cocaine abusers.

Lack of effect of hypertonic sodium bicarbonate on QRS duration in patients taking therapeutic doses of class IC antiarrhythmic drugs

Hypertonic sodium bicarbonate (HSB) has been reported to reduce the toxicity of Class IC antiarrhythmic agents in rats and, anecdotally, in patients. A pilot study was conducted of the safety and efficacy of HSB for reversing the electrocardiographic effects of therapeutic doses of encainide or flecainide in ten patients taking these drugs for chronic ventricular arrhythmias. Patients had a mean drug-induced QRS prolongation before treatment of 27.6 +/- 8.8%. Each patient received a single dose of HSB 100 mEq or normal saline IV over 5 minutes on two separate occasions. The administration of treatments was blinded and balanced. There were no important side effects of HSB. Venous blood pH, CO2 content and sodium concentration were all significantly increased by HSB in comparison to saline. No differences were found during the 2-hour observation period in the primary endpoint, QRS duration, the PR or QT intervals, or the frequency of premature ventricular beats. It was concluded that HSB 100 mEq does not reduce QRS duration in patients taking therapeutic doses of flecainide or encainide. Because HSB was well tolerated, investigation of its use in higher doses or in patients with overt toxicity due to Class IC drugs is feasible.

Effects of alpha-1-acid glycoprotein administration on propranolol binding and beta blockade in rats

Alpha-1-acid glycoprotein (AAG), 750 mg/kg, was administered to rats to determine its effect on propranolol binding and beta blockade. Anesthetized rats received [3H]propranolol i.v., followed in 15 min by human AAG or bovine serum albumin, 750 mg/kg. AAG treatment produced a human AAG concentration in serum of 7.76 +/- 1.17 mg/ml, several times higher than the endogenous serum AAG concentration in stressed rats. AAG treatment significantly increased the heart rate response to isoproterenol, compared to albumin (95.4 +/- 19.6 vs 28.3 +/- 16.7% of baseline value, measured 45 min after propranolol, P less than 0.001). AAG-treated rats had greater [3H]propranolol binding in serum (93.0 +/- 3.2 vs 76.7 +/- 3.0%, P less than 0.01) and a lower calculated unbound [3H]propranolol concentration in serum (2.7 +/- 1.3 vs 7.4 +/- 3.1 X 10(6) dpm/ml, P less than 0.001) than albumin-treated rats. These data demonstrate that AAG can alter propranolol pharmacokinetics and pharmacodynamics even when administered after the propranolol effect is evident. Because the reported affinity of propranolol for cardiac beta receptors is 10,000 times greater than its affinity for AAG, these data suggest that AAG acted by altering propranolol disposition rather than by directly competing with beta receptors for drug.

Individual variability in the blood pressure response to intravenous phenylpropanolamine: a pharmacokinetic and pharmacodynamic investigation

The intersubject variability in blood pressure response to 0.44 mg/kg intravenous phenylpropanolamine (d,l-norephedrine) was studied in 10 normal subjects. A phenylpropanolamine or placebo infusion was administered over 45 minutes on separate days according to a double-blind, balanced protocol. Blood pressure increased by 24 +/- 13/16 +/- 7 mm Hg (systolic/diastolic, mean +/- SD) after the phenylpropanolamine infusion and was statistically different from the placebo infusion response (7 +/- 5/8 +/- 3 mm Hg). Phenylpropanolamine infusions were terminated early in two subjects (hyperresponders) after 0.31 and 0.23 mg/kg because of excessive increases in blood pressure (52/30 and 34/21 mm Hg, respectively). The hyperresponders had the lowest peak serum phenylpropanolamine concentrations. These data suggest that considerable intersubject variability exists in the blood pressure response to intravenous phenylpropanolamine. A pharmacokinetic basis for the variability in response to racemic phenylpropanolamine was not observed. A relationship did not exist within the group between blood pressure effect and serum concentration but did exist within each subject. Therefore phenylpropanolamine's blood pressure effect in an individual cannot be predicted solely from a serum concentration of racemic drug.

Suppression of human peripheral blood mononuclear cell function by methadone and morphine

Recent studies have shown that in vitro exposure of peripheral blood mononuclear cells (PBMC) to morphine results in suppressed respiratory-burst activity of monocytes and impaired interferon-gamma (IFN-gamma) production by lymphocytes. To investigate the potential in vivo effect of an opiate on these cell functions, PBMC were obtained from patients maintained on methadone. These freshly isolated mononuclear cells had a significantly impaired capacity to generate superoxide anion (O2-) in response to phorbol myristate acetate (PMA), while production of IFN-gamma by concanavalin A-stimulated cells was intact. After cell culture for 48 h, the defective O2- generating capacity was sustained. Also, culturing PBMC from healthy controls in the presence of methadone or morphine at concentrations as low as 10(-12) M caused significant suppression of PMA-stimulated O2- release. Because reactive oxygen intermediates produced by PBMC may participate in host defense against opportunistic pathogens in AIDS, these results underscore the need for investigations of the biological consequences of opiate-mediated immunosuppression.

Hypertonic sodium bicarbonate partially reverses QRS prolongation due to flecainide in rats

Hypertonic (1M) sodium bicarbonate can partially reverse the cardiac toxicity of some Class IA antiarrhythmic agents, presumably by antagonizing sodium channel inhibition. We studied the effects of 1M sodium bicarbonate on toxicity due to the Class IC drug flecainide. Anesthetized rats received i.v. loading and maintenance doses of flecainide to produce QRS prolongation of 76% that was stable over the 60 min study period. 20 min after the start of the maintenance infusion, groups of 8 rats received an i.v. infusion of 1M sodium bicarbonate (6 meq/kg) or an equal volume of 0.9% saline. QRS prolongation was reduced by 1M sodium bicarbonate but not only 0.9% saline (% change -12.2 +/- 10.0 v. +3.0 +/- 2.7, p = 0.001). Expressed as a percent of the flecainide-induced QRS prolongation, bicarbonate reduced this prolongation by 26.5%. The improvement in QRS duration persisted until sacrifice 30 min later. Compared to controls, the bicarbonate group had a significantly higher blood pH (7.55 +/- 0.06 v. 7.44 +/- 0.05) and serum sodium concentration (149 +/- 5 v. 137 +/- 2 meq/l). Serum flecainide concentrations were similar. These data suggest that 1M sodium bicarbonate can partially reverse flecainide-induced conduction delay in rats. This effect may be due to changes in the extracellular pH and sodium concentration. 1M sodium bicarbonate may be useful in assessing the role of sodium channel inhibition in mediating the toxicity of flecainide or other Class IC drugs.

The effect of phenylpropanolamine on blood pressure during upright bicycle exercise in normal subjects

We studied the effect of a single oral dose (37.5 mg immediate release) of phenylpropanolamine (PPA) in six normal subjects during upright incremental bicycle exercise to symptom tolerance. The study was double blind and placebo controlled with a crossover design. The patients rested supine for 90 minutes after dosing, sat quietly on the bicycle for five minutes, and then began exercise. After PPA systolic blood pressure was increased during supine rest (135 +/- 14 mmHg vs 120 +/- 10 mmHg for placebo; P less than 0.01) and while sitting on the bicycle prior to exercise (131 +/- 8 mmHg vs 121 +/- 8 mmHg for placebo; P less than 0.05). However, PPA did not alter the expected increase in systolic blood pressure with progressive exercise. Systolic blood pressure at maximum exercise (254 +/- 43 watts) was 185 +/- 14 mmHg for PPA vs 182 +/- 13 mmHg for placebo (difference not significant). This suggests that the physiological factors controlling blood pressure during exercise (increased cardiac output, vasodilation of vessels in exercising muscles and vasoconstriction of non-exercising vascular beds) are not altered significantly by a dose of PPA sufficient to increase resting systolic blood pressure, at least in normal subjects.

Clinical features and consequences of seizures due to cyclic antidepressant overdose

The charts of 30 patients with seizures caused by cyclic antidepressant toxicity were reviewed to determine the natural history of these seizures and their effect on outcome. Mortality was 10%. Onset of seizures was within 1.5 hours of admission in all but two patients. Both patients with seizures occurring later than 1.5 hours had also ingested ethanol. Antecedent mental status did not predict the occurrence of seizures; 23% of seizures occurred when patients were alert. The majority of patients (25 of 30) had one or two brief (less than 2 minute), self-limited, generalized, tonic/clonic seizures that terminated without anticonvulsant therapy. Six patients had prolonged or repetitive seizures. Most seizures, even when multiple or sustained, had no sequela. However, four patients developed marked cardiovascular deterioration immediately following seizures, with hypotension and ventricular tachycardia or bradyarrhythmias. All four had a QRS duration greater than 200 msec at some time during their hospitalization. It is concluded that most seizures associated with cyclic antidepressant toxicity occur soon after admission, are brief, and terminate without specific anticonvulsant therapy. Even when brief, however, seizures may occasionally lead to abrupt cardiovascular deterioration and death. It is not clear which patients are at risk for severe cardiovascular complications.

Flu-like symptoms associated with fluoxetine overdose: a case report

The serotonin reuptake inhibitor zimelidine may cause flu-like symptoms and Guillain-Barre syndrome. Guillain-Barre syndrome has not been reported with the use of the structurally related serotonin reuptake inhibitor fluoxetine. Flu-like symptoms are described in the manufacturer's literature on fluoxetine but are absent from published studies. We describe a patient who developed flu-like symptoms, urticaria and angioedema 2 days after fluoxetine overdose. There were no neurologic sequelae. This case confirms that fluoxetine may be associated with flu-like symptoms and suggests that, because these symptoms occurred after overdose, they may be dose related. The additional findings in this patient suggestive of allergy, an unusual feature of drug overdose, are discussed.

Effects of high dose alpha-1-acid glycoprotein on desipramine toxicity in rats

Tricyclic antidepressants are bound extensively to alpha-1-acid glycoprotein (AAG) in serum. It has been suggested that the extent of drug-protein binding may influence the magnitude of cardiotoxicity associated with tricyclic antidepressant overdose. To study this question, desipramine (DMI), 45 mg/kg, was administered i.p. to anesthetized rats, producing an increase in QRS duration of 86.4 +/- 20.5% and a decrease in systolic blood pressure of 28.1 +/- 13.2%. Groups of six rats then received human AAG, 2.2 g/kg i.v. over 30 min, 4.4 g/kg i.v. over 60 min or control infusions of albumin. The serum AAG concentration increased to 29 times normal after AAG, 2.2 g/kg, and 46 times normal after AAG, 4.4 g/kg. The serum DMI concentration did not change with albumin infusion but increased 4.7-fold (1.9 +/- 0.7 to 9.0 +/- 1.6 micrograms/ml) after AAG, 2.2 g/kg, and 6.7-fold (2.3 +/- 0.5 to 15.7 +/- 7.0 micrograms/ml) after AAG, 4.4 g/kg. AAG infusion at either dose had no effect on systolic blood pressure compared to albumin. Animals treated with AAG, 2.2 g/kg, had less QRS prolongation 30 min after AAG infusion than animals treated with albumin (41 +/- 13% vs. 64 +/- 8%, P less than .01) but no difference in QRS duration was observed between groups after the higher dose (4.4 g/kg) of AAG or albumin. Cardiac DMI concentrations 90 min after AAG or albumin treatments did not differ.(ABSTRACT TRUNCATED AT 250 WORDS)