Liver transplantation is increasingly used in patients with end-stage liver disease or acute liver failure, leading to an increased demand for donor livers. However, there is currently a shortage in donor livers, and some patients die while waiting for a liver transplant. A possible source of donor organs is the livers of patients with brain death due to drug overdose. In deciding whether the liver of an intoxicated patient can be used for transplantation, one has to consider (i) that the donor liver might carry over a relevant amount of ingested drug, causing possibly dangerous adverse effects in the recipient and (ii) that the ingested drug might be hepatotoxic, potentially damaging the donor liver. For example, Brems et al.(1) report the case of a patient who experienced severe cardiac arrhythmia, deteriorating mental function, and severe hepatic dysfunction after receiving a liver from a donor who had ingested a large amount of amitriptyline. The recipient recovered rapidly after retransplantation, performed some days later. The authors of this case report therefore suggested that the cardiac arrhythmia and hepatic dysfunction were caused by amitriptyline contained in the donor liver resulting in a tricyclic antidepressant overdose of the recipient. We report the case of a patient with a fatal trimipramine overdose who was evaluated as an organ donor.
A 38-year-old woman who had been treated with the tricyclic antidepressant trimipramine for several years had a cardiac arrest after ingesting a trimipramine overdose of approximately 8.4 g. After cardiopulmonary resuscitation, the patient required a temporary pacemaker, after which her cardiopulmonary situation stabilized. However, the patient stayed in coma with large, nonreacting pupils, missing brainstem reflexes, and no response to pain. All peripheral organs were undamaged and functioned normally. Her relatives agreed that the patient's organs might be used for transplantation and the patient was therefore referred to the University Hospital of Zurich. The diagnosis of brain death could not be established since the serum trimipramine concentration was elevated and some cerebral perfusion was still present by angiography. On the following day, signs of elevated intracranial pressure and cardiovascular instability developed and the patient died of cardiopulmonary arrest. Both kidneys were immediately removed and were successfully transplanted. Liver tissue was obtained at the time of laparotomy and immediately frozen for later analysis of the trimipramine tissue content.
During the first days after ingestion of the trimipramine overdose, the serum aspartate aminotransferase and alanine transferase activities were elevated approximately 10 times the upper normal limit. They gradually decreased during the following days and had almost normalized on the day before death. The serum γ-glutamyl transpeptidase, alkaline phosphatase, bilirubin, and the prothrombin time were all in the normal range. At autopsy the liver weighted 1400 g and showed no macroscopic pathology, with the exception of some congestion. Microscopically hepatic venous congestion was confirmed, but otherwise normal liver tissue was found. Notably, only mild macrovesicular and no microvesicular steatosis was found.
The concentrations of trimipramine, 2-hydroxy-trimipramine, and N-desmethyl-trimipramine in serum and liver tissue were measured by high-performance liquid chromatography (2, 3). The trimipramine and metabolite serum concentrations measured at 4, 5, 6, 9, and 10 days after drug ingestion are shown in Figure 1. The serum trimipramine concentration fell about linearly on a semilogarithmic scale from day 4 to day 10 after drug ingestion with a half-life of approximately 39 hr. This estimate lies in the upper normal range for trimipramine (4), which is metabolized by cytochrome P450 isoenzymes and exhibits a large interindividual variability of clearance and half-life (4, 5). Based on the half-life estimate of 39 hr, the serum concentration of trimipramine on the day of drug ingestion can be estimated to have been between 6000 and 7000 μg/L. On the last day, serum trimipramine concentration had fallen to concentrations generally measured in patients receiving therapeutic trimipramine dosages(5). In the first serum sample studied, the metabolite concentrations were about one fourth and one seventh of the trimipramine concentration. The peak concentration of the 2-hydroxy metabolite had probably occurred between the sixth and the ninth day after drug ingestion. The serum concentration of the N-desmethyl metabolite had decreased from the fourth to tenth day with a somewhat slower rate than trimipramine and had fallen to values below the detection limit (50 μg/L) on day 9. Except for the last day, when the drug concentration had already fallen to therapeutic values, metabolite concentrations were always considerably lower than drug concentrations. Only limited information is available about the therapeutic and toxic activity of these metabolites. Musa suggests that for desmethyl-trimipramine the activity is the same as for trimipramine(5). If we assume a similar activity for the metabolite as for trimipramine, the low metabolite concentrations measured suggest a rather limited clinical relevance.
The concentrations of trimipramine, 2-hydroxy-trimipramine, and N-desmethyl-trimipramine in the liver tissue sample taken on the tenth day after ingestion of the trimipramine overdose were 1750 μg/kg, 850 μg/kg, and 225 μg/kg, respectively. Thus, at that time, the hepatic trimipramine concentration was 22 times and the hepatic 2-hydroxy-trimipramine concentration was 17 times higher than the respective serum concentration. If one had decided to make the liver available for transplantation at that time, one would have carried over a dose of 2.5 mg of trimipramine, 1.2 mg of 2-hydroxy-trimipramine, and 0.3 mg of N-desmethyl-trimipramine to the recipient. This is a small and probably negligible amount of trimipramine, if we compare it with the standard initial therapeutic dose of 50-100 mg/day and the maximal dose of 400 mg/day. Based on these results and considering the absence of major histologic changes in the liver, we think that in this case the liver could have been used for transplantation. Unfortunately, we cannot prove this suggestion, since the patient died before the diagnosis of brain death was established and therefore the liver was not made available as a transplant organ. However, our hypothesis is supported by the fact that both kidneys were transplanted successfully without any adverse effects of the tricyclic antidepressant carried over to the recipients, despite the much higher tissue fixation ratio of antidepressants in kidney compared with liver tissue (6).
Considering the large volume of distribution of trimipramine (30.9 L/kg)(7), it is striking that the amount of trimipramine found in the liver was so small. The total body content of trimipramine at a serum concentration of 80 μg/L should be about 170 mg, but only 2.5 mg were found in the liver. A possible explanation for this discrepancy is that trimipramine accumulates mainly in tissues other than the liver. In a study with radioactive labeled imipramine, Bickel et al. (8) showed that in rats the drug and its metabolites are present in the first hour after intravenous administration mainly in the lung, the kidneys, the brain, and the small intestine. Hilberg et al. (6) studied tissue concentrations in rats killed 2 hr after subcutaneous administration of amitriptyline, a structurally closely related tricyclic antidepressant. The ratio of postmortem tissue to antemortem femoral blood concentration was 56 in the lung, 33 in the kidney, 22 in the brain, 16 in the heart, and 7 in the thigh muscle. In the liver they found only a 2.5 times higher concentration than in the femoral blood. Postmortem release of amitriptyline and possibly other tricyclic antidepressants from the lung can result in a high variability of the measured blood concentrations, depending on the site of blood sampling and the time elapsed between death and sampling (6). In these experiments, it was shown that blood concentrations changed considerably during the first 2 hr after death. In contrast, liver tissue concentrations were found to be unchanged during the first 2 hr after death; thereafter, an increase in tissue concentrations over the next 4 days was observed. Thus, concentrations of tricyclic antidepressants in serum obtained at any time point after death and in liver tissue obtained more than 2 hr after death should be interpreted with caution.
The literature and the results of our case report clearly demonstrate that the liver is not the main tissue where tricyclic antidepressants accumulate and that at therapeutic serum trimipramine concentrations the amount of trimipramine present in the liver is negligible and does not put a possible liver transplant recipient at risk of antidepressant-induced toxicity. Because in our case it was possible to measure liver tissue concentrations at one time point only, it is difficult to judge whether in case of confirmed brain death the liver could have been used for transplantation at an earlier time point. Assuming that the ratio of the liver to serum concentration of trimipramine is independent of its serum concentration, i.e., that the half-life of trimipramine in the liver is the same as in the serum, we can extrapolate that, at a serum concentration of 2000 μg/L, the total amount of trimipramine and its metabolites in the liver would have been about 100 mg, i.e., in the range of a standard initial therapeutic dose. In our case, similar serum concentrations were probably reached on the third day after the overdose. If we extrapolate back to the day of drug overdose, the total amount of drug and metabolites in the liver would have been in the range of 346 mg. This represents a relevant dose of trimipramine, but still lies in the upper range of therapeutic daily dosages. Based on these extrapolations, we would suggest that at a serum trimipramine concentration of 2000 μg/L or less, the amount of trimipramine carried over with the donor liver does not put the recipient at risk for adverse drug effects.
As mentioned above, the second major concern in obtaining a transplant liver from a person dying from any drug overdose is the possible hepatotoxicity of the drug ingested. It is well known that therapeutic doses of tricyclic antidepressants can cause hepatic injury(9). Severe liver damage due to tricyclic antidepressants is rare, however, and the hepatotoxicity of tricyclic antidepressants seems not to be dose related. Correspondingly, there are very few reports of hepatic injury due to an overdose of antidepressants. In most cases the drug ingested was amineptin, a tricyclic antidepressant with a heptanoic side chain that inhibits hepatic mitochondrial β-oxidation and can cause microvascular steatosis of the liver (10).
In contrast to our findings, Brems et al. (1) suggest in their case report that the amitriptyline carried over to the recipient was responsible for the arrhythmia and the liver failure that occurred in a patient after transplantation, making retransplantation necessary. Unfortunately, the only drug and metabolite concentrations reported in this case report are amitriptyline and metabolite concentrations in the first liver transplant after retransplantation. The amounts of amitriptyline and nortriptyline present at that time were 0.3 and 2 mg, respectively. Brems et al. suggest, based on a normal half-life of 12 hr for amitriptyline, that the patient probably had very high tissue concentrations at the time of transplantation 6 days earlier. However, in a patient with liver failure, it seems difficult to justify a half-life in the lower range of healthy volunteers for a drug mainly eliminated by hepatic metabolism. If we assume a longer half life, e.g., 24 hr, which corresponds to a half-life in the higher normal range for amitriptyline (4), extrapolation results in an amount of drug and active metabolites in the liver in the upper therapeutic dose range, which would not be sufficient to cause the adverse events reported. Also, since amitriptyline has a strong extrahepatic tissue fixation, one would expect that the postulated amitriptyline overdose carried over to the recipient would have redistributed after liver transplantation to other tissues of the recipient. Correspondingly, one would expect that removing the transplant liver would have removed only a minor part of the amitriptyline dose carried over; therefore, the adverse effects would have persisted for some time after removal of the transplant. However, compared with our case, the transplant was harvested very early after drug overdose(only 24 h after hospital admission) and another tricyclic drug was ingested, which might possibly accumulate to a higher extent in liver tissue at this early time point.
In conclusion, our case report shows that the amounts of trimipramine and its metabolites in the liver are rather small at trimipramine serum concentrations in the therapeutic range. This implies that harvesting a liver from a donor who was on long-term trimipramine treatment and had no signs of liver disease is safe for the recipient. Extrapolations from the data presented suggest that a liver transplant obtained from a donor who died from a trimipramine overdose is safe, if the serum concentration in the donor has fallen below 2000 μg/L before the transplant is harvested. At trimipramine concentrations above 2000 μg/L, we recommend delaying organ harvesting until the serum concentration has fallen below this value. If waiting is not an option, the decision to use the donor liver must be made after weighing the risks of the trimipramine dose carried over to the recipient. This dose may be calculated by multiplying the liver mass (as estimated from an abdominal CT scan) by the liver trimipramine concentration including metabolites (as estimated by multiplying the trimipramine serum concentration by a factor of 35). Because animal data suggest that tricyclic antidepressants accumulate considerably more in lung and possibly also in heart, we suggest that tissue concentrations should be determined in these organs in patients dying from overdoses of tricyclic antidepressants before they are considered for transplantation.
1. Brems JJ, Merenda GO, Hayek ME, Kane RE, Flynn MF, Kaminski DL. Orthotopic liver transplantation resulting in amitriptyline toxicity in the recipient. Transplantation 1989; 48: 159.
2. Fraser AD, Isner AF, Perry RA. Distribution of trimipramine and its major metabolites in a fatal overdose case. J Anal Toxicol 1987; 11: 168.
3. Binder SR, Regalia M, Biaggi-McEachern M, Mazuar M. Automated liquid chromatographic analysis of drugs in urine by on-line sample cleanup and isocratic multi-column separation. J Chromatogr 1989; 473: 325.
4. Schulz P, Turner TK, Smith G, Giacomini KM, Blaschke TF. Amitriptyline disposition in young and elderly normal men. Clin Pharmacol Ther 1983; 33: 360.
5. Musa MN. Nonlinear kinetics of trimipramine in depressed patients. J Clin Pharmacol 1989; 29: 746.
6. Hilberg T, Morland J, Bjorneboe A. Postmortem release of amitriptyline from the lungs: a mechanism of postmortem drug redistribution. Forensic Sci Int 1994; 64: 47.
7. Abernethy DR, Greenblatt DJ, Shader RI. Trimipramine kinetics and absolute bioavailability: use of gas-liquid chromatography with nitrogen-phosphorus detection. Clin Pharmacol Ther 1984; 35: 348.
8. Bickel MH, Graber BE, Moor M. Distribution of chlorpromazine and imipramine in adipose and other tissues of rats. Life Sci 1983; 33: 2025.
9. Stricker BHC. Drug-induced hepatic injury, 2nd ed. Amsterdam: Elsevier, 1992.
10. Fromenty B, Pessayre D. Inhibition of mitochondrial beta-oxidation as a mechanism of hepatotoxicity. Pharmacol Ther 1995; 67: 101.