2025 年 50 巻 10 号 p. 547-554
Venlafaxine was the first serotonin/noradrenaline reuptake inhibitor used to treat major depressive disorders. Its overdose can cause cardiovascular toxicity and life-threatening cardiogenic shock. We present the case of a 48-year-old woman who experienced venlafaxine overdose-induced cardiogenic shock. Initial treatment included gastric lavage, blood purification therapy, and ventricular assist device use. The serum venlafaxine concentration was 21.4 μg/mL at 12–24 hr after ingestion, which subsequently decreased to 11.0 and 8.4 μg/mL after 1 and 2 days, respectively. This trend in blood concentration exhibited a biphasic elimination pattern. In addition to venlafaxine-induced cardiotoxicity, the patient exhibited peripheral vascular unresponsiveness to catecholamines. Notably, this vascular dysfunction resolved more rapidly than the cardiotoxic effects. Ultimately, the patient was transferred to a psychiatric ward without sequelae. Although venlafaxine overdose-induced cardiotoxicity has been reported, reports on the unresponsiveness of peripheral blood vessels to catecholamines remain lacking. In cases of venlafaxine overdose-induced cardiogenic shock, both left ventricular function may be impaired and peripheral blood vessels may also be unresponsive to catecholamines. Therefore, rapid initiation of extracorporeal life support and multimodal removal of venlafaxine tailored to the clinical situation may contribute to patient survival.
Venlafaxine was the first serotonin/noradrenaline reuptake inhibitor used to treat major depressive disorders (Kent, 2000). The effectiveness and tolerability of venlafaxine in patients with major depressive disorders have been demonstrated (Cipriani et al., 2016; Kishi et al., 2023). The risk of sudden death or near death with venlafaxine is lower than that with other antidepressants, such as fluoxetine, dosulepin, or citalopram (Martinez et al., 2010). However, cardiotoxicity, neurological toxicity, and gastrointestinal abnormalities are well documented in cases of massive venlafaxine overdose (Blythe and Hackett, 1999). Specifically, venlafaxine overdose-related cardiotoxicity can be fatal (Batista et al., 2013). Schulz et al. reported therapeutic, toxic, and comatose-fatal venlafaxine plasma concentration ranges as 0.1–0.4 mg/L, 1.0–1.5 mg/L, and 3.2–24 mg/L, respectively (Schulz et al., 2020).
We present a case report of venlafaxine overdose-induced severe catecholamine-refractory shock treated with gastric lavage, blood purification therapy, and a ventricular assist device.
Case reportA 48-year-old woman (163.0 cm, 66.8 kg, and body mass index 25.1 kg/m2) was found unconscious in a public restroom and brought to the emergency department with altered consciousness due to medication overdose. The suspected drugs included venlafaxine, chlorpromazine, mirtazapine, nitrazepam, alprazolam, flunitrazepam, lorazepam, etizolam, haloxazolam, fenofibrate, and aspirin. The exact quantity of the ingested substances and the time since ingestion were unknown.
Upon arrival at the emergency department (day 0, 22:52), her vital signs were as follows: respiratory rate, 26 breaths/min; SpO2, 89%; blood pressure, 87/64 mmHg; heart rate, 103 bpm; body temperature, 36.7°C; and warm extremities. The Glasgow Coma Scale score was E1V1M1. Tremors, clonus (inducible, spontaneous, and ocular), or akathisia was not observed, and serotonin syndrome was not suspected. Venous blood gas analysis (while on 10 L/min oxygen via reservoir mask) showed a pH of 7.34, a venous carbon dioxide partial pressure of 52.7 mmHg, HCO3- level of 27.9 mmol/L, blood glucose level of 68 mg/dL, and lactate level of 30 mg/dL. Electrocardiography (ECG) showed sinus tachycardia with a PR interval of 180 ms, QRS duration of 118 ms, and QTc interval of 440 ms.
Flumazenil (0.5 mg) did not improve consciousness level, and the patient underwent endotracheal intubation, gastric lavage, and activated charcoal administration. Ampicillin/sulbactam (9 g/day) was administered intravenously to prevent aspiration pneumonia. Subsequently, extracellular fluid was infused at 500 mL/hr, but a decrease in blood pressure (from 94/62 mmHg to 78/68 mmHg) was observed. Noradrenaline was initiated at 0.05 μg/kg/min and titrated up to 0.25 μg/kg/min. Adrenaline was initiated at 0.1 μg/kg/min and gradually titrated up to 0.2 μg/kg/min, and hydrocortisone (200 mg/day) was administered. Regardless, the patient remained in circulatory failure, with a heart rate of 107 bpm, blood pressure of 55/36 mmHg, oliguria, and an elevated lactate level of 47 mg/dL. Transthoracic echocardiography revealed diffuse wall motion abnormalities in the left ventricle, with an ejection fraction (EF) of 20%. Based on these data, the patient’s condition was consistent with cardiogenic shock (SCAI Shock Classification C or D). The benefits of treating the overdose were judged to outweigh the risks, and a single 100 mL dose of 20% lipid emulsion was administered. Subsequently, an emergency coronary angiography revealed no thrombosis or coronary artery stenosis, and myocarditis was ruled out based on endomyocardial biopsy. A pulmonary artery catheter (PAC) was placed, and an Impella CP® ventricular assist device (Abiomed Inc., Danvers, Massachusetts, USA) was implanted. The systemic vascular resistance index (SVRI) was 1058 dynes·sec/cm5/m2, and laboratory tests revealed that the aspartate aminotransferase, alanine aminotransferase, creatinine, and blood urea nitrogen levels were 824 IU/L, 2586 IU/L, 3.06 mg/dL, and 26.2 mg/dL, respectively, indicating hepatic and renal dysfunction at the time (day 1, 14:34).
After transfer to the intensive care unit (ICU), hemoadsorption (HA) with Hemosorba CHS-350® (porous molded body and bead-like activated carbon; Asahi Kasei Medical Co., Ltd., Nagoya, Japan) was conducted for 3 hr (blood flow: 150 mL/min, day 1, 18:00–21:00). This resulted in a rapid resolution of acidemia, with the pH rising from 7.22 to 7.42. Additionally, HA did not alter the sinus rhythm on ECG. Continuous hemodiafiltration (CHDF) was initiated at 22:00 on day 1. The CHDF settings were as follows: blood flow rate, 100 mL/min; filtration pump flow rate, 1000 or 1100 mL/hr; dialysate pump rate, 300 or 500 mL/hr; and replacement fluid pump flow rate, 500 or 700 mL/hr. Subsequently, the ampicillin/sulbactam dosage was reduced from 9 to 6 g/day. On day 2, from 06:30, her blood pressure increased, and the levels of liver enzymes and creatine kinase also peaked but began to decrease later. As her hemodynamics stabilized, noradrenaline dose was reduced, and she was successfully weaned off the Impella CP® device (day 4, 16:00). Subsequently, the PAC was removed (day 5, 09:00). Furthermore, following the spontaneous urination, CHDF was discontinued, and the EF was 70% at the time of weaning. The patient was discharged from the ICU to the emergency department ward and then to the psychiatric ward without sequelae. The laboratory and blood gas data during the clinical course are shown in Tables 1 and 2, respectively.
Day 0 23:00 |
Day 1 14:34 |
Day 2 06:30 |
Day 3 06:30 |
Day 4 06:00 |
||
---|---|---|---|---|---|---|
White blood cell | 103/uL | 9.7 | 13.4 | 13.3 | 13.2 | 12.1 |
Hemoglobin | g/dL | 12.1 | 11.7 | 10.5 | 8.8 | 9.6 |
Platelet | 103/uL | 298 | 237 | 129 | 79 | 72 |
Total protein | g/dL | 6.5 | 4.5 | 4.8 | 4.6 | 4.7 |
Albumin | g/dL | 3.5 | 2.4 | 2.5 | 2.3 | 2.5 |
C-reactive protein | mg/dL | 0.33 | 3.61 | 8.41 | 9.51 | 5.7 |
Creatine kinase | U/L | 4376 | 14500 | 17246 | 10899 | 3508 |
Aspartate aminotransferase | U/L | 57 | 2586 | 12404 | 5268 | 2354 |
Alanine aminotransferase | U/L | 27 | 824 | 2359 | 2121 | 1798 |
Lactate dehydrogenase | U/L | 384 | 2511 | 8417 | N/D | 2009 |
Creatine | mg/dL | 0.81 | 3.06 | 3.11 | 2.12 | 1.88 |
Blood urea nitrogen | mg/dL | 16.3 | 26.2 | 33.6 | 20.2 | 22.3 |
Total bilirubin | mg/dL | 0.3 | 0.3 | 0.8 | 0.6 | 0.8 |
Estimated glomerular filtration rate | mL/min/1.73 m2 | 45.4 | 13.9 | 13.6 | 20.7 | 23.7 |
N/D: no data
Venous blood | Arterial blood | ||||||||
---|---|---|---|---|---|---|---|---|---|
Day 0 23:00 |
Day 1 06:30 |
Day 1 15:48 |
Day 1 22:55 |
Day 2 05:08 |
Day 3 06:16 |
Day 4 06:31 |
|||
pH | 7.34 | 7.21 | 7.22 | 7.42 | 7.408 | 7.401 | 7.419 | ||
PCO2 | mmHg | 52.7 | 63.3 | 46.9 | 33.7 | 30.1 | 41.2 | 39.6 | |
PO2 | mmHg | 48.8 | 28.8 | 46.2 | 261 | 171 | 134 | 122 | |
HCO3- | mmol/L | 27.9 | 25.1 | 18.4 | 21.5 | 18.6 | 25 | 25.2 | |
Actual base excess | mmol/L | 1.8 | -4.1 | -8.8 | -1.9 | -4.7 | 0.7 | 1.2 | |
Lactate | mg/dL | 30 | 41 | 59 | 15 | 34 | 8 | 8 |
One month after discharge, nine different drugs were detected in the patient’s serum using liquid chromatography–mass spectrometry. The trends in serum concentrations are summarized in Table 3. Based on the compilation of the therapeutic and toxic plasma concentration ranges of more than 1,100 drugs and other xenobiotics (Schulz et al., 2020), venlafaxine was the only drug exhibiting a comatose-fatal blood plasma concentration. Chlorpromazine, mirtazapine, nitrazepam, and alprazolam were in the toxic blood plasma concentration range, whereas the other drugs were within their therapeutic blood plasma concentrations. Therefore, although the contribution of other drugs to the toxicity symptoms could not be entirely excluded, the serum concentration trends and clinical course supported venlafaxine as the cause of catecholamine-refractory cardiogenic shock with peripheral vasoconstriction disorder in this case. The clinical course and changes in venlafaxine blood concentration are summarized in Fig. 1.
Day 1 06:30 |
Day 1 16:00 |
Day 2 06:30 |
Day 2 08:40 |
Day 3 06:30 |
|
---|---|---|---|---|---|
Venlafaxine | 21.4** | 17.0** | 11.0** | 10.2** | 8.4** |
Chlorpromazine | 0.60* | 0.41 | 0.38 | 0.36 | 0.45 |
Mirtazapine | 0.48* | 0.30* | 0.23* | 0.22* | 0.18* |
Nitrazepam | 0.38* | 0.24* | 0.13 | 0.16 | 0.10 |
Alprazolam | 0.11* | 0.049 | 0.018 | 0.025 | 0.003 |
Flunitrazepam | 0.015 | 0.006 | 0.005 | 0.006 | 0.003 |
Lorazepam | 0.11 | 0.070 | 0.050 | 0.051 | 0.037 |
Etizolam | 0.016 | 0.005 | 0.002 | 0.002 | ND |
Caffeine | 1.1 | 0.89 | 0.40 | 0.40 | 0.07 |
ND: not detected
Blood plasma concentration reported by Schulz et al.
**comatose-fatal
*toxic
(A) Venlafaxine concentration changes and evolution of systemic vascular resistance index. Blood samples for venlafaxine concentration measurement were collected from days 1 to 3. Systemic vascular resistance index was monitored using a pulmonary artery catheter from days 1 to 3. Impella CP® was performed from 15:00 on day 1. Hemoadsorption was performed from 18:00 to 21:00 on day 1, followed by alternating continuous hemodiafiltration and hemodiafiltration. The downward triangles indicate the ejection fraction at each time point. (B) Clinical course of changes in blood pressure. The top line shows the systolic blood pressure, and the bottom line shows the diastolic blood pressure. Blood pressure was measured using non-invasive blood pressure from 0:00 to 10:00 on day 1 and arterial blood pressure from 11:00 on day 1 onwards. (C) Changes in noradrenaline, adrenaline, and vasopressin doses during the clinical course. Administration of noradrenaline began at 01:52 on day 1 and ended at 15:39 on day 3. Administration of adrenaline began at 10:50 on day 1 and ended at 20:24 on day 2. Administration of vasopressin began at 15:41 on day 1 and ended at 13:26 on day 2. HA, hemoadsorption; CHDF, continuous hemodiafiltration; HDF, hemodiafiltration; SVRI, systemic vascular resistance index; EF, ejection fraction; NIBP, non-invasive blood pressure; ABP, arterial blood pressure; NA, noradrenaline; AD, adrenaline
Figure 1A illustrates the interventions for detoxification, installation of supportive devices, and changes in the blood serum venlafaxine concentration, SVRI, and EF. The venlafaxine concentration at 06:30 on day 1 (21.4 µg/mL) is presumed to reflect the level 12–24 hr after ingestion. SVRI and blood pressure fluctuated due to Impella CP® and CHDF until around 06:30 on day 2 but were not affected by vasopressor drugs and extracellular fluids (Fig. 1A, B, C). At a venlafaxine concentration of 8.4–11 µg/mL, venlafaxine clearance appeared to decrease, showing a biphasic change in blood concentration. At that time, systolic blood pressure and subsequently started to increase, indicating that the peripheral blood vessels were responding to the catecholamines (Fig. 1A, B, C).
This case report describes a case of venlafaxine overdose-induced cardiogenic shock. The combination of detoxification and multimodal mechanical support, including gastric lavage, HA, HDF, CHDF, and Impella CP®, enabled the patient to survive this life-threatening condition. A notable aspect of this case is that during the venlafaxine overdose-induced cardiogenic shock, high blood concentration of venlafaxine was associated with both left ventricular dysfunction and peripheral vascular unresponsiveness to catecholamines.
However, the mechanisms underlying venlafaxine-related cardiotoxicity have not been elucidated. Hoffmann et al. discussed the mechanisms of venlafaxine toxicity based on several references (Hoffmann et al., 2024). Venlafaxine-induced cardiotoxicity results in left heart failure, which may mimic Takotsubo cardiomyopathy and is life-threatening (Höjer et al., 2008; Neil et al., 2012). Excessive adrenergic stimulation is hypothesized to cause myocardial stunning, which is considered a cause of Takotsubo cardiomyopathy (Christoph et al., 2010; Neil et al., 2012; Vasudev et al., 2016). Alternatively, at high concentrations, venlafaxine may reduce sodium influx currents, suppressing action potentials and causing myocardial dysfunction (Khalifa et al., 1999; Pancrazio et al., 1998). These physiological processes may contribute to venlafaxine-associated cardiotoxicity.
Several treatments may have contributed to venlafaxine elimination from the blood. First, single-dose activated charcoal and gastric lavage, alone or in combination, potentially decreased the area under the plasma concentration–time curve after venlafaxine overdose. Moreover, whole bowel irrigation may be more beneficial because it significantly reduces peak concentrations (Kumar et al., 2009; Kumar et al., 2011). Additionally, in a case of venlafaxine overdose, a pharmacobezoar likely caused delayed absorption and a subsequent increase in venlafaxine levels (Laurent et al., 2021). In our case, the contribution of gastric lavage could not be evaluated from the blood concentration data; however, its absence might have led to a higher or prolonged peak concentration.
After gastric lavage, HA with Hemosorba CHS-350® was conducted for 3 hr. The half-life of venlafaxine at therapeutic doses is approximately 3–7 hr (Klamerus et al., 1992; Schulz et al., 2020). In cases of overdose, the half-life of venlafaxine may be prolonged due to decreased cardiac output and abnormal hepatocellular function (Couderc et al., 2024; Langford et al., 2002). In this case, the half-life was prolonged due to the overdose. Venlafaxine generally has a large distribution volume, making its removal via dialysis difficult. Although the efficacy of Hemosorba CHD-350® for venlafaxine removal has not been demonstrated, its activated carbon may reversibly adsorb small molecule drugs (Schroeder et al., 2017).
CHDF and HDF were implemented as part of blood purification therapy to manage acidemia and renal failure, and effectively improved the acidemia and facilitated a gradual recovery of renal function, with withdrawal on day 5. These blood purification therapies may play a crucial role in sustaining patients with venlafaxine intoxication until spontaneous recovery occurs.
Couderc et al. reported the results of tracking venlafaxine and its major metabolite, O-desmethylvenlafaxine (ODV), and EF recovery (Couderc et al., 2024). The clinical course in this case confirms that the toxic effects of venlafaxine on cardiac function are potent, sustained, and reversible. In this case, catecholamine-resistant peripheral vascular toxicity caused by venlafaxine was also observed. However, during the clinical course, the response to catecholamine was restored, blood pressure increased, and liver enzyme and creatine kinase levels decreased—even when venlafaxine concentrations were within the lethal range. Although the serum venlafaxine concentration decreased to 8.4 µg/mL, the patient still required Impella CP® support. These results suggest that the concentration range at which venlafaxine exhibits toxicity varies across organs, with particularly strong and prolonged toxic effects on the heart.
Chlorpromazine reverses the vasoconstrictive effects of adrenaline (Foster et al., 1954; Higuchi et al., 2014). However, we were unable to assess the extent to which chlorpromazine levels influenced peripheral vasodilation in this case. Although chlorpromazine was initially present at toxic plasma levels, it subsequently reached therapeutic levels (Schulz et al., 2020), suggesting that the clinical symptoms observed in this case were primarily due to venlafaxine. Given the hypothesis that venlafaxine-induced Takotsubo cardiomyopathy results from excessive adrenergic stimulation (Christoph et al., 2010; Neil et al., 2012; Vasudev et al., 2016), we postulate that adrenergic overflow may similarly impair vasoconstrictive responses in the peripheral vascular system. Venlafaxine toxicity may not only impair the heart but also reduce the responsiveness of peripheral blood vessels to catecholamines, potentially inducing and prolonging shock.
A limitation of this case is that the active metabolite, ODV, was not measured. Additionally, CYP2D6 polymorphism, which is involved in venlafaxine metabolism, was not assessed, leaving the impact of ODV in this case unknown. However, ODV has a longer half-life than venlafaxine (Castanares-Zapatero et al., 2016; Couderc et al., 2024), and the possibility that prolonged cardiotoxicity was influenced by ODV cannot be ruled out.
Therapeutic strategies for venlafaxine overdose have not been established. In this case, Impella CP® implantation was decided based on persistent hypoperfusion according to the Japanese guidelines (Nishimura et al., 2024). Furthermore, CHDF was implemented to address renal and hepatic impairments associated with cardiogenic shock. Although only a low-to-moderate level of evidence supports this claim, extracorporeal life support appears effective in improving survival in critically ill patients with cardiac arrest and severe cardiogenic shock (Voicu et al., 2023). In this case, venlafaxine overdose-induced toxicity was managed with Impella CP®, HDF, and CHDF. In addition, venlafaxine overdose can be successfully managed by introducing extracorporeal life support (Couderc et al., 2024; Forsberg et al., 2021; Hoffmann et al., 2024; Magny et al., 2023; Marquetand et al., 2020; Murphy et al., 2021; Napp et al., 2017; Schroeder et al., 2017; Thomas et al., 2017). Notably, such patients recovered without significant sequelae. Therefore, the use of such extracorporeal devices may improve treatment outcomes in patients with venlafaxine overdose-induced cardiogenic shock.
In this case, considering the potential benefits, a single administration of lipid emulsion was applied; however, no clinical improvement was observed. Schroeder et al. reported the successful treatment of venlafaxine overdose using a combination of extracorporeal life support and CytoSorb®, along with lipid emulsion (Schroeder et al., 2017). Lipid emulsions effectively relieve local anesthetic toxicity; however, evidence for their use in neuropsychiatric drug toxicity remains insufficient (Hwang and Sohn, 2024). A previous case report has demonstrated lipid emulsion administration for severe venlafaxine overdose (de Wit et al., 2022), but this alone does not establish its efficacy. Additionally, the risks associated with lipid emulsion therapy—including agglutination, which may lead to mechanical complications, as well as lipid deposition in extracorporeal life support systems, fat embolism, and fat overload syndrome—should be considered. Further analyses of numerous case reports and additional large-scale studies are required.
In summary, this case highlights the potential to save lives in cases of severe cardiogenic shock induced by a massive dose of venlafaxine through traditional detoxification and multimodal mechanical support. In cases of venlafaxine overdose-induced cardiogenic shock, peripheral vascular unresponsiveness to catecholamines may be observed, making it crucial to promptly initiate circulatory support and provide additional robust support based on the patient’s condition. Furthermore, meticulous management to prevent complications can improve the treatment outcomes.
We acknowledge the assistance of the Research Equipment Sharing Center at Nagoya City University.
Conflict of interestThe authors declare that there is no conflict of interest.