A Case of High-Dose Intravenous MgSO₄ and Hemoperfusion for Aconite Poisoning with Chronic Kidney Disease

ABSTRACT

Aconite contains four highly toxic diester-diterpene alkaloids, including aconitine, mesaconitine, hypaconitine, and jesaconitine, in all plant parts. Aconite has been used as for suicide, murder, and as an arrow poison since ancient Greek and Roman times. Ventricular tachyarrhythmias are the most common cause of death in aconite poisoning, and antiarrhythmic drugs and cardioversion are ineffective. A 61-year-old woman ingested the crushed raw roots of a single aconite plant. An ambulance brought her to the Tottori University Hospital 30 min after ingestion. She had a history of chronic stage 5 kidney disease but was not on dialysis. Her heart rate (HR) was 120 bpm upon arrival. The patient developed sustained supraventricular tachycardia (SVT) at an HR of 165 bpm with frequent premature ventricular contractions (PVCs) 15 min after arrival. She then developed sustained monomorphic ventricular tachycardia (VT) at an HR of 200 bpm 20 min after arrival, which progressed to pulseless polymorphic VT. Cardioversion was unsuccessful. External cardiac massage restored spontaneous circulation; however, her underlying rhythm remained sustained SVT with frequent PVCs. These arrhythmias repeatedly led to circulatory arrest. She was administered six intravenous boluses of 2 g of MgSO₄ in the emergency department, which prevented her from going into sustained pulseless VT. Hemoperfusion (HP) with activated charcoal was performed 1.5 h after arrival. The aconitine, mesaconitine, and hypaconitine plasma concentrations were high at 8.9, 23.5, and 5.5 ng/mL, respectively, before the start of HP but decreased to 1.7, 4.0, and 2.7 ng/mL, respectively, after 7 h of HP. She returned to sinus rhythm on the second day of hospitalization; however, the patient required maintenance hemodialysis. We concluded that high-dose IV MgSO₄ is an effective treatment for fatal tachyarrhythmias due to aconite poisoning, and that in cases of renal failure, HP may be required to remove aconite toxins from the body.

Aconite contains four highly toxic diester-diterpene alkaloids (DDAs), including aconitine, mesaconitine, hypaconitine, and jesaconitine, in all plant parts.^{1, 2} Aconite has been used as a suicide, murder, and arrow poison since ancient Greek and Roman times.^2,3,4 In East Asia, dried aconite roots have been used in herbal medicine after its toxicity has been reduced by boiling or processing,⁵ which hydrolyzes the DDAs in aconite roots to less toxic monoester diterpene alkaloids.³ Aconite poisoning is common in China, Taiwan, India, and other countries owing to the ingestion of improperly processed herbal medicines. Over 100 species of aconite are native to the temperate and subarctic zones of the Northern Hemisphere,⁶ and occasional cases of aconite poisoning in Europe and North America have been reported because of the ingestion of aconite mistaken for an edible wild plant.⁴ VF and sustained PVT were observed in 15 out of 17 patients with aconite poisoning.⁷ Ventricular tachyarrhythmias (VTAs) are the most common cause of death in aconite poisoning, and antiarrhythmic drugs and cardioversion are ineffective; the mortality for this poisoning is 15%.⁸

A woman with chronic stage 5 kidney disease (CKD) ingested raw aconite root, resulting in an electrical storm (ES). Although she was at risk for hypermagnesemia, she was injected with an intravenous (IV) bolus of 2 g of MgSO₄ at each onset of pulseless ventricular tachycardia (VT) to prevent it from becoming sustained, ultimately saving her life. Charcoal hemoperfusion (HP) was performed to remove the four DDAs. We concluded that high-dose IV MgSO₄ is an effective treatment for fatal tachyarrhythmias due to aconite poisoning, and that in cases of renal failure, HP may be required to remove aconite toxins from the body.

PATIENT REPORT

Clinical history

The patient was a 61-year-old woman, 155 cm tall and weighing 60 kg. The patient was diagnosed with CKD at the age of 54 but had not yet undergone dialysis. She also had a history of alcoholism. She had purchased and grown a single ornamental aconite in advance to use for suicide. The patient crushed the raw aconite root in a blender, mixed it with milk, and ingested it. Her husband found her vomiting and in a disturbed state. She confessed to her husband that she had ingested aconite. An ambulance brought her to the Tottori University Hospital 30 min after ingesting the aconite roots.

On arrival, she had vomiting, diarrhea, and cold sweats. Her consciousness level was a Glasgow Coma Scale score of 11 (E2V4M5), her systolic blood pressure (BP) was 70 mmHg, and her heart rate (HR) was 120 bpm with an irregular rhythm. She received 5 mg IV diazepam for sedation. The patient suddenly developed narrow and wide QRS tachyarrhythmia with an HR of 165 bpm 15 min after arrival (Fig. 1a). A 12-lead electrocardiogram (ECG, Fig. 2) showed that her underlying rhythm was sustained supraventricular tachycardia (SVT) with frequent premature ventricular contractions (PVCs) and aberrant ventricular conduction. Adenosine triphosphate (20 mg), lidocaine (100 mg), aprindine (100 mg), and disopyramide (100 mg) were administered intravenously in that order but were ineffective.

Fig. 1.

 Lead II rhythm strips recorded in the ED. 15 min after arrival, the patient sustained SVT at an HR of 165 bpm with frequent PVCs (a). 20 min after arrival, the patient progressed from non-sustained VT (b) to sustained MVT at an HR of 200 bpm (c, d), and cardioversion was unsuccessful, resulting in pulseless PVT (e) and VF (f). The patient returned to sustained SVT at an HR of 157 bpm with ventricular bigeminy during ECM (g). At one point, the PVCs disappeared for approximately 1 min, yet SVT persisted. In her SVT at that time, the QT interval was prolonged to occupy most of the RR interval, and the SVT was also accompanied by alternating LAH (h). 40 min after arrival, the patient’s PVC frequency increased, and she again developed sustained pulseless PVT at an HR of 164 bpm (i).

Fig. 2.

 12-lead ECG recorded in the ED. The patient’s underlying rhythm was sustained SVT, accompanied by frequent PVCs and aberrant ventricular conduction. The characteristic twisting QRS configuration of TdP was absent.

Her HR and PVC frequency gradually increased (Fig. 1b), and she developed sustained monomorphic ventricular tachycardia (MVT) with an HR of 200 bpm 20 min after arrival (Figs. 1c and d) and respiratory arrest. Cardioversion was attempted but failed, and she developed pulseless polymorphic ventricular tachycardia (PVT, Fig. 1e) and ventricular fibrillation (VF, Fig. 1f). The patient was intubated and ventilated; after intubation, cardioversion was attempted again but was also unsuccessful. External cardiac massage (ECM) was immediately initiated. One minute after starting ECM, she returned to pulseless MVT with an HR of 170 bpm. Seven minutes after starting ECM, she returned to sustained SVT at an HR of 157 bpm with ventricular bigeminy (Fig. 1g). Her spontaneous circulation returned, and ECM was discontinued.

The PVCs then disappeared for approximately 1 min; however, her underlying rhythm was sustained SVT at an HR of 200 bpm. Figure 1h shows a lead II rhythm strip of the SVT with transient disappearance of PVCs, indicating that the QT interval was prolonged to occupy most of the RR interval, and SVT was accompanied by alternating left anterior hemiblock (LAH). Her underlying rhythm reverted to sustained SVT with frequent PVCs, and her HR returned to approximately 160 bpm.

The PVC frequency of the patient increased, and she again developed PVT at an HR of 164 bpm 40 min after arrival (Fig. 1i). Cardioversion was attempted but was still unsuccessful, and pulseless PVT progressed to VF. ECM was continued for 5 min, and she returned to sustained SVT at an HR of 150 bpm with frequent PVCs. Her spontaneous circulation returned again.

An IV bolus of 2 g of MgSO₄ was injected before the recurrent pulseless VT became sustained because various antiarrhythmic drugs were ineffective in sustained SVT with frequent PVCs, and cardioversion was unsuccessful in sustained VT. The IV bolus of 2 g of MgSO₄ terminated pulseless VT; however, her underlying rhythm remained sustained SVT with frequent PVCs. She was then administered 2 g of MgSO₄ intravenously over 5 to 30 min each time pulseless VT was likely to be sustained. Initially, pulseless VTs occurred approximately every 10 min; however, their frequency gradually decreased. One milligram IV epinephrine was administered five times during the previous several cardioversions and ECMs.

Although the 12-lead ECG (Fig. 2) did not show the characteristic twisting QRS configuration of torsades de pointes (TdP), we postulated that her pulseless VT was caused by early afterdepolarization (EAD) due to prolonged action potential duration (APD), by the same mechanism as that of TdP. Serum K⁺ and creatinine phosphokinase were normal at 4.5 mmol/L and 45 U/L, respectively. The patient was diagnosed with fatal tachyarrhythmias due to aconite poisoning. Blood urea nitrogen and serum creatinine levels were high at 69 and 4.2 mg/dL, respectively. The estimated glomerular filtration rate was low at 9.2 mL/min/1.73 m², resulting in a diagnosis of stage 5 CKD.

The patient’s underlying rhythm remained sustained SVT with frequent PVCs, and HP with activated charcoal was performed to remove the four DDAs because she had stage 5 CKD and circulatory instability. An 11.5 French dual lumen catheter (Quinton–Mahurkar catheter, Quinton Instrument Company, Seattle, WA) was inserted into her right femoral vein. HP was initiated 1.5 h after arrival, which was 2 h after ingestion. DHP-1 (Kuraray, Osaka, Japan) was used for the activated charcoal column. Immediately after introducing HP, her BP decreased substantially, and ECM was performed temporarily; however, her BP recovered when the circuit flow rate was temporarily slowed. HP was performed for 3 h in the emergency department (ED) at a circuit flow rate of 100 mL/min. After initiating HP, a gastric tube was inserted, followed by gastric lavage with 10 L of tepid water. Activated charcoal (100 g) and 200 mL of 13.6% Mg citrate solution (89.3 mmol of Mg) were then injected into her stomach.

The patient was transferred to the intensive care unit (ICU) 5 h after arrival because her HR gradually decreased to approximately 140 bpm, and pulseless VT tended to terminate voluntarily. After a 30-min pause, HP was performed for 4 h in the ICU using a second column. As a result, HP was performed from 2 h to 9.5 h after ingestion, and the platelet count decreased from 372,000 to 111,000/μL before and after 7 h of HP.

Eight hours after ingestion during HP, the 12-lead ECG showed that her underlying rhythm was atrial tachycardia (AT), and her PVCs had disappeared (Fig. 3). This AT had 1:1 atrioventricular (AV) conduction, an atrial rate of 140 bpm and was accompanied by a right bundle branch block (BBB). The AV conduction time was prolonged. The QT interval was extended to occupy most of the RR interval. In addition, the shape of the right BBB changed cyclically with each heartbeat.

Fig. 3.

 12-lead ECG recorded during HP in the ICU 8 h after aconite ingestion. The patient’s underlying rhythm was AT, and her PVCs disappeared. This AT had a 1:1 AV conduction, an atrial rate of 140 bpm, and was accompanied by a right BBB. The ventricular rhythm was regular, and some P waves were visible in the V1–V3 leads (arrows). The AV conduction time was prolonged. The QT interval was extended to occupy most of the RR interval. In addition, the shape of the right BBB changed cyclically with each heartbeat.

Fifteen hours after ingestion, her underlying rhythm reverted from AT to sinus rhythm (SR) with an HR of 98 bpm. Occasional AT was present until the first half of the second hospital day, even after the return to SR. Twenty-three hours after ingestion, the Bazett-corrected QT interval was 520 ms, markedly above the upper normal limit. The QTc interval normalized on the third hospital day. She received continuous IV infusions (CIVIs) of dopamine and norepinephrine to maintain her hemodynamics until the second hospital day.

Figure 4 shows the change in serum Mg level after arrival. The serum Mg level was not measured before the first IV bolus of 2 g of MgSO₄. However, the serum Mg level was 1.4 mmol/L after the first dose of 2 g of MgSO₄. The maximum serum Mg level was 4.8 mmol/L at 3.5 h after arrival. By this time, we had administered 10 or 12 g of MgSO₄ intravenously. During a 5-h stay in the ED, six IV boluses of 2 g of MgSO₄ were injected to terminate sustained pulseless VT. Two additional 2g IV boluses of MgSO₄ were then injected within 6 h of transfer to the ICU. The patient finally returned to SR 15 h after ingestion. The serum Mg level at this time was 3.8–4.2 mmol/L (Fig. 4). The serum Mg level did not normalize in 24 h and took a week to decrease to approximately 1.8 mmol/L and remained above the upper normal limit after that (Fig. 4).

Fig. 4.

 Time course of serum Mg levels. No serum Mg level was measured before the first IV bolus of 2 g of MgSO₄. The serum Mg level was 1.4 mmol/L after the first dose of 2 g of MgSO₄. The maximum serum Mg level was 4.8 mmol/L at 3.5 h after arrival. By this time, 10 or 12 g of MgSO₄ had been administered intravenously. MgSO₄ was administered intravenously only on the first hospitalization day, and the total dose was 16 g. The serum Mg level did not normalize within 24 h; instead, it took a week to decrease to approximately 1.8 mmol/L and remained above the upper normal limit after that.

Her endotracheal tube was removed on the fourth hospital day. Her serum Mg level was 3.1 mmol/L at that time. However, the patient was reintubated on the fifth hospital day because of atelectasis caused by sputum retention. Shortly after that, a tracheotomy was performed, and the ventilator was removed. The patient’s life was saved; however, because her renal function deteriorated, she received hemodialysis (HD) for the first time on the 23rd hospital day and then maintenance HD after that.

Plasma concentration measurements of four DDAs

The patient’s plasma samples were collected over time, including during the HP session, and her gastric content was collected before gastric lavage in the ED and stored at -80°C. The concentrations of the four DDAs—aconitine, mesaconitine, hypaconitine, and jesaconitine—were measured using gas chromatography-selected ion monitoring.⁹ Three DDAs were detected—aconitine, mesaconitine, and hypaconitine—in her plasma samples and gastric content; however, jesaconitine was below the lower quantification limit in all samples. Figure 5 shows the changes in each plasma concentration of the three DDAs over time. The first sample was collected 1.5 h after aconite ingestion, i.e., 30 min before the start of the HP session. The total concentration of the three DDAs in the first sample was 37.9 ng/mL (aconitine, 8.9 ng/mL; mesaconitine, 23.5 ng/mL; and hypaconitine, 5.5 ng/mL). The three DDA concentrations of the first sample were the highest during the entire measurement period. The concentrations of the three DDAs decreased gradually but were still detectable on the sixth hospital day. Therefore, the maximum total concentration of the three DDAs was 37.9 ng/mL (maximum concentration, C_max), and the time to reach the maximum total concentration of the three DDAs was 1.5 h (time to maximum concentration, T_max). The total concentration of the three DDAs 9.5 h after aconite ingestion, or 7 h after the start of HP, was 8.4 ng/mL (aconitine, 1.7 ng/mL; mesaconitine, 4.0 ng/mL; and hypaconitine, 2.7 ng/mL). The total concentration of the three DDAs 20 h after ingestion was 5.5 ng/mL (aconitine, 2.0 ng/mL; mesaconitine, 2.6 ng/mL; and hypaconitine, 0.9 ng/mL). The patient returned to SR 15 h after ingestion, and the total concentration of the three DDAs at this time should be between 5.5 and 8.4 ng/mL, which is the concentration 9.5 to 20 h after ingestion (Fig. 5). The QTc interval normalized on the third hospital day when the total concentration of the three DDAs was 5.1 ng/mL. The DDA concentrations in the gastric content initially collected in the ED were 700, 2,180, and 370 ng/mL for aconitine, mesaconitine, and hypaconitine, respectively.

Fig. 5.

 Time course of plasma concentrations of the three DDAs. On arrival, i.e., before HP, aconitine, mesaconitine, and hypaconitine concentrations were high at 8.9, 23.5, and 5.5 ng/mL, respectively. The three DDAs decreased rapidly less than 24 h after ingestion, followed by a gradual decline to below the quantitation limit by the seventh hospitalization day. HP was performed from 2 to 9.5 h after ingestion.

DISCUSSION

The content of each of the four DDAs in aconite varies depending on the part, species, collection season, and growing region.^{2, 5, 10} All parts of aconite are toxic, with the roots being the most harmful.^{1, 2} For subcutaneous injection in mice, the lethal doses 50 of mesaconitine, jesaconitine, and hypaconitine are 0.5-, 0.5-, and 3.5-fold higher than those of aconitine, respectively.¹¹ In guinea pigs, mesaconitine is more arrhythmogenic than aconitine at the same dose.¹³ The lethal oral dose in humans is 1 g for aconite and 2 mg for pure aconitine.¹² For Aconitum japonicum, each gram of root contains 2–4 mg of the four DDAs.¹⁰ Given the amount and means of ingestion, the present case should have been fatal.

Aconitine inhibits Na channel inactivation, prolongs APD, causes EAD, and leads to ventricular arrhythmias.¹³ APD corresponds to the QT interval on the surface ECG. Long QT syndrome (LQTS)-3 is a congenital abnormality of the Na channel that produces EAD-induced triggered activity and causes TdP.¹⁴ Incidentally, TdP is not a single arrhythmia, but a syndrome of PVT characterized and caused by repolarization delay, repolarization dispersion and EAD.¹⁵ Although the present case did not show the characteristic twisting QRS configuration of TdP, her pulseless VT possibly developed via the same mechanism as that of TdP. In the present case, the QTc interval was still high at 520 ms 8 h after return to SR. Serum Mg levels were also high at approximately 3.7 mmol/L. IV MgSO₄ has a therapeutic effect on typical TdP without shortening the QT interval.^{16, 17} Class Ia and III antiarrhythmic drugs are usual causes of acquired LQTS.¹⁸ In the present case, disopyramide administered during the initial treatment may have enhanced aconite-induced arrhythmogenic effects.

In the present case, SVT was present from the first visit and sustained until the patient returned to SR. Even when her VT was sustained, her atria were probably also contracting frequently. Furthermore, SVT was accompanied by aberrant ventricular conduction, such as LAH and a right BBB. The sustained, regular, wide QRS tachycardia triggering circulatory arrest was diagnosed as sustained MVT, which may also have been sustained SVT with aberrant ventricular conduction. When the underlying rhythm changed to AT, AT was accompanied by right BBB, and the right BBB cyclically changed shape with each heartbeat. The cyclic changes in the right BBB were possibly caused by concomitant VT originating from the left posterior bundle branch, which occurred at a rate similar to atrial excitation and fused with the underlying rhythm, AT. In other words, VPC was thought to be gone, but VT was still occurring at this point.

In addition to APD prolongation and EAD development, aconitine also induces intracellular Na⁺ accumulation, leading to intracellular Ca²⁺ overload via the Na/Ca exchanger, resulting in delayed afterdepolarization, another cause of aconitine-induced arrhythmias.¹⁹ Aconitine induces arrhythmias by acting on Ca channels, intracellular Ca²⁺ signaling, and even K channels.³ However, even with these fundamental studies, it is still unclear which of the existing antiarrhythmic drugs and which dosing regimen is effective against aconite-induced fatal arrhythmias.

One to two grams of IV MgSO₄ is the first-line therapy for typical TdP caused by acquired LQTS, and IV MgSO₄ should be adjusted to maintain a serum Mg level of 2.0 mmol/L as a reference value.¹⁷ An IV bolus of 2 g of MgSO₄ over 1–3 min is recommended for typical TdP.^{16, 20} Furthermore, even if the first 2 g IV bolus of MgSO₄ is ineffective, a second or third bolus is recommended at 5–15 min intervals, followed by a CIVI of 2–20 mg/min MgSO₄. In addition to typical TdP, preeclampsia is another indication for IV MgSO₄. IV MgSO₄ has been the first-line therapy for preeclampsia since the 1950s in the United States, and its safety is well established. The therapeutic range for serum Mg levels in preeclampsia is 2–4 mmol/L.^{21, 22}

Sawanobori et al.^{23, 24} examined the effect of extracellular Mg²⁺ concentration ([Mg²⁺]o) on aconitine-induced VTAs in rabbit Langendorff heart and Purkinje muscle preparations. EAD, triggered activity, and automaticity disappeared, and PVT returned to SR after 30 min when they maintained [Mg²⁺]o as high as 5 mmol/L. However, PVT improved to MVT after 15 min but did not return to SR when they maintained [Mg²⁺]o at 3.0 mmol/L. IV MgSO₄ returns aconite-induced PVT to SR; however, serum Mg levels may need to be maintained as high as 5 mmol/L to return to SR in 30 min. However, IV MgSO₄ could improve aconite-induced PVT to some extent in 15 min if the serum Mg level is maintained at 3 mmol/L.

Notably, Coulson et al.⁸ determined that IV MgSO₄ restored SR in only two out of nine patients. However, the effectiveness of IV MgSO₄ should be evaluated in terms of whether the MgSO₄ dose was sufficient. Future studies should first assess whether the IV MgSO₄ dose was sufficient in individual cases when determining the effectiveness of IV MgSO₄ for aconite-induced tachyarrhythmias. In the present case, an IV bolus of 2 g of MgSO₄ rapidly terminated the VT every time pulseless VT developed and prevented progression to VF. An IV bolus of 2 g of MgSO₄ returns a typical TdP to SR in 1–5 min but may not return aconite-induced VTAs to SR in such a short time. However, it may rapidly terminate aconitine-induced sustained pulseless VT.

Adverse effects such as moderate decreases in respiratory rate and depth^{25, 26} and dyspnea²⁷ occur when serum Mg levels exceed 6 mmol/L with higher IV MgSO₄ doses. In the present case, the serum Mg level peaked at 4.8 mmol/L 3.5 h after arrival in the ED. By then, the patient had received 10–12 g of MgSO₄ intravenously and was on mechanical ventilation when the serum Mg level peaked. The patient was extubated on the fourth hospital day and reintubated on the fifth day, and IV MgSO₄ was likely not the direct cause of reintubation because her serum Mg level on the fourth hospital day was 3.1 mmol/L, which was not markedly high.

Patients should be immediately injected with an IV bolus of 2 g of MgSO₄ if emergency physicians (EPs) suspect aconite poisoning as the cause of irregular wide QRS tachycardia, and evaluation for QT prolongation can wait. No specific conditions are known for which IV MgSO₄ is contraindicated, and serum Mg levels are the primary concern. In previous preeclampsia case reports, the relationship between the IV bolus dose of MgSO₄ and the maximum serum Mg level was as follows: 2.2 mmol/L at 2 g,²⁸ 3.1 mmol/L at 3 g,²⁹ 4.5 mmol/L at 4 g,²⁷ 5 mmol/L at 8 g,²⁵ 8.1 mmol/L at 10 g,³⁰ and 16.1 mmol/L at 20 g.³¹ Considering this dose-serum level relationship and the studies by Sawanobori et al.,^{23, 24} an IV loading dose of 2 g of MgSO₄ may be too low for aconite-induced VTAs. In cases of suspected aconite-induced VTAs, EPs should try an IV bolus of 2 g of MgSO₄ up to three times at approximately 10-min intervals, even if no apparent effect is observed. As in the present case, an IV bolus of 2 g of MgSO₄ will not restore SR at once but may rapidly terminate pulseless VT and prevent progression to VF.

Sawanobori et al.²³ suggested a long-lasting affinity of aconitine for cardiac tissue. Considering the affinity time of aconitine for cardiac tissue, a CIVI of MgSO₄ is necessary to maintain serum Mg levels in the therapeutic range for aconite-induced VTAs over the long term. The CIVI rate should then be adjusted to maintain a target serum Mg level of 3.5–4 mmol/L with ventilatory support until the ES resolves. Unlike conventional antiarrhythmic drugs, which have a narrow therapeutic range and proarrhythmic effects, IV MgSO₄ has no marked proarrhythmic impact, and the intensity of the primary side effect, neuromuscular blockade, depends on serum Mg levels. In addition, EPs can quickly measure serum Mg levels, unlike traditional antiarrhythmic drugs, and adjusting the CIVI rate of MgSO₄ would be easy.

In renal failure or anuria, serum Mg levels are highly dependent on the Vd of Mg.³² The maximum serum Mg level immediately after an MgSO₄ IV bolus is similar in renal failure and normal renal function because the Vd of Mg is relatively large.³² An IV bolus of 3 g MgSO₄ followed by an intramuscular injection of 10 g MgSO₄ did not cause the serum Mg level to exceed 3 mmol/L, even in patients with oliguria.³² Furthermore, even after IV administration of 10 g MgSO₄ over 30–60 min, serum Mg levels ranged from 2.2–5.5 mmol/L, regardless of renal failure, if previous values were normal.²⁶ In addition, IV infusion of 20 g MgSO₄ over 2–8 h did not cause serum Mg levels to exceed 3.5–6.2 mmol/L, even during renal failure and anuria.²⁵ Importantly, once IV MgSO₄ loading causes serum Mg levels to exceed normal values, serum Mg levels return to normal within 24 h with normal renal function, whereas in renal failure, serum Mg levels take several days to return to normal. In renal failure, only 5–8% of the loading Mg dose is excreted in the urine over 24 h.^{25, 26} Therefore, an IV loading dose of 4–6 g MgSO₄ followed by MgSO₄ CIVI is recommended for aconite-induced VTAs, even in the presence of oliguria caused by circulatory failure or pre-existing CKD. EPs should then measure serum Mg levels frequently during MgSO₄ CIVI and adjust serum Mg levels. MgSO₄ CIVI should be extremely tapered or discontinued in the first 24 h, especially in renal insufficiency. Otherwise, hypermagnesemia may persist for several days, and neuromuscular blockade may delay ventilator weaning. Because most aconite-poisoning survivors return to SR approximately 24 h after successful ES treatment,^{4, 33, 34} IV MgSO₄ will unlikely be necessary for more than 24 h. The present case had a low estimated glomerular filtration rate of 9.2 mL/min/1.73 m² at the first medical examination. MgSO₄ was administered intravenously only on the first hospital day, and the total dose was 16 g. The serum Mg level peaked at 4.8 mmol/L 3.5 h after arrival and finally decreased to < 2 mmol/L on the eighth hospital day. The patient was injected with 200 mL of 13.6% Mg citrate solution (89.3 mmol Mg²⁺) through the gastric tube on the first hospital day in the ED, which may have also contributed to the prolonged hypermagnesemia.

Plasma levels of the four DDAs have been reported in 13 patients with aconite poisoning.^{33, 35,36,37,38,39,40,41} Seven of these patients survived^{33, 39,40,41} and six died.^35,36,37,38 Excluding one patient who died of post-resuscitation encephalopathy six days after ingestion,³⁵ the maximum total concentration of the four DDAs (C_max) ranged from 0.7–6.4 ng/mL in survivors and from 15.1– 498.1 ng/mL in non-survivors. The C_max of the three DDAs in the present case was 37.9 ng/mL; therefore, the case is considered fatal.

In five aconite-poisoning survivors,^{33, 40} total plasma concentrations of the four DDAs at SR recovery were 0.5–1.5 ng/mL. In the present case, the total plasma concentration of the three DDAs at SR recovery was substantially higher at 5.5–8.4 ng/mL. The patient recovered to SR even at high plasma concentrations of the three DDAs, probably because serum Mg concentrations were maintained at high levels, which were 3.8–4.2 mmol/L at the time.

HP does not remove electrolytes such as Mg²⁺, but it does remove disopyramide.⁴² In the present case, removing disopyramide using HP may have been a factor that facilitated SR recovery. In general, because disopyramide is renally excreted, lower doses and longer dosing intervals are recommended in the presence of renal impairment. HP is less commonly used for toxin removal in North America than in other countries because of the development of high-flux dialyzers and the high cost of HP columns.⁴³ HD is the definitive treatment for the acute phase of hypermagnesemia; the Mg²⁺ concentration in standard dialysate is 0.5 mmol/L. The Mg²⁺ concentration in the dialysate should be high when treating fatal arrhythmias caused by aconite poisoning with HD and IV MgSO₄ because decreased serum Mg levels in HD may reproduce aconite-induced tachyarrhythmias.

This study suggests that the use of IV bolus MgSO₄ effectively terminated pulseless VT, implying that high-dose IV MgSO₄ can be a viable initial treatment option for life-threatening arrhythmias induced by aconite poisoning. Furthermore, the pharmacokinetics (including absorption, distribution, metabolism, and excretion) of DDAs in aconite poisoning is not yet well understood. Therefore, this case report alone is insufficient to prove the efficacy of HP in aconite poisoning. Additional simulation studies on HP should be conducted to validate the findings of this case study and refine treatment protocols. Lastly, the patient’s return to SR within 15 h post-ingestion, coupled with the substantial reduction in DDA levels within 9.5 h, illustrates the potential effectiveness and timeline for clinical recovery with appropriate treatment.