In order to set up the analytical method, blood specimens were collected from 80 patients who were admitted to Chungnam National University Hospital from February to July in 2015. The collected specimens were respectively stored in EDTA solution at 4 ℃ until analysis. GCS score, mental change, elapsed time from intoxication and estimated dose of patients were recorded at the time of admission.

Blank human whole blood was purchased from the Korean Red Cross (Daejeon, Korea) and used for validation of the analytical procedure. Zolpidem standard and trimipramine-d₃, taken as the internal standard (IS), were purchased from Cerilliant (Round Rock, TX, USA). All solvents were products from Burdick & Jackson (Muskegon, MI, USA), of HPLC grade.

Stock solution of zolpidem was prepared at 1 and 10 μ g/mL in methanol, and trimipramine-d₃ was prepared at 5 μ g/mL. Working solutions were prepared by diluting stock solution to appropriate concentrations. All solutions were stored at 4 ℃.

Zolpidem metabolite, collected from patients in hospital.

Solid-Phase Extraction (SPE) cartridges (Bond Elut Certify, 130 mg/3 mL, Agilent Technologies, Santa Clara, CA, USA) were used for sample preparation.

3.3.1 Sample preparation for screening of patients samples

For each sample, a 1.0 mL aliquot of blood sample and 3 mL of 0.1 mol/L phosphate buffer (pH 6.0) were added into a test tube. The mixture then followed to undergo: 2 minutes of vortex, 10 minutes of sonification and 10 minutes of centrifugation at 3500 rpm. 3 mL supernatant of the mixture was loaded into an SPE cartridge preconditioned with 2 mL methanol and 2 mL of 0.1 mol/L phosphate buffer. The cartridge was then washed with 1 mL water and 1 mL 0.2 mol/L acetic acid. Vacuum (15-in Hg) was applied for 4 minutes to dry. 50 μ L of methanol was applied, and then dried for 1 minute (vacuum, 15-in Hg).

The acidic analytes were eluted with 3 mL of chloroform/acetone (1:1, v/v) (Acidic fraction). The basic analytes were eluted with 3 mL of 2% ammoniated ethyl acetate (Basic fraction). The eluted substances were evaporated using nitrogen gas (50 ℃). The extract was reconstituted with 100 μ L of methanol, from which 1 μ L was injected into the GC-MS for analysis.

3.3.2 Sample preparation for quantification of zolpidem

A 1.0 mL aliquot of blood sample was added into a test tube, together with 30 μ L of IS (trimipramine-d₃) and 3 mL of 0.1 mol/L phosphate buffer (pH 6.0). The following steps were same as those for patients samples in 3.3.1 above, with the exception of second dryness at vacuum of 1-in Hg.

3.4.1 GC-MS condition for screening of patients samples

An Agilent 7890B GC system coupled with an Agilent 5977A MSD (Agilent Technologies, Santa Clara, CA, USA) was used in an assemblage of the HP-5MS fused silica capillary column (30 m × 0.25 mm i.d., 0.25 μ m film thickness, Agilent Technologies). Splitless mode was selected. Helium was chosen as the carrier gas at the flow rate of 1.0 mL/min. The oven temperature was set at 80℃ initially for 1 min, and then programmed to increase to 300 ℃ by 20 ℃/min to hold for another 15 min. The temperatures at injection port and transfer line were 300 ℃ and 250 ℃, respectively. The mass spectrometer was operated at 70 eV on electron impact mode. Analysis was performed in scan mode (m/z 45 ~ 550).

3.4.2 GC-MS condition for quantification of zolpidem

Identical apparatus was applied as in 3.4.1. Split mode (20꞉1) was selected. The carrier gas and running were same as those in 3.4.1. The oven temperature was set at 120° C initially for 1 minute, and then programmed to increase to 300° C by 30° C/min to hold for another 3 min. Following the same temperature setup at injection port and transfer line as in 3.4.1 plus the mass spectrometer operating, the analysis was performed in selected ion monitoring (SIM) mode, with the parameters of m/z 235, 307 and 219 for zolpidem and m/z 61, 249 and 297 for trimipramine-d3.

The method for quantification was validated by measuring linearity, extraction efficiency, limit of detection (LOD), limit of quantification (LOQ), precision, and accuracy ^[13].

Calibration curve was constructed for quantification by injecting 6 different concentrations (20, 50, 100, 200, 500 and 1000 ng/mL) of zolpidem into blank specimens. Extraction efficiency was estimated by comparing the injected solution with the extracted one. LOD and LOQ were determined through signal-to-noise (S/N) at 3:1 and 10:1. Precision and accuracy were tested by replicate analyses (n=3) of repeating 3 times for 3 days. Precision was calculated by the percentage of either relative standard deviation or accuracy.

Patients admitted to the emergency room of Chungnam National University Hospital from February to July in 2015 were examined. 80 patients (46 males and 34 females) were intoxicated, and among them, 15 patients (18.7%, 12 males and 3 females) were intoxicated with zolpidem (Fig. 1). The ages ranged from 20 to 88, with the average of 60.

Fig.1

	Figure Option ViewDownloadNew Window
	Fig.1 Demographics of patients intoxicated with zolpidem

Through analysis of 80 patient`s samples from Chungnam National University Hospital, zolpidem was the most frequent drug, identified in 15 cases, followed by 9 cases of diphenhydramine, 8 cases of tramadol and acetaminophen, 6 cases of chlorpheniramine, 5 cases of quetiapine, and 4 cases of doxylamine. Imipramine, temazepam and trazodone were of 3, respectively. Diazepam, phenobarbital, dihydrocodeine and carbamazepine were of 2, individually.

Most of the drugs such as benzodiazepine were analyzed through basic fraction, but acetaminophen and phenobarbital were analyzed in acid fraction.

The method was validated before analyzing zolpidem from actual samples. Linearity, LOD and LOQ, extraction efficiency, and precision and accuracy of zolpidem out from GC-MS were the tested parameters.

The calibration curve was linear over the concentration range of 20 - 1000 ng/mL, with the correlation coefficient of r² =0.9997 (Fig. 2). The extraction efficiency ranged from 64.4% to 77.7%. LOD and LOQ values were 1 and 10 ng/mL, respectively (Table 1). The intra- and inter-day precision and accuracy were acceptable, ranging from 0.4% - 1.85% and 86.82% - 113.72%, respectively (Table 2).

Fig.2

	Figure Option ViewDownloadNew Window
	Fig.2 Calibration curve of zolpidem

Table 1

Table 1

Table 1 Method validation for zolpidem Compound RT /min Quantifier Ion /(m/z) Qualifier Ion /(m/z) R2 LOD/( ng/mL) LOQ/( ng/mL) Efficiency/% Zolpidem 9.47 235 219, 307 0.9997 1 10 71.80

Table 1 Method validation for zolpidem

Table 2

Table 2

Table 2 Intra- and inter-day precision and accuracy Concentration Precision/ (RSD, %) Accuracy/ % /(ng/mL) Intra-day Inter-day Intra-day Inter-day Zolpidem 50 1.85 0.71 113.72 113.33 500 0.44 0.72 86.82 87.39 1000 0.4 0.42 104.92 105.09

Table 2 Intra- and inter-day precision and accuracy

The blood concentration of zolpidem in 15 patients ranged from 19.63 - 3605.85 ng/mL (Table 3). In cases of ingesting zolpidem alone (n=5), the blood concentration was spanning from 20.04 - 3605.85 ng/mL, while bestriding across 19.63 - 2195.85 ng/mL in the cases involving multiple drugs (n=10). According to Schulz ^[12], the blood concentration of zolpidem amid 80 - 200 ng/mL is therapeutic concentration, toxic for 120 - 700 ng/mL, and comatose-fatal across 1600 - 7700 ng/mL. Therefore, 3 cases were classified as comatose-fatal, 4 as toxic, 2 as therapeutic and 5 as less than therapeutic in this study. In terms of fatal concentration of zolpidem, it has been reported that the blood concentration of the zolpidem spiked with alcohol was 800 - 900 ng/mL, while in two cases resulting in death from zolpidem ingestion alone, the blood concentrations were even 4300 and 7900 ng/mL ^[14]. In addition, according to Chung ^[11], blood concentration of zolpidem was 100 - 2620 ng/mL in 16 deaths. In Jones ^[8], the average blood concentration was 300 ng/mL in deaths related to zolpidem. Jonsson ^[9] reported that the average blood concentration was 1500 ng/mL in deaths by zolpidem alone. Tominaga ^[10] discovered that the blood concentration of zolpidem ranged from 60 - 240 ng/mL in 3 deaths. In other studies ^{[15, 16, 17, 18]}, 520 - 5000 ng/mL of zolpidem was detected in post-mortem cases. Thus, fatal concentration of zolpidem differs significantly by case from 60 to 7900 ng/mL. These results indicate that various factors such as dose, elapsed time after intoxication and co-administration with other drugs have their own and/or combinatorial effect. Therefore, toxicokinetics studies on zolpidem should be carried out.

Zolpidem is absorbed readily from the gastrointestinal tract. First-pass hepatic metabolism results in oral bioavailability of 67%, and 92% of the absorbed drug binds with plasma proteins. Zolpidem has a half-life of 2.2 + 0.4 hours, but shorter in children (less than 1.4 hours) and longer for elders (about 2.8 hours). Patients with hepatic cirrhosis show longer half-life of about 9.9 hours ^{[19, 20]}. After single oral administration of 10 mg, the peak blood concentration was 120 ng/mL in average at 1.6 hours, and 230 ng/ml on 2.1 hours for 20 mg ^{[21, 22, 23, 24, 25, 26, 27]}. In the case of overdosing 300 mg of zolpidem, added with 600 mg of prothipendyl (a neuroleptic) and/or ethanol, the zolpidem blood concentrations at 3, 4, and 5 hours after ingestion were 550, 435, and 290 ng/mL, respectively ^[28].

Out of the 15 patients, 5 (33%) had ingested zolpidem for suicide attempt; 2 patients suffered from depression; schizophrenia, anxiety disorders and chronic alcoholism had 1 case for each. There was 1 patient suffering from Alzheimer’ s and Parkinson’ s diseases. The elapsed time after intoxication varied, ranging from 1.5 to 25 hours until analysis, which was carried out within 12 hours for 9 cases and 24 hours for 14 cases. The longest time elapsed was 25 hours.

In this study, 7 were elderly patients (over 65 years old). The average concentration of zolpidem in the 5 patients of the 80s (year-old) was 1571.81 ng/mL, 2 patients in the 70s was 214.73 ng/mL. And 2 patients in the 50s was 523.79 ng/mL, 5 patients in the 40s was 291.17 ng/mL, indicating the highest zolpidem concentration appearing in the highest age group (Fig. 3). The average concentration of patients over 65 years old was 1184.08 ng/mL (21.64 - 3605.84 ng/mL), and 317.96 ng/mL (19.63 - 1027.54 ng/mL) in patients under 65 years old.

Table 3

Table 3

Table 3 Clinical status and blood concentrations Case No. Age/Sex Elapsed time after intoxication/h GCS score Mental change Zolpidem /(ng/mL) Alcohol /% Estimated dose /mg Single drug 1 42/M 8.5 E2, V3, M5 (10) drowsy 414.25 - 500 4 52/M 4.5 E4, V5, M6 (15) alert 20.04 0.104 187.5 6 80/M 25 E1, V1, M4 (6) stupor 3605.85 - 50 11 71/M 1.5 E4, V5, M6 (15) drowsy 407.83 - 160 14 82/M 12 E2, V2, M5 (9) stupor 357.18 - unknown Multiple drugs 2 43/M 17.5 E4, V1, M4 (9) confusion 19.63 - 6.3 3 53/M 5 E4, V4, M6 (14) drowsy 1027.54 0.106 100 5 46/M 16 E1, V1, M5 (7) deep drowsy 199.9 - 40 7 79/F 24 (estimated) E4, V5, M6 (15) alert 21.64 - 18.8 8 81/M 13 E4, V4, M6 (14) alert 1627.77 - 250 9 81/F 9 E2, V1, M5 (8) drowsy 2195.85 - 187.5 10 20/M 14 E2, V3, M4 (9) drowsy 40.3 - 12.5 12 88/F 11 E1, V1, M1 (3) semicoma 72.42 - 10 13 46/M 1 E2, V2, M4 (8) drowsy 692.55 0.112 250 15 41/M 11 E2, V4, M6 (12) drowsy 129.51 - 25

Table 3 Clinical status and blood concentrations

Fig.3

	Figure Option ViewDownloadNew Window
	Fig.3 Correlation between zolpidem blood concentration and patient age

In reviewing the GCS score at the time of admission, it ranged from 3 to 15. Semi-coma was observed at GCS 3, stupor at GCS 6, and deep drowsy at GCS 7. Drowsy, stupor and confuse were seen at GCS 8 and 9. Alert and drowsy were both observed at GCS 12 and 15. In comparing GCS score with blood concentration of zolpidem, the two measurements did not agree. The average concentration of zolpidem was 149.84 ng/mL at GCS 15, 1327.65 ng/mL at GCS 14, 129.51 ng/mL at GCS 12, 414.25 ng/mL at GCS 10, 139.04 ng/mL at GCS 9, 1444.2 ng/mL at GCS 8, 199.9 ng/mL at GCS 7, 3605.85 ng/mL at GCS 6, and 72.42 ng/mL at GCS 3. Additionally, mental change did not agreed with the blood concentration of zolpidem. Thus, comparing the clinical symptoms in accordance with the blood concentration is inadequate. Various factors must be taken into consideration in order to determine intoxication patterns of zolpidem.

The blood concentration of the patient who allegedly ingested 500 mg of zolpidem was 414.5 ng/mL, whereas that of the patient who allegedly ingested 50 mg was 3605.85 ng/mL. There was no correlation between estimated dose and blood concentration of zolpidem. Even in the cases of 8 and 13, the administration of 250 mg resulted in different concentrations of 1627.77 and 692.55 ng/mL. Also, two cases of 187.5 mg ingestion caused significant difference of 20.04 and 2195.85 ng/mL. In comparing blood concentration with the elapsed time after intoxication, 692.55 ng/mL after 1 hour, 407.83 ng/mL after 1.5 hours and 3605.85 ng/mL after 25 hours were observed. The blood concentration was roughly but not exactly in proportion getting higher when the elapsed time elongated, indicating lack of accuracy on information regarding estimated dose.

Alcohol test was conducted on patients who ingested zolpidem, with 3 patients being tested positive. The blood concentration of alcohol ranged from 0.104 to 0.112% (Table 3). In the case of alcohol level of 0.104%, the blood concentration of zolpidem was low, his mental state was alert and no effects shown of alcohol. For the two cases with the alcohol concentration of 0.106 and 0.112% was showed the mental state of drowsiness, indicating no significant interaction between alcohol and zolpidem. For cases of ingesting zolpidem with alcohol, some alcohol interaction was observed with the ingestion of 10 mg of zolpidem, while very little interaction was observed with the ingestion of 15 mg or greater, coinciding with previous studies ^{[21, 29, 30, 31]}.

Among patients intoxicated with zolpidem, 10 cases involved ingestion of other drugs. The most frequently taken concurrently was tramadol, identified in 4 cases, followed by acetaminophen of 3, and trazodone and chlorpheniramine of 2, respectively. Alprazolam, clozapine, diphenhydramine and quetiapine had 1 case reported for each (Fig. 4).

Fig.4

	Figure Option ViewDownloadNew Window
	Fig.4 Zolpidem combined with other drugs

In cases where zolpidem was ingested alone, the mental change was alert at blood concentration of 20.04 ng/mL and drowsy around 407.83 - 414.25 ng/mL. Despite some consistency in the observed patterns, mental change of stupor is also observed at 357.18 and 3605.85 ng/mL, showing no correlation between blood concentration and mental change. Establishing correlation between the two components was even more difficult in cases of multiple drug ingestion.

In order to correlate the blood levels with clinical symptoms, it is a must to develop one method to define GCS score and mental change along with obtaining the accurate information of amount ingested.

The concentrations of zolpidem in certain patients from, e.g., Cases 6, 8 and 9, revealed 3605, 1627 and 2195 ng/mg, which could be comatose-fatal according to Schulz ^[12]. However, mental changes of the three involved patients were stupor, alert and drowsy when admitted, demonstrating not serious with comparison to the zolpidem level in blood. This result is compatible with the report by Suh et al. ^[32]. They reported that the clinical course couldn’ t be predicted by zolpidem concentration, concluding that the acute zolpidem intoxication is mild in progress. They determined the blood concentration of zolpidem in 30 patients ranging from 200 to 7400 ng/mL by REMEDi approach, observing that all patients recovered to full consciousness. According to Garnier ^[33], 4 cases of coma plus only 1 case of respiratory trouble were observed out of 344 cases. In Wyss ^[34], only 1 case of coma was observed out of 54 cases. These studies indicate relatively moderate intoxication symptoms and easy recovery from zolpidem. Even the patients, who showed fatal concentration of zolpidem, recovered and were discharged within a few days. As Garnier stated ^[33], major intoxication symptom of zolpidem is decreasing in consciousness with moderate clinical outcomes.