Acute cholestatic liver injury caused by polyhexamethyleneguanidine ...

10 downloads 0 Views 316KB Size Report
clinical cases of cholestatic hepatitis caused by pharmaceuti- cals such as phenothiazines, 2 clindomycine, 3 ciprofloxacin, 4 azathioprine, 5 cyproterone acetate ...
Clinical Toxicology (2011), 49, 471–477 Copyright © 2011 Informa Healthcare USA, Inc. ISSN: 1556-3650 print / 1556-9519 online DOI: 10.3109/15563650.2011.592837

CRITICAL CARE TOXICOLOGY

Acute cholestatic liver injury caused by polyhexamethyleneguanidine hydrochloride admixed to ethyl alcohol Y. N. OSTAPENKO1, K. M. BRUSIN2, Y. V. ZOBNIN3, A. Y. SHCHUPAK4, M. K. VISHNEVETSKIY5, V. G. SENTSOV2, O. V. NOVIKOVA2, S. A. ALEKSEENKO4, O. A. LEBED’KO4, and Y. B. PUCHKOV4 1Research

and Applied Toxicology Center of the Federal Medical-Biological Agency, Moscow, Russia State Medical Academy, Ekaterinburg, Russia 3Irkutsk State Medical University, Irkutsk, Russia 4Far-Eastern State Medical University, Khabarovsk, Russia 5MSCh 9, Perm, Russia

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

2Ural

Introduction. Polyhexamethyleneguanidine hydrochloride (PHMG) is an antimicrobial biocide of the guanidine family. In the period from August 2006 to May 2007, more than 12 500 patients were admitted to hospital with a history of drinking illegal cheap “vodka” in 44 different regions in Russia, of whom 9.4% died. In reality, the “vodka” was an antiseptic liquid composed of ethanol (≈ 93%), diethyl phthalate, and 0.1–0.14% PHMG (brand name “Extrasept-1”). Material and methods. We performed an analysis of the clinical features and outcome in four poisoning treatment centers in the cities of Perm, Ekaterinburg, Irkutsk, and Khabarovsk. A total of 579 patients (215 females and 364 males) with similar symptoms were included. Results. The main symptoms on admission included jaundice (99.7%), skin itch (78.4%), weakness (96%), anorexia (65.8%), dizziness (65.3%), nausea (54.8%), vomiting (22.6%), stomach ache (52.7%), diarrhea (32%), and fever (50%). Mild symptoms were found in 2.5% of cases, moderate in 63%, and severe in 34.5%. Laboratory results were (mean ⫾ SD): total bilirubin 249 ⫾ 158 μmol/L, direct bilirubin 166 ⫾ 97 μmol/L, cholesterol 14 ⫾ 8 mmol/L, alanine aminotransferase 207 ⫾ 174 IU/L, aspartate aminotransferase 174 ⫾ 230 IU/L, alkaline phosphatase 742 ⫾ 751 IU/L, and gamma-glutamyltranspeptidase 1199 ⫾ 1095 IU/L. Patients generally recovered over a period of 1–5 months, although high levels of alkaline phosphatase and gammaglutamyltranspeptidase were still found in all patients examined after 6 months. Sixty-one patients (10.5%) died between 23 and 150 days after poisoning. Local cholestasis, inflammatory infiltration, and fibrosis developing into cirrhosis were found by liver biopsy. Conclusion. Acute liver injury caused by PHMG-hydrochloride or PHMG in combination with either ethanol or diethyl phthalate can be characterized as cholestatic hepatitis with a severe inflammatory component causing high mortality. Keywords

Cholestatic hepatitis; Poisoning; Polyhexamethyleneguanidine hydrochloride

market. It contained ethanol (93 ⫾ 0.5%), diethyl phthalate (DEP) (0.08–0.15%), water (up to 100%), and the antiseptic polyhexamethyleneguanidine hydrochloride (PHMG) 0.1– 0.14 ⫾ 0.01%. PHMG [C7H16ClN3]n is a water-soluble, cation-active polymeric compound, a guanidine derivative with a molecular weight ranging from 1000 to 3000 Daltons. For a long time, it had been used as an indoor disinfectant in aqueous solution and was considered to have a low toxicity (class III-toxicity). It showed only mild cumulative toxic effects in mice.1 Following intragastric dosing in rat acute toxicity tests, 50% mortality (LD-50) was 700–1022 and 300 mg/kg following intraperitoneal injection. The LD-50 intragastric dose in mice was 419 mg/kg. We report a mass poisoning outbreak in Russia in 2006– 2007 with a high mortality due to this product.

Introduction During the period from August 2006 to May 2007, a mass outbreak of acute poisoning occurred in Russia involving more than 12 500 patients, of whom 9.4% died. This poisoning, with its typical clinical features, spread to involve 44 of the 89 Russian regions. It began in the regions located closest to Moscow and, over a couple of weeks, spread to the Leningradskiy region, the Volga, Ural, and Siberian regions, finally reaching the Far East. A large proportion of the victims were regular consumers of cheap illegal alcohol, and they also became the first victims of the outbreak. Analyses of liquor confiscated from the different regions where many poisonings occurred, revealed “Extracept-1” as the source of toxicity. “Extrasept-1” was produced as an antiseptic liquid for indoor disinfection, and was brought into the illegal alcohol

Materials and methods

Received 24 February 2011; accepted 27 May 2011.

This is a retrospective analysis of the clinical symptoms, laboratory data, histology, and outcomes reported from four poisoning treatment centers in the cities of Perm,

Address correspondence to Yuri Ostapenko, Federal Medical-Biological Agency, Research and Applied Toxicology Center, Moscow, Russian Federation. E-mail: [email protected]

471

472 Y. N. Ostapenko et al.

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

Ekaterinburg, Irkutsk, and Khabarovsk, and the Research and Applied Toxicology Centre, Moscow. The study included 579 poisoned patients treated at these four poisoning treatment centers: 45 patients in Perm, 66 in Ekaterinburg, 96 in Khabarovsk, and 372 in Irkutsk. Locations of the regions involved in the outbreak and the centers participating in this study are shown in Fig. 1. Laboratory data are presented on admission in all 579 patients, 1 month after admission in 312 cases, and 6 months after admission in 38 cases. Ultrasound examination was performed in 568 cases. Liver biopsy was included in the protocol of examination only in the Khabarovsk Poisoning Treatment Centre and was performed on admission and 1 month later for 62 patients. Postmortem examination of all deceased patients was performed in local forensic departments. Definitions and classifications Severity of poisoning was evaluated according to a special classification, which included bilirubin level, duration of symptoms, complications, and outcome. Mild poisonings were defined as a normal or close to normal bilirubin level with a duration of symptoms that did not exceed 1 month. Moderate cases corresponded to bilirubin levels between 100 and 300 μmol/L, with a duration of symptoms that did not exceed 2 months and with no other complications. Severe poisonings were characterized by a bilirubin level that exceeded 300 μmol/L, a duration of symptoms of more than 2 months and the presence of complications such as bleeding, encephalopathy, or oliguria. Hepatic encephalopathy was graded in four stages: 1. Mild confusion, disordered sleep pattern 2. Drowsy, inappropriate behavior

3. Stuporous, confused, disorientation 4. Coma Alcohol dependence was defined in patients who consumed alcohol constantly for 1 year or longer, that is every day or several days a month an amount that exceeded 100 mL absolute (96%) ethyl alcohol. Statistics All clinical and laboratory data of patients in the four poisoning treatment centers were collected in a single database. The data are presented as mean ⫾ standard deviation, median, and range for the continuous variables and in frequency (in percentage) for categorical variables. Statistical analysis was performed with EXCEL and Biostat software and all p values below 0.05 were considered statistically significant. The differences among groups were compared with the Student’s t-test and with z criterion.

Results Table 1 shows the age and gender characteristics of 579 poisoned patients. As for social status, the majority of patients, 396 (68.4%), were unemployed, 51 of whom were homeless. Eighty-nine patients (15.4%) were employed, 58 (10.0%) were retired, and 34 (5.9%) were invalids. A history of alcohol dependence was confirmed in 554 patients (95.7%) who consumed illegal alcohol regularly, whereas 25 patients (4.3%) were poisoned after a single drinking episode. A history of viral hepatitis A was known in 29 cases, HBV was found in 12 cases, HCV in 11, and both in one case.

Regions of Russia involved in the poisoning (total 44) Poisoning Centres participated in the Study

Fig. 1. Location of the regions involved to the outbreak and the centres participated in this study on the map of Russia (see colour version of this figure online). Clinical Toxicology vol. 49 no. 6 2011

Cholestatic liver injury 473 Table 1. Age and gender characteristics of patients. Age (years)

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

Male Female All patients (%)

17–20

21–30

31–40

41–50

51–60

61–70

Total

1 1 2 (0.3%)

41 28 69 (11.9%)

96 58 154 (26.6%)

133 64 197 (34.0%)

74 57 131 (22.6%)

19 7 26 (4.5%)

364 (62.9%) 215 (37.1%) 579 (100%)

It was almost impossible to determine the dose of alcohol contaminated with PGMH as people took alcohol from different sources over a few days. In 62 cases with good data, the median dose of alcohol was 300 mL with a range of 100–500 mL from 1 to 120 days, median 5 days. No patients were certain that all alcohol they consumed was contaminated with PHMG. About half the patients were admitted to hospital during the first week of their illness (47%), 25% during the second week, and 28% after more than 2 weeks. Clinical manifestations The main clinical manifestations were jaundice in 577 cases (99.7%), fatigue in 556 (96.0%), and pruritus in 454 (78.4%). Jaundice was observed during the first day after what was believed to be drinking this product in 62 patients (11%) and during the second and third day in 41 patients (7%). A latent period of more than 3 days was found in 103 cases (18%) and the remaining 373 patients (64%) were not aware the moment their skin became yellow. Frequent symptoms were dizziness in 380 cases (65.6%), fever in 290 cases (50%), and sinus tachycardia in 402 cases (69.4%). Gastrointestinal symptoms included anorexia in 381 cases (65.8%), nausea in 317 (54.8%), abdominal pains in 305 (52.7%), vomiting in 131 (22.6%), and diarrhea in 185 cases (32.0%). Some 142 patients (24.5%) were hypotensive and 47 (8.1%) manifested oliguria on admission. CNS complications that developed during hospital stay were observed in 49 patients (8.5%), among them 27 were confused, 13 developed psychosis, and 9 coma. Pneumonia developed in 20 patients (3.5%). Bleeding manifested in 16 cases (2.8%), including gastrointestinal bleeding in 5, nasal bleeding in 11, petechial hemorrhage in 11, and other bleeding disturbances in 3 cases. The duration of the clinical features varied; average duration of pruritus was 16 ⫾ 6 days, fatigue 30 ⫾ 5 days, dizziness 3 ⫾ 0.2 days, fever 9 ⫾ 2 days, anorexia 5 ⫾ 1 days, nausea 3 ⫾ 0.2 days, vomiting 2 ⫾ 0.4 days, diarrhea 3 ⫾ 1 days, and abdominal pain 6 ⫾ 0.2 days. Jaundice reached a maximum average on day 28 ⫾ 14 (range 7–60). The duration of jaundice was more than 1 month in 88% of cases and as long as 5 months in some cases. Jaundice was the most prolonged feature. Overall severity of poisoning was estimated as mild in 14 cases (2.5%), moderate in 365 cases (63%), and severe in 200 cases (34.5%). Laboratory findings High bilirubin level, both total and direct fraction, was the first laboratory feature found by general practitioners. Only Copyright © Informa Healthcare USA, Inc. 2011

two patients had a normal bilirubin level on admission. Bilirubin levels exceeding 300 μmol/L (17.54 mg/dL) were found in 27.8%. Median value of total bilirubin was 215.8 μmol/L (n ⫽ 572) with a range of 10.4–960.0 μmol/L (normal range 2–18 μmol/L). Median value of direct bilirubin was 150.2 μmol/L (n ⫽ 564) with a range of 3.5–561.7 μmol/L. Alanine aminotransferase (ALT) was increased in 91.5% and aspartate aminotransferase (AST) in 82.8%, but only two patients developed a rise of transaminases exceeding 1000 IU/L. Median value of ALT was 167 U/L (n ⫽ 199) with a range of 10–1162 U/L, median value of AST was 107 U/L (n ⫽ 198) with a range of 5–1998 U/L. High activity of alkaline phosphatase (AlPh) and gamma-glutamyl transpeptidase (GGT) were commonly found with median AlPh 529 U/L (n ⫽ 401, range 27.7–4250.0) and GGT 952.2 U/L (n ⫽ 332, range 34.8–9259.0). Elevated cholesterol (⬎ 6 mmol/L) was found in 89.4%, median value was 13.0 mmol/L (n ⫽ 463) with a range of 1.0–72.2 (normal range 3.1–5.6 mmol/L). The total protein concentration was in the normal range in 74.7%, but albumin concentration was ⬍ 3.5 g/dL in 59.3% (median value 3.3 g/dL, range 1.9–5.6, n ⫽ 145). Serum potassium (K) of ⬍ 3 mmol/L was found in only 20.8%; but in one case, extreme potassium depletion was restored only by continuous hemofiltration with normal potassium solution. The international normalized ratio (INR) was normal on admission in the majority of cases (median 1.0 mmol/L, range 0.8–3.2, n ⫽ 54), values exceeding 1.2 were found in only 16.7% of the patients. There were no patients with a platelet level less than 100*109/L (median 278*109/L, range 125–680, n ⫽ 127). Thrombin time was on average 17.6 ⫾ 5.8 sec. A low hemoglobin level (3.8–9.9 g/dL) was found in 19%. After one month, follow-up was performed among those who were still enrolled and treated in the four hospitals. Results of the laboratory blood tests on admission and 1 month after admission are shown in Table 2. Cases without 1-month follow-up examination were excluded from further analysis. Anemia was found more often in 29% (90/312) in comparison to 16% (49/312) on admission (z ⫽ 3.8, p ⬍ 0.0001) and median values of erythrocytes and hemoglobin were significantly lower (p ⬍ 0.001), leukocytosis exceeding 10*109/L was estimated in 28% (87/312) in comparison to 18% (56/312) on admission (z ⫽ 2.7, p ⫽ 0.007), and median leukocytes value was significantly higher (p ⬍ 0.001). Cholesterol level did not significantly change. AST, ALT, AlPh, and GGT significantly decreased but remained elevated (Table 2). There was no difference in total bilirubin level (median changed from 213.7 to 196.9 μmol/L), but direct fraction fell (from 154.8 to 101.5, p ⬍ 0.001). Median INR was 0.98 with a range of 0.74–2.14. There was only one

474 Y. N. Ostapenko et al. Table 2. Laboratory blood tests of patients on admission and 1 month later.

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

Total bilirubin, μmol/L (2–18) Direct bilirubin, μmol/L (0–4.3) ALT, IU/L (5–40) AST, IU/L (5–40) AlPh, IU/L (20–130) GGT, IU/L (7–60) Total protein, g/dL (6.5–8.7) Albumin, g/dL (3.5–5.0) Glucose, mmol/L (3.5–6.4) Cholesterol, mmol/L (3.1–5.6) Potassium, mmol/L (3.4–5.1) Erythrocytes, *1012/L Hemoglobin, g/dL Leukocytes, *109/L Platelets, *109/L

Admission

1 month

213.7 (8.0–960.0) 154.8 (3.5–452.0) 131.0 (11.0–1050.0) 106.5 (5.0–1998.0) 482.5 (16.9–3150) 950.0 (45.3–8893.1) 7.4 (4.6–10.0) 3.7 (2.0–5.6) 5.5 (3.1–15.5) 13.1 (1.0–43.6) 3.9 (2.2–5.8) 3.8 (1.4–5.3) 12.0 (3.8–15.8) 6.6 (2.0–28.6) 295.0 (202.0–680.0)

196.9 (10.8–938.3) 101.5 (0–673.9)*** 64.6 (9.0–461.9)*** 62.5 (4.0–237.0)** 377.0 (27.7–2488.0)* 818.0 (30.0–4180.0)** 7.2 (5.6–10.5) 3.4 (2.6–5.2) 5.1 (3.6–11.9)* 12.8 (1.4–44.5) 4.0 (2.7–6.2) 3.5 (1.4–4.8)*** 11.2 (3.7–16.2)*** 8.0 (3.0–45.8)*** 261.5 (251.0–339.0)

*p ⬍ 0.05, **p ⬍ 0.01, ***p ⬍ 0.001. Median (range) n ⫽ 312.

patient with an INR ⬎ 2 (initial INR 1.1), who survived but underwent one session of albumin dialysis. Patients who underwent examination 6 months later showed normal bilirubin level in 74% (28/38), others showed only a mild rise. All patients had a high AlPh and almost all had a high GGT (Table 3).

of fatal poisonings was a low potassium level. On admission, median serum potassium was 2.5 mmol/L (range 1.9–4.8) in the group of patients who eventually died and 3.6 mmol/L (range 1.6–4.9) in the surviving patients (p ⬍ 0.01). Potassium level remained the same at 1 month, 2.8 (1.5–4.1) and 3.6 (3.2–4.7), p ⬍ 0.01.

Mortality Sixty-one patients died within 23–150 days of the poisoning; the average time from intake to death being 55 ⫾ 26 days. The causes of death were determined in 40 cases and the findings were: lung edema in 19, pneumonia in 15, brain edema in 27, and acute hepatitis in 21 cases (including 7 cases of local necrosis in liver parenchyma and 3 of cirrhosis). Pancreatic necrosis was found in five cases and pancreatic abscess in one other case. Esophageal bleeding and subarachnoid hemorrhage were found in one case each. The average age of the patients who died was significantly older than survivors, 48 ⫾ 12 vs 43 ⫾ 11 years, respectively (p ⬍ 0.01). The level of total and direct bilirubin was significantly higher (p ⬍ 0.01) both on admission and at 1 month after admission in the group of patients who died (Fig. 2). Another important laboratory feature in the group

Ultrasound findings Ultrasound examination revealed liver enlargement in 85.6% of the cases, with splenic enlargement in 11.6%. The liver parenchyma was hyperechoic in 64.8%. The portal vein was dilated in 26.5% of cases; mild dilatation (13–14 mm) was found in 11.4% and severe (more than 14 mm) in 15.1%. The splenic vein was dilated in 5.3% of cases. Ascites was found in 2.8% of cases. Pathologic data Liver biopsy was performed in 62 patients hospitalized in the Khabarovsk Poisoning Treatment Centre. A transdermal biopsy was made under ultrasound control with the BARD biopsy system and G-16 needles. Typical biopsy pictures are shown in Fig. 3. The most common features were fat dystrophy of

Table 3. Laboratory blood tests of patients on admission, 1 and 6 months later. Admission Total bilirubin, μmol/L (2–18) Direct bilirubin, μmol/L (0–4.3) ALT, IU/L (5–40) AST, IU/L (5–40) AlPh, IU/L (20–130) GGT, IU/L (7–60) Total protein, g/dL (6.5–8.7) Glucose, mmol/L (3.5–6.4) Cholesterol, mmol/L (3.1–5.6) Erythrocytes, *1012/L Hemoglobin, g/dL Leukocytes, *109/L *Significant

188.0 (8–735.0) 112.0 (38.0–299.0) 113.0 (22.0–475.0) 85.0 (24.0–531.0) 827.5 (310.0–3200.0) 800.0 (107.8–4000.0) 7.6 (5.5–10.0) 5.8 (3.8–10.4) 8.0 (1.0–26.0) 3.7 (1.9–4.7) 11.7 (5.7–15.1) 7.8 (4.6–21.2)

1 month (11.0–430.0)*

96.9 60.3 (5.0–33.0) 87 (20–461.9) 68.8 (8–155.9) 951.0 (283.0–2330.0) 935.0 (112.0–2792.0) 7.3 (2.6–5.2)* 4.8 (3.6–11.9)* 7.1 (1.4–18.6) 3.7 (1.9–4.6) 12.0 (5.6–14.3) 10.2 (5.2–20.6)

6 month 14.2 (7.0–33.0)*** 5.5 (2.0–17.6)*** 69.4 (15.0–420.0) 47.4 (9.0–403.0) 397.0 (103.0–3000.0)* 1100.0 (25.0–3200.0) 7.5 (6.3–9.1) 4.4 (2.5–5.8)*** 6.2 (2.6–11.6) 4.0 (3.0–4.6) 13.3 (8.7–15.4)* 7.2 (4.7–11.0)

differences with admission, p ⬍ 0.05, **p ⬍ 0.01, ***p ⬍ 0.001. Median (range) n ⫽ 38. Clinical Toxicology vol. 49 no. 6 2011

Cholestatic liver injury 475

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

Fig. 2. Levels of the bilirubin fraction in the groups of patients who died (-d) and survived (-s) on admission and in a one month (see colour version of this figure online).

hepatocytes, mostly large droplets, which was present in 86% of biopsies, and protein dystrophy in 74%. Mallory bodies were found in 31% of cases. Portal tract white cell infiltrates, mostly including lymphocytes, were found in 98.2% of cases. A lymphocytic-macrophagal infiltrate was found in 60% of the liver biopsies, lymphocytic-histiocytic in 25.5%, lymphocytic-neutrophilic in 12.7%, and isolated lymphocytic infiltration in 1.8%. Histopathological samples revealed necrosis of the hepatic cells in 53.4% of cases, including centrilobular in 34.5%, local in 17.2%, and portal in 1.7%. Marked intracellular cholestasis was found in 27.6%, and both intracellular and bile duct cholestasis in 20.7%. Portal and periportal fibrosis were found in 94.8 and 79.3%, respectively, and there were also signs of chronic hepatitis. Cirrhosis was found in 10% of cases. Re-examination after 30 days showed polymorphonuclear cell infiltrates in 77.8% with lymphocytes and macrophages prevalent in 66.7%. Intralobular necrosis was revealed in 44.4%. However, in fatal poisonings, necrosis of the hepatic cells was found in five of six cases. Thus, the pathological processes still remained active 1 month after admission. Management Management generally included intravenous fluids and glucocorticoids, with plasmapheresis used empirically. Glucocorticoids appeared to have no therapeutic effect. Khabarovsk Poison Centre staff used ursodeoxycholic acid (UDCA) and felt it had benefit. According to Paumgartner, UDCA protects the biliary epithelium by reducing the toxicity of bile, stimulates hepatobiliary secretion by upregulating transporters, and inhibits apoptosis.11 Seventeen patients admitted to Khabarovsk Poison Treatment Centre received the basic therapy and further 11 patients were treated by UDCA for 1 month with a dose of 10 mg/kg in addition to the basic therapy. Total and direct bilirubin decreased in both groups in 1 month after admission, but the differences were not significant (Table 4). ALT levels decreased significantly in both groups, but AST only in the UDCA-treated group. Alkaline phosphatase and cholesterol did not change in the basic therapy group and decreased in the UDCA-treated group, but the difference was not significant. Albumin dialysis (molecular absorbent recirculation system, MARS) was performed on seven patients in Ekaterinburg Poisoning Treatment Centre; three of seven patients died. The Copyright © Informa Healthcare USA, Inc. 2011

indication for the MARS therapy was third or fourth degree hepatic encephalopathy (psychosis or coma) with a bilirubin level that exceeded 500 μmol/L. The patients received from one to four sessions of MARS therapy (overall 14) with an average duration of 8.3 ⫾ 0.5 h. Total bilirubin before the MARS therapy was on average 570 ⫾ 22.1 μmol/L, and after the treatment it was 433 ⫾ 17 μmol/L (p ⬍ 0.001). After the treatment, bilirubin in three patients returned to pre-treatment values within 2–3 days and these patients died within 1 month. In four patients, bilirubin did not increase; each had one session and recovered.

Discussion Cholestatic liver injury due to the toxic effects of chemicals or drugs is less commonly found than necrotic injury. Through a literature search, we found only a few reports on clinical cases of cholestatic hepatitis caused by pharmaceuticals such as phenothiazines,2 clindomycine,3 ciprofloxacin,4 azathioprine,5 cyproterone acetate,6 pyritinol,7 some nutritional supplements,8 and herbal products.9,15 Anabolic steroid exposure resulted in bland cholestasis causing an inhibition of bile salt and bilirubin excretion. Cholestasis with periductal inflammation may be the result of inhibition of Na-K ATPase, which results in decreased biliary flow.16 Cholestatic hepatitis caused by drugs is considered to follow hypersensitivity reactions when immune system tolerance is impaired. Cytotoxic lymphocytes CD4 ⫹, CD8 ⫹, and natural killer cells are thought to be involved in these reactions.3 In a study by Maria et al.,7 six previously healthy subjects who developed cholestatic hepatitis during pyritinol treatment had a high in vitro CD4 ⫹ T-cell response to this drug. According to Zimmerman and Lewis,2 sites of lesions that contribute to cholestasis include the sinusoidal and canalicular plasma membrane, the pericanalicular network, the tight junction, and, possibly, microtubules, microfilaments, and Golgi apparatus. The mass poisoning cases we report resulted in severe cholestatic hepatitis that we believe was caused by the disinfectant PHMG, which had not previously been reported as a cause of acute poisoning.

Portal tract infiltrate

Periportal infiltrate

Fig. 3. Liver histology of pat. N, 57 y.o. on admission. Marked portal lymphocytic and hystiocytic infiltration, polymorphonuclear infiltration, intracellular and intraductal cholestasis, acidophilic degeneration of hepatocytes, local necrosis with perifocal lymphocytic and hystiocytic infiltration (moderate histological activity). Stained with hematoxelineosin, ⫻140 (see colour version of this figure online).

476 Y. N. Ostapenko et al. Table 4. Laboratory blood tests of patients treated with basic therapy and UDCA on admission and 1 month later. Admission Total bilirubin, μmol/L (2–18) Direct bilirubin, μmol/L (0–4.3) ALT, IU/L (5–40) AST, IU/L (5–40) AlPh, IU/L (20–130) GGT, IU/L (7–60) Cholesterol, mmol/L (3.1–5.6)

1 month

Basic therapy (n ⫽ 17)

UDCA (n ⫽ 11)

Basic therapy (n ⫽ 17)

UDCA (n ⫽ 11)

134 (8–408) 105 (38–200) 76 (22–475) 52 (24–188) 666 (310–2485) 1000 (250–4000) 6.5 (1.0–24.5)

122 (35–256) 70 (13–200) 186 (35–453) 181 (30–493) 907 (215–3150) 620 (80–3400) 10.9 (5.0–22.0)

95 (11–430) 68 (5–310) 44 (20–172)* 70 (8–125)† 667 (294–1421) 379 (112–1400) 7.2 (1.4–16.9)

62 (5–324) 40 (3–170) 95 (26–289)* 49 (26–143) 569 (159–1453) 406 (65–1600) 6.3 (3.1–31.8)

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

*Significant differences on admission, p ⬍ 0.05. †Significant differences between two groups, p ⬍ 0.05. Median (range).

Several samples of the “cheap vodka” that was consumed before developing jaundice were collected in different territories of the country involved in this mass poisoning episode and subjected to chemical-toxicological investigation using GLC and HPLC methods. Ethanol, DEP, and PHMG were found in all samples, and these were identical to the ingredients of the antiseptic liquid “Extrasept-1” produced in Russia. There have been a few experimental studies of DEP hepatic toxicity in mice and rats. Higher concentrations of DEP resulted in mitochondrial proliferation as well as in accumulation of glycogen, cholesterol, and triglycerides within the mouse liver.12 Serum total cholesterol levels in DEP and ethanol ⫹ DEP treated rats were significantly higher as compared to control and ethanol alone-treated rats. Thus, DEP alone led to severe impairment of lipid metabolism and toxic liver injury, as judged by altered lipid and enzyme levels in the liver and serum.13 A human risk assessment of DEP in alcoholic beverages accepted the potential hepatic toxicity of DEP was possible and noted a higher incidence of liver cirrhosis in regions with high consumption of unregulated alcohol, but felt that there was lack of sufficient evidence to postulate a real public health threat.14 Information about PHMG metabolism and toxicity is very limited. This chemical is a strong antiseptic with broadspectrum antimicrobial activity, including mycobacterium tuberculosis. Thus, the precise mechanism of PHMG toxicity remains unclear. In 2005, Muller and Kramer considered guanidine derivates as nontoxic, but studied a very low concentration (0.005%).10 We believe PHMG was key in the development of cholestatic hepatitis. From our experience, people who often consume cheap illegal alcohol take DEP admixed to ethanol with no manifestation of jaundice; thus, we believe that the addition of PHMG dramatically changed the toxicity of the mixture. There was a varying susceptibility to symptoms. As a rule, people consumed liquid containing PHMG in a group, but not all of them developed symptoms. We were not able to define the actual dose of poison as most of the patients consumed alcohol for several days and did not know which bottles contained PHMG. Median daily dose was 300 mL in cases with more or less certain doses of contaminated alcohol. Accordingly, the dose of PHMG might be 0.3–0.4 mL and PHMG could accumulate during continuous consumption.

Therefore, we cannot confirm whether hepatic toxicity was dependent on the dose ingested. Almost all poisoned people were regular drinkers and the presence of earlier alcoholinduced hepatitis or cirrhosis seemed to contribute to a more severe course of the poisoning. Twenty-five patients (4.3%) were not regular drinkers, but consumed the PHMG-containing alcohol once at a party. They also developed symptoms of poisoning described above. We did not find any obvious difference in hepatitis clinical course and outcome in these cases. Cholestatic hepatitis was accompanied by a marked inflammatory component confirmed by liver biopsy data, but we were not able to confirm whether the immune system was involved in the pathogenesis of this poisoning. In the majority of cases, the liver injury was not accompanied by hypocoagulation or renal failure. Similarly, severe encephalopathy was infrequent. Liver biopsy revealed marked cholestasis in 48% of cases, while features of portal fibrosis were found in 95% and cells necrosis in 53% of cases. Therefore, liver damage was probably multifactorial and often resulted in the combined action of ethanol, DEP, and PHMG, accompanied by chronic alcohol liver disease and chronic viral hepatitis in some cases. While cytolysis and inflammation were involved in the toxic process, cholestasis was the main clinical syndrome. A 6-months observational study of 38 patients confirmed the presence of a mild cholestatic syndrome without the features of liver failure. Hypokalaemia which did not recover despite intravenous potassium administration and high bilirubin level (>400 μmol/L) increased during the month after the poisoning onset were associated with poor prognosis. We found no benefit of corticosteroids, and our experience of UDCA is insufficient to make a clear recommendation on its efficacy. Limitations There were several limitations to our study. Despite the large number of patients we reported (n ⫽ 579), this represented only about 5% of the outbreak. This study was performed in only 4 of 44 regions involved in outbreak. It was primarily the more severely poisoned patients that were admitted to toxicology centers, whereas the rest were treated in general hospitals and not included in this report. Therefore, the cases described may not give a representative picture of the whole outbreak. Clinical Toxicology vol. 49 no. 6 2011

Cholestatic liver injury 477

Clinical Toxicology Downloaded from informahealthcare.com by 188.226.96.102 on 08/12/11 For personal use only.

Laboratory and ultrasound data were collected from four different hospitals, and were performed with different kinds of equipment. Availability of some methods of diagnostics and treatment was quite different in the four centers involved in the study. For example, liver biopsy was performed only in the Khabarovsk Centre and albumin dialysis only in the Ekaterinburg Centre. The time between PHMG intake and arrival to hospital varied greatly. Some patients were first treated in local hospitals, and then transferred to the toxicology centers after their condition worsened. This explains why not all patients had their peak bilirubin level at the first reported blood test. One-month follow-up examination was performed only among those patients who were still inpatients or under active follow-up. Finally, we were only able to gather 38 patients for examination after 6 months, which is nevertheless important given the population being studied.

Conclusion Antiseptic PHMG in complex with ethanol and DEP is a highly toxic compound. Acute liver injury caused by PHMG could be characterized as cholestatic hepatitis with a severe inflammatory component causing high mortality. The effect of previous chronic alcohol abuse and liver injury may have made the damage due to PHMG more severe in this case series. At this stage, we cannot be sure of the mechanism of the effect of PHMG. Further experimental investigations are necessary in order to evaluate the direct hepatotoxic risk and to estimate the dose-dependent effect of PHMG alone, and the interaction between ethanol, PHMG, and DEP in animal model.

Acknowledgments The authors are very thankful to Dr Knut Erik Hovda from the Medical Division of Oslo University Hospital, Norway, for providing constructive commentary and initial English revision of the manuscript.

Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.

Copyright © Informa Healthcare USA, Inc. 2011

References 1. Oule MK, Azinwi R, Bernier AM, Kablan T, Maupertuis AM, Mauler S, et al. Polyhexamethylene guanidine hydrochloride-based disinfectant: a novel tool to fight meticillin-resistant Staphylococcus aureus and nosocomial infections. J Med Microbiol 2008; 57: 1523–1528. 2. Zimmerman HJ, Lewis JH. Drug-induced cholestasis. Med Toxicol 1987; 2:112–160. 3. Aygen C, Kocaman O, Gerbez Y, enterk C, Helage S. Clindamycininduced acute cholestatic hepatitis. World J Gastroenterol 2007; 13:5408–5410. 4. Dichiara AJ, Atkinson M, Goodman Z, Sherman KE. Ciprofloxacininduced acute cholestatic liver injury and associated renal failure. Case report and review. Minerva Gastroenterol Dietol 2008; 54: 307–315. 5. Romanguolo J, Sadowski DC, Lalor E, Jewell L, Thomson AB. Cholestatic hepatocellular injury with azathioprine: a case report and review of the mechanisms of hepatotoxicity. Can J Gastroenterol 1998; 12:479–483. 6. Savidou I, Deutsch M, Soultati AS, Koudouras D, Kafiri G, Dourakis SP. Hepatotoxicity induced by cyproterone acetate: a report of three cases. World J Gastroenterol 2006; 12:7551–7555. 7. Maria V, Albuquerque A, Loureiro A, Sousa A, Victorino R. Severe cholestatic hepatitis induced by pyritinol. BMJ 2004; 328:572–574. 8. Whitt KN, Ward SC, Deniz K, Liu L, Odin JA, Qin L. Cholestatic liver injury associated with whey protein and creatine supplements. Semin Liver Dis 2008; 28:226–231. 9. Tarantino G, Pezzullo MG, di Minno MND, Milone F, Pezzullo LS, Milone M, Capone D. Drug-induced liver injury due to “natural products” used for weight loss: a case report. World J Gastroenterol 2009; 15:2414–2417. 10. Muller G, Kramer A. Effect of selected wound antiseptics on adult articular cartilage (bovine sesamoid bone) in the presence of Escherichia coli and Staphylococcus aureus. J Orthop Res 2005; 23:127–133. 11. Paumgartner G. Pharmacotherapy of cholestatic liver diseases. J Dig Dis 2010; 11:119–125. 12. Pereira C, Mapuskar K, Rao CV. Chronic toxicity of diethyl phthalate in male Wistar rats – a dose-response study. Regul Toxicol Pharmacol 2006; 45:169–177. 13. Sonde V, D’souza A, Tarapore R, Pereira L, Khare MP, Sinkar P, Krishnan S, Rao CV. Simultaneous administration of diethylphthalate and ethyl alcohol and its toxicity in male Sprague-Dawley rats. Toxicology 2000; 147:23–31. 14. Leitz J, Kuballa T, Rehm J, Lachenmeier DW. Chemical analysis and risk assessment of diethyl phthalate in alcoholic beverages with special regard to unrecorded alcohol. PLoS One 2009; 4:e8127. Published online PMCID:10.1371/journal.pone.0008127. 15. Brent J. Three new herbal hepatotoxic syndromes. Clin Toxicol 1999; 37:715–719. 16. Goldfrank LR. Goldfrank’s Toxicologic Emergencies. 7th ed. New York: McGraw-Hill; 2002: 216–230.