Int J Clin Exp Med 2015;8(9):16670-16675 www.ijcem.com /ISSN:1940-5901/IJCEM0011996

Original Article Clinical features of ischemic caused by with four different types: a retrospective study of 328 cases

Gang Guo1, Xian-Zheng Wu1, Li-Jie Su1, Chang-Qing Yang2

1Department of Emergency Internal , Tongji Hospital Affiliated to Tongji University, Shanghai 200065, P.R. China; 2Department of , Tongji Hospital Affiliated to Tongji University, Shanghai 200065, P.R. China Received June 27, 2015; Accepted August 11, 2015; Epub September 15, 2015; Published September 30, 2015

Abstract: The aim of the study was to investigate the clinical features of ischemic hepatitis due to shock with four dif- ferent types (allergic shock, , , and cardiogenic shock). A total of 328 patients (200 males, 128 females, mean age, 65.84 ± 15.21 years old, range, 15-94 years) diagnosed with shock in Tongji Hos- pital were retrospectively investigated from Jun 2008 to Feb 2010. The parameters of function test, including alanine aminotransferanse (ALT), aspartate aminotransferanse (AST), lactate dehydrogenase (LDH), total bilirubin (TB), alkaline phosphatase (ALP) and γ-glutamyltransferase (γ-GT), were recorded and analyzed. Besides, the serum levels of C-reactive protein (CRP) and brain natriuretic peptide (BNP) were also measured and relevant correlation analysis was conducted. Among all the cases, 242 (73.8%) patients developed ischemic hepatitis. The mortality of shock patients combined with ischemic hepatitis was significantly higher than the total mortality (26.0% vs 23.8%, P < 0.05). The incidence of hepatic damage was highest in the septic shock (87.5%), while the lowest in thehypo- volemic shock (49.4%). The sensitivity of ALT elevation was higher than that of AST. In addition, CRP was positively correlated with the levels of liver function parameters in the septic shock and BNP was positively correlated with that in the cardiogenic shock. Ischemic hepatitis is a common complication of shock, increasing the mortality of shock patients. The septic shock is the most common cause of hepatic damage in shock patients. CRP may be a useful predictor for septic shock, while BNP may be a useful predictor for cardiogenic shock.

Keywords: Ischemic hepatitis, shock, allergic, hypovolemic, septic, cardiogenic

Introduction “ischemic hepatitis” [8]. After that, a large amount of studies concerning ischemic hepati- Ischemic hepatitis, also named hypoxic hepati- tis were followed, including animal model exper- tis or shock liver, describes diffuse liver dam- iments and clinical investigations [9, 10]. In age due to the acute and sharp reduction of clinical practice, it has been suggested that the hepatic blood perfusion, with the histological most common cause of ischemic hepatitis was characteristics of the necrosis of hepatic cen- shock, among the various complex clinical set- trilobular cell [1, 2]. It has been reported that tings [4]. The majority of the previous studies the prevalence of ischemic hepatitis was 2.4- are concerned about only one of shock types 11% in the department of intensive care unit, [11, 12]. However, little is known about the clini- frequently occurring in patients with critically ill cal study of ischemic hepatitis caused by shock diseases, such as severe anemia, chronic peri- cardial constriction, cardiac arrest, and shock with four different types among a large popu- [3-5]. Emerging evidence has shown that the lation. presence of ischemic hepatitis predicts a poor In order to improve our clinical understanding of prognosis in critically ill patients [4, 6, 7]. the underlying pathogenesis and clinical man- In 1979, Bynum et al. reported that 7 patients agement of ischemic hepatitis caused by shock, with cardiovascular disease had marked eleva- a retrospective study of the clinical features of tions of transaminase and first used the term ischemic hepatitis caused by shock with four Ischemic hepatitis in different type shocks different types (allergic shock, hypovolemic infectious disease history, were recorded with- shock, septic shock, and cardiogenic shock) in the first week on admission. was conducted. In addition, the effect of shock type on the hepatic function was analyzed. In Liver blood laboratory tests were performed on the final, the relevant correlation analyses were occasion of the diagnosis of shock, the first 24 also analyzed. h of the diagnosis of shock, and the first week of the confirmed diagnosis of shock. Liver blood Materials and methods laboratory data included hepatic enzymes, bili- ary enzymes and bilirubin. The hepatic enzymes Patients includedserum alanine aminotransferanse (ALT), aspartate aminotransferanse (AST), and From Jun 2008 to Feb 2010, a total of 328 lactate dehydrogenase (LDH), while biliary patients diagnosed with shock of different types in Tongji Hospital Affiliated to Shanghai enzymes included alkaline phosphatase (ALP) Tongji University were retrospectively investi- and γ-glutamyltransferase (γ-GT). Besides, gated. The enrolled patients fulfilled the follow- other routine lab data, including total bilirubin ing inclusion criteria, which was the presence (TB), C-reactive protein (CRP) and brain natri- of shock, including: (1) specific reasons for the uretic peptide (BNP), were also obtained and induction of shock; (2) two or more of the fol- analyzed. lowing terms: the presence of abnormal con- sciousness; heart rate higher than 90 beats/ Definition min or inability to touch the heat rate; clammy The ischemic hepatitis wasdefined as in previ- peripheries, positive result of the digital mark- ing test on the sternal skin, patterned skin, pale ous studies [13, 14], including the following or cyanotic mucosa, urinary output less than conditions: (1) a specific cardiac or circulatory 30 ml/h or anuria; (3) one or more of the follow- shock or other clinical settings; (2) a large but ing conditions: systolic blood pressure less rapidly reversible increase of serum amino- than 80 mmHg, pulse pressure less than 20 transferases levels; (3) exclusion of other fac- mmHg, 30% reduction of systolic blood pres- tors of damaging liver function. The diagnosis sure when compared with the primary systolic of ischemic hepatitis largely depends on the blood pressure among the patients with prima- liver biochemical tests in clinical practice. In ry hypertension. The exclusion criteria included our study, the diagnosis of ischemic hepatitis previously known history of hepatic dysfunc- was performed according to clinical criteria and tion, acute hepatobiliary diseases or the opera- clinical experience. All the patients were tion history of hepatobiliary diseases in recent dynamically performed with hepatobiliary bio- times, past history of chronic liver benign or chemical tests two times within 24 h on admis- malignant diseases, traumatic history of liver sion, and the diagnosis of ischemic hepatitis during this disease course, other factors lead- was made according to condition that the two ing to liver dysfunction (drug abuse, drinking, tests both showed abnormal live function. We poison, hepatotoxic , etc.). defined abnormal live function as serum ALT higher than 40 U/L, AST higher than 50 U/L, All the patients were divided into four groups ALP higher than 150 U/L, γ-GT higher than 50 according to the shock types: allergic shock, U/L, according to the criteria of the machine hypovolemic shock, septic shock, and cardio- Beckman DXC800 (Beckman Coulter Inc., genic shock. Kraemer Boulevard Brea, CA, USA) used in our For experiments involving human subjects, hospital. study approval was obtained from the institu- tional review board of our hospital ethics com- Statistical analysis mittee and the study was performed according to the Declaration of Helsinki. All the data were analyzed using SPSS version 12.0 (SPSS Inc., Chicago, IL, USA). Continuous Data collection data are presented as mean ± SD. Categorical variables are presented as numbers and per- The baseline data, including age, sex, smoking centages. A chi-square analysis or Fisher exact and drinking history, -taking history, test was used for categorical variables.

16671 Int J Clin Exp Med 2015;8(9):16670-16675 Ischemic hepatitis in different type shocks

Table 1. Clinical features of ischemic hepatitis in Comparison of the effects of shock on hepatic the shock with four different types enzymes and biliary enzymes Hepatic aP- Shock type Cases damage value Of all the shock patients, it was found that the Allergic shock 24 (7.3%) 18 (75.0%) < 0.05 incidence of abnormal hepatic enzymes levels was significantly higher than that of abnormal Hypovolemic shock 83 (25.3%) 41 (49.4%) biliary enzymes levels, suggesting that eleva- Septic shock 136 (41.5%) 119 (87.5%) tion of hepatic enzymes was more sensitive in Cardiogenic shock 85 (25.9%) 64 (75.3%) the diagnosis of ischemic hepatitis caused by a represents the significant difference P( < 0.05) when com- shock (P < 0.01). For different types of shock, paring any two of the four groups, with the exception of the the same result was found in allergic shock, allergic shock and cardiogenic shock. cardiogenic shock and septic shock, with exception of hypovolemic shock (P > 0.05) Correlation analysis was performed using (Table 2). Spearman rank-correlation test. Differences were considered statistically significant if the P The sensitivity analysis of AST and ALT in the terms of hepatic damage value was less than 0.05.

Results Among all the patients, our results showed that incidence of abnormal ALT level was significant- Patient characteristics ly higher than that of abnormal ASTlevel, sug- gesting that abnormal elevation of ALT was Of all the 328 shock patients, the mean age more sensitive in the diagnosis of ischemic was 65.84 ± 15.21 years old, with an age range hepatitis caused by shock when comparing of 15-94 years old. There were 200 males and with ALT (P < 0.01). In the terms of shock type, 128 females included in our study. Among all it was found that abnormal elevation of ALT was more sensitive in the diagnosis of ischemic the cases, 242 (73.8%) patients developed hepatitis caused by hypovolemic and septic ischemic hepatitis. The total in-hospital mortal- shock, with the exceptions of allergic and car- ity in shock patients was 23.8% (78/328). The diogenic shock (Table 2). mortality of shock patients combined with isch- emic hepatitis was 26.0% (63/242). It was Correlation analysis ofage, CRP and BNP with found that the mortality of shock patients com- hepatic damage bined with ischemic hepatitis was significantly higher than the total mortality of the whole An inverse correlation was noted between age shock patients (26.0% vs 23.8%, P < 0.05). and ALT (r = -0.28, P = 0.04), TB (r = -0.31, P = 0.02), AST (r = -0.02, P = 0.44), ALP (r = -0.80, Distribution of ischemic hepatitis in the four P < 0.01) and γ-GT (r = -0.41,P = 0.03) levels, different type shocks while a positive correlation was noted between age and LDH (r = 0.22, P = 0.07) (Table 3). Among all the 328 cases, there were 24 cases (7.3%) with allergic shock, 83 cases (25.3%) CRP was considered as the traditional marker with hypovolemic shock, 136 cases (41.5%) of sepsis. In our study, it was found that there with septic shock, and 85 cases (25.9%) with was a positive correlation between serum CRP cardiogenic shock. For the relationship of shock level and ALT (r = 0.96, P < 0.01), AST (r = 0.83, type and incidence of hepatic damage, our P < 0.01), LDH (r = 0.69, P < 0.01), TB (r = 0.66, study results showed that the incidence of P = 0.02), ALP (r = 1.00, P < 0.01) and γ-GT (r = hepatic damage was the highest in the septic 0.97, P < 0.01) levels (Table 4). shock, while the hypovolemic shock presented BNP was also regarded as an indicator of heart with the lowest incidence of hepatic damage. failure. Our study results demonstrated that The statistical results showed that there were there was a positive correlation between serum significant differences P( < 0.05) between two BNP level and ALT (r = 0.76, P < 0.01), AST (r = groups among the four groups, with the excep- 0.85, P < 0.01), LDH (r = 0.68, P < 0.01), TB (r = tion of the comparison of the allergic shock and 0.26, P < 0.01), ALP (r = 0.85, P < 0.01) and cardiogenic shock (P > 0.05) (Table 1). γ-GT (r = 0.77, P = 0.03) levels (Table 5).

16672 Int J Clin Exp Med 2015;8(9):16670-16675 Ischemic hepatitis in different type shocks

Table 2. Relevant sensitive analyses in the terms of hepatic damage Biochemical test type Hepatic enzyme type Items P value P value Hepatic enzyme Billary enzyme ALT AST Total shock 202 (61.6) 83 (25.2) < 0.01 98 (63.7) 62 (50.8) < 0.05 Allergic shock 12 (60.0) 7 (35.0) < 0.01 12 ( 60.0) 12 (60.0) > 0.05 Hypovolemic shock 5 (16.7) 15 (16.7) > 0.05 5 ( 63.3) 25 (16.7) < 0.01 Septic shock 36 (60.0) 18 (30.0) < 0.01 36 (60.0) 52 (55.0) < 0.05 Cardiogenic shock 8 (80.0) 1 (10.0) < 0.01 8 (80.0) 8 (80.0) > 0.05 The values are expressed as n (%). ALT, alanine aminotransferanse; AST, aspartate aminotransferanse.

Table 3. Correlation analysis of age with the The pathogenesis of ischemic hepatitis is con- levels of liver function test in the total shock sidered to be multifactorial and poorly under- patients stood. The main mechanism was the acute Enzymes r P-value reduction of hepatic blood flow, leading to the hypoxia or anoxia of hepatic cells [8]. Previous ALT -0.28 0.04 studies have reported that the shock with dif- AST -0.02 0.44 ferent types usually presents with lower cardiac LDH 0.22 0.07 output, leading to the reduction of hepatic TB -0.31 0.02 blood flow [16-18]. Besides, there may be other ALP -0.80 < 0.01 potential mechanisms, which were greatly γ-GT -0.41 0.03 associated with the variable causes of isch- ALT, alanine aminotransferanse; AST, aspartate emic hepatitis. Henrion et al. reported that con- aminotransferanse; LDH, lactate dehydrogenase; gestive could present with the TB, total bilirubin; ALP, alkaline phosphatase; γ-GT, increased central venous pressure, leading to γ-glutamyltransferase. the formation of congestive liver, which may result in the hepatic injury [10]. In septic shock, Discussion the pathologic mechanism is more complex. Increased needs of oxygen in liver and the Many studies have suggested that ischemic inability of liver to utilize oxygen contributed hepatitis is a common complication of shock, more to the hepatic injury [10]. This may well associated with a high mortality and poor clini- explained the reason why the incidence of isch- cal outcome [12, 15]. The clinical management emic hepatitis was the highest in the patients of ischemic hepatitis due to shock remains still with septic shock patients in our study (Table challenging. Our retrospective study in a single 1). center showed that the incidence of ischemic hepatitis in shock patients was 73.8%, and the Raurich et al. has reported that the incidence of presence of ischemic hepatitis significantly ischemic hepatitis in the patients with septic increased the mortality of shock patients (26% shock was 13.8% [12]. In contrast, our results vs 23.8%, P < 0.05). The incidence of ischemic showed much higher incidence of ischemic hepatitis was the highest in the septic shock, hepatitis in the patients with septic shock while the hypovolemic shock presented with (87.5%), which was the highest among the four the lowest incidence of ischemic hepatitis. different types of shock (Table 1). The possible During the development of liver damage caused reason may be due to the sample size and by shock, the sensitivity of the hepatic enzymes regional difference, because our study was a elevation was higher than that of biliary single-center study. In addition, it was also enzymes, with the exception of hypovolemic found that the incidence of ischemic hepatitis shock. Among the two types of hepatic caused by shock was 73.8%, suggesting that enzymes, the sensitivity of ALT elevation was shock is the main cause of ischemic hepatitis higher than that of AST. In addition, it was also in clinical practice [4]. It has been reported that found that CRP was positively correlated with the mortality of ischemic hepatitiswas 25-73% the levels of liver function parameters in the [6, 10]. In accordance with the lower limit of the septic shock patients and BNP was positively above range, it was found that the mortality of correlated with that in the cardiogenic shock shock patients combined with ischemic hepati- patients. tis was 26.0%. In addition, our results also

16673 Int J Clin Exp Med 2015;8(9):16670-16675 Ischemic hepatitis in different type shocks

Table 4. Correlation analysis of CRP with the CRP is an acute phase protein, which is mainly levels of liver function test in the septic shock used as a marker of inflammation or sepsis patients [21, 22]. In our study, our study results demon- Enzymes r P-value strated that serum CRP level was positively cor- ALT 0.96 < 0.01 related with the levels of all the liver function AST 0.83 < 0.01 parameters with a significant difference P ( < Table 4 LDH 0.69 0.01 0.05) ( ), suggesting that CRP may be useful in the prediction of the progress of isch- TB 0.66 0.02 emic hepatitis in septic shock patients. ALP 1.00 < 0.01 Similarly, BNP is considered as a marker for γ-GT 0.97 < 0.01 heart failure in clinical practice, which was use- CRP, C-reactive protein; ALT, alanine aminotransferanse; ful in the screening and prognosis of heart fail- AST, aspartate aminotransferanse; LDH, lactate dehydro- genase; TB, total bilirubin; ALP, alkaline phosphatase; ure [23-25]. In the patients of ischemic hepati- γ-GT, γ-glutamyltransferase. tis in cardiogenic shock, our data showed that there was a significantly positive correlation between serum BNP level and the levels of liver Table 5. Correlation analysis of BNP with the function testing parameters (Table 5), suggest- levels of liver function test in the cardiogenic ing that BNP might be a beneficial predictor in shock patients the risk evaluation of the progress of ischemic Enzymes r P-value hepatitis. ALT 0.76 < 0.01 AST 0.85 < 0.01 There were some limitations of our study, of which we were aware. For one thing, the study LDH 0.68 < 0.01 was limited by the nature of the retrospective TB 0.26 < 0.01 study. For another, our study was also limited ALP 0.85 < 0.01 by the small number of patients and the single- γ-GT 0.77 < 0.01 center study. Thus, a study of large population BNP, brain natriuretic peptide; ALT, alanine aminotrans- and multiple centers is of great need to further feranse; AST, aspartate aminotransferanse; LDH, lactate dehydrogenase; TB, total bilirubin; ALP, alkaline phospha- explore the potential pathogenesis and clinical tase; γ-GT, γ-glutamyltransferase. features of ischemic hepatitis caused by shock for better clinical management. showed that the mortality of shock patients To sum up, ischemic hepatitis is a common combined with ischemic hepatitis was signifi- complication of shock, increasing the mortality cantly higher than the total mortality of the of shock patients. Of all the four different types whole shock patients. This result was consis- of shock, the septic shock is the most common tent with many previous reports showing that cause of hepatic damage. ALT is a more sensi- ischemic hepatitis is a risk factor for the mor- tive parameter for the diagnosis of hepatic dys- tality of critically ill patients [7, 12, 19]. function caused by shock. CRP may be a useful predictor for septic shock, while BNP may be a It is generally acknowledged that the diagnosis useful predictor for cardiogenic shock. of ischemic hepatitis mainly depends on the relevant lab tests, such as ALT, AST, LDH, ALP, Acknowledgements TB, γ-GT, coagulation factor and albumin levels [20]. In our study, the main parameters of ALT, We would like to acknowledge the reviewers for AST, LDH, ALP, TB, and γ-GT were tested. Among their helpful comments on this paper. all the parameters, our results showed that the sensitivity of the abnormal hepatic enzyme Disclosure of conflict of interest (including ALT and AST) elevation was signifi- cantly higher than that of biliary enzyme (includ- None. ing ALP and γ-GT) in the diagnosis of hepatic damage caused by shock. In addition, among Address correspondence to: Dr. Chang-Qing Yang, the two types of hepatic enzymes, the sensitiv- Department of Gastroenterology, Tongji Hospital ity of ALT elevation was higher than that of AST Affiliated to Tongji University, No. 389, Xincun in the diagnosis of hepatic damage caused by Road, Shanghai 200065, P.R. China. Tel: +86- shock (Table 2). 13817802801; E-mail: [email protected]

16674 Int J Clin Exp Med 2015;8(9):16670-16675 Ischemic hepatitis in different type shocks

References [14] Rawson JS, Achord JL. Shock liver. South Med J 1985; 78: 1421-1425. [1] Seeto RK, Fenn B and Rockey DC. Ischemic [15] Maynard ND, Mason RC, Bihari DJ, Dalton RN, hepatitis: clinical presentation and pathogen- Beale R and Smithies MN. Liver function and esis. Am J Med 2000; 109: 109-113. splanchnic ischemia in critically III patients. [2] Buhac I and Bynum T. Hepatic centrilobular ne- CHEST J 1997; 111: 180-187. crosis in congestive heart failure. Digest Dis [16] Riddez L, Johnson L and Hahn RG. Central and Sci 1979; 24: 732-733. regional hemodynamics during crystalloid fluid [3] Nomura T, Keira N, Urakabe Y, Naito D, Na- after uncontrolled intra-abdominal kayama M, Kido A, Kanemasa H, Matsubara H bleeding. J Trauma 1998; 44: 433-439. and Tatsumi T. Chronic pericardial constriction [17] Ruokonen E, Takala J and Kari A. Regional induced severe ischemic hepatitis manifesting blood flow and oxygen transport in patients as hypoglycemic attack. Circ J 2009; 73: 183- with the low cardiac output syndrome after car- 186. diac . Crit Care Med 1993; 21: 1304- [4] Raurich JM, Llompart-Pou JA, Ferreruela M, Co- 1311. lomar A, Molina M, Royo C, Ayestarán I and [18] Spapen H. Liver perfusion in sepsis, septic Ibáñez J. Hypoxic hepatitis in critically ill pa- shock, and multiorgan failure. Anat Rec 2008; tients: incidence, etiology and risk factors for 291: 714-720. mortality. J Anesthesia 2011; 25: 50-56. [19] Fuhrmann V, Kneidinger N, Herkner H, Heinz [5] Miry Blich M, Shimon Edelstein M, Roy Man- G, Nikfardjam M, Bojic A, Schellongowski P, An- sano M and Yeouda Edoute M. Ischemic Hepa- germayr B, Kitzberger R and Warszawska J. titis Induced by Severe Anemia. Isr Med Assoc Hypoxic hepatitis: underlying conditions and J 2003; 5: 208-209. risk factors for mortality in critically ill patients. [6] Birrer R, Takuda Y and Takara T. Hypoxic hepa- Intensive Care Med 2009; 35: 1397-1405. topathy: pathophysiology and prognosis. In- [20] Strassburg CP. Shock liver. Best Pract Res Clin ternMed (Tokyo, Japan) 2006; 46: 1063-1070. Gastroenterol 2003; 17: 369-381. [7] Fuhrmann V, Kneidinger N, Herkner H, Heinz [21] Povoa P, Almeida E, Moreira P, Fernandes A, G, Nikfardjam M, Bojic A, Schellongowski P, An- Mealha R, Aragao A and Sabino H. C-reactive germayr B, Schöniger-Hekele M and Madl C. protein as an indicator of sepsis. Intensive Impact of hypoxic hepatitis on mortality in the Care Med 1998; 24: 1052-1056. intensive care unit. Intensive Care Med 2011; [22] Wiwanitkit V. C reactive protein in sepsis. J Vet 37: 1302-1310. Emerg Crit Care 2010; 20: 1-1. [8] Bynum T, Boitnott J and Maddrey W. Ischemic [23] Bhalla V, Willis S and Maisel AS. B-Type Natri- hepatitis. Dig Dis Sci 1979; 24: 129-135. uretic Peptide: The Level and the Drug-Part- [9] Bulkley GB, Oshima A and Bailey RW. Patho- ners in the Diagnosis and Management of Con- physiology of hepatic ischemia in cardiogenic gestive Heart Failure. Congest Heart Fail 2004; shock. AmJ Surg 1986; 151: 87-97. 10: 3-27. [10] Henrion J, Schapira M, Luwaert R, Colin L, [24] Tung RH, Garcia C, Morss AM, Pino RM, Fifer Delannoy A and Heller FR. Hypoxic hepatitis: MA, Thompson BT, Lewandrowski K, Lee-Le- clinical and hemodynamic study in 142 con- wandrowski E and Januzzi JL. Utility of B-type secutive cases. Medicine 2003; 82: 392-406. natriuretic peptide for the evaluation of inten- [11] Bailey RW, Brengman ML, Fuh KC, Hamilton sive care unit shock. Crit Care Med 2004; 32: SR, Herlong FH and Bulkley GB. Hemodynamic 1643-1647. pathogenesis of ischemic hepatic injury follow- [25] Hejmdal A, Boesgaard S, Lindholm MG and ing cardiogenic shock/. Shock Goetze JP. B-type natriuretic peptide and its 2000; 14: 451-459. molecular precursor in myocardial infarction [12] Raurich JM, Pérez O, Llompart-Pou JA, Ibáñez complicated by cardiogenic shock. J Card Fail J, Ayestarán I and Pérez-Bárcena J. Incidence 2007; 13: 184-188. and outcome of ischemic hepatitis complicat- ing septic shock. Hepatol Res 2009; 39: 700- 705. [13] Gitlin N and Serio KM. Ischemic hepatitis: wid- ening horizons. AmJ Gastroenterol 1992; 87: 831-836.

16675 Int J Clin Exp Med 2015;8(9):16670-16675