Advertisement

The Prognostic Significance of the Platelet Count in Alcoholic Hepatitis

Open AccessPublished:August 29, 2022DOI:https://doi.org/10.1016/j.gastha.2022.07.022

      Background and Aims

      Thrombocytopenia is present in up to 76% of patients with chronic liver disease, and lower platelet counts (PCs) are associated with greater severity of portal hypertension. In this study, we assess the relationship of PC in patients with a clinical diagnosis of severe alcoholic hepatitis (SAH) with clinical severity and response to corticosteroid (CS) therapy.

      Methods

      Clinical characteristics, treatment, and hospital outcomes for patients admitted with SAH were analyzed from an electronic health record system. Patients were categorized based on admission PC (k/uL) into 5 categories: <50, 50–99, 100–149, 150–199, and ≥200. Frequency of complications (acute kidney injury, ascites, and hepatic encephalopathy), length of stay, and admission to an intensive care unit were analyzed across PC categories. Characteristics of patients who did and did not receive at least 4 days of CS therapy were compared.

      Results

      Among 159 patients, 15 (9.4%) were in the PC < 50 category, 42 (26.4%) in PC 50-99, 51 (32%) in PC 100–149, 23 (14.5%) in PC 150–199, and 28 (17.6%) in PC ≥ 200. A higher admission PC was associated with greater white blood cell count, absolute neutrophil count, and total bilirubin (P < .05). Patients with higher PC on admission were more likely to receive steroids. PC was inversely associated with Lille score at treatment day 4 (P < .05).

      Conclusion

      A higher PC in SAH was associated with a greater inflammatory response and total bilirubin. Patients with a higher PC were more likely to receive CS and have a favorable treatment response.

      Keywords

      Abbreviations used in this paper:

      AH (alcoholic hepatitis), ANC (absolute neutrophil count), CP (Child-Turcotte-Pugh), CS (corticosteroid), DF (Maddrey discriminant function), HD (hospital day), IL-6 (interleukin-6), INR (international normalized ratio), MELD (Model for End-Stage Liver Disease), NST (no specific treatment), PC (platelet count), SAH (severe alcoholic hepatitis), SI (splenic index), TBili (total serum bilirubin level), TCP (thrombocytopenia), TD (treatment day), TPO (thrombopoietin), US (United States), WBC (White Blood Cell)

      Introduction

      Alcoholic hepatitis (AH) develops in up to 35% of patients with heavy, sustained alcohol use
      • Basra S.
      • Anand B.S.
      Definition, epidemiology and magnitude of alcoholic hepatitis.
      and is a cause of substantial morbidity and mortality. It accounts for approximately 18.4% of alcohol-associated liver disease hospitalizations in the United States (US)
      • Shirazi F.
      • Singal A.K.
      • Wong R.J.
      Alcohol-associated cirrhosis and alcoholic hepatitis hospitalization trends in the United States.
      and was estimated in 2010 to be responsible for 1% of all US hospital admissions.
      • Thompson J.A.
      • Martinson N.
      • Martinson M.
      Mortality and costs associated with alcoholic hepatitis: a claims analysis of a commercially insured population.
      Frequent emergency department visits, inpatient admissions, and high mortality further contribute to health-care-related costs, with total expenditure exceeding an estimated $1.5 billion for insured patients in the US in 2013.
      • Thompson J.A.
      • Martinson N.
      • Martinson M.
      Mortality and costs associated with alcoholic hepatitis: a claims analysis of a commercially insured population.
      This estimate grossly underestimates the current economic and social burden of AH which has been steadily increasing annually and rose dramatically during the COVID-19 pandemic.
      • Moon A.M.
      • Curtis B.
      • Mandrekar P.
      • et al.
      Alcohol-associated liver disease before and after COVID-19-an overview and call for ongoing investigation.
      AH presents with a wide clinical spectrum of disease severity, ranging from asymptomatic cases to severe liver failure with jaundice, fever, malaise, and tender hepatomegaly.
      • Crabb D.W.
      • Im G.Y.
      • Szabo G.
      • et al.
      Diagnosis and treatment of alcohol-associated liver diseases: 2019 practice guidance from the American association for the study of liver diseases.
      Patients with mild disease have a favorable outcome. Although prolonged abstinence can lead to marked improvement in liver function among those with severe AH (SAH), severe disease is associated with 1- and 6-month mortality approaching 30% and 40%–70%, respectively.
      • Thompson J.A.
      • Martinson N.
      • Martinson M.
      Mortality and costs associated with alcoholic hepatitis: a claims analysis of a commercially insured population.
      Prognostic models for AH that are employed for therapeutic decisions include the total serum bilirubin level (TBili) and international normalized ratio (INR). Severity is mostly commonly defined by the Maddrey discriminant function (DF) and more recently by the Model for End-Stage Liver Disease (MELD). The mainstay of treatment is corticosteroids (CS), and in select cases, they significantly reduce short-term mortality but also carry the risk of adverse events such as infection, sepsis, and gastrointestinal bleeding.
      • Thursz M.R.
      • Richardson P.
      • Allison M.
      • et al.
      Prednisolone or pentoxifylline for alcoholic hepatitis.
      Thrombocytopenia (TCP) is the most common laboratory abnormality in patients with chronic liver disease and is present in up to 76% of patients.
      • Mitchell O.
      • Feldman N.
      • Diakow M.
      • et al.
      The pathophysiology of thrombocytopenia in chronic liver disease.
      Multiple pathophysiologic processes lead to TCP in the cirrhotic patient, the most important of which are reduced platelet production due to impaired hepatic thrombopoietin (TPO) production and splenic sequestration due to portal hypertension. In patients with AH, the platelet count (PC) can also be affected by the acute effects of alcohol with direct impairment of megakaryopoiesis, platelet toxicity, and accelerated platelet apoptosis.
      • Mitchell O.
      • Feldman N.
      • Diakow M.
      • et al.
      The pathophysiology of thrombocytopenia in chronic liver disease.
      A low PC is an important prognostic marker in patients with chronic liver disease. There is an inverse relationship between PC and severity of fibrosis, portal hypertension, and prognosis. Adding the PC to the albumin-bilirubin grade is more accurate than the Child-Turcotte-Pugh (CP) score in predicting survival of patients with hepatocellular carcinoma undergoing resection.
      • Roayaie S.
      • Jibara G.
      • Berhane S.
      • et al.
      PALBI-an objective score based on platelets, albumin & bilirubin stratifies HCC patients undergoing resection & ablation better than Child’s classification.
      A low PC is also an independent risk factor for varices and may be used to predict the presence of varices with a high bleeding risk.
      • Thomopoulos K.C.
      • Labropoulou-Karatza C.
      • Mimidis K.P.
      • et al.
      Non-invasive predictors of the presence of large oesophageal varices in patients with cirrhosis.
      To date, there is only limited information on the PC in patients presenting with AH. In this study, we assess the spectrum of platelet counts in patients hospitalized with a clinical diagnosis of SAH and correlate its level with clinical severity and response to CS therapy.

      Methods

      Subjects

      A single-center, retrospective cohort study was conducted on patients with a discharge diagnosis of alcoholic liver disease (International Classification of Diseases, 9th or 10th revision, clinical modification codes K70.10, K70.11, K70.40, K70.41, 571.5, 572.52, 572.8, 789.59) who were admitted from January 2016 to June 2020. Data were extracted from an electronic health record system (EPIC) using Clinical Looking Glass, (Emerging Health Information Technology, Yonkers, NY). EPIC is a computerized patient database that contains comprehensive data, including patient demographics, hospitalizations, discharge diagnoses (International Classification of Diseases codes), laboratory and imaging results, histopathology, endoscopic and surgical procedures, and medications. Clinical Looking Glass is a proprietary software program that permits exploration of the database contained within EPIC.
      Eligible patients satisfied SAH criteria including age ≥18 years, documentation of ongoing heavy alcohol consumption, onset of jaundice within prior 8 weeks, <60 days of abstinent before onset of jaundice, serum total bilirubin ≥3.0 mg/dL, aspartate aminotransferase ≥50, aspartate aminotransferase/alanine aminotransferase ≥1.5, and both values <400 IU/L for enrollment into a clinical trial.
      • Crabb D.W.
      • Bataller R.
      • Chalasani N.P.
      • et al.
      NIAAA Alcoholic Hepatitis Consortia. Standard definitions and common data elements for clinical trials in patients with alcoholic hepatitis: recommendation from the NIAAA alcoholic hepatitis Consortia.
      Exclusion criteria included liver transplantation or other underlying liver disease, absence of documentation of ongoing heavy alcohol consumption, hepatocellular carcinoma or other active malignancies, biliary obstruction, clinically significant upper gastrointestinal bleeding, HIV infection, pre-existing hemodialysis, kidney transplantation, kidney injury with serum creatinine >2.5 mg/dL, sepsis, other serious medical conditions that would significantly alter the hospital course, active substance abuse, patients who were transferred from outside facilities, and absence of laboratory criteria. Only the first index case was analyzed for those with more than 1 admission.

      Admission PC Category and Clinical Characteristics

      The study cohort was stratified into 5 admission PC categories (k/uL): <50, 50–99, 100–149, 150–199, and ≥200.
      • Brown Jr., R.S.
      Current management of thrombocytopenia in chronic liver disease.
      PC categories were based on standard clinical definitions of mild, moderate, and severe TCP. PCs within the normal range were divided into 2 categories to maintain a standardized PC distribution of 50 cells per microliter. Clinical characteristics of subjects were recorded, and DF, MELD, Na-MELD, and CP scores were calculated. Clinical characteristics between the various PC categories were then compared.

      Clinical Characteristics Based on PC Category and HD4 Admission Status

      Hospital status at day 4 after admission (HD4) was recorded as discharge before vs at HD4 or later. For those discharged before HD4, the reason for discharge was recorded as improvement, discharge against medical advice, or death. Clinical characteristics for patients based on PC category and hospital status at HD4 were compared.

      Baseline Characteristics and Outcomes Based on PC Category and SAH Treatment

      Administration of specific treatment for SAH (no specific treatment [NST], CS, granulocyte colony-stimulating factor, or pentoxifylline) during the first 4 hospital days was recorded. For patients receiving NST, baseline and HD4 laboratory test results were recorded, and DF, MELD, Na-MELD, and 4-day Lille (observation) scores were calculated.
      • Foncea C.G.
      • et al.
      Day-4 Lille score is a good prognostic factor and early predictor in assessing therapy response in patients with liver cirrhosis and severe alcoholic hepatitis.
      Differences in changes in the parameters were compared between the various PC categories. For patients who received CS for at least 4 days, liver tests at the start (TD0) and at treatment day 4 (TD4) were recorded, and 4-day Lille (CS treatment) scores calculated.

      Estimation of Spleen Size

      Spleen size was assessed by calculating the splenic index (SI) on abdominal computed tomography or magnetic resonance imaging obtained during hospitalization.
      • Kucybała I.
      • Ciuk S.
      • Tęczar J.
      Spleen enlargement assessment using computed tomography: which coefficient correlates the strongest with the real volume of the spleen?.
      Abdominal ultrasound was used to determine SI if cross-sectional imaging was not available. Analysis was performed controlling for height using the SI-to-height ratio. SI was analyzed by correlation with PC category at admission and gender.

      Statistical Analysis

      Comparison of PC category characteristics at admission was performed. Hospital outcomes were analyzed based on admission PC, PC category, and treatment provided. Demographic and clinical characteristics of the study cohort were tabulated as means and standard deviations or counts and percentages. The distribution of categorical variables was compared using Spearman’s chi-squared test or Fisher’s exact test. Continuous variables were compared using the analysis of variance. A linear regression model was used to examine the association between continuous day-4 Lille score and the continuous PC. A logistic regression model was used to examine day-4 Lille score and its association with the continuous PC. Statistical significance was set at P value <.05. All analyses were performed using SPSS (IBM).

      Results

      Subjects

      Records of 906 patients with a diagnosis of alcoholic liver disease were evaluated. One hundred fifty-nine (17.5%) patients satisfied criteria for SAH after excluding 116 (12.8%) for liver transplantation or other underlying liver disease; 82 (9.1%) for absence of documentation of ongoing heavy alcohol consumption; 85 (9.4%) for hepatocellular carcinoma or other active malignancies; 31 (3.4%) for biliary obstruction; 75 (8.3%) for clinically significant upper gastrointestinal bleeding; 16 (1.8%) for HIV infection; 9 (1.0%) for pre-existing hemodialysis, kidney transplantation, or kidney injury with serum creatinine >2.5 mg/dL; 69 (7.6%) for sepsis; 34 (3.8%) for other serious medical conditions that would significantly alter the hospital course; 3 (0.3%) for active substance abuse; 94 (10.4%) patients who were transferred from outside facilities, and 133 (14.7%) due to absence of laboratory criteria. All patients included in the study were admitted to the hospital with alcoholic liver disease as the major diagnosis and met criteria for SAH.

      Admission PC Category and Clinical Characteristics

      Fifteen patients (9.4%) were in PC <50 category, 42 (26.4%) in the 50–99, 51 (32.1%) in the 100–149, 23 (14.5%) in the 150–199, and 28 (17.6%) in the ≥200 category. Increasing PC category was associated with greater white blood cell count (WBC) (P < .0001), greater absolute neutrophil count (ANC) (P < .0001), greater TBili (P = .0054), and higher Na-MELD scores (P = .0446). Admission diagnoses are presented in Table 1. There was no significant association between PC category and admission diagnoses.
      Table 1Admission Platelet Categories and Clinical Characteristics
      Characteristic mean (SD)PC < 50 (N = 15)PC 50–99 (N = 42)PC 100–149 (N = 51)PC 150–199 (N = 23)PC ≥ 200 (N = 28)P value
      Age (y)51.3 (8.3)51.8 (10.9)52.1 (10.8)48.6 (.9)45.5 (10.9).0656
      Male (%)80.0% (12)69.0% (29)56.9% (29)65.2% (15)64.3% (18).3086
      PC (k/uL)38.5 (8.9)74.4 (12.3)122.8 (14.4)171.6 (18.2)283.4 (77.2)-
      Hgb (k/uL)11.8 (2.5)10.2 (3.4)10.5 (3.4)10.1 (4.7)10.4 (7.0).1666
      WBC (k/uL)5.3 (2.6)6.7 (3.4)8.2 (3.4)11.7 (4.7)17 (7.0)<.0001
      ANC (k/uL)3.8 (2.5)4.7 (3.2)5.9 (3.2)9.0 (4.6)13.6 (6.7)<.0001
      Na (mEq/L)135.7 (5.0)135.5 (5.2)134.8 (4.8)132.7 (6.0)133 (4.2).0919
      BUN (mg/dL)9.2 (4.2)11.9 (11.1)12.1 (6.8)12.4 (10.2)13.2 (7.3).7085
      Cr (mg/dL)0.75 (0.28)0.86 (0.43)0.83 (0.41)0.87 (0.35)0.90 (0.53).8227
      Alb (g/dL)3.2 (0.62)2.9 (0.71)2.9 (0.57)2.8 (0.44)2.9 (0.63).4380
      TBili (mg/dL)7.6 (6.5)7.9 (5.0)9.2 (6.9)10.9 (6.3)13.8 (9.3).0054
      ALT (U/L)44.8 (13.2)45.6 (34.5)60.7 (40.7)56.6 (51.1)59.8 (30.8).2553
      AST (U/L)166.3 (68.4)151.4 (88.4)185.9 (96.5)174.4 (104.3)201.2 (92.3).2207
      PT (s)18.8 (6.9)17.4 (3.7)16.4 (3.9)17.3 (6.8)17.9 (5.2).4723
      INR1.8 (0.64)1.7 (0.37)1.6 (0.41)1.7 (0.59)1.7 (0.55).6188
      DF38.7 (36.5)30.9 (22.4)29.4 (21.9)35.5 (34.7)41.1 (30.0).3591
      MELD19.5 (5.0)20.2 (4.5)19.6 (4.6)20.9 (5.5)22.3 (6.6).2312
      MELD-Na20.5 (5.9)21.8 (4.8)21.6 (4.8)23.9 (5.7)24.6 (6.4).0446
      Child-Pugh9.5 (2.2)9.9 (2.2)10 (1.8)9.7 (1.6)10.2 (1.9).8370
      AKI/HRS6.7% (1)16.7% (7)11.8% (6)13% (3)17.9% (5).51
      Ascites, mild53.3% (8)61.9% (26)64.7% (33)70% (16)60.7% (17).74
      Ascites, moderate/large6.7% (1)23.8% (10)29.4% (15)4.3% (1)14.3% (4).25
      HE, mild13.3% (2)14.3% (6)11.8% (6)8.7% (2)21.4% (6).38
      HE, overt20.0% (3)14.3% (6)11.8% (6)0.0% (0)10.7% (3).17
      Values represented as mean (± SD) or % (count). Significant difference between baseline characteristic and PC, category on ANOVA, and chi-squared (for continuous and categorical variables, respectively) (<.05) are represented by P values in bold.
      AKI/HRS, acute kidney injury or hepatorenal syndrome; Alb, albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; BUN, blood urea nitrogen; Cr, creatinine; HE, hepatic encephalopathy; Hgb, hemoglobin; PT, prothrombin time; SD, standard deviation.

      Clinical Characteristics Based on PC Category and HD4 Status

      Nineteen patients were discharged prior to HD4 (10%). Sixteen patients were discharged due to improvement, and 3 patients left against medical advice. Baseline characteristics of patients discharged before vs ≥HD4 are presented in Table 2. Patients with length of stay <HD4 had lower WBC, ANC, TBili, prothrombin time, and INR levels and lower DF, MELD, Na-MELD, and CP scores.
      Table 2Patient Admission Characteristics, Without Specific Therapy, Stratified by LOS
      Characteristic, mean (SD)LOS < 4 (n = 19)LOS ≥ 4 (n = 116)P value
      Age (y)51.4 (11.6)51.1 (10.2).916
      Male73.7% (14)64.6% (75).432
      PC (k/uL)143.5 (60)129.0 (81.82).377
      Hgb (k/uL)10.7 (1.9)19.0 (2.36).551
      WBC (k/uL)7.1 (2.4)9.0 (5.35).006
      ANC (k/uL)4.6 (1.9)7.0 (4.861)<.0005
      Na (mEq/L)135.0 (5)135.0 (5.366).585
      BUN (mg/dL)11.0 (5)12.0 (9.41).326
      Cr (mg/dL)0.83 (0.18)1.00 (0.45).325
      Alb (g/dL)3.1 (0.6)3.0 (0.643).115
      TBili (mg/dL)5.7 (2.2)9.0 (6.9)<.0005
      ALT (U/L)76.0 (51)52.0 (36.28).069
      AST (U/L)188.0 (97)170.0 (94.1).446
      PT (s)14.5 (2.3)17.0 (4.95)<.0005
      INR1.4 (0.2)2.0 (0.472)<.0005
      DF16.0 (18.11)31.6 (26.26)<.0005
      MELD16.0 (2)20.0 (5.08)<.0005
      MELD-Na18.2 (4)22.0 (5.57)<.0005
      Child-Pugh8.6 (1.5)10.0 (1.96)<.0005
      T-test for unequal variance. Values presented as mean (SD). Significant values are presented in bold (P < .05).
      Alb, albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; BUN, blood urea nitrogen; Cr, creatinine; Hgb, hemoglobin; LOS, length of stay; PT, prothrombin time; SD, standard deviation.

      Baseline Characteristics and Outcomes Based on PC Category and AH Treatment

      One hundred and thirty-five patients received NST during the first 4 hospital days, 24 patients received CS, 2 granulocyte colony-stimulating factors, and 1 pentoxifylline. Baseline and HD4 assessments for patients who received NST are presented in Table 3. Increased WBC, ANC, and TBili correlated with increased PC category at HD0. PC increased from HD0 to HD4 in the PC <50 and 50–99 categories and decreased in the PC 100-149, 150–199, and ≥200 categories, but the changes were only significant in the PC 100-149 and 150–199 categories. ANC decreased from HD0 to HD4 in all PC categories but was only significant in the PC <50 category. Alanine aminotransferase and aspartate transaminase levels decreased in all categories. There were no significant changes in DF, MELD, or Na-MELD scores between HD0 and HD4.
      Table 3Baseline and HD4 Assessment for Patients With LOS ≥4 Receiving NST
      Characteristic, mean (SD)PC < 50 (n = 9)PC 50–99 (n = 30)PC 100–149 (n = 28)PC 150–199 (n = 10)PC 200 (n = 25)P value across PC category at HD0 (ANOVA)
      HD0HD4HD0HD4HD0HD4HD0HD4HD0HD4
      Age (y)50.6 (8.9)50.6 (11.5)53.7 (11.3)53.7 (10.2)41.2 (8)
      Male (%)77.8 % (7)73.3% (22)57.1% (16)50.0% (5)52.0% (13)-
      PC (k/uL)37.0 (9)55.0 (28)73.0 (12)78.0 (20)126.0 (14)110.0 (38)∗179.0 (19)141.0 (58)∗223.0 (114)214.0 (177.5)-
      Hgb (k/uL)12.7 (2.5)11.4 (2.3)10.2(2.6)9.8 (2.1)10.5 (2.1)9.9 (1.5)10.3 (2.7)9.8 (2.5)10.0 (1.6)9.6 (3).015
      WBC (k/uL)5.2 (3.2)5.0 (3.2)6.6 (3.8)5.8 (2.3)8.5 (4)8.3 (3.5)11.3 (4.7)10.2 (6.3)15.0 (5.9)13.9 (7.3)<.0005
      ANC (k/uL)3.7 (3.1)3.2 (2.7)∗4.6 (3.5)3.8 (1.9)6.4 (4)5.7 (3.2)9.0 (4.2)8.0 (5.6)12.0 (5.4)10.7 (6.6)<.0005
      Na (mEq/L)137.0 (3)138.0 (3)135.0 (5)136.0 (4)134.0 (6)136.0 (5)∗131.0 (6)133.0 (6)133.0 (5.8)135.2 (5.1)∗.053
      BUN (mg/dL)8.0 (3)9.0 (4)13.0 (13)12.0 (6)12.0 (6)10.0 (4)16.0 (14)12.0 (10)12.0 (5.9)9.9 (5.9).612
      Cr (mg/dL)0.64 (0.17)0.60 (0.15)0.90 (0.46)0.87 (0.4)0.89 (0.49)0.76 (0.27)0.94 (0.41)0.84 (0.45)1.00 (0.46)0.70 (0.28)∗.928
      Alb (g/dL)3.4 (0.5)3.0 (0.5)∗2.9 (0.7)2.7 (0.7)∗2.8 (0.5)2.5 (0.3)∗3.0 (0.2)2.8 (0.4)∗3.0 (0.72)2.6 (0.65)∗.703
      TBili (mg/dL)5.7 (2.4)4.6 (1.5)6.4 (4.8)7.5 (5.3)7.8 (3.9)8.8 (5.3)8.3 (6.2)8.2 (6.3)13.0 (8.1)14.1 (8.2)0.047
      ALT (U/L)46.0 (15)35.0 (14)∗49.0 (40)40.0 (22)∗54.0 (38)47.0 (33)∗67.0 (64)57.0 (46)55.0 (29.7)44.0 (23.9)∗.783
      AST (U/L)156.0 (74)105.0 (50)∗143.0 (94)90.0 (46)∗177.0 (100)124.0 (70)∗181.0 (121)123.0 (75)∗197.0 (100)134.0 (50.5)∗.258
      PT (s)16.6 (2.3)17.8 (3.8)17.0 (3.4)17.7 (5.3)15.5 (3.2)18.3 (4.3)∗13.0 (4.7)15.9 (3.8)20.0 (7.2)21.0 (6.5).171
      INR1.6 (0.2)1.7 (0.4)1.7 (0.4)1.7 (0.5)1.5 (0.3)1.8 (0.4)∗1.4 (0.2)1.5 (0.4)2.0 (0.72)2.0 (0.69).0965
      DF26.9 (6.3)31.0 (13)28.0 (16)35.0 (30)23.6 (13.9)35.0 (19)19.0 (12)28.0 (27)44.8 (29.6)59.6 (33.4).594
      MELD18.0 (2)19.0 (2)20.0 (5)19.0 (3)20.0 (4)20.0 (4)20.0 (4)18.0 (3)24.0 (5.9)24.0 (6.2).413
      MELD-Na19.0 (3)18.0 (4)21.7 (5.6)17.1 (2.9)21.0 (5)18.0 (5)22.7 (5.3)16.7 (2.8)29.0 (6.6)27.0 (6.9).152
      Values represented as mean (SD). Analysis performed through paired t-test within PC categories. Significant change from HD0 to HD4 (P < .05) represented with ∗. Statistically significant values on ANOVA are represented in bold.
      Alb, albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; BUN, blood urea nitrogen; Cr, creatinine; HD0, hospital day at admission; HD4, hospital day 4 after admission; Hgb, hemoglobin; PT, prothrombin time; SD, standard deviation.
      Thirty-six patients received at least 4 days of CS treatment during their hospital stay. Baseline characteristics of patients who received CS vs NST are presented in Table 4. Patients with a higher PC on admission were more likely to receive CS (173 vs 131, P = .08). Patients who received CS were younger; had greater WBC, ANC, TBili, prothrombin time, and INR levels; and had greater DF, MELD, and Na-MELD scores (Table 4). Assessments for the patients treated with CS on TD0 and TD4 are presented in Table 5. A higher PC was significantly associated with a lower day-4 Lille score (Table 6).
      Table 4Admission Characteristics, Stratified by Therapy
      Characteristic, mean (SD)No specific therapy (n = 135)Corticosteroids (n = 24)P value
      Age (y)51.2 (12)45.4 (10.32).03
      Male (%)65.958.3.473
      PC (k/uL)131.0 (79.09)173.0 (100).08
      Hgb (k/uL)10.5 (2.30)10.5 (2).88
      WBC (k/uL)9.0 (5.09)10.6 (8.2).03
      ANC (k/uL)7.6(4.63)9.1 (7.7).05
      Na (mEq/L)135.0 (5.26)134.0 (4).50
      BUN (mg/dL)12.0 (8.93)11.0 (7).40
      Cr (mg/dL)0.88 (0.42)0.72 (0.38).09
      Alb (g/dL)2.9 (0.64)2.8 (0.5).20
      TBili (mg/dL)8.8 (6.56)15.7 (7.5)<.0005
      ALT (U/L)56.0 (39.37)48.0 (30).31
      AST (U/L)172.0 (94.43)194.0 (86).27
      PT (s)16.7 (4.78)19.8 (5).008
      INR1.6 (0.46)1.9 (0.5).007
      DF28.8 (27.82)51.7 (27.16).0006
      MELD20.0 (4.96)24.0 (5).001
      MELD-Na22.0 (5.52)26.0 (5).0006
      Child-Pugh9.8 (1.96)10.4 (1.7).14
      T-test assuming unequal variance used for continuous variables. Significant values presented in bold.
      Alb, albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; BUN, blood urea nitrogen; Cr, creatinine; Hgb, hemoglobin; PT, prothrombin time; SD, standard deviation.
      Table 5Treatment Day 0 and Treatment Day 4 Characteristics in Patients Receiving CS for ≥4 days
      Characteristic, mean (SD)PC < 50 (n = 2)PC 50–99 (n = 9)PC 100–149 (n = 5)PC 150–199 (n = 10)PC ≥ 200 (n = 10)
      TD0TD4TD0TD4TD0TD4TD0TD4TD0TD4
      Age (y)62.0 (9.19)51.0 (13)49.8 (4)42.0 (10)46.7 (10)
      Male (%)100.033.360.07040.0
      PC (k/uL)46.0 (0.71)74.5 (14.85)84.0 (13.82)198.0∗ (24.13)292.0 (48.5)252.0 (87.9)174.0 (13.81)159.0 (39.42)292.0 (48.5)252.0 (87.9)
      Hgb (k/uL)10.0 (0.35)10.1 (0.71)9.0 (1.06)9.9 (1.02)10.0 (1.96)10.0(25.53)9.9(1.48)9.2 (1.32)10.0 (1.96)10.0(25.53)
      WBC (k/uL)3.6 (1.41)4.7 (0.64)8.9 (3.87)20.6 (6.33)33.0 (46.1)20.0 (6.38)16.0 (6.81)12.8 (6.62)33.0 (46.1)20.0 (6.38)
      ANC (k/uL)2.0 (0.85)3.4 (0.78)5.7 (2.41)16.8 (5.66)16.0 (8.45)17.0 (5.7)13.0 (6.47)9.6 (6.53)16.0 (8.45)17.0 (5.7)
      Na (mEq/L)139.0 (1.41)136.0 (2.83)136.0 (3.16)134.0 (6.10)133.0 (2.69)134.0 (2.9)130.0 (6.29)136.0 (4.52)133.0 (2.69)134.0 (2.9)
      BUN (mg/dL)6.0 (0.85)12.0 (5.66)10.0 (9.10)18.0 (9.75)7.0 (2.8)15.0 (10)15.0 (5.05)15.0 (6.64)7.0 (2.8)15.0 (10)
      Cr (mg/dL)0.71 (0.01)0.80 (0.16)0.58 (0.18)0.77 (0.11)1.00 (0.16)1.00 (0.45)1.01 (0.51)0.61 (0.15)∗1.00 (0.16)1.00 (0.45)
      Alb (g/dL)2.6 (0.92)2.4 (0.57)8.1 (0.29)2.6 (0.32)2.0 (0.38)3.0 (1)2.8 (0.72)2.5 (0.56)2.0 (0.38)3.0 (1)
      T. Bili (mg/dL)9.3 (8.56)6.4 (6.93)12.4 (5.01)16.2 (6.21)17.0 (9.4)16.0 (9.96)17.5 (6.75)11.8 (7.76)∗17.0 (9.4)16.0 (9.96)
      ALT (U/L)48.0 (2.83)38.5 (0.71) ∗42.0(27.05)59.0 (27.44)37.0 (13.66)57.0 (19.2)∗46.0 (32.70)49.0 (24.47)37.0 (13.66)57.0 (19.2)∗
      AST (U/L)198.0 (9.90)93.5(0.71)142.0 (60.92)121.0 (54.14)138.0 (37.8)164.0 (34.6)161.0 (102.18)121.0 (66.24)138.0 (37.8)164.0 (34.6)
      PT (s)18.2 (5.23)21.4 (2.83)22.3 (3.61)20.8 (4.28)21.0 (5.6)20.0 (2.97)22.0 (6.73)19.0 (4.26)∗21.0 (5.6)20.0 (2.97)
      INR1.9 (0.49)2.0 (0.49)2.2 (0.34)2.1 (0.45)2.0 (0.65)2.0 (0.3)2.1 (0.66)1.8 (0.45)∗2.0 (0.65)2.0 (0.3)
      DF37.8 (19.30)49.2 (6.01)56.0 (16.84)57.0 (20.31)56.5 (31.20)50.1 (24.33)48.9 (34.21)37.6 (22.96)∗56.5 (31.20)50.1 (24.33)
      MELD∗22.0 (2.83)23.0 (0.71)26.0 (3.77)24.0 (3.66)22.0 (5.82)24.0 (4.95)22.0 (6.28)22.0 (4.12)∗22.0 (5.82)24.0 (4.95)
      MELD-Na20.3 (0.71)19.0 (0.71)29.0 (2.5)26.0 (4.21)30.0 (4.63)25.0 (5.06)25.0 (5.84)21.0 (4.03)∗30.0 (4.63)25.0 (5.06)
      Values represented as mean (SD). Analysis performed through paired t-test within PC categories. Significant change from TD0 to TD4 (P < .05) represented with ∗.
      Alb, albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; BUN, blood urea nitrogen; Cr, creatinine; CS, corticosteroid; Hgb, hemoglobin; PT, prothrombin time; SD, standard deviation; TD, treatment day.
      Table 6Regression for Baseline PC and Day-4 Lille Score
      VariableEstimateSEP value
      Intercept0.6300.1071.255E-6
      PC TD0−0.0010.001.020
      Significance observed through linear regression of Lille score with continuous PC. Increasing PC associated with a lower Lille score. Significant values are in bold.
      SE, standard error.

      Spleen Size Based on Platelet Category

      SI recorded for 136 patients is presented in Table A1 stratified by PC category and gender. Imaging in the remaining 23 patients did not include the spleen. Among men (n = 88), the mean SI was 1.89 (standard deviation 1.21). Greater SI was correlated with a lower baseline PC (P = .016) and PC category (P = .01). No correlation was present among women.

      Discussion

      The important findings in our study include a direct relationship between baseline PC with markers of inflammatory response (WBC, ANC) and greater disease severity (TBili, Na-MELD). As expected, there was an inverse relationship with the SI. Patients with a higher PC were more likely to receive CS. Importantly, a higher PC was associated with a favorable response to CS as indicated by a lower day-4 Lille score.
      Only 2 studies to date have assessed the significance of the PC in patients with AH. In a study of 58 patients with a high clinical suspicion for the diagnosis who underwent liver biopsy for confirmation, PC was higher in the biopsy-confirmed group (178 × 109/L) than in those without AH (98.4 × 109/L) (P = .0005).
      • Hardy T.
      • Wells C.
      • Kendrick S.
      • et al.
      White cell count and platelet count associate with histological alcoholic hepatitis in jaundiced harmful drinkers.
      A PC of >147.5 × 109/L had a sensitivity of 56% and a specificity of 93% for detecting AH.
      • Hardy T.
      • Wells C.
      • Kendrick S.
      • et al.
      White cell count and platelet count associate with histological alcoholic hepatitis in jaundiced harmful drinkers.
      Although a subsequent study assessing cytokine levels in the diagnosis of AH also reported elevated PC in patients with biopsy-proven AH compared to that in those without confirmation, the predictive value of PC was significantly lower.
      • Bissonnette J.
      • Altamirano J.
      • Devue C.
      • et al.
      A prospective study of the utility of plasma biomarkers to diagnose alcoholic hepatitis.
      Neither study correlated inflammatory markers of PC with disease severity.
      A possible relationship between a higher PC and severity of AH is a shared pathophysiologic mechanism involving the inflammatory cytokine interleukin-6 (IL-6). The inflammatory response in AH is intimately linked to IL-6 expression. In patients with AH, IL-6 levels correlate with WBC, ANC, and clinical severity.
      • Fujimoto M.
      • Uemura M.
      • Nakatani Y.
      • et al.
      Plasma endotoxin and serum cytokine levels in patients with alcoholic hepatitis: relation to severity of liver disturbance.
      In a study of 127 patients, the plasma IL-6 level was the most precise predictor of morality.
      • Sheron N.
      • Bird G.
      • Goka J.
      • et al.
      Elevated plasma interleukin-6 and increased severity and mortality in alcoholic hepatitis.
      ,
      • Rachakonda V.
      • Gabbert C.
      • Raina A.
      • et al.
      Stratification of risk of death in severe acute alcoholic hepatitis using a panel of adipokines and cytokines.
      In most situations, the PC is determined by hepatic TPO production which, in turn, is regulated by the circulating PC mass. Multiple factors have the potential to suppress the PC in AH, including impaired TPO production, splenic sequestration, and alcohol-induced bone marrow suppression. However, a link that has not previously been explored is the potential for IL-6 to affect the PC. IL-6 expression in inflammatory states enhances hepatic TPO mRNA transcription,
      • Kaser A.
      • Brandacher G.
      • Steurer W.
      • et al.
      Interleukin-6 stimulates thrombopoiesis through thrombopoietin: role in inflammatory thrombocytosis.
      ,
      • Liang X.
      • Xiu C.
      • Liu M.
      • et al.
      Platelet-neutrophil interaction aggravates vascular inflammation and promotes the progression of atherosclerosis by activating the TLR4/NF-κB pathway.
      and administration of IL-6 results in a corresponding increase in TPO plasma levels and platelet counts. As a result, IL-6 could be driving both clinical severity as well as thrombocytosis in patients with severe disease.
      There are several important limitations of our study. First, biopsy for diagnostic confirmation was not available. In a trial in which liver biopsy was required for confirmation, 25% of patients with clinically suspected AH did not have histologic confirmation.
      • Hardy T.
      • Wells C.
      • Kendrick S.
      • et al.
      White cell count and platelet count associate with histological alcoholic hepatitis in jaundiced harmful drinkers.
      The duration of abstinence prior to hospitalization was also not available. Abstinence from alcohol in alcohol-dependent patients leads to bone marrow recovery and increased platelet counts in peripheral blood.
      • Cowan D.H.
      • Hines J.D.
      Thrombocytopenia of severe alcoholism.
      The average PC in patients with a lower baseline PC (<100) increased during the first 4 days of hospitalization with confirmed abstinence from alcohol. Although patients with greater portal hypertension as indicated by lower platelet counts and greater splenomegaly might have influenced the decision to administer CS, the presence and severity of ascites with the concern for the development of spontaneous bacterial peritonitis was similar in all groups. Finally, the small sample size per PC category and uneven distribution among PC categories are significant limitations.

      Summary and Conclusion

      Our study is the first to emphasize the relationship between admission PC with systemic inflammation and liver damage. Confirmation use of PC as a predictor for response to CS should be assessed further in future studies.

      Authors' Contributions

      Jessie A. Birnbaum contributed to conceptualization, investigation, methodology, and writing – original draft preparation, review, and editing. Howard Herman contributed to conceptualization, investigation, methodology, and writing – original draft preparation, review, and editing. Qi Gao contributed to data curation, formal analysis, data interpretation, critical review, writing – review and editing. Mordecai Koenigsberg contributed to conceptualization, investigation, and writing – review and editing. Samuel H. Sigal contributed to conceptualization, investigation, methodology, and writing – original draft preparation, review, and editing.

      Supplementary Materials

      References

        • Basra S.
        • Anand B.S.
        Definition, epidemiology and magnitude of alcoholic hepatitis.
        World J Hepatol. 2011; 3: 108-113
        • Shirazi F.
        • Singal A.K.
        • Wong R.J.
        Alcohol-associated cirrhosis and alcoholic hepatitis hospitalization trends in the United States.
        J Clin Gastroenterol. 2021; 55: 174-179
        • Thompson J.A.
        • Martinson N.
        • Martinson M.
        Mortality and costs associated with alcoholic hepatitis: a claims analysis of a commercially insured population.
        Alcohol. 2018; 71: 57-63
        • Moon A.M.
        • Curtis B.
        • Mandrekar P.
        • et al.
        Alcohol-associated liver disease before and after COVID-19-an overview and call for ongoing investigation.
        Hepatol Commun. 2021; 5: 1616-1621
        • Crabb D.W.
        • Im G.Y.
        • Szabo G.
        • et al.
        Diagnosis and treatment of alcohol-associated liver diseases: 2019 practice guidance from the American association for the study of liver diseases.
        Hepatology. 2020; 71: 306-333
        • Thursz M.R.
        • Richardson P.
        • Allison M.
        • et al.
        Prednisolone or pentoxifylline for alcoholic hepatitis.
        N Engl J Med. 2015; 372: 1619-1628
        • Mitchell O.
        • Feldman N.
        • Diakow M.
        • et al.
        The pathophysiology of thrombocytopenia in chronic liver disease.
        Hepat Med. 2016; 8: 39-50
        • Roayaie S.
        • Jibara G.
        • Berhane S.
        • et al.
        PALBI-an objective score based on platelets, albumin & bilirubin stratifies HCC patients undergoing resection & ablation better than Child’s classification.
        62. HEPATOLOGY, 2015: 631A-632A
        • Thomopoulos K.C.
        • Labropoulou-Karatza C.
        • Mimidis K.P.
        • et al.
        Non-invasive predictors of the presence of large oesophageal varices in patients with cirrhosis.
        Dig Liver Dis. 2003; 35: 473-478
        • Crabb D.W.
        • Bataller R.
        • Chalasani N.P.
        • et al.
        NIAAA Alcoholic Hepatitis Consortia. Standard definitions and common data elements for clinical trials in patients with alcoholic hepatitis: recommendation from the NIAAA alcoholic hepatitis Consortia.
        Gastroenterology. 2016; 150: 785-790
        • Brown Jr., R.S.
        Current management of thrombocytopenia in chronic liver disease.
        Gastroenterol Hepatol. 2019; 15: 155-157
        • Foncea C.G.
        • et al.
        Day-4 Lille score is a good prognostic factor and early predictor in assessing therapy response in patients with liver cirrhosis and severe alcoholic hepatitis.
        J Clin Med. 2021; 10
        • Kucybała I.
        • Ciuk S.
        • Tęczar J.
        Spleen enlargement assessment using computed tomography: which coefficient correlates the strongest with the real volume of the spleen?.
        Abdom Radiol (NY). 2018; 43: 2455-2461
        • Hardy T.
        • Wells C.
        • Kendrick S.
        • et al.
        White cell count and platelet count associate with histological alcoholic hepatitis in jaundiced harmful drinkers.
        BMC Gastroenterol. 2013; 13 (55–55)
        • Bissonnette J.
        • Altamirano J.
        • Devue C.
        • et al.
        A prospective study of the utility of plasma biomarkers to diagnose alcoholic hepatitis.
        Hepatology. 2017; 66: 555-563
        • Fujimoto M.
        • Uemura M.
        • Nakatani Y.
        • et al.
        Plasma endotoxin and serum cytokine levels in patients with alcoholic hepatitis: relation to severity of liver disturbance.
        Alcohol Clin Exp Res. 2000; 24: 48s-54s
        • Sheron N.
        • Bird G.
        • Goka J.
        • et al.
        Elevated plasma interleukin-6 and increased severity and mortality in alcoholic hepatitis.
        Clin Exp Immunol. 1991; 84: 449-453
        • Rachakonda V.
        • Gabbert C.
        • Raina A.
        • et al.
        Stratification of risk of death in severe acute alcoholic hepatitis using a panel of adipokines and cytokines.
        Alcohol Clin Exp Res. 2014; 38: 2712-2721
        • Kaser A.
        • Brandacher G.
        • Steurer W.
        • et al.
        Interleukin-6 stimulates thrombopoiesis through thrombopoietin: role in inflammatory thrombocytosis.
        Blood. 2001; 98: 2720-2725
        • Liang X.
        • Xiu C.
        • Liu M.
        • et al.
        Platelet-neutrophil interaction aggravates vascular inflammation and promotes the progression of atherosclerosis by activating the TLR4/NF-κB pathway.
        J Cell Biochem. 2019; 120: 5612-5619
        • Cowan D.H.
        • Hines J.D.
        Thrombocytopenia of severe alcoholism.
        Ann Intern Med. 1971; 74: 37-43