Corticosteroids and pentoxifylline for the treatment of alcoholic hepatitis: Current status

Table 1 Randomized studies to assess corticosteroids for treatment of acute alcoholic hepatitis

Ref.	Study design	Sample size	Mean age (yr) males (%)	Drug schedule	Main outcome /findings	Secondary findings	Causes of death
Helman et al[19] 1971	Randomized controlled trial: 3 groups: severe, moderate without encephalopathy, and ambulatory	37 (20)	48 (32)	Prednisolone 40 mg/d × 4 wk	Mortality benefit seen only for group I with severe alcoholic hepatitis (1/15 in treated vs 6/15 in untreated, P < 0.01)	No difference on histology at 4 wk and no effect on prevention to cirrhosis. Improved caloric intake was seen with steroids	Treated: (n = 1): liver failure. Untreated: (n = 6): hepatorenal syndrome (4), lower gastrointestinal bleed (1), variceal bleed (1)
Porter et al[20] 1971	Double blind Randomized controlled trial	20 (11)	45 (64)	6-Methylprednisolone 40 mg/d in 3 d × 10 d followed by oral if possible	Survival 45% vs 22%, P = NS	No effect on biochemical parameters	Treated: (n = 1): tuberculosis
Campra et al[21] 1973	Prospective controlled trial	45 (20)	43 (75)	Prednisone 0.5 mg/kg per day × 3 wk then 0.25 mg/kg per day × 3 wk	Survival was no different (36% vs 35%). Trend for improved survival with encephalopathy (P = 0.2)	No effect on biochemical parameters	Treated: (n = 7): hepatic failure. Untreated (n = 9): hepatic failure, GIB (5), renal failure (4)
Blitzer et al[22] 1977	Prospective double blind	28 (16)	48.4 (not reported)	Prednisolone 40 mg × 14 d then tapering × 2 wk	Overall mortality higher in the treated group	No effect on biochemical parameters. Prothrombin time higher in non-survivors	Treated: (n = 11): GIB, Hepatorenal syndrome, spontaneous bacterial peritonitis fungal infections (33% cases): disseminated aspergillosis, candidemia disclosed on autopsy; Untreated (31%): GIB, hepatorenal syndrome, spontaneous bacterial peritonitis, fungal infections
Shumaker et al[23] 1978	Randomized controlled trial	27 (12)	44 (75)	6 -methyl prednisolone 80 mg/d × 4-7 d po	No change in survival (50% vs 53%)	Patients with contraindication to steroids had higher mortality. Causes of death in the two groups were similar with > 50% dying from GIB	Treated: (n = 3): GIB, (n = 2): hepatic failure, (n = 1): acute pancreatitis. Untreated: (n = 3): GIB, (n = 2): sepsis, (n = 1): not reported
Maddrey et al[24] 1978	Randomized controlled trial	55	40 (60)	prednisolone 40 mg/d × 28-32 d	Improved short-term mortality but no effect on development of portal hypertension even in short term	Serum bilirubin > 20, prothrombin time > 8 s, prolonged and encephalopathy predicted mortality	Treated: (n = 2): hepatic failure, (n = 1): cytomegalic inclusion disease and pneumocystis carinii pneumonia mono-lineal esophagitis, (n = 2): severe liver disease. Untreated: (n = 5): hepatic failure, coma and hepatorenal syndrome
Lesesne et al[25] 1978	Randomized controlled trial	14 (7)	49 (not reported)	Prednisolone 40 mg/d × 30 d then 2 wk of tapering	Improved survival of the treated group. Improved nutrition alone is not a factor for better survival	Infrequent complications from steroids could be cause of death	Treated: (n = 2): hepatic failure and hepatorenal syndrome, (n = 1): hemorrhagic pancreatitis, (n = 1): pneumococcal pneumonia. Untreated: (n = 7): hepatic failure, (n = 4): GIB, (n = 3): hepatorenal syndrome, (n = 1): aspiration pneumonia, (n = 1): klebsiella bacteremia
Depew et al[26] 1980	Randomized controlled trial	28 (15)	49 (66)	Prednisolone 40 mg/d × 28 d then taper × 14 d	Mortality in both groups similar (53% vs 54%). LOS: 66 d with prednisolone and 56 d with placebo	No effect on biochemical parameters and complications higher with steroids	Treated: (n = 7): urinary tract infection, (n = 3) pneumonia, (n = 2) septicemia, (n = 1) perinephric abscess. Untreated: (n = 6): urinary tract infection, (n = 1) pneumonia
Ramond et al[27] 1992	Randomized controlled trial	61 (32)	48 (not reported)	prednisolone 40 mg/d × 28 d (IV if unable to take orally)	Improved survival at 6 mo (84% vs 45%, P = 0.002) irrespective of encephalopathy for patients with discriminant function > 32 (21/23 vs 10/19, P < 0.001)	Death in steroids group occurred early. Patients should be started on steroids while awaiting biopsy results	Treated: (n = 2): gastritis and GIB, (n = 2): septicemia lung. Untreated: (n = 16): GIB, ascites, variceal rupture, pancreatitis, (n = 1 each): septicemia
Theodossi et al[28] 1982	Randomized controlled trial	55 (27)	Not reported, 70% (treatment group), 30% (control group)	Methylprednisolone 1 g/d × 3 d	Patients survival predicted by: encephalopathy, discriminant function > 93, bilirubin 20 mg/dL, creatinine 3 mg/dL, and histological evidence of cirrhosis	Not reported	Treated: (n = 7):septicemia, (n = 2): pancreatitis. Untreated: (n = 6): septicemia, (n = 2): pancreatitis % mortality in patients of hepatic encephalopathy: 94% (treatment group) 69% (control group)
Richardet et al[29] 19931	Randomized controlled trial	23 (12)	Not reported	Prednisolone 40 mg/d × 8 d	Tumor necrosis factor, interleukin-6, interleukin-8 decreased in treated group significantly at Day 8 from their baseline Day 0 levels	Not reported	Not reported

¹Abstract publication. NS: Not significant; GIB: Gastrointestinal bleeding.

Table 2 Randomized controlled trial controlled trial studies to assess pentoxifylline for treatment of acute alcoholic hepatitis

Ref.	Sample size	Mean age (yr) males (%)	Drug schedule	Main outcome/findings	Secondary findings	Causes of death
McHutchison et al[30] 19911	22 (12 pentoxifylline)	Not reported	Pentoxifylline 400 mg tid× 10 d	Renal impairment more with placebo (mean creatinine change -0.3 vs +2.1	No difference on other biochemical parameters. Plasma tumor necrosis factor increased in controls only. Survival trend better with pentoxifylline (3 deaths vs 1 death)	Not reported
Akriviadis et al[15] 2000	101 (49 pentoxifylline)	42 (71% males)	Pentoxifylline 400 mg tid× 28 d	Mortality during the admission 25% vs 46% (P = 0.037)	Age, creatinine at randomization, and pentoxifylline treatment predicted survival. Tumor necrosis factor levels were no different with pentoxifylline and placebo. However, among non-survivors tumor necrosis factor levels decreased more in pentoxifylline group	Hepatorenal syndrome: treated vs untreated (50% vs 92%, P = 0.009)
Paladugu et al[31] 20061	30 (14)	50 (100%)	Pentoxifylline	Mortality at 28 d: 29% vs 46% (P = 0.09). Time to death 21 d vs 18 d (P = 0.041)	Tumor necrosis factor levels unchanged in both groups	Hepatorenal syndrome: treated vs untreated (50% vs 86%, P = 0.1)
Sidhu et al[32] 20061	50	Not reported	Pentoxifylline 400 mg tid× 28 d	Mortality at 28 d (24% vs 40%, P = NS)	Pentoxifylline reduced creatinine, tumor necrosis factor, discriminant function index, prothrombin time	Hepatorenal syndrome: treated vs untreated (83% vs 60%)
Lebrec et al[33] 20071	132	Not reported	Pentoxifylline	Mortality at 2 mo (14% vs 16%, P = 0.77) and at 6 mo (27% vs 31%, P = 0.3) were similar	No difference for serious adverse effects between the 2 groups. Subgroup with renal dysfunction also did not get benefit with pentoxifylline	Not reported

¹Abstract publication.