ONLINE MANUSCRIPT SUBMISSION
Editorial Board | Instructions for Authors | Main Page | Table of Contents | Archive | Search
The Turkish Journal of Gastroenterology
1999, Volume 10, No 2, Page(s) 96-100
[ Summary ] [ Similar Articles ] [ Mail to Editor ]
The diagnostic role of ultrasonography in liver steatosis
Ergün Yılmaz
Çukurova Üniversitesi, Tıp Fakültesi, İç Hastalıkları Anabilim Dalı, Gastroenteroloji Bilim Dalı, Adana
Keywords: IHepatic steatosis, steatonecrosis, alcoholic cirrhosis, ultrasonography,Hepatik steatoz, steatonecrosis, alkolik siroz, ultrasonografi.
Summary
Evaluation of 4536 patients who underwent abdominal ultrasonographic investigation at the Gastroenterology department of Çukurova University revealed hepatic steatosis in 672 patients. Only 52 of these patients gave informed consent to enter the study and all 52 patients had percutaneous liver biopsies performed. Histologic findings were consistent with steatosis in all but one patient who had chronic persistent hepatitis. Predisposing conditions were: manifested diabetes mellitus in 27.45 %, impaired diabetes mellitus in 21.56 %, potential diabetes mellitus in 9.8%,alcohol abuse in 13.72 %,hyperlipidemia in 13.72 %, chronic estrogen use in 3.92 %,chronic corticosteroid use in 5.98 % and chronic anemia (iron deficiency) in 3.92 % the patients. Fourty six (90 %) patients complained of right upper quadrant pain. The positive predictive value of ultrasonographic evaluation was 98.07 % in this group. A new ultrasonographic staging was performed and these stages were correlated with laboratory and histopathologic findings.
  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • Introduction
    In healthy subjects, five percent of the liver weight consists of lipids mostly being phospholipids, triglycerides and other lipid fractions. However, lipid content, especially triglycerides may increase up to 40-45 % in liver steatosis (1). Lipid accumulation may occur either in microvesicular or macrovesicular form (2). Steatosis may involve all of the hepatocytes and liver weight could reach 6000 grams in severe cases.

    The most common form of steatosis is macrovesicular fatty accumulation. Diabetes mellitus and chronic alcoholism are the most common predisposing conditions. The amount of alcohol is as important as the duration of abuse. Fatty infiltration occurs in zone 2-3, in addition to collagen accumulation in zone 3, phlebitis and gradual obliteration in the terminal hepatic and sublobular veins (2). In 25 % of cases, serum bilirubin and transaminase levels are increased (3). On the other hand liver histology ¦n diabetes mellitus is similar to that of chronic alcoholism except for vascularization in hepatocyte nuclei and periportal lesions. Fatty liver is predominant in type II diabetes mellitus, whereas minimal fatty accumulation involving mostly zone I occurs in type I diabetes mellitus. Other common reasons for macrovesicular fatty infiltration are Kwashiorkor disease, inflammatory bowel disease, pancreatic disease and intestinal by-pass operations.

    Microvesicular fatty infiltration occurs in zone 3 and accompanies centrizonal necrosis. The most common causes of microvesicular fatty infiltration are as follows (2,4,5): acute lipid metamorphosis of pregnancy, Reye syndrome and tetracycline and salicylate toxicities.

    The aim of this study was to correlate the sonographic, histopathologic and biochemical findings in patients with fatty liver infiltrations.

    The diagnostic role of ultrasonography in liver steatosis

    Figure 1:

    Figure 2:

  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • Materials And Methods
    This was a prospective clinical study, carried out between September 1990 and October 1995 at the department of Gastroenterology, Çukurova University Medical School, Adana. The male /female ratio was 30/22 . Ultrasonographic evaluation was performed using Toshiba SAL- 38 ultrasonography by a gastroenterologist specializing in gastroenterologic. Following ultrasonographic evaluation, a complete physical examination, past medical history and informed consent was taken from the patients in this study. Fasting blood glucose level, prothrombin time, markers of viral hepatitis, SMA 12, serum lipid profile, lipoprotein electrophoresis and whole blood count (WBC) were evaluated in addition to the liver biopsy. Patients with positive hepatitis markers were excluded from the study.

    Criteria used for ultrasonographic grading of hepatosteatosis were as follow.

    Grade Ia : Phomogenous hyperechogenic changes in liver parenchyma are seen ( Figure 1).

    Grade Ia': There is a homogenous hyperechogenity with fine and spark architecture (Figure 2).

    Grade Ib : Hepatomegaly and thickening of the liver edge in addition to hyperdensity and excessive hyperechogenicity in liver architecture are detected (Figure 3).

    Grade Ic : Hepatomegaly, thickening of the liver edge, uniform, dense and coarse excessively hyperechoic pattern (Figure 4).

    Figure 3:

    Figure 4:

    Grade II : Dense hyperechogenity of the liver parenchyma, moderate nonhomogenous changes, punctutions and patchy hypoechoic regions, thickening of the liver edge and hepatomegaly are the typical findings (Figure 5)

    Grade III: Nonhomogeneous hyperechogenictiy, coarse and patchy hypoechoic regions, regeneration nodules in the hyperechoic liver parenchyma, and thickening of the liver edge are defined. The sonographic findings indicat portal hypertension and ascites are also detected in this grade of fatty liver (Figure 6).

    Criteria used for the histopathologic grading of hepatosteatosis were as follow.

    Grade Ia : There is a moderate fat accumulation in the hepatocytes

    Grade Ia': Centrolobular microvesicular steatosis of the hepatocytes and mild centrozonal necrosis are seen

    Grade Ib : Severe fat accumulation in the hepatocytes, vascularizatition in the cytoplasm, cleaved nuclei and intercellular lipid cysts are demonstrated.

    Garde Ic : Severe fat accumulation also occurs in other liver cells and cytoplasmic ballooning is seen in addition to the findings at grade Ib.

    Grade II : Significant hepatocellular necrosis, mono nuclear infiltration, paracentral sclerosis and sinusoidal capillarization are the predominant findings. In addition to these findings, leucocyte infiltration, Mallory bodies and perisinusoidal colloid fibers are detected in chronic alcoholism .

    Grade III: Significant degeneration of the liver parenchyma, fibrosis and regenerating nodules are seen, in addition to the findings seen in grade II.

    In statistical analysis for the comparability tests we used kappa statistics and chi square, for trend test for testing the existence of linear trend on SGOT, SGPT, triglyceride and cholesterol (SPSS Statistical software Inc., Chicago 1990 ).

    Figure 5:

    Figure 6:

  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • Discussion
    There are several predisposing diseases in hepatosteatosis (2,4,5) . Of those, the most common are diabetes mellitus and chronic alcohol abuse. In one study, autopsy of 268 alcoholic patients revealed hepatosteatosis in 78 % of them (6). In other studies, the incidence of hepatomegaly was 80% and of cirrhosis 20% in autopsy specimens of patients with alcoholic hepatitis (1,2).

    In another group of 69 patients, the causes of hepatosteatosis were as follows: chronic alcoholism in 40.57%, obesity in 24.63%, diabetes mellitus in 10.14 % and not determined in 24.65% (7). In the current study, causes of liver steatosis were as follows: manifested diabetes mellitus in 14(27.45%), impaired diabetes in 11(21.56%), potential diabetes in 5(9.8%), alcoholism in 7(13.72%), hyperlipoproteinemia in 7(13.72%), long term use of prednisone in 3(5.98%), estrogen use in 2(3.92) and anemia in 2 patients (3.92%). Alcoholism has been the most frequently accounted disease in western populations, but diabetic hepatosteatosis was predominant in our group. It has been suggested that there is a difference between manifested and chemical diabetes mellitus concerning the development of hepatic steatosis (8), which is in concordant with our data. The rate of hepatic steatosis was 27.45% in manifested and 21.56% in chemical diabetes mellitus in our study group (p= 0.05). It has been reported that hepatomegaly and macrovesicular fat infiltration develops in 100% of patients with type I diabetes mellitus and this proqresses to cirrhosis (2). In our group, there were six cases of grade III hepatosteatosis, with manifested diabetes ¦n one and potential diabetes in five, but no deterioration in serum transaminase levels or any other predisposing conditions. One of the interesting findings in our group was the presence of two patients with severe iron deficiency anemia and special so-nographic findings indicating a grade Ia' hepatosteatosis. This finding has not been previously reported.

    The percentage of the liver steatoses has been reported as 66.99% in obesity, with a positive correlation between the severity of obesity and steatosis (8). There was no excessively obese patient in our group but nine female and five male patients were moderately overweight. Cushing syndrome is not frequently associated with liver steatosis (9), but long term corticosteroid use could lead to a fatty liver causing an increase in free fatty ascites (10). In our group only 3/51 patients had a history of corticosteroid use for several months.

    It has also been reported that hepatic steatosis could develop in patients using estrogen for prostatic carcinoma (11). However, we had only two female patients with a history of long term estrogen use no male and in our study population.

    Right upper quadrant pain has been noted in patients with the hepatic steatosis but the frequency of this tendency has not been defined as yet (1,2). In our patient population this frequency was 612/672 (91.07 %) in the group with sonographic evidence of hepatic steatosis and 46/ 51 (90.19%) in the patients who had histopathologically proven hepatic steatosis.

    Diffuse fatty infiltration of the liver has also been demonstrated in type IV and V hyperlipoproteinemias (12,13). In our study group there were five cases wich type II- B and two cases with type II- A hyperlipoproteinemias. In the current study hypertriglyceridemia and hypercholesterolemia accompanied diabetes mellitus in 17/24 cases and chronic alcohol abuse in 3/7 cases.

    It was concluded that specificity and sensitivity of ultrasonographic imaging is high in the diagnosis of hepatic steatosis. The new sonographic grading system was found to be valuable, but it needs to be further investigated in large patient populations. Chronic anemia may be a coexisting factor in the development of a fatty liver, and a blunt pain in the right upper quadrant may frequently accompany this entity.

    Table 1: The correlation of ultrasonographic and histopathologic grades

    Table 2: Correlations between ultrasonographic grades of steatosis and abnormalities of liver function tests.

  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • Conclusion
    During the five year period of this study 2,203 men and 2,333 women had been evaluated by abdominal ultrasonography and hepatosteatosis was found in 672 of them(14.81%). The male/females ratio was 302/370. Informed consent was obtained from only 52/672 patients to enter this study. Clinical and laboratory evaluation revealed manifested diabetes mellitus in 14 (27. 45%), impaired diabetes mellitus in 11 (21.56 %), potential diabetes in 5 (9.8 %),alcoholism in 7(13.72 %) and hyperlipoproteiniemia in 7(13.72 %, which were type II-B in 5,and type II-A in 2 patients). A medical history of long term prednisone use was reported in 3(5.98%) patients and long term estrogen use 2(3.92 %). Two the patients (3.92%)had iron deficiency anemia. Biochemical abnormalities in the different grades of steatosis are shown in Table 2. Forty six of the patients complained of a blunt pain in the right upper quadrant of the abdomen and were referred to Department of Gastroenterology, Çukurova university with suspected cholelithiasis. The positive predictive value of sonography was (51/52) 98.07%. In addition,there was a significant (p=0.001) correlation between the histologic and sonographic grading criteria (% 88.46), which is shown in Table 1. The high value of the kappa statistics (K:0.79, p=0.00) indicate the high degree of agreement between these two systems of grading.

    Another interesting finding in our study was a positive correlation between the sonographic grade of fatty liver and abnormalities in serum transaminase levels. Serum glutamic oxaloacetic transaminase (SGOT) levels were within the normal limits in grade Ia-b-a', slightly increased in grade Ic' and moderately increased in grade II and grade III liver steatosis (p= 0.004) (Table 2). Serum bilirubin levels were within the normal range in patients with grade Ia, a',b and c;but slightly increased in 11/17 patients with grade II and high in 2/6 patients with grade III hepatic steatosis ( p= 0.02). Serum alkaline phosphatase level was increased in 25% of grade II and 40% of grade III hepatic steatosis (p=0.002). Prolongation of prothrombin time was demonstrated in 47% of grade II and in 80% of grade III hepatic steatosis (p= 0.0001) ( Table 2). The serum triglyceride and cholesterol levels showed a slight downward trend which was not significant.

  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • References

    1) Schiff, L and Schiff, E. Diseases of the liver, Sixth Edition, Philadelphia, J.B. Lippincott Company, 1987; 25: 949-74.

    2) Sherlock S and Dooley J. Diseases of the liver and biliary system, Ninth edition. Oxford: Blackwell Scientific publications, 1993; 18: 328-33.

    3) Lelbach WK. Cirrhosis in the alcoholic and the relation to the volume of alcohol abuse. Ann NY Acad Sci 1975; 252: 85.

    4) Braillon A, Herse MA, Degott C et al. Liver in obesity. Gut 1985; 26: 133.

    5) Clain DJ, Lefkowitch JH. Fatty liver disease in morbid obesity. Gastroenterol. Clin N Am 1987; 16: 239.

    6) Lynch MJC et al. Fat embolism in chronic alcoholism.Arch. Pathol 1959; 67: 68.

    7) Samarasinghe D, Tasman-Jones C: The clinical associations with hepatic steatosis. A retrospective study. N.Z. Med J 1992; 26: 105: 57-8.

    8) Creutzfeld W et al: Liver diseases and diabetes mellitus.Prog.Liver Dis. 1970; 3: 371.

    9) Soffer LJ et al. Cushing's syndome, a study of 50 patients. Am J Med 1961 ; 30 :129.

    10) Hill RBJr et al. Hepatic lipid metabolism in the cortisone treated nat. Exp. Mol. Pathol 1965; 4: 320.

    11) Poupon R, Rosensztaj C, de Saint-Maur RD et al : Perhexilene maleate-associated hepatic injury: prevalence and characteristics. Digestion 1980 ; 20:145

    12) Leanerts J, Verresan L, Van Steenbergen W, Fevery S: Fatty liver hepatitis and type V hiperlipoproteinemia in juvenile diabetes mellitus. Case report and review of literature. J Clin Gastroenterol 1990; 12: 93-97

    13) Hoyumpa AM et al: Fattt liver. Biochemical and clinical considerations. Am J Dig Dis 1975; 20: 1142.

  • Top
  • Summary
  • Introduction
  • Materials And Methods
  • Discussion
  • Conclusion
  • References
  • [ Summary ] [ Similar Articles ] [ Mail to Editor ]
    Editorial Board | Instructions for Authors | Main Page | Table of Contents | Archive | Search