혈청과 복수 알부민 농도차 측정의 임상적 의의

Abstract

[한글]

일반적으로 복수의 감별진단에 있어 복수의 총단백질양, 세포수와 비증율 기준으로 복수를 여출액과 삼출액으로 구분하여 왔다. 간경변에 의한 복수는 총단백질양이 2.5 g/㎗ 이하인 삼출액으로 생각되고 있으나 약 20%에서는 총단백질양이 2.5 g/㎗ 이상으로 복수

의 원인질환을 규명하는데 명확한 지표가 되지 못하고 있다. 이에 비하여 혈청과 복수에서 동시에 알부민 농도를 측정하여 혈청의 알부민 농도에서 복수의 알부민 농도를 뺀 값인 혈청과 복수알부민 농도차(Serum Ascites Albumin Gradient: 이하 SAAG로 약함)는 SAA

G가 1.1 g/㎗ 이상인 경우는 간문맥압의 상승에 의한 복수로, 1.1 g/㎗ 미만인 경우는 간문맥압의 상승에 의하지 않은 복수로 나누어 복수의 원인질환을 규명하는데 유용함이 보고되고 있고, 악성복수의 감별진단에도 유용함이 보고되고 있다. 만성간질환 환자에서 간

문맥압의 상승은 복수, 식도정맥류 등의 다양한 합병증을 유발하고 정맥류 출혈등으로 생존에 유의한 예후인자로 작용할 수 있으나 간문맥압의 측정은 간정맥 도자술등 잔혈적인 검사방법으로 실제 임상에서 이용하는데는 많은 제약이 있다. 이에 간문맥의 혈역동학을

비관혈적으로 보다 간편하게 조사하기 위한 많은 연구가 있었으며 최근 도플러 초음파와 SAAG를 이용하여 간문맥압을 간접적으로 예측하기 위한 연구가 진행되고 있다. 그러나 알코올성 간질환과는 달리 B형 및 C형 간염바이러스에 의한 비알코올성 간질환에서는 SAAG

와 간문맥압과의 연관관계가 정립되어 있지 않고 악성복수의 진단적 가치도 아직 이론이 있어서 SAAG의 측청이 복수의 감별진단과 간문맥압의 반영지표로 유용한지 알아보기 위하여 본 연구를 시행하였다. 연구대상은 복수를 동반한 질병으로 연세대학교 의과대학 세브란스병원에 입원한 100명의 환자를 대상으로 하였고 총146회 SAAG를 측정하였다. 복수의 원인질환별로는 간경변이 39예, 간암이 30예, 간전이암이 9예, 복막암종증이 14예, 결핵성 복막염이 6예, 기타 2예 이었고 원인질병에 따른 SAAG는 간경변군 2.09 0.44 g/㎗, 간암군 2.09 0.42g/㎗, 복막염군 1.84 0.35 g/㎗, 간암파열군 1.58 0.35 g/㎗, 간전이암군 1.51 0.61 g/㎗, 암종증군 0.74 0.35 g/㎗, 결핵군 0.58 0.16 g/㎗ 이었다. 간경변, 간암, 복막염, 간암파열, 간전이암군은 SAAG가 1.1 g/㎗ 이상으로 간문맥압의 항진에 의한 복

수로 분류되었고 암종증, 결핵군은 SAAG가 1.1 g/㎗ 이하로 복막의 질환에 의해 유발된 복수로 분류되어 양 군의 SAAG는 서로 유의한 차이가 있었고(p<0.01) 간경변군과 간암군은 간암파열 및 간전이암군과 유의한 차이가 있었으나(p<0.05) 간경변군과 간암군 및 암종증군과 결핵군간에는 각각 유의한 차이가 없었다. 복수의 총단백질양은 간경변군 1.33 0.56 g/㎗, 간암군 1.56 0.89 g/㎗, 복막염군 1.63 0.84 g/㎗, 간암파열군 2.83 0.97 g/㎗, 간전이암군 2.53 1.26 g/㎗, 암종증군 4.16 1.06 g/㎗, 결핵군 4.97 0.87 g/㎗로 복수의 총단백질양 2.5 g/㎗를 기준으로 간경변과 간암, 복막염군은 삼출액으로 간암파열, 간전이암, 암종증, 결핵군은 여출액으로 구분되었으나 여출액인 간전이암군은 삼출액인 간경변, 간암, 복막염군과 복수의 총단백질양의 유의한 차이가 없었고, 여출액인 간암파열군 또한 여출액인 간전이암, 삼출액인 복막염군과 유의한 차이가 없어 복수의 생성기전, 원인질환등의 감별과 각 질병별 감별진단에 어려움이 있음을 알 수 있었다. SAAG가 시간경과 및 이뇨제 투여에 따른 영향을 받는지 조사하기 위하여 이뇨재 투여후 SAAG를 추적 검사하여 비교하여 보았으나 유의한 차이는 없어서 SAAG값은 단기적으로는 시간 및 이뇨제 투여에 따라 변하지 않고 일정한 같을 유지함을 알 수 있었다. 17예에서 간문맥압을

측정하여 이를 SAAG와 비교하여 본 결과 상관계수 0.52로 통계학적으로 유의한(p=0.032) 상관관계가 있음을 알 수 있었고 간문맥압 = 5.215 x SAAG + 2.02 의 회귀방정식을 얻어

서 SAAG를 이용한 간접적인 간문맥압의 예측이 가능함을 알 수 있었으나 도플러 초음파를 이용하여 측정한 간문맥직경 및 간문맥 혈류속도는 간문맥압 및 SAAG와 유의한 상관관계가 없었다. 식도정맥류가 있었던 군과 없었던 군의 SAAG는 각각 1.99 0.41 g/㎗, 1.28 0.67 g/㎗로 유의한 차이가 있었고(p<0.05) 식도정맥류 출혈이 있었던 군과 없었던 군의 SAAG는 각각 2.14 0.39 g/㎗ 1.63 0.63 g/㎗ 로 유의한 차이가 있었으나(p<0.0001) 식도정맥류의 정도와 적색소견의 유무 및 출혈횟수에 따른 유의한 차이는 없었고 간경변의 원인 및 정도와도 유의한 차이가 없었다. SAAG의 간문맥압 항진에 의한 복수와 복막질환에 의한 복수의 감별진단능은 예민도 95.24%, 특이도 98.28%, 정확도 98.47%이었고(n=79), 악성복수와 양성복수의 감별진단능은 예민도 93.75%, 특이도 90.38%, 정확도 91.17%로(n=67

) 복수의 총단백질양, 콜레스테롤, CEA보다 우수하였다. 또한 SAAG와 복수의 콜레스테롤과는 상관계수 0.67로(p=0.0001) 유의한 상관관계가 있음을 알 수 있었으나 SAAG와 CEA 및 글로블린과는 유의한 상관관계가 없었고 콜레스테롤과 간문맥압과도 유의한 상관관계

가 없었다. 이상의 결과로 SAAG는 다양한 복수의 감별진단에 유용하며 간문맥압과 유의한 상관관계가 있어서 간문맥압의 반영지표로 유용하게 사용될 수 있음을 알 수 있었으나 상관계수가 높지 않아 정밀한 반영은 어려울 것으로 생각되었다. 또한 SAAG는 식도정맥류

의 존재, 식도정맥류 출혈의 유무에 따른 유의한 차이가 있었고 간문맥압의 상승에 의한 복수와 복막질환에 의한 복수의 감별진단에 유용하며 악성복수의 감별진단에도 기존의 다른 화학적 검사보다 우수함을 알 수 있었다. 복수가 있는 환자에서 SAAG의 측정은 복수의

감별진단, 간접적인 간문맥압의 예측, 식도정맥류 및 식도정맥류 출혈의 예측, 악성복수의 진단등에 이용될 수 있으리라고 생각된다.

Clinical value of serum ascites albumin gradient in the prediction of portal

hypertension and differential diagnosis of ascites

Young Jun Shin

Department of Medical Science The Graduate School, Yonsei University

(Directed by associate professor Chae Yoon Chon)

Transudate and Exudate concept was applied in the differential diagnosis of

ascites according to the ascitic fluid total protein content cell count and

specific gravity. Cirrhotic ascites was regarded as transudate but about 20% of

cirrhotic ascites had more than 2,5 g/dl total protein within the ascites. Serum

ascites albumin gradient(SAAG) which is the difference of serum and ascites albumin

concentration can differentiate the portal hypertensive and nonportal hypertensive

ascites with the cut off value 1.1 g/dl, and is reported as an useful method in the

differential diagnosis of malignant ascites. In chronic liver disease patients,

portal hypertension can cause various complications and is a significant prognostic

factor due to variceal bleeding, but measurement of portal vein pressure needs

invasive hepatic vein catheterization procedure and had the limitation in the

clinical application. Another metaled of portal vein pressure estimation is an

ultrasonic measurement of portal hemodynamics with doppler ultrasound, but the

accuracy of doppler ultrasonic measurement needs more verification. SAAG is also

reported as an useful index of portal vein pressure in alcoholic liver disease

patients, but the correlation of SAAG and portal vein pressure in postnecrotic

liver disease patients was poor in some reports and need more investigation for the

clinical application in Korea because most liver cirrhosis patients were

postnecrotic.

SAAG, ascitic fluid total protein, cholesterol and carcinoembryonic antigen were

measured to evaluate the accuracy in the differential diagnosis of ascites and the

hepatic vein catheterization was performed to compare the portal vein pressure with

SAAG and doppler measurement of portal blood flow velocity and diameter. SAAG was

measured 146 times in 100 ascitic patients. The underlying cause of ascites was

liver cirrhosis in 39 cases(alcoholic 5, hepatitis B and C 32, cause unknown 2),

hematoma in 30 cases(alcoholic 3, hepatitis B and C 25, cause unknown 2), liver

metastasis 9 cases, peritoneal carcinomatosis 14 cases, tuberculous peritonitis 6

cases and miscellaneous 2 cases. The measured SAAG was 2.09 ± 0.44 g/dl in liver

cirrhosis, 2.09 ± 0.42 g/dl in hepatoma, 1.84% ± 0.35g/dl in spontaneous

bacterial peritonitis, 1.58±0.35 g/dl in hepatoma rupture, 1.51±0.61 g/dl in

liver metastasis, 0.74±0.35 g/dl in peritoneal carcinomatosis, 0.58±0.16 g/dl in

tuberculous peritonitis patients and showed significant difference between portal

hypertensive ascites such as liver cirrhosis, hematoma, spontaneous bacterial

peritonitis, hepatoma rupture, liver metastasis and nonportal hypertensive ascites

such as peritoneal carcinomatosis, tuberculous peritonitis(p<0.01). Liver

cirrhosis, hepatoma and spontaneous bacterial periyonitis were classified as

trasudate and hepatoma rupture was classified as exudate when classified with

ascitic fluid total protein content, and showed the limitations in the differential

diagnosis of ascites. There were no difference of SAAG after diuretcs treatment and

showed stable value when measured 2 weeks later after diuretics treatment. The

correlation of SAAG and portal veinpressure measured in 17 patients showed

significant correlation(r=0,52:p=0.032) and could predict the portal vein pressure

with SAAG using the regression equation(Portal vein pressure=5.215 x SAAG + 2.02).

There were no correlation between doppler measured portal vein blood flow velocity

and diameter with SAAG and Portas vein pressure. The SAAG of ascites patients with

esophageal varix was 1.99t10.41 g/dl, and without esopahgeal varix was 1.28±0.67

g/dl and there were sinigficant difference between the two groups(p<0.05). The SAAG

of patients with the episode of variceal bleeding was 2.14±0.39 g/dl and the

patients without bleeding episode was 1.63±0.63 g/dl, and showed the significant

difference also(p<0.01). However, there were no significant difference of SAAG

according to the varix grading, red color sign and Child-Pugh class of liver

cirrhosis. In the differential diagnosis of malignant ascites and nonneoplastic

ascites, the accuracy, sensitivity and specificity of SAAG was 91.2%, 93.8%, 90.4%

respectively and was the best one among the measured humoral parameters of ascitic

fluids(SAAG, total protein, cholesterol and CEA). SAAG and cholesterol showed

significant correlation(r=0.67) but there were no significant correlation with CEA

and globulin.

In conclusion, SAAG can differentiate the portal hypertensive and nonportal

hypertensive ascites and can be used in the diagnosis of malignant ascites. Also

SAAG measurement can give us the useful information of esophageal varix and

variceal bleeding. The correlation between SAAG and portal vein pressure was 0.52

and showed significant correlation and can be used as a representatives of portal

vein pressure.

[영문]

Transudate and Exudate concept was applied in the differential diagnosis of ascites according to the ascitic fluid total protein content cell count and specific gravity. Cirrhotic ascites was regarded as transudate but about 20% of cirrhotic ascites had more than 2,5 g/dl total protein within the ascites. Serum

ascites albumin gradient(SAAG) which is the difference of serum and ascites albumin concentration can differentiate the portal hypertensive and nonportal hypertensive ascites with the cut off value 1.1 g/dl, and is reported as an useful method in the

differential diagnosis of malignant ascites. In chronic liver disease patients, portal hypertension can cause various complications and is a significant prognostic factor due to variceal bleeding, but measurement of portal vein pressure needs

invasive hepatic vein catheterization procedure and had the limitation in the clinical application. Another metaled of portal vein pressure estimation is an ultrasonic measurement of portal hemodynamics with doppler ultrasound, but the accuracy of doppler ultrasonic measurement needs more verification. SAAG is also

reported as an useful index of portal vein pressure in alcoholic liver disease patients, but the correlation of SAAG and portal vein pressure in postnecrotic liver disease patients was poor in some reports and need more investigation for the clinical application in Korea because most liver cirrhosis patients were

postnecrotic.

SAAG, ascitic fluid total protein, cholesterol and carcinoembryonic antigen were measured to evaluate the accuracy in the differential diagnosis of ascites and the hepatic vein catheterization was performed to compare the portal vein pressure with SAAG and doppler measurement of portal blood flow velocity and diameter. SAAG was measured 146 times in 100 ascitic patients. The underlying cause of ascites was liver cirrhosis in 39 cases(alcoholic 5, hepatitis B and C 32, cause unknown 2),

hematoma in 30 cases(alcoholic 3, hepatitis B and C 25, cause unknown 2), liver metastasis 9 cases, peritoneal carcinomatosis 14 cases, tuberculous peritonitis 6 cases and miscellaneous 2 cases. The measured SAAG was 2.09 ± 0.44 g/dl in liver cirrhosis, 2.09 ± 0.42 g/dl in hepatoma, 1.84% ± 0.35g/dl in spontaneous bacterial peritonitis, 1.58±0.35 g/dl in hepatoma rupture, 1.51±0.61 g/dl in liver metastasis, 0.74±0.35 g/dl in peritoneal carcinomatosis, 0.58±0.16 g/dl in tuberculous peritonitis patients and showed significant difference between portal hypertensive ascites such as liver cirrhosis, hematoma, spontaneous bacterial peritonitis, hepatoma rupture, liver metastasis and nonportal hypertensive ascites such as peritoneal carcinomatosis, tuberculous peritonitis(p<0.01). Liver

cirrhosis, hepatoma and spontaneous bacterial periyonitis were classified as trasudate and hepatoma rupture was classified as exudate when classified with ascitic fluid total protein content, and showed the limitations in the differential diagnosis of ascites. There were no difference of SAAG after diuretcs treatment and showed stable value when measured 2 weeks later after diuretics treatment. The correlation of SAAG and portal veinpressure measured in 17 patients showed significant correlation(r=0,52:p=0.032) and could predict the portal vein pressure with SAAG using the regression equation(Portal vein pressure=5.215 x SAAG + 2.02).

There were no correlation between doppler measured portal vein blood flow velocity and diameter with SAAG and Portas vein pressure. The SAAG of ascites patients with esophageal varix was 1.99t10.41 g/dl, and without esopahgeal varix was 1.28±0.67

g/dl and there were sinigficant difference between the two groups(p<0.05). The SAAG of patients with the episode of variceal bleeding was 2.14±0.39 g/dl and the patients without bleeding episode was 1.63±0.63 g/dl, and showed the significant difference also(p<0.01). However, there were no significant difference of SAAG according to the varix grading, red color sign and Child-Pugh class of liver cirrhosis. In the differential diagnosis of malignant ascites and nonneoplastic ascites, the accuracy, sensitivity and specificity of SAAG was 91.2%, 93.8%, 90.4% respectively and was the best one among the measured humoral parameters of ascitic fluids(SAAG, total protein, cholesterol and CEA). SAAG and cholesterol showed significant correlation(r=0.67) but there were no significant correlation with CEA and globulin.

In conclusion, SAAG can differentiate the portal hypertensive and nonportal hypertensive ascites and can be used in the diagnosis of malignant ascites. Also SAAG measurement can give us the useful information of esophageal varix and variceal bleeding. The correlation between SAAG and portal vein pressure was 0.52

and showed significant correlation and can be used as a representatives of portal vein pressure.