| HOME
|
Cirrhosis
|
Complications of Cirrhosis This material does not cover all information and is not intended as a substitute for professional care. Please consult with your physician on any matters regarding your health. Written by Russell Wiesner M.D. Medical Director liver transplant
program, Mayo clinic Rochester Mn Over time , however , usually decades following HCV infection liver
disease can progress to a point where the damage is too great for the body
to compensate. At this time serious complications can occur, and individuals
are said to be Decompensating or to have Decompensated liver disease. In this article, I will review the likelihood of a persons developing decompensated cirrhosis related to HCV, as well as clinical manifestations and warning signs. I will also discuss how complications can be treated, and how new therapies are offering greater hope for patients who progress to this point. How Likely are you to develop decompensated
cirrhosis related to HCV? In most cases, it takes about 15 to 25 years for a person with chronic HCV to develop cirrhosis. A study done in Europe found that the probability of developing decompensated cirrhosis after being diagnosed with cirrhosis was 12 percent at three years, 18 percent at five years and 29 percent at 10 years. The probability of survival after diagnosis of compensated cirrhosis was 96 percent at three years,91 percent at five years, and 79 percent at 10 years. The 5-year probability of survival after the appearance of the first major complication of decompensated cirrhosis was 50 percent at five years. Appropriate and timely care can make it possible for a patient to recover from complications and return to their previous level of health, but treatments has to begin as soon as possible. For this reason, it is important for patients to understand what possible complications can occur, how to prevent them, and when they can be managed with routine doctor visits and when emergency help is required. Complications warning signs and treatments Portal hypertension occurs when scare tissues builds up in the liver to
the point that it blocks the normal flow of blood from the intestinal area
through the portal vein to the liver and then through the vena caca back to
the heart. As blood is blocked from flowing through the liver, pressure in
the liver and portal vein increases. This known as "Portal hypertension".
The increased blood pressure in the vein forces fluid to weep from the liver
into the peritoneal (Abdominal) cavity. This causes an accumulation of fluid
in the abdomen referred to as "ascites". Ascites Ascites can range from mild to severe. The first sign of ascites is
abdominal swelling. Ascites also can be detected by an abdominal ultrasound.
If the abdomen becomes swollen, a patient should contact a physician and
arrange for a evaluation. Mild ascites can be treated with dietary salt
restriction and diuretics such as furosemide(Lasix) and
spironolactone(Aldactone). Edema Pedal edema also involves a build up of fluid in the body: however, in this condition the fluid collects in the ankles and feet rather than in the abdomen. Warning signs include swollen ankles or feet. Gradual swelling indicates a mild condition that is treated with dietary salt restriction and diuretics. But if the swelling rapidly worsens, the patient should seek emergency help. Spontaneous bacterial pertionitis (SBP) Varcies Varcies and blood vessels, usually in the esophagus or stomach, which
have become swollen as a result of portal hypertension. Varices can be
diagnosed by viewing the blood vessels lining the esophageal wall with an
endoscope. Mild varices are treated with beta-blockers, which lower portal
venous pressure in the portal vein. A patient with varices should avoid
using aspirin, which can increase the risk of bleeding. If variceal swelling
worsens, varices can be banded using an endoscopic approach to prevent
further swelling. Coagulopathy Coagulopathy (abnoral blood clotting) results when the liver is unable to
produce clotting factors that are needed to control bleeding. The lack of
clotting factors results in so called THIN BLOOD and increases the risk of
bleeding. Warning signs of mild coagulopathy problems include frequent nose
bleeds, prolonged bleeding in of the gums, and bruising easily. A patient
should call a physician if any of these warning signs are present. Renal insufficiency (worsening kidney function) or hepatorenal syndrome
(HRS) occurs as a result of a complex series of events. When blood vessels
enlarge (dilate) in the gut as result of portal hypertension, low blood
pressure results. To compensate for the low blood pressure, the kidneys act
to conserve (retain) sodium in the body, which increases blood pressure. But
when no more sodium is available and the kidneys can no longer compensate,
the falling blood pressure leads to kidney damage secondary to the liver
(hepatic) damage. Encephalopathy Hepatic encephalopathy (HE) is defined as a liver related change in the
brain activity and consciousness. This results when the liver is unable to
perform its normal function and hepatic toxins (poisons) such as ammonia
accumulate in the blood and adversely affect the brain activity. Signs of
mild encephalopathy range from forgetfulness and tiredness to an inability
to focus. What you can do now As always , there are many important things you can do to help minimize
the likelihood of potential damage associated with serious complications as
a result of decompensated liver disease. Stay informed about your condition
and about new and developing therapies. Get support form family, friends,
counselors and other patients to maintain the very best quality of life.
http://www.dr-zhang.com/columns/c26_bleeds.htm
http://www.liverfoundation.org Research Topics http://www.ikp.unibe.ch/lab2/lab2.html Regeneration and cell death in cirrhosis http://www.ikp.unibe.ch/lab2/REG/index.htm Latest Report on Cirrhosis http://www.reutershealth.com/wellconnected/doc75.html Best Message boards on Cirrhosis http://forums.delphiforums.com/liverfailure/start Clinical Significance of Variceal Hemorrhage in Recent Years in Patients With Liver Cirrhosis and Esophageal Varices
Posted 08/27/2004
Dong Kyu Park; Soon Ho Um; Jae Won Lee; Jung Bok Lee; Young Sun Kim;
Chul Hee Park; Yoon Tae Jin; Hoon Jai Chun; Hong Sik Lee; Sang Woo Lee;
Jai Hyun Choi; Chang Duck Kim; Ho Sang Ryu; Jin Hai Hyun
Abstract and IntroductionAbstractBackground and Aims: Recent progress in the treatment of
variceal bleeding might have reduced the impact of variceal bleeding
on survival in patients with esophageal varices. We conducted a
retrospective cohort study in an attempt to re-evaluate the clinical
significance of variceal bleeding. IntroductionEsophageal varices are the most common clinical manifestation of portal hypertension in patients with liver cirrhosis. It is known that once bleeding occurs in patients with varices the prognosis is extremely poor, with 30-50% of patients dying within 6 weeks of the first variceal hemorrhage. Among those who survive the first hemorrhage, 47-84% show recurrent bleeding and 70% die within the first year.1-5 Thus, various prophylactic measures such as shunt operation, use of beta-blockers, endoscopic sclerotherapy and endoscopic variceal ligation have been attempted to increase the survival rate by preventing the first variceal hemorrhage.4-6 Based on the results of these trials, oral administration of beta-blockers is widely used as a primary prophylactic method against first variceal bleeding. The effectiveness, however, of each prophylactic measure is still controversial, and the survival benefit of prophylaxis has not yet been clearly shown. In one study, beta-blockade was shown to increase the bleeding-free survival only in patients without ascites, whereas in another study it was only effective in patients with ascites.7,8 Notable progress in the treatment of variceal bleeding has possibly reduced the fatality rate of variceal hemorrhage and might weaken the necessity for prophylaxis. In the present study, we re-evaluated the clinical significance of variceal bleeding throughout the natural history of cirrhotic patients, who were managed in our department during the last decade, and tried to determine the impact of variceal bleeding on survival. MethodsSelection of PatientsThe study was conducted with a retrospective cohort of 304 patients with liver cirrhosis and esophageal varices who were newly diagnosed at our department (Korea University Anam Hospital, Seoul, Korea) from March 1991 to March 1999 and followed up consecutively. This included all patients who fulfilled the following inclusion criteria: (i) presence of esophageal varices without liver cancer or life-threatening extra-hepatic diseases at the time of diagnosis of liver cirrhosis; (ii) absence of previous gastrointestinal bleeding; and (iii) no prophylactic treatment against variceal bleeding during the study period. The entry time was defined as the date of diagnosis of esophageal varices. A total of 1139 patients with previously unrecognized liver cirrhosis were admitted to our department during the same period. Among them, 395 already had hepatocellular carcinoma (HCC) when cirrhosis was diagnosed, 220 presented no varices in endoscopic study, 141 had previous or ongoing gastrointestinal bleeding at first presentation, and two were suffering from chronic renal failure. Thus, the remaining 381 patients had esophageal varices without previous variceal bleeding or any concurrent life-threatening diseases at the time of diagnosis of liver cirrhosis. However, 77 of these were given prophylactic treatments against variceal bleeding immediately after diagnosis of varices. Therefore, 304 patients were finally considered to be suitable for the purpose of this study. Among the final cohort, 58 patients at first presentation, and an additional 57 during follow up, showed large varices with red color signs. Thirty-six patients received the prophylaxis against first variceal bleeding at 1-79 months after entry, and we considered them to be censored at the time of prophylaxis. The remaining 79 patients were not given active prophylactic measures because they served as a control group in the clinical trial for endoscopic prophylaxis. Assessment of Characteristics at EntryDiagnosis of liver cirrhosis and esophageal varices was made during admission for work-up of abnormal physical findings or abnormal biochemical data suggesting advanced chronic liver disease. At admission, all patients had a full history taken, a complete physical examination and endoscopic examination. Also, a blood biochemistry test, prothrombin time, blood cell count and viral markers were checked. The presence of ascites was confirmed with abdominal ultrasonography or computed tomography (CT). Diagnosis of liver cirrhosis was made by liver biopsy or imaging. When the patients consumed more than 80 g of alcohol per day for more than 5 years, they were considered to have a heavy alcohol consumption habit. The overall degree of hepatic decompensation was assessed using Child-Turcotte-Pugh grades.9 The varices were described according to the general rules for recording endoscopic findings on esophageal varices proposed by the Japanese Research Societies for Portal Hypertension, with some modifications.10 Briefly, the varices were classified by form (F), location (L), color (C), red color sign (RC sign) and gastric varices. Form of varices was divided largely into two groups; small (F1) or large (F2 or F3) varices, and RC sign was also divided into two groups; positive or negative. Positive was determined if there was clear evidence of a cherry red spot, red whale marking or a hemocystic spot. Follow UpFollow up was usually conducted in the outpatient clinic at 3-monthly intervals, but more frequently if needed for the patients with signs of decompensation such as ascites, encephalopathy or jaundice. At the time of follow up, the patients underwent physical and laboratory examinations to assess the change in hepatic function and the development of complications. For surveillance of HCC, serum alpha-fetoprotein and ultrasonography were checked every 3 and 6 months, respectively. Follow up was continued until death or until the time of prophylaxis against variceal bleeding. The study was terminated on 31 August 1999. During the study period, 111 of 304 patients died, 171 remained alive, and 22 (7.2%) patients dropped out. For the whole series, the median follow-up duration was 30 (range: 1-101) months; 42 (13.8%) were followed for less than 12 months, 68 (22.4%) for 12-23 months; 69 (22.7%) for 24-35 months; 48 (15.8%) for 36-47 months, 34 (11.2%) for 48-59 months, and 43 (14.1%) for 60 months or more. For patients who dropped out, median follow-up duration was 22 months (range: 8-50). Variceal BleedingIf patients had any evidence of gastrointestinal bleeding, they were hospitalized immediately and underwent an emergency endoscopy. These patients were considered to have variceal bleeding if blood was emanating from the varix or clots were adherent on the surface of the varix, and blood was seen in the stomach without any other source of bleeding besides the varix.11,12 In case of variceal bleeding, patients underwent emergency endoscopic treatment or supportive treatment (balloon tamponade and/or somatostatin or vasopressin) followed by elective endoscopic treatment. Endoscopic treatment was aimed at the eradication of varices using band ligation, sclerotherapy, or both. Additional treatments were repeated if needed. Death was regarded as being related to variceal bleeding when it occurred within 6 weeks of the first variceal bleed.11,12 Management of Other ComplicationsAscites was controlled with sodium restriction and, if needed, with diuretics and/or therapeutic paracentesis. Encephalopathy was usually managed as an in-patient with protein restriction, withdrawal of diuretics, lactulose and oral antibiotics, as well as correction of precipitating factors. When mass lesions had been suggested by ultrasonography or serum alfa-fetoprotein, the patients were hospitalized for further examinations such as CT, hepatic angiographies and biopsy. Forty patients, in whom 38 the cirrhosis was related to chronic viral hepatitis, were found to have HCC during follow up. These patients underwent specific treatments, such as partial hepatic resection in two, transarterial oily chemoembolization in 16, combined treatments of percutaneous ethanol injection and chemoembolization in 13, and conservative treatments in nine patients. Data AnalysisUnivariate and multivariate analysis were carried out to evaluate the prognostic roles of various clinicalparameters in first variceal bleeding or survival. In univariate analysis, we screened the data of 20 variables at entry by comparing cumulative bleeding rates or cumulative survival rates among subgroups that were defined by different levels of a given variable. Survival curves were computed by the Kaplan-Meier method and compared using the log-rank test. Candidate variables were chosen according to their known or potential relevance to variceal bleeding and survival. Comorbidities were included in the candidate variables because these or relevant treatments might affect variceal bleeding or death. The variables yielding continuous values were categorized into two levels according to their median value in univariate analysis, while they were introduced intact as continuous values in multivariate analysis. To identify independent predictors of the development of first variceal hemorrhage or survival, the time-fixed variables satisfying proportional hazard assumption and reaching statistical significance (P < 0.05) in the univariate analysis were introduced in the multivariate analysis using the Cox regression model or an extended model fitted by Cox and Oakes,13 respectively. In addition, the prognostic impact of variceal bleeding and HCC on survival was also analyzed by including them as time-dependant variables in Cox and Oakes' model. In this model, the survival time after bleeding was introduced as an additional time-dependent covariate to elucidate the potential increase in the risk of dying immediately after variceal bleeding. Basically, multivariate analysis was conducted using step-wise regression, in which both significance levels for entering and for staging in the model were 0.15. Risk factors for death from variceal bleeding were assessed using a contingency table. Comparison of proportions was determined by the chi-squared test or by Fisher's exact test when the number in a cell was 5 or less. All analyses were performed using the SAS program (SAS Institute, Cary, NC, USA). The significance level was set at 0.05, and the two-tailed test was conducted. ResultsData at PresentationIn the whole series, 222 (73%) out of 304 patients were male and the median age was 52 years (range: 22-82 years). Liver cirrhosis was related to hepatitis B virus in 192 (63%) patients, alcohol ingestion in 65 (21%) and HCV in 31 (10%) patients; cirrhosis was cryptogenic in 16 (5%) patients. According to Child-Pugh grading, 153 (50%) were classified as Child A, 113 (37%) as Child B and 38 (13%) as Child C. The varix was large in 140 (46%); RC sign was observed in 59 (19%); and gastric varices were found in 76 (25%) patients. The main comorbidities found were diabetes mellitus in 43 (14%), cholelithiasis in eight (3%), and hypertension in 10 (3%) patients. Incidence and Source of BleedingSixty-seven of 304 patients bled from the upper gastrointestinal tract during follow up, and 64 of these were given emergency endoscopies. The sources of bleeding were varices in 52 (81%), gastric ulcer in seven (11%) and duodenal ulcer in five (8%) patients. Variceal hemorrhages originated from esophageal varices in 38 and from gastric varices in 14 patients (cardia in 13, fundus in one). In the three patients who had not undergone endoscopies at the time of bleeding, variceal bleeding was highly suggested in the later endoscopies. Thus, the total number of variceal bleeds was regarded as 55 (52 verified endoscopically at time of bleeding and three attributed to varices). First variceal bleeding occurred within 2 years after inclusion in 39 (71%) of 55 bleeders. The cumulative incidence of first variceal bleeding was 6.6%, 21.9% and 26.2% at 1, 3 and 5 years after inclusion, respectively. Risk Factors for First Variceal BleedingUnivariate analysis revealed that 10 of 20 clinical or endoscopic variables were significantly associated with first variceal bleeding (Table 1). These were sex, presence of heavy alcohol consumption, etiology of liver cirrhosis, serum albumin level, ascites, Child-Turcotte-Pugh grade, form of varix, location of varix, RC sign and the presence of gastric varix. Of these, form of varix, RC sign and the presence of heavy alcohol consumption were only found to have independent relationships with first variceal bleeding in multivariate analysis by Cox's regression model (Table 2). Using three independent risk factors, we divided the whole cohort into three risk groups for first variceal hemorrhage. If the patient had all of the three following risk components: a large varix, positive RC sign and the presence of heavy alcohol consumption, he or she was assigned to the 'very-high-risk group'. A patient with two components was allocated to the 'high-risk group', and a patient with a single component or none fell into the 'low-risk group'. The cumulative incidences of first variceal bleeding and variceal bleeding related death were significantly (P < 0.0001) different among the three risk groups (Fig. 1, Table 3). For example, 3-year incidences for first variceal bleeding and bleeding related deaths were 9.5 times and 8.9 times higher in the 'very-high-risk group' than in the 'low-risk group', respectively.
In addition, among 132 patients with a habit of heavy alcohol
ingestion at entry, those who stopped drinking during follow up
showed a significantly lower bleeding incidence than those who
continued drinking (4/43vs31/89,P= 0.04 by log-rank
test). Of the patients with alcoholic cirrhosis, all the variceal
bleeders continued to abuse alcohol (20/20), whereas 24% (11/45)
of the non-bleeders abstained from alcohol after entry.
Clinical Course After First Episode of Variceal BleedingSix variceal bleeders succumbed to uncontrollable hemorrhage at their first episode. Within 6 weeks of their first hemorrhage, an additional five patients died from liver failure worsened by hemorrhage, and six patients experienced early rebleeding, three of whom died. Thus, the mortality related to first variceal bleeding was 25% (14/55). Twelve of the remaining 41 survivors had late recurrent bleeds, three of whom died, and 22 died of other causes during follow-up. The cumulative rebleeding rates were 12.2%, 24.4%, 30.2%, 42.9% and 59.8% at 6 weeks, 6 months, 1, 3 and 5 years, respectively. The cumulative survival rate after first variceal bleeding was 75%, 67%, 58%, 37% and 27% at 6 weeks, 6 months, 1, 3 and 5 years, respectively. Characteristics of Patients With Death Related to First Variceal BleedingOf 14 patients who died within 6 weeks of their first variceal bleed, four (29%) already suffered from end-stage HCC infiltrating the main trunk of the portal vein at the time of bleeding; 11 (79%) had Child grade C cirrhosis; 11 (79%) had refractory ascites; and seven (50%) had recurrent encephalopathy before variceal bleeding developed. Univariate analysis clearly showed that the mortality within 6 weeks of bleeding was associated with the degree of liver decompensation at the time of bleeding (Table 4). For example, there was no death in the patients corresponding to Child grade A, whereas those with Child grade C showed approximately 55% mortality. Cause of Death in the Whole SeriesIn the whole series, a total of 111 patients died during follow up. As shown in Table 5, hepatic failure accounted for 50 deaths (45%); progression of HCC for 24 deaths (22%); variceal bleeding for 17 deaths (15%); and other minor causes for 20 deaths (18%). Overall, variceal bleeding was less responsible for death than HCC, but it was a more dominant cause of death in patients with alcoholic cirrhosis. Survival and prognostic indicatorsThe cumulative survival rate in the whole group of patients was 93.7%, 67.4% and 49.5% at 1, 3 and 5 years after inclusion, respectively. Univariate analysis in the whole series revealed that 12 of the 20 clinical variables considered at entry were significantly associated with survival. These included age, heavy alcohol consumption, serum levels of albumin, bilirubin, and aspartate aminotransferase, prothrombin time, ascites, encephalopathy, Child-Turcotte-Pugh grade, platelet count, the form of varix and RC sign (Table 1). Also, the development of two major complications, variceal bleeding and HCC, during follow up significantly correlated with a decreased survival rate (Table 6). Of these, age, Child-Turcotte-Pugh grade or its three components (albumin, ascites and encephalopathy), platelet count, variceal hemorrhage and HCC were found to have statistically significant correlations with death in multivariate analysis using Cox-Oakes' regression model (Table 7). As far as variceal hemorrhage is concerned, it was associated with increased risk of death by a factor of 3.9 in the whole cohort. However, the hazard ratio due to variceal bleeding declined exponentially as time passed after bleeding, as shown in Table 7. The prognostic impact of variceal bleeding was further analyzed in subgroups of patients according to the etiology of cirrhosis or liver function at baseline, because different clinical courses are expected depending on such characteristics. When other prognostic factors were adjusted in the Cox-Oakes model, variceal bleeding was more strongly linked with death in patients with alcoholic cirrhosis (hazard ratio: 26.0) than in those with non-alcoholic cirrhosis (hazard ratio: 3.2), and showed a significant association with mortality for patients in Child grade B, but not in those in grade A or C (Table 7). Furthermore, the difference in hazard ratios associated with variceal bleeding between alcoholic and non-alcoholic cirrhosis proved to be statistically significant (P= 0.004) in the model for the whole cohort, including a covariate denoting etiology of cirrhosis and its interaction with variceal bleeding. Discussion
In the present study, we were able to confirm several previous
observations concerning variceal bleeding in cirrhotic patients,
although there were some differences in the etiology of
cirrhosis between the study groups. We found that, as reported
in Western studies,14-17 variceal bleeding accounted
for more than 80% of first upper gastrointestinal bleeds in
patients with esophageal varices, and most of these occurred
within the initial 2 years from entry. Also, we demonstrated
that three well-known risk factors for variceal bleeding; the
size of varix, RC sign and heavy alcohol consumption, had
independent correlations with first variceal hemorrhage. In our
series, the patients with varices larger than F2 showed
approximately three times higher bleeding risk than those with
F1 varices; RC sign increased the bleeding risk up to four
times; and the patients with heavy alcohol consumption carried
double the risk of those without it. These results are
comparable to those of earlier reports.16-18
Generally, the size of varix and RC sign have been thought to be
closely related to the tension in the variceal wall,19
and alcohol has been proposed to increase both the intrahepatic
vascular resistance and portal blood flow, and thereby to
elevate portal blood pressure and intravariceal pressure.19,20
It is of note in this study that the incidence of variceal
bleeding in heavy drinkers significantly decreased when they
stopped drinking. This confirms the previous observation by
Dagradiet al.21 and supports an important role
of alcohol in variceal bleeding. Conversely, Child grade and
gastric varices, which were independent indicators of bleeding
in some studies,17,22 did not provide any more
independent information than the above three factors in the
present study.
The recognition of risk factors for variceal bleeding enabled us to discriminate a group of patients at high risk of bleeding. In the present study, we graded each risk factor into two levels for simplicity, and sorted patients into three major risk groups depending on the presence of a large varix, RC signs, or heavy alcohol consumption at entry. Seventy-six per cent of the patients with all three risk factors and 45% of those with two risk factors developed bleeding from varices during the 3 years from entry, as compared with 8% in the remaining patients. This suggests that, considering the distressing nature of variceal hemorrhage, it would be worthwhile to try primary prophylaxis for patients with two or more risk factors. By preventing first variceal hemorrhage, how much could we expect survival to increase? Disappointingly, our results suggest that the overall survival benefit expected in all cirrhosis patients would not be so high, even if assuming we had the perfect prophylactic measures. In this study, the difference in 3-year survival between bleeders and non-bleeders was no more than 17% when calculated by the Kaplan-Meier method. Although the risk of death in variceal bleeders was initially 3.6-fold higher than that in non-bleeders in Cox-Oakes' model, it decreased exponentially over time after bleeding. Perhaps this result might reflect the low bleeding related mortality and the low rebleeding incidence in our study. In this series, the mortality of the first variceal bleeding was 21% (11/55) when excluding the cases with concurrent terminal liver cancer, and the 1-year rebleeding rate was only 19% in the survivors of the first bleed. These figures are considerably lower than those reported in the past (mortality 30-70%; rebleeding rate 47-84%),1,4,5 and confirm a similar trend in a recent large study.23 We speculate that this low impact of variceal bleeding on mortality might, in part, explain why beta-blockers have failed to improve survival in many trials to date, despite an appreciable reduction in bleeding risk. It should, however, be noted that the prognostic impact of variceal bleeding was dissimilar in different subsets of patients depending on the etiology of cirrhosis or on the grade of hepatic decompensation at entry. Variceal bleeding had a more potent association with survival in patients with alcoholic cirrhosis than in those with non-alcoholic cirrhosis, and had a significant prognostic impact only in patients of Child grade B, but not in those of grade A or C. These results are interesting because they suggest that the prophylaxis against variceal bleeding might be more cost-effective in such groups of patients. Indeed, Sauerbruchet al. reported that prophylactic sclerotherapy improved survival only in patients with alcoholic cirrhosis and in those of Child grade B.24 The strong prognostic impact of variceal hemorrhage in patients with alcoholic cirrhosis might, to some extent, be ascribed to a low incidence (1/65) of HCC in these patients, which might augment the relative influence of variceal bleeding on mortality. The lack of prognostic impact of variceal bleeding in Child grade A or C patients might be, in part, due to a lack of statistical power derived from small sample sizes in this study. This phenomenon, however, could be explained by the low bleeding mortality in the Child grade A group and by the grave prognosis in the Child grade C group. In the present study, there were no deaths among those in Child grade A at the time of variceal bleeding, and only two (9%) bleeding related deaths among 23 variceal bleeders who had been in Child grade A at entry. Conversely, the chance of long-term survival in Child grade C patients appears to be very slim because of severe hepatic decompensation, even if they do not encounter variceal bleeding. In fact, their survival was not affected by HCC, either. In conclusion, this study validated the usefulness of risk factors for variceal bleeding by discriminating a subgroup with a high risk of bleeding. It showed that variceal bleeding still had a significantly adverse impact on survival in cirrhotic patients with esophageal varices, particularly those with alcoholic cirrhosis and moderately decompensated cirrhosis, despite the decrease in fatality of variceal bleeding. These results suggest that prophylaxis against variceal bleeding would be most cost-effective in such patient groups.
Reprint Address
Correspondence: Dr Soon Ho Um, Department of Internal Medicine, College of Medicine, Korea University, 126–1, 5 ga Anam-Dong, Sungbuk-Gu, Korea University Anam Hospital, Seoul, 136–705 Korea. Email: umsh@korea.ac.kr
|
Reviewed Sep 03 2004
