Prevalence of hepatitis A antibody in children and adolescents with chronic liver disease

Prevalência do anticorpo da hepatite A em crianças e adolescentes com hepatopatia crônica
Cristina T. Ferreira, Adriano N. R. Taniguchi, Sandra M. Vieira,
Jorge P. Lima, Themis R. da Silveira
J Pediatr (Rio J) 2002;78(6):503-8


Acute viral hepatitis A (HAV) is a ubiquitous infection that results in significant morbidity. From an epidemiological point of view, the rapid dissemination and the large outbreaks of HAV, in addition to the fact that there is no evolution to a chronic form, distinguish it from Hepatitis B and C. (1) Infection by the A virus can also cause fulminant hepatitis and be fatal, with an estimated mortality rate of 0.14 to 2%. (2)

There is growing evidence in medical literature that individuals with chronic liver disease exhibit an increased risk of developing more serious forms of HAV or even fulminant hepatitis. (1,3)

Recipients of liver transplants should avoid further lesions to the graft, and should be immunized against HAV, even when the risk of exposure is low. (4) Furthermore, patients with chronic liver disease, who require liver transplants, should undergo all vaccinations in advance of the surgical procedure, since afterwards they will be more immunocompromised than before the transplant. For this reason, this specific group of patients is indicated for vaccination against viral hepatitis A. (4,5,6)

Children with chronic liver disease would theoretically benefit from the HAV vaccine, since it has been shown to be immunogenic and safe in controlled studies with immunocompetent children. (6,7,8) The fact that the vaccine is safe and effective in healthy adults and children does not guarantee similar results in populations with chronic liver disease. (1) Studies performed with adults suggested that while patients with chronic liver diseases are not at increased risk for infection, they have a greater risk of fulminant hepatitis caused by the A virus. (2,9,10) American death records indicate a greater prevalence of chronic liver disease in people who died from fulminant HAV than in people who died from other causes. (5,6,9,10)

Neither an immunization scheme nor a seroconversion rate for populations with chronic hepatic diseases have been defined yet. (1)

The purpose of this study was to evaluate the prevalence of HAV in children and adolescents with chronic liver diseases at a pediatric hepatology service.

Patients and Methods

A study with 60 children and adolescents with chronic liver disease was conducted between May 1999 and February 2001. The patients belonged to the Pediatric Gastroenterology Unit and the Pediatric Liver Transplant Project of the Pediatric Service of the Hospital de Clínicas de Porto Alegre. This is a tertiary care hospital, considered a benchmark in Pediatric Hepatology.

Patients with chronic liver disease between one and 16 years of age, with no previous history of hepatitis A or vaccination against HAV, and not carrying the human immunodeficiency virus (HIV) were included. They were all being assessed or had already been assessed for liver transplants, some of them were already on the waiting list, others yet to be referred at the time of assessment. The patients suffered from chronic liver diseases confirmed histologically by liver biopsy or ultrasound scans and laboratory tests.

This was a cross-sectional study which evaluated the presence of the anti-HAV (total) antibody, a marker of hepatitis A infection. The total anti-HAV was determined and assessed according to age, sex, color, etiological diagnosis of liver diseases, and patient's family income.

Ages varied between 12 months and 16 years, with an average (+/- SD) of 7 years (+/- 4.9). There were 25 children between one and four years old, nine between 5 and 8, 19 between 9 and 12 and seven adolescents between 13 and 16 years old.

Thirty-three children (55%) were female and 92% were white.

The etiologic diagnoses of the liver diseases are shown in Table 1.

Table 1 -
Etiologic diagnoses of the chronic liver diseases

The average age of patients with extrahepatic biliary atresia (EHBA) was 3.9 years; for the patients in the autoimmune group it was 11.8 years; for those with cryptogenic cirrhosis it was 6.2 years and for the others it was nine years.

The family income of these patients varied between 150 and 3,000 reais (40 to 800 US dollars), with a mean average of 595.50 reais (US$ 165). The family income of 73% of the patients was less than or equal to 500 reais (US$ 138), that of 63% of the patient's families was below R$ 300 (US$ 83). The minimum salary at this time was R$ 136 (US$ 50). The median of family incomes was 300 reais with interquartile ranges of R$ 200 (25th percentile) and R$ 675.00 (75th percentile).

The anti-HAV was obtained using a laboratory test which is commercially available in our setting (Abbott - HAVAB MEIA - AXSYM system). Venous blood was collected normally in combination with the patient's routine blood sampling, and was centrifuged and frozen for later evaluation. The anti-HAV (total) tests were all performed in the same clinical analysis laboratory (Laboratório Weinmann), in Porto Alegre. The results were considered positive, negative or indeterminate according to positive and negative controls. A cutoff value was calculated, below which the sample was positive, and above which it was negative. Tests with an absorbance within the gray scale, which corresponds to more or less 10% of the cutoff value, were considered indeterminate. These tests were repeated, using a fresh sample, and if they continued to yield values within the gray scale, i.e. indeterminate, they were eliminated from the study.

The two groups (anti-HAV positive and negative) were compared using the chi-squared test with Yates' correlation (for the sex variable), Fisher's exact test (for color), the Student's t test (for age) and the Mann-Whitney U test (for family income). In addition, the chi-squared test for linear trend was used to check the association between anti-HAV and increase in the age of the patients. The significance level was set at 0.05.

The study was approved by the Ethics and Research Committee of the Hospital de Clínicas de Porto Alegre. All of the patients' parents or legal guardians signed consent forms after being informed of the nature of the study.


Of the 60 children and adolescents with chronic liver disease, only one was excluded because, on two different occasions, the anti-HAV antibody result was indeterminate. Another boy, one year old, also had indeterminate results for the first test, but, when repeated two weeks later, the result was negative.

Of the 59 remaining children and adolescents, 14 (24%) showed positive results for (total) anti-HAV.

The ages of patients with anti-HAV positive results varied from one to 16 years (x = 7.7 years and median of 8.5 years). There was a tendency for the percentage of positive results to rise with age, although there was no statistical evidence ( 2 for linear trend = 0.6) (Table 2). The comparison between the positive and negative groups is shown in Table 3.

Of the children who were anti-HAV positive, chronic autoimmune liver disease was detected in five, cryptogenic cirrhosis in four, EHBA in three, Budd-Chiari syndrome in one and biliary cirrhosis with associated histiocytosis in one.

Table 2 -
Positivity of anti-HVA related to age

Table 3 -
Patients' characteristics (positive and negative anti-HVA)

The family incomes of the group of anti-HAV positive patients were between 150 and 2,000 reais, the mean average being 441.00, a little below that of the group of 45 anti-HAV negative patients (R$ 642.62). Twelve anti-HAV patients (86%) had a family income of less than 500 reais, while 69% of the negative group was within this range. Eleven of the anti-HAV positive children (79%) had family incomes below R$ 300. The median family income of anti-HAV positive patients was R$ 300 (p-25 R$ 150 - p-75 R$ 350).

The ages of the five anti-HAV positive patients with autoimmune liver disease varied between seven and 13 years, with a mean average of 11 years and a median of 12. All were girls and one was black. The mean average of family income for this group was R$ 380 while the median was R$ 300. Two of the four positive patients with cryptogenic cirrhosis were male. Their ages were 1, 3, 6 and 10 years with a mean of 5. Their families' incomes ranged from R$ 150 to R$500 with a mean average of R$ 267.50. The three anti-HAV positive patients with EHBA were 1, 2 and 16 years old. They were all white males. Their family incomes were R$ 300 for two of them and R$ 2.000 for the other (Table 4).

Table 4 -
Patients' age and family income according to the etiology of the chronic liver disease


Different anti-HAV antibody prevalence patterns have been described, showing variations in economic development. (11) In areas with a high endemicity, 90% of children are infected by around ten years of age. The infections are asymptomatic, and viral hepatitis A is not a clinical problem. In areas of medium endemicity the 90% seroprevalence level is not reached before early adulthood. In these communities the disease occurs in different age groups (children, adolescents and young adults). These populations exhibit large epidemics, at regular intervals, which last for long periods of time, or have high sustained levels of the disease for many years. (6,12,13)

Although it is convenient to think of these different patterns as mutually exclusive, they are a simplification of a very complex epidemiology. In many countries the three patterns can occur within different sections of a single community. (14) These general endemicity patterns, therefore, can vary within a single country, city or region. In 1994 we studied the prevalence of HAV in 387 children and adolescents between one and 19 years old in Porto Alegre. (13) In this first study we found a highly significant difference in prevalence between children from a low socioeconomic background and those from a high level (51% against 11%). This shows that, while HAV is endemic in our setting, there is a significant number of people who are susceptible to infection: 89% of the children up to 19 years old from a high socioeconomic background, and 49% of those who live under less favorable socioeconomic conditions. (13) In the same study, the prevalence of anti-HAV was 60% and 11% for children up to 14 years old, from low and high socioeconomic classes, respectively. Other studies have been performed in Brazil which show different prevalence rates, depending upon location, all of which are higher than those found in Rio Grande do Sul. (15) In the current study, the prevalence of anti-HAV was 24% up to 16 years of age, lower than the earlier study, probably due to differences in the socioeconomic background of the individuals studied. As the criteria employed to evaluate socioeconomic level in each study were different, comparison becomes difficult.

HAV, following chronic liver disease, has been associated with more serious cases or with fulminant forms. (1,2,9,10,16,17) It is possible that the acute insult caused by viral hepatitis can result in substantial involvement of the hepatic functions already compromised by chronic lesions. The exact incidence of hepatic insufficiency due to HAV in chronic liver disease is not known. (17)

Emmet Keefe (9) reviewed cases of acute hepatitis A in populations with underlying chronic liver disease in epidemics which occurred in Taiwan, in 1988, and in the United States, from 1983 to 1988. He also revised smaller series and case reports. According to this author, the literature suggests that HAV, following a chronic B infection, is associated with altered laboratory results and more serious illness, including the fulminant form, and also a higher mortality rate amongst such patients. Furthermore, there also appears to be a higher mortality rate related to HAV in patients with other types of chronic liver disease. (9,17)

The analysis of epidemiological data reported to the CDC (Centers for Disease Control and Prevention), in the United States, from 1983 to 1988, revealed a hepatitis A mortality rate of 11.7% in patients diagnosed with chronic hepatitis B, and 4.6%, in patients with other pre-existing chronic liver diseases. Deaths occurred mainly within the older population (72% of those who died were over 49 years old). These rates are, respectively, 58 and 23 times higher than in the population without underlying chronic liver disease. (9,10,18)

All of these studies have methodological problems which limit the possibilities of comparison and generalizations. Nevertheless, there appears to be evidence of an increased risk of death in patients with chronic hepatitis when they contract HAV, particularly in population-based studies which survey deaths. (10)

The association between death from fulminant hepatitis A and chronic liver disease was recently demonstrated in a study based upon death certificates between 1981 and 1997 from the National Health Statistics Center. (18) In this case-control study, 63% of the deaths due to HAV mentioned chronic liver disease on the death certificate, compared to 8 to 11% of deaths due to other gastrointestinal causes.

Considering the 1988 epidemic in Taiwan, in which more than 300 thousand people were contaminated by seafood with the A virus, the mortality rate for individuals carrying the B virus was 0.05%, which is 5.6 times higher than for patients without chronic liver disease type B. In this epidemic, the affected population was younger, which perhaps explains the lower mortality. (9)

Sandro Vento et al. (02) performed a seven-year prospective study of 595 adults with either chronic hepatitis B or C, previously seronegative for HAV. Twenty-seven patients (10 with HBV and 17 with HCV) presented acute hepatitis A during this period. Fulminant hepatitis occurred in 7 hepatitis C patients, and only one individual survived. They concluded that patients with chronic hepatitis C run a substantial risk of death when associated with a superinfection by the A virus.

Evangelos Akriviadis and Redeker (19) describe four adult patients who died after developing HAV. All four, on necropsy, displayed underlying chronic liver disease.

In contrast to these and other series of individuals, certain authors do not indicate any difference in prognosis for patients who had chronic liver disease when they acquired the A virus, mainly in those who are younger. (16) Viola et al. (20) and Zachoval et al. (21) did not encounter differences in bilirubin or transaminases, nor in the evolution of chronic patients with B virus acutely infected by HAV. Both suggested that there could be some type of viral interference, since they recorded reduction in the patients' HBsAg levels during HAV. Helbling and Kammerlander, (22) in Switzerland, did not find an association between fulminant hepatitis A and underlying chronic hepatitis B or C.

In over 6,000 cases of HAV notified between 1992 and 1996 to SEIEVA, acute viral hepatitis notification system in Italy, no deaths were recorded. (23) Amongst these HAV cases there were 179 individuals with chronic hepatitis B or C.

Since 1996, the ACIP (Advisory Committee on Immunization Practices to the CDC in the United States) has recommended the immunization of all patients with chronic liver disease against HAV. (5,6,24) The recommendation is based upon the assumption that if an individual already has a chronic liver disease, an additional hepatic lesion might be poorly controlled.

Currently, immunization guides clearly state that patients with chronic liver disease should receive the HAV vaccine. (5,6) The World Health Organization and the CDC recommend immunization against HAV in people with chronic liver disease due to the increased rate of mortality amongst those who contract an infection by the A virus. (5,6,24). The American Academy of Pediatrics has indicated vaccination against HAV for patients with chronic liver disease since 1996, (25) and so has the Brazilian Society of Pediatrics. (26,27)

In order to indicate vaccination against HAV for patients with chronic liver disease, just as for any other group of individuals, it is necessary to know the prevalence of infection for that specific group. (28,29)

In the studied sample, of 59 patients with etiologically different chronic liver diseases, 24% tested positive for the anti-HAV antibody. This is a low percentage considering age and the fact that the majority of the group have family incomes of less than R$ 500.00. There was a slight increase in the percentage of anti-HAV positive patients with relation to age, however this was not statistically significant. In addition, there appears to be a tendency for those patients who are positive to belong to low-income families. However the difference is not statistically significant, which probably results from the small number of patients. The relationship with diagnosis is probably explained rather by age than by etiology. Thus the patients with autoimmune hepatitis, with a higher average age, showed a higher rate of positive anti-HAV tests. Although the group of patients with biliary atresia is larger, it showed less positive results, as the average age was lower than those of the other two groups (3.9 years against 11.8 in the autoimmune hepatitis group and 6.2 years in the cryptogenic cirrhosis group).

Koçak et al. (28), from Turkey, found a seropositivity rate for anti-HAV of 44% amongst 403 children and adolescents with chronic liver diseases, between 1.5 and 20 years old. Of these, 267 suffered from chronic hepatitis B, and 136 had other chronic liver diseases. The 140 children who had no liver disease showed 31% seropositivity for anti-HAV. The prevalence of anti-HAV antibodies increased with age: 16% of the children less than 5 years old, 35.5% in those between 6 and 10, 59.5% in those between 11 and 15 years old and 67% in patients older than 15.

Diago et al. (29), in Spain, studied the prevalence of anti-HAV in 425 adult patients with liver disease (x = 40 years old), and found that 75% were positive, ranging from 20% at 19, to 93% in those who were over 40 years old.

The reduction of a population's susceptibility, through vaccination, can eliminate diseases and their complications, but periodical updating of seroepidemiological data is necessary to assess natural immunity levels and to identify those who are susceptible to infection.

Vaccinating patients with chronic liver diseases, in order to protect them from hepatic lesions caused by the A virus, appears to be a correct and intelligent decision. (1,10,22) This study shows that the majority (76%) of pediatric patients with chronic liver disease are anti-HAV negative, which makes them susceptible to infection by the A virus and to its possible complications. It should also be taken into account that, in countries such as Brazil, the risk of acquiring the infection is very high, since the virus is endemic and there exist people who are susceptible. (13,15)

Neither an immunization scheme nor a seroconversion rate for patients with chronic liver diseases are well established. Studies of anti-HAV prevalence, with later vaccination of patients with chronic liver diseases, can assist in the establishment of immunization criteria for this group of individuals.


Thanks to Laboratório Weinmann for performing the anti-HAV tests and to Vânia Hirakata for the statistical analysis.