Helicobacter pylori Infection in Children: A Consensus Statement

On September 1–2, 1998, a group of pediatric gastroenterologists, mainly from Europe, and other specialists working in the field met in Budapest to discuss the management of Helicobacter pylori infection in children. It was recognized that there is still very little information available about the clinical relevance of H. pylori infection in most children. However, there is a need for some guidelines. The guidelines that have been published by several groups in relation to adults (1,2) are not applicable to children. Furthermore, the increasing availability of diagnostic tests for H. pylori infection, for use in primary care practice, is likely to result in the identification of an increasing number of children with this infection. This conference was not intended to provide a comprehensive overview of all matters relating to H. pylori infection in children. At this meeting the discussion was focused on the indications for investigating children for H. pylori, and the role of noninvasive tests in clinical practice. Consensus was sought initially on a number of fundamental issues. Subsequently, various statements were proposed, and the arguments in favor of and against these statements were presented by different speakers. This was followed by a general discussion and vote in relation to the statement. RAPID CONSENSUS Consensus was readily achieved on these fundamental points: •H. pylori infection causes chronic gastritis in children. •H. pylori infection is associated with gastric and duodenal ulcer disease in children. • Eradication of H. pylori leads to healing of chronic gastritis. • Eradication of H. pylori leads to long-term healing of duodenal ulcer disease. INDICATIONS FOR INVESTIGATING CHILDREN FOR H. PYLORI INFECTION Statement I In the paediatric age group, there is no specific clinical picture of H. pylori infection, indicating a need to use noninvasive tests to screen children with dyspeptic symptoms for H. pylori. There was consensus that to date there is no evidence demonstrating a link between H. pylori–associated gastritis and abdominal pain except in those rare cases in which gastric or duodenal ulcer disease is present. In addition, no data have been published that H. pylori eradication in infected children without ulcer disease is superior to placebo or nontreatment in producing symptom relief. The group therefore considered that screening for H. pylori infection should not be performed routinely in children with upper gastrointestinal symptoms, including abdominal pain. In populations with a high prevalence of infection, such screening leads to treatment of large numbers of children. Although such a “test and treat policy” has been recommended for young adults with dyspepsia but without alarming symptoms (1,2), the relation of benefits to risks and costs of such a policy differ between adults and children, because H. pylori–related ulcer disease occurs with a much lower frequency in the paediatric population. However, assessment of H. pylori status could be performed as part of the evaluation when symptoms were suggestive of organic disease such as peptic ulcer or esophagitis, and the child therefore should undergo an upper gastrointestinal endoscopy. Consensus All agreed with this statement. It was also agreed that further studies are required to establish whether there are any specific symptoms associated with H. pylori gastritis alone and whether infected children without ulcer disease benefit by reduction of their symptoms after H. pylori eradication. Statement II Children should be investigated for H. pylori only when their symptoms are severe enough to justify the risks of therapy. Arguments against the statement: • Noninvasive tests can establish a diagnosis of H. pylori infection easily and accurately. • The antibiotics used for treatment of H. pylori are widely used to treat children for other infections. • Treatment of H. pylori may prevent serious diseases later in life. Arguments supporting the statement: • Studies using the same treatment in children as in adults indicate that some H. pylori eradication regimens may have lower efficacy in children than in adults, although more studies are needed. • Fewer drugs are approved for use in children (e.g., some proton pump inhibitors, bismuth preparations, and tetracyclines are not approved). • There is a risk of side effects from the drugs used. • There is a risk of generating resistant strains of H. pylori that may make it more difficult to treat H. pylori infection in these patients later in life and may lead to an increase in resistant strains in the community. • There is a risk of selective resistance in other bacterial species to the antibiotics used that could limit the choice of antibiotics for other infections in children. • In children, the main goals of H. pylori eradication therapy are to heal ulcer disease, if present, and to relieve symptoms. In infected children who do not show an immediate benefit from eradication, therapy for prevention of serious H. pylori–related disease could be postponed to a later time when safer and cheaper therapeutic options are likely to be available. Consensus Although most conferees recognized that there have been no studies of sufficient size or quality in which side effects and risks of therapy in children have been examined, the statement was accepted by the majority. It is possible, however, that further knowledge of measurable health risks in children or other treatment options such as vaccinations will change the benefit–risk relation and will lead to different recommendations. Statement III Endoscopy is the preferred method of investigation in children with upper digestive symptoms suggestive of organic disease after exclusion of other causes with noninvasive methods (i.e., lactose maldigestion, celiac disease, constipation, liver and biliary system disease). Arguments against the statement: • Endoscopy is an invasive method that may induce psychological trauma, especially in children. • Although the risk of complications with endoscopy related to general anesthesia, sedation, or aspiration is low (less than 1%), this risk cannot be ignored. • The cost benefit of the examination is not obvious, because ulcer disease is rare in children. •H. pylori–related ulcer disease heals when the infection is treated successfully. Arguments supporting the statement: • Endoscopy provides more complete information on the disease process. It allows the identification of causes of pain such as esophagitis and peptic ulceration. • It is the only method by which biopsies can be performed to diagnose gastritis (including H. pylori), eosinophilic gastroduodenitis or esophagitis. It also allows for small intestinal specimens to be obtained to look for celiac disease and other enteropathies. • A gastric biopsy can be cultured to grow H. pylori and subsequently to test for susceptibility to antibiotics, which may be important in choosing treatment. • The necessity for performing endoscopy, which has to be justified by symptoms indicating organic disease, would attenuate the misuse of noninvasive tests and the prescribing of anti-H. pylori therapy in children with nonspecific symptoms. • Compliance with H. pylori eradication therapy may be better after endoscopy, compared with diagnosis of the infection by noninvasive testing. Better compliance results in a lower rate of treatment failures and, therefore, a reduced risk of therapy-induced antibiotic resistance of the surviving bacteria. Consensus The statement was accepted by a majority. Emphasis was placed on the fact that children should be investigated for H. pylori infection only if symptoms are suggestive of organic disease. Statement IV If H. pylori is identified as a result of the endoscopy, treatment for this infection should be offered to the patient. Arguments against the statement: • After exclusion of organic disease by endoscopy, children with H. pylori–related gastritis only should not receive anti-infective therapy because, so far, no benefits have been demonstrated regarding symptomatic improvement after therapy compared with nontreatment. Arguments supporting the statement: •H. pylori–infected persons have a 10% to 15% lifetime risk of development of H. pylori–associated serious organic diseases. Because in a given patient the individual risk for complications can not be estimated, a proven infection should not be left untreated. • In children with symptoms severe enough to justify upper endoscopy the risks of therapy are justified even in the absence of evidence that the symptoms are related to the gastritis. • The knowledge of the infection with gastritis may induce psychological fear of serious consequences (ulcer disease, gastric cancer, spreading of the infection) in the child and the family. Consensus There was consensus that the physician should offer treatment for the infection if a child undergoes endoscopy and H. pylori is identified. However, in the absence of ulcer disease, the patient and the parents should be fully informed about the treatment involved and should know that eradication of H. pylori does not necessarily lead to relief of symptoms. They should be given the option to refuse treatment. THE ROLE OF NONINVASIVE TESTING IN CLINICAL PRACTICE Statement V Serologic tests (enzyme-linked immunosorbent assay, immunoblot), especially quick tests, are not as reliable in children as in adults. Arguments supporting the statement: • In pediatric studies published to date, the results obtained in serologic testing are very inconsistent. In most studies results were not categorized according to different age groups. This is crucial, because in young children the serologic response may be weak or absent. • Most commercial kits have not been validated in children. The cutoff point validated in adults may be too high for children, resulting in low sensitivity of the test (3,4). • Sensitivity, specificity, positive, and negative predictive value of the same test kit may differ in different ethnic or geographic populations (5). • Serology does not distinguish between actual and previous infection, because the antibody titer decreases very slowly after cure. Serology is therefore not appropriate for monitoring the response to treatment because of the slow decrease in antibody after successful treatment. • The interlaboratory reproducibility of the commercially available kits has been questioned (6). • False-positive results can occur after spontaneous eradication. Consensus The statement was accepted. It is possible, however, that improvement and validation of commercial serologic tests in children will make them a useful tool for the diagnosis of H. pylori infection in children and epidemiologic studies in the future. Statement VI The13C urea breath test (UBT) is a reliable diagnostic test. Arguments against the statement: • The UBT has not been validated in a sufficient number of infants and young children below the age of 6 years. • The UBT is likely to be unreliable when antibiotics or acid-suppressing drugs have been taken within 2 to 4 weeks before testing. • The UBT is very costly in some countries, and registered kits are not available everywhere. Arguments supporting the statement: • In the validation studies performed in children 6 years of age or more, the test has given concordant and excellent results regarding sensitivity and specificity, irrespective of the different test protocols that were used (7–10). These protocols were different in relation to: • the dose of tracer used (fixed dose or based on weight of the patient), • the duration of fasting (2 to 4 hours after meals or after an overnight fast), • the need to delay gastric emptying (ice cream, juice, a citric acid solution, or no test meal or drink), • the type of breath sampling (bag or straw), • the time and frequency of breath sampling (between 15 and 40 minutes after tracer ingestion), • the type of breath analysis (isotope ratio mass spectrometry or nondispersive infrared spectrometry). Consensus The urea breath test is an excellent diagnostic test, but there is a need to evaluate the test further in young children, especially in those less than 2 years of age in whom no studies have been conducted. Statement VII In children treated for H. pylori infection, the response to treatment should be monitored with a reliable noninvasive test. Arguments against the statement: • The costs of UBT are high. • If the infection is not cured in a patient with gastritis only but the child becomes asymptomatic after therapy, it leads to a dilemma: Should this child not receive further therapy as other infected children without symptoms, or should the child be re-treated with or without an additional endoscopy? Arguments supporting the statement: • The UBT is reliable and allows for confirmation or not of eradication 1 month after the end of the treatment. However, further studies are needed to increase the number of children evaluated after treatment. • The posttreatment evaluation is an important step, because disappearance of symptoms is not a reliable predictor of successful eradication. • In children with gastric or duodenal ulcer disease, there is a high recurrence rate if the infection persists. • Endoscopy is not justified when only the presence of H. pylori has to be determined. • The physician gets feedback on the efficacy of the treatment regimen. Consensus The statement that the success of treatment should be monitored was accepted. However, there were different opinions about how to treat symptomatic and asymptomatic children who remain infected after treatment (see flowchart, Fig. 1).FIG. 1.: Proposed treatment strategy for children who have upper abdominal symptoms and are infected with Helicobacter pylori. *Dashed lines indicate lack of consensus regarding management of children who remain infected but asymptomatic after treatment.DISCUSSION H. pylori infection is mainly acquired in childhood. Cross-sectional studies indicate that most infections are even acquired before 5 years of age. Infection mainly occurs among children living in developing countries and in children living in poor socioeconomic circumstances in developed countries. H. pylori infection is associated with chronic gastritis in all infected children, and ulcer disease develops in a small number of children. Recently, H. pylori has been classified as a group 1 carcinogen by the World Health Organization. During the past few years, consensus conferences have been held in Europe and North America regarding the diagnosis and treatment of H. pylori infection in adults. To date, no consensus report has been published on the specific issue of H. pylori infection in children. Considering that H. pylori infection is acquired in early childhood, it is important that a consensus be achieved on the management of this infection in the pediatric population. The consensus group included participants from a broad range of countries across western and eastern Europe. There were also a number of people in attendance from North America. Although most of the group comprised pediatric gastroenterologists, there was also input from gastroenterologists, microbiologists, and epidemiologists. It was agreed that in the pediatric age group, there is no specific clinical picture indicating a need to screen for H. pylori. Most infected children are asymptomatic. Consensus was reached on the point that at present there is no evidence to suggest a link between H. pylori gastritis and abdominal pain in the absence of ulcer disease. Recurrent abdominal pain occurs in up to 15% of school-aged children. It was agreed that children with recurrent abdominal pain should not undergo noninvasive or endoscopy-based tests that seek evidence of H. pylori infection. The participants were unanimous in suggesting that children with abdominal pain should undergo investigations for H. pylori only in a situation in which upper endoscopy is performed to look for organic disease such as ulcer disease or esophagitis. Concern was expressed by several members of the consensus group that a decision not to treat children who have chronic gastritis and abdominal pain might result in leaving untreated a small group of children whose symptoms are related to H. pylori infection. In adults, the issue is controversial. Two recent randomized, double-blind, placebo-controlled studies in adults with nonulcer dyspepsia did not show convincing evidence that eradication of H. pylori relieves the symptoms (11,12). However, investigators in a third trial reported a benefit of eradication treatment in a small subgroup of patients (13). Some members of the consensus group thought that a similar subset of children may exist. It is difficult to distinguish this group from the large number of infected children. Therefore, the group suggested that if a child undergoes endoscopy because of severe symptoms and H. pylori is identified, the physician should offer treatment of the infection. Parents should be fully informed that eradication of H. pylori does not necessarily lead to any change in symptoms. They should also be informed of the potential adverse effect of the treatment involved and should be given the option of refusing treatment. It was recognized that there is a contradiction in this approach, because at the same time, it was suggested that screening for H. pylori should not be performed in children with abdominal pain. The group suggested that further studies on symptoms and nonulcer H. pylori infection should be undertaken. The consensus group was particularly concerned about the potential for overuse of noninvasive tests for the detection of H. pylori in children. The availability of 13C-UBT and rapid serology kits places pressure on pediatricians and to screen children for this infection when the patient has abdominal symptoms. There was that such screening should not be because no specific symptoms have been associated with H. pylori gastritis If a child has symptoms suggestive of ulcer or other disease, an endoscopy is required to make a Therefore, a noninvasive diagnosis of H. pylori is not in such on the sensitivity and of serology and for epidemiologic studies in children and of children after treatment. It was agreed that serologic testing is not developed for use in children, especially in young children. infected children may not H. antibody are or years of age that serology in this age group will be very difficult to However, further development of serologic for use in children should be with a placed on serology by using from children than from adults. It was accepted that the 13C-UBT is and specific in children who are more than 6 years of age. in children 5 years of age and particularly in is and should be Furthermore, the 13C-UBT is in the of children to the success of treatment. Concern was expressed by several members of the consensus group that the 13C-UBT may not be a reliable test in children less than 2 years of age. members of the group reported about with results in children less than 2 years of age. There was a consensus that the 13C-UBT to be further validated in very young children before it could be recommended for use in this age The issue on which consensus could not be achieved related to how an infected child should be treated who infected after treatment but whose symptoms have It was agreed that if there is a of previous ulcer disease this child should undergo endoscopy and biopsy of the with and susceptibility testing for and followed by treatment. However, no was reached on how to treat such children in whom no of ulcer disease Some of the group considered that these children should undergo endoscopy in of the decision that been to treat the infection. However, other members of the consensus group thought that if a child were further investigation or treatment should not be symptoms The consensus group did not discuss the treatment for H. pylori infection in childhood, because there are data in the in relation to this There is a need to further studies in of size to different treatment regimens in children. Concern was expressed about the high prevalence of strains that have been identified among children in various countries, as as It was suggested that the lack of validated treatment regimens and the risk of development of resistant strains were further to be about screening and treatment of children for H. pylori infection. The participants at the Budapest Consensus did not discuss the relevance of H. pylori infection in to the development of gastric in life. There was that the World Health statement H. pylori as a group 1 carcinogen would result in pressure for screening of children and treatment if H. pylori is to be present. The of H. pylori infection in children in of the risk of gastric in life further because it is likely to be the issue in whether screening and treatment are among children. At present, however, it is considered that a screening and treatment policy in children would be cost in gastric This issue will have to be in the future. Consensus is needed on the of in children in relation to the risk of developing gastric In this consensus group agreed that there is at present no for screening for H. pylori infection in symptomatic or asymptomatic children. However, it was agreed that when children undergo endoscopy as a result of symptoms suggesting organic disease, could be for H. pylori infection, and in such if H. pylori is identified, it should be The contradiction between these statements was recognized but accepted as a The group that further studies are needed to to whether a subgroup of children infected with H. pylori have symptoms to the infection. The 13C-UBT is an excellent test for use in epidemiologic studies in children, but further validation in young children and especially in those less than 2 years of age. Serologic testing for H. is not in children. Consensus is needed about the of H. pylori infection in children in relation to the risk of development of gastric in life. Consensus was achieved on the •H. pylori infection causes chronic gastritis in children. •H. pylori infection is associated with duodenal ulcer disease in children. • Eradication of H. pylori leads to healing of chronic gastritis. • Eradication of H. pylori leads to long-term healing of duodenal ulcer disease. • In children, so far, there is no evidence demonstrating a link between H. pylori–associated gastritis and abdominal pain or dyspeptic symptoms, except in those rare cases in which gastric or duodenal ulcer disease is present. • In H. pylori–infected children with nonulcer treatment of the infection has no proven benefit in symptom relief. Therefore, screening children with with dyspeptic symptoms for H. pylori infection with noninvasive tests is not • Children should be investigated for H. pylori only when their symptoms are suggestive of organic disease and are severe enough to justify the risks of therapy. • gastrointestinal endoscopy with biopsies is the preferred method of investigation in children with upper digestive symptoms suggestive of organic disease (e.g., esophagitis, peptic celiac disease). • If H. pylori is identified at endoscopy, treatment should be offered to the patient. • In children treated for H. pylori infection, the response to treatment should be monitored with a reliable noninvasive test. • Serologic tests for H. pylori are not reliable for use in children. • The 13C-UBT is very reliable in children but further evaluation in children, especially in those less than 2 years of age.