300 Hien Q. Huynh gastric cancer and H. pylori infection has been confirmed further in animal studies using the Mongolian gerbil [27]. In a prospective cohort study of Japanese patients, gastric cancer was shown to develop in 2.9% of 1246 adult patients infected with H. pylori. Mean follow-up was 7.8 years. No gas- tric cancer was found in those not infected or in a subgroup of patients who received eradication therapy for H. pylori [28]. However, in another pro- spective randomized placebo-controlled population-based primary study of 1630 healthy Chinese patients, carriers of H. pylori infection in a region of China with a high prevalence of gastric cancer, 817 received eradication therapy and 813 a placebo. No difference was found in terms of the inci- dence of gastric cancer development between the two groups over a period of 7.5 years. In a subgroup analysis of patients without precancerous lesions, such as atrophic gastritis and intestinal metaplasia, eradication seemed to decrease the development of gastric cancer [29]. A recent trial suggests that eradication of H. pylori may reduce the incidence of precancerous lesions Figure 1. Schematic representation of gastric pathology and disease outcome. Adapted from [19], with kind permission of the Massachusetts Medical Society Helicobacter pylori infection in children 301 [30]. Eradication of H. pylori in the older individuals with precancerous lesions appeared not to be effective in preventing gastric cancer. Since these precancerous lesions are rarely seen in children [31], perhaps eradication should take place in childhood to prevent the development of these pre- cancerous lesions later in life. Currently, the effect of H. pylori eradication on the incidence of gastric cancer is unknown and conclusive trials will take many years. Most experts favor eradication in first-degree relatives of gas- tric cancer patients [32–34]. Mucosal associated lymphoid tissue (MALT) lymphoma is rare in those infected with H. pylori, with a lifetime risk of less than 1%. Helicobacter pylori and abdominal pain The association between recurrent abdominal pain and H. pylori infection remains controversial. Some studies have supported the link and others have not [35]. Of interest, a recent study published by Malaty et al. [36] demonstrates that younger children suffering from recurrent abdominal pain are more likely to be infected with H. pylori than older children with recurrent abdominal pain. Another study from Taiwan found that H. pylori infection is more commonly found in children with short-term (2 weeks to 3 months) recurrent abdominal pain, suggesting that perhaps short-term abdominal pain may be a feature of acute H. pylori infection [37]. On the other hand, a recent community-based cross-sectional study from Sweden of 695 children between ages 10 and 12 years showed that 18% of children were infected with H. pylori based on positive anti-H. pylori antibody tests, and that there was no increase in recurrent abdominal pain reported in this age group of children with H. pylori infection [38]. In a double-blind randomized placebo-controlled trial, symptomatic response to H. pylori eradication was determined in children with recurrent abdominal pain. The control group was put on Omeprazole and the treatment group received eradication triple therapy; there were 10 children in each group. Bacteria eradication was achieved an 8 out of 10 children in the treatment group and none in the placebo group. After 52 weeks, there was a similar reduction in the symptom index observed in both groups. A limitation of this study was the small number of patients enrolled [39]. A recent Japanese study showed that children with recurrent abdominal pain that fulfilled the Room II cri- teria are more likely to have H. pylori infection and a psychiatric disorder [40]. All these studies suggest that recurrent abdominal pain of childhood is a heterogeneous syndrome with unclear etiology. H. pylori infection is likely to represent only a very minor cause of recurrent abdominal pain, perhaps affecting those who are younger and have recent onset of abdominal pain. There is therefore no indication for the test-and-treat strategy for H. pylori in children with recurrent abdominal pain. 302 Hien Q. Huynh H. pylori and gastroesophageal reflux disease Currently there is no evidence that H. pylori eradication worsen gastro- esophageal reflux disease (GERD) in adults [41]. Limited data are also available in children. However, there is a theoretical risk, even though the data are conflicting, that long-term proton pump inhibitor (PPI) treatment could increase the development of H. pylori-associated atrophic gastritis and increase the risk of gastric cancer [42, 43]. Some experts recommend testing and eradicating H. pylori if the child or adolescent is undergoing endoscopy for GERD, but not for those with clinically diagnosed GERD [44]. Extragastric manifestations of H. pylori infection in children With the increasing awareness among clinicians of H. pylori infection over the last decade, there have been a number of reports on the consequences of adverse effects from H. pylori infections outside the gastrointestinal tract (Tab. 1). Studies purporting the association between these manifestations of H. pylori infections are weak in terms of design and are not reviewed in this chapter. However, there are two manifestations that need to be mentioned: the first is refractory iron-deficiency anemia and the second is idiopathic thrombocytopenia (ITP) [45]. Refractory iron-deficiency and H. pylori infection There are a number of potential biological explanations for iron deficiency observed in H. pylori infection apart from bleeding secondary to peptic ulcer disease. Currently, there is no evidence to support that chronic gastritis secondary to H. pylori results in occult blood loss. However, H. pylori infec- tion in some settings might give rise to hypochlorhydria, low ascorbic acid levels, and increased lactoferrin (a host iron-binding protein) sequestration by the organisms. Also, H. pylori possesses multiple iron acquisition systems in its genome, which makes it an avid competitor for iron uptake with the host in the gastric microenvironment [46]. There are a number of case reports as well as case control studies, and recently a population study, supporting the role of H. pylori infection as a potential cause of otherwise unexplained refractory iron deficiency [47–49]. H. pylori infections, especially those with atrophic gastritis, are more likely to have unexplained refractory iron-deficiency anemia compared to the age- and sex-matched controls without iron deficiency [50]. Baysoy et al. [51] describe a group of children undergoing investigation for upper gastrointestinal symptoms and found that those with H. pylori were more likely to have lower iron stores compared those without infection. However, eradication of H. pylori has not consistently increased hemoglobin levels Helicobacter pylori infection in children 303 in control studies. One study confirmed an increase in hemoglobin level in Korean female athletes [52], whereas a large control household randomized open-label trial involving 290 Alaskan children with iron deficiency and H. pylori infection in a population that has a high prevalence of this infection, failed to show that H. pylori eradication improved isolated iron deficiency or mild anemia up to 14 months after treatment initiation. The limitation is that this study was not designed to detect small effects, with only 2 patients had hemoglobin less than 100 g/L [53]. Further randomized control trials in different populations, with age groups with varying degree of anemia are still needed to confirm the association between H. pylori and refractory iron deficiency. Despite this, the Canadian consensus conference on H. pylori still recommends, in the absence of other causes of iron deficiency, H. pylori should be tested for and treated [33]. Idiopathic thrombocytopenia and H. pylori infection In 1998, Akiyama and Onozawa [54] demonstrated that a PPI administra- tion was associated with an increase in platelet count in patients with chron- ic idiopathic thrombocytopenia (ITP). Subsequently, a Lancet paper in a pilot study by Gasbarini et al. [55] described a significant increase in platelet count in 8 of 11 patients with H. pylori, which was successfully eradicated. Since then, there have been numerous case reports and case series reporting that eradication of H. pylori is accompanied by platelet increase in adults patients with ITP. A review by Franchini and Vener [56]i summarized the adult literature. Of a total of 1126 patients with ITP, 64% were infected with H. pylori, eradication occurred in 81% with the platelet response rate occur- ring in approximately 50%. Subsequently, there has been a randomized con- trol trial looking at the effect of H. pylori eradication in adult patients with chronic ITP involving 36 Japanese patients, 25 of whom were positive for H. pylori; 13 of these 25 were randomized to the eradication group and 12 to the non-eradication group. Of the 13 patients in the eradication group, 6 had Table 1. Purported extragastric manifestations of H. pylori infection Manifestations Cardiovascular Atherosclerotic heart disease, stroke Neurological Parkinson disease, migraine Autoimmune Autoimmune thrombocytopenia purpura, Reynaud’s phenomenon, Sjögren’s syndrome, diabetes mellitus Dermatological Chronic urticaria, angioedema, rosacea, alopecia areata Others Refractory iron deficiency, halitosis, hyperemesis gravidarum, anorexia, glaucoma, oral ulcers, urethritis 304 Hien Q. Huynh either partial or complete response in their platelet count, whereas none of the patients in the non-eradicated group responded [57]. In another small randomized control trial there was no difference when comparing PPI ver- sus H. pylori eradication therapy in the treatment of ITP [58]. The potential explanation for this is likely molecular mimicry, where anti-platelet antibod- ies in the serum recognize the CAG protein of H. pylori [59]. The data are conflicting for children [60, 61]. There are a number of case reports of children with ITP and increased platelet count following H. pylori eradication [62–64]. Most of these reports came from the Far East. Other reports demonstrated no association between these two conditions [65]. These reports suggest that much larger randomized placebo-controlled trials need to be performed in children from different ethic backgrounds to determine whether in fact there is an association between the two con- ditions. This study needs to be conducted in areas with both low and high prevalence of H. pylori infection. Investigations Non-invasive test Currently, there are two non-invasive tests that are becoming extremely reliable in detecting H. pylori infection. The more established one is the urea breath test and the other is the stool antigen test. Serology test is not recommended as diagnostic tool because of its poor sensitivity and test-to- test variability [33]. Urea breath test The urea breath test utilizes the essential enzyme urease, which is pro- duced by H. pylori. Urease converts urea to ammonia and CO 2 . If the CO 2 is labeled with a stable isotope, this can be detected in the expired air (Fig. 2). In the non-infected individual, urea will leave the stomach unchanged. This test is essentially a detection of urease activity, which can also be produced by other bacteria in the oral cavity, as in the setting of bacterial overgrowth. 13 C and 14 C are the two isotopes that are well validated [66]. Only the 13 C urea breath test has been extensively tested in children [67]. 14 C is radioactive and is not acceptable to be used in chil- dren. The 13 C test is more expensive than the 14 C test, because it requires mass spectrometry or infrared spectroscopy equipment for analysis of the expired breath. The results of the urea breath test are reported as 6/base- line (DOB), which is a measure of the ratio of 13 C CO 2 , to 12 C CO 2 . If the DOB exceeds a certain point, the patient is considered to be infected with H. pylori infection. The test is best performed when the patient has fasted Helicobacter pylori infection in children 305 and has not been on a PPI for at least 2 weeks. The patient should not have taken antibiotics for 4 weeks prior to testing because this can reduce the H. pylori load. The use of an acid solution as part of the test solution, either citric acid or orange or apple juice, is ideal because urea activity is highest in an acid environment. Expired breath can be collected between 15 and 45 min depending on the laboratory. Despite the variability in the dosage of urea used and the different cut off points, this study consistently shows that the urea breath test has a sensitivity of over the 96% and a specific- ity of over 90% [68]. Infants and toddlers are much more likely to have a positive result in comparison to school-age children and adolescents [69]. Reducing the tracer dose and changing the DOB value increase the speci- ficity of this test in younger children [70, 71]. Also, technically, it is much more difficult to collect reliable expired air from infants and toddlers compared to school age children. Stool antigen test This is a non-invasive test for H. pylori antigen in stool in the pediatric population. Unlike with the urea breath test, there is no collection dif- ficulty, particularly in the younger age group. There are two types of stool antigen tests on the market: one is polyclonal, and the other a monoclonal, antibody enzyme immunoassay. In a recent meta-analysis of 22 studies including 2499 patients, the monoclonal stool antigen test was found to have sensitivity and specificity of 94% and 97%, respectively. In 13 of the Figure 2. Carbon in Urea is labeled with either 13 C or 14 C and exhaled after being converted to label CO 2 by urease produced by Helicobacter pylori in the gastric mucosa. 306 Hien Q. Huynh studies in which the monoclonal was compared to the polyclonal stool anti- gen tests, the monoclonal test had a higher sensitivity of 95% versus 83% for the polyclonal test. In terms of eradication, analyzing 12 studies with 957 patients post treatment, the sensitivity and specificity for the mono- clonal test were 93% and 96%, respectively. In 8 of the studies where both monoclonal and polyclonal tests were used, sensitivity was higher for the monoclonal (91%) than for the polyclonal test (76%), demonstrating that the monoclonal stool antigen test is a much more accurate, non-invasive test for both diagnosis and confirmation of eradication of H. pylori infec- tion post treatment [72]. A recent European multi-centered study comparing urea breath tests with stool and serology tests, as well as antibody detection in urine, found that the urea breath test is the most sensitivity and specific. The stool anti- gen test used in this study was polyclonal (Meridian). The urea breath test had a sensitivity and a specificity exceeding 96%, whereas the stool test has sensitivity and specificity of 92%. This study only had a small number of children under the age of 6 (48 children accounted for 15% of the study population) [73]. However, another study performed in Egypt compared the urea breath test, monoclonal stool antigen test, and serology test to endoscopy with biopsy and rapid urease test. In this population 53 of the 108 children tested were under the age of 6. Overall, the sensitivity and specificity of the urea breath test was 98% and 89%, respectively, and those of the monoclonal stool antigen test were 94% and 81%, respectively. Interestingly, the urea breath test sensitivity was not affected by age but the specificity was lower in those under the age of 6 (86% versus 95%). With regard to the monoclonal stool antigen test, those performed on children under the age of 6 showed a sensitivity of 94% and specificity of 81%. However, above the age of 6, the sensitivity remained about the same at 92% but the specificity increased to 100%. Serology had the worst outcome in this study, giving a sensitivity of 50% and specificity of 80% [69]. Overall, the urea breath test remained the most sensitive and specific non-invasive test for H. pylori. There is still some conflicting data on how reliable this test is in the younger age group. The monoclonal stool antigen test is also proving to be quite a sensitive and specific test , and, again, its reliability in the younger age group requires further study. The use of these diagnostic tests needs to be interpreted in the local context, particularly whether these tests have been validated to the population that is under investigation, in particular with regard to the population age group as well as ethnicity and geography. Invasive tests Upper gastrointestinal endoscopy with mucosal biopsy remains the gold standard for the diagnosis of H. pylori infection in children. It has the advan- Helicobacter pylori infection in children 307 tage of being able to detect upper gastrointestinal pathology including the complications of H. pylori infection such as nodular gastritis, peptic ulcer disease, gastric cancer, and MALT lymphoma. In pediatrics, the primary indication for upper GI endoscopy is the presence of persistent, severe upper abdominal symptoms and not simply the presence of H. pylori [33]. It is difficult to differentiate symptoms secondary to the complication of H. pylori infection such as peptic ulcer disease and functional dyspepsia. The most common endoscopic finding in children with H. pylori infection is nod- ular gastritis, which is seen most commonly in the antrum with an irregular (cobblestone) appearance, which is highlighted with blood from a bleeding biopsy site. When nodular gastritis is found, it has high specificity (98%) for H. pylori infection, and therefore a high predictive indicator for H. pylori infection, but it has low sensitivities (44%) [17, 74]. In naïve patients, antral biopsy had the highest yield, particularly in the mid antrum region of the lesser curvature [75]. For a patient who has been on acid suppression therapy or antibiotics, a biopsy from the transitional zone and body are also required to improve the yield [76, 77]. For patients who have developed complications from H. pylori infection such as peptic ulcer disease, multiple biopsies from different regions of the stomach are required. H. pylori can often be seen using hematoxylin and eosin staining; immunohistochemistry using polyclonal and H. pylori antibodies is likely to be the most reliable detection method on biopsy sections, although this method is expensive and time consuming. Among other stains that are often used is the Giemsa stain, which is less reliable but widely available and affordable [78]. The optimal staining method is often guided by local expertise. With a biopsy of gastric mucosa, a rapid urease test can also be used. This test essentially detects the presence of urease produced by H. pylori. The test is highly specific and sensitivity in adults, but less so in children [79]. The accuracy of this test also depends on the number of biopsies taken, site of biopsy and the use of antibiotics and PPIs. However, one of the major advantages of biopsy in H. pylori infection is the ability to culture this bacterium. In clinical trials, the success rate is up to 80% of infected children [80]. Bacterial culturing is time consuming and expensive, but it does allow for antibiotic sensitivity to be determined. This is particularly useful for those who failed previous eradication therapy. In addition, for a positive culture, the genotype of clini- cal isolate can be investigated for specific bacterial virulent factors. There are now a number of molecular techniques that can be used to detect the presence of H. pylori and the presence of a number of point mutations in the bacterial genome that determines antibiotic resistance genotypes. For example, the predominant cause of clarithromycin resistance is a point mutation in the peptidyl transferase of the 23S rRNA gene. There are also a number of inactivating mutations involving some reductase genes that convert Metronidazole from a harmless product to an anti-bacterial agent, inactivating the gene responsible for a portion of H. pylori resistance to Metronidazole [81]. 308 Hien Q. Huynh Antibiotic resistance Studies of antibiotic resistance in children are small in number; an US network for antibiotic resistance that tracked the national incidence rate of H. pylori and microbial resistance reported 340 clinical H. pylori iso- lates collected over a period of 4 years that demonstrated a 29% rate of antibiotic resistance to at least one antimicrobial agent and 5% resistant to two or more antimicrobial agents: 25% were resistant to Metronidazole and 12.9% were resistant to Clarithromycin. Only a very small number of cases (0.9%) were resistant to Amoxicillin [82]. However, more recently, the results of a larger prospective multi-center study from Europe on the rate of antibiotic resistance in H. pylori strain in 1233 children have been reported. These patients came mostly from western and southern Europe. Most of the isolates were obtained prior to any treatment, and overall the resistance rate for Clarithromycin was 24%. This increased to 42% in those who had previously received treatment that had failed. The resistance rate to Metronidazole was 25% and higher (35%) in those who received previ- ous failed treatment. Resistance to both antibiotics only occurred at 6.9%; however, this increased to 15.3% in those who received previous failed treatment. Resistance to Amoxicillin was exceptionally low at 0.6%. These results confirm that antibiotic usage in children represents a major risk fac- tor for developing treatment resistance [80]. Treatment Unfortunately, there have not been many randomized placebo-controlled studies in the pediatric population looking at eradication of H. pylori. One study involving 73 children with dyspeptic symptoms demonstrated an eradication rate of 74% using Amoxicillin and Clarithromycin with a PPI, and 9.4% using dual therapy of Amoxicillin and Clarithromycin for 7 days using intention to treat analysis [83]. Another study by Oderda et al. [84] used Lansoprazole, Amoxicillin and Tinidazole triple therapy versus place- bo plus Amoxicillin and Tinidazole dual therapy for 1 week; after 6 months the eradication rate was 72% for triple therapy and remained at 71% for dual therapy, showing no difference between the two treatments. A recently developed 10-day sequential treatment for H. pylori eradication was studied in 78 children. They were either randomized to receiving Omeprazole plus Amoxicillin for 5 days, followed by Omeprazole plus Clarithromycin and Tinidazole for another 5 days, compared to triple therapy of Omeprazole plus Amoxicilline and Tinidazole for 1 week. Sequential treatment had an eradication rate of 97.3% and triple therapy an eradication rate of 75.7%, demonstrating that sequential treatment is superior to triple therapy, consis- tent with results from adult studies [85, 86]. This sequential treatment needs to be further studied in different populations to determine its efficacy and Helicobacter pylori infection in children 309 safety, and to confirm its higher eradication rate in comparison to a 2-week course of triple therapy. A recent Canadian Helicobacter Study Group Consensus Conference still recommends the use of triple therapy for 2 weeks using a PPI with Clarithromycin and Amoxicillin or Metronidazole given for 14 days. The duration of treatment of 2 weeks is likely to be optimal, but not conclusive; there is a 7–9% increase in the eradication rate with 14 days of treatment versus 7 days [87]. Tetracycline should be avoided in children under the age of 12 because it may cause staining of the children’s enamel. In addi- tion, treatment failure is increased with antibiotic resistance [88]. A recent Russian study by Nijevitch et al. [89] treated 76 children, who had failed triple therapy, using quadruple therapy. These children were randomized to receive a 2-week course of bismuth subcitrate, Amoxicillin with Nifuratel or Furazolidone plus Omeprazole. The eradication rate was 89% for Nifuratel and 87% for Furazolidone. Nifuratel is preferred because of a lower fre- quency of side effects. Potentially, this could be a treatment of choice for those who have failed eradication. It is vital that reference laboratories are available to monitor the population H. pylori antibiotic sensitivity and test those with treatment failure. Conclusion H. pylori is generally acquired in childhood and the prevalence in developed countries is now decreasing. The clinical manifestations of disease are a result of the host, bacteria and environment interaction, and are only seen in a subset of infected individuals. Its association with peptic ulcer disease, gastric cancer and MALT lymphoma is beyond dispute. A test-and-treat strategy is not indicated for children with recurrent abdominal pain. New indications in children are now emerging advocating its eradication, such as refractory iron deficiency, ITP and a strong family history of gastric cancer, although further studies are needed. Stool antigen tests and urea breath tests have emerged as some of the best non-invasive tests for H. pylori. Antibiotic resistance is on the rise, and novel treatment strategies are need- ed. Improving the social situation of children such as better housing, sanita- tion and hygiene remain one of the key pillars in reducing the prevalence of this infection in childhood. References 1 Parsonnet J (2005) Clinician-discoverers – Marshall, Warren, and H. pylori. N Engl J Med 353: 2421–2423 2 Rowland M, Daly L, Vaughan M, Higgins A, Bourke B, Drumm B (2006) Age- specific incidence of Helicobacter pylori. Gastroenterology 130: 65–72 [...]... thrombocytopenic purpura associated with Helicobacter pylori infection Pediatr Int 48: 76 78 Hayashi H, Okuda M, Aoyagi N, Yoshiyama M, Miyashiro E, Kounami S, Yoshikawa N (2005) Helicobacter pylori infection in children with chronic idiopathic thrombocytopenic purpura Pediatr Int 47: 292–295 314 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 Hien Q Huynh Yetgin S, Demir H, Arslan D, Unal S, Kocak N (2005) Autoimmune... hMPV/ NL/1/99 (subgroup B1), hMPV/CAN/ 97/ 83 (subgroup A2), and hMPV/ CAN/98 /75 (subgroup B2) A nearly complete genome sequence was determined for the prototype NL/1/00 strain of hMPV, and complete genome sequences were determined for two Canadian strains, CAN 97/ 83 and CAN98 /75 [2, 3] These studies also confirmed that the 3’–5’ hMPV gene order is N-P-M-F-M2,1-M2,2-SH-G-L None of these proteins have been... 03–Oct 04 Oct 02–May 03 Oct 03–May 04 Nov 02–May 03 Nov 03–May 04 Study period RT-PCR(1) PCR RT-PCR(2) RT-PCR(2) RT-PCR(2) RT-PCR(2) RT-PCR(1) Method . of Helicobacter pylori. Lancet 352: 878 56 Franchini M, Veneri D (2006) Helicobacter pylori-associated immune thrombo- cytopenia. Platelets 17: 71 77 57 Suzuki T, Matsushima M, Masui A, Watanabe. therapy with nifuratel to furazoli- done. Aliment Pharmacol Ther 22: 881–8 87 Pediatric Infectious Diseases Revisited 3 17 ed. by Horst Schroten and Stefan Wirth © 20 07 Birkhäuser Verlag Basel/Switzerland Human. and CAN98 /75 [2, 3]. These studies also confirmed that the 3’–5’ hMPV gene order is N-P-M-F-M2,1-M2,2-SH-G-L. None of these proteins have been identified or characterized by direct biochemical