Helicobacter pylori-Specific Immune Responses of Children: Implications for Future Vaccination Strategy

Abstract

We analyzed the specific anti-Helicobacter pylori immunoglobulin G (IgG) antibody profile for a sample of 824 asymptomatic schoolchildren in southern Germany (mean age, 10.7 ± 0.65 years) with an H. pylori-specific IgG enzyme-linked immunosorbent assay and Western blot analysis. The prevalence of infection was 19.8% (95% confidence interval, 17.1 to 22.7%). The immunoresponses were characterized predominantly by antibodies against low-molecular-mass antigens of 14 and 29 kDa, with a significant difference between children of German and Turkish nationalities (P = 0.0012 and P < 0.0001, respectively).

Helicobacter pylori plays an etiologic role in gastritis, peptic ulcer disease, gastric cancer, and mucosa-associated lymphoid tissue lymphoma (2), and acquisition of the infection seems to occur mainly in the first years of life (26).

Genetic typing has shown that H. pylori is extremely diverse as a species and possesses a number of antigens that induce a specific immune response in infected individuals (1, 10, 13, 16, 17, 27). The specific immune response is part of the natural history of the infection in children, which remains poorly characterized, and may mirror the specific interaction of the host with the bacterium (8, 12, 15, 21). This immune response may be of special importance with regard to vaccination, as it may determine the antigens considered for a vaccine. However, the few existing studies of the infection in children are based mainly on symptomatic and small clinical samples (4, 7, 14, 18, 23).

The aim of this study was to analyze the specific anti-H. pylori immunoglobulin G (IgG) antibody profile in a sample of asymptomatic schoolchildren. This study is part of a population-based investigation performed by the Baden-Württemberg State Health Office between October 1999 and April 2000 with a sample of schoolchildren (fourth graders; mean age ± standard deviation, 10.7 ± 0.65 years; range, 9 to 13 years; median, 11 years) living in southern Germany in four different areas of Baden-Württemberg (Stuttgart, Mannheim, Kehl, and Ravensburg). Participation in the study was voluntary. The study was approved by the Ethics Board of the University of Ulm. The parents of 1,118 children filled out a questionnaire which requested information regarding sociodemographic factors, family characteristics, and medical history. Serum samples from 824 (73.7%) of those 1,118 children were available.

H. pylori IgG antibodies were determined by enzyme-linked immunosorbent assay (GAP; Bio-Rad Laboratories Diagnostics Group, Munich, Germany). According to the manufacturer's instructions, levels of >20 U/ml were considered positive. This test has previously been evaluated with 310 children (compared with the histology, culture, and urease test, the sensitivity of our test was 94.9% and the specificity was 92.4%) (28). The serum samples of the IgG-positive children were further analyzed for the presence of antibodies against H. pylori-specific antigens with a commercial Western blot method (Helicobacter pylori Western blot; AID, Strasberg, Germany) as described previously (24).

Of the 824 children included in the study, 53.3% were boys. About two-thirds, or 63.5%, of the schoolchildren were of German nationality, 12.5% were Turkish, and 12.7% were of other nationalities (for 89 of the children (10.8%), there were no data available concerning nationality). Information gathered on family characteristics showed that 10.2% of the children had no siblings, 43.3% had one, and 38.1% had two siblings; 45.5% of the fathers and 37.5% of the mothers had 9 or fewer years of school education.

One hundred sixty-three children (19.8% of the total number, 95% confidence interval [CI], 17.1 to 22.7%) had positive serology results for H. pylori (Table 1). The seroprevalence in children of German nationality was 13.9% (95% CI, 11.0 to 17.1%) and varied from 12.4% (95% CI, 9.6 to 15.6%) for the German children born in Germany to 42.3% (95% CI, 23.4 to 63.1%) for the German children born elsewhere (mostly children of families who had lived in countries of Eastern Europe for generations and who had returned to Germany quite recently). The prevalence of H. pylori in children of Turkish nationality was 37.9% (95% CI, 28.5 to 48.0%) (in Turkish children, place of birth was not associated with seroprevalence), and for children of other nationalities, it was 30.5% (95% CI, 21.9 to 40.2%). These strong differences among ethnic groups were independent of socioeconomic factors, such as family size and school education of parents, and persisted after these factors had been controlled for. On the other hand, duration of school education of parents and number of siblings were not factors associated with seroprevalence if ethnic origin was considered in the analysis (data not shown).

TABLE 1.

Prevalence of H. pylori infection in children according to nationality

Nationality	Total no. of children	H. pylori-positive children
		n	%	95% CI
German	527	73	13.9	11.0-17.1
Born in Germany	501	62	12.4	9.6-15.6
Born elsewhere	26	11	42.3	23.4-63.1
Turkish	103	39	37.9	28.5-48.0
Other	105	32	30.5	21.9-40.2
Unknown	89	19	21.3	13.4-31.3
Overalla	824	163	19.8	17.1-22.7

^aAll children.

Western blots were used to detect and analyze eight different H. pylori-specific immunoreactive bands (Table 2).

TABLE 2.

Prevalence of specific H. pylori IgG antibodies in H. pylori-infected children (n = 163)

H. pylori-specific antibody and molecular mass	Prevalence
	n	%	95% CI
CagA, 120 kDa	40	24.5	18.1-31.9
VacA, 87 kDa	26	16.0	10.7-22.5
UreB, 66 kDa	39	23.9	17.6-31.2
UreA
29 kDa	71	43.6	35.8-51.5
25 kDa	59	36.2	28.8-44.1
19 kDa	10	6.1	3.0-11.0
17 kDa	6	3.7	1.4-7.8
14 kDa	73	44.8	37.0-52.8

Table 3 shows that antibodies to the 25-kDa antigen and the 14-kDa antigen were detected much more frequently in infected children of Turkish or other nationality than in infected children of German nationality (P values of 0.0012 and <0.0001, respectively).

TABLE 3.

Prevalence of H. pylori-specific antibodies identified by immunoblotting among infected children according to nationality of the children (n = 144)^a

H. pylori-specific antibody and molecular mass	Nationality						P value
	German		Turkish		Other nationality
	n	%	n	%	n	%
CagA, 120 kDa	15	20.6	9	23.1	11	34.4	0.31
VacA, 87 kDa	8	11.0	8	20.5	7	21.5	0.25
UreB, 66 kDa	16	21.9	13	33.3	9	28.1	0.41
UreA
29 kDa	29	39.7	12	30.8	17	53.1	0.16
25 kDa	15	20.6	18	46.2	17	53.1	0.0012
19 kDa	4	5.5	3	7.7	1	3.1	0.70
17 kDa	1	14	3	7.7	2	6.2	0.22
14 kDa	17	23.3	23	59.0	19	59.4	<0.0001

^aNineteen H. pylori-positive children were excluded because of missing data concerning nationality.

This study, conducted with a population-based sample of asymptomatic schoolchildren in Germany, demonstrated that the specific systemic immune response in children to H. pylori infection is particularly pronounced against low-molecular-weight proteins of H. pylori. Furthermore, there was a large variation in both the prevalence of infection and the specific antibody patterns with respect to the nationalities of the children, although the children were living in the same geographic region. As shown in a previous study, differences between ethnic groups are not explained by differences in socioeconomic factors (25) and most likely mirror strain differences among these groups.

A recent French study of 80 pediatric patients found the highest prevalence (88.7%) for antibodies to a 26-kDa antigen (22). Little is known about the function of the 26-kDa antigen, which may be identical to the 25-kDa antigen in our study. However, recently it has been shown that a 25-kDa outer membrane protein acts in a lectin-like manner with lipopolysaccharide to mediate attachment of H. pylori to laminin and may therefore be important for the colonization process in the early phase of the infection (11, 29).

Much interest exists in the seroprevalence in children of CagA antibodies, which, according to a study by Mitchell et al. (14), correlated with age and increased up to the age of 15 years. There are few data available for children, and the results of recent studies have been inconsistent concerning the prevalence and association with disease of CagA-positive strains in both adults (16, 19, 28) and children (3, 6, 7, 9, 12, 28, 30). This may be due partly to the considerable variation in the frequencies of vacAsl subtypes and iceA alleles in children originating from different geographic regions (19, 20). As for adults, a high prevalence of CagA-positive H. pylori strains was found in Japanese children; however, there was no association of this prevalence with nodular gastritis or peptic ulcer disease (7) in Japanese children (mean age, 12 ± 3 years). This result challenges the role of CagA as a marker for more serious gastroduodenal disease in children as well as in adults in Japan (7). The highest prevalence of CagA antibodies in Europe (80%), has been found in a study of infected children in Poland (5); the prevalence in infected schoolchildren in Estonia (9 to 15 years old) was 46% (30). Raymond et al. described the prevalence of CagA antibodies in children in France who had abdominal pain to be 43% (22).

In conclusion, the differences in the antibody patterns between children of various ethnic groups or found in our study may indicate strain differences and should be considered if antigens are chosen as targets for a vaccination approach.