1.

Abstract-Design, setting, participants & measurements: One hundred and eight patients presented to the University College Hospital Galway, Ireland (UCHG) with upper gastro-intestinal symptoms were included in the study. There were 52 males with age range 18-82 years, mean age 49 years, and 56 females, age range 20-83 years, mean age 52 years. They were grouped according to the endoscopic findings into 4 groups (i) Duodenal Ulcer (DU), (ii) Gastric Ulcer (GU), (iii) Gastritis (GS), & (iv) Non-Ulcer Dyspepsia (NUD). Five milliliter of venous blood was taken in a sterile plain tube; the serum was separated and stored at -70 C 0 for ELISA assay. Endoscopy was carried out and at least 4 biopsies were obtained, 2 from the antrum and 2 from the first part of the duodenum (duodenal bulb). One biopsy each from the antrum and duodenum was transferred immediately into a sterile container containing 2 mL of RPMI 1640. The other biopsy specimens one from the antrum and one from the duodenum were processed for histological examination. The serum IgG antibodies to H. pylori and local IgA and IgG were measured by an ELISA test. Serum IgG antibody levels of the four groups of patients with various degrees of inflammation (0-3 degree) were compared, a highly significant association was found between the level of IgG antibodies to H. pylori and the severity of inflammatory reaction in the biopsy specimens (0 &2 and 0 & 3, p=0.0001) and between grade (1 & 2, p=0.045) and (1 &3, p=0.002) but not between (0&1, p=0.310), and (2&3, p=0.282) grades of inflammation.

Duodenal and antral IgA and IgG antibodies to H. pylori were assessed in 108 patients. A higher number of antral compared to duodenal biopsies in culture produce both IgA and IgG antibodies to H. pylori in all groups. 21

elicobacter pylori are now recognized as a primary cause of active chronic gastritis in humans. Most infected humans remain asymptomatic, but are at increased risk for developing peptic ulcer disease and possibly gastric cancer (1)(2)(3)(4)(5)(6)(7)(8). The pathogenesis of this infection is well understood and the motility and urease activities are virulent factors in an animal model (9,10). The urease enzyme also elicits a strong immune response during acute infection, suggesting that this abundant antigen is readily available to the immune system. An increase in serum IgG titers is predictive of ongoing infection (11,12). About 50% of H. pylori isolates produce vacuolating toxins in vitro, which may be an important determinant of virulence (12)(13)(14)(15). The epidemiology of the infection by H. pylori correlates with the prevalence of superficial type B gastritis. Infection is associated with active (presence of neutrophils), chronic (presence of lymphocytes and plasma cells), or active-chronic (neutrophils, lymphocytes, and plasma cells) inflammatory reactions (16)(17)(18). Gastric inflammation nearly always precedes the development of peptic ulceration, and is an important component in initiating the multi-step progression towards gastric carcinogenesis (19). In this study we have demonstrated that the severity of inflammation in the antrum is strongly associated with the presence and density of the bacterial colonization of the antral mucosa and the level of the serum IgG antibody to H. pylori.

One hundred and eight patients presented to the University College Hospital Galway, Ireland (UCHG) with upper gastro-intestinal symptoms were included in the study. There were 52 males with age range 18-82 years, mean age 49 years, and 56 females, age range 20-83 years, mean age 52 years. They were grouped into 4 groups according to the endoscopic findings; (i) Duodenal Ulcer group (DU), (ii) Gastric Ulcer group (GU), (iii) Gastritis group (GS), and (iv) Non-Ulcer Dyspepsia group (NUD).

Endoscopy was carried out and at least 4 biopsies were obtained, 2 from the antrum and 2 from the first part of the duodenum (duodenal bulb). One biopsy each from the antrum and duodenum was transferred immediately into a sterile container containing 2 mL of RPMI 1640 (Gibco Life Technology Ltd. UK) supplemented with 10% fetal calve serum (FCS) and 40 ug/ml gentamicin (sigma). The biopsy specimens were cultured for 3 days at 37C 0 ; the unspun supernatants were removed and stored at -70C 0 for estimation of IgA and IgG antibodies against H. pylori (34). The other biopsy specimens one from the antrum and one from the duodenum were placed in formalin and processed for histological examination. Five milliliter of venous blood was taken in a sterile plain tube; the serum was separated and stored at -70 C 0 for ELISA assay.

The serum IgG antibodies to H. pylori were measured by a commercially available ELISA kits (Biometra, Germany). Local IgA and IgG antibodies to H. pylori in the supernatants were measured by ELISA kit (KPL Kierkegaard & Perry Laboratories, Inc.) and the results were read on Dynatech MR 5000 automatic micro-plate reader at wave length 410 nm. Endoscopy: all endoscopies were performed by two experienced endoscopists (CLL and SFM). Endoscopic findings were registered and stored in a computer data base. Biopsy specimens for histological examination, culture of the organisms, and estimation of local antibodies to H. pylori in the cultured biopsy materials were taken from the antrum and duodenal bulb. The endoscopes (Olympus Gastrscope type GIF-Q10 and GIF-Q20) and biopsy forceps were cleaned and disinfected in a commercially available 2% glutaraldehyde solution (Totacide 28 Coventry Chemicals Ltd) for a period of 10 minutes between each endoscopy.

ELISA assay for estimation of systemic antibodies: H. pylori antibody test is an enzyme linked immune-sorbent assay (ELISA) for the quantitative determination of H. pylori IgG antibodies in human serum or plasma. The methodology is based on using the solid phase technique on the micro-wells. The purified H. pylori specific antigens are immobilized into the surface of the micro-wells. The antigens react with H. pylori IgG antibodies present in both the standards and patient samples. The resulting antigen-antibody complex binds the second enzyme-conjugate antihuman IgG antibody, creating a sandwich. The enzyme conjugated antigenantibody complex reacts with added substrate to develop a colored solution. The intensity of the color is dependent on the amount of the enzyme coupled to the complex and proportional to the H. pylori antibody concentration in the patient's samples. The optical density (OD) of the color in each well is read using Dynatech MR 5000 automatic micro-plate reader at 450 nm. Serum IgG antibodies to H. pylori were measured using the computer constructed formula for the semi-log curve fit. Y= a+b Logx Where (Y) is the mean of the OD of the sample, (a)-is the Y intercept of the graph and (b) is the constant factor for (Logx). Having obtained the value of Logx from the equation the inversion of that value represents the amount of the IgG in the sample expressed in ELISA units.ml sample (E.U/ml sample). The assay procedure is done at room temperature (18-22C 0 ), higher temp up to 35C 0 does not interfere with the results of the test but the incubation time for the TMB enzyme substrate reaction should be reduced according to the specifications of the manufacturer. ELISA assays for estimation of local antibodies: Local IgA and IgG antibodies in the supernatants of the cultures biopsy materials from the duodenal bulb and the antrum were assayed using the ELISA assay as specified by the manufacture (KPL Kierkegaard & Perry Lab Inc.)

The analysis of the results of the serum IgG antibody levels to H. pylori were assessed using One Way Analysis of Variance (ANOVA) with the addition of the Tukey HSD test option to the statistics. The additions of the Tukey option to the statistic command in the oneway analysis of variance will protect from declaring pairs of means differ when they could differ by chance. Analysis of variance offers a standard method for comparing various groups when there is no presumption beforehand that they differ. The Tukey's method exemplifies those tests for difference among group means by using the difference between the largest and smallest means, often called the range, as a measure of their dispersion. We have used the SYSTAT computer package (1990, SYSTAT inc.) for ANOVA with Tukey's option. Chi square was used to analyze the difference of the local and systemic humoral immune response to H. pylori infection. Pearson's correlation method was used for statistical analysis of the relationship between the density of H. pylori and serum IgG antibody levels. (0-3 degree) were compared, a highly significant association was found between the level of IgG antibodies to H. pylori and the severity of inflammatory reaction in the biopsy specimens. One was analysis of variance with Tukey HSD multiple comparisons was applied for statistical analysis. There were a significant difference of serum IgG antibody levels to H. pylori between grade (0 &2 and 0 & 3, p=0.0001) and between grade (1 & 2, p=0.045) and (1 &3, p=0.002) but not between (0&1, p=0.310), and (2&3, p=0.282) grades of inflammation, Table-1.

In vitro production of duodenal and antral IgA and IgG antibodies to H. pylori were assessed in 108 patients with dyspepsia. There were divided into four subgroups according to the endoscopic findings (i) DU group (25 patients), (ii) GU group (6 patents), (iii) GS group (31 patients), (iv) NUD group (56 patients), Tables -2-6. The results were analyzed using the Chi square test and the p values were shown in Table -6.

A higher number of antral compared to duodenal biopsies in culture produce both IgA and IgG antibodies to H. pylori in all groups (tables 2-6). This is an expected finding as H. pylori only colonize duodenum if gastric metaplasia is present. 21/25 (84%) of DU patients had IgG antibodies to H. pylori, 22/25 (88%) had antral IgA antibodies, 19/25(76%) had antral IgG antibodies, 12/25(48%) had duodenal IgA antibodies and 7/25(28%) had duodenal IgG antibodies to H. pylori, Table 2.

In the GU group although 6/6(100%) patients had serum IgG antibodies to H. pylori, only 4/6(66.7%) had antral IgA antibodies, and 3/5(60%) had antral IgG antibodies. Duodenal antibodies were much less common than in the DU group, only 1/5(20%) had duodenal IgA antibodies and 0/5(0.0%) had IgG antibodies to H. pylori infection, Table 3 5.

The difference between the local (antral and duodenal) IgA and IgG antibodies in the DU, GU, GS, and NUD groups are shown in Tables 2-6.

The level of serum IgG to H. pylori was significantly different among the 4 groups when tested by the Kruskal-Wallis One Way Analysis of variance (p=0.0008). However, when the multiple comparisons using the Tukey HSD One way analysis of variance was applied the significant difference was found to be residing in the GS and NUD groups (p=0.013). There were no significant difference between the other groups.

Chronic gastritis with a prominence of lymphocytes and plasma cells is considered to be a morphologic indicator of an immune response to H. pylori infection (175). Indeed, a local immune response and production of antibodies against H. pylori have been demonstrated in patients with gastritis by Grabtree and others (19)(20)(21) and have been confirmed in this study also (Tables 2-6). Berstad et al (22) have demonstrated that the characteristics (severity and cell type) of gastritis associated with H. pylori infection are influenced by geographical factors that may be similar to those that modify infection rates for different geographical locations. The pathogenesis of the bacterium depends on the production of several virulence factors. The most important ones are CagA (cytotoxic associated gene A) and VacA (vacuolizing cytotoxin A) (23)(24)(25)(26)(27). We have found a relatively good correlation between the serology and the histological findings in antral biopsies despite the low sensitivity (63%) of our ELISA test compared to histology and culture. Our results agreed with Bertram et al (28,29) that atrophic gastritis was associated with low detection rate of H. pylori by histology.

The reliability of culture of H. pylori as a diagnostic test varies considerably from center to center. Many workers found a sensitivity of 85% as we did. Our results showed lower sensitivity and specificity (77.4% and 75%) of histology. ELISA assay has been widely utilized in the sero-epidemiological studies and in the follow up of treated patients with H. pylori infection (30). We recommended determining IG antibody as a pre-endoscopic screening test. Further improvement and refinement of H. pylori antigens and proper exploitation of this test may have a great implication on the work load of busy endoscopic units and on the economy of health services. In H. pylori both IgA and IgG antibody titers are significantly elevated. On the other hand, the IgM antibodies are similar in H. pylori positive and negative subjects (31,32). Probably because the IgM sero-conversion in the early phase of infection is rapidly transit and usually will not be detected. Acid-glycin extraction and more purified antigens to H. pylori raise the sensitivity and specificity of ELISA to H. pylori to 95-96% and 74-83% respectively (33)(34)(35)(36)(37). Although the antigen we used for ELISA assay was purified antigen, our ELISA lacks a sufficient sensitivity (63%, but has a high specificity (97%) for H. pylori.

Talley et al (38) sensitivity and specificity 96 and 94% respectively. Newell et al (39) also reported high sensitivity and specificity of ELISA (100%) in 47 patients who had endoscopy and confirmed by histology and culture.

Prieto et al (40) found that nodular anteritis was a frequent (67%) and a specific finding in H. pylori infection. Their presence was associated with lymphoid follicles in histopathological examination. Our results agreed with Prieto et al. Moderate inflammation were present in 9/25(36%) of our patients, all were H. pylori positive. We concluded that the majority of the elderly (>61 yrs) had a prevalence of H. pylori ranging from 45-64% depending on the mode used to detect H. pylori infection. More than 50% of this group had only mild inflammation and H. pylori was isolated in 43%, however, moderate to severe degree of inflammation was present in 44% of them and H. pylori was always associated with gastritis. If we interpret this association in the light of serum IgG antibodies to H. pylori we will find that 95% of the elderly people with moderate to severe degree of inflammation will have an elevated serum IgG antibody levels of 220-250 EU for moderate inflammation and >250 EU for severe inflammation.

The evolving knowledge of the immune reaction to H. pylori infection allows us to study the interrelationship between (i) the presence of H. pylori and the density of the organisms on the histological section from the antral biopsies, (ii) the severity of inflammation, and (iii) the serum IgG antibodies to H. pylori as a marker of established infection. The presence of the bacteria was associated with increased amounts of mononuclear inflammatory cells, neutrophilic and eosinophilic leukocytes in the antral biopsies, thus confirming the results of Karttunen et al (41) and Alam et al (42). In these patients the infection with H. pylori produces a higher number of local IgA and IgG antibodies to H. pylori in the antral compared to duodenal biopsies in culture, in all groups (p=0.0296), Tables 2-6.

Helicobacter pylori are now recognized as a primary cause of active chronic gastritis in humans. Most infected humans remain asymptomatic, but are at increased risk for developing peptic ulcer disease and possibly gastric cancer. H. pylori is capable of inflecting local and systemic immune response with production of local (antral and duodenal) IgA and IgG and systemic IgG that correlate with the degree of inflammation in the antrum and duodenum. Successful treatment of the infection with antibiotics and proton pump inhibitors may influence the degree of inflammation, the local and systemic immune response to the organisms.