Epidemiology and Diagnosis of Helicobacter pylori infection

Prevalence of H. pylori Infection

Table 1 presents data obtained from the prevalence studies concerning adults published during the last year and carried out since the year 2000. The studies concerning children are presented in the corresponding article of the same issue of the journal.

Two studies, in Africa (Uganda) and South America (Chile), included pregnant women. In Uganda, Helicobacter pylori prevalence was very different according to the region considered (18.2–60.5%). In the northern part of the country (Apac district), home of the Langi population, a low prevalence was found despite a low socioeconomic status 1. The authors question whether the situation might be comparable to that of the Malay population in Malaysia.

In Chile, the authors focused on dyspeptic symptoms, especially hyperemesis occurring during pregnancy, and they found an association with a positive H. pylori status 2.

The prevalence in two rural villages in Bolivia was very high (80%) despite the inclusion of children suggesting an early acquisition in childhood in the country 3.

Among the studies from Asia, one focused on senior citizens (>60 years) from Beijing, China, and not surprisingly, a high prevalence (83.4%) was found 4.

The cross-sectional study performed in seven different areas in Japan began before 2000 (1997) and was heterogeneous with regard to the tests used, that is, serology, stool antigen test (SAT), and urine antibody test. An important heterogeneity of prevalence was found, ranging from 29.4% in Hokkaido to 54.5% in Yamagata. Indeed, the diagnostic test used in Hokkaido was the urine antibody test which is well known to have a much lower sensitivity than the others, explaining at least partially the low prevalence observed in this area. A higher prevalence was also found in men versus women (43.2 vs 37.6%, respectively) and was statistically significant, a situation which is not the rule but has already been observed. According to age, the highest prevalence was in those born between 1940 and 1949, and then it decreases. This is a clear birth cohort effect, that is, the acquisition rate decreased in successive generations of children as the living standards improved. There is a striking decrease in prevalence in the young Japanese population 5.

In Taiwan, a survey was carried out on a quarter of the population of Lanyu Island, where most of the people are of the Yami ethnicity, living in a closed environment. Their H. pylori prevalence was high (76.5%) versus the Han ethnic group (47.9%) 6.

Two African studies included patients, either at endoscopy in Butare, Rwanda 7, or dyspeptic patients in Ethiopia 8, and H. pylori prevalence was higher than 70%.

A cross-sectional study in Germany was published this year including subjects recruited between 1985 and 1989 as well as in 2002. H. pylori prevalence has evolved considerably over the last 20–30 years, but this study had the added interest of a multiplex serology (15 proteins tested). A major finding was the qualitative increase in the immune response to certain H. pylori antigens with age, indicating a lifelong stimulation of the immune system 9.

Two studies showed a lasting high prevalence in Southern and Eastern Europe. The first was carried out in 2002 in the Calabrian region of Italy in a village of 932 inhabitants. Among the adults, 459 were tested by urea breath test (UBT) with a high rate of positivity (72.1%) 10. The second was a national survey in Poland on 3307 adults. The prevalence of H. pylori infection was as high as 84.2% highlighting the traditional socioeconomic risk factors 11.

Another interesting study from Europe, published previously 12 and based on the status of mothers and their children, showed the same proportion of intergenerational decrease in H. pylori prevalence (76%) across nine ethnic groups living in Rotterdam, despite a higher prevalence in non-Dutch ethnicities 13.

A systematic review of the studies related to the prevalence of H. pylori infection worldwide with national coverage was also published last year 14. Thirty-seven studies were eligible from 22 countries reporting data from 1968 to 2011. The highest prevalences were found in Asia and Latin America. There was no study from Africa. Countries with a high gastric cancer rate had a two-fold higher prevalence of H. pylori infection. A decrease in prevalence was noted these last years, except in some countries with an already low prevalence like the Netherlands where the level remained stable.

Another systematic review focused on the studies performed in the Middle East. It included 26 articles from eight countries reporting data from 1989 to 2013. There were two studies on the healthy population in Iran published in 2009 and 2007, and the prevalence was 69 and 67.5%, respectively. Among the other countries, Egypt had the highest prevalence in healthy adults. In all of these countries, the prevalence was 70% or more 15.

Incidence of H. pylori Infection

Only one study from Italy 10 reported incidence data on 157 H. pylori-negative individuals in 2002 who participated in a survey in 2012 (of 207 according to the 2011 census). Four of them (2.5%) became infected with H. pylori during the 10-year period, that is, three adults aged 34, 60, and 71 years, and an adolescent aged 14 years, giving a yearly incidence rate of 0.25% (0.10–0.63). Based on the differences in VacA serology (positive) and UBT (negative) especially in the younger age groups, the authors claim that in these cases, the host was able to clear the infection spontaneously 10.

Risk Factors for H. pylori Infection

Last year's studies did not bring much new information concerning the risk factors for infection.

They confirmed the importance of the low socioeconomic status of the families as a risk for contamination.

In the Chinese study, surprisingly they found that a vegetarian diet was associated with a high H. pylori prevalence, in contrast to all of the other previously published data 4.

In Uganda, the risk of environmental factors, that is, water of diverse origin used in the household, was pointed out 1. Drinking spring water was also found to be a risk factor in a study of H. pylori infection in children in Madagascar after multivariate analysis 16. However, in a comprehensive study of environmental exposure in Canadian Arctic communities, untreated drinking water was not associated with increased H. pylori prevalence. Their multilevel models showed H. pylori clustering within households, with environmental exposure accounting for little of this clustering 17.

The data of the US National Health and Nutrition Examination Survey were examined for environmental risk factors. They found an important burden of infection associated with poor health and crowded living conditions. They also noted an increased risk with using a well as source of home water in participants >20 years 18. In addition, two European studies claimed positive detection of H. pylori by PCR in river water 19 and in drinking water 20.

Another risk factor of person-to-person transmission can also be attendance of children at day care centers. In a study on 1047 children aged 4–5 years in Portugal, the prevalence of H. pylori infection was 30.6%. It varied from 13.2% for those who never attended day care centers to 40.2% for those attending for >3 years. The OR after adjustments was 4.88 (95% CI 2.55–9.35) 21.

Transmission

It remains that transmission of H. pylori is essentially interhuman and mainly intrafamilial. Two Japanese studies used multilocus sequence typing and rapid amplified polymorphic DNA to show the strain transmission in families.

Of 35 index pediatric patients, the allele match was found between the index child and the mother for 25 (60%) and between the index child and both father and mother for 9 (25.7%) 22. In the other study, 19 isolates from five families were analyzed. Mother-to-child transmission was documented in four of the five families, father to child in two families and sibling to sibling in one family 23.

The transmission may be prevented by breast feeding in less economically developed families as shown in the study of Carreira et al. 24.

Later in life, antibiotic treatment appears to affect the prevalence of H. pylori infection 25.

Stool antigen test (SAT)

Endoscopy

A meta-analysis of 18 studies conducted by Chen et al. 38 on the appropriateness of prompt endoscopy as an initial strategy for the management of dyspepsia revealed that this is suitable for investigation in dyspeptic Asian patients. Novel endoscopic techniques enable endoscopists to observe microscopic structures and cellular morphology of the gastric mucosa in real time. Using narrow-band imaging-magnifying endoscopy, Liu et al. categorized the microstructure of the gastric mucosa of 90 patients into five types (A-E) and were able to predict the histologic types in a real-time manner. The sensitivity and specificity of the method were 85.4 and 81.7% for gastric inflammation, 71.8 and 95.2% for gastric intestinal metaplasia, and 80 and 98.9% for gastric carcinoma, respectively 39. In another study, endocytoscopy had 100% PPV for differentiating between normal tissue and intestinal metaplasia; however, the PPV for chronic gastritis and atrophic gastritis was only 62.5 and 40%, respectively 40.

Rapid Urease Test

Recent studies in a variety of clinical settings reconfirm that additional biopsies improve the performance of the rapid urease test (RUT). In a study comprised 116 patients with PU bleeding, Lee et al. evaluated the sensitivity of RUT using one piece of biopsy compared to more than one biopsies from the gastric antrum. The sensitivity increased from 61% using one biopsy to 74% using four antral biopsies 41. Similar results were observed in children by Seo et al. 42.

Culture

Improvements in culture methods of Helicobacter spp. have been described recently. A culture method based on microcapillaries in an aerobic environment has been developed with reported high sensitivity and moderate specificity (80%) 43. A new biphasic medium system, using a urea agar slant with overlaying Bolton broth supplemented with 10% defibrinated horse blood, detected more positive samples than a commercial RUT 44. Liang et al. 45 were able to obtain pure cultures of Helicobacter suis by growing bacteria as colonies on 1% brucella agar plates, followed by purification and enrichment by biphasic subculture. Interestingly, they observed differences in the bacterial genotype and phenotype of the isolated strains compared to their parental strains. Testing of bacterial colonies by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) is a fast and reliable method for bacterial identification and typing. Xiao et al. 46 added the data of protein components of 56 different H. pylori isolates to the original reference database of MALDI-TOF-MS Biotyper system (Bruker, Bremen, Germany) and used the extended database to identify 92 H. pylori strains. They observed a considerable improvement in the identification efficiency at the species level with the extended database.

Histology

The usefulness of special staining for H. pylori in gastric biopsies was assessed in recent studies. Chitkara evaluated the Diff-QuiK and immunohistochemical staining in gastric biopsy specimens and concluded that upfront staining is not cost-effective and should be performed only in selected specimens with chronic active, or moderate inactive gastritis 47. Comparing Giemsa, Warthin-Starry, acridine orange, and immunohistochemical stains to hematoxylin and eosin (H&E), Panarelli et al. 48 showed that ancillary staining is not necessary in most cases and should be considered only for biopsies with mild inflammation and for biopsies from patients under proton-pump inhibition or antibiotic treatment. On the other hand, a low cost, modified toluidine blue staining for H. pylori detection in gastric biopsies showed comparable sensitivity to immunoperoxidase staining, but significantly better sensitivity compared to H&E 49.

Molecular Methods

Real-time PCR and other molecular methods are increasingly used for the detection and typing of H. pylori and other Helicobacter species as well as the antibiotic susceptibility of the bacterium. Kalach et al. 50 in their evaluation of a quantitative real-time PCR (qPCR) for the diagnosis of H. pylori infection in gastric biopsies of French children concluded that H. pylori qPCR is a more precise test than routine culture, histology, or RUT alone and allows the detection of low bacterial loads. Novel qPCRs and nested PCR have been used to detect H. pylori in the oral cavity in specimens such as inflamed dental pulp, saliva, and dental plaque 51, 52. A novel method for the detection of 16S rRNA and ureC genes of H. pylori based on the GenomeLabTM Genetic Analysis System (Beckman Coulter) showed high performance when it was compared with conventional methods 53.

Formaldehyde and paraffin, used for tissue preservation, may impair microbial DNA yield and amplification efficiency in downstream molecular applications. Rabelo-Gonçalves et al. 54 compared five DNA extraction commercial methods and phenol–chloroform for the detection of H. pylori using 16S rRNA gene amplification by PCR from formalin-fixed paraffin-embedded liver tissue. Although the highest percentage of positive cases (70%) was found in samples extracted with phenol–chloroform, no statistical significant difference was found among the methods.

Omics-based approaches are increasingly used for H. pylori infection. Kim et al. 55 were able to detect H. pylori by pyrosequencing 16S rDNA amplicons of gastric biopsy specimens in all of the conventionally H. pylori-positive samples and surprisingly in 60% of the H. pylori-negative samples. Matsui et al. 56 performed SOLiD sequencing on two H. suis genomes to design corresponding PCR primers and were successful in detecting H. suis DNA in gastric biopsy specimens of experimentally infected mice. Two-dimensional (2D) gel electrophoresis was used 57 to compare the cellular proteomes of non-pylori Helicobacters (Helicobacter mustelae, Helicobacter felis, Helicobacter cinaedi, Helicobacter hepaticus, Helicobacter fennelliae, Helicobacter bilis and Helicobacter cholecystus) against H. pylori. The different species showed distinctive 2D protein profiles, and it was possible to clearly discriminate between gastric and enterohepatic Helicobacter species.

Trespalacios et al. designed new primers to detect the N87I mutation in the gyrA gene of H. pylori and used them in the allele-specific PCR assay to detect the levofloxacin resistance of the bacterium directly in gastric biopsies. The use of the new primers improved the sensitivity of the method for levofloxacin resistance prediction from 52 to 100% 58. The performance of the GenoType HelicoDR kit for the rapid detection of the A2143G mutation of the 23S rRNA conferring resistance to clarithromycin and the N87K mutation responsible for fluoroquinolone resistance was evaluated in 42 H. pylori isolates. The test had a >80% sensitivity and specificity for the detection of clarithromycin and fluoroquinolone resistance when compared with sequencing; however, it presented a sensitivity of 55% for clarithromycin resistance when compared to the MIC results 59.