Association of Helicobacter pylori and protozoal parasites in patients with chronic diarrhoea.

Introduction An association of Helicobacter pylori and common protozoal parasites in patients with abdominal discomfort and chronic diarrhoea is unclear and may be pathological. Materials and methods One hundred and sixty-one patients with diarrhoea were compared to 114 age and sex matched controls. Stool samples were examined by microscopy and DNA extracted for PCR with specific primers for H. pylori and protozoal parasites Blastocystis sp., Entamoeba sp. (Entamoeba histolytica, Entamoeba dispar and Entamoeba moshkovskii) and Giardia duodenalis (G. duodenalis). Results There was a marked difference in the presence of parasites between patients and controls: no parasite 42/75%, one parasite 42/15%, two or more parasites 16/10%, respectively (p < 0.001). Patients with diarrhoea were more likely to be infected with Blastocystis sp (p < 0.001), E. histolytica (p = 0.027) and E moshkovskii (p = 0.003). There was no difference in the frequency of H. pylori (p = 0.528), G duodenalis (p = 0.697) or E dispar (p = 0.425). Thirty-three patients and 27 controls had H. pylori infection. Of these, 22 patients and 6 controls were infected with Blastocystis sp (p = 0.001), 6 patients and no controls were infected with E. histolytica (p = 0.02), whilst 7 patents and 9 controls were infected with E dispar (p = 0.292). Conclusion In this population, diarrhoea is linked to infection with Blastocystis sp, E. histolytica and E moshkoviskii. In H. pylori infection, diarrhoea is linked to Blastocystis sp and E. histolytica infection. These associations may be linked pathogenically.

Gastric lymphoma: association with Helicobacter pylori outer membrane protein Q (HopQ) and cytotoxic-pathogenicity activity island (CPAI) genes.

B-cell non-Hodgkin lymphoma (B-cell NHL) is the second commonest malignancy in the stomach. We determined the distribution of Helicobacter pylori outer membrane protein Q (HopQ) allelic type, cytotoxin-associated gene (cag)-pathogenicity activity island (cag-PAI) and vacuolation activating cytotoxin A (vacA) genes, respectively, in patients with B-cell NHL. We also compared them with their distribution in non-ulcer dyspepsia (NUD). H. pylori was cultured from gastric biopsy tissue obtained at endoscopy. Polymerase chain reaction was performed. Of 170 patients enrolled, 114 (63%) had NUD and 56 (37%) had B-cell NHL. HopQ type 1 was positive in 66 (58%) in NUD compared with 46 (82%) (P = 0·002) in B-cell NHL; HopQ type 2 was positive in 93 (82%) with NUD compared with 56 (100%) (P < 0·001) in B-cell NHL. Multiple HopQ types were present in 46 (40%) in NUD compared with 46 (82%) (P < 0·001) in B-cell NHL. CagA was positive in 48 (42%) in NUD vs. 50 (89%) (P < 0·001) in B-cell NHL; cagT was positive in 35 (31%) in NUD vs. 45 (80%) (P < 0·001) in B-cell NHL; left end of the cagA gene (LEC)1 was positive in 23 (20%) in NUD vs. 43 (77%) (P < 0·001) in B-cell NHL. VacAs1am1 positive in B-cell NHL in 48 (86%) (P < 0·001) vs. 50 (44%) in NUD, while s1am2 was positive in 20 (17%) in NUD vs. 46 (82%) (P < 0·001) in B-cell NHL. H. pylori strains with multiple HopQ allelic types, truncated cag-PAI evidenced by expression of cagA, cagT and cag LEC with virulent vacAs1 alleles are associated with B-cell NHL development.

Helicobacter pylori outer membrane protein and virulence marker differences in expatriate patients.

We studied the prevalence of Helicobacter pylori virulence markers, e.g. cytotoxin associated gene (cagA), cagA promoter, vacuolating associated cytotoxin A (vacA) alleles induced by contact with epithelium (iceA type), and outer membrane protein Q (hopQ) in expatriates and compared them with those in local residents. Gastric biopsies were obtained at endoscopy for culture, histology and PCR for virulence marker and hopQ. Of 309 patients, 236 (76%) were males with a mean age of 45 years. A total of 102 patients were expatriates. hopQ type 1 was present in 98 (47%) local residents compared to 88 (86%) expatriates (P < 0·001), while hopQ type 2 was present in 176 (85%) local residents, compared to 60 (59%) expatriates (P < 0·001). H. pylori virulence marker cagA was positive in 97 (47%) local residents compared to 86 (84%) expatriates (P < 0·001) while cagA-P was positive in 72 (35%) local residents compared to 87 (85%) expatriates (P < 0·001). iceA type 1 was positive in 157 (76%) local residents compared to 45 (44%) expatriates (P < 0·001), while iceA type 2 was positive in 81 (39%) local residents compared to 86 (84%) expatriates (P < 0·001). Distribution of H. pylori cagA, cagA promoter, iceA and hopQ type in local residents and expatriates was different. H. pylori virulence markers were associated with severe pathology in expatriates.

Helicobacter pylori: correlation of the virulence marker iceA allele with clinical outcome in a high prevalence area.

The association of Helicobacter pylori virulence marker 'induced by contact with epithelium A' (iceA) allele types was determined in H. pylori-related diseases and virulence markers. Gastric biopsies were obtained at EGD from patients for culture, histopathology and polymerase chain reaction (PCR) for iceA types, cagA and vacA alleles. Two hundred and eighty-four H. pylori isolates were examined. iceA type 1 was positive in 177 (62%) and iceA type 2 in 158 (56%). In iceA type 2, gastric ulcer was present in 34 (21%) (P < 0.001) and carcinoma in 28 (25%) (P = 0.002), compared to nine (8%) and 2 (2%) in iceA type 2-negative cases. For iceA type 2, 139 (88%) were associated with chronic active gastritis compared to 95 (75%) (P = 0.006) in iceA type 2-negative. H. pylori cagA was positive in 101 (64%) iceA type 2 strains compared to 57 (45%) in negative strains (P = 0.002). H. pylori iceA type 2 was dominant and associated with cagA, chronic active inflammation, gastric ulcer and carcinoma.

Seroprevalence of hepatitis E and Helicobacter pylori in a low socioeconomic area of a metropolitan city in a developing country.

This study aims to determine the prevalence of coinfection of H. pylori and hepatitis E virus (HEV) in the paediatric age group in an urban slum area of Karachi and identify risk factors associated with co-infection. Five hundred and forty children aged one to 15 years were investigated. Blood samples were collected and questionnaires completed on socio-demographic characteristics. Anti-H. pylori, HEV IgG and IgM antibodies were analysed by enzyme immunoassays (EIAs). The seroprevalence of H. pylori antibody was 47.2%, while that of HEV IgG and IgM was 14.4% and 2.4%, respectively. 12.4% exhibited seroprevalence for both H. pylori and HEV (IgG). In 67 (26%) cases positive for H. pylori IgG, HEV IgG positivity was also seen (P < 0.001). Only 13 (5%) positive for H. pylori were also positive for HEV IgM (P < 0.001). Only 11 (4%) HEV IgG-positive cases were H. pylori antibody-negative (P < 0.001). Hepatitis E virus was common in children who had access to municipal piped water (P = 0.025). H. pylori was common in children who used a non-flush toilet system (P < 0.001). Children exposed to H. pylori infection were also exposed to the risk of HEV.

Virulence markers of Helicobacter pylori in patients with diarrhoea-dominant irritable bowel syndrome.

Recent studies suggest that irritable bowel syndrome (IBS) is associated with low-grade inflammation. This study aims to determine the distribution of Helicobacter pylori cytotoxin-associated gene A (cagA) and vacuolating cytotoxin A (vacA) alleles (e.g., s1 and s2) in patients with diarrhoea-dominant IBS (IBS-D) as the latter causes vacuolation in colonic epithelial cells in vitro. One hundred and seventy patients meeting Rome III criteria for IBS-D (mean age: 40 +/- 15 years) were enrolled. Gastric biopsy was assessed histologically and DNA extraction was performed by polymerase chain reaction (PCR) for H. pylori genus 16S ribosomal DNA (16S rDNA), cagA and vacA allele s1 and s2. There was no age- or gender-related difference in H. pylori positivity in IBS-D compared to the control group. H. pylori was positive in 116 (68%) with IBS-D compared to 88 (55%) in the control group (P=0.01). cagA was positive in 73 (63%) with IBS-D compared to 42 (48%) in the control group (P=0.03). vacA s1 was positive in 61 (53%) with IBS-D compared to 32 (36%) in the control group (P=0.02). cagA s1 was positive in 39 (34%) with IBS-D compared to 13 (15%) in the control group (P=0.002).

Entamoeba species associated with chronic diarrhoea in Pakistan.

We determined the prevalence of Entamoeba (E.) histolytica, E. dispar and E. moshkovskii in patients with chronic diarrhoea associated with abdominal pain or discomfort mimicking irritable bowel syndrome. Stool samples were collected from 161 patients with chronic diarrhoea and from 157 healthy controls. Stool microscopy with modified trichrome stain, culture and polymerase chain reaction (PCR) for Entamoeba spp. differentiation was performed. Microscopy demonstrated Entamoeba cysts in 44% (57/129) of patients with diarrhoea compared to 29% (44/151) of controls (P=0·009). In patients with diarrhoea, PCR for E. histolytica was positive in 9% (11/129) (P=0·008), E. dispar in 19% (24/129) (P=0·117) and E. moshkovskii in 19% (24/129) (P<0·001). E. histolytica and E. moshkovskii were significantly associated with diarrhoea while E. dispar was found equally in both groups.

Microsporidial infections due to Encephalitozoon intestinalis in non-HIV-infected patients with chronic diarrhoea.

We determined the prevalence of microsporidia Enterocytozoon (Ent.) bieneusi and Encephalitozoon (E.) intestinalis infection in patients with chronic diarrhoea and hepatocellular carcinoma (HCC). A total of 330 stool samples were examined from 171 (52%) patients with chronic diarrhoea, 18 (5%) with HCC while 141 (43%) were controls. Stool microscopy, polymerase chain reaction (PCR) with species-specific primers for Ent. bieneusi and E. intestinalis and sequencing were carried out. Microsporidia were found by trichrome staining in 11/330 (3%) and E. intestinalis by PCR in 13/330 (4%) while Ent. bieneusi was not detected. PCR for E. intestinalis was positive in 8/171 (5%) stool samples from patients with chronic diarrhoea, 2/141 (1·4%) samples from healthy controls and in 3/18 (17%) samples from patients with HCC. In the chronic diarrhoea group, E. intestinalis was positive in 4/171 (2·3%) (P=0·69) stool samples compared to 2/18 (11%) (P=0·06) in the HCC group and 2/141 (1·4%) from healthy controls. E. intestinalis infection was significantly associated with chronic diarrhoea and HCC in these patients who were negative for HIV. Stool examination with trichrome or species-specific PCR for microsporidia may help establish the cause of chronic diarrhoea.

Helicobacter pylori: association with gall bladder disorders in Pakistan.

Helicobacter species colonise the biliary tract and therefore this study explores the relationship between of Helicobacter pylori and cholecystitis. Bile and gall bladder tissue samples were obtained from 144 patients who underwent cholecystectomy. Of these, 89 had chronic cholecystitis with cholelithiasis, 44 had gall bladder carcinoma and 11 had gall bladder polyps. Histopathology examination included special staining and immunohistochemistry (IHC), while Helicobacter species (H. pylori, H. bilis and H. hepaticus) were detected by the polymerase chain reaction (PCR). Sequencing and BLAST query of PCR products was undertaken and samples were considered to contain H. pylori if both PCR and IHC were positive. Immunohistochemistry for H. pylori was positive in 22 (25%) cases compared to five (9%) in the control group (P=0.02). Testing (PCR) for 16S rDNA was positive in 23 (26%) cases compared to six (11%) controls (P=0.03). Negative PCR results were obtained for H. bilis and H. hepaticus. Twenty-four (89%) were positive by both 16S rDNA PCR and IHC for H. pylori (P<0.001). Both PCR for 16S rDNA and IHC were positive in 21 (24%) cases compared to five (9%) controls (P=0.03). Sequencing of 16S rRNA and glmM PCR products were consistent with H. pylori. In conclusion, H. pylori DNA was demonstrated in cases of chronic cholecystitis and gall bladder carcinoma associated with cholelithiasis, but this association requires further study.

Risk factors associated with Helicobacter pylori infection treatment failure in a high prevalence area.

Triple therapy is commonly used for the treatment of Helicobacter pylori infection. We determined risk factors associated with its failure in compliant patients focusing on H. pylori density, virulence marker and 23S ribosomal RNA (rRNA) point mutations associated with clarithromycin resistance. H. pylori infection was diagnosed by 14C urea breath test (14C UBT) and rapid urease test or histology. Triple therapy with esomeprazole 20 mg b.i.d., amoxicillin 1 g b.i.d. and clarithromycin 500 mg b.i.d. was prescribed for 10 days. 14C UBT was repeated 4 weeks after treatment. In total, 111 patients [69 (62%) males] with a mean age of 46±16 years were enrolled. The mean age of treatment failure was 39±14 years compared to 48±16 years with eradication (P=0·002). Treatment failure was associated with younger mean age, point mutations in the 23S rRNA gene of H. pylori and vacA s1a and m1 when associated with cagA negativity.

Prevalences of Giardia lamblia and Cryptosporidium parvum infection in adults presenting with chronic diarrhoea.

Giardia lamblia and Cryptosporidium parvum are both waterborne pathogens associated with diarrhoea in developing countries. In a recent study based at the Aga Khan University in Karachi, 334 adults aged 16-83 years (178 patients with chronic diarrhoea and 156 diarrhoea-free volunteers who acted as controls) were checked for infection with these parasites, using stool microscopy and/or PCR. Overall, 21 (6.3%) and 29 (8.7%) of the subjects were found positive for G. lamblia by microscopy and PCR, respectively, while the corresponding values for C. parvum were 13 (3.9%) and 14 (4.2%). Although, compared with the diarrhoea-free controls, the patients with diarrhoea were not significantly more likely to be found infected with Giardia, either by microscopy [15 (8.4%) v. six (3.8%); P=0.085] or PCR [19 (10.7%) v. 10 (6.4%); P=0.167], they were significantly more likely to be found infected with C. parvum, both by microscopy [11 (6.2%) v. two (1.3%); P=0.024] and by PCR [12 (6.7%) v. two (1.3%); P=0.014]. The 19 patients found PCR-positive for Giardia comprised 10 (67%) of the 15 found smear-positive for the same parasite but only nine (5%) of the 163 found smear-negative (k=0.545; P<0.001). Similarly, the 12 patients found PCR-positive for Cryptosporidium comprised all 11 (100%) patients found smear-positive for the same parasite but only one (0.6%) of the 167 found smear-negative (k=0.954; P<0.001). Although C. parvum was associated with chronic diarrhoea in the present study, the carriage of G. lamblia often appeared asymptomatic.

Antibiotic susceptibility patterns of Helicobacter pylori and triple therapy in a high-prevalence area.

This study aims to determine primary Helicobacter pylori resistance and its effect on eradication of the organism. Ninety-two patients with dyspeptic symptoms were enrolled. H. pylori was cultured and antibiotic sensitivity was determined by the Epsilometer test (Etest) for clarithromycin (CLR), amoxicillin (AMX) and metronidazole (MTR). 23S ribosomal RNA (rRNA) point mutations associated with clarithromycin resistance were also detected. Patients were treated with omeprazole (40 mg daily), CLR (500 mg) and AMX (1g twice a day) for 14 days. A 14C-urea breath test (14C-UBT) was repeated four weeks after completion of treatment to confirm eradication. Triple therapy failure was seen in 30 (33%) patients. The resistance rates were: CLR 33% (30/92), MTR 48% (44/92) and AMX 2% (2/92). Clarithromycin resistance (CLR-R) was present in the 16-39 age group in 21 (47%) (P = 0.007) compared to nine (19%) in the 40-79 age group. CLR resistance was seen in 30 H. pylori isolates, 20 (67%) from patients with non-ulcer dyspepsia (NUD), six (20%) with gastric ulcer (GU) and four (13%) with duodenal ulcer (DU). Triple therapy failure was associated with CLR-R in 28 (93%) (P < 0.001). CLR-R mutations were present in 30 (33%) and were associated with treatment failure in 27 (90%; P < 0.001). They were present in 20 (44%) isolates obtained from patients in the 16-39 age group (P = 0.018). Treatment failure was associated with A2142G mutation in 20 (67%; P < 0.001), A2143G mutation in 12 (40%; P < 0.001) and A2142C mutation in five (17%; P = 0.003). In conclusion, triple therapy failure was associated with CLR-R. Metronidazole resistance exceeded that of CLR, hence it cannot be substituted for CLR in a triple therapy.

Helicobacter pylori infection in children: population-based age-specific prevalence and risk factors in a developing country.

AIM: We estimated the prevalence, age of acquisition and risk factors for Helicobacter pylori (H. pylori) seroprevalence in children aged 1-15 years. METHODS: Exposure was assessed using ELISA. Parents responded to a questionnaire regarding number of individuals sharing house, rooms, water source, latrines, housing and assessment of socioeconomic status (SES) by Hollingshead Index. RESULTS: Serum of 1976 children was tested. Helicobacter pylori seropositivity in children aged 11-15 years was 53.5% (OR: 2.0, 95% CI: 1.58-2.5). It increased with moderate crowding index (CRI) of 2-4 to 45.9% (OR: 1.23, 95% CI: 0.92-1.63) and to 51.2% with CRI >4 (OR: 1.52, 95% CI: 1.12-2.06). In middle SES, seropositivity was 50.5% (331/655) (OR: 1.7, 95% CI: 1.29-2.35), whereas in lower SES, it was 47.1% (500/1062) (OR: 1.5, 95% CI: 1.1-2.0). Multivariate analysis showed that Helicobacter pylori seroprevalence was high in children aged 6-10 and 11-15 years (OR: 1.5, 95% CI: 1.2-1.9 and OR: 1.9, 95% CI: 1.56-2.47 respectively), in lower-middle SES (OR: 1.6, 95% CI: 1.2-2.1 and OR: 1.5, 95% CI: 1.10-2.0 respectively) and in uneducated fathers (OR: 1.58, 95% CI: 1.27-1.95). CONCLUSION: Helicobacter pylori seropositivity increases with age, in low-middle SES and is related to father's educational status. Reducing H. pylori seroprevalence will require improvement in sanitary conditions and educational status of the population.

Low prevalence of the intact cag pathogenicity island in clinical isolates of Helicobacter pylori in Karachi, Pakistan.

Clinical diseases that follow Helicobacter pylori infection are associated with expression of the cagA gene, a part of cytotoxin-associated gene pathogenicity island (cag-PAI). This study aims to determined whether or not the presence of cagA is associated with the presence of complete cag-PAI and to evaluate inflammatory changes associated with the five loci in the cag-PAI of H. pylori comprising cagA, cagA promoter region (cagAP), cagE, cagT and the left end of the cagA gene (LEC). H. pylori isolates were obtained from patients with dyspeptic symptoms. Clinical strains of H. pylori were screened by the polymerase chain reaction (PCR) for respective genes of the cag-PAI. Of 115 H. pylori isolates, 31 (28%) were positive for the five cag-PAI loci. H. pylori isolates with intact cag-PAI were associated with gastric carcinoma (GC; n=9 [60%]) and gastric ulcer (GU; n=5 [45%]) compared to non-ulcer dyspepsia (NUD; n=14 [18%]) (P=0.001 and P=0.049, respectively). In patients with intact cag-PAI, acute on chronic inflammation was present in 25 (81%) and was more common than chronic inflammation (P=0.013). The cagE and cagAP had deletions in 25 (37%) and 23 (35%) cases, respectively. The cagAP region was significantly associated with GC (n=12 [80%], P<0.001) and GU (n=9 [82%], P=0.001) compared to NUD (n=24 [30%] and with significant acute on chronic inflammation (n=40 [80%], P=0.007). The distribution of vacAs1a with intact cag-PAI in GC was 9 (60%) and in NUD was 10 (13%) (P<0.001). The presence of the cagA gene does not signify presence of an intact cag-PAI. Most of the H. pylori isolates studied had partial cag-PAI with missing cagE and cagA promoter regions.

Comparison of biopsy-based methods for the detection of Helicobacter pylori infection.

Various biopsy-based methods for the detection of Helicobacter pylori are evaluated to determine their sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV), followed by polymerase chain reaction (PCR) for the 16S ribosomal RNA (rRNA) gene of H. pylori (16S PCR) to confirm the results. Seventyfive patients (65% [49] males, age range: 17-77 years, mean 42+/-14.6 years) with dyspeptic symptoms are included in the study. Gastric antrum biopsy specimens collected during endoscopy are tested using a urea agar base enriched with 40% urea solution (eUAB, Oxoid)), a commercial rapid urease test (Pronto Dry, Medical Instrument Corp, Switzerland), histopathology and 16S PCR. The eUAB test showed 97% sensitivity, 86% specificity, 84% PPV, 97% NPV and 91% accuracy when the diagnosis of H. pylori infection was made with positive Pronto Dry and histopathology. Pronto Dry showed 100% sensitivity, 82% specificity, 80% PPV, 100% NPV and 89% accuracy when the diagnosis of H. pylori infection was made on positive histopathology and eUAB. Thus, the eUAB can be used as a rapid urease test. It is economical and has a sensitivity and specificity comparable to a commercially available rapid urease test to detect urease activity of H. pylori in gastric biopsy.

In vitro susceptibility of Blastocystis hominis isolated from patients with irritable bowel syndrome.

This study aims to determine the growth pattern and in vitro susceptibility of clinical isolates of Blastocystis hominis to different concentrations of metronidazole, furazolidone and ciprofloxacin. Stool specimens from 25 consecutive patients with irritable bowel syndrome presenting to the gastroenterology department of Aga Khan University Hospital between January and May 2003 are examined by microscopy and cultured for B. hominis. Drug susceptibility assays are performed for metronidazole, furazolidone, and ciprofloxacin using final concentrations of 0.01 mg/mL and 0.1 mg/mL. The effect of the drugs is assessed after B. hominis culture for 48 h. With furazolidone and metronidazole, 68% (17/25) and 60% (15/25) of B. hominis isolates, respectively, failed to grow at drug concentrations of both 0.01 mg/mL and 0.1 mg/mL. However, ciprofloxacin failed to suppress growth completely at both concentrations. B. hominis resistance to furazolidone, metronidazole and ciprofloxacin at 0.01 mg/mL was 32% (8/25), 40% (10/25) and 100% (25/25), respectively. B. hominis isolates varied in their degree of susceptibility to the three drugs studied, being greater with furazolidone than with metronidazole, and complete resistance with ciprofloxacin.

Irritable bowel syndrome in health care professionals in Pakistan.

OBJECTIVE: To evaluate the symptomatology of irritable bowel syndrome (IBS) among health care professionals attending an IBS symposium in a tertiary care university hospital. METHOD: A questionnaire designed to incorporate Manning and Rome II criteria was distributed among participants of an IBS symposium, most of them were health care professionals. A total of 100 questionnaires were distributed, 41 had symptoms fulfilling criteria of IBS. In these patients male: female ratio was 28:13 with age range 18-68. RESULTS: The predominant symptom was abdominal pain 87.8 % (36/41) which was aggravated post-prandially 72.2% (29/41), relieved following defecation in 87% (35/41) with a sense of incomplete evacuation 85.3% (35/41) and distention after defecation in 80.4% (33/41). Anxiety and depression was present in 80% (33/41) as an extraintestinal symptom. CONCLUSION: Irritable bowel syndrome is common in health care workers with intestinal and extraintestinal manifestations being equally common.

Helicobacter pylori cagA and vacA cytotoxin genes in Changsha, China.

Cytotoxin-associated protein (cagA) and the vacuolating cytotoxin (vacA) encoded by cagA and vacA genes are virulence determinants of Helicobacter pylori. In earlier studies among Chinese patients, all H. pylori strains were cagA-positive and vacAs1a/m2 type. Here, we determine the cagA, vacA and allele status of H. pylori strains isolated from patients with upper gastrointestinal symptoms in Changsha, China. Forty strains of H. pylori isolated from patients with peptic ulcer disease between March 1997 and August 1999 were recovered from storage at -80 degrees C and studied by the polymerase chain reaction (PCR) for cagA and vacA genotypes. cagA was positive in 75% of H. pylori isolates. Patients with peptic ulcer demonstrated cagA in 83% (15/18), compared with 68% (15/22) patients with superficial gastritis. vacAs1 allele was carried in 82.5% (33/40) isolates, of which 52.5% (21/40) were subtype vacAs1a/m2 and 17.5% (7/40) were subtype vacAs1b/m2.

Antibiotic susceptibility of Helicobacter pylori in the Chinese population.

AIM: To assess antibiotic susceptibility of Helicobacter pylori (H. pylori) strains to metronidazole, clarithromycin and tetracycline in the Chinese population, and to test the stability of antibiotic resistance in H. pylori 1 year after storage at -80 degrees C. METHODS: Fifty H. pylori strains isolated from patients with peptic ulcer disease were recovered from storage at -80 degrees C. Susceptibility of these strains to metronidazole, clarithromycin and tetracycline was determined by using validated disk diffusion tests, which was repeated 1 year after storage at -80 degrees C. The DNA profiles of each strain were determined by using the polymerase chain reaction-based-random amplified polymorphic DNA fingerprinting technique (PCR-RAPD). This was repeated if any change in antibiotic susceptibility pattern was noticed. RESULTS: The resistance rate was 50% to metronidazole and 8% to clarithromycin. None of the strains was resistant to tetracycline. A dual resistance to metronidazole and clarithromycin was demonstrated in three H. pylori strains. The antibiotic susceptibility test reproduced itself in 92% (36 of 39) of the strains 1 year later; the three strains with dual resistance exhibited susceptibility to both antibiotics. Variation in antibiotic susceptibility pattern in the three H. pylori strains was associated with change in the RAPD fingerprint. CONCLUSION: The prevalence of resistance in H. pylori is high to metronidazole but low to clarithromycin in the Chinese population. The disk diffusion test appears to be a simple and reliable test, while antibiotic resistance in some H. pylori strains may disappear after long-term storage at -80 degrees C.