Recurrence of Helicobacter pylori infection after several eradication therapies: long-term follow-up of 1000 patients

Abstract

Summary

Aim

To study the incidence of Helicobacter pylori recurrence, its chronological aspects, and the variables that might influence it.

Methods

A total of 1000 patients in whom H. pylori had been eradicated were prospectively studied. Therapies were classified as low and high efficacy regimens. Four to eight weeks after completion of therapy, ¹³C-urea-breath-test was performed, and it was repeated yearly up to 5 years. In some patients, endoscopy with biopsies was also performed to confirm H. pylori eradication.

Results

A total of 1000 patients were included, giving 2744 patient-years of follow-up. Seventy-one H. pylori recurrences were observed (2.6% per patient-year). Probability of being H. pylori-negative at 1 year was 94.7%, and at 5 years 90.7%. In the multivariate analysis, low age (OR: 1.84; 95% CI: 1.04–3.26) and low efficacy therapies (OR: 2.5; 1.23–5.04) correlated with 1-year H. pylori recurrence. Differences were observed when Kaplan–Meier curves were compared depending on age and therapy regimen.

Conclusion

Risk of posteradication H. pylori recurrence is higher during the first year, which suggests that most recurrences during this period are recrudescence and not true reinfections. H. pylori recurrence is more frequent in younger patients and in those treated with low efficacy therapies, but is exceptional if high efficacy therapies are used, in which case post-therapy eradication can be safely confirmed at 4 weeks with ¹³C-urea-breath-test.

Introduction

Nowadays no one can doubt the importance of Helicobacter pylori infection in the development of gastritis, gastroduodenal ulcer disease and gastric cancer. The most solid argument supporting the causal role of this micro-organism in peptic ulcer disease is the evidence that eradicating it is associated with a dramatic decrease in the rate of ulcer relapses.¹ Consequently, a definitive cure could be on the point to be brought about – and not just a temporary scarring – of gastroduodenal ulcer disease by eradicating H. pylori. Being able at last to change the natural course of peptic ulcer disease is one of the most significant advances in gastroenterology in recent decades. On the other hand, H. pylori infection has been shown to be closely related with gastric adenocarcinoma and with low-grade gastric mucosa associated lymphoid tissue (MALT) lymphoma. Furthermore, regression of MALT lymphoma has been demonstrated following H. pylori eradication.²

Given that a recurrence of infection with H. pylori seems to be a practically essential requisite for an ulcer relapse to occur subsequently, and that reappearance of the micro-organism explains some of the MALT lymphoma recurrences, the study of the incidence of the organism's recurrence represents an extremely important issue. The availability of therapies with optimal eradication rates would be of little value if the infection is again identified in a relatively short period, with the consequent risk for recurring ulcer or lymphoma.

Both the overall incidence of H. pylori recurrence and its chronologic aspects – the exact moment when it appears – have scarcely been studied. The influence of different variables is neither known on this recurrence; yet these are precisely the factors that are important in differentiating between true reinfection (i.e. when the patient is infected with a new strain of H. pylori) and recrudescence (when the pre-treatment strain of H. pylori becomes detectable again).³

Therefore, the aim of our study was not only to evaluate the incidence of recurrence of H. pylori infection in a 5-year follow-up study of 1000 patients, but also to assess the different variables that could influence it; among these variables, the most noteworthy are the following: (i) chronology of recurrences (do recurrences appear at a constant rate over the time?); (ii) age at reinfection (does age affect recurrences?); (iii) waiting time to define eradication of H. pylori after antibiotic treatment finalizes (is 4 weeks enough?); (iv) diagnostic methods to check for H. pylori eradication (is ¹³C-urea breath test sufficient?); and (v) efficacy of eradication regimen (does the efficacy of therapy influence further recurrences?). The response to the aforementioned questions will help to clarify the progression of post-therapy H. pylori infection and this will clarify our questions as to the true benefit of H. pylori eradication therapy.

Patients and methods

Patients

A prospective multicentre study was made of 1000 patients in whom H. pylori had been eradicated. Mean age (±s.d.) was 52 ± 19 years (median, 52 years), 63% were males and 28% were smokers. Seventy-nine per cent of patients were diagnosed with gastroduodenal ulcer disease and 21% with functional dyspepsia. Informed consent was obtained from all patients.

Therapy

Several eradication therapies were employed, and they were classified as low efficacy therapies (omeprazole plus amoxicillin, 32% eradication rate; omeprazole plus amoxicillin and metronidazole, 56% eradication rate) and high efficacy therapies (omeprazole plus clarithromycin and either amoxicillin or metronidazole, 85% eradication rate; bismuth subcitrate, tetracycline chlorhydrate and metronidazole, 77% eradication rate). The above-mentioned therapies were employed in several prospective protocols or studies (aimed to assess efficacy of different H. pylori eradication regimens), and we pooled them to obtain a larger sample for studying H. pylori recurrence (the main outcome in this study), as therapy efficacy in eradicating H. pylori was not the aim of the present study. The therapy regimen, as explained in the statistical analysis section, was included in the multiple logistic regression model and in the log-rank model to know whether the type of therapy influenced recurrence rate.

Diagnostic methods to confirm H. pylori eradication and follow-up

In all patients, ¹³C-urea breath (with citric acid and 100 mg of urea, as previously reported)⁴ was performed 4 weeks (380 patients) or 8 weeks (620 patients) after completion of therapy. In 810 patients, ¹³C-urea breath test was the only method for confirming H. pylori eradication. In the remaining 190 patients, endoscopy was also performed to confirm H. pylori eradication 4–8 weeks after completion of therapy, and at that time two biopsy specimens were obtained from both gastric antrum and body and were sent for histological examination (haematoxylin and eosin stain). Both ¹³C-urea breath test and histology were carried out by operators unaware of therapy and patients’H. pylori status. Eradication was defined as the absence of H. pylori by all performed diagnostic methods (that is, a negative ¹³C-urea breath test when only this test was performed, and both a negative ¹³C-urea breath test and histology when the two diagnostic procedures were performed). ¹³C-urea breath test was repeated yearly up to 5 years in order to study the incidence of H. pylori recurrence.

Statistical analysis

The 95% confidence interval was calculated. A P-value <0.05 was considered statistically significant. Categorical variables were compared with the χ² test, and quantitative variables with the Wilcoxon test (because distribution of values was not normal). A multiple logistic regression analysis was performed. The dependent variable was H. pylori recurrence at 1 year, and independent variables were: age (more or less than 52 years, the median age value), sex (male or female), smoking habit (smokers or non-smokers), diagnostic methods to confirm eradication (only ¹³C-urea breath test or breath test plus histology), therapy regimen (low or high efficacy; see ‘Therapy’ section), waiting time to define eradication of H. pylori after antibiotic treatment finalizes (4 or 8 weeks) and centre (hospital where the patient was included). We used a backward modelling strategy. The log-likelihood ratio was the statistic used for model comparison. In order to study the ‘time of H. pylori recurrence’ the Kaplan and Meier method was used, and differences between curves were evaluated with the log-rank test. Censoring during each yearly interval occurred at the end of the interval.

Results

Follow-up

In total, 1000 patients were followed up (357 for 1 year, 189 for 2 years, 23 for 3 years, 215 for 4 years and 216 for 5 years), giving 2744 patient-years of follow-up. Mean follow-up was similar for all the six centres participating in the study.

Recurrence of the infection

A total of 71 H. pylori recurrences were observed (53 at 1 year, 10 at 2 years, one at 3 years, three at 4 years and four at 5 years of follow-up), giving a yearly recurrence (rate of incidence, in per patients-year) of 2.6% (95% CI, 2.06–3.25%). Cumulative recurrence rates were 5.3% at 1 year, 6.8% at 2 years, 7% at 3 years, 7.6% at 4 years and 9.3% at 5 years (Figure 1). In other words, the probability of being H. pylori-negative at 1, 2, 3, 4 and 5 years was, respectively, 94.7% (95% CI, 93–96%), 93.2% (92–95%), 93% (91–95%), 92.4% (91–94%) and 90.7% (88–93%).

Figure 1. Kaplan–Meier curve for Helicobacter pylori recurrence.

Association between different variables and H. pylori recurrence rate

The characteristics of the studied variables with respect to H. pylori recurrence at 1 year are summarized in Table 1. The recurrence rate at 1 year in patients <52 years was 7.1% (95% CI, 4.7–9.4%), and only 3.9% (95% CI, 2.5–5.9%) when patients were ≥52 years (P = 0.028). Helicobacter pylori recurrence at 1 year in patients previously treated with low efficacy therapies was 11.3% (95% CI, 6.4–19.2%), and only 4.7% (95% CI, 3.5–6.3%) when therapies used were of high efficacy (P = 0.006). In the multivariate analysis, patient's age and therapy regimen were the only variables which correlated with H. pylori recurrence at 1 year (Table 2). However, sex, smoking, the number/characteristics of diagnostic methods for confirming H. pylori eradication (breath test ± histology), waiting time to define eradication of H. pylori after antibiotic treatment finalizes (4 vs. 8 weeks) and centre where the patient was included did not correlate with recurrence of the infection. Differences were observed when Kaplan–Meier curves were compared (log-rank test) depending on age (<52 or ≥52 years; P = 0.016; Figure 2) and therapy regimen (high or low efficacy; P = 0.002; Figure 3).

Table 1.  Characteristics of the studied variables with respect to Helicobacter pylori recurrence at 1 year

Variable	No recurrence	Recurrence	P-value
2 Methods, ^13C-urea breath test and histology; 1 method, ^13C-urea breath tests only; high efficacy, omeprazole plus clarithromycin and either amoxicillin or metronidazole; bismuth subcitrate, tetracycline chlorhydrate and metronidazole; Low efficacy, omeprazole plus amoxicillin; omeprazole plus amoxicillin and metronidazole; N.S., non-statistically significant differences.
Patients (number)	929	53
Age (mean ± s.d.)	52 ± 19	46 ± 16	0.005
Sex (% males)	62	72	N.S.
Smoking (% smokers)	28	30	N.S.
Diagnosis of eradication (2 methods/1 method) (%)	19/81	15/85	N.S.
Therapy regimen (high efficacy/low efficacy) (%)	91/9	79/21	0.006

Table 2.  Association between recurrence at 1 year and age and therapy regimen by multiple logistic regression

Variable	OR	95% CI	P-value
χ^2 model = 10.4; P = 0.005. OR, odds ratio; 95% CI, 95% confidence interval. * <52 years (the median age value). † Low efficacy therapies: omeprazole plus amoxicillin; omeprazole plus amoxicillin and metronidazole. Using high efficacy therapies as reference level: omeprazole plus clarithromycin and either amoxicillin or metronidazole; bismuth subcitrate, tetracycline chlorhydrate and metronidazole.
Age*	1.84	1.04–3.26	0.037
Therapy regimen†	2.5	1.23–5.04	0.011

Figure 2. Comparison of Kaplan–Meier curves for Helicobacter pylori reinfection depending on age.

Figure 3. Comparison of Kaplan–Meier curves for Helicobacter pylori reinfection depending on therapy regimen.

Discussion

In this study, we have evaluated the incidence of recurrence of H. pylori infection in 1000 patients in a 5-year follow-up study. This is, to our knowledge, the largest study up to now evaluating recurrence of H. pylori infection.³ Although other authors have previously included a higher number of patients in their studies, we have followed up a large number of subjects for several years – e.g. more than 200 patients were followed up to 5 years – resulting in the remarkably high (total) follow-up of 2744 patient-years.

A total of 71 recurrences were observed in our study, giving a yearly recurrence rate of 2.6%. This value is similar to that reported in most developed countries, but lower than in developing ones.³ The socioeconomic level of the country in question might be a predictor of reinfection risk. A recently performed systematic review calculated a mean annual reinfection rate of 3.4% and 8.7%, respectively in developed and developing countries.³ Therefore, the ‘cured once, cured forever’ statement⁵ may not hold true for H. pylori-infected patients living in developing countries. Nevertheless, although the overall reinfection rate is considerably higher in developing than in developed countries, several discrepancies show that reinfection is not consistently higher in areas of high H. pylori prevalence than in areas of low prevalence.⁶

Cumulative recurrence rates were 5.3% at 1 year, 6.8% at 2 years, 7% at 3 years, 7.6% at 4 years and 9.3% at 5 years (Figure 1), which indicates that the probability of recurrence is higher during the first year. Other authors have also observed that the risk of being reinfected with H. pylori seems to be mainly limited to a few months or 1 year after being given eradication treatment, as the incidence of delayed reinfection is substantially lower.⁷ If all H. pylori recurrences were true reinfections, one would expect these recurrences to appear at a constant rate over time, and this is not the case. In fact, some studies have evaluated the patients who had remained H. pylori negative 12 months after eradication and have shown that the recurrence rates are relatively low.^8–12 Beyond the first year, the rates come close to the rate of natural acquisition of H. pylori infection in adulthood.¹³ These observations suggest that most of the recurrences detected during the first months after treatment are actually late recrudescences. Several studies using molecular fingerprinting techniques (polymerase chain reaction) favour this hypothesis, as they confirm that the identified micro-organisms (before and after therapy) are usually genetically identical.^{3, 13}

The impossibility of detecting infection 1–2 months after completing therapy could be explained by several hypotheses: (i) lower bacterial density after antibiotic administration;⁸ (ii) persistence of micro-organisms in areas with gastric metaplasia in the duodenum,¹⁴ in inter- or intra-cellular regions;¹⁵ (iii) transitory modification of its morphology from spiral to coccoid by the H. pylori;¹⁶ and (iv) the micro-organism might lodge in reservoirs outside the gastrointestinal system, such as saliva or dental plaque.¹⁷

Mean age of our reinfected patients at 1 year was lower than that of patients without recurrence, and age correlated with H. pylori recurrence in the multivariate analysis. Thus, the recurrence rate at 1 year in patients <52 years was 7.1%, and only 3.9% when patients were ≥52 years. Furthermore, differences were observed when Kaplan–Meier curves were compared depending on age. Some observations suggest that both primary infection and reinfection with H. pylori is acquired mainly in childhood and only rarely at an adult age.^18–21 This hypothesis is supported by the results of the study by Rowland et al.²² which found that risk of reinfection amongst children under 5 years old was 26 times higher than that amongst older individuals. Likewise, some studies have observed that the average age of reinfected patients is lower than that of patients who remain H. pylori negative,^23–26 although other studies performed in children could not confirm that the risk of reinfection depends on age.²⁷ The reason for this relationship between lower age and H. pylori recurrence is unknown, nor is it known whether it is due to differences in the distribution of other risk factors among age groups.

Helicobacter pylori recurrence at 1 year in patients previously treated with low efficacy therapies was 11.3%, and only 4.7% when therapies used were of high efficacy. In our study, we classified as low efficacy therapies those combining omeprazole–amoxicillin (32% eradication rate) or omeprazole–amoxicillin–metronidazole (56% eradication rate). Although some authors have achieved relatively encouraging results with this last triple regimen, mean eradication rate of only approximately 70% was calculated in a systematic review.²⁸ Therapy regimen correlated with H. pylori recurrence in the multivariate analysis, and differences were observed when Kaplan–Meier curves were compared depending on type of therapy. Therefore, the lower the efficacy of an antibiotic therapy, the greater the likelihood that recurrence will occur, again suggesting that in these cases temporary ‘clearance’ has been achieved rather than true eradication.³ However, other data are in contrast with the concept that unsuccessful treatment regimens are associated to a high recurrence rate. Thus, some authors have shown that infection relapse does not differ between patients treated with dual treatments (with low efficacy) or triple regimens (highly effective),²⁹ and others have reported very low reinfection rates even when treatments with a very low H. pylori eradication efficacy have been prescribed.³⁰ Nevertheless, a different interpretation can be given for this relationship:³¹ the negative predictive value of a diagnostic method will be low when the prevalence of infection is high (as occurs after low efficacy eradication therapies) and, consequently, false negative results will be more frequent as the therapeutic efficacy lowers. Therefore, recurrences would not be influenced by the therapy per se, but would just reflect the inadequate negative predictive value of diagnostic methods when they are used after ineffective therapy. Nevertheless, whatever the interpretation, the clinical relevance of this observation on the prognosis of the patient is similar.

The optimum time for confirming success of eradication therapy is a matter of debate. In 1992, the American Food and Drug Administration recommended that this time should oscillate between 10 and 14 weeks.³² In addition, some authors have suggested postponing the confirmation of eradication until several months after therapy.^{11, 33–37} However, waiting time to define eradication of H. pylori after antibiotic treatment finalizes (4 vs. 8 weeks) did not correlate with recurrence of the infection in our study. Furthermore, several other studies suggest that a 4-week time span is enough to confirm eradication,^{1, 38} and this convention is generally accepted by the scientific community.^{7, 39} In summary, although the current functional definition for eradication (4-week rule) is arbitrary, post-therapy eradication can be safely confirmed at that time with ¹³C-urea breath test in clinical practice.

The number and characteristics of the diagnostic methods for confirming eradication also seem to affect future recurrence of H. pylori infection.^{34, 35} Obviously, less sensitive techniques (with more false-negative results) will favour the misinterpretation of the ‘eradication’ concept. For example, the rapid urease test is not reliable for confirming H. pylori eradication if it is the only diagnostic test used.^{40, 41} On the other hand, diagnostic methods capable of making an overall evaluation of the gastric mucosa (as the urea breath test) are especially appropriate for confirming eradication, in contrast to diagnostic tests based on gastric biopsy specimens, which, at least in theory, can be associated with sample errors. Results of the present study, where the number and characteristics of diagnostic methods to confirm H. pylori eradication (breath test vs. both breath test plus histology) did not correlate with incidence of recurrences, indicate that ¹³C-urea breath test is sufficient to safely confirm H. pylori eradication success, probably due to the remarkable accuracy of this method for the confirmation of H. pylori eradication.⁴²

Finally, a clinically relevant question is whether it is necessary to perform routine controls (for example with annually ¹³C-urea breath test) in patients with H. pylori eradication success, to detect recurrence of the infection. As the incidence of reinfection is relatively low, at least in industrialized countries as ours, it seems unnecessary to perform periodic controls after eradication success. However, in patients with bleeding peptic ulcers and with MALT lymphoma, where the clinical relevance of reinfection is especially evident – as in both cases reappearance of the disease after reinfection has been reported – control for recurrence of H. pylori may be suggested.³

Acknowledgements

This work was supported in part by a Grant from the Instituto de Salud Carlos III (C03/02).

Conflict of interest

None declared.