Does Helicobacter pylori Eradication Reduce
the Risk of Open
Angle Glaucoma in Patients With Peptic Ulcer
Disease?
Hsin-Yi Chen, MD, Cheng-Li Lin, MSc, Wen-Chi Chen, MD, and Chia-Hung
Kao, MD
INTRODUCTION
G
laucoma is the leading cause of blindness, but manypathogenesis remain to be clarified.1 Helicobacter pylori
infection has been noted to play some role in pathogenesis of glaucoma.1,2 Some molecular mechanisms were proposed to
explain the relationship between glaucoma and H pylori infection,
3 including proinflammatory and vasoactive materials
release, platelet and platelet-leukocyte aggregation promotion, and apoptotic cascades influence.4–7
Previous studies provide important evidence for the potential roles of H pylori infection in glaucoma pathogenesis; however, there is still no clear answer as to whether H pylori eradication therapy prevents future glaucoma development.4,8 Based on the
hypothesis that H pylori eradication may be a feasible method for glaucoma prevention, we conducted a population-based retrospective cohort study of patients with peptic ulcer disease who
received H pylori eradication therapy over a 10-year period. The primary outcome of interest was whether early H pylori eradication is associated with decreased risk of primary open angle
glaucoma (POAG) in patients with peptic ulcer disease.
METHODS
Data Source
This retrospective cohort study was retrieved from the
Longitudinal Health Insurance Database (LHID2000), derived from the Taiwan National Health Insurance (NHI) program. The
NHI program, launched in 1995, covers more than 99% of the population, which is currently 23 million (http://www.nhi. gov.tw/english/index.aspx). This mandatory universal program offers comprehensive medical coverage, including outpatient, inpatient, emergency, dental, and traditional Chinese medicine services and prescription drugs, to all Taiwanese residents. The LHID2000 contained 1 million insurant randomly selected from the year 2000 Registry for Beneficiaries under the NHI program. This secondary dataset was encrypted prior to its release for research purposes to protect privacy. The study was approved from full review by the Institutional Research Ethic Committee (CMU-REC-101–012). Diagnostic codes based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).
Sampled Participants
From the LHID 2000, patients who were diagnosed of
peptic ulcer (ICD-9-CM codes: 531, 532, and 533) and received H pylori eradication therapy at the same time were recruited. Helicobacter pylori eradication with triple or quadruple therapy was defined as proton pump inhibitor or H2 receptor blocker,
plus clarithromycin or metronidazole, plus amoxicillin or tetracycline, with or without Bismuth (details for all eligible H pylori
eradication regimens are reported previously).9 These drug
combinations were prescribed within the same prescription order, and the duration of therapy was between 7 and 14 days. One year was chosen as the cutoff value based on the distribution of H pylori eradication date after diagnosed of peptic
ulcer. Patients who received H pylori eradication therapy within the first year after the diagnosis of peptic ulcer were included in the ‘‘early eradication cohort.’’ Patients included in the ‘‘late eradication cohort’’ were given H pylori eradication therapy 1 year or more after diagnosed of peptic ulcer. The date of the first received H pylori eradication therapy was used as the index date. We excluded patients diagnosed with POAG (ICD-9-CM 365.1) at the baseline and <20 years old and those without information on age and sex. Fourfold of control patients were randomly selected from LHID 2000 beneficiaries without peptic
applied the same exclusion criteria used in selecting H pylori
eradication cohort. The control cohort patients were frequencymatched with the H pylori eradication cohort by age (every
5-year span), sex, and the year of receiving H pylori eradication therapy.
Outcome and Comorbidities
All patients were followed up until a diagnosis of POAG, loss to follow-up, death, the date of withdrawing from the insurance, and the end of 2011, whichever date came first. We considered several well-known risk factors of POAG including hypertension 9-CM codes 401–405), diabetes
(ICD-9-CM code 250), hyperlipidemia (ICD-(ICD-9-CM code 272), coronary artery disease (CAD) (ICD-9-CM codes 410–414), to be
the comorbidities.
Statistical Analysis
The differences between the covariates of the H pylori
eradication and control cohorts were analyzed using the Chisquared test for categorical variables or t test for continuous
variables. The incidence of POAG was calculated for each H pylori eradication subgroup (early and late) and for the control cohort. Poisson regression model was used to assess glaucoma incidence rate ratios (IRRs) for H pylori eradication cohort when compared to control cohort. Multivariable Cox proportional
hazard regression analysis was performed to estimate the relative hazard ratios (HRs) of POAG development for H pylori eradication cohort with adjustment of age, sex, and comorbidities of
hypertension, diabetes, hyperlipidemia, and CAD when compared with the control cohort. The risk changed over time was
further evaluated by stratifying the follow-up duration into 2 periods (_5 and >5 years), with both IRR and HR measured for H pylori eradication cohort when compared with control cohort. All models were also used to estimate the risk of POAG for late
H pylori eradication cohort when compared with the early H pylori eradication cohort. All analyses were performed using
the SAS statistical package (version 9.2 for Windows; SAS Institute, Inc., Cary, NC). A 2-tailed P value of <0.05 indicated the statistical significance level.
In total, we included 6061 subjects receiving H pylori eradication therapy, among whom 2876 were early H pylori eradication cohort, and 3185 were late H pylori eradication cohort. The control cohort consisted of 24,244 subjects. The baseline characteristics of the patients in the 3 cohorts are presented in Table 1. Among the 3 cohorts, the majority of patients were <49 years (52.9% in early H pylori eradication, 36.2% in late H pylori eradication, and 44.1% in control cohort, respectively) and were male (59.7% in early H pylori eradication, 53.0% in late H pylori eradication, and 56.2% in control
cohort, respectively). The mean age was 52.5_14.6 years in the H pylori eradication cohort and 52.0_14.9 years in the control cohort. Compared with the control cohort, the H pylori eradication cohort was more likely to have all of the listed comorbidities in Table 1. Compared with the patients in early H pylori eradication, those in late H pylori eradication were more likely to have comorbidities. The mean duration of follow-up for control cohort was 6.71 years, approximately 0.5 year longer than that for early H pylori eradication (6.08 years), and late H pylori eradication (6.26 years) (data not shown).
The higher incidence of POAG was observed in the late
H pylori eradication cohort and in the early H pylori eradication cohort than in the control cohort (1.57, 1.32, and 0.95, per 1000 person-year, respectively; Table 2). The POAG incidence was greater in women in the late H pylori eradication cohort (2.10, per 1000 person-year). The glaucoma incidence increased with age in all 3 cohorts. The age-specific early H pylori eradication to control relative risk was greater for the ages 50 to 65 years (adjusted HR¼1.94; 95% confidence interval [CI] ? 1.05– 3.60). The glaucoma incidence increased with comorbidity in all 3 cohorts. We further analyzed the association between late H pylori eradication and the risk of POAG stratified by comorbidity and the result shows that 2.93-fold of POAG risk was
significantly observed in patients without comorbidity (95% CI ? 1.52–5.64).
Table 3 shows that the overall risk of glaucoma was not significantly higher in late H pylori eradication cohort than in
CI¼0.48–1.53). The age-specific adjusted HR of glaucoma in the late H pylori eradication cohort to early H pylori eradication cohort was not significant for the younger group (adjusted HR¼3.68; 95% CI ? 0.95–14.3). Furthermore, we compared incidence densities and HRs of POAG among 3 cohorts by follow-up duration (Table 4). The incidence of glaucoma was greater in the late H pylori eradication cohort when follow-up duration _5 years (1.59, per 1000 person-years). On the contrary, when follow-up duration >5 years, the incidence of
POAG was greater in the early H pylori eradication cohort (1.68, per 1000 person-years). However, these relationships were not associated with a significantly increased or decreased risk of POAG in multivariable analyses.
DISCUSSION
The risk of glaucoma among patients with H pylori infection has been widely studied. Higher prevalence of H pylori infection was found in the sera of patients with POAG and PXFG patients compared with control group.5 A possible
association between H pylori infection and glaucoma was supported again that glaucoma-related parameters were
improved in the subgroup of patients in whom H pylori eradication was successful.8 There are additional published data further
reinforcing the association glaucoma and H pylori infection in other ethnicities, including Greece, Turkey, Iran, Korea, India, and China, by using serology, aqueous humor and anti-H pylori IgG antibodies and/or histology.10–12 In Zavos’s study,12
H pylori bacteria have been detected histologically in eye biopsies of POAG patients. On the contrary, in a Canadian study, the authors suggest H pylori infection is not associated with POAG based on serum test.13 And in a Israel study,14 no
association between H pylori infection or CagA-bearing strains and glaucoma was noted. Because of different study designs and methods, it is not suitable to directly compare the outcomes among all these studies. Here, in our study, we aim to clarify the role of H pylori infection in glaucoma; therefore, the study cohort was subdivided into 2: early eradication cohort (2876 subjects) and late eradication cohort (3185 subjects). Logically, early eradication cohort should have shorter period of H pylori
exposure time than late eradication cohort. Slightly higher incidence of POAG was observed in the late H pylori eradication cohort and in the early H pylori eradication cohort than in the control cohort (1.57, 1.32, and 0.95, per 1000 person-year, respectively). This finding first confirms our hypothesis that longer period of H pylori exposure would have slightly higher POAG risk compared to shorter period exposure of H pylori. The glaucoma incidence increased with age and with comorbidity in all 3 cohorts. The age-specific early H pylori eradication to control relative risk was greater for the ages 50 to 65 years (adjusted HR¼1.94; 95% CI ? 1.05–3.60). We further analyzed the association between late H pylori eradication and the risk of POAG stratified by comorbidity, and the result shows that 2.93-fold of POAG risk was significantly observed in patients without comorbidity (95% CI ? 1.52–5.64). This result again supports our presumed thought that longer exposure of H pylori would increase POAG risk even without the effect of comorbidity.
Interesting finding was noted that the overall risk of
glaucoma was not significantly higher in late H pylori eradication cohort than in early H pylori eradication cohort (adjusted
HR¼0.85, 95% CI¼0.48–1.53); and the age-specific adjusted HR of glaucoma in the late H pylori eradication cohort was not significant higher for the younger group (adjusted HR¼3.68; 95% CI ? 0.95–14.3) when compared to early H pylori
eradication cohort. All these results suggest that glaucoma risk was not significantly different between early eradication cohort and late eradication cohort.
To understand whether different follow-up duration would
influence glaucoma risk, further analysis was performed. Logically, follow-up duration means the H pylori exposure duration.
Although incidence of POAG was greater in the late H pylori eradication cohort when follow-up duration _5 years (1.59, per 1000 person-years) and was greater in the early H pylori eradication cohort (1.68, per 1000 person-years) when follow-up duration >5 years. However, all these results were
not significantly different related to the follow-up duration (H pylori exposure duration).
To solve the hypothesis if H pylori eradication plays some role in glaucoma treatment, the optimal timing for eradication is an important issue. Although our result infers the slightly higher
risk of POAG in H pylori eradication cohort but not achieve statistically meaningful result compared to normal controls,
glaucoma risk is not significantly different among 3 cohorts. This finding has important value because it is the few one which addressed the risk of POAG and H pylori eradication in patients with peptic ulcer disease based on a large nationalized health insurance database in a Chinese population. Based on some limitations of present study design and method, it is a pity that we could not achieve definite conclusion regarding the role of H pylori eradication in glaucoma treatment.
There are several limitations to our study. First, our
observations were a retrospective cohort study based on claimed database from Taiwan NHIRD. Certain selection biases may exist, and caution must be taken in extrapolating our results to other ethnic populations. Second, the definitions of early and late eradication cohorts (1 year) was referred from another published work also from Taiwan NHIRD.9 Different definition
for early or late should have different outcome. Third, we did not analyze the relative risk of glaucoma in patients with peptic ulcer diseases without H pylori eradication therapy. According to our observations about general clinical practice pattern in Taiwan, patients with H pylori infection almost receive eradication therapy; therefore, the H pylori infection case without
eradication therapy is quite few. Fourth, we were unable to get data to verify whether H pylori eradication was effective based on this database. Therefore, we can only calculate and compare the IRRs and HRs in early and late eradication cohorts, not
among those in whom H pylori was eradicated and not eradicated.
9 As we know so well that eradication therapy with proton
pump inhibitor, clarithromycin and amoxicillin has been extensively used for years, although it fails in a considerable percentage of patients.15 Also in a community-based study from
early and late eradication cohorts
should be similar, the influence of the uncalculated failure rates may still bias the statistics. However, this is an inevitable study limitation in this kind of claims database study. Fifth, the glaucoma definition was completely relied on ICD coding (365.1) based on NHIRD; and it is potentially not completely accurate. Miscoding problem might exist in this study. In conclusion, in this nationwide, long-term cohort study,
glaucoma risk is not significant different among normal population, early eradication cohort and late eradication cohort. The
role of H pylori eradication in glaucoma treatment is not yet clear based on the current result, further large-scale prospective study should be carried on to further elucidate this important issue.
