Risk Factors and Incidence of Gastric Cancer After Detection of Helicobacter pylori Infection: A Large Cohort Study

Abstract

Background & Aims:

Nearly all studies of gastric adenocarcinoma in the United States have relied on national cancer databases, which do not include data on Helicobacter pylori infection, the most well-known risk factor for gastric cancer. We collected data from a large cohort of patients in the United States to calculate the incidence of and risk factors for non-proximal gastric adenocarcinomas after detection of H pylori. Secondary aims included identifying how treatment and eradication affect cancer risk.

Methods:

We performed a retrospective cohort study, collecting data from the Veterans Health Administration on 371,813 patients (median age, 62 years; 92.3% male) who received a diagnosis of H pylori infection from January 1, 1994 through December 31, 2018. The primary outcome was a diagnosis of distal gastric adenocarcinoma 30 days or more after detection of H pylori infection. We performed time to event with competing risk analysis (death before cancer a competing risk).

Results:

The cumulative incidence of cancer at 5, 10, and 20 years after detection of H pylori infection was 0.37%, 0.5%, and 0.65%, respectively. Factors associated with cancer included older age at time of detection of H pylori infection (sub-hazard ratio [SHR], 1.13; 95% CI, 1.11–1.15; P<.001), black/African American race (SHR, 2.00; 95% CI, 1.80–2.22), Asian race (SHR, 2.52; 95% CI, 1.64–3.89) (P<.001 for race), Hispanic or Latino ethnicity (SHR, 1.59; 95% CI, 1.34–1.87; P<.001), or history of smoking (SHR, 1.38; 95% CI, 1.25–1.52; P<.001). Women had decreased risk of gastric adenocarcinoma compared with men (SHR, 0.52; 95% CI, 0.40–0.68; P<.001); patients whose H pylori infection was detected based on serum antibody positivity also had a reduced risk of cancer (SHR 0.74; 95% CI, 0.54–1.04; P=0.04). Patients who received treatment for their H pylori infection still had an increased risk of gastric cancer (SHR, 1.16; 95% CI, 0.74–1.83; P=.51), but confirmed H pylori eradication after treatment reduced risk of gastric cancer (SHR, 0.24; 95% CI, 0.15–0.41; P<.001).

Conclusions:

In a study of 371,813 veterans with a diagnosis of H pylori infection, we found significantly higher risks of gastric cancer in racial and ethnic minorities and smokers. Treatment of H pylori infection only decreased risk if eradication was successful. Studies are needed on the effects of screening high-risk persons and to identify quality measures for diagnosis, resistance patterns, and treatment efficacy.

Among US patients with H pylori, racial/ethnic minorities and smokers have a higher risk of future gastric cancer. Eradication of H pylori infection decreases the risk.

INTRODUCTION

Gastric cancer is the third most common cause of cancer death worldwide.¹ Although less common in the United States, with an annual incidence of 26,000 cases, it is among the most fatal, with 5-year survival under 30%.^{2, 3} Helicobacter pylori (HP), a class I carcinogen, is a causative agent in the cascade leading to gastric adenocarcinoma (GAC), particularly non-proximal cancers.^4–6 Yet infection alone is not sufficient for carcinogenesis, illustrated by the discordant association of high HP prevalence with low GAC incidence in sub-Saharan Africa (the “African enigma”) or disparate rates of GAC throughout Middle Eastern countries despite high HP burden.^7–10

Despite robust data on GAC and the role of HP from Asia, nearly all large studies in the US have used national cancer databases, which contain limited demographic and environmental exposures, medical history, and most importantly, no data on HP.^11–16 As a result, there are critical knowledge gaps in our understanding of the incidence of non-proximal GAC in the setting of HP in the US, risk factors associated with future GAC, and the impact of HP treatment. Modeling studies have been performed to assess which populations in the US may benefit from screening, but were limited by a lack of robust data on the incidence of GAC in a US population.^{17, 18}

We aim to determine the incidence of non-proximal GAC after detection of HP infection in a large US cohort and further define how demographics, environmental factors, and treatment of HP infection impact the incidence of GAC. In doing so, we sought to overcome current knowledge gaps in GAC, including potentially modifiable risk factors in a US population, to best inform screening strategies.

METHODS

This retrospective cohort study was conducted within the Veterans Health Administration (VHA) Corporate Data Warehouse (CDW), which includes data from the unified electronic medical record of all VHA facilities (i.e., hospitals and outpatient) since 10/01/1999.

Study cohort

We first identified patients with HP infection based on any of the following: 1) HP infection on endoscopic pathology by natural language processing (Supplemental Methods), 2) positive stool antigen test, 3) positive urea breath test, 4) prescription for one of 11 accepted eradication regimens for HP therapy as recommended by the American College of Gastroenterology, or 5) HP-associated International Classification of Diseases (ICD) 9/10 codes (ICD-9: 041.86; ICD-10: B96.81.^19–22

For patients with multiple criteria, the criterion with the earliest date was used. (Unique identifiers assured no duplications.) For those who had a prescription or administrative code as initial HP diagnosis without a diagnostic test that confirmed infection, we queried to identify if serum antibody was tested within 90 days of HP diagnosis.

Study outcome

The outcome for all analyses was GAC, identified using Veterans Affairs Central Cancer Registry and/or ICD 9/10 codes for non-proximal gastric adenocarcinoma (ICD-9: 151.1–151.9; ICD-10: C16.1–C16.9).²³ The Central Cancer Registry is a comprehensive, national database of cancers diagnosed and treated in the VHA since 1995. The diagnosis of GAC was minimum 30 days after HP diagnosis to ensure testing was for HP, not a malignancy workup. We filtered to include intestinal type non-cardia GACs, to avoid capturing non-adenocarcinomas and cardiac/gastroesophageal junction tumors, which are less clearly associated with HP.^24–26 Additional natural language processing was performed (Supplemental Methods).

Study exposures and statistical analysis

We performed a time to event analysis using competing risk models, with start time the date of HP diagnosis. Follow-up time ended at development of GAC, death prior to GAC, or end of follow-up. Death was considered a competing risk to GAC as patients could die from another cause, precluding the development of GAC, and factors associated with mortality might also be associated with GAC. Among the entire HP cohort, we evaluated covariates shown to be associated with GAC^{15, 27, 28}: age at HP diagnosis, gender, race, ethnicity, history of ever smoking (current or prior diagnostic code)²⁹, and zip code-level poverty at HP diagnosis. Zip code-level poverty was based on 2010 census data, categorized based on percentage of people within a zip code below the federal poverty line. Stata/IC 15.1 (College Station, TX) was used to perform backward selection, with inclusion of all clinically significant sub-hazard ratios (SHRs), where P<0.10.

After evaluating baseline factors associated with GAC, we sought to evaluate the association between treatment and eradication status on the incidence of GAC. Because one of our larger cohort inclusion criteria was receipt of eradication therapy, these analyses were restricted to those for whom the HP diagnosis was made using the gold standard of histology, stool antigen, and/or urea breath testing. Receipt of HP treatment was defined as receiving a recommended antibiotic regimen after HP diagnosis at the VHA using prescription filling data at any inpatient or outpatient VHA facility (Supplemental Methods Table 1). Because guidelines recommend that all patients treated for HP have confirmed eradication (i.e., cure) given the presence of resistant HP strains, we sought to evaluate whether HP treatment only decreased the risk of GAC among those with confirmed eradication (i.e., those who were successfully treated). Eradication was based on having either a negative stool antigen, urea breath test, and/or pathology (gastric biopsy on endoscopy) upon repeat testing. Failed eradication was defined as a positive stool antigen, urea breath test, and/or pathology, or a positive HP test after a prior negative test given that true re-infection is exceedingly rare. Patients without any eradication testing were considered as ‘unknown’ eradication status. HP status on pathology was determined by repeat natural language processing (Supplemental Methods). For this analysis, we excluded patients who had eradication testing via endoscopy within 90 days of eventual cancer diagnosis, as this was possibly performed for alarm symptoms, versus eradication testing alone. The Institutional Review Boards of the Corporal Michael J. Crescenz VA Medical Center and the University of Pennsylvania approved this study.

RESULTS

We identified 371,813 patients with HP: 26,873 with endoscopic pathology positive for HP, 11,262 with stool antigen, 400 with urea breath test, 266,216 with prescription of an eradication regimen, and 67,062 with administrative codes (Figure 1a). Of 266,216 with a prescription for an eradication regimen for HP, 7,854 (3.0%) had a positive serum antibody within 90 days. Of 67,062 with administrative code for HP, 57,508 (85.8%) had a positive serum antibody within 90 days.

Figure 1a–c:

Cohort of patients: a) with H pylori; b) who developed non-proximal gastric adenocarcinoma; and c) among those a positive diagnostic test (endoscopic pathology, stool antigen, urea breath test) who underwent treatment and eradication testing ICD = International Classification of Diseases (administrative codes)

Among 371,813 patients with HP, 2,024 (0.54%) developed GAC, with median follow-up 7.4 years (Figure 1b). As compared to patients with HP who did not develop cancer, patients with HP who developed cancer were older (median age at HP diagnosis: 65.1 vs 62 years, P<0.001), male (97% vs 92.3%, P<0.001), more likely to be African American (31.7% vs 23.8%, P<0.001), less likely to be Hispanic/Latino (74.5% vs 78.5%, P=0.002), and more likely to have ever smoked (30.7% vs 26.4%, P<0.001). Those who developed cancer were more likely to have died, 67.7% vs 37.2%, P<0.001. Median age at cancer diagnosis was 69 years.

Of the 38,535 with a positive diagnostic test (stool antigen, urea breath test, or pathology), 28,818 (74.8%) were prescribed an eradication regimen and 8,020 (20.8%) then underwent re-testing (Figure 1c). Among those who underwent eradication testing, the infection was successfully eradicated in 7,292 (90.9%).

Multivariable analysis of risk factors associated with GAC

Several factors were associated with incident diagnosis of GAC in multivariable competing-risk models (Table 2). The method of HP diagnosis was associated with GAC, with the lowest risk among those identified as having HP based on only an ICD diagnosis of HP with a positive serum antibody in the adjacent 90 days (SHR, 0.75; 95% CI, 0.62–0.90). For each 5-year increase in age at HP diagnosis, the SHR for future GAC increased 1.13 (95% CI, 1.11–1.15; P<0.001). Black or African American race (SHR, 2.00; 95% CI, 1.80–2.22) and Asian race (SHR, 2.52; 95% CI, 1.64–3.89) were associated with significantly increased risks of GAC, as was Hispanic ethnicity (SHR, 1.59; 95% CI, 1.34–1.87; P<0.001). Smoking history was associated with GAC (SHR, 1.38; 95% CI, 1.25–1.52; P<0.001), while zip code level poverty was not. Figures 2a–2d depict the cumulative incidence curves for each significant covariate, adjusted for other covariates. Sub-analysis of those with a true positive diagnostic test found 5,10, and 20-year incidence of future GAC was 0.40%, 0.55%, and 0.71%, respectively, comparable to the larger cohort.

Table 2.

Risk factors for development of gastric cancer using multivariable competing risk time to event model

	SHR (95% Cl)	P- value
Agea	1.13 (1.11–1.15)	P<0.001
Method of H pylori diagnosis
Positive diagnostic test	REFERENCE	0.006
RX, no serum Ab	0.95(0.82–1.10)
ICD, no serum Ab	0.89(0.65–1.23)
RX, with serum Ab	0.83(0.59–1.17)
ICD, with serum Ab	0.75 (0.62–0.90)
Ethnicity
Not Hispanic or Latino	REFERENCE
Hispanic or Latino	1.59(1.34–1.87)	P<0.001
Unknown	1.30 (1.08–1.55)
Race
White	REFERENCE
Black or African American	2.00 (1.80–2.22)	P<0.001
American Indian or Alaskan Native	1.36(0.84–2.19)
Asian	2.52(1.64–3.89)
Native Hawaiian or other Pacific Islander	0.82(0.48–1.42)
Unknown	1.08(0.91–1.29)
History of smoking	1.38(1.25–1.52)	P<0.001
Female gender	0.52 (0.40–0.68)	P<0.001
Poverty level of zip code where patient resided at H pylori diagnosis
< 10% residing below poverty level	REFERENCE
10 − 24.9% residing below poverty level	0.86 (0.77–0.97)	0.09
25 − 49.9% residing below poverty level	0.95(0.84–1.09)
≥50% residing below poverty level	0.99(0.76–1.29)
Unknown	0.83(0.65–1.06)

^a.Age is per 5-year incremental increase in year

RX = prescription therapy; ICD = International Classification of Diseases (administrative codes); Ab = antibody

Figure 2a–d:

Cumulative incidence curves for a) race; b) ethnicity; c) gender; d) method of H pylori diagnosis, amongst entire cohort, adjusted for other covariates RX = prescription therapy; ICD = International Classification of Diseases (administrative codes); Ab = antibody

Secondary analyses of treatment and eradication among those with confirmed HP

Among the 38,535 individuals with a positive diagnostic test within the VHA, receipt of HP treatment was not associated with subsequent development of GAC (SHR, 1.16; 95% CI, 0.74–1.83; P=0.51). Similar to models of the entire cohort, increasing age, tobacco use, and race/ethnicity were associated with GAC in this restricted cohort (Figure 3a–c and Supplemental Table 1).

Figures 3a–d:

plotted sub-hazard ratios for development of gastric cancer after positive diagnostic test (endoscopic pathology, stool antigen, urea breath test), considering a) treatment status and b) eradication status, and cumulative incidence curves for c) treatment of H pylori and d) eradication of H pylori, adjusted for other covariates HR = hazard ratio

Among patients with confirmed HP who had eradication testing, there was a significantly lower risk of GAC among those with confirmed HP eradication (SHR, 0.24; 95% CI, 0.15–0.41; P<0.001). Age, race/ethnicity, and tobacco history were also still associated with development of GAC (Figure 3b–d and Supplemental Table 2).

Sensitivity analyses

Sensitivity analyses were conducted excluding incident cancers within 6 months and 12 months, versus the 3-month exclusion data presented above. The incidence of GAC after HP diagnosis excluding cancers within 6 months at 5, 10, and 20 years was 0.19%, 0.33%, and 0.49%. The incidence of GAC after HP diagnosis excluding cancers within 12 months at 5, 10, and 20 years was 0.24%, 0.38% and 0.53%. The results of subsequent analyses did not vary results significantly, and are presented in Supplemental Tables 3–8.

DISCUSSION

In the largest study of gastric cancer among US patients with diagnosed HP, the incidence of GAC after HP diagnosis at 5, 10, and 20 years was 0.37%, 0.5%, and 0.65%. Non-modifiable risk factors included older age, Black/African American and Asian race, Hispanic ethnicity, and male gender, while modifiable risk factors included tobacco history and eradication of HP. These data provide the first population-level estimates of GAC risk in a US cohort with HP infection which are needed to identify at-risk patients and to inform large-scale gastric cancer screening protocols in the US population.

Most importantly, we demonstrate a stark difference in incidence after HP detection in US patients as compared to patients in countries with high rates of GAC. In Japan, which has among the highest rate of GAC in the world, and where HP eradication is a strong health focus, 2.9% of patients with HP had future GAC, five-fold greater than our findings.^{30, 31} The reasons for this are not completely clear. The widely touted “hygiene hypothesis” is often used to explain the decreased incidence of GAC in the US, where a decrease in HP due to improved sanitation paralleled the decrease in GAC, but does not explain why we have a significantly lower incidence of GAC among those with known HP.^{32, 33} There have been findings of HP-negative GAC in the US, with some hypothesizing that this is due to causes of negative testing versus a lack of HP as a requirement for future GAC, but countries with high rates of GAC also have this finding, so it would not account for the discrepancy either.^{24, 34–36} It is possible that different virulence factors are associated with HP in the US, or that there is some other, not yet elucidated, protective environmental factor in the US, similar to the “African enigma” noted above. These are avenues that should be explored by future studies.

Prior US studies using national level data have demonstrated an increased incidence of GAC among racial/ethnic minorities.^{17, 37–41} This was hypothesized to be due to a higher prevalence of HP in these populations, but our study offers important insights in that it was restricted to patients with HP, and therefore clearly demonstrates that a higher prevalence of HP in these minority populations alone does not explain their increased incidence of GAC.^{22, 42} This underscores the need to evaluate other potential genetic, environmental, or HP-specific factors that increase the incidence of HP in these patients. We also found that tobacco smoking was associated with an increased risk of GAC, which speaks to the importance for behavioral modification to mitigate the risks of GAC among smokers, as well as to identify a higher-risk cohort of patients. Male gender was also associated with an increased risk of GAC, consistent with previous trends seen in the US patients with GAC, and posited to be related to a protective effect of estrogen.^43–45 Lastly, older age was associated with future GAC, and is likely a proxy for duration of infection and associated inflammation.²⁸

Based on our data, modeling studies could be performed to evaluate the benefit of screening, when targeted towards those patients we can identify as high-risk. Our findings suggest that a larger screening or surveillance program is likely not warranted given the low incidence of GAC after detection of HP, since even those patients we can identify as high-risk have an absolute risk that is small for future GAC. And while some may argue that the absolute risk is so low that no patient should be considered high-risk, our findings have some parallel with esophageal adenocarcinoma after Barrett’s esophagus, where the risk is approximately 2.5% at 18 years among patients with established intestinal metaplasia (an arguably higher risk group than our cohort), but robust screening and surveillance protocols are recommended.^{46, 47} Moreover, the incidence of gastric cancer is higher than the incidence of esophageal cancer, underscoring the disparity in screening protocols for these two malignancies.⁴⁸

Importantly, patients who were apparently given a diagnosis of HP based on serum antibody positivity, which has been explicitly recommended against, had a lower risk of GAC, emphasizing the importance of testing with appropriate methods to establish a diagnosis and facilitate appropriate treatment.²² While we are not able to conclusively demonstrate that many patients are treated without a diagnostic test, as they may have undergone testing outside the VHA, there seems to be continued reliance on the serum antibody. Quality measures are important to ensure appropriate testing, and the widespread availability of the serum antibody could be reconsidered, given its lack of a role in diagnosing HP.

Among patients with positive diagnostic test, subsequent treatment was not associated with decreased incidence of future GAC, but failing eradication testing was associated with increased risk. This speaks to the ability of HP eradication to modify future risks of GAC, and the need to not only treat those diagnosed with HP, but to confirm eradication, and re-treat those who fail eradication. Treatment was prescribed in almost three-quarters of patients with a positive diagnostic test, and the majority (90.9%) of patients who were prescribed therapy and tested for eradication had their infections eradicated. Yet only one in five had post-treatment testing, an essential component of treatment. Though confirming eradication is guideline recommended, the rate of eradication testing is known to be subpar.²² Testing for eradication is especially important given the now alarming rates of HP antibiotic resistance, a well-known issue.^{22, 49–52} Two recent studies from Sweden and Hong Kong suggest that in comparison to the background population or matched controls, treatment decreases risk of future GAC, but other studies have presented limited and conflicting evidence for which populations would most benefit from treatment of HP.^{16, 21, 33, 53–59} Our study adds to these findings, and further suggests that inadequate eradication may explain disparate findings. Our findings that failed eradication is associated with future GAC also lend further credence to the importance of HP in GAC.

It is important to note, there have been proponents cautioning against the mass eradication of HP, given its role in microecology and antibiotic resistance, suggesting that eradication may fuel unintended diseases, including esophageal adenocarcinoma.^60–62 In our study, those confirmed to have their infections eradicated and those who had unknown eradication status both had a lower risk of GAC. Those patients with unknown eradication status may have been re-tested outside of the VHA, treated successfully with a course of antibiotics (though not re-tested), or have some other reason to be lower-risk for GAC than those persons who were tested for eradication, apart from HP. Future studies should continue to assess the global risk-benefit of eradication of HP infection.

There are several limitations to this study. First, its retrospective nature diminishes the ability to determine causality. Second, the cohort has some inherent selection bias (those tested for HP are being tested for some clinical reason), and we are unable to compare to it a confirmed background or control population without HP. The majority were also included based on prescription therapy or administrative code, however, we evaluated those with a true positive diagnostic test, and found incidence of future GAC was comparable to the larger cohort, as noted above. Furthermore, we adjusted for method of diagnosis in the model (Figure 2d). Third, despite its robust nature, the VHA is overwhelmingly male, and our cohort was > 90% male, limiting generalizability, though our findings are consistent with known trends in GAC. Fourth, we were unable to determine indication for HP testing, country of origin, or family history of GAC, potential confounders. Fifth, there are possibilities for false-negative/positive testing. Lastly, there are measurement issues, including patients receiving care outside the VHA (limiting HP diagnosis, testing, treatment ascertainment, lack of endoscopic histopathology, as well as oncologic diagnoses and cancer registry inputs) and limitations in longitudinal follow up. However, misclassification of HP status would only affect inclusion into the cohort, and misclassification of the outcome, inadvertently including a proximal GAC, would only bias towards the null, since risk factors for distal and proximal cancer differ in terms of age, race, and ethnicity.⁶³ Misclassification of treatment/eradication are also possible, with outside treatment and eradication testing under-capturing those who were treated and eradicated. Patients may have not obtained or taken their HP eradication regimen appropriately, which would explain the non-differential finding for GAC after treatment prescription. However, even with fill data, there would be no way to confirm that patients completed their treatment regimen. There may be misclassifications in our available data within the VHA, limiting those patients we are able to identify as diagnosed, treated, or cured, leading us to underestimate the number of patients diagnosed, treated, tested thereafter, and confirmed cured.

The strengths of our study are primarily related to the unique nature of the cohort, being the largest study of HP-related GACs in the US. National level data lacks the ability to identify preceding HP status, which we were able to do. We were able to define that HP status even further in granularity by identifying positive testing and those who were deemed positive by serology. With our diverse, nationwide population, and up to 20 years of follow up, this is the first study to identify risk factors and the importance of HP infection among US individuals. We demonstrate that in a diverse group of US patients with HP infection, future GAC incidence is low. Those at highest risk are males, Black or African Americans, Asians, and Hispanic and Latino patients. Smoking is an important modifiable risk factor for future GAC development. Treatment and eradication of infection are not uniformly performed at the VHA, and testing for HP continues to be performed by serum antibody, despite recommendations against this modality. Eradication of HP infection is associated with decreased future GAC. Future studies should focus on consideration of screening of high-risk persons, identification of what host and HP factors predispose to increased risk, quality measures to ensure appropriate diagnosis of HP infection, and resistance patterns and efficacy of treatment of HP infection.

Table 1.

Comparison of patients with H pylori who did and did not develop gastric cancer

	H pylori (n=369,789)	H pylori with future non-proximal gastric adenocarcinoma (n = 2,024)	P-value
Age in years at H pylori, median (IQR)	62.0 (52.7, 70.9)	65.1 (57.4, 73.3)	P<0.001
Male (%)	341,367(92.3%)	1,963 (97.0%)	P<0.001
Race (%)			P<0.001
White	214,090 (57.9%)	984 (48.6%)
Black or African American	87,989 (23.8%)	641 (31.7%)
American Indian or Alaskan Native	3,322 (0.9%)	17 (0.8%)
Asian	2,483 (0.7%)	21 (1.0%)
Native Hawaiian / Pacific Islander	3,837 (1.0%)	13 (0.6%)
Unknown / Declined to Answer	58,068 (15.7%)	348 (17.2%)
Ethnicity			P<0.001
Hispanic / Latino	290,286 (78.5%)	1,507 (74.5%)
Not Hispanic or Latino	33,306 (9.0%)	213 (10.5%)
Unknown / Declined to Answer	46,197 (12.5%)	304 (15.0%)
Ever smoker (%)	97,571 (26.4%)	621 (30.7%)	P<0.001
Method of H pylori diagnosis			0.009
Pathology	26,725 (7.2%)	148 (7.3%)
Stool Antigen	11,205 (3.0%)	57 (2.8%)
Urea Breath Test	399 (0.1%)	1 (0.05%)
RX, no serum Ab	256,883 (69.5%)	1,479 (73.1%)
ICD, no serum Ab	9,507 (2.6%)	47 (2.3%)
RX, with serum Ab	7,815 (2.1%)	39 (1.9%)
ICD, with serum Ab	57,255 (15.5%)	253 (12.5%)
Poverty level of zip code where patient resided at H pylori diagnosis			P<0.001
< 10% residing below poverty level	69,754 (18.9%)	386 (19.1%)
10 − 24.9% residing below poverty level	184,848 (50.0%)	912 (45.1%)
25 − 49.9% residing below poverty level	88,986 (24.1%)	572 (28.3%)
≥50% residing below poverty level Unknown	9,880 (2.7%) 16,321 (4.4%)	76 (3.8%) 78 (3.9%)
Follow up time in years, median (IQR)	7.4 (3.3, 12.3)	1.8 (0.4,5.8)	P<0.001
Deceased	137,575 (37.2%)	1,371 (67.7%)	P<0.001

IQR = inter-quartile range; RX = prescription therapy; ICD = International Classification of Diseases (administrative codes); Ab = antibody

WHAT YOU NEED TO KNOW.

BACKGROUND AND CONTEXT:

We collected data from a large cohort of patients in the United States to calculate the incidence of and risk factors for non-proximal gastric adenocarcinomas after detection of Helicobacter pylori.

NEW FINDINGS:

In a study of 371,813 veterans with a diagnosis of H pylori infection, we found significantly higher risks of gastric cancer in racial and ethnic minorities and smokers. Treatment of H pylori infection only decreased risk if eradication was successful.

LIMITATIONS:

This was a retrospective study of US veterans.

IMPACT:

Eradication of H pylori infection reduces the risk of gastric cancer. Studies are needed on the effects of screening high-risk persons and to identify quality measures for diagnosis, resistance patterns, and treatment efficacy.

Abbreviations:

CDW

Corporate Data Warehouse

GAC

gastric adenocarcinoma

HP

Helicobacter pylori

ICD

International Classification of Diseases

VHA

Veterans Health Administration