Acid suppressive drugs and gastric cancer: A meta-analysis of observational studies

Table 1 Characteristics of observational studies on acid suppressive drugs and gastric cancer

Ref.	Country	Design	Study period	Range of age (yr)	Population	Drugs	Adjusted OR or RR (95%CI)	Study quality1	Adjustment for covariates
La Vecchiaet al[23]	Italy	CC	1985-1989	27-74	563 cases and 1501 controls	Cimetidine, Ranitidine	1.80 (1.20 -2.70)	5	Age, sex, education, smoking, alcohol, coffee
Johnson et al[24]	United States	CC	1988-1992	NA	113 cases and 452 matched controls	Cimetidine, Ranitidine	2.00 (1.00 -3.90)	5	Age, sex, first pharmacy use
Møller et al[25]	Denmark	PCC	1977-1990	NA	134 cases among 16739	Cimetidine	2.30 (1.60-3.10)	6	Age, sex, diagnosis, method of diagnosis
Schumacher et al[26]	United States	CC	1981-1987	17-94	99 cases and 365 controls	Cimetidine	2.30 (0.80 -6.90)	6	Age, sex, date of first pharmacy use
Chow et al[27]	United States	MCC	1986-1992	NA	196 cases and 200 controls	Cimetidine, Ranitidine, Famotidine, Nizatidine	0.60 (0.30-1.30)	6	Race, smoking, BMI, number of composite conditions2
Farrow et al[52]	United States	PBCC	1993-1995	30-79	293 esophageal adenoca, 221 esophageal SCC, 261 gastric cardia adenoca, 368 non-cardia gastric adenoca and 695 controls	Cimetidine, Ranitidine, Famotidine, Nizatidine	1.00 (0.60-1.90)	7	Age, sex, study center, smoking, BMI, history of gastric or duodenal ulcer, GERD symptom frequency, alcohol
Suleiman et al[53]	United Kingdom	NCC	1990-1992	NA	231 cases among 9876 controls	Cimetidine, Ranitidine	2.56 (0.17-38.09)	7	Age, sex, MI, antacid, steroid, smoking, alcohol, social class, height, weight
García Rodríguez et al[54]	United Kingdom	NCC	1994-2001	40-84	287 esophageal adenoca, 195 gastric cardiac adenoca, 327 gastric non-cardia adenoca and 10000 controls	H2RAs, PPIs	1.24 (0.88-1.75)	8	Age, sex, calendar year, smoking, alcohol, BMI, UGI disorder, hiatal hernia, GU, DU, dyspepsia
Tamim et al[55]	Canada	NCC	1995-2003	NA	1589 cases and 12991 controls	H2RAs, PPIs	1.37 (1.22-1.53)	8	Age, sex number of prescriptions to any drug, total length of hospitalizations and number of visit to GPs, specialist, emergency rooms
Duan et al[56]	United States	PBCC	1992-1997	NA	220 esophageal adenoca, 277 gastric cardiac adenoca, 441 distal gastric adenoca and 1356 controls	H2RAs, PPIs	1.15 (0.58 -2.29)	7	Age, sex, race, birthplace, education, smoking, BMI, history of UGI symptom
Poulsen et al[57]	Denmark	PCC	1990-2003	NA	161 cases among users of 18790 PPIs or 17478 H2RAs	Omeprazole, Lanzoprazole, Esomeprazole, Pantoprazole, Rabeprazole, Cimetidine, Ranitidine, Nizatidine, Famotidine	1.30 (0.70-2.30)	8	Age, sex, calendar period, history of H.pylori, gastroscopy year, COPD, alcohol, NSAID

¹Study quality was judged based on the Newcastle-Ottawa Scale (range, 1-9 stars). The mean value for the methodological quality of 11 studies was 6.64 stars;

²Number of composite conditions were created to include gastroesophageal reflux, hiatal hernia, esophagitis/esophageal ulcer, or difficulty swallowing. OR: Odds ratio; RR: Relative risk; CC: Case-control; MCC: Medical record based case-control; PBCC: Population-based case-control study; NCC: Nested case-control study; PCC: Prospective cohort study; NA: Not available; Adenoca: Adenocarcinoma; SCC: Squamus cell carcinoma; H₂RAs: Histamine 2-receptor antagonists; PPIs: Proton pump inhibitors; BMI: Body mass index; GERD: Gastroesophageal reflux disease; MI: Myocardial infarction; UGI: Upper gastrointestinal; GU: Gastric ulcer; DU: Duodenal ulcer; GPs: General practitioners; COPD: Chronic obstructive pulmonary disease; NSAID: Nonsteroidal anti-inflammatory drug.

Table 2 Association between acid suppressive drugs use and gastric cancer risk in subgroup meta-analysis

Category	No. of studies	Adjusted OR/RR (95%CI)	Heterogeneity, I2 % (95%CI)	Model used	Heterogeneity between groups
Type of drugs					P = 0.01
H2RAs	10[23-27,52-55,57]	1.40 (1.24-1.59)	59.5 (19.0-80.0)1	Fixed-effects
PPIs	3[54,55,57]	1.39 (1.19-1.64)	0.0 (0.0-90.0)	Fixed-effects
Location of gastric cancer					P = 0.03
GE junction	2[24,26]	2.28 (0.97-5.35)	0.0 (NA)	Fixed-effects
Gastric cardia	4[52,54-56]	0.88 (0.63-1.24)	9.2 (0.0-86.0)	Fixed-effects
Non-cardia	6[24,26,52,54-56]	1.42 (1.12-1.79)	0.0 (0.0-75.0)	Fixed-effects
Duration of drugs use					P = 0.27
Within 5 yr	7[23,24,26,27,52,55,57]	1.58 (1.35-1.81)	60.2 (9.0-83.0)1	Fixed-effects
Over 5 yr	6[23,24,26,27,52,57]	1.24 (0.84-1.84)	25.4 (0.0-69.0)	Fixed-effects
Study quality					P = 0.01
High-quality2	6[52,53,54,55-57]	1.34 (1.21-1.48)	0.0 (0.0-75.0)	Fixed-effects
Low-quality	5[23-27]	1.86 (1.49-2.32)	63.5 (4.0-86.0)1	Fixed-effects

¹I² value greater than 50% represents substantial heterogeneity. We conducted sensitivity analyses by removing each study to determine the studies that contributed most to the heterogeneity;

²High quality study was considered a study awarded 7 or more stars, as standard criteria have not been established. OR: Odds ratio; RR: Relative risk; H₂RAs: Histamine 2-receptor antagonists; PPIs: Proton pump inhibitors; GE junction: Gastroesophageal junction; NA: Not available.