Bisphosphonate use and gastrointestinal tract cancer risk: Meta-analysis of observational studies

Table 1 Search strategy

Search strategy for MEDLINE	Search strategy for EMBASE	Search strategy for Cochrane reviews of the Cochrane Library
1. Diphosphonates [MH] OR diphosphonates [ALL] OR diphosphonate [ALL]	1. Bisphosphonates/de	1. Bisphosphonates [ALL]
2. Bisphosphonates [ALL]	2. Diphosphonate/de	2. Diphosphonate [ALL]
3. Alendronate [MH] OR alendronate [ALL]	3. Alendronate/de	3. Alendronate [ALL]
4. Clodronic acid [MH] OR clodronic acid [ALL] OR clodronate [ALL]	4. Clodronate/de	4. Clodronate [ALL]
5. Etidronic acid [MH] OR etidronic acid [ALL] OR etidronate [ALL]	5. Etidronate/de	5. Etidronate [ALL]
6. Ibandronic acid [NM] OR ibandronic acid [ALL] OR ibandronate [ALL]	6. Ibandronate/de	6. Ibandronate [ALL]
7. Minodronate [ALL] OR YM 529 [NM] OR YM 529 [ALL]	7. Minodronate/de	7. Minodronate [ALL]
8. Neridronate [ALL] OR 6-amino-1-hydroxyhexane-1,1-diphosphonate [NM] OR 6-amino-1-hydroxyhexane-1,1-diphosphonate [ALL]	8. Neridronate/de	8. Neridronate [ALL]
9. Olpadronic acid [NM] OR olpadronic acid [ALL] OR olpadronate [ALL]	9. Olpadronate/de	9. Olpadronate [ALL]
10. Pamidronate [NM] OR pamidronate [ALL]	10. Pamidronate/de	10. Pamidronate [ALL]
11. Risedronic acid [NM] OR risedronic acid [ALL] OR risedronate [ALL]	11. Risedronate/de	11. Risedronate [ALL]
12. Tiludronic acid [NM] OR tiludronic acid [ALL] OR tiludronate [ALL]	12. Tiludronate/de	12. Tiludronate [ALL]
13. Zoledronic acid [NM] OR zoledronic acid [ALL] OR zoledronate [ALL]	13. Zoledronate/de	13. Zoledronate [ALL]
14. 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13	14. 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13	14. 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13
15. Observational studies [ALL]	15. Observational studies/de
16. CohORt studies [ALL]	16. CohORt studies/de	[ALL] - in all text
17. Case control studies [ALL] OR case-control studies [ALL]	17. Case control studies/de OR case-control studies/de
18. Case referent studies [ALL] OR case-referent studies [ALL]	18. Case referent studies OR case-referent studies
19. 15 OR 16 OR 17 OR 18	19. 15 OR 16 OR 17 OR 18
20. 14 OR 19	20. 14 OR 19
21. Humans [MH]	21. [Humans]/lim
22. 20 AND 21	22. [Embase]/lim OR [embase classic]/lim
	23. 20 AND 21 AND 22
Date of search: April 29, 2011 (1973 April 2011); Result: 1709 articles found	Date of search: April 29, 2011 (1977 April 2011);Result: 2129 articles found	Limitation: cochrane database of systematic reviews;Date of search: April 29, 2011 (1971 April 2011); Result: 47 articles found

ALL: All fields; MH: MeSH terms; NM: Supplementary concept/substance name; PT: Publication type; /de: Mapped terms; /lim: Limitation.

Table 2 Characteristics of cohort studies (n = 3) included in meta-analysis regarding use of bisphosphonates and risk of cancer

Ref.	Country	Cohort formation (population)	Study period	Mean F/U duration, yr (range)	Age, yr (mean±SD)	Sex ratio (reference group)	Type of cancer	Type of drug (reference group)	Risk estimate	Adjustment	No. of cancer development/ no. of study population		No. of total cohort
											Bisphosphonate user	Reference group
Abrahamsen et al[10]	Denmark	Patients with fracture who had prescription for oral bisphosphonate (national registers)	E:1995-2005; F/U: -2005	2.8 (NA)	74.3 ± 8.8	M: 11; F: 89	Esophageal cancer, gastric cancer	Any bisphosphonate use (no bisphosphonate use)	Esophageal or gastric cancer 0.78 (0.49-1.26); esophageal cancer 0.35 (0.14-0.85); gastric cancer 1.23 (0.68-2.22)	Age, sex, fracture type, Charlson index, number of concomitant medication, time to event	NA/13 678	NA/27 356	41 034
Solomon et al[11]	United States	Patients who had prescription for oral bisphosphonate (health care utilization records of medicare beneficiaries)	NA	NA (NA)	NA	NA	Esophageal cancer	Any bisphosphonate use (other osteoporosis medications use)	Esophageal cancer 0.55 (0.06-4.72)	NA	Incidence rate 26.7/100 000	Incidence rate 48.4/100 000	NA
Cardwell et al[7]	United Kingdom	Patients receiving prescription for oral bisphosphonate (GPRD)	E:1996-2006; F/U: -2008	4.5 (0.5-12.9)	70.0 ± 11.4	M: 19; F: 81	Esophageal cancer, gastric cancer	Any bisphosphonate use (regardless of bisphosphonate use)	Gastric cancer 0.78 (0.50-1.23); esophageal cancer 1.07 (0.77-1.49)	Age, sex, general practice, BMI, cigarette smoking, alcohol intake, hormone therapy, NSAID use, Barrett's esophagus, GERD, H2 receptor antagonist use, proton pump inhibitor use	Esophageal or gastric cancer 116/41 826 (gastric cancer 37/41 826; esophageal cancer 79/41 826)	Esophageal or gastric cancer 115/41 826 (gastric cancer43/41 826; esophageal cancer 72/41 826)	83 652

E: Enrollment; F/U: Follow up; BMI: Body mass index; GERD: Gastroesophageal reflux disease; NSAID: Nonsteroidal antiinflammatory drugs; M: Male; F: Female; NA: Not available; GPRD: General Practice Research Database.

Table 3 Case-control studies (n = 3) included in meta-analysis regarding use of bisphosphonates and risk of cancer

Ref.	Country (study type)	Case selection method	Control sampling method	Medication data collection method (period)	Age, yr (mean±SD)	Sex ratio of cases (reference group)	Site of cancer	Type of drug (reference group)	Odd’s ratio (95% CI)	Adjustment	No. of cases/no. of controls
											Exposed	Unexposed
Green et al[8]	United Kingdom (nested case-control)	Review of computerized information (within participants of GPRD between 1995 -2005)	Matched on age (within 2 yr), sex, observation period in the database and general practice attended)	Review of computerized medical records (from 1995 until cancer diagnosis)	72 ± 11	M: 57; F: 43	Gastrointestinal cancer (esophageal, gastric, colorectal)	Any bisphosphonate use (no bisphosphonate use)	Esophageal cancer 1.30 (1.02-1.66); Gastric cancer 0.87 (0.64-1.19); colorectal cancer 0.87 (0.77-1.00)	Age, sex, observation period, general practice, BMI, cigarette smoking, alcohol intake	Esophageal cancer 90/345; gastric cancer 49/270; colorectal cancer 276/1555	Esophageal cancer 2864/14 376; gastric cancer 1969/9737; colorectal cancer 10 365/51 467
Nguyen et al[12]	United States (nested case-control)	Review of computerized information (within patients with Barrett’s esophagus in the national veterans affair database between 2000-2002)	Matched on age(interval of 5 yr) and Barrett’s esophagus index date	Review of computerized medical records (from Barrett’s esophagus diagnosis until 3 mo before cancer diagnosis)	65.0 ± 10.3	M: 2.6; F: 97.4	Esophageal cancer	Any bisphosphonate use, mostly alendronate (no bisphosphonate use)	0.81 (0.18-3.72)	Age, Barrett’s esophagus index date, race, non cancer disease comorbidity index, NSAID use, PPI use	2/13	114/683
Rennert et al[13]	Israel (conventional case-control)	Review of a computerized information (within postmenopausal women in CHS database between 2000-2006)	Matched on age, sex, residence, and ethnic group in CHS database	Review of CHS pharmacy records	71.1 ± NA	F: 100	Colorectal cancer	Any bisphosphonate use more than 1 yr (no bisphosphonate use)	0.41 (0.25-0.67)	BMI, family history of colorectal cancer, vegetable consumption, sports participation, use of low-dose aspirin, statins, vitamin D, postmenopausal hormones	53/100	880/833

E: Enrollment; F/U: Follow up; CI: Confidence interval; M: Male; F: Female; NA: Not available; BMI: Body mass index; NSAID: Nonsteroidal antiinflammatory drugs; PPI: Proton pump inhibitors; CHS: Clalit Health Service; GPRD: General Practice Research Database.

Table 4 Assessment of study quality

Study		Quality assessment items
Representativeness	Ascertainment of exposure: Secure record or structured interview	Demonstration that outcome of interest		Assessment of outcome: Independent blind assessment or record linkage was not present at start of study	Study controls for age, cigarette smoking, BMI status
Cohort studies
Steinbuch et al[9]	- (female)	+	+	+	- (randomized controlled trials)
Abrahamsen et al[10]	+	+	-	+	+ (age)
Solomon et al[11]	- (medicare beneficiaries)	+	-	+	- (not reported)
Cardwell et al[7]	+	+	+	+	+++ (age, cigarette smoking, BMI)
Case-control studies
Green et al[8]	+	+	-	+	+++ (age, cigarette smoking, BMI)
Nguyen et al[12]	- (Barrett’s esophagus)	+	+	+	+ (age)
Rennert et al[13]	- (female)	+	+	+	++ (age, BMI)

BMI: Body mass index.