Abstract
Background
Bisphosphonates are common medications for the treatment of osteoporosis in older populations. Several studies, including the Women’s Health Initiative (WHI), have found inverse associations of bisphosphonate use with risk of breast and endometrial cancer, but little is known about its association with other common malignancies. The objective of this study was to evaluate the association of bisphosphonate use on the incidence of lung cancer in the WHI.
Patients and methods
The association between oral bisphosphonate use and lung cancer risk was examined in 151 432 postmenopausal women enrolled into the WHI in 1993–1998. At baseline and during follow-up, participants completed an inventory of regularly used medications including bisphosphonates.
Results
After a mean follow-up of 13.3 years, 2511 women were diagnosed with incident lung cancer. There was no evidence of a difference in lung cancer incidence between oral bisphosphonate users and never users (adjusted hazard ratio = 0.91; 95% confidence intervals, 0.80–1.04; P = 0.16). However, an inverse association was observed among those who were never smokers (hazard ratio = 0.57, 95% confidence interval, 0.39–0.84; P < 0.01).
Conclusion
In this large prospective cohort of postmenopausal women, oral bisphosphonate use was associated with significantly lower lung cancer risk among never smokers, suggesting bisphosphonates may have a protective effect against lung cancer. Additional studies are needed to confirm our findings.
Keywords: lung cancer, bisphosphonates, postmenopausal women, epidemiology, Women’s Health Initiative
Key Message
We examined the effects of oral bisphosphonate use on the risk of lung cancer among postmenopausal women.
There were no significant associations between lung cancer incidence and oral bisphosphonate use.
Bisphosphonate use was associated with reduced lung cancer risk among never smokers.
Introduction
Lung cancer is the second most common cancer and the leading cause of cancer death among American women [1]. Although tobacco smoking is the leading cause, identifying other modifiable factors is important and may be useful in developing novel prevention strategies for this highly fatal disease.
Bisphosphonates are commonly prescribed medications for the prevention and treatment of osteoporosis in older populations, particularly among postmenopausal women [2]. Experimental studies have shown that bisphosphonates have a range of direct and indirect antitumor effects, including inhibition tumor angiogenesis and cellular proliferation, prevention tumor cell adhesion and invasion, and induction of tumor cell apoptosis [3]; however, it is still unclear whether they play a role in primary prevention of cancer. A number of recent studies, including the Women’s Health Initiative (WHI) have explored the relationships between bisphosphonate use and risk of breast [4–8], gastrointestinal tract [9–11], endometrial and ovarian cancer [12, 13]. A few studies, with inconsistent results have investigated the association of oral bisphosphonate use with lung cancer risk [14–17]. Nonsignificant inverse associations were observed in retrospective studies nested within the United Kingdom (UK) General Practice Research Database (GPRD) [hazard ratio (HR) = 0.85, 95% confidence interval (CI), 0.70–1.03 and odds ratio (OR) = 0.97, 95% CI, 0.88–1.08] [14, 15], while results from combined GPRD and practice-based QRsearch databases showed no association (OR = 1.04, 95% CI, 0.97–1.12) [15]. Results from retrospective studies using Taiwan health insurance system database were conflicting; one suggested no association (HR = 1.17, 95% CI, 0.05–1.43) [16], but another one found positive association (HR = 1.47, 95% CI, 1.00–2.17) [17]. However, these previous studies had very limited information on risk factors of lung cancer including smoking, and none of them explored whether the associations between bisphosphonate use and lung cancer vary by smoking status. Meanwhile, some but not all previous studies have observed associations between exposure to tobacco smoke or tobacco products and low bone mineral density [18–21] and some clinical studies reported that bisphosphonate treatment failure was associated with current smoking status among postmenopausal women with osteoporosis [22]. These findings suggest a potential interaction between bisphosphonate use and smoking. Therefore, we hypothesized that oral bisphosphonate use may have effects on lung carcinogenesis in older women, particularly among those who are never smokers.
To test these hypotheses, we evaluated the relationship between oral bisphosphonates use and lung cancer risk and its potential interaction with smoking status on lung cancer risk in the well-described WHI study, a large prospective cohort of postmenopausal women involving three randomized clinical trials (WHI-CTs) and a large observational study (WHI-OS).
Materials and methods
WHI overview
Details of the WHI study methods have been published elsewhere [23–25]. Briefly, WHI study enrolled 161 808 postmenopausal women in an observational study (WHI-OS) and three clinical trials (WHI-CTs) between 1993 and 1998 from 40 clinical centers across the United States. Eligible women for the WHI study were 50 to 79 years old at baseline with anticipated survival of at least 3 years, accessible for follow-up. Each WHI trial had its own set of eligibility criteria. WHI-OS participants were ineligible or unwilling to be included in a randomized trial. All participants provided informed consent, and the study protocol was approved by the Institutional Review Boards of all respective clinical centers.
Measurement of exposures
Standardized, self-administered questionnaires were completed by participants at the baseline visit, including queries regarding demographic factors, smoking history, medical history, alcohol intake, dietary intake, personal habits, and physical activity.
A medication inventory of all current, regularly used medications, including bisphosphonates, was taken at the baseline and at year 3 clinic visits for all WHI participants, and additionally at years 1 and 6 after random assignment for WHI-CT participants. Participants were instructed to bring all medication containers for those prescribed and over the counter medications taken at least twice per week during the previous 2 weeks. The medication information (type and frequency) was obtained directly from the medication containers and entered into the WHI database by trained research staff. Each medication was matched to the Medi-Span® (Indianapolis, IN) Master Drug Database to ascertain detailed ingredient information. For those reporting use of bisphosphonates, the type of the compound and the duration of use was recorded and validated by checking pill box labels. No information on dose was collected. A validity study was conducted among 223 WHI participants for whom we had bisphosphonate use information from both medication inventory and pharmacy records; the agreement between sources was high (current bisphosphonate use: sensitivity = 80%; duration of bisphosphonate use: κ = 0.95) [26].
Extensive information on smoking was collected at baseline in the WHI study; participants were asked whether they had ever smoked ≥100 cigarettes in their lifetime and whether they smoked currently to identify current and former smokers. A woman was defined as former smoker if she did not smoke currently but had ever smoked ≥100 cigarettes in the past. The number of cigarettes smoked per day and the number of years as a regular smoker were also queried. Never smokers were defined as participants who reported having ever smoked <100 cigarettes in their lifetime. The 10-year probability of hip fracture was calculated using a WHO developed fracture Risk Assessment Tool (FRAX)® to evaluate a hip fracture risk [27]. It is based on individual patient models that integrate the risk associated with clinical risk factors measured at baseline [27].
Ascertainment of outcomes
Self-reported health outcomes, including lung cancer, were ascertained semiannually for WHI-CT or annually for WHI-OS using a health update questionnaire until 2005 and then annually for all WHI participants thereafter. Lung cancer diagnoses that were reported were confirmed and adjudicated using medical records and pathology reports reviewed by centrally trained study physicians blinded to study components, medication, or randomization assignments [24]. Tumor characteristics were coded according to the Surveillance, Epidemiology, and End Results (SEER) guidelines at the WHI Coordinating Center [24]. Follow-up was censored at the earliest of the following event: last known follow-up or 30 September 2013.
Study participants
We excluded participants who had not turned 50 years old at baseline, had a prior history of lung cancer at baseline, or had missing data on follow-up time, or lacked information of medication inventory at baseline and at year 3 for all participants and additionally at year 1 and 6 for WHI-CT participants. The current analyses finally included 151 432 WHI participants.
Statistical analysis
We estimated HRs and 95% CIs using Cox proportional hazards regression models [28] to examine the associations between time since initiation of oral bisphosphonate use and lung cancer incidence. Oral bisphosphonate use was, in general, dichotomized as a time-varying never/ever variable by updating baseline use at years 1, 3 or 6 clinic visits among women at risk of lung cancer in the CT women and year 3 in the OS women (nonusers could become ever users over time). Therefore, nonusers in the Cox regression model referred to women who never reported oral bisphosphonate use during those follow-up visits before the occurrence of lung cancer. Quartile distributions among oral bisphosphonate users were used to categorize total duration of use. In the age-adjusted analyses, the Cox proportional hazards models were adjusted for age and stratified on WHI study component and for those in the WHI-CT’s randomization arm. In more general multivariate analyses, the Cox regression models were adjusted for covariates measured at baseline, including age, ethnicity, education, smoking status, number of cigarettes smoked per day, total years of smoking, alcohol use status, physical activity, body mass index (BMI), personal prior hormone therapy use, statin use, total calcium intake, and total vitamin D intake. Previous studies have shown significant correlation between calculated hip fracture risk score with BMD [8]. Therefore, in current study, we also assessed potential confounding by 10-year probability of hip fracture and history of bone fracture after age 54 years, and these two factors did not alter point estimates by ≥10% and were excluded from final models.
Lung tumor histology, stage, and grade were compared between users of any type of oral bisphosphonate and nonusers with χ2 tests. Tumors with missing information were not included in these analyses.
Multiplicative interactions between oral bisphosphonate use and other covariates of interest measured at baseline, including smoking status and continuous total calcium intake, were tested using the Wald test. All statistical tests were based on two-sided probability. Statistical analyses were conducting using SAS, Version 9.4 (SAS Institute, Cary, NC).
Results
Oral bisphosphonate use at baseline was uncommon (n = 3146 users, 2.1%) but had increased to 9.7% (14 721 users) by year 6 (Table 1). About 72.3% of baseline users of bisphosphonate reported a physician diagnosis of osteoporosis compared with only 6.6% of nonusers. Compared to nonusers, baseline users were older, more likely to be white, had higher educational level, and substantially higher estimated 10-year probability of hip fracture. Baseline users also reported higher physical activity, had lower BMI, higher total calcium and vitamin D intakes, were less likely to be current smokers, have less total pack-years of smoking and were less likely to use hormone therapy before enrollment than nonusers (Table 1). Moreover, baseline characteristics of WHI study participants by their smoking status are presented in Table A1.
Table 1.
Baseline characteristics by oral bisphosphonate use in the WHI study, 1993–2013
| Characteristic at baseline | Oral bisphosphonate use |
|||
|---|---|---|---|---|
| At baseline |
At baseline, year 1, 3 or 6a |
|||
| No | Yes | No | Yes | |
| (n = 148 286) | (n = 3146) | (n = 136 413) | (n = 14 721) | |
| Age at screening (years) (%) | ||||
| <50–59 | 48 681 (32.8) | 406 (12.9)b | 46 242 (33.9) | 2799 (19.0) |
| 60–69 | 66 993 (45.2) | 1462 (46.5) | 60 970 (44.7) | 7326 (49.8) |
| 70–79 | 32 612 (22.0) | 1278 (40.6) | 29 201 (21.4) | 4596 (31.2) |
| Race-ethnicity (%) | ||||
| Non-Hispanic white | 123 276 (83.3) | 2837 (90.3)b | 112 666 (82.8) | 13 186 (89.8) |
| Black/African American | 13 380 (9.1) | 44 (1.4) | 13 026 (9.6) | 376 (2.6) |
| Hispanic/Latino | 5399 (3.7) | 86 (2.7) | 5115 (3.8) | 365 (2.5) |
| American Indian | 632 (0.4) | 1 (0.03) | 607 (0.4) | 26 (0.2) |
| Asian/Pacific islander | 3576 (2.4) | 140 (4.5) | 3126 (2.3) | 585 (4.0) |
| Other/Unknown | 2023 (1.1) | 38 (1.1) | 1873 (1.1) | 183 (1.0) |
| Education (%) | ||||
| High school or less | 33 169 (22.5) | 606 (19.4)b | 30 860 (22.8) | 2843 (19.5) |
| School after high school | 56 154 (38.2) | 1029 (33.0) | 51 963 (38.4) | 5086 (34.8) |
| College degree or higher | 57 880 (39.3) | 1481 (47.5) | 52 606 (38.8) | 6667 (45.7) |
| Missing | 1083 | 30 | 984 | 125 |
| Smoking status (%) | ||||
| Never | 74 302 (50.8) | 1660 (53.6)b | 68 121 (50.6) | 7789 (53.6) |
| Former smoker | 62 063 (42.4) | 1306 (42.1) | 57 202 (42.5) | 6000 (41.3) |
| Current smoker | 10 012 (6.8) | 134 (4.3) | 9330 (6.9) | 742 (5.1) |
| Missing | 1909 | 46 | 1760 | 190 |
| Pack-years of smoking (%) | ||||
| Never | 74 302 (52.0) | 1660 (55.0)b | 68 121 (51.8) | 7789 (54.9) |
| <5 | 20 738 (14.5) | 391 (13.0) | 19 161 (14.6) | 1949 (13.7) |
| 5–20 | 20 662 (14.4) | 413 (13.7) | 19 060 (14.5) | 1983 (14.0) |
| ≥20 | 27 287 (19.1) | 555 (18.4) | 25 186 (19.1) | 2471 (17.4) |
| Missing | 5297 | 127 | 4885 | 529 |
| Physical activity (MET/week) (%) | ||||
| 0–3.00 | 40 623 (28.7) | 692 (22.2)b | 37 712 (29.0) | 3502 (24.5) |
| >3.00–<11.75 | 44 801 (31.7) | 940 (30.1) | 41 126 (31.7) | 4529 (31.6) |
| ≥11.75 | 55 923 (39.6) | 1488 (47.7) | 51 036 (39.3) | 6284 (43.9) |
| Missing | 6939 | 26 | 6539 | 406 |
| Alcohol intake (%) | ||||
| Never | 15 843 (10.8) | 397 (12.7)b | 14 667 (10.8) | 1558 (10.7) |
| Past drinker | 27 776 (18.9) | 541 (17.3) | 25 892 (19.1) | 2366 (16.2) |
| <1 drink per month | 18 331 (12.4) | 287 (9.2) | 16 963 (12.5) | 1612 (11.0) |
| <1 drink per week | 30 217 (20.5) | 675 (21.6) | 27 736 (20.5) | 3102 (21.2) |
| 1–<7 drinks per week | 37 818 (25.7) | 827 (26.5) | 34 444 (25.4) | 4121 (28.2) |
| ≥7 drinks per week | 17 224 (11.7) | 393 (12.6) | 15 714 (11.6) | 1859 (12.7) |
| Missing | 1077 | 26 | 997 | 103 |
| BMI, kg/m2 (%) | ||||
| <25 | 50 435 (34.3) | 1714 (54.9)b | 44 563 (33.0) | 7480 (51.2) |
| 25–30 | 51 162 (34.8) | 1000 (32.0) | 47 212 (34.9) | 4843 (33.2) |
| ≥30 | 45 398 (30.9) | 410 (13.1) | 43 456 (32.1) | 2271 (15.6) |
| Missing | 1291 | 22 | 1182 | 127 |
| Hormone-therapy use (%) | ||||
| Never used | 61 746 (41.6) | 1472 (46.8)b | 55 976 (41.0) | 7099 (48.2) |
| <5 years | 32 788 (22.1) | 702 (22.3) | 30 241 (22.2) | 3195 (21.7) |
| 5 to <10 years | 19 560 (13.2) | 341 (10.8) | 18 214 (13.4) | 1653 (11.2) |
| ≥10 years | 34 187 (23.1) | 631 (20.1) | 31 977 (23.4) | 2774 (18.8) |
| Missing | 2 | 0 | 5 | 0 |
| Statin use (%) | 11 567 (7.8) | 375 (11.9)b | 10 608 (7.8) | 1312 (8.9) |
| History of osteoporosis (%) | 9602 (6.6) | 2251 (72.3)b | 7975 (5.9) | 3849 (26.5) |
| 10-year probability of hip fracture (%) | ||||
| <0.53 | 48 881 (31.6) | 270 (8.6)b | 45 022 (33.0) | 2089 (14.2) |
| 0.53–1.68 | 49 782 (33.6) | 752 (23.9) | 45 867 (33.6) | 4557 (31.0) |
| ≥1.69 | 51 623 (34.8) | 2124 (67.5) | 45 524 (33.4) | 8075 (54.8) |
| Total calcium intake (supplements and diet) (mean ± SD, mg/day) | 1171.8 ± 741.0 | 1585.8 ± 867.5b | 1162.7 ± 741.4 | 1345.2 ± 771.1 |
| Total vitamin D intake (supplements and diet) (mean ± SD, U/day) | 370.7 ± 278.4 | 500.2 ± 314.2b | 367.8 ± 277.6 | 424.8 ± 294.9 |
| WHI study component (%) | ||||
| OS | 84 511 (57.0) | 2312 (73.5) | 79 428 (58.2) | 7285 (49.5) |
| Hormone CT | 25 071 (16.9) | 298 (9.5) | 22 372 (16.4) | 2898 (19.7) |
| Dietary modification CT | 46 076 (31.1) | 585 (18.6) | 41 372 (30.3) | 5181 (35.2) |
| Calcium/vitamin D CT | 34 250 (23.1) | 354 (11.3) | 30 736 (22.5) | 3778 (25.7) |
| Enrollment in WHI hormone CT (%) | ||||
| No | 123 215 (83.1) | 2848 (90.5) | 114 041 (83.6) | 11 823 (80.3) |
| E-alone placebo group | 5067 (3.4) | 53 (1.7) | 4584 (3.4) | 523 (3.6) |
| E-alone active group | 4960 (3.3) | 45 (1.4) | 4562 (3.3) | 418 (2.8) |
| E+P placebo group | 7361 (5.0) | 103 (3.3) | 6317 (4.6) | 1124 (7.6) |
| E+P active group | 7683 (5.2) | 95 (3.0) | 6909 (5.1) | 833 (5.7) |
| Bisphosphonate type (%) | ||||
| Alendronate sodium | – | 2913 (92.6) | – | 13 467 (91.5) |
| Risedronate sodium | – | 13 (0.4) | – | 898 (6.1) |
| Othersc | – | 220 (7.0) | – | 356 (2.4) |
| Duration of bisphosphonate use (mean ± SD, years) | 1.2 ± 1.7 | 2.5 ± 3.0 | ||
Including participants remaining at risk of lung cancer at year 6.
χ2 tests for categorical variables and Student’s t-tests for continuous variables were used to compare oral bisphosphonate users with nonusers. P value < 0.01.
Others include etidronate disodium, tiludronate disodium, pamidronate disodium and zoledronic acid.
WHI, Women’s Health Initiative; CT, clinical trials; OS, observational study; E, estrogen; E + P, estrogen plus progestin; SD, standard deviation.
After 13.3 years’ (mean) cumulative follow-up, 2511 women were diagnosed with incident lung cancer (2257 nonusers and 254 ever users of oral bisphosphonates) (Table 2). The crude incidence of lung cancer was similar among nonusers and women reporting any type of oral bisphosphonates use. Based on the multivariable-adjusted model, the HR for use of any type of oral bisphosphonates compared with never use was not significantly different (HR = 0.91, 95% CI, 0.80–1.04, P = 0.16). The magnitude of the associations was stronger for long-term users of bisphosphonate than for short-term users, particularly for 1.50–2.99 years use (HR = 0.65, 95% CI, 0.46–0.93; P = 0.02); however, there was no dose–response relationship for duration of use (P for trend = 0.25). Alendronate was the most common type of oral bisphosphonate in the study population during the follow-up (>90% of users); however, risk estimates for risedronate and other types of bisphosphonate use were not very precise. Additionally, we conducted sensitivity analyses by further excluding lung cancer cases diagnosed within one year of enrollment of the WHI, and similar results were observed (data not shown).
Table 2.
Relative risk of lung cancer comparing oral bisphosphonate user to nonusers, Women’s Health Initiative (1993–2013)
| Oral bisphosphonate usea | No. of lung cancer | Person years | Incidence (% per year) | HR (95% CI)b | P | HR (95% CI)c | P |
|---|---|---|---|---|---|---|---|
| Any use | |||||||
| Never | 2257 | 1 811 696 | 0.125 | Ref. | Ref. | ||
| Ever | 254 | 200 060 | 0.127 | 0.89 (0.78–1.02) | 0.08 | 0.91 (0.80–1.04) | 0.16 |
| Duration of use, yearsd | |||||||
| Never | 2257 | 1 811 696 | 0.125 | Ref. | Ref. | ||
| <0.67 | 65 | 51 991 | 0.125 | 0.90 (0.70–1.15) | 0.39 | 0.91 (0.78–1.17) | 0.45 |
| 0.67–1.49 | 76 | 48 844 | 0.156 | 1.11 (0.88–1.39) | 0.39 | 1.18 (0.94–1.49) | 0.15 |
| 1.50–2.99 | 32 | 35 918 | 0.089 | 0.63 (0.45–0.90) | 0.01 | 0.65 (0.46–0.93) | 0.02 |
| ≥3.00 | 81 | 63 308 | 0.128 | 0.87 (0.70–1.09) | 0.22 | 0.86 (0.69–1.08)e | 0.19 |
| Type | |||||||
| Never | 2257 | 1 811 696 | 0.125 | Ref. | Ref. | ||
| Alendronate sodium | 227 | 183 344 | 0.124 | 0.87 (0.76–1.00) | 0.05 | 0.89 (0.77–1.02) | 0.09 |
| Risedronate sodium | 13 | 12 236 | 0.106 | 0.73 (0.42–1.26) | 0.26 | 0.78 (0.45–1.35) | 0.37 |
| Othersf | 14 | 4480 | 0.313 | 2.16 (1.28–3.66) | 0.01 | 2.02 (1.19–3.42) | 0.01 |
Users reported at least 2 weeks of use; nonusers included never users and those who used for <2 weeks. Baseline oral bisphosphonate use was updated at years 1, 3 and 6 for women in the WHI-CT and at year 3 for women in the WHI-OS.
Adjusted for age and WHI study component.
Adjusted for baseline age, ethnicity, education, smoking status, number of cigarettes per day, duration of regular smoking in years, alcohol use status, BMI, physical activity, total calcium intake, total vitamin D intake, statins use, and hormone treatment status and stratified on WHI study component.
Quartile cutoff points for total duration of oral bisphosphonate use among ever user (years): 0.67, 1.5 and 3.0.
Likelihood ratio tests were used to test difference between different strata of duration of bisphosphonate use. P value < 0.05.
Others include etidronate disodium, tiludronate disodium, pamidronate disodium, and zoledronic acid.
Two subgroup analyses were carried out (Table 3). We found that the association patterns with bisphosphonate use differed in never smokers and ever smokers (P interaction = 0.02). Among never smoking women, oral bisphosphonate use was inversely associated with lung cancer risk in the multivariate adjusted models (HR = 0.57, 95% CI, 0.39–0.84, P < 0.01); the association was stronger for total duration of use ≥1.5 years (HR = 0.36, 95% CI, 0.18–0.73). Among ever smokers, bisphosphonate use was not related to lung cancer risk. The reduced risk associated with oral bisphosphonate use was stronger among women with lower baseline total calcium intake, but the multiplicative interaction test was not statistically significant (P interaction = 0.13).
Table 3.
Association between oral bisphosphonate use and lung cancer by selected baseline characteristic factors (1993–2013)
| Parameter | Oral bisphosphonate use |
HR (95% CI) | P for interaction | |||
|---|---|---|---|---|---|---|
| Never |
Evera |
|||||
| Cases | Person years | Cases | Person years | |||
| Smoking | ||||||
| Never | 366 | 913 445.0 | 29 | 106 658.0 | 0.57 (0.39–0.84)b* | 0.02 |
| Ever | 1860 | 877 056.6 | 220 | 90 969.5 | 0.96 (0.83–1.11)c | |
| Former smokers | 1233 | 763 562.6 | 162 | 81 502.8 | 1.05 (0.89–1.24)c | |
| Current smokers | 627 | 113 494.0 | 58 | 9466.7 | 0.85 (0.64–1.12)c | |
| Total calcium intake (mg/day) | 0.13 | |||||
| <800 | 887 | 626 538.2 | 68 | 50 092.0 | 0.73 (0.57–0.95)d‡ | |
| 800–1200 | 513 | 446 347.3 | 54 | 45 243.3 | 0.97 (0.72–1.29)d | |
| ≥1200 | 857 | 738 810.2 | 132 | 102 724.9 | 1.00 (0.83–1.20)d | |
Users reported at least 2 weeks of use; nonusers included never users and those who used for <2 weeks. Baseline oral bisphosphonate use was updated at years 1, 3 and 6 for women in the WHI-CT and at year 3 for women in the WHI-OS.
P ≤0.01;
P = 0.02.
Adjusted for baseline age, ethnicity, education, alcohol use status, BMI, physical activity, total calcium intake, total vitamin D intake, statins use, and hormone treatment status and stratified on WHI study component.
Additionally adjusted number of cigarettes per day, duration of regular smoking in years among smokers.
Adjusted for baseline age, ethnicity, education, smoking status, number of cigarettes per day, duration of regular smoking in years, alcohol use status, BMI, physical activity, total vitamin D intake, and hormone treatment status and stratified on WHI study component.
The incidences of small cell lung cancer and nonsmall-cell lung cancer (NSCLC) were similar in bisphosphonate users. However, among never smoking women, oral bisphosphonate use was inversely associated with NSCLC risk (HR = 0.62, 95% CI, 0.41–0.92; P = 0.02). Similarly, a nonsignificant inverse association with adenocarcinoma was found among never smokers (HR = 0.70, 95% CI, 0.46–1.07; P = 0.10). No association of bisphosphonate use with NSCLC was found in ever smokers (data not shown). There were no apparent differences in the stage or grading of lung cancer in oral bisphosphonate users (Table 4).
Table 4.
Lung cancer tumor characteristics by oral bisphosphonate use (1993–2013)
| Characteristics | Oral bisphosphonate use |
P valueb | HR (95% CI)c | |
|---|---|---|---|---|
| Never (%) | Ever (%)a | |||
| Histological typed | ||||
| Small cell | 214 (10.4) | 26 (11.1) | 0.49 | 1.10 (0.72–1.67) |
| Nonsmall-cell | 1844 (89.6) | 209 (88.9) | 0.90 (0.78–1.04) | |
| Adenocarcinoma | 1112 (49.5) | 140 (55.1) | 0.97 (0.81–1.17) | |
| Squamous cell carcinoma | 315 (14.0) | 31 (12.2) | 0.88 (0.61–1.29) | |
| SEER stage | ||||
| Localized | 572 (25.5) | 80 (31.5) | 0.22 | 1.16 (0.91–1.48) |
| Regional | 550 (24.5) | 63 (24.8) | 0.93 (0.71–1.22) | |
| Distant | 839 (37.4) | 87 (34.2) | 0.83 (0.66–1.04) | |
| Unknown | 284 (12.6) | 24 (9.5) | – | |
| Grading | ||||
| Well differentiated | 221 (9.8) | 31 (12.2) | 0.44 | 1.20 (0.82–1.77) |
| Moderately differentiated | 409 (18.2) | 53 (20.9) | 1.06 (0.79–1.42) | |
| Poorly differentiated | 459 (20.4) | 54 (21.3) | 1.00 (0.75–1.34) | |
| Anaplastic | 108 (4.8) | 11 (4.3) | 0.84 (0.45–1.58) | |
| Unknown | 1050 (46.7) | 105 (41.3) | – | |
Users reported at least 2 weeks of use; nonusers included never users and those who used for <2 weeks. Baseline oral bisphosphonate use was updated at years 1, 3 and 6 for women in the WHI-CT and at year 3 for women in the WHI-OS.
χ2 tests are for categorical variables comparing oral bisphosphonate users with nonusers.
Adjusted for baseline age, ethnicity, education, smoking status, number of cigarettes per day, duration of regular smoking in years, alcohol use status, BMI, physical activity, total calcium intake, total vitamin D intake, statins use, and hormone treatment status and stratified on WHI study component.
Histologic subtypes were classified according to International Classification of Disease for Oncology and WHO Classification of Tumors for tumors of the lung. Nonsmall-cell lung cancer includes squamous cell carcinoma, adenocarcinoma, large cell carcinoma, sarcomatoid carcinoma and pleomorphic carcinoma.
Discussion
We did not find statistically significant association between ever use oral bisphosphonate and lung cancer risk in this large prospective cohort study of postmenopausal women. However, an inverse association of oral bisphosphonate use with lung cancer risk was found among never smokers. Furthermore, our results suggested associations between longer duration of use and reduced lung cancer risk.
Our results are consistent with those from Cardwell et al. [14], where a nonsignificant inverse association was observed (HR = 0.85, 95% CI, 0.70–1.03). However, there are some differences between their analyses and ours. WHI had larger sample size (2511 lung cancer cases among 151 432 postmenopausal women at risk), confirmed bisphosphonate use, type and duration through assessments of medication packages, and collected detailed information on risk factors for lung cancer and other potential confounding factors. Cardwell et al. [14] determined bisphosphonate use from prescription or medical claim records, which limited potential recall bias. However, their pharmacy databases had limited ability to adequately control for confounding such as smoking habits. Moreover, their analysis may have failed to estimate an association among women alone due to limited sample size.
Usage of oral alendronate, a nitrogen-binding bisphosphonate and subsequent lung cancer risk has been evaluated among osteoporosis patients based on a population-based health insurance system database of Taiwan [16, 17]. Chiang et al. [16] reported no association between oral alendronate and risk of lung cancer among postmenopausal women with osteoporosis. In the same database, a later report included both male and female osteoporosis patients, and found a positive association with risk of lung cancer among osteoporosis patients in Taiwan, and the association was stronger among patients with higher dosage of alendronate (≥1.0 g/year) [17]. The inconsistent results from the same database may be partially due to sample differences in the two analyses. The studies from Taiwan only adjusted for age and sex due to little information on possible confounding factors in their health insurance databases.
In our study, a reduction in lung cancer risk was seen in women using bisphosphonate for ≥1.5 year, particularly for use of 1.50–2.99 years. This finding is consistent with previous reports from UK, which reported a nonsignificant inverse association after 1 year of use (HR = 0.86, 95% CI, 0.63–1.17) [14]. Although no information was collected on dose of the medication in the WHI Study prevented us to explore the dose–response relationship for cumulative prescribed intakes, the likelihood ratio test suggested different effects of long-term use. Consistent with our hypothesis, we firstly found that reduction in lung cancer risk with bisphosphonate use tended to be limited to never smoking women, those who have an overall lower risk of developing lung cancer in this study. However, relative small number of incident lung cancer cases among never smokers limited our ability to further explore precise dose–response relationship for duration of use. Further studies are needed to evaluate different patterns of bisphosphonate use on lung cancer risk and to better understand the underlying mechanisms of bisphosphonate use on lung carcinogenesis among never smokers. Also, we found an inverse association between oral bisphosphonate use and lung cancer risk among postmenopausal women with low baseline calcium intake. Previous studies suggested that calcium intake interferes with bisphosphonate absorption and can interplay with bisphosphonate to impact efficacy of osteoporosis treatment [29]. It is possible that bisphosphonate may interact with homeostasis of calcium to influence proliferation of lung cancer cells. Further studies are needed to elucidate the potential interactions between bisphosphonate and calcium on lung carcinogenesis, and to better understand the possible molecular mechanisms.
Preclinical and substantial emerging clinical evidence have indicated that bisphosphonates have antitumor activities against a broad range of cancers including breast and prostate cancers [30, 31]. Preclinical studies have shown that bisphosphonates can induce apoptosis, and inhibit cell invasion and tumor angiogenesis [30, 32, 33] through their activation of gamma delta T cells, induction cytokines production, and contribution to immune response to tumor cells [31, 34, 35]. In vitro studies showed that zoledronic acid, a third generation bisphosphonate, inhibits the prenylation of small G-proteins such as Ras, reduces the signals and prevents the growth of lung cancer cells [36, 37]. In vivo studies also revealed that bisphosphonates show the antitumor and antiangiogenic effects through their ability to inhibit tumor cell growth and vascular endothelial growth factor release in NSCLC cell lines [38, 39]. Furthermore, bisphosphonates were shown to inactive human epidermal growth factor receptors (EGFRs) in both NSCLC cell lines and in mice with NSCLC [36, 40, 41]. The EGFR pathway is particularly relevant to progression of lung adenocarcinoma [42]. It is possible that bisphosphonates may have a protective effect on NSCLC, especially adenocarcinoma, through inactivation of EGFR pathways. In our study, bisphosphonate use was found to be associated with reduced risk of NSCLC in never smokers. More studies with larger sample size of NSCLC cases, particularly adenocarcinoma, are needed to help confirm this association.
The strengths of the study include prospective cohort design, the large study population of postmenopausal women, comprehensive measurements of lung cancer risk factors, detailed information on oral bisphosphonate use, the standardized central adjudication of lung cancers, and a relatively large number of incident lung cancer cases allowing subgroup analyses by smoking status and total calcium intake. Nevertheless, several issues regarding the current study should be noted. First, this analytic study design is an observational analysis. Lung cancer was not a predefined study outcome in the WHI. Since the protocol did not require participants to undergo chest radiology imaging tests, some cancers may have been missed. Therefore, nondifferential misclassification of the outcome is still possible. In addition, and perhaps more importantly, there were substantial differences in the characteristics including smoking status between oral bisphosphonate users and nonusers. Although we adjusted for many possible factors that could confound the association, we were unable to eliminate possible residual confounding. Secondhand smoke (SHS) exposure at home and at work since age 18 years was collected only among the WHI-OS participants at the baseline, so we were unable to eliminate potential confounding from SHS exposure. When limiting our analyses in the WHI-OS and additionally adjusting of SHS exposure in the multivariable-adjusted models, we found similar but attenuated results. Subgroup analyses limited to WHI-OS participants who were never smokers generated similar findings. WHI started enrollment before the time bisphosphonates began being widely prescribed; only women enrolled toward the end of the recruitment had the opportunity to use bisphosphonate at baseline. Therefore, the usage was low at baseline (2%). Although the prevalence of bisphosphonate use increased during follow-up, half of these women used bisphosphonates for less than 1.5 years and majority of users took alendronate. No information on dose of medications prevented us to further evaluate the dose–response relationship between oral bisphosphonate use and lung cancer risk. The relative small number of lung cancer cases still limited our ability to explore the duration–response relationship, associations for other bisphosphonate types, associations in different ethnicity groups, or heterogeneity in associations according to histological type of lung cancer. Finally, as the WHI study enrolled volunteer postmenopausal women who may have led a healthier lifestyle than women who were nonvolunteers and over 80% of participants were white, we may have limited ability to generalize our findings to the general US population of postmenopausal women [43].
In this large population of postmenopausal women, oral bisphosphonate use was associated with lower lung cancer risk among never smokers. These observational study findings need to be confirmed. As the U.S. Food and Drug Administration issued safety announcements related to potential risk of long-term bisphosphonates use [44, 45], further studies are warranted to investigate how duration of bisphosphonate use may influence risk of lung cancer and evaluate optimal dose of oral bisphosphonates for lung cancer prevention in older women. Further research to explore the mechanisms by which risk is reduced is warranted.
Acknowledgements
We acknowledge the dedicated efforts of investigators and staffs at the WHI clinical centers, the WHI Clinical Coordinating Center, and the Nation Hear, Lung and Blood program office (listing available at http://www.whi.org). We also recognize the WHI participants for their extraordinary contributions to the WHI program. The study sponsor had no role in the design and conduct of the study, the data collection, analyses, and interpretations, the writing of the article, or the decision to submit for publication. The views expressed in this article are those of the authors and do not necessarily represent the views of the NCI.
Funding
This work was primarily supported by a grant from the National Cancer Institute at the National Institutes of Health (R03CA184327 to MHT). The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C and HHSN271201100004C.
Disclosure
HW received research grant support from ACEA Biosciences (Inst), AstraZeneca/MedImmune (Inst), Bristol-Myers Squibb (Inst), Celgene (Inst), Clovis Oncology (Inst), Exelixis (Inst), Eli Lilly (Inst), Genentech/Roche (Inst), Gilead Sciences (Inst), Novartis (Inst), Pfizer (Inst), Parmacyclics (Inst) and Xcovery (Inst). HW has served a consultant to Peregrine, ACEA Biosciences, Pfizer, Genentech/Roche, Helsinn Therapeutics. XM is an employee of and holds stock in United Health Group. All remaining authors have declared no conflicts of interest.
Appendix
Table A1.
Baseline characteristics by smoking status in the WHI study, 1993–2013
| Baseline smoking statusa |
||
|---|---|---|
| Characteristic at baseline | Never (n = 75 962) | Ever (n = 73 515) |
| Age at screening (years) (%) | ||
| <50–59 | 23 644 (31.1) | 24 974 (34.0)b |
| 60–69 | 33 832 (44.5) | 33 773 (45.9) |
| 70–79 | 18 486 (24.3) | 14 768 (20.1) |
| Race-ethnicity (%) | ||
| Non-Hispanic white | 62 075 (81.9) | 62 532 (85.3)b |
| Black/African American | 6485 (8.6) | 6661 (9.1) |
| Hispanic/Latino | 3354 (4.4) | 2036 (2.8) |
| American Indian | 297 (0.4) | 320 (0.4) |
| Asian/Pacific islander | 2643 (3.5) | 1050 (1.4) |
| Other/Unknown | 1108 (1.2) | 916 (1.0) |
| Education (%) | ||
| High school or less | 18 094 (24.0) | 15 223 (20.9)b |
| School after high school | 27 464 (36.4) | 28 970 (39.7) |
| College degree or higher | 29 849 (39.6) | 28 797 (39.4) |
| Physical activity (MET/week) (%) | ||
| 0–3.00 | 21 062 (29.0) | 19 783 (28.2)b |
| >3.00–<11.75 | 23 627 (32.5) | 21 637 (30.8) |
| ≥11.75 | 27 926 (38.5) | 28 845 (41.0) |
| Alcohol intake (%) | ||
| Never | 13 896 (18.4) | 2241 (3.1)b |
| Past drinker | 13 096 (17.4) | 14 917 (20.4) |
| <1 drink per month | 10 300 (13.6) | 8174 (11.2) |
| <1 drink per week | 16 015 (21.2) | 14 544 (20.0) |
| 1–<7 drinks per week | 16 795 (22.3) | 21 382 (29.2) |
| ≥7 drinks per week | 5391 (7.1) | 11 972 (16.4) |
| BMI, kg/m2 (%) | ||
| <25 | 26 352 (35.0) | 25 138 (34.5) |
| 25–30 | 26 030 (34.6) | 25 454 (34.9) |
| ≥30 | 22 947 (30.5) | 22 261 (30.6) |
| Hormone-therapy use (%) | ||
| Never used | 32 496 (42.8) | 29 853 (40.6)b |
| <5 years | 16 455 (21.7) | 16 608 (22.6) |
| 5 to <10 years | 9676 (12.7) | 10 016 (13.6) |
| ≥10 years | 17 333 (22.8) | 17 036 (23.2) |
| Statin use (%) | 5749 (7.6) | 6019 (8.2)b |
| 10-year probability of hip fracture (%) | ||
| <0.53 | 23 727 (31.2) | 22 902 (31.2)b |
| 0.53–1.68 | 24 829 (32.7) | 25 074 (34.1) |
| ≥1.69 | 27 406 (36.1) | 25 539 (34.7) |
| Total calcium intake (supplements and diet) (mean ± SD, mg/day) | 1187.7 ± 772.1 | 1174.4 ± 719.7b |
| Total vitamin D intake (supplements and diet) (mean ± SD, U/day) | 376.8 ± 280.7 | 370.4 ± 278.9b |
| Oral bisphosphonate use (%) | 1660 (2.2%) | 1440 (2.0%)b |
| Duration of bisphosphonate usec (mean ± SD, years) | 1.22 ± 1.6 | 1.21 ± 1.8 |
Missing value were not included in the table.
χ 2 tests for categorical variables and Student’s t-tests for continuous variables were used to compare oral bisphosphonate users with nonusers. P value < 0.01.
Among baseline bisphosphonate users.
References
- 1. Siegel RL, Miller KD, Jemal A.. Cancer statistics, 2017. CA Cancer J Clin 2017; 67(1): 7–30. [DOI] [PubMed] [Google Scholar]
- 2. Whitaker M, Guo J, Kehoe T, Benson G.. Bisphosphonates for osteoporosis–where do we go from here? N Engl J Med 2012; 366(22): 2048–2051. [DOI] [PubMed] [Google Scholar]
- 3. Van Acker HH, Anguille S, Willemen Y. et al. Bisphosphonates for cancer treatment: mechanisms of action and lessons from clinical trials. Parmacol Ther 2016; 158: 24–40. [DOI] [PubMed] [Google Scholar]
- 4. Rennert G, Pinchev M, Rennert HS.. Use of bisphosphonates and risk of postmenopausal breast cancer. J Clin Oncol 2010; 28(22): 3577–3581. [DOI] [PubMed] [Google Scholar]
- 5. Newcomb PA, Trentham-Dietz A, Hampton JM.. Bisphosphonates for osteoporosis treatment are associated with reduced breast cancer risk. Br J Cancer 2010; 102(5): 799–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Liu Y, Zhao S, Chen W. et al. Bisphosphonate use and the risk of breast cancer: a meta-analysis of published literature. Clin Breast Cancer 2012; 12(4): 276–281. [DOI] [PubMed] [Google Scholar]
- 7. Chlebowski RT, Col N.. Bisphosphonates and breast cancer incidence and recurrence. Breast Dis 2011; 33(2): 93–101. [DOI] [PubMed] [Google Scholar]
- 8. Chlebowski RT, Chen Z, Cauley JA. et al. Oral bisphosphonate use and breast cancer incidence in postmenopausal women. J Clin Oncol 2010; 28(22): 3582–3590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Oh YH, Yoon C, Park SM.. Bisphosphonate use and gastrointestinal tract cancer risk: meta-analysis of observational studies. World J Gastroenterol 2012; 18(40): 5779–5788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Passarelli MN, Newcomb PA, LaCroix AZ. et al. Oral bisphosphonate use and colorectal cancer incidence in the Women's Health Initiative. J Bone Miner Res 2013; 28(9): 2043–2048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Rennert G, Pinchev M, Rennert HS, Gruber SB.. Use of bisphosphonates and reduced risk of colorectal cancer. J Clin Oncol 2011; 29(9): 1146–1150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Newcomb PA, Passarelli MN, Phipps AI. et al. Oral bisphosphonate use and risk of postmenopausal endometrial cancer. J Clin Oncol 2015; 33(10): 1186–1190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Rennert G, Rennert HS, Pinchev M, Lavie O.. The effect of bisphosphonates on the risk of endometrial and ovarian malignancies. Gynecol Oncol 2014; 133(2): 309–313. [DOI] [PubMed] [Google Scholar]
- 14. Cardwell CR, Abnet CC, Veal P. et al. Exposure to oral bisphosphonates and risk of cancer. Int J Cancer 2012; 131(5): E717–E725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Vinogradova Y, Coupland C, Hippisley-Cox J.. Exposure to bisphosphonates and risk of common non-gastrointestinal cancers: series of nested case-control studies using two primary-care databases. Br J Cancer 2013; 109(3): 795–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Chiang CH, Huang CC, Chan WL. et al. Oral alendronate use and risk of cancer in postmenopausal women with osteoporosis: a nationwide study. J Bone Miner Res 2012; 27(9): 1951–1958. [DOI] [PubMed] [Google Scholar]
- 17. Lee WY, Sun LM, Lin MC. et al. A higher dosage of oral alendronate will increase the subsequent cancer risk of osteoporosis patients in Taiwan: a population-based cohort study. PLoS One 2012; 7(12): e53032. [doi: 53010.51371/journal.pone.0053032]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ayo-Yusuf OA, Olutola BG.. Epidemiological association between osteoporosis and combined smoking and use of snuff among South African women. Niger J Clin Pract 2014; 17(2): 174–177. [DOI] [PubMed] [Google Scholar]
- 19. Quandt SA, Spangler JG, Case LD. et al. Smokeless tobacco use accelerates age-related loss of bone mineral density among older women in a multi-ethnic rural community. J Cross Cult Gerontol 2005; 20(2): 109–125. [DOI] [PubMed] [Google Scholar]
- 20. Lucas R, Fraga S, Ramos E, Barros H.. Early initiation of smoking and alcohol drinking as a predictor of lower forearm bone mineral density in late adolescence: a cohort study in girls. PLoS One 2012; 7(10): e46940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Waugh EJ, Lam MA, Hawker GA. et al. Risk factors for low bone mass in healthy 40–60 year old women: a systematic review of the literature. Osteoporos Int 2009; 20(1): 1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Cairoli E, Eller-Vainicher C, Ulivieri FM. et al. Factors associated with bisphosphonate treatment failure in postmenopausal women with primary osteoporosis. Osteoporos Int 2014; 25(4): 1401–1410. [DOI] [PubMed] [Google Scholar]
- 23. Anderson GL, Manson J, Wallace R. et al. Implementation of the Women's Health Initiative study design. Ann Epidemiol 2003; 13(9 Suppl): S5–17. [DOI] [PubMed] [Google Scholar]
- 24. Curb JD, Mctiernan A, Heckbert SR. et al. Outcomes ascertainment and adjudication methods in the women's health initiative. Ann Epidemiol 2003; 13(9 Suppl): S122–S128. [DOI] [PubMed] [Google Scholar]
- 25. Hays J, Hunt JR, Hubbell FA. et al. The women's health initiative recruitment methods and results. Ann Epidemiol 2003; 13(9 Suppl): S18–S77. [DOI] [PubMed] [Google Scholar]
- 26. Drieling RL, LaCroix AZ, Beresford SA. et al. Validity of self-reported medication use compared with pharmacy records in a cohort of older women: findings from the women's health initiative. Am J Epidemiol 2016; 184(3): 233–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. FRAX® Fracture Risk Assessment Tool. http://www.shef.ac.uk/FRAX (1 February 2017, date last accessed).
- 28. Cox DR. Regression models and life-tables (with discussion). J R Statist Soc B 1972; 34: 187–220. [Google Scholar]
- 29. Boonen S, Bouillon R, Haentjens P, Vanderschueren D.. Optimizing the benefits of bisphosphonates in osteoporosis: the importance of appropriate calcium intake. Treat Endocrinol 2006; 5(6): 375–383. [DOI] [PubMed] [Google Scholar]
- 30. Neville-Webbe HL, Holen I, Coleman RE.. The anti-tumour activity of bisphosphonates. Cancer Treat Rev 2002; 28(6): 305–319. [DOI] [PubMed] [Google Scholar]
- 31. Coleman R, Cook R, Hirsh V. et al. Zoledronic acid use in cancer patients: more than just supportive care? Cancer 2011; 117(1): 11–23. [DOI] [PubMed] [Google Scholar]
- 32. Fournier P, Boissier S, Filleur S. et al. Bisphosphonates inhibit angiogenesis in vitro and testosterone-stimulated vascular regrowth in the ventral prostate in castrated rats. Cancer Res 2002; 62: 6538–6544. [PubMed] [Google Scholar]
- 33. Winter MC, Holen I, Coleman RE.. Exploring the anti-tumour activity of bisphosphonates in early breast cancer. Cancer Treat Rev 2008; 34(5): 453–475. [DOI] [PubMed] [Google Scholar]
- 34. Meraviglia S, Eberl M, Vermijlen D. et al. In vivo manipulation of Vgamma9Vdelta2 T cells with zoledronate and low-dose interleukin-2 for immunotherapy of advanced breast cancer patients. Clin Exp Immunol 2010; 161: 290–297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Castella B, Riganti C, Fiore F. et al. Immune modulation by zoledronic acid in human myeloma: an advantageous cross-talk between Vgamma9Vdelta2 T cells, alphabeta CD8+ T cells, regulatory T cells, and dendritic cells. J Immunol 2011; 187(4): 1578–1590. [DOI] [PubMed] [Google Scholar]
- 36. Xie F, Li P, Gong J. et al. The bisphosphonate zoledronic acid effectively targets lung cancer cells by inhibition of protein prenylation. Biochem Biophys Res Commun 2015; 467(4): 664–669. [DOI] [PubMed] [Google Scholar]
- 37. Yuasa T, Kimura S, Ashihara E. et al. Zoledronic acid – a multiplicity of anti-cancer action. Curr Med Chem 2007; 14: 2126–2135. [DOI] [PubMed] [Google Scholar]
- 38. Di Salvatore M, Orlandi A, Bagalà C. et al. Anti-tumour and anti-angiogenetic effects of zoledronic acid on human non-small-cell lung cancer cell line. Cell Prolif 2011; 44(2): 139–146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Li YY, Chang JW, Chou WC. et al. Zoledronic acid is unable to induce apoptosis, but slows tumor growth and prolongs survival for non-small-cell lung cancers. Lung Cancer 2008; 59(2): 180–191. [DOI] [PubMed] [Google Scholar]
- 40. Yuen T, Stachnik A, Iqbal J. et al. Bisphosphonates inactive human EGFRs to exert antitumor actions. Proc Natl Acad Sci USA 2014; 111(50): 17989–17994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Stachnik A, Yuen T, Iqbal J. et al. Repurposing of bisphosphonates for the prevention and therapy of nonsmall cell lung and breast cancer. Proc Natl Acad Sci USA 2014; 111(50): 17995–18000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Sholl LM. Biomarkers in lung adenocarcinoma: a decade of progress. Arch Pathol Lab Med 2015; 139(4): 469–480. [DOI] [PubMed] [Google Scholar]
- 43. Langer RD, White E, Lewis CE. et al. The Women's Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol 2003; 13(9): S107–S121. [DOI] [PubMed] [Google Scholar]
- 44. Jha S, Wang Z, Laucis N, Bhattacharyya T.. Trends in media reports, oral bisphosphonate prescriptions, and hip fractures 1996–2012: an ecological analysis. J Bone Miner Res 2015; 30(12): 2179–2187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Kim SC, Kim DH, Mogun H. et al. Impact of the U.S. Food and Drug Administration’s Safety-related announcements on the use of bisphosphonates after hip fracture. J Bone Miner Res 2016; 31(8): 1536–1540. [DOI] [PMC free article] [PubMed] [Google Scholar]