Adenocarcinoma, Esophageal Cancer, Squamous Cell Carcinoma
Conditions
Brief summary
This is a 2-phase retrospective database study, using both case-cohort and inception (intention-to-treat) cohort analyses to evaluate any association between oral treatments for osteoporosis and the risk of esophageal cancer in women.
Interventions
Sponsors
World Health Information Science Consultants, LLC
Organon and Co
Study design
Observational model
COHORT
Time perspective
RETROSPECTIVE
Eligibility
Sex/Gender
FEMALE
Age
55 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Cases: * Women in the database aged 55 or older between 1996 and 2008 with diagnosis of esophageal cancer * Comparator Controls: * Each case was matched to all women in the random subcohort of 25,000 who had the same year of birth as the case and were in the database at the time of diagnosis.
Exclusion criteria
* Women with diagnosis of any other cancer or Paget's Disease or who have received oral or intravenous steroids before the index date
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Exposure to study drug at least 720 days before disease onset | To determine the use of study drugs (alendronate, etidronate, ibandronate, risedronate, and raloxifene) among female participants with esophageal cancer (cases) and a comparison subcohort, a case-cohort analysis was performed using women meeting criteria from the General Practice Research Database (GPRD, United Kingdom). Exposure to osteoporosis drugs administered 720 days before cancer onset was determined in cases and compared to contemporaneous assessments in a comparison subcohort matched by year of birth and membership in the GPRD on the case's onset date. Cases were confirmed and case onset dates determined by electronic algorithm (based on electronic medical record data) or by medical record review. |
| Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | Up to approximately 7.3 years of follow-up | To assess the relative risk of esophageal cancer associated with osteoporosis study drugs (alendronate, etidronate, ibandronate, risedronate, and raloxifene), initiators of osteoporosis drugs and non-initiators (comparators, women sharing match criteria with the initiator) entered an inception cohort for every three-month period, beginning in the first quarter of 1996. Assignment to study drug exposure group remained fixed from the start of follow-up, analogous to an intent-to-treat analysis. The risk of esophageal cancer among initiators of study drug compared to non-initiators of study drug was estimated through calculation of a hazard ratio. For calculation of 721+ day hazard ratios, only esophageal cancer cases occurring at least 721 days from initiation of study drug were used. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug were used. |
Participant flow
Recruitment details
Participants were not recruited nor enrolled in this study. This study is a retrospective observational study. Data from the GPRD were anonymized and used to develop participant cohorts. All diagnoses and treatments were recorded in the course of routine medical practice.
Pre-assignment details
684,815 women in the General Practice Research Database (GPRD) formed the Overall Study Population. These women met the demographic criteria of being age 55 or older, were born between 1922 and 1955, had more than 720 days experience in the GPRD and at least one of those days fell between 1996 and 2008.
Participants by arm
| Arm | Count |
|---|---|
| Overall Study Population | 684,815 |
| Total | 684,815 |
Baseline characteristics
| Characteristic | Overall Study Population |
|---|---|
| Age, Customized <55 | 0 Participants |
| Age, Customized ≥55 | 684815 Participants |
| Sex: Female, Male Female | 684815 Participants |
| Sex: Female, Male Male | 0 Participants |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | — / — |
| other Total, other adverse events | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 |
Outcome results
Primary
Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis)
To assess the relative risk of esophageal cancer associated with osteoporosis study drugs (alendronate, etidronate, ibandronate, risedronate, and raloxifene), initiators of osteoporosis drugs and non-initiators (comparators, women sharing match criteria with the initiator) entered an inception cohort for every three-month period, beginning in the first quarter of 1996. Assignment to study drug exposure group remained fixed from the start of follow-up, analogous to an intent-to-treat analysis. The risk of esophageal cancer among initiators of study drug compared to non-initiators of study drug was estimated through calculation of a hazard ratio. For calculation of 721+ day hazard ratios, only esophageal cancer cases occurring at least 721 days from initiation of study drug were used. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug were used.
Time frame: Up to approximately 7.3 years of follow-up
Population: Inception Cohort came from the Overall Study Population beginning treatment with an osteoporosis study drug (initiators, 78,630 women) and 300,610 matched control women, who did not receive study drug (noninitiators). Participants may have been exposed to more than one study drug. Also, one comparator may have been used for multiple study drugs.
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Esophageal Cancer Cohort | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 32 Number of cases per 100,0000 woman-years |
| Comparison Sample (Case Cohort) | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 32 Number of cases per 100,0000 woman-years |
| Etidronate | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 42 Number of cases per 100,0000 woman-years |
| Ibandronate | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 46 Number of cases per 100,0000 woman-years |
| Risendronate | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 47 Number of cases per 100,0000 woman-years |
| Raloxifene | Number of Cases of Esophageal Cancer Per 100,000 Woman-Years (Intent-to-Treat Analysis) | 29 Number of cases per 100,0000 woman-years |
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of alendronate~compared to non-initiators of alendronate. For calculation of this hazard ratio, esophageal cancer cases occurring during the entire follow-up period were used.95% CI: [0.7, 1.3]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of etidronate~compared to non-initiators of etidronate. For calculation of this hazard ratio, esophageal cancer cases occurring during the entire follow-up period were used.95% CI: [0.8, 2]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of ibandronate~compared to non-initiators of ibandronate. For calculation of this hazard ratio, esophageal cancer cases occurring during the entire follow-up period were used.95% CI: [0.5, 3.4]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of risendronate~compared to non-initiators of risendronate. For calculation of this hazard ratio, esophageal cancer cases occurring during the entire follow-up period were used.95% CI: [0.9, 2]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of raloxifene~compared to non-initiators of raloxifene. For calculation of this hazard ratio, esophageal cancer cases occurring during the entire follow-up period were used.95% CI: [0.3, 2.3]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of~alendronate compared to non-initiators of alendronate. For calculation of 721+ day hazard ratios, only esophageal cancer cases occurring at least 721 days from initiation of study drug (data not shown) were used.95% CI: [0.7, 1.5]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of~etidronate compared to non-initiators of etidronate. For calculation of 721+ day hazard ratios, only~esophageal cancer cases occurring at least 721 days from initiation of study drug (data not shown) were used.95% CI: [0.7, 1.9]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of~ibandronate compared to non-initiators of ibandronate. For calculation of 721+ day hazard ratios,~only esophageal cancer cases occurring at least 721 days from initiation of study drug (data not shown) were used.95% CI: [0.8, 11.1]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of~risedronate compared to non-initiators of risedronate. For calculation of 721+ day hazard ratios,~only esophageal cancer cases occurring at least 721 days from initiation of study drug (data not shown) were used.95% CI: [1.1, 3]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of~raloxifene compared to non-initiators of raloxifene. For calculation of 721+ day hazard ratios, only~esophageal cancer cases occurring at least 721 days from initiation of study drug (data not shown) were used.95% CI: [0.1, 2.7]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of alendronate~compared to non-initiators of alendronate. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug (data not shown) were used.95% CI: [0.5, 1.6]Proportional Hazards Regression Model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of etidronate~compared to non-initiators of etidronate. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug (data not shown) were used.95% CI: [0.6, 2]Proportional hazards regression model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of risedronate~compared to non-initiators of risedronate. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug (data not shown) were used.95% CI: [0.4, 2.5]Proportional hazards regression model
Comparison: A hazard ratio was calculated to estimate the risk of esophageal cancer among initiators of raloxifene~compared to non-initiators of raloxifene. For calculation of 1441+ day hazard ratios, only esophageal cancer cases occurring at least 1441 days from initiation of study drug (data not shown) were used.95% CI: [0.2, 3.9]Proportional hazards regression model
Primary
Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis)
To determine the use of study drugs (alendronate, etidronate, ibandronate, risedronate, and raloxifene) among female participants with esophageal cancer (cases) and a comparison subcohort, a case-cohort analysis was performed using women meeting criteria from the General Practice Research Database (GPRD, United Kingdom). Exposure to osteoporosis drugs administered 720 days before cancer onset was determined in cases and compared to contemporaneous assessments in a comparison subcohort matched by year of birth and membership in the GPRD on the case's onset date. Cases were confirmed and case onset dates determined by electronic algorithm (based on electronic medical record data) or by medical record review.
Time frame: Exposure to study drug at least 720 days before disease onset
Population: Case-Cohort analysis population came from the Overall Study Population and comprised 929 women with esophageal cancer (cases) and a Comparison Sample of 338,911 matched control women. Participants may have been exposed to more than one study drug. Also, one comparator may have been used for multiple study drugs.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Esophageal Cancer Cohort | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Etidronate | 3.7 Percentage of participants |
| Esophageal Cancer Cohort | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Risedronate | 2.6 Percentage of participants |
| Esophageal Cancer Cohort | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Ibandronate | 0.3 Percentage of participants |
| Esophageal Cancer Cohort | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Raloxifene | 0.3 Percentage of participants |
| Esophageal Cancer Cohort | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Alendronate | 4.6 Percentage of participants |
| Comparison Sample (Case Cohort) | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Raloxifene | 0.4 Percentage of participants |
| Comparison Sample (Case Cohort) | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Alendronate | 2.7 Percentage of participants |
| Comparison Sample (Case Cohort) | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Etidronate | 2.1 Percentage of participants |
| Comparison Sample (Case Cohort) | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Ibandronate | 0.03 Percentage of participants |
| Comparison Sample (Case Cohort) | Percentage of Participants With Exposure to Study Drugs (Case-Cohort Analysis) | Risedronate | 0.9 Percentage of participants |
Comparison: A hazard ratio estimate was calculated from the case-control matched set to measure the risk association between exposure to alendronate treatment and incidence of esophageal cancer.95% CI: [0.7, 1.5]Logistic Regression Model
Comparison: A hazard ratio estimate was calculated from the case-control matched set to measure the risk association between exposure to etidronate treatment and incidence of esophageal cancer.95% CI: [0.9, 2]Logistic Regression Model
Comparison: A hazard ratio estimate was calculated from the case-control matched set to measure the risk association between exposure to ibandronate treatment and incidence of esophageal cancer.95% CI: [1.4, 16.7]Logistic Regression Model
Comparison: A hazard ratio estimate was calculated from the case-control matched set to measure the risk association between exposure to risedronate treatment and incidence of esophageal cancer.95% CI: [1, 2.5]Logistic Regression Model
Comparison: A hazard ratio estimate was calculated from the case-control matched set to measure the risk association between exposure to raloxifene treatment and incidence of esophageal cancer.95% CI: [0.2, 2.2]Logistic Regression Model