This study aims to investigate if hormonal treatments given for breast cancer are associated with musculoskeletal conditions such as arthritis, painful tendon conditions (tendinopathies), and carpal tunnel syndrome- a condition causing pain and numbness in the hand and wrist due to compression of a nerve at the wrist. The aim is to investigate this association in women who have already gone through the menopause in order to investigate the role of tablet hormonal treatments on the risk of these conditions after their naturally occurring hormones have stopped being produced inside the body.
The proposed study uses primary care (GP) and linked hospital data as part of a global collaboration using existing healthcare data. The study aims to use global collaboration in order to identify if associations are seen in other countries, and to better understand why associations may occur.
Objectives: To investigate if there is an association between aromatase inhibitor (AI) use compared to tamoxifen and the incidence of carpal tunnel syndrome (CTS), tendinopathy of the wrist, shoulder, Achilles tendon and osteoarthritis (OA) of the hand, shoulder, hip and knee in post- menopausal women treated for breast cancer.
Design
Cohort study using non-identifiable CPRD data mapped to the OHDSI Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) to evaluate the risk of adverse musculoskeletal events in post-menopausal women taking AI compared to tamoxifen to treat hormone receptor positive breast cancer. Study undertaken using OMOP CDM as planned replication in other international datasets, and to enable comparison between countries.
Participants
All post-menopausal (defined as over 55 years old) women with an incident diagnosis of breast cancer treated with either AI or tamoxifen. Those with a prevalent cancer, hormone receptor negative breast cancer (where possible to determine), breast cancer not treated with AI or tamoxifen, or an outcome event prior to breast cancer date of diagnosis are excluded.
Variables and Measurements
Drug exposure will be determined from the first recorded drug exposure; outcomes will be defined as the first incident event for each individual outcome (based upon the code list given in the appendix, generated in the OHDSI CDM using the Atlas tool).
Propensity score adjustment of patients will be undertaken to minimise confounding (with the choice of propensity score adjustment between matching and stratification made a priori). Cumulative incidence of outcomes will be plotted; cox proportional hazards modelling to estimate hazard ratios for each of the outcomes according to drug use.
Expected results
Incidence of adverse musculoskeletal outcomes at a population level associated with AI treatment in comparison to tamoxifen in post-menopausal women.
Outcome will be defined as the first incident CTS, tendinopathy or osteoarthritis record within the CPRD database. Outcome cohorts will be identified using OMOP CDM codes for diagnoses, based upon some codes that have been previously validated in datasets included in OHDSI, but also using the ATLAS diagnostics tool to investigate orphan codes.[1,2] Cohort entry and exit is defined as per the time at risk window: entry is at the index date, exit at date of death or migration, date of outcome, loss to follow up or where possible, 1825 and 3650 days after the index date.
Negative control outcomes
Negative control outcomes will be added into the analysis in order to investigate the validity of the propensity score adjustment for confounding (Appendix 3)
In order to undertake this subsequent analysis for the competing risk of mortality in this cohort, mortality will also be used as an outcome in this study.
Daniel Prieto-Alhambra - Chief Investigator - University of Oxford
Jennifer Lane - Corresponding Applicant - University of Oxford
Antonella Delmestri - Collaborator - University of Oxford
Edward Burn - Collaborator - University of Oxford
HES Admitted Patient Care