Association Between Aspirin Use and Biliary Tract Cancer Survival

This database study analyzes the association between aspirin use and survival in patients with biliary tract cancer.


Introduction
Biliary tract cancers (BTCs) are rare, with a worldwide incidence of <2/100,000 individuals. 1e five-year survival rate is ~5-15%, with a median survival of <1 year. 1 Between 60-70% of patients present with late-stage disease (e.g.inoperable or metastatic tumors) owing to lack of symptoms. 2Consequently, there is a critical need for treatments that improve BTC survival.
Aspirin has been proposed as a treatment to reduce cancer mortality as it may slow cancer growth through the inhibition of both cyclooxygenase-2, which promotes inflammation and cell proliferation, 3 and platelet aggregation, which may slow the metastatic spread of cancer. 4We investigated post-diagnosis aspirin use and BTC survival.

Methods
We obtained data, including all-cause deaths, on adult patients diagnosed with BTC between 1990-2017 from the United Kingdom's Clinical Practice Research Datalink (CPRD), an electronic medical record database.We identified cancers using Read codes for gallbladder cancer (GBC), cholangiocarcinoma, ampulla of Vater cancer (AVC), and overlapping lesions of the biliary tract.We excluded patients with previous cancer, except for non-melanoma skin cancer.
Ever use of post-diagnosis aspirin was defined as ≥1 prescription recorded in the CPRD on or after the BTC diagnosis date.We used Cox proportional hazards regression models to estimate the cancer site-specific hazard ratios (HRs) and 95% CIs for the association between time-dependent post-diagnosis aspirin use and overall survival.Patients who received an aspirin prescription within 30 days of diagnosis entered the model as users.The time scale began at diagnosis until death, exit, or end of follow-up (truncated at 10 years).We adjusted for the following covariates a priori: age at diagnosis, sex, comorbidities, statin use at diagnosis, indicators of a healthy lifestyle, and year of diagnosis.We fit separate models for each BTC type and stratified the baseline hazard by pre-diagnosis aspirin use (yes/no).We estimated adjusted survival curves using a marginal approach to remove the sex and age effects on aspirin use, accounting for the time-dependent exposure. 5We conducted analyses using SAS (version 9.4; SAS Institute Inc) and survival curves in R Studio (version 1.1.453).

Discussion
We observed a reduced risk of death for post-diagnosis aspirin users across all BTC types.Platelet activation protects tumor cells from elimination, enhances metastatic cell growth, and enables cancerous cells to spread via the bloodstream. 4,6Aspirin may slow metastatic spread of cancer cells through inhibition of platelet aggregation, improving BTC survival. 1A limitation of our analysis is the lack of data on stage and chemotherapy regimens received (if any).
However, most BTCs are diagnosed at late stage 2 with <10% of patients presenting with resectable tumors and 50% of tumors metastasizing to the lymph nodes. 1 The survival benefit of aspirin observed in our study are on par with the current standard of care. 2 Author Contributions: Drs Sarah S. Jackson and Jill Koshiol had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.Role of the Funder/Sponsor: The funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.b The results presented used Cox regression where aspirin was modeled as time-dependent (e.g.individuals could switch between use and non-use status), the numbers (n) represent any aspirin use after BTC diagnosis.c Prevalent users were defined as patients with two or more aspirin prescriptions before BTC diagnosis.Incident users were defined as patients who only initiated aspirin use on or after BTC diagnosis date.P-values for interaction was estimated by putting a cross-product term in the models for post-diagnosis use and pre-diagnosis use Study concept and design: Jill Koshiol Acquisition, analysis, or interpretation of data: Sarah S. Jackson, Ruth M. Pfeiffer, Zhiwei Liu, Lesley Anderson, Huei-Ting Tsai, Shahinaz M. Gadalla, and Jill Koshiol Drafting of the manuscript: Sarah S. Jackson Critical revision of the manuscript for important intellectual content: Ruth M. Pfeiffer, Zhiwei Liu, Lesley Anderson, Huei-Ting Tsai, Shahinaz M. Gadalla, and Jill Koshiol Statistical analysis: Sarah S. Jackson and Ruth M. Pfeiffer.Obtained funding: Jill Koshiol Administrative, technical, or material support: Sarah S. Jackson, Zhiwei Liu, and Huei-Ting Tsai Study supervision: Jill Koshiol Conflict of Interest Disclosures: None reported.Funding: Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics.Disclaimer: Dr. Tsai is currently an employee of the US Food and Drug Administration.This manuscript reflects the views of the author and does not necessarily represent FDA's views or policies.This study is based on data from the CPRD database September 2018 release, obtained from the UK Medicines and Healthcare Products Regulatory Agency (Copyright © 2018).The interpretation and conclusions contained in this study are those of the authors alone.This study was approved by the National Institutes of Health Human Research Protection Program and the Independent Scientific Advisory Committee of the CPRD (Protocol 17_160.R)

Figure 1 .
Figure 1.Adjusted survival curves of overall survival among post-diagnosis aspirin users and non-users by cancer site

Table 1 .
Time-dependent associations between post-diagnosis aspirin use and overall survival for each biliary tract cancer site a Adjusted for sex, history of heart disease, statin use (current, former, never), presence of comorbidities, age at diagnosis, and year of diagnosis.Aspirin use was modeled as time-dependent and the baseline hazard was stratified by pre-diagnosis aspirin use.