Cardiotoxicity of Fluorouracial (Fu-Based) Adjuvant Chemotherapy for Resected Colorectal Cancer: a Nationwide Cohort Study in Taiwan

Over the past two decades, improvements in early-stage cancer detection and treatment have significantly improved prognoses of several major cancers, especially colorectal cancer (CRC). However, non-cancer death (especially cardiotoxicity) of treatment is a major challenge for cancer survivors. In recent years, the impact of long-term cardiotoxicity of adjuvant chemotherapy for resected CRC have called for more real-world data by many high-impact journals. A population-based study, published in 2018, reported the 10-year cumulative incidence of new-onset cardiovascular disease and heart failure were 57.4% and 54.5% among patients with stage I-III CRC when compared with 22% and 18% for matched cohort without cancer, respectively. In Taiwan, the National Health Insurance Administration reimburses the intravenous bolus (IV-) administering 5-FU, oral UFT and oral capecitabine as an adjuvant chemotherapy for resected and non-metastasis CRC. Because the result of the 2018 study is limited to patients with age ＞ 65 years and only compared chemotherapy monotherapy (5-FU alone vs. capecitabine alone). It is not clear whether or not the oral fluoropyrimidine (UFT) has a lower rate of cardiotoxicity.In this study, we will compare the cardiotoxicity of adjuvant chemotherapy for resected non-metastasis colorectal cancer: UFT, capectiabine, 5-FU, mixed, using real-world evidence. The cardiotoxicity includes acute myocardial infarction (AMI), life-threatening arrhythmia (LTA), congestive heart failure (CHF), ischemic stroke (IS). These findings will be helpful for oncologists in daily practice. We will establish a nationwide cohort of patients with a newly diagnosed CRC from 2004-2016, receiving fluorouracial (FU-based) adjuvant chemotherapy (CT) for resection by linking Taiwan Cancer Registry (TCR), National Health Insurance Research Database (NHIRD), and Taiwan Death Registry (TDR). We anticipate that there will be approximately 52000 eligible patients (7800 in UFT only, 7800 in capecitabine only, 13000 in 5-FU only, and 23400 in mixed chemotherapy). All eligible patients will be followed from the half year after CRC first diagnosis (index date) till the occurrence of cardiotoxicity (AMI, LTA, CHF, IS, independently), death, loss follow-up, Dec. 31, 2018, whichever came first. The main statistical analysis is balancing all covariates among four chemotherapy groups by stabilized inverse probability of treatment weighting using propensity score (stabilized IPTW). Survival analysis (log-rank test in univariate analysis and Cox’s proportional hazard model/ cause-specific hazard model in multivariate analysis) will then be conducted to examine the association between the study outcomes and chemotherapy groups. Note that chemotherapy grouping is the only covariate which is included in either the Cox's model or cause-specific hazard model, because the four chemotherapy groups are balanced after stabilize IPTW. Other statistical analysis will also be conducted, including Cox's/Cause-specific model by including other covariates and chemotherapy grouping as static or time-dependent, landmark analysis, subgroup analysis, and computing E-values.

Project ID:PC10907-1133
External Project ID:MOST109-2314-B182-035