{"doi":"10.1002/hep.32051","title":"REPLY:","abstract":"Potential conflict of interest: Nothing to report. We thank Zhang et al. for their interest in our article.(1) A large‐scale registry database provides an exceptional opportunity to investigate biliary tract cancer (BTC), a rare cancer.(1,2) Although large databases can provide a robust resource to investigate rare cancers, these analyses may still suffer from a lack of detailed clinical information, such as liver function and coagulation disorders, as well as from confounding due to compliance and over‐the‐counter use of concomitant medications. The study subjects’ health conditions and indications of drug prescriptions were limited in the registry database; thus, prescription bias could not be ruled out. The patients ultimately prescribed aspirin may have a period of unexposed immortal time before starting aspirin use(3); in other words, patients must survive long enough to receive aspirin during the study period. We performed repeated analyses by setting the landmark times to 0.25, 0.5, 1.0, and 1.5 years to address this bias. With respect to stage, beneficial effects of aspirin on BTC‐specific mortality were observed after adjusting for stage in multivariate models restricted to patients with data on stage (n = 2,518; P < 0.001). Additionally, we created two pseudo‐populations with aspirin use or nonuse with balanced characteristics, and the benefits of aspirin still appeared be relevant. In reality, the patients with BTC with postdiagnosis aspirin use compared to nonusers tended to be older (71.2 vs. 67.6 years old, P < 0.0001) and more likely to receive surgery (51% vs. 38%, P < 0.0001). After accounting for these confounders through modeling and stratification analyses, the beneficial effect of postdiagnosis aspirin use on BTC‐specific mortality remained. Beyond estimation of BTC‐specific mortality, we examined the effects of aspirin on the overall mortality among these patients with BTC. The findings were in line with the BTC‐specific death results, and postdiagnosis aspirin use showed improved overall survival for each BTC subtype (P < 0.0001; age‐sex adjusted survival curves in Fig. 1). However, we feel that cancer‐specific mortality provides better evidence of causal relationships between aspirin and target cancers than overall mortality, because aspirin also prevents cardiovascular diseases, which could affect overall mortality.FIG. 1: Age‐adjusted and sex‐adjusted survival curves for postdiagnosis aspirin use and cancer‐specific mortality: BTC (A), gallbladder cancer (GBC) (B), ampulla of Vater cancer (AVC) (C), and cholangiocarcinoma (CC) (D).Observational studies to investigate the effects of drugs on the occurrence or morality of diseases are challenging. Although limitations still exist, the potential causal relation between aspirin and BTC‐specific mortality is supported by temporality, dose‐response relationships, strength of the associations, and biologic plausibility, as per Hill’s criteria of causal inference.(4) We agree that clinical trials are critical to evaluate the effect of this inexpensive and well‐tolerated drug on BTC survival and to determine the optimal dosage, initiation time, potential harms, adverse effects, and the magnitude of net benefits.","journal":"Hepatology","year":2021,"id":225752,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":0,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.8981,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2021-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":457659,"name":"Yi‐Hsiang Huang","orcid":"0000-0001-5241-5425","position":1,"is_corresponding":false},{"id":291448,"name":"Jill Koshiol","orcid":"0000-0002-3832-6204","position":2,"is_corresponding":false},{"id":317486,"name":"Mei‐Hsuan Lee","orcid":"0000-0002-0222-2251","position":0,"is_corresponding":true}],"reference_count":4,"raw_metadata":null,"created_at":"2026-07-18T23:54:30.292454Z","pmid":"34240448","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}