Reply

Abstract

We thank Dr. Risch for his comments regarding our recent manuscript investigating the association between prediagnosis aspirin use and risk of pancreatic cancer among participants in the Health Professionals Follow-up Study (HPFS) and Nurses’ Health Study (NHS).1Khalaf N. et al.Gastroenterology. 2018; 154: 1380-1390Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar Dr. Risch noted different pancreatic cancer case numbers between our recent study1Khalaf N. et al.Gastroenterology. 2018; 154: 1380-1390Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar and a previously published study in the same cohorts by our colleagues and coauthors Cao et al.2Cao Y. et al.JAMA Oncol. 2016; 2: 762-769Crossref PubMed Scopus (228) Google Scholar All pancreatic cancer cases described in Cao et al.’s study were included in our case population. We were also able to identify additional cases by: (1) performing reviews of more recently collected medical records in both cohorts, (2) including participants who had been diagnosed with multiple cancers, and (3) extending our follow-up period for the HPFS cohort. Cao et al, examined site-specific cancer risk, with each participant being followed until their first cancer diagnosis.2Cao Y. et al.JAMA Oncol. 2016; 2: 762-769Crossref PubMed Scopus (228) Google Scholar Although we excluded persons with a cancer diagnosis at baseline, those who developed cancer of any site throughout the follow-up period were eligible for study inclusion. Given the large case population and extensive analyses specifically of pancreatic cancer, our recent study was well powered to evaluate the association between prediagnosis aspirin use and risk of incident pancreatic cancer. We thank Dr Risch for pointing out that our study provides important information about prediagnosis aspirin use and risk of pancreatic cancer.1Khalaf N. et al.Gastroenterology. 2018; 154: 1380-1390Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar We agree that a quantitative summary of all prior studies was not undertaken as part of this analysis. The primary aims of our work were to evaluate the prediagnosis use of aspirin and nonsteroidal anti-inflammatory drugs in 2 large, prospective cohorts and investigate prediagnosis plasma salicylurate levels and risk of pancreatic cancer. We also considered multiple exposure classifications with differing lag times between aspirin use and incident pancreatic cancer to thoroughly evaluate for potential associations. Risch suggests a possible decreased risk of pancreatic cancer with aspirin use in more recent time periods.3Risch H.A. Cancer Epidemiol Biomarkers Prev. 2017; 26: 1155-1156Crossref PubMed Scopus (3) Google Scholar To investigate this in the Health Professionals Follow-up Study and the Nurses’ Health Study participants, we divided the follow-up period into 1980-2000 and 2000-2010 using the 2000 questionnaire return date as the cutoff between the 2 groups. We found similar multivariable-adjusted relative risks between prediagnosis aspirin use and incident pancreatic cancer when analyzed by time period (follow-up period 1980-2000, relative risk, 0.97 [95% confidence interval, 0.80-1.17]; follow-up period 2000-2010, relative risk, 0.92 [95% confidence interval, 0.79-1.08]). Re: NSAID Use and Pancreatic Cancer RiskGastroenterologyVol. 155Issue 3PreviewDrs Khalaf et al1 have provided an analysis of regular use of aspirin and risk of pancreatic cancer in the Health Professionals Follow-up Study and the Nurses’ Health Study (NHS), two substantial cohort studies. Given that these 2 studies were recently analyzed by Cao et al2 for the same association, to interpret these reports it would help to clarify some of the data involved in them. In particular, because the 2 analyses seem to cover essentially the same follow-up periods of the 2 cohorts, it is unclear why appreciably different numbers of pancreatic cancer cases were reported. Full-Text PDF