Assessment of a Resource and Descriptive Analysis of Participants with Multiple Primary Cancers

Abstract

We conducted an analysis of participants in the Nurses’ Health Study (NHS) and Nurses’ Health Study II (NHSII) cohorts, longitudinal studies with 34-47 years of follow-up, conducted jointly by Brigham and Women’s Hospital and Harvard School of Public Health. Our focus was on participants who had multiple primary cancers, confirmed by medical record review. Due to resource and time limitations, we further narrowed our study focus to participants who were diagnosed with various combinations of breast, colorectal, and ovarian cancer. We first assessed the resource, as we suspected large gaps existed in the data collection and disease confirmation processes and went on to identify a loss of a significant portion of reported multiple primary cancers during the disease follow-up and medical record confirmation process. This was due to deliberate decisions given limited resources within the NHS and NHSII staff and an aging study population whose participants were reporting cancer at a higher frequency. We conducted a descriptive analysis of characteristics of NHS and NHSII participants who were diagnosed with multiple primary cancers (breast, colorectal, and ovarian cancers). Variables assessed in our descriptive analysis included mean age of diagnosis at time of first cancer, mean age of diagnosis at time of second cancer, mean latency period between cancer diagnoses, mean body mass index (BMI) at age 18, mean BMI at time of diagnosis, historical alcohol consumption, race, and smoking status at time of diagnosis. We identified trends suggesting that increased alcohol consumption between the ages of 23 and 40 may have an impact on developing a second primary cancer after a first colorectal cancer. We also identified a trend suggesting that smoking status at time of diagnosis of first cancer may influence the development of a second cancer. It is important to note that the gaps in the multiple primary cancer confirmation process within the NHS and NHSII cohorts dissuaded us from conducting traditional statistical analyses on these factors. As a result, we provide these data as descriptive characteristics of participants who have multiple primary cancers. This analysis also sets up current and future NHS and NHSII cohort researchers with a backdrop for further study of multiple primary cancers. We also make recommendations for methods to help fill in the gaps in operational processes and data management.