Wang, Alice T.

Abstract Background: Essential thrombocythemia (ET), polycythemia vera (PV) and myelofibrosis (MF) are collectively classified as myeloproliferative neoplasms (MPN). Given that little is known about their etiology, we investigated the role of anthropometric, medical and lifestyle factors in the development of MPNs and the common subtypes (ET & PV) in a population-based cohort of women. Methods: The IWHS is a prospective cohort of Iowa women aged 55-69 years at enrollment in 1986. Participants completed 6 self-administered questionnaires between 1986-2004. This analysis includes 27370 women who completed the 1993 questionnaire, had no history of cancer, and were enrolled in Medicare at least 1 year. MPN cases through 2004 were identified by linkage to Medicare. Relative risks (RR) and 95% confidence intervals (CI) were estimated using Cox proportional hazard regression, with age as the underlying time variable. Results: After >250,000 person-years of follow-up, 257 cases of MPNs were identified (172 ET, 64 PV and 21 MF). Height, body mass index (BMI) and waist-to-hip ratio were not associated with risk of MPN. In subtype analysis, BMI was positively associated with risk of ET (RR=1.52 for >29.3 vs <23.5 kg/m2; p-trend=0.042) but not PV. Physical activity (PA) showed a suggestive association with MPN (RR=0.74 for high vs low activity; p-trend=0.063), specific only to ET (RR=0.66; p-trend=0.04). Diabetes (DM) was positively associated with risk of MPN (RR=1.51; 95% CI 1.02-2.24), also specific only to ET (RR=1.82; 95% CI 1.16-2.84). Regular use of aspirin, but not other non-steroidal anti-inflammatory agents, was associated with a lower risk of MPN (RR=0.71; 95% CI 0.55-0.92); slightly stronger for ET (RR=0.68; 95% CI 0.50-0.93) than for PV (RR=0.84; 95% CI 0.50-1.43). There was no dose-response with frequency of use. Cigarette smoking was associated with risk of MPN, which was stronger for PV (RR=0.94 for former and RR=2.83 for current; p-trend=0.016) than for ET (RR=1.28 for former and RR=1.32 for current; p-trend=0.15). There were no associations with alcohol use. In a multivariate model for ET, the associations for BMI (RR=1.28; p-trend=0.21) and DM (RR=1.51; 95% CI 0.92-2.46) attenuated, while there was little change for PA (RR=0.66; p-trend=0.062) or aspirin use (RR=0.70; 95% CI 0.51-0.96). In PV, there was little impact on the association with smoking (RR=0.92 for former and RR=2.67 for current; p-trend=0.028) after adjustment for BMI, PA, DM and aspirin use. Conclusions: Risk factors were unique to the 2 most common types of MPN. ET was associated with obesity-related factors (PV was not), and PV was associated with current smoking (ET was not). Aspirin use had a stronger inverse association for ET. Our results suggest distinct etiologies of these PMN subtypes, and raises mechanistic hypotheses related to inflammation for ET and smoking-related carcinogenic pathways for PV. Citation Format: Alexis D. Leal, Carrie A. Thompson, Alice Wang, Robert A. Vierkant, Thomas M. Habermann, Julie A. Ross, Beth A. Virnig, James R. Cerhan. Anthropometric, medical history and lifestyle risk factors for myeloproliferative neoplasms in the Iowa Women's Health Study (IWHS) cohort. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2536. doi:10.1158/1538-7445.AM2013-2536

Abstract Background: Cigarette smoking (CS) is an established, although relatively modest, colorectal cancer (CRC) risk factor. Emerging data suggest that CS might affect distinct pathways of colorectal carcinogenesis. In this study, we examined CS-associated CRC risks by KRAS mutation status in the population-based Iowa Women's Health Study (IWHS). Methods: The IIWHS recruited 41,836 randomly selected Iowa women, ages 55-69 years, with a valid driver's license at study entry (1986). CS and other exposure variables were assessed at baseline, by self-report. Incident CRC cases were ascertained by annual linkage with the Iowa Cancer Registry. Archived, paraffin-embedded tissue specimens were collected from 732 representative CRC cases diagnosed through 12/31/2002. CS was categorized by smoking status (current, former), average number of cigarettes per day (1-19, 20-39, > 40) and pack-years (1-19, 20-39, > 40); never smokers were defined as the reference group. KRAS mutation status (wildtype [wt] vs. mutated [mut]) was determined by direct sequence analysis of codons 12 and 13 (exon 2). Multivariate Cox regression models were fit to estimate relative risks (RRs) and 95% confidence intervals (CIs). Results: Of the 732 available CRC cases, 507 (69%) had complete CS and KRAS data. Each of the CS-related variables was more strongly associated with KRAS-wt vs. KRAS-mut tumors (see Table). However, event rates were relatively small in some cells, and tests for heterogeneity between KRAS-defined subtypes were not statistically significant (p > 0.05 for each comparison). Conclusions: These findings suggest that CS may affect colorectal carcinogenesis through KRAS-independent pathways, but additional data are needed from larger, more diverse study populations. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 1891. doi:10.1158/1538-7445.AM2011-1891