Characteristics of Synchronous and Multiple Primary Ovarian Cancer Cases in SEER and Kentucky

OBJECTIVES: The primary aim of this study was to investigate whether ovarian cancer patients in Kentucky are more likely to have multiple primary and synchronized primary cancers compared to Non-Kentucky SEER registry cancer patients. A secondary aim of the study was to determine if there are other factors that may be associated with an increased risk of having multiple primary and synchronized cancer as well as determining if having multiple primary and synchronous cancer diagnoses changes survival rates in ovarian cancer patients. Additionally, another objective was to identify the most frequently observed subsequent and synchronous cancers sites that are diagnosed in ovarian cancer patients within Kentucky. METHODS: The data for this retrospective, population-based cohort study of 72,491 primary ovarian cancer patients were obtained from the Kentucky Cancer Registry. To be included in the study population, the patients from Kentucky and Non-Kentucky areas within the SEER registry had to have been diagnosed between January 1, 2000 and December 31, 2014. Subjects were classified as a multiple primary cancer case based on whether they were diagnosed with a second primary cancer in addition to their first primary of ovarian cancer within the years 2000-2014. Subjects were classified as a synchronous cancer case if they were diagnosed with their secondary cancer diagnosis within 6 months or less or their first diagnosis. Logistic and Cox regression were utilized to identify factors associated with multiple primary cancer and synchronized cancer. RESULTS: After much investigation, the focus of this analysis turned to early stage disease and the comparison to all stage for multiple primary and synchronous ovarian cancers. This is due to the fact that in the analysis, an interaction was found between early stage ovarian cancer and the location variable with Kentucky having 1.25 greater odds of being diagnosed with early stage ovarian cancer (adjusted OR=1.25 [95% CI (1.04, 1.5)], p=0.0364). There were statistically significant differences in the distribution of early stage and late stage cancer cases in patients who had multiple primaries but not for synchronous cancers between Kentucky and SEER (early stage multiple primary p=0.0071, early stage synchronous p=0.1008). There was also a significant association between multiple primary cancer status between Kentucky and SEER with Kentucky patients having a 1.23 times the odds of developing multiple primaries compared to SEER patients; adjusted OR=1.23, [95% CI (1.02, 1.47)]. The Cox proportional hazard model for multiple primary cancer adjusted for early stage revealed that there were no statistically significant differences between Kentucky and SEER in regards hazard of being diagnosed with a multiple primary cancer (adjusted OR=1.04 [95% CI (0.91, 1.19)]). There were no survival differences seen between Kentucky and SEER for multiple primary cases (log-rank test=0.7093). There were also no survival differences seen between Kentucky and SEER for synchronous cancer cases (log-rank test: 0.8419). Survival curves were also constructed for early stage of disease for both multiple primary and synchronous cancer cases and no differences were seen (log-rank test: 0.5805 and log-rank test: 0.9833). Survival differences were seen when comparing Kentucky and SEER for both multiple primary and synchronous cancer cases with Kentucky having much lower survival in both cases (log-rank test: 0.0329 and log-rank test: 0.0353). CONCLUSIONS: Based on the results from the Cox proportional hazard model, primary ovarian cancer patients living in Kentucky are not at a greater risk of having multiple primary cancers than those residing in Non-Kentucky SEER areas, regardless of stage of disease. When observing the results from the logistic model for early stage of disease, the results indicate that Kentucky is at a greater risk for developing early stage multiple primary cancers in comparison to Non-Kentucky SEER. This difference could be due to the fact that a large number of observations had to be deleted due to the fact that date of diagnosis for the secondary cancer for these patients could not be obtained because the information was not in the database. High risk groups identified in this study are women who were diagnosed in earlier years, were older, white, single, lived in an urban area, had grade I cancer, and those who had sex chord-stromal tumors. Further analyses are needed to determine the definitive implications of the higher proportions of early stage cancer within Kentucky