Disparities in Plasma Cell Neoplasms in Kentucky, Appalachia and Other States: A Population Based Study

Background:

Historically, Kentucky has had one of the highest rates of cancer mortality in USA. In non-hematologic malignancy, the Appalachian region of the eastern United States is also associated with poor outcomes, however, the relationship with hematologic malignancy is poorly understood. We aim to study the disparities of Plasma Cell Neoplasms (PCN) in this region utilizing the Surveillance, Epidemiology, and End Results Program (SEER) database.

Methods:

We identified patients with PCN (multiple myeloma [MM], solitary plasmacytomas [PC], and plasma cell leukemia [PCL]) from the SEER database between 2000-2015. Data obtained included demographics, state, residence in an Appalachian region, rural/urban continuum code, median annual household income, and overall survival (OS) outcomes. Kentucky and Georgia are the only states that report to SEER which have populations from Appalachia. Therefore, patients were classified into 5 groups: Appalachia/Kentucky, non-Appalachia/Kentucky, Appalachia/Georgia, non-Appalachia/Georgia, and other (non-Appalachian) states. We used Kaplan-Meier & Cox regression to analyze survival outcomes. Income was analyzed as a continuous variable. Variables with a p value < 0.1 in univariate analysis were included in a stepwise multivariate Cox proportional hazard ratio (HR) model.

Results

A total of 68,627 patients were identified and included in the study (5.5% [n=3806] in Kentucky and 94.5% in other states). 3028 were identified as Appalachian (1969 in Georgia and 1059 in Kentucky). Baseline characteristics were comparable between Kentucky and other states except for income and rural/urban code. Percentages of patients with an income of <50k/year in Kentucky and other states were 56% and 22%, respectively (p<0.0001). In Kentucky, 55% of the patients were classified as living in a metro area, compared to 91% in other states (p<0.0001). All patients from Appalachia/Kentucky had an income of <50k/year, compared to 38% from Appalachia/Georgia. On univariate analysis, age >70 (HR=2.83, vs <60), age 60-70 (HR=1.44), diagnosis of MM/PCL vs PC (HR=1.85), black vs white race (HR=1.06), residence in Appalachia/Kentucky (HR=1.24) or non-Appalachia/Kentucky (HR=1.11) vs other states, year of diagnosis 2000-2005 (HR=1.4) or 2006-2010 (HR=1.12) vs 2011-2015, residence in a rural/urban area vs metro area (HR=1.16) and income (HR=1.029 for each 5000$ decrease) were significantly associated with worse OS (Table 1). Figure 1 shows the survival curves according to the region. On multivariate analysis (Table 2), age > 70 (HR=2.76) or 60-70 (HR=1.93), diagnosis of MM/PCL (HR=1.67), year of diagnosis 2000-2005 (HR=1.39) or 2006-2010 (HR=1.13), and income (HR=1.03 for each 5000$ decrease) persisted as significant risk factors for worse OS. Race, rural/urban code and being from non-Appalachia/Kentucky were no longer predictive of OS, while there was a non-significant trend towards worse OS for being from Appalachia/Kentucky (HR=1.07, CI=0.99-1.16, p=0.08).

Conclusions

Overall survival of patients with PCN in Kentucky in general, and even more prominently of patients from the Kentucky Appalachia area, is worse when compared to other states. However, when adjusted for income, this disparity corrects. This data highlights the importance of improving health outreach to this at-risk population. More studies focusing on underlying causes, such as education, compliance, co-morbidities and access to a care are warranted.

View largeDownload slide

View largeDownload slide

Close modal