(22.11) Uncover Effects of Glycemic Levels on Mortality Rate of Cancer Patients Without Diabetes as related to Sociodemographic Characteristics

Saturday, May 11, 2024

12:50 PM – 1:05 PM CDT

Location: ePoster Showcase, Empire Ballroom Foyer, Level 2

Submitter(s)

Samuel H. Fantaye, MD, MPH, FACP

Assistant Clinical Professor
University of Texas Health San Antonio
San Antonio, Texas, United States

Objective : Hyperglycemia, often resulting from cancer treatments, impacts both diabetic and non-diabetic patients. Managing hyperglycemia in cancer patients is challenging as cancer treatment often takes precedence over glucose management, and there's a lack of specific protocols for glycemic control. Studies are needed to understand hyperglycemia's full impact on cancer patients without prior diabetes. We explored the relationship between blood glucose levels and mortality in cancer patients without a diabetes history, considering the influence of race/ethnicity on survival.

Methods: Retrospective study of adult cancer patients without history of diabetes, who were treated in City of Hope Comprehensive Cancer Center between 2010 and 2020. Data on survival and patient characteristics was retrieved from electronic medical records. Demographics and glucose levels categories 0 - 3 ( < 140, 140 to 180, 180 to 200, >200) were stratified by vital status (alive or deceased), and associations were tested with the chi-squared test for categorical data or the Wilcoxon test for continuous data. The effect of glucose on the vital status was estimated using logistic regression adjusting for age, gender, race, and ethnicity. Statistical testing was two-sided at significance level .05. All analyses were performed with R (v4, Vienna Austria).

Results: There were 25, 641patient included and roughly half 11,423 (45%) had normal glucose with 6,110 (24%) having glucose above 200. Glucose was notably associated with deceased status with normal glucose have the highest level of survival at 88% and glucose above 200 having a survival rate of just 12% (Chi-square p= < .001). There was broad representation of gender (52% female), ethnicity (25%Latino), and race. The Hawaiian natives and Pacific islander race were the most likely to have elevated glucose (64%). Latinos (40%) were less likely than non-Latino’s (46%) to have normal glucose. Results of the logistic regression of the mortality outcome showed Age, male gender, and Black race were all positively associated with mortality. Elevated glucose is strongly associated with mortality risk relative to normal glucose such that Category 1, 2, and 3 elevations have increasing risk with odd ratios of 2.11, 2.61, and 4.26, respectively(p < .001).

Discussion/Conclusion:

Hyperglycemia is common in cancer patient without prior history of diabetes and associated with significant risk of mortality, particularly minority groups, specifically black individuals. The impact of hyperglycemia in non-diabetic cancer patient needs further investigation to identify those at greatest risk for hyperglycemia and develop standardized glycemic control practices to improve cancer patient survival.