THU320 Uncontrolled Diabetes May Increase The Risk Of COVID-19-related Cytokine Storm

Abstract

Abstract Disclosure: S. Sadiq: None. K.E. Criner: None. A.D. Rao: Other; Self; Advisory Board Member for Horizon Pharmaceuticals, but no conflict of interest. S. Zheng: None. A. Vargas: None. A. Amalfitano: None. Introduction: COVID-19 infection can lead to a hyperinflammatory state resulting in cytokine storm and in turn, mortality [1]. Furthermore, persistent glucose elevations in the setting of cytokine storms have been associated with increased mortality [2]. We hypothesize that among patients with diabetes hospitalized with COVID-19 infection, those with higher baseline HbA1c are more predisposed to cytokine storm and mortality. Methods: A retrospective chart review was performed of 477 patients with COVID-19 infection hospitalized between March 1, 2020, to May 31, 2020, at a large urban center. Patients > 18 years of age were included. Baseline characteristics obtained included age, gender, BMI, race, comorbidities, diagnosis of diabetes, type of diabetes, admission HbA1c, length of stay (LOS), and endocrinology consult. Cytokine storm was identified by rheumatology consult based on clinical condition, inflammatory markers, and fibrinogen level or by the use of immunosuppressants (tocilizumab, anakinra or sarilumab). Results: Baseline characteristics showed that 191 (41%) were women, 286 were men (59%). Majority of patients were African American 38% (n=183); 23% were Hispanic (n=110), 12% White (n=61), 2.5% (n=12) Asian/ Pacific Islander, 24.5% (n=111) were mixed/other race. 34% of these patients had diabetes. Out of 477 patients, 152 had an admission HbA1c, and of those 34% had an HbA1c > 8%. 16% of patients with diabetes had cytokine storm (n=67) compared to 12% of patients without diabetes (n=267). Mean pre-admission HbA1c in patients with diabetes with cytokine storm was significantly higher at 10.2% compared to 7.8% in patients with diabetes without cytokine storm (p=0.01), indicating a possible link between diabetes control and cytokine storm. Patients with cytokine storm had a higher rate of endocrinology consults (31%) compared to those without (10%) (p<0.01). Patients with endocrine consults had a longer LOS (median of 11.2 days vs 4.25 days), which may indicate a higher severity of the illness. Additionally, there was a higher use of steroids, with a mean dose of methylprednisolone (1625 +/−1205 mg) in all patients with cytokine storm compared to (433 +/−514 mg) in those without cytokine storm over the course of hospitalization. Conclusion: Patients with diabetes hospitalized with COVID-19 infection having higher baseline HbA1c had a greater incidence of cytokine storm, which is well known to effect COVID-19-related mortality. Our study showed the importance of admission baseline HbA1c to identify those patients at risk and early interventions to decrease morbidity.