COVID-19 in patients with advanced stages of diabetic kidney disease: new onset renal replacement therapy as one of the predictors of adverse outcome

Background: patients with Diabetes Mellitus 2 (DM2) and advanced stages of Diabetic Kidney Disease (DKD) are at high risk for the lethal outcome of COVID-19. The causes of high mortality and the prognostic signifi cance of the new onset of renal replacement therapy (hemodialysis de novo, HD de novo) among these patients are still points of debate. Aim: the identifi cation of risk factors (RF) of lethal outcome in patients with DKD 4-5D stages and evaluation of the prognostic value of HD de novo in patients not receiving HD at the time of hospital admission. Methods: the patients with COVID-19 and advanced stages of DKD were included in a retrospective observational study from 04.01. to 10.30.2020. The endpoints were the outcome of hospitalization (discharge/death) and HD de novo initiation during the inpatient course. Several demographic, DM2, DKD, and COVID-19-associated signs and laboratory parameters were analyzed as independent variables. The subgroup of patients with HD de novo was selected from the general cohort. Results: 120 patients with DKD 4-5D stages were included, with a mean age of 69±10 y, females - 52%. Initially, the observation cohort was divided into subgroups: DKD 4-5 and DKD 5D on maintenance hemodialysis (MHD). The mortality among patients with DKD 4-5 was comparable with the patients on MHD (38,2% vs 38,5%, р=0,975). The independent predictors of lethal outcome in group DKD 4-5 were: age ≥65 y (OR 12,30;95% CI 1,40-33,5;р=0,009), initial prandial glycemia ≥10 mmol/l (OR 14,5;95% CI 3,7-55,4;р<0,001), albuminemia at admission ≤35 g/l (OR 5,17;95% CI 1,52-17,50;р=0,012), Charlson comorbidity index (CCI) ≥10 (OR 6,69;95% CI 1,95-23,00;р=0,002), News2 >4 at admission (OR 7,58;95% CI 2,18-26,37;р=0,001), lung damage CT 3-4 at admission (OR 3,39;95% CI 1,09-10,58;р=0,031). In subgroup DKD 5D the independent predictors of lethal outcome were prandial glycemia at admission ≥10 mmol/l (OR 28,5;95% CI 7,1-33,5;р<0,001), lung damage at admission CT 3-4 (OR 8,35;95% CI 2,64-26,40;р<0,001), CCI ≥10 (OR 6,00;95% CI 1,62-22,16;р=0,006). To determine the risk of lethal outcome predictive models were created using identifi ed risk factors and variables. The predictive value for DKD 4-5 group was 93%, and for DKD 5D was 88%. The assessment of the overall predictive value of these models was carried out using ROC analysis. The mortality among patients with DKD 4-5 without HD de novo was 21,6% vs 72,2% in patients with initiated HD de novo (р<0,001). The independent predictors of HD de novo during the inpatient course were: prandial glycemia at admission ≥10 mmol/l (OR 3,38;95% CI 1,04-10,98;р=0,050), albuminemia at admission ≤35 г/л (OR 3,41;95% CI 1,00-11,55;р=0,050), News2 >4 at admission (OR 5,60;95% CI 1,67-19,47;р=0,006), eGFR ≤20 ml/min/1,73 m2 at admission (OR 4,24;95% CI 1,29-13,99;р=0,020). HD de novo was identifi ed as an independent predictor of adverse outcomes (OR 9,42;95% CI 2,58-34,4;р=0,001). The analysis of cumulative survival demonstrated comparable results in DKD 4-5 without HD de novo group and DKD 5D group. The cumulative 55-day survival in the subgroup with HD de novo was only 10%. Conclusion: the need to start HD de novo is one of the most powerful predictors of adverse outcomes of COVID-19 in patients with advanced DKD. The comparable mortality rate in DKD 4-5 and DKD 5D groups is due to extremely high mortality in the subgroup with HD de novo. The strict control and correction of HD de novo risk factors could turn them into modifi able ones and thus improve the survival prognosis of patients with advanced stages of DKD. © 2023 JSC Vidal Rus. All rights reserved.