key: cord-0977694-yb23xjtz
authors: Mondal, Sunetra; DasGupta, Riddhi; Lodh, Moushumi; Gorai, Ramprasad; Choudhury, Brojen; Hazra, Arindam Kumar; Ganguly, Arunangshu
title: Predictors of new-onset diabetic ketoacidosis in patients with moderate to severe COVID-19 receiving parenteral glucocorticoids: A prospective single-centre study among Indian type 2 diabetes patients
date: 2021-03-28
journal: Diabetes Metab Syndr
DOI: 10.1016/j.dsx.2021.03.022
sha: f427a7740233798209d5d709788a3190c6d4f269
doc_id: 977694
cord_uid: yb23xjtz

BACKGROUND AND AIM: COVID-19 infection predisposes to diabetic ketoacidosis(DKA); whether glucocorticoids enhances this risk is unknown.We aimed to study the occurrence of DKA after initiating glucocorticoids in patients with type 2 diabetes mellitus(T2DM) and moderate-to-severe COVID-19, and identify predictors for it. METHODS: Patients with T2DM and moderate or severe COVID-19 infection were prospectively observed for development of new-onset DKA for one week following initiation of parenteral dexamethasone. Clinical and biochemical parameters were compared between those who developed DKA (Group A) and those who didnot (Group B). Logistic regression was done to identify independent risk-factors predicting DKA; ROC-curve analysis to determine cut-offs for the parameters in predicting DKA. RESULTS: Amongst 302 patients screened, n = 196 were finally included, of whom 13.2% (n = 26,Group A) developed DKA. Patients in Group A were younger, had lower BMI, increased severity of COVID-19 infection, higher HbA1c%, CRP, IL-6, D-dimer and procalcitonin at admission (p(all) < 0.02). Further, admission BMI (OR: 0.43, CI: 0.27–0.69), HbA1c % (OR: 1.68, CI: 1.16–2.43) and serum IL-6 (OR: 1.02, CI: 1.01–1.03) emerged as independent predictors for DKA. Out of these, IL-6 levels had the highest AUROC (0.93, CI: 0.89–0.98) with a cut-off of 50.95 pg/ml yielding a sensitivity of 88% and specificity of 85.2% in predicting DKA. CONCLUSION: There is significant incidence of new-onset DKA following parenteral glucocorticoids in T2DM patients with COVID-19, especially in those with BMI <25.56 kg/m(2), HbA1c% >8.35% and IL-6 levels >50.95 pg/ml at admission.

Background and Aim: COVID-19 infection predisposes to diabetic ketoacidosis(DKA); whether glucocorticoids enhances this risk is unknown.We aimed to study the occurrence of DKA after initiating glucocorticoids in patients with type 2 diabetes mellitus(T2DM) and moderate-to-severe COVID-19, and identify predictors for it. • Risk is high for lean T2DM with severe COVID-19, high pro-inflammatory cytokines.

• Serum IL-6 level at admission has the best predictive potential for DKA. Following the preliminary results of the RECOVERY trial, dexamethasone is recommended in the treatment of cases with severe COVID-19 infection or moderate cases with persistent hypoxia. 4 The occurrence of diabetic ketoacidosis (DKA) in patients with T2DM is rare , and though viral infections and glucocorticoids are known to act as precipitating factors, the occurrence of DKA following glucocorticoid therapy has been sparsely reported in literature. 5, 6 There have been several reports of ketoacidosis precipitated by COVID-19 infection in patients with known DM and even in previously normoglycemic individuals, [7] [8] [9] [10] with impairment in both insulin secretion and sensitivity being implicated in the pathogenesis . 9,10 Whether initiation of parenteral glucocorticoids can potentially worsen the ketoacidotic milieu and increase the risk of DKA in these patients remains unknown.

J o u r n a l P r e -p r o o f Increased mortality and RAAS-mediated fluid and electrolyte alterations reported with DKA in COVID-19 patients present unique clinical challenges necessitating meticulous management. [11] [12] [13] Thus, analysing the association of parenteral glucocorticoids with newonset DKA and identifying its potential predictors can have significant therapeutic ramifications.

The current study was done with the primary objective of evaluating the occurrence of newonset DKA with initiation of parenteral glucocorticoid therapy in patients with T2DM and moderate to severe COVID-19 infection, and to identify clinical and biochemical factors that influence the development of DKA in them. The secondary objective was to derive cutoff values for these factors that can help in predicting the increased risk of DKA at admission .

The screening population comprised of patients with known T2DM and rt-PCR confirmed SARS-CoV2 admitted to a tertiary care hospital of Eastern India between 1 st of June 2020 to 30 th of November 2020 ( n = 302 ) .

J o u r n a l P r e -p r o o f

For the current study , we screened consecutive patients with T2DM and rt-PCR confirmed SARS-CoV2. As outlined in Fig 1, patients conforming to the inclusion criteria and filling the informed consent forms, underwent baseline clinical and biochemical evaluation and, once initiated on parenteral dexamethasone therapy, were monitored with capillary blood glucose levels (done before and 2 hours after every meal) , Arterial Blood Gas analysis (ABG) and urine ketones done daily for the subsequent one week to detect onset of DKA.

Clinical and biochemical parameters were eventually compared between those who developed DKA ( n = 26 ) and those who did not develop DKA ( n = 170 ) after dexamethasone therapy . Ethical clearance for the current project was obtained from the institutional ethical committee reference no HWH/IEC-BMHR/001/2020. Fig.1 , were followed to select the study subjects . Patients requiring immediate endotracheal intubation or mechanical ventilation, those with known autoantibody positivity for GAD-65 or IA-2, secondary or pancreatic diabetes and those who had already received or were on concurrent glucocorticoids prior to admission were excluded. Further, patients with prolonged starvation , chronic alcoholism , decompensated chronic liver disease or advanced chronic kidney disease ( CKD stages 4 and 5) , known malignancies, pregnancy and patients unable/ unwilling to give informed consent were also excluded. All the patients received usual standard of care for management of moderate to severe COVID-19 infection as per Indian national guidelines. 14 Additionally , parenteral dexamethasone was administered to all at a standard dose of 0.1 mg/kg for six consecutive days following admission. DKA and severity of DKA were defined using standard criteria . 15 Management of hyperglycemia and DKA was done in accordance with standard of care recommendations . 15 Biochemical investigations done at admission included serum levels of C reactive protein (CRP) , Interleukin-6 (IL-6) , D-dimer and Procalcitonin, in addition to complete blood count, electrolytes, renal and liver function tests. Serum CRP was measured using particle-J o u r n a l P r e -p r o o f enhanced immunoturbidimetry by Integra 400 + analyser (Roche Diagnostics, Rotkreuz, Switzerland), serum IL-6 and procalcitonin by electrochemiluminiscence assay using Cobas e411 analyser (Roche Diagnostics) and D -dimer using quantitative ELISA by Vidas Ddimer (Biomerieux) . Semiquantitative estimation of urine ketones was done using KetoDiastix reagent strips by Bayer Diagnostics, India and simultaneous urine sample analysis for ketone body estimation using Cobas -u411 analyser. Capillary blood glucose measurements were done before and 2 hours after every meals using Accu-Check Active glucometer and strips following standard measures and precautions. ABG analysis was done using Gem Premier 3500 from Instrumentation Laboratory.

Statistical analysis was performed using SPSS version 22.0 (SPSS Inc.,Chicago,Illinois).

The two groups were compared using unpaired t-test or Mann-Whitney's U test for quantitative variables and Chi-square test, with Fisher's correction where appropriate, for categorical variables . A p value <= 0.05 was considered significant. Multiple logistic regression to identify the independent predictors of DKA was done in two steps -univariate analysis (step 1) followed by multivariate analysis using independent variables which were found to be significant (p<0.05) or suggestive of significance (p<0.100) in step 1. ROC curves were constructed for variables that can predict DKA and cut-offs were determined J o u r n a l P r e -p r o o f using Youden's index in conjunction with ROC analysis. AUC >=0.5 was considered significant.

For the current study, n = 302 patients with T2DM and COVID-19 infection were screened during the study period of six months, of which n = 196 were included for the final analysis Overall , the mean age of the study population (n = 196) was 59·4 ( +/-10·8) years and 37·9% were females.. The mean baseline HbA1c was 8·1 ( +/-1·7) % (65 mmol/mmol) . Comparative analysis of clinical and biochemical characteristics were made between these two groups. (Table 1) . Those in Group A were younger in age , had lower BMI and higher mean HbA1c% than group B ( 54·8 vs 60·1 years , p = 0·02; 24·85 vs 27·36 kg/m 2 , p < 0·0001 and 10·11 % vs 7·72 % [87 mmol/mol vs 61 mmol/mol], p < 0·0001 respectively) .

Severe COVID-19 infection was more common in group A (n = 11, 42·3%) than in group B ( n = 38, 22·8%). The median CRP ( mg/l ), IL-6 ( pg/ml ), procalcitonin ( ng/ml ) and Ddimer levels ( ng/ml ) at admission were all higher in group A than group B ( p all < 0·001). The SpO2% at admission of the patients and the PaO2, PaCO2 in the initial J o u r n a l P r e -p r o o f ABG in the two groups were not significantly different between the two groups. There were no differences in the rates of pre-admission use of insulin or any oral hypoglycemic agents including SGLT2 inhibitors in the two groups. (Table 1) Amongst those in group A, DKA was severe in n = 4 ( 15 . 

ROC curves were constructed for BMI, HbA1c % and serum IL-6 levels as predictors of DKA at admission. Serum IL-6 levels showed the highest AUROC of 0·93 ( CI : 0·89 -0·98, p = 0·001 ) amongst the three [ Table 2 ]. At a cut-off of 50·95 pg/ml , serum IL-6 levels at admission predicted the occurrence of DKA with a sensitivity of 88%, specificity of 85·2% and LR of 5·95 . Similarly, cut-offs were derived for HbA1c% and BMI , though they displayed lower sensitivity and specificity than serum IL-6 levels ( Table 2, Figure 2 ) .

In our cohort of 196 Indian T2DM patients with moderate or severe COVID-19 infection , SARS-CoV-2 is known to enter pancreatic β cells via ACE2 receptor and cause downregulation of ACE2, which in turn potentiates Ang-II levels , with the latter exerting pro-inflammatory responses and beta cell cytotoxic responses through its receptors (AT1 or AT2) to impede insulin secretion. [20] [21] [22] Moderate to severe COVID 19 infection further increases the susceptibility to DKA through deterioration in peripheral glucose uptake mediated by release of pro-inflammatory cytokines . Though glucocorticoids are known to cause hyperglycemia, the risk of DKA following glucocorticoid use is mostly limited to case reports and is seen more with acute than chronic use. 5, 6, 17 . Glucocorticoids worsen insulin resistance through a myriad of pathways involving reduced muscle GLUT4 transporters, increased lipolysis and proteolysis and enhanced effects of counter-regulatory hormones such as glucagon and epinephrine. 18, 19 Simultaneously , there is also downgradation of the hyperinsulinemic response which is usually seen in the pancreatic β cells as a compensatory mechanism to the increase in insulin resistance . In extreme settings , such as those with moderate to severe COVID-19 infection , the absence of the anti-lipolytic effect of insulin may result in DKA. Thus the combined adverse metabolic effects of glucocorticoids and COVID-19 infection can explain the increased propensity of DKA seen in our study population.

In our study, the patients who developed DKA following parenteral dexamethasone had lower BMI, higher HbA1c %, higher prevalence of severe COVID-19 infection and higher serum levels of pro-inflammatory cytokines including CRP, IL-6 and D-dimer. Low BMI, high HbA1c % and serum IL-6 levels at admission were independent predictors of the risk for DKA. Low BMI and higher HbA1c % have previously been identified as risk factors for DKA in young adults with type 1 DM. 23, 24 Another study on DKA in COVID-19 patients highlighted that patients were younger and had higher rates of ARDS, acute liver injury and requirement for mechanical ventilation, underpinning more severe COVID-19 infection in those with DKA than in those without. 10 Severe COVID-19 disease is accompanied by higher levels of inflammatory cytokines, which are also elevated in the setting of DKA, independent of accompanying illnesses. 25, 26 IL-6 has been postulated to play a very important role in the maladaptive immune response to SARS-CoV-2 virus and is being studied as a possible treatment target. 27 Our finding of IL-6 as an independent predictor of DKA strongly suggests that in patients with moderate to severe COVID-19 infection , pro-inflammatory cytokines serve as one of the most important pathogenetic factors . Elevated IL-6 levels can impair insulin sensitivity , through SOCS-mediated pathways , as also insulin secretion , predominantly through macrophage-induced inflammation and subsequent beta-cell apoptosis, amyloidosis and fibrosis . A longer time to resolution of DKA , as seen in the current study , has been reported in previous studies in COVID-19 patients . 13, 28 The daily insulin requirement for resolution of DKA in our study J o u r n a l P r e -p r o o f population was higher than the usually recommended dosage required in type 1 DM subjects. 15 This can probably be attributed to increased insulin resistance in our patients as a result of the combined effects of parenteral glucocorticoid therapy and elevated pro- Upon ROC-analysis, we found admission serum IL-6 levels were superior to HbA1c% or BMI in predicting the occurrence of DKA . Further, we demonstrated that admission serum IL-6 greater than 50·95 pg/ml, HbA1c % greater than 8·35% and BMI less than 25·56 kg/m 2 can aid in early identification of patients at higher risk for DKA , who in turn would require more stringent monitoring and intensive care for better outcomes. Previous studies done amongst children and young adults with type 1 DM have demonstrated the association of lower BMI with increased risk for DKA, but failed to provide definite cut-off values. 29, 30 Similarly , raised levels of IL-6 have shown predictive potential in adult patients with DKA, especially in the presence of an underlying infection. 31, 32 Though decline in IL-6 levels have been seen with correction of DKA, there is a lack of established cut-off values that can predict DKA. 26, 30 In a study on patients with T2DM developing DKA, J o u r n a l P r e -p r o o f ketoacidosis was more likely with HbA1c values at ≥10·1% in patients with newly diagnosed T2DM and and at ≥8·6% in patients with a known history of T2DM respectively. 32 However, such HbA1c% cut-offs have not previously been reported for T2DM patients with concomitant COVID-19 infection.

The strengths of our study include a prospective design , stringent inclusion and exclusion criteria and meticulous monitoring of all study subjects for the development of DKA during the entire duration of parenteral dexamethasone use. We have put forth a set of clinically relevant predictors with suitable cut-off values that will aid in early identification and management of those susceptible to develop DKA following parenteral dexamethasone. 

The authors thank all the nurses, critical care technicians and laboratory technicians working in the critical care settings and emergency wards dedicated to management of COVID-19 infected patients in our hospital. Sodium-Glucose co-transporter-2, ,CRP = C reacrive protein, IL-6 = Interleukin-6 . p<0.05 taken as significant J o u r n a l P r e -p r o o f 

Are patients with hypertension and diabetes mellitus at increased risk for COVID-19 infection

Coronavirus infections and type 2 diabetes-shared pathways with therapeutic implications. Endocrine reviews

Dexamethasone in hospitalized patients with Covid-19-preliminary report. The New England journal of medicine

Steroid-induced diabetic ketoacidosis in a patient with type 2 diabetes mellitus

Clinical characteristics and outcomes of diabetic ketoacidosis in Pakistani adults with Type 2 diabetes mellitus. Diabetic medicine

Diabetic ketoacidosis precipitated by COVID-19: a report of two cases and review of literature

Euglycemic diabetic ketoacidosis in a patient with type 1 diabetes and SARS-CoV-2 pneumonia: case report and review of the literature

Diabetic ketoacidosis precipitated by Covid-19 in a patient with newly diagnosed diabetes mellitus

COVID-19 infection may cause ketosis and ketoacidosis. Diabetes, Obesity and Metabolism

Diabetic ketoacidosis in COVID-19: unique concerns and considerations

Unexpected high mortality in COVID-19 and diabetic ketoacidosis

Clinical profile and outcomes in COVID-19 patients with diabetic ketoacidosis: a systematic review of literature

Management of diabetic ketoacidosis: a summary of the

Joint British Diabetes Societies guidelines

High prevalence of COVID-19-associated diabetic ketoacidosis in UK secondary care. Diabetes research and clinical practice

Steroid-Induced Diabetes Ketoacidosis in an Immune Thrombocytopenia Patient: A Case Report and Literature Review

Glucocorticoid-induced hyperglycemia (glucocorticoid-induced hyperglycemia)

Glucocorticoid-induced hyperglycemia. Endocrine Practice

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

SARS-CoV-2 infections and ACE2: clinical outcomes linked with increased morbidity and mortality in individuals with diabetes

Binding of SARS coronavirus to its receptor damages islets and causes acute diabetes. Acta diabetologica

Factors associated with the presence of diabetic ketoacidosis at diagnosis of diabetes in children and young adults: a systematic review

Severe hypoglycemia and diabetic ketoacidosis in adults with type 1 diabetes: results from the T1D Exchange clinic registry

Inflammatory cytokine and plasma C-reactive protein response to ketoacidosis in adults with type 1 diabetes: Egyptian multicenter study

Cytokine response to diabetic ketoacidosis and its treatment

Complement as a target in COVID-19?

Protracted ketonaemia in hyperglycaemic emergencies in COVID-19: a retrospective case series. The Lancet Diabetes & Endocrinology

Diabetic ketoacidosis at diagnosis in Austrian children in 1989-2008: a populationbased analysis

Age-related differences in the frequency of ketoacidosis at diagnosis of type 1 diabetes in children and adolescents. Diabetes care

Interleukin-6 and C-reactive protein as early markers of sepsis in patients with diabetic ketoacidosis or hyperosmosis

HbA 1c as a Screening tool for Ketosis in Patients with Type 2 Diabetes Mellitus. Scientific reports

J o u r n a l P r e -p r o o f

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: