Clinical, biological and radiological characteristics of diabetic patients hospitalized for COVID-19 in a Belgian tertiary care center

Orioli Laura; Servais Thomas; P. Hermans Michel; Thissen Jean-Paul; Vandeleene Bernard; Yombi Jean-Cyr; Maiter Dominique

doi:10.1530/endoabs.71.021

Introduction: In Belgium, 21.4% of patients hospitalized with COVID-19 are diabetic¹. In this study, we describe the characteristics of diabetic patients hospitalized for COVID-19 in a Belgian tertiary care center.

Patients and methods: The Clinics Ethics Committee approved the systematic registration of anonymized data from hospitalized patients with confirmed COVID-19 (N° CEHF 2020/22MAI/290), i.e. a positive SARS-CoV-2 PCR test on nasopharyngeal swab and/or computed tomography (CT) consistent with SARS-CoV-2 pneumonia in symptomatic patients. We retrospectively reviewed demographics, clinical, biological, and radiological data on admission from patients with known or newly-diagnosed diabetes. Survivors were compared to non-survivors to study prognostic factors of SARS-CoV-2 infection severity in diabetic patients.

Results: Fifty-six diabetic patients were identified among a total of 336 patients hospitalized with COVID-19 between March 11, 2020 and April 23, 2020, i.e. a 16.7% prevalence. Fifty-five had positive SARS-CoV-2 PCR (98.2%). Patient’s characteristics are summarized in Table 1. None had type 1 diabetes. Most patients had comorbidities (93%), half of them having ≥ 3. The most frequently-used glucose-lowering and blood pressure-lowering agents were: metformin (71.5%), insulin (37.5%), sulfonylurea/glinides (25%), diuretics (44.5%), angiotensin-converting enzyme inhibitors (ACEI), and angiotensin II receptor blockers (ARBs) (64.5%). DDP4 inhibitors were used in 5.5% of patients. Fever (73%), cough (80%) and dyspnea (65.5%) were the commonest symptoms on admission. Treatment consisted of azithromycin (16.5%) and/or antibiotics (23%) and/or hydroxychloroquine (89%), and/or glucocorticoids (13%). Non-invasive ventilation was required in 18 patients (32%). Twelve patients (21.4%) were admitted to an ICU, 7 (58.0%) of them requiring invasive mechanical ventilation. The overall case-fatality rate was 17.9%, as 10 patients died from COVID-19. Compared to survivors, non-survivors were older (P=0.021), had more severe pneumonia on the basis of lung surface area involvement on CT (P=0.025), and lower cycle threshold (Ct) of PCR test (P=0.024), the latter used as a proxy for viral shedding. Non-survivors were also less often treated with metformin (P=0.024) but required more often antibiotics (P=0.041) for suspected or proven secondary lung infections (P=0.037). No statistically significant difference was found regarding ethnicity, HbA1c, diabetes duration, body mass index (BMI), chronic diabetes-related complications, COVID-19-related symptoms, biological values, and other therapies. Limitations of this study include its retrospective design, small population size, and missing data regarding diabetes duration and HbA1c.

Table 1 Clinical characteristics and COVID-19-related biological values and radiological data on admissionaccording to vital outcome
	Number of	All	Survivors	Non-survivors	P value
Variables	Patients with	(N=56)	(N=45)	(N=10)
	available data
Age (years)	56	66±13	64±12	74±16	0.021
Female sex	56	26 (46.0)	22 (48.0)	4 (40.0)	0.737
BMI (kg/m²)	55	30.5±6.0	30.0±5.0	31.0±8.0	0.633
Stratification of BMI	55				0.101
<25 kg/m²		9/55 (16.5)	6/46 (13.0)	3/9 (33.0)
25–29 kg/m²		17/55 (31.0)	16/46	(35.0)	1/9 (11.0)
30–34 kg/m²		19/55 (34.5)	16/46	(35.0)	3/9 (33.0)
35–39 kg/m²		6/55 (11.0)	6/46 (13.0)	0/9 (0.0)
≥40 kg/m²		4/55 (7.0)	2/46	(4.0)	2/9 (22.0)
Diabetes type	56				0.551
Type 2		51/56 (91.0)	41/46	(89.0)	10/10 (100.0)
Secondary		3/56 (5.5)	3/46	(6.5)	0/10 (0.0)
Newly-diagnosed^a		2/56 (3.5)	2/46	(4.5)	0/10 (0.0)
Diabetes Duration (years)	49	9 [0–30]	9 [0–30]	9 [0–16]	0.743
HbA1c (%)	46	7.0 [4.1–11.7]	7.0 [5.4–11.7]	6.7 [4.1–9.4]	0.368
Dyslipidemia	56	43/56 (77.0)	34/46 (74.0)	9/10 (90.0)	0.424
Current Smoking	52	2/52 (4.0)	1/43 (2.5)	1/9 (11.0)	0.319
Chronic glucose -related	54	30/54 (55.5)	25/46 (54.5)	5/8 (62.5)	0.720
vascular complications^b
Ground glass	45	44/45 (98.0)	37/38 (97.5)	7/7 (100.0)	1.000
opacities/crazy paving
Chest Imaging severity	41				0.025
Mild to moderate (<25%)		20/41 (49.0)	20/35	(57.0)	0/6 (0.0)
Extensive (25–50%)		16/41 (39.0)	12/35	(34.5)	4/6 (67.0)
Severe to critical (>50%)		5/41 (12.0)	3/35 (8.5)	2/6 (33.0)
Ct SARS-CoV-2 RT-PCR^c	53	31 [18–39]	32 [19–39]	23 [18–38]	0.024
Laboratory values^d
Plasma glucose (mg/dl)	54	167 [37–349]	181 [37–349]	133 [87–288]	0.266
C-reactive protein (mg/l)	56	74 [2–413]	74 [2–413]	85 [18–263]	0.881
GFR (ml/min/1.73 m²)^e	55	68 [12–110]	71 [16–110]	58 [12–104]	0.647
Lymphocytes (10³/mm³)	56	0.9 [0.1–3.1]	0.9 [0.1–2.1]	0.9 [0.7–3.1]	0.369
Neutrophils/lymphocytes	56	5.2 [0.7–25.3]	5.5 [0.7–25.3]	3.0 [1.2–17.9]	0.164
ratio
Eosinophils (10³/mm³)	56	0.0 [0.0–0.9]	0.0 [0.0–0.9]	0.0 [0.0–0.1]	0.346
Data are expressed as means (±S.D.), medians [min-max], and numbers (%). Differences between groups were assessed by Student’s t-Test/Mann-Whitney U Test, or Chi-Square test/Fischer’s exact test according to distribution. Abbreviations: BMI, body mass index; HbA1c, glycated hemoglobin A1c; Ct, cycle threshold, GFR, glomerular filtration rate^aHbA1c ≥6.5% on admission^bIncluded retinopathy, nephropathy, neuropathy, foot ulcer, ischemic heart disease, stroke/transient ischemic attack, peripheral arterial disease^cRS-CoV-2 RNA in nasopharyngeal swabs was detected using COVID-19 genesig® RT-PCR assay (Primerdesign Ltd, Chandler’s Ford, United Kingdom) in a LightCycler 480 instrument (Roche Diagnostics, Mannheim, Germany). Probes and primers target RNA-dependent RNA polymerase (RdRp) gene. A cycle threshold <40 was considered positive.^dAlso included platelet count, lactate dehydrogenase (LDH), creatine kinase (CK), aspartate-aminotransferase (AST), alanine-aminotransferase (ALT); non-significant, P>0.05.^eCalculated with CKD-EPI formula.

Conclusion: In this monocentric cohort, we showed that diabetic patients hospitalized in Belgium for COVID-19 had mostly type 2 diabetes, and chronic hyperglycaemia-related complications. Moreover, half of them were obese. Age, but not BMI nor chronic glycaemic control, adversely influenced mortality. Although these findings are comparable to data from the multicentric CORONADO study², they need confirmation in larger series from other Belgian centers.

References: 1. Sciensano, Coronavirus Covid-19, https://covid-19.sciensano.be/fr, [accessed 6 August 2020].

2. Cariou B, Hadjadj S, Wargny M et al. Phenotypic characteristics and prognosis of inpatients with COVID-19 and diabetes: the CORONADO study, https://diabetologia-journal.org/wp-content/uploads/2020/05/20-0610-Cariou-in-press.pdf

Endocrine Abstracts

021