Clinical characteristics and outcome in patients with combined diabetic ketoacidosis and hyperosmolar hyperglycemic state associated with COVID-19: A retrospective, hospital-based observational case series.

AIM: One of the risk factors for poor outcome with SARS-CoV-2 infection is diabetes mellitus; diabetic ketoacidosis (DKA) and hyperosmolar hyperglycemic state (HHS) are the most serious complications of diabetes mellitus. We aimed to explore the clinical characteristics and outcomes of COVID-19 patients presenting with combined DKA/HHS to our institution.
METHODS: A retrospective, hospital based observation case series was performed on patients with SARS-CoV-2 admitted to Intensive Care Unit between 3/20/2020 and 4/20/2020. Inclusion criteria were: (1) Blood Glucose >250 mg/dL; (2) Serum bicarbonate <18 mmol/L; (3) Anion Gap >10; (4) serum pH <7.3; (5) ketonemia or ketonuria; (6) effective/calculated plasma osmolality >304 mOsm/kg and (7) positive SARS-CoV-2 RT-PCR.
RESULTS: We reported 6 patients who presented during this period with combined DKA/HHS. Their median age was 50 years, all males, three Hispanic, and three African American. Hispanic patients, had more severe acidosis, and multiple comorbidities, with a higher mortality. The striking feature was that combined DKA/HHS was the initial presentation for COVID-19 for most of the cases. DISCUSSION: Our observational retrospective case series shows that diabetic patients are at risk of developing combined DKA/ HHS associated with COVID-19 and a substantial mortality. To our knowledge, we are first to report the clinical characteristics and outcome in this group of patients.

Introduction

A novel enveloped RNA beta-coronavirus was discovered in December 2019 and was found to belong to the same family as severe acute respiratory syndrome coronavirus (SARS-CoV), hence was called SARS-CoV-2 [1]. It is responsible for the current global pandemic with escalating cases and fatalities worldwide with over three million cases reported as of May 5th, 2020 [2]. According to the Chinese Center for Disease Control and Prevention (CDC), only 5 percent developed critical disease with multiple organ failure [3].

To our knowledge, combined DKA and HHS has not been reported in association with COVID-19 as of May 2020, despite the fact that DM is a well-known predisposing factor leading to severe COVID-19 disease [4]. In this retrospective, hospital based cohort study, we are reporting 6 patients who were admitted to the intensive care unit (ICU) for combined DKA and HHS with positive SARS-CoV-2 that had no other predisposing causes.

Methodology

A retrospective analysis was done on patients admitted to our intensive care unit between March 20th and April 20th, 2020. Inclusion criteria were: 1) Blood Glucose > 250 mg/dL; 2) Serum bicarbonate ≤ 18 mmol/L; 3) Anion Gap > 10; 4) pH < 7.3; 4) ketonemia or ketonuria; 5) effective/calculated plasma osmolality > 304 mOsm/kg. and 6) positive SARS-CoV-2 RT-PCR. The only exclusion criteria was hyperglycemia without features of DKA/HHS.

Case definition of hyperglycemic crises - combined DKA and HSS on admission was: fulfilling the criteria above with pH < 7.3, bicarbonate < 18 mmol/L, Anion Gap > 10 and ketonemia with serum osmolality > 304 mOsm/kg.

Demographic analysis was performed, and data was expressed as counts, percentages or median and inter-quartile range (IQR). Cases were presented including history, hospital course, and outcome. SARS-CoV-2 RT-PCR was positive for all cases with elevated inflammatory markers (Table 1 ). All cases were managed with DKA/HHS protocol with aggressive hydration, insulin drip and electrolytes repletion. A waiver of HIPAA privacy authorization have been obtained through the institutional review board.

Table 1

Demographic characteristic, inflammatory markers and outcome.

Cases	Age	Sex	Ethnicity	BMI	HbA1c	Admitting BG	pH	Bicarb	Anion gap	BHOB	Calculated serum osmolarity	CRP	LDH	D-dimer	Ferritin	Outcome
1	48	M	AA	25.2	13.6	1130	7.2	9.1	33	>4.5	365	8.3	263	4120	909	Stable
2	52	M	AA	23.6	12.1	683	7.24	11	21	>4.5	318	11.2	691	1220	1650	Stable
3	19	M	AA	58.9	10.9	1112	7.1	8	32	>4.5	307	8.1	695	1234	1628	Expired
4	38	M	Hispanic	24.2	N/A	1070	6.77	<5	>29	>4.5	347	2.6	275	834	807	Expired
5	62	M	Hispanic	24.1	13.3	604	7	6	29	N/A	323	13.2	397	1060	6671	Expired
6	62	M	Hispanic	30.5	N/A	959	6.86	6.6	>29	>4.5	358	N/A	N/A	N/A	3400	Expired
Average	46.8			31.1	12.5	926.3	7.03	8.1	28.8		336.3	8.7	464.2	1693.6	2510.8
Median	50			24.7	12.7	1014.5	7.05	8.0	30.5		335.0	8.3	397.0	1220	1639

M (Male), BMI (Body Mass Index), HbA1c (Hemoglobin A1c, normal range 4–5.6%), BG (blood glucose, normal range 70–140 mg/dL), pH (normal range 7.34–7.44), Bicarb (Bicarbonate, normal range 20–31 mmol/L), CRP (C-reactive protein, normal range 0–0.8 mg/dL), LDH (lactate dehydrogenase, normal range 122–222 U/L), D-dimer (0–500 ng/mL), Ferritin (24–336 ng/mL), Anion gap (5–15 mmol/L), BHOB (beta-hydroxybutyrate, normal range 0.02–027 mmol/L), Serum osmolality (278–305 mOsm/kg), N/A (not available).

Case presentation

Case 1: A 48-year-old male with past medical history of type 2 diabetes mellitus (T2DM) presented to the emergency department (ED) with abdominal pain, nausea, vomiting and confusion. The patient was alert and awake but not oriented to time, place and person. Initial vital signs showed temperature 98.5 °F, heart rate (HR) 130/min, respiratory rate (RR) 12/min, blood pressure (BP) 110/68 mmHg, and oxygen saturation 99% on room air. Physical exam was notable for dry mucous membranes and cool extremities. Basic metabolic panel (BMP) and arterial blood gas (ABG) indicated combined DKA and HHS (Table 1). Electrolytes were within normal limit. Complete blood counts (CBC) showed Hgb 16.3 g/uL (13.5–16.5 g/dL), white blood count (WBC) 18,000/uL (4000–11,000/uL), with absolute neutrophil count (ANC) 15,600/uL (1700–7000/uL) and absolute lymphocyte count (ALC) 500/uL (900–2900/uL). Chest X-ray (CXR) was normal. Electrocardiogram (ECG) showed normal sinus rhythm. Computed tomography (CT) head was within normal limits. He was admitted to intensive care unit (ICU) and was started on DKA/HHS protocol, hydroxychloroquine and doxycycline along with therapeutic anticoagulation. He recovered and was discharged home.

Case 2: A 52-year-old male with past medical history of T2DM, hypertension and coronary artery disease, presented to the ED complaining of generalized weakness, polyuria and polydipsia of two days duration. He claimed compliance with his medications. He denied any nausea, vomiting, cough, shortness of breath, fever, or chest pain. Initial vital signs showed temperature 97.9 °F, HR 104/min, RR 17/min, BP 121/54 mmHg, and oxygen saturation 98% on room air. Physical examination was within normal limits. BMP and ABG indicated combined DKA and HHS (Table 1). Electrolyte panel showed Na 130 mmol/L (135–145 mmol/L) and K 7.2 mmol/L (3.5–4.5 mmol/L). CBC were within normal limits. Urine drug screen (UDS) was positive for cocaine. CXR showed diffuse patchy bilateral opacities (Fig. 1 a). ECG showed normal sinus rhythm (HR 96/min). He was treated with DKA/HHS protocol, one dose of ivermectin and 5 days course of doxycycline and metronidazole. He recovered and was discharged home.

Fig. 1

(a-e) Abnormal CXR on admission showing bilateral infiltrates consistent with COVID-19 disease.

Case 3: A 19-year-old male with no past medical history presented to ED complaining of abdominal pain, increased thirst, nausea and multiple episodes of vomiting. He denied any respiratory symptoms. No history of smoking or illicit drug use. Initial vital signs showed temperature 99.1 °F, HR 135/min, RR 45/min, BP 86/50 mmHg, and oxygen saturation 92% on room air. Physical examination was within normal limits except for morbid obesity (BMI 58.9 kg/m2). He was intubated in the ED due to altered mental status. BMP and ABG indicated combined DKA and HHS (Table 1). Electrolyte panel showed Na 117 mmol/L, K 5.7 mmol/L, and creatinine 3.2 mmol/L. CBC was significant for WBC 14,400/uL with ALC 900/uL. CXR showed low lung volumes with bilateral infiltrates (Fig. 1b). ECG showed atrial tachycardia. He was treated with DKA/HHS protocol, vasopressor and meropenem as well as doxycycline initially. He also received COVID-19 convalescent plasma. His renal function progressively deteriorated, ultimately requiring Continuous Veno-Venous Hemodialysis (CVVHD). His clinical status worsened with increasing oxygen requirements and persistent atrial flutter unresponsive to maximal therapy. He developed multi-organ failure and expired after 13 days of mechanical ventilation.

Case 4: A 38-year-old male with past medical history of T2DM was found to be unresponsive at home, cold extremities, and self-measurement blood glucose at home of more than 600 mg/dL. Initial vital signs showed temperature 88.9 °F, HR 59/min, RR 20/min, BP 80/26 mmHg, and oxygen saturation 90% on room air. On examination, he was lethargic with BMP and ABG indicated combined of DKA and HHS (Table 1). Electrolyte panel showed Na 123 mmol/L, K 4.4 mmol/L, UDS and ethanol levels were negative. CBC showed Hgb 9.1 g/dL, WBC 5500/uL with ANC 3700/uL and ALC 1000/uL. CXR showed bilateral infra-hilar infiltrates (Fig. 1c). He was started on DKA protocol, active rewarming, ceftriaxone and doxycycline, and vasopressor. He was also intubated for airway protection. He ultimately expired with cardiopulmonary arrest.

Case 5: A 62-year-old male patient with past medical history of T2DM presented to the ED complaining of cough and shortness of breath. On initial encounter, he was alert and awake but disoriented. Initial vital signs were significant for BP 156/83 mmHg, HR 120/min, RR 33/min, and oxygen saturation 95% on room air. Initial physical exam was remarkable for reduced breath sounds all over the lung fields. BMP and ABG indicated combined DKA and HHS (Table 1). Electrolytes were within normal limits. UDS was negative. CBC showed WBC 15,100/uL, with ANC 13,100/uL and ALC 1300/uL. ECG showed sinus tachycardia and CXR showed diffuse bilateral patchy opacities (Fig. 1d). Blood cultures were negative. The patient was admitted to ICU and was started on DKA/HHS protocol, ceftriaxone and doxycycline. His DKA resolved but his respiratory status worsened in the ICU and he required mechanical ventilation. Unfortunately, he expired after three days.

Case 6: A 62-year-old male with past medical history of T2DM, hypertension, coronary artery disease and heart failure was brought to the ED by an ambulance after he was found to be unresponsive for an unknown duration. He had a cardiac arrest on his way to the hospital, basic life support was performed for a short period of time. Upon arrival to the ED, he was bradycardic with heart rate 30–35/min, which was treated with atropine and intubated. His initial vital signs showed temperature 91.1 F, BP was 77/40 mmHg with HR 130/min and RR 33/min. He had sluggish pupillary reflexes, absent gag and corneal reflexes and was not responding to painful stimulus. BMP and ABG indicated combined DKA and HHS (Table 1). Electrolyte panel showed Na 142 mmol/L and K 5.7 mmol/L. Cardiac enzymes and ethanol level were negative. Lactic acid 7.6 mmol/L, WBC 15,000/uL and ALC 1200/uL. CXR showed diffuse bilateral lung opacities (Fig. 1e). Blood and urine cultures were negative. He was started on DKA/HHS protocol, vasopressor, active rewarming, and ceftriaxone and doxycycline. However, a few hours later he became bradycardic with persistent hypotension and expired.

Result

Between March 20th to April 20th, 2020, 6 patients were admitted to our intensive care unit with diabetic crisis associated with COVID-19. BMP and ABG on all the patients were consistent with combined DKA and HHS. All patients had laboratory values fulfilling DKA/HHS defining parameters as tabulated in Table 1. The median age for the patient was 50 years old (range from 19 to 62 years old). All patients are male. Of the 6 patients, 5 patients (83%) had a history of DM, one patient was newly diagnosed DM, presenting as DKA/HHS, presumptively precipitated by COVID-19. Only 2 patients (33%) had a BMI ≥ 30 kg/m2. As for ethnicity; three were Hispanic (50%), and three African American (50%).

The average hemoglobin A1c was 12.5% (ranged from 10.9 to 13.6%, no data for two patients). The average admitting plasma glucose was 926.3 mg/dL (ranged from 604 to 1130 mg/dL). The average serum osmolality was 336.3 mOsm/kg (ranged from 307 to 358 mOsm/kg). Though statistical analysis could not be done due to small number size, patients with Hispanic ethnicity, severe acidosis with multiple comorbidities tended to have higher mortality. Four of six (67%) of the patients expired. Nonetheless, the extent of inflammatory markers elevation does not seem to correlate with the mortality. Those who required mechanical ventilation had a worse prognosis. There were four patients who required mechanical ventilation, and all expired. Another notable finding from this case series is that, the majority of patients presented with DKA/HHS symptoms rather than respiratory symptoms, and SARS-CoV-2 was detected mainly because our institution in the epicentre of the pandemic and all admissions were screened for COVID-19.

Discussion

Infection is a well-documented precipitating factor for combined DKA and HHS in both new-onset and established diabetes. Patients with long-standing diabetes frequently have other co-morbidities including hypertension, cardiovascular disease, obesity, and chronic kidney disease. All of which have been established as risk factors for severe COVID-19 [4], [5]. A retrospective analysis conducted in China observed an increased risk of mortality in diabetic patients with COVID-19 infection. Better glucose control has been associated with improved outcomes in this group of patients as compared to those with poorly glucose control [4].

The mechanism of how SARS-CoV-2 could precipitate DKA is poorly understood. SARS-CoV-2 belong to the family of SARS coronavirus which has been shown to affect both exocrine and endocrine cells of pancreas [6]. Diabetes has been associated with poor outcomes in prior coronavirus epidemics like SARS-CoV [7] and Middle East Respiratory Syndrome (MERS) Coronavirus [8]. It has been postulated that diabetes predisposes to dysregulated immune response (with elevated Interleukin-17α) resulting in lung pathology [9].

It has been established that Angiotensin-Converting Enzyme 2 (ACE2) serves as a functional receptor for SARS-CoV-2 S-protein, similar to SARS-CoV [10]. ACE2 is known to be expressed in lungs, heart, kidneys, gastrointestinal system, and pancreas [10]. ACE2 receptors are also found in the exocrine and islet cells of the pancreas [11]. The canine pancreas was shown to have local angiotensin peptide generating system, with Angiotensin II predominant [12]. ACE2 typically degrades Angiotensin II to Angiotensin (1–7), one of the active heptapeptides in the renin angiotensin system. Angiotensin II can delay insulin secretion and decrease blood flow to islet-cells resulting in hyperglycemia, local inflammation, apoptosis and decreased proliferation of islet cells[13]. Angiotensin (1–7) counteracts this effect by increasing insulin secretion and vasodilatation. SARS-CoV-2 binding to ACE2 may result in downregulation of ACE2, leading to an unopposed action of Angiotensin II. SARS-CoV-2 could also have a direct effect on islet-cells, leading to acute islet-cell dysfunction, decreased insulin release and subsequent acute hyperglycemic crisis and even ketosis [11]. Nonetheless, there is still lack of data on the mechanism of SARS-CoV-2 effect on the pancreatic angiotensin system and islet-cells.

Studies have shown that patients with diabetes are at risk for severe COVID-19 [4]. In our patients, there was 67% mortality and 67% required mechanical ventilation. All ventilated patients ultimately expired. Hispanics, severe acidosis and multiple comorbidities tended to be associated with higher mortality, which is consistent with retrospective cohort analysis reported in China and United States [3], [4], [14], [15]. Age in particular, in our case series, did not appear to influence the mortality, however, the sample size is relatively small. Pasqual et al. reported that combined DKA and HHS has higher mortality as compared to isolated DKA and HHS alone [16]. In this hospital based retrospective study, we observed the same trend with 67% mortality in patients with combined DKA and HHS.

Managing DKA with or without HHS associated with COVID-19 poses a great challenge. Currently, there is no consensus guideline on how to manage this group of patients. Fluid management is a particular concern, in addition to maintaining adequate hydration to prevent acute renal failure, there is also a need to maintain negative fluid balance to prevent patient progressing into Acute Respiratory Distress Syndrome (ARDS), which is a major complication in patients with COVID-19. All our patients were managed with fluids hydration, insulin drip and electrolytes replacement. In addition, bicarbonate was given if the pH was ≤6.9. Kidney function and electrolytes were monitored to ensure adequate hydration while at the same time trying to prevent volume overload.

To our knowledge, this is the first retrospective, observational case series summarizing the clinical characteristics and outcome of combined DKA and HHS patients associated with COVID-19. A comprehensive cohort study is paramount to explore the degree of blood sugar control and its impact on the severity of infection in COVID-19 patients. In addition, more studies are needed to dissect the interplay between SARS-CoV-2 and pancreatic angiotensin system, in order for us to have a better understanding on the pathogenesis of SARS-CoV-2 induced hyperglycemia, and combined DKA and HHS.

Conclusion

Our observational retrospective case series shows that diabetic patients are at risk of developing combined DKA and HHS associated with COVID-19. Hispanic, poor blood glucose control, mechanical ventilation and severe acidosis seems to be the poor prognostic features in this group of patients. It is paramount to have a tight glucose control in both inpatient and outpatient setting, when managing diabetic patients. Further studies are needed to established the degree of hyperglycemia as the predictors for mortality and morbidity in COVID-19 patients. Moreover, we need to understand the pathogenesis of SARS-CoV-2 on the pancreatic angiotensin system, so as to monitor the effect of SARS-CoV-2 on the pancreas in long term.

Funding

The authors received no funding from an external source.

Declaration of Competing Interest

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.