Diabetic ketoacidosis, respiratory distress and myocardial dysfunction

Abstract

This case report describes a patient with respiratory distress, myocardial dysfunction, elevated troponin level, ECG changes and pulmonary oedema secondary to new onset diabetic ketoacidosis (DKA). This case may be unique, but it is possible that less severe cases occur and are underappreciated in the paediatric setting. This report demonstrates the need to closely evaluate and monitor cardiac function in patients with DKA.

Background

We report the case of a patient with new onset diabetic ketoacidosis (DKA) presenting in respiratory distress with myocardial dysfunction, elevated troponin level, ST depression and prolonged QTc interval on electrocardiogram (ECG), and pulmonary oedema. This report demonstrates that cardiac dysfunction may be a complication of DKA and indicates the need for accurate cardiac monitoring in the evaluation and management of these patients.

Case presentation

A previously well 9-year-old child presented to her medical provider’s office following 4 days of progressive fatigue, fever, sore throat, headache and abdominal pain. She was alert but ill-appearing. Physical examination was significant for tachypnoea, laboured breathing, inspiratory and expiratory stridor, diffuse rhonchi, capillary refill >15 s and oxygen saturation 94% in room air. Diagnoses were airway obstruction, bacterial tracheitis and septic shock. Treatment was initiated with oxygen, albuterol, methylprednisone (60 mg IV), a 20 ml/kg fluid bolus of 0.9% sodium chloride solution (NSS) and transport to the local emergency department.

In the emergency department, the patient was in severe respiratory distress with stridor and staccato speech. Initial blood gas values were pH 6.86, HCO₃ 1.3 mmol/l and base deficit 32 mmol/l (table 1). Chest roentgenograph was normal. Bacterial tracheitis was suspected. She received racemic epinephrine, albuterol and ipratropium bromide nebuliser treatments, methylprednisolone (20 mg IV), cefuroxime (50 mg/kg) and a second 20 ml/kg bolus of NSS. She was taken to the operating room and upon direct visualisation noted to have mild erythaema of the trachea and mild subglottic swelling, but no exudates. She was intubated and transported to the Children’s Hospital at Dartmouth. During transport she received sodium bicarbonate (50 mEq IV) and 55 ml/kg of NSS for septic shock.

Table 1.

Selected laboratory values during the first 3 days of admission*

Time	pH	pCO2 (mm Hg)	HCO3 (mmol/l)	pO2 (mm Hg)	BE (mmol/l)	Na (mmol/l)	K (mmol/l)	Cl (mmol/l)	CO2 (mmol/l)	Ca (mg/dl)	Mg (mmol/l)	Phos (mg/dl)	Glucose (mg/dl)	Ketones (mmol/l)
Local emergency department	6.86	8	1.3	204	−32	130	3.5	98	<5	8.8			379
	6.73	30	3.8	578	−34
	6.66	45	5	213	−35
CHaD
Admission	6.87	30	6	223	−28	142	2.5	112	<5	6.3	0.67	3.8	329	8.43
Day 2	7.21	28	10.4	100	−17	140	2.8	110	10	7.9	0.56	1.9	386	2.83
	7.41	23	14.1	106	−10	141	2.2	114	13	7.4	0.59	3.2	67	0.34
Day 3	7.41	29	18	174	−7	148	2.9	118	17	7.3	0.8	4.9	110	0.3
	7.35	37	20	87	−6	144	3.1	114	18	7.4	0.7	4.2	200	<0.1

*Selected laboratory values for Children’s Hospital at Dartmouth (CHaD) admission are approximately every 12 h. Initial BUN, 14 mg/dl, creatinine 0.8 mg/dl. Normal pH 7.35–7.45; pCO₂ 35–45 mm Hg; HCO₃ 20–36 mmol/l; pO₂ 85–104 mm Hg; BE 0 mmol/l; Na 135–145 mmol/l; Cl 98–107 mmol/l; CO₂ 22–31 mmol/l; Ca 7.6–10.4 mg/dl; Mg 0.69–1.07 mmol/l; phos 2.8–5.6 mg/dl; glucose 60–99 mg/dl; serum ketones 0–0.3 mmol/l.

On arrival, the patient was intubated and sedated, pupils were reactive to light, lungs were clear bilaterally, there was no murmur or gallop, the abdomen was slightly distended with mild guarding and absent bowel sounds, and extremities were cool with a capillary refill >3 s. Her weight was 28.4 kg. Differential diagnoses included acute abdomen, toxic ingestion, sepsis, meningitis, new onset DKA and congenital metabolic disorder. Bacterial tracheitis was excluded due to findings on direct visualisation at the local hospital. Blood, urine and CSF were cultured. Abdominal roentgenograph revealed mildly distended loops of the small bowel and colon, with no free air or obstruction. Upon further questioning, her parents stated she had increased thirst and frequent urination for 2 weeks. New onset DKA was suspected and confirmed when laboratory tests revealed serum glucose 329 mg/dl (normal, 60–99 mg/dl), serum ketones 8.4 mmol/l (normal, 0–0.3 mmol/l), urine glucose >1000 mg/dl (normal, negative), urine ketones >80 mg/dl (normal, negative) and haemoglobin A₁C of 16.2% (normal, 4.3–6.1%) (table 1). Treatment for DKA with continued haemodynamic instability included a 20 ml/kg NSS bolus, 150% maintenance IV fluids of NSS, sodium bicarbonate (1 mEq/kg IV slowly, given twice), KCl and insulin.

Approximately 6 h after admission, endotracheal secretions were frothy pink. Chest roentgenograph revealed bilateral pulmonary oedema (figure 1). ECG showed sinus tachycardia with non-specific ST abnormalities and a prolonged QTc interval (489 ms) (figure 2A). Troponin level was elevated at 0.12 ng/ml (normal, <0.03 ng/ml) and CPK level was normal at 134 U/l (normal, 0–250 U/l). At this time, the patient had received a total of 4369 ml of fluid (156 ml/kg; rate of 11.7 l/m²/24 h) over 9 h. There was no suspicion of cerebral oedema by clinical signs and magnetic resonance imaging of the brain was negative.

Figure 1.

Portable chest x-ray showing bilateral pulmonary oedema obtained after frothy tracheal secretions were noted 6 h after admission.

Figure 2.

(A) ECG showing sinus tachycardia, non-specific ST abnormalities and prolonged QTc interval. (B) ECG showing ST depression in the inferior and lateral leads and non-specific T wave abnormalities.

Administration of furosemide resulted in significant diuresis but no change in haemodynamics or chest roentgenograph. Repeat ECG showed ST depression in the inferior and lateral leads and non-specific T wave abnormality (figure 2B). Echocardiogram at 28 h after admission showed global right and left ventricular dysfunction with hypokinetic systolic contraction of the base and mid-section of the left ventricle. The areas of dysfunction did not correlate with distribution of the coronary arteries. Both atria and ventricles were normal size. Vital signs at the time of the echocardiogram were heart rate 96 bpm, blood pressure 111/52 mm Hg, central venous pressure 6 cm H₂O, respirations 25 bpm (mechanical), and temperature 37.9°C. The patient received dopamine at 5 μg/kg/min and multiple infusions of KCl, calcium gluconate, magnesium sulfate, and potassium phosphate over the next 48 h. Toxicology screens and blood, urine and CSF cultures were negative. Throat culture obtained at the local hospital was positive for Haemophilus influenza, β-lactamase negative. Cefuroxime was continued for a 10-day course. On the third hospital day the patient was extubated.

Outcome

ECG was normal 6 days after admission and echocardiogram was normal 7 days after admission. Repeat troponin level was normal (0.04 ng/ml). The patient was discharged home on insulin 7 days after admission.

Our institution does not require Institutional Review Board approval for case reports.

Discussion

Children are frequently admitted to the paediatric intensive care unit secondary to DKA. Most children do well with treatment protocols designed to provide fluid resuscitation, electrolyte replacement, and insulin infusion rates that do not result in rapid changes in serum osmolality. Septic shock and/or myocardial dysfunction are not disorders commonly associated with DKA and pulmonary oedema is rarely reported. The combination of DKA and presumed septic shock and myocardial dysfunction poses a challenge in terms of fluid administration. Children with DKA or septic shock may require a degree of fluid resuscitation which may not be tolerated in the presence of myocardial dysfunction.

We know of only one report of myocardial dysfunction in a paediatric patient with DKA. This patient had severe hyperosmolarity and experienced an acute myocardial infarction likely secondary to alterations in regional blood flow and a hypercoagulable state.¹ Our report differs in that our patient was not severely hyperosmolar and the pattern of myocardial dysfunction was not in the likely distribution of the coronary arteries.

Myocardial dysfunction may occur secondary to cardiac, pulmonary, metabolic or haemodynamic abnormalities. Severe acidosis, hypophosphataemia, hypokalaemia, hypocalcaemia and hypomagnesaemia have been shown to contribute to poor myocardial performance and/or arrhythmias.² Treatment with β agonists or sodium bicarbonate may worsen ventricular function by decreasing ventricular filling time (β agonists) or reducing ionised calcium in the serum and inducing paradoxical intracellular acidification, especially in myocardial cells, resulting in decreased left ventricle function (sodium bicarbonate).² However, despite the frequency of electrolyte abnormalities and dehydration, myocardial dysfunction is not often diagnosed in patients with DKA.

Our patient experienced respiratory distress, myocardial dysfunction, elevated troponin level, ECG changes and pulmonary oedema which did not respond to administration of furosemide. While pulmonary oedema could be due to iatrogenic fluid overload, the patient’s haemodynamic instability at initial presentation warranted treatment with sufficient fluid resuscitation to preserve infusion of vital organs including the brain. Pulmonary oedema has been reported in both paediatric³ and adult patients with DKA, but is thought to be due to a generalised inflammatory state and not fluid overload. In addition, fluid overload would not cause ECG and troponin abnormalities. While sufficient fluid administration is recommended for patients in septic shock,⁴ fluid restriction is recommended for DKA, a difficult situation in a patient who may have both conditions. Fluid restriction in the treatment of DKA is recommended to prevent cerebral, not pulmonary, oedema.⁵

Sepsis has also been shown to have adverse effects on the myocardium.⁴ Our patient had a positive throat culture, but blood, urine and CSF cultures were negative. However, her initial presentation was consistent with septic shock, and culture negative sepsis remains likely. While viral myocarditis is also a possibility, we would not expect her to recover so quickly and completely without more specific therapy.

Although this case of DKA associated with respiratory distress, myocardial dysfunction and pulmonary oedema may be unique, it is possible that less severe cases occur and these may be underappreciated in the paediatric setting. We believe cardiac function needs to be closely evaluated and monitored in patients with DKA.

Learning points

• Patients with diabetic ketoacidosis (DKA) may have multiple morbidities.

• Myocardial dysfunction may occur in paediatric patients in DKA.

• Fluid management for DKA may conflict with fluid management for possible sepsis.