A clinical alert: Oral glucose tolerance test triggering diabetic ketoacidosis in pregnancy

Abstract

Gestational diabetes mellitus is a common medical disorder of pregnancy. Diabetic ketoacidosis is a complication that may affect both maternal and perinatal wellbeing adversely. It is rare, most often involving women with type 1 or type 2 diabetes, but occasionally can be seen in gestational diabetes mellitus. Here are two cases of ketoacidosis seemingly triggered by glucose ingestion for the oral glucose tolerance test in previously normoglycemic women, posing a diagnostic and therapeutic challenge. Prevention of such complications must be considered when treating high-risk pregnant women> 40 years of age, pregnant as a result of assisted reproductive techniques. Fasting blood glucose checked before ingestion of the glucose in a selected group of women may be one way of avoiding this complication. This suggestion may put women at risk of prolonged fasting and stretching services. Glucose tolerance test is a diagnostic test, and these cases demonstrate a rare complication.

Introduction

Diabetic ketoacidosis (DKA) is a rare complication of diabetes. It occurs in 0.5–3% of pregnancies complicated by diabetes. ¹ A single episode of DKA is related to maternal mortality in nearly 1% and fetal loss in 9–35% of the cases.^1,2 Prompt recognition and treatment may improve the maternal and fetal prognosis—the risk of developing DKA increases in later pregnancy. Potential precipitants in pregnancy include infections (e.g. urinary tract), hyperemesis, dehydration, β₂ mimetic drugs, corticosteroids, non-compliance with treatment, and insulin pump failure. New-onset diabetes accounts for 30% of the cases of DKA in pregnancy. ¹

The 75 g oral glucose tolerance test (OGTT) is the gold standard diagnostic test used worldwide. There have been no case reports of women developing ketoacidosis following a diagnostic test.

Here we describe two women with normal serum random blood glucose (RBS) and glycosylated hemoglobin (HbA1c) values in the first trimester who presented with DKA following OGTT in the second trimester.

Glucose tolerance test (GTT) normal value used according to IADPSG. ³

OGTT—fasting 92 mg/dl (≥ 5.2 mmol/l) or 1-h ≥ 180 mg/dl (≥ 10 mmol/l) or 2-h ≥ 153mg/dl (≥ 8.5 mmol/l).

Case 1

A 35-year-old woman in her first pregnancy, conceived with in-vitro fertilization (IVF) following eight years of infertility. Her blood glucose before the stimulation protocol was normal. She had a family history of diabetes mellitus. Her pre-pregnancy body mass index was 21.9 kg/m². In the first-trimester RBS was 111 mg/dl (6.1 mmol/l) and HbA1c 5.2% (both normal). OGTT at 25 weeks showed fasting of 203 mg/dl (11.5 mmol/l), 1st-hour value 335 mg/dl (18.6 mmol/l) and 2nd-hour value of 338 mg/dl (18.7 mmol/l). She was admitted to the high dependency unit (HDU) with giddiness and excessive thirst for further evaluation and glucose control.

On admission, her urinary ketones were 3+, RBS—360 mg/dl (20 mmol/L), serum pH was 7.2, blood pressure was 120/84 mmHg, pulse rate 106 beats/min, and she was afebrile. The anion gap was 24, and bicarbonate was 14.5 mmol/l, which was consistent with metabolic acidosis. Ketones were elevated, potassium was 4.2 mmol/l, and base excess was −10.3. Fetal status was reassuring. She was diagnosed with DKA.

Management

She was kept nil by mouth as she had vomiting and started on intravenous sodium chloride (10 ml/kg/h) (hourly blood glucose monitoring). Since her RBS on admission was 251 mg/dl (13.9 mmol/l), she was given regular intravenous insulin of 12 units, and 1 h RBS was 247 mg/dl (13.7 mmol/l) additional four units of regular insulin were added. Once she was stabilized, post-meal blood glucose was taken 2 h after the meal. The target value of fasting blood glucose was 95–100 mg/dl (5.1–5.5 mmol/l), and 2 h post-meal, blood glucose was 110–120 mg/dl (6.1–6.6 mmol/l).

She was discharged on day seven after euglycemia. The growth scan showed sufficient growth in the fetus. At 32 weeks of gestation, she had preterm prelabor rupture of membranes and was delivered by caesarean section. On postoperative day three, fasting (83 mg/dl or 4.6 mmol/l) and post-prandial blood glucose (120 mg/dl or 6.6 mmol/l) were normal on treatment. Insulin was withdrawn during the postpartum period as the blood glucose was normalized. She was discharged with metformin 500 mg once a day. At six weeks, her post-prandial glucose was 220 mg/dl (12.2 mmol/l) and fasting 110 mg/dl (6.1 mmol/l). The glycosylated hemoglobin was 6.9%. She was diagnosed with type 2 diabetes. Metformin was increased to 500 mg three times daily.

Case 2

A 37-year-old woman was pregnant for the first time with MCDA twins resulting from IVF. Her blood glucose before the stimulation protocol was normal. RBS was 140 mg/dl (7.7 mmol/l), and HbA1c was 4.5%. She had no family history of diabetes mellitus. OGTT with 75 g glucose load, done at 26 weeks, was—fasting 292 mg/dl (16.2 mmol/l), 1st-hour value of 511 mg/dl (28.3 mmol/l), and 2nd-hour value of 555 mg/dl (30.8 mmol/l). Soon after the glucose intake, the patient complained of giddiness, increased thirst, and increased frequency of micturition. She was admitted to the HDU. Her investigations showed RBS 320 mg/dl (17.7 mmol/l), urine ketones 4 +  and serum pH 7.1, bicarbonate 14.4 along with an anion gap of 22, potassium was 4.4 mmol/L, and base excess was 10.9, and she was diagnosed with DKA. Her fetal status was reassuring during her admission.

Treatment of DKA

She was treated with sodium chloride with 20 mmol/l potassium at 100 ml/h. Intravenous insulin (Actrapid) 6 ml/h was started. Insulin was titrated; accordingly, electrolytes and bicarbonates were monitored.

Once the ketosis was corrected, she was started on oral nutrition. The blood glucose level returned to normal. Fetal status was reassuring. She was discharged on subcutaneous Actrapid insulin (28 units with each meal) and Basalog (long-acting insulin), 20 units with her evening meal.

At 37 weeks, her twins were delivered by caesarean section.

Fasting 87 mg/dl (4.8 mmol/l) and post-prandial blood glucose 118 mg/dl (6.5 mmol/l) were noted postoperative day 3. The patient was discharged with metformin 500 mg three times daily, and at six weeks, her blood glucose was 113 mg/dl (6.2 mmol/l) fasting and 220 mg/dl (12.2 mmol/l) after a meal. Insulin was started in addition to metformin, and type 2 diabetes was diagnosed.

Type 2 diabetes has been assumed in both women, but further tests for type 1 diabetes have not been performed.

Discussion

DKA is seen in type 1 diabetes and can occur in type 2 diabetes, too; however also seen in women with gestational diabetes with the use of tocolytics and steroids. ¹ Our patients were normoglycemic before conception and in the first trimester, as both RBS and glycosylated hemoglobin were normal and reliable. Himuro et al. reported two cases of DKA with its onset in the third trimester in women who were previously normoglycemic. ⁴ On the day of the OGTT, the fasting blood glucose value in both patients exceeded 200 mg/dl (11.11), which was not checked before giving the glucose load of 75 g. However, these patients did not have any signs of infection either clinically or by laboratory parameters, suggesting diabetes to be the reason for elevated fasting blood glucose value. Diabetic pregnant women after 20 weeks of gestation are prone to ketoacidosis at lower blood glucose level. ¹

Both patients had elevated glycosylated hemoglobin, which was done at the diagnosis of DKA, suggesting undiagnosed diabetes before GTT. This indicates that they had developed diabetes much earlier than we routinely screen. We routinely do glycosylated hemoglobin in the first trimester. We can consider repeating it at the beginning of the second trimester, especially for women at risk of diabetes, like women > 35 years like both our patients with a family history of diabetes as in one of our patients, had it. However, this could pose a large number of women having tests repeated, increasing the cost and the number of visits considering the rarity. In patients with elevated glycosylated hemoglobin, GTT should not be done. All women at risk of diabetes should be warned of the symptoms of dizziness, increased thirst, and weakness following the glucose intake during the GTT and report to the hospital. It has been observed that IVF increases the risk of gestational diabetes mellitus (GDM).^3,5 Hence, it would be wise to screen for GDM early in the second trimester in IVF patients.

It is, therefore, an important learning point to check the fasting blood glucose value before the glucose load. The routine practice in our hospital and the hospitals around our locality is that the fasting and the post-prandial blood samples are analyzed together for GTT. Since there is no increased visit to the patient nor burden of the cost, probably worth waiting for the fasting blood glucose value before the glucose load in a selected number of patients. However, this could also increase the risk associated with prolonged fasting. Hence, if it were possible to identify a higher-risk group of women, these women might benefit from waiting for the fasting blood glucose value before ingestion of glucose.

Any degree of glucose intolerance with onset or recognized first during pregnancy is GDM. This definition applies to whether insulin therapy or only diet alteration is used for treatment and whether or not the high blood glucose prevails after pregnancy.

Conclusion

DKA may occur in women who are euglycemic before and during pregnancy. A glucose load given during the OGTT may predispose to this complication. Checking fasting hyperglycemia on the day of the test before the glucose load may be helpful; repeating the glycosylated hemoglobin in the early second trimester may also be beneficial in early recognition of GDM and reducing the need for further OGTT, thus avoiding the potential complication of DKA, especially in patients with a high risk of GDM.