Abstract
Usually, hyperglycaemia crisis presents with acidotic pH, but ketoalkalosis is a rare and unheard entity presenting in diabetic ketoacidosis. We describe three unique cases where the patients present with hyperglycaemia >250 mg/dL, normal or alkalotic pH, and bicarbonate >20 meq/L, which does not meet criteria for diabetic ketoacidosis. However, once these patients were supplemented with intravenous fluids, diagnosis of diabetic ketoacidosis was evident in laboratory analysis. These case series provide a learning opportunity in diagnosing and management of this rare phenomenon.
Keywords: diabetes, metabolic disorders, adult intensive care, fluid electrolyte and acid-base disturbances
Background
Diabetes mellitus (DM) is a disease with significant morbidity and mortality, which continues to burden the healthcare system. Many complications can arise from uncontrolled DM, including diabetic ketoacidosis (DKA), a medical emergency. DKA can result from poor compliance with an insulin regimen or a state of stress on the body that increases glucose demand creating a relative insulin deficiency. This state of stress gives rise to an increase in counter-regulatory hormones (growth hormone, epinephrine, cortisol and glucagon) progressing to uncontrolled hyperglycaemia. Insulin deficiency leads to an alternate source of energy, ultimately the breakdown of fatty acids and the creation of ketones. DKA is defined as blood glucose >250 mg/dL, blood pH <7.3, serum bicarbonate <18 meq/L and the presence of elevated anion gap metabolic acidosis. It typically manifests with an acidaemia, however, it could present as an alkalaemia in atypical presentation with diuretics, intractable vomiting and hyperaldosteronism.1 In these case series, we present three patients with hyperglycaemia, alkalotic pH and normal bicarbonate that were treated with volume expansion fluids, which unmasked their underlying DKA.
Case presentation
Case 1
A 25-year-old Hispanic female presenting with complaints of intractable vomiting, malaise and abdominal pain. She has a history of type 1 DM for last 10 years with no previous records of haemoglobin A1C and takes insulin glargine 25 units daily at home. Six hours prior to admission, she started to experience abdominal pain around the periumbilical area and subsequently developed intractable non-bilious vomiting. On initial evaluation, physical examination was positive for lethargy, dry oral mucosa, Kussmaul respirations and intense periumbilical tenderness. The patient was afebrile, blood pressure was 107/58 mm Hg, respiratory rate 26 breaths/minute, pulse 99 beats/minute and saturating 99% on ambient air, height 61 inches, weight 58.2 kg and body mass index (BMI) 24.24.
Case 2
A 24-year-old White man with a history of type 1 DM presented to the hospital for the 7-hour duration of abdominal pain and vomiting. He was diagnosed with type 1 DM about 9 years ago, for which he was prescribed NPH insulin 17 units daily based on his last haemoglobin A1C 9.8 about 3 months prior to admission. He had a significant history of non-adherence with his insulin regimen causing the precipitation of the symptoms. On examination, patient’s blood pressure was 124/86 mm Hg, heart rate 83 beats/minute, temperature 97.9 degrees Fahrenheit, respiratory rate 22 breaths/minute, pulse oximetry 98% on ambient air, height 71 inches, weight 68.04 kg and BMI 20.92. The physical examination was positive for epigastric tenderness, dry mucous membranes and decreased skin turgor.
Case 3
A 49-year-old African woman with a medical history of type I DM, major depression disorder and hypertension presented with symptoms of intractable vomiting, nausea and epigastric abdominal pain for the last 16 hours. In addition, she complained of fatigues for the day prior to presentation. She was diagnosed with type 1 DM about 8 years ago and currently prescribed intermediate-acting NPH (Novolog 70/30) insulin 45 units daily with last haemoglobin A1C recorded 8.6 (4 months prior to admission). On admission, her blood pressure was 189/98 mm Hg, respiratory rate 26 breaths/minute, saturating 100% on ambient air, heart rate 98 beats/minute, temperature 98 degrees Fahrenheit, height 67 inches, weight 77.4 kg and BMI 26.73. The physical examination was positive for dry oral mucosa, tenderness around the epigastric area and positive costovertebral angle tenderness.
Investigations
Case 1
Initial laboratory investigations revealed: blood glucose 534 mg/dL, serum sodium 133 meq/dL and large number of serum ketones. The corrected anion gap was 21.25 meq/L. Blood gas analysis showed pH 7.43, pCO228 mm Hg and bicarbonate 19 mmol/L. Urinalysis showed high ketones and urine glucose >500 mg/dL (table 1).
Table 1.
Biochemical profile | Admission | After intravenous fluids resuscitations |
Sodium (meq/L) | 133 | 130 |
Potassium (meq/L) | 5.1 | 3.3 |
Chloride (meq/L) | 93 | 102 |
Bicarbonate (meq/L) | 20 | 11 |
Serum glucose (mg/dL) | 534 | 263 |
Albumin (G/L) | 3.5 | 3.3 |
Phosphorus (meq/L) | – | 3.5 |
Magnesium (meq/L) | – | 1.6 |
Creatinine (mg/dL) | 0.94 | 0.83 |
Urine ketones | Moderate | – |
Blood gas pH (venous) | 7.43 | 7.27 |
pCO2 (mm Hg) | 28 | 19 |
pO2 (mm Hg) | 101 | 129 |
Arterial bicarbonate (mmol/L) | 17.7 | 9.1 |
Corrected anion gap | 21.25 | 18.75 |
Delta–Delta gap | 5 | −6.25 |
Case 2
Laboratory investigation revealed glucose of 494 mg/dL, positive ketones in blood and urine, and metabolic alkalosis on venous blood gas (table 2).
Table 2.
Biochemical profile | On admission | After intravenous fluid resuscitation |
Sodium (meq/L) | 130 | 140 |
Potassium (meq/L) | 3.6 | 4.6 |
Chloride (meq/L) | 88 | 105 |
Bicarbonate (meq/L) | 21 | 16 |
Serum glucose (meq/dL) | 695 | 337 |
Albumin (G/L) | 5.2 | 4.4 |
Phosphorus (meq/L) | 2.6 | 2.6 |
Magnesium (meq/L) | 2 | 1.9 |
Creatinine (mg/dL) | 1.26 | 1.05 |
Urine ketones | Moderate | – |
Blood gas pH (venous) | 7.45 | 7.32 |
pCO2 (mm Hg) | 28 | 29 |
pO2 (mm Hg) | 28 | 55 |
Bicarbonate (mmol/L) | 18.2 | 13.9 |
Corrected anion gap | 18 | 18 |
Delta–Delta gap | 3 | -2 |
Case 3
Initial laboratory investigation revealed positive ketones, hyperglycaemia with high anion gap, but pH on the venous gas analysis indicated the presence of metabolic alkalosis. Urine analysis confirmed positive ketones and urinary tract infection (UTI). The additional biochemical analyses are shown in table 3 along with follow-up laboratory values 5–6 hours after admission.
Table 3.
Biochemical profile | On admission | After intravenous fluid resuscitation |
Sodium (meq/L) | 136 | 136 |
Potassium (meq/L) | 4 | 4.9 |
Chloride (meq/L) | 97 | 100 |
Bicarbonate (meq/L) | 20 | 12 |
Serum glucose (meq/dL) | 408 | 387 |
Albumin (G/L) | 3.9 | 4 |
Phosphorus (meq/L) | – | – |
Magnesium (meq/L) | 1.3 | – |
Creatinine (mg/dL) | 0.64 | 0.68 |
Urine ketones | Small | – |
Blood gas pH (venous) | 7.58 | 7.29 |
pCO2 (mm Hg) | 25 | 30 |
pO2 (mm Hg) | 47 | 48 |
Bicarbonate (mmol/L) | 23.4 | 16.1 |
Corrected anion gap | 19 | 24 |
Delta–Delta gap | 3 | 0 |
Differential diagnosis
As all the patients presented with severe hyperglycaemia and anion gap, DKA was the top differential diagnosis suspected, though low-normal bicarbonate on the chemistry panel and metabolic alkalosis (on blood gas analysis) deterred and evaded to meet the diagnostic criteria for DKA. Reviewing the causes of anion gap, all the three patients had a negative screen for toxic alcohol, salicylates or lactic acid. There was no significant difference in the value of anion gap among the three patients. Anion gap with metabolic alkalosis was hinting about the volume depletion that may cause contraction alkalosis. Therefore, a hyperglycaemic crisis with metabolic alkalosis was a differential diagnosis.
Treatment
Case 1
The patient was treated with intravenous saline, intravenous insulin and replenishment of electrolytes. After receiving 2 L of intravenous saline, the patient was awake and alert and she was given 10 units of insulin levemir.
Case 2
Due to the patient’s severe hyperglycaemia, he was admitted to the medical intensive care unit (ICU) with an initiation of intravenous fluids, continuous intravenous insulin infusion and electrolytes supplementation.
Case 3
The patient was admitted to the ICU, treated with intravenous insulin, intravenous normal saline and electrolyte repletion. The patient was also started on antibiotics for UTI. Table 3 provides additional biochemical profile after resuscitation.
Outcome and follow-up
Case 1
Subsequent metabolic panel after 5 hours from admission showed: sodium 130 meq/dL, potassium 3.3 meq/dL, chloride 102 meq/dL, bicarbonate 11 meq/dL, glucose 263 mg/dL and subsequent anion gap improved to 18.75 meq/L. Blood gas analysis showed pH 7.27, pCO219 mm Hg and bicarbonate 9 mmol/L. Subsequently, after reviewing the laboratory values, the patient was admitted to the ICU and continued with intravenous fluids and insulin infusion. Within a day, the patient recovered from the hyperglycaemia crisis and was eventually discharged. Additional biochemical profile in table 1.
Case 2
After the fluid resuscitation, patient’s follow-up laboratory values conducted showed ketoacidosis with acidotic pH and decreased bicarbonate from admission. The patient fully recovered within 8 hours of admission, intravenous insulin and fluids were discontinued and transitioned to subcutaneous insulin regimen with diabetic education.
Case 3
Within the next 8–9 hours, the patient successfully recovered and started tolerating oral diet as the emesis resolved. She was started on subcutaneous insulin before intravenous insulin was discontinued, and the patient was eventually discharged home.
Discussion
Diabetic ketoalkalosis was first described by Bleicher in 1967.2 Since then there have been >30 cases of diabetic ketoalkalosis reported and continues to be a rare pathology for the clinicians to recognise.3 DKA can present with profuse vomiting, which leads to metabolic alkalosis due to hydrogen ion loss from the gastrointestinal tract and contraction alkalosis due to volume depletion. Volume depletion will result in activation of the renin–angiotensin–aldosterone system which stimulates renal tubules for sodium reabsorption, bicarbonate reabsorption and new bicarbonate generation, the latter two effects are accomplished by the secretion of hydrogen ions into the tubular lumen.4 Hyperaldosteronism will, in turn, induce hypokalaemia which contributes to the generation of metabolic alkalosis by shifting hydrogen ions from the extracellular fluid to the intracellular fluid in exchange for potassium and by increasing renal bicarbonate reabsorption.4 5 This multifactorial effect leads to alkalaemia instead of the typical acidaemia seen in DKA due to the ketoacids.
In our three patients diagnosed with DKA, aetiology of the metabolic alkalosis was postulated as intractable vomiting that snowballed acid balance chaos. Diffuse vomiting caused the volume contraction masking innate acidosis from excess hyperglycaemia. In a case series by Svart et al, six patients with diabetic ketoalkalosis were studied and all had symptoms of severe nausea and vomiting.6 Calculating the anion gap provides insight into suspicion of underlying DKA, masked with metabolic alkalosis, as all the three patients had evident anion gap.
Treatment of diabetic ketoalkalosis does not differ from the conventional management of DKA. These patients should be treated with aggressive intravenous fluids, electrolytes supplementation (especially potassium) and hourly intravenous insulin infusion. Administration of intravenous fluids remains of utmost priority with timely replenishment of electrolytes. In a case reported by Jerrard and Hanna, adding potassium to an intravenous saline solution aids in quickly reversing the alkalotic state.7
Due to limited literature for diabetic ketoalkalosis, there has been no reported risk factors or biochemical variables that directly associate with the presentation. We suspect that diabetic ketoalkalosis may be underdiagnosed because clinicians are unaware of this phenomenon. Therefore, these cases will provide awareness among clinicians and help us understand ketoalkalosis to a greater degree.
Patient’s perspective.
Patient 1: ‘My diabetes has been a real problem because I am in and out of the hospital non-stop. Already have tried many medications and as of now, I am admitted in the hospital for my diabetes. Its been a lot of effort for my family’. Patient’s daughter: ‘My mom has been through so much that its unreal and nonstop. We don’t even know what to do with diabetes and only thing left is to may be try another doctor’.
Patient 2: ‘Thank you for all the help and I was struggling when I came in. But you guys (ICU team) were amazing. Thanks once again’.
Patient 3: On requesting perspective during hospitalisation, she added, ‘I just thankful that you guys are treating and helping real well. I came in a bad shape, feeling worse and if it weren’t your treatment with all these (referring to IV), I wouldn’t be talking’.
Learning points.
Patients with type 1 diabetes may present with diabetic ketoacidosis (DKA) masked by alkalaemia.
It is theorised that diabetic ketoalkalosis develops secondary to the development of volume contraction, and the loss of chloride (Cl−) in the form of HCl via vomiting. This stimulates bicarbonate reabsorption and decreasing secretion in the collecting ducts of the kidneys.
Suspecting DKA in patients with type 1 diabetes, the clinician should calculate for triple disorder derangement, especially anion gap and the delta–delta ratio. In patients with diabetic ketoalkalosis, an underlying anion gap metabolic acidosis, respiratory alkalosis and metabolic alkalosis may be detected.
Treatment of diabetic ketoalkalosis is similar to DKA→ expansion of extracellular volume and intravenous insulin remains the mainstay treatment.
Correction of the electrolyte abnormalities, especially potassium, remains pivotal by adding to saline in a reversal of the alkalotic state.
Acknowledgments
The authors would like to thank Dr Chandran, Dr Sandra Gibiezaite and Dr Patrick Michael who provided their extensive support and guidance.
Footnotes
Contributors: SN, VK, GM: contributions in the manuscript for the introduction, case presentation and discussion. MS: mentor guiding in writing the case and discussion.
Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests: None declared.
Patient consent: Obtained.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
- 1.Kumar V, Nanavati SM, Komal F, et al. . Ketoalkalosis: Masked Presentation of Diabetic Ketoacidosis With Literature Review. J Endocrinol Metab 2017;7:194–6. doi:10.14740/jem472w [Google Scholar]
- 2.Bleicher S. Ketosis not always acidosis: ‘‘heartburn’’ can be relevant. Diabetes 1967;2:3–4. [Google Scholar]
- 3.Huggins EA, Chillag SA, Rizvi AA, et al. . Diabetic ketoalkalosis in children and adults. South Med J 2014;107:6–10. 10.1097/SMJ.0000000000000040 [DOI] [PubMed] [Google Scholar]
- 4.Kamel KS, Halperin ML. Acid-base problems in diabetic ketoacidosis. N Engl J Med 2015;372:546–54. 10.1056/NEJMra1207788 [DOI] [PubMed] [Google Scholar]
- 5.Elisaf MS, Tsatsoulis AA, Katopodis KP, et al. . Acid-base and electrolyte disturbances in patients with diabetic ketoacidosis. Diabetes Res Clin Pract 1996;34:23–7. 10.1016/S0168-8227(96)01332-0 [DOI] [PubMed] [Google Scholar]
- 6.Svart MV, Voss TS, Bayat M, et al. . Rare presentations of ketoacidosis: diabetic ketoalkalosis and ketoacidosis secondary to fasting and muscular dystrophy. Clin Diabetes 2015;33:37–9. 10.2337/diaclin.33.1.37 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jerrard D, Hanna J. Diabetic ketoacidosis with alkalemia. Am J Emerg Med 2001;19:521–2. 10.1053/ajem.2001.27167 [DOI] [PubMed] [Google Scholar]