Extreme alkalaemia with mixed alkalosis due to suspected acute-on-chronic respiratory alkalosis

Abstract

Herein we present a case of severe alkalaemia (pH 7.81) due to suspected acute-on-chronic respiratory alkalosis in a patient with chronic anxiety and metabolic alkalosis secondary to emesis. The patient was managed in the intensive care unit with significant improvement and discharged in stable condition. The case report emphasises considering a broad differential of aetiologies that can cause acid–base status derangements and identifying the appropriate therapeutic approach.

Background

Acid–base disorders (ABDs) are common in hospitalised patients, with a prevalence from 62% to 71% in cross-sectional studies of a pooled population in the intensive care unit (ICU) and emergency department (ED).^{1 2} A prospective observational study identified that an elevated serum pH (alkalaemia) was associated with higher levels of morbidity and mortality, with a mortality rate as high as 48.5% in patients with a pH greater than 7.60.³ Alkalaemia can lead to life-threatening arrhythmias.⁴ Thus, it is crucial to recognise and explain the underlying conditions that led to alkalaemia and provide a focused treatment plan.

Case presentation

Our patient is a man in his 60s with a medical history of coronary artery disease requiring a coronary artery bypass graft 30 years prior, arterial hypertension, hyperlipidaemia, general anxiety disorder, obstructive sleep apnoea on home continuous positive airway pressure and permanent pacemaker placement because of sick sinus syndrome. He presented to the ED with haematemesis, haematochezia and generalised weakness. The patient was tachycardic at 104 beats/min with a regular rhythm; blood pressure was 137/93 mm Hg, a respiratory rate of 18 breaths/min and a pulse oximeter saturation reading (SpO₂) of 98%. The patient reported hyporexia in the past 2 weeks with increased emesis (up to 10 times per day) and several bouts of bloody diarrhoea. The patient stated he was feeling shaky and diaphoretic. His travel history was non-significant, and he denied having recent sick contacts. He denied having any previous upper endoscopies or colonoscopies. His current medications were atorvastatin, duloxetine, losartan–hydrochlorothiazide, metoprolol succinate, nitroglycerine, clopidogrel and aspirin.

We observed melaena in the rectal vault with a positive stool guaiac test. Initial haemoglobin was 14.8 g/L (reference range: 13.5–17.5 g/L), while white cell count was slightly elevated at 12 000 10⁹/L (reference range: 4000–11 000 10⁹/L). Liver enzymes were within normal ranges, while total serum bilirubin was nearly normal (direct and indirect) at 1.5 mg/dL (reference range: 0.0–1.2 mg/dL). International normalised ratio was within normal limits. There was a slight increase in creatinine at 1.24 mg/dL (baseline creatinine pre-admission was 1.0 mg/dL; reference range: 0.7–1.3 mg/dL). A random urine analysis showed mild ketonuria (40 mg/dL; reference value: negative) and proteinuria (30 mg/dL; reference value: negative). One hour later, in the ED, he was hyperaemic and tachypnoeic at 27 breaths/min, a heart rate of 87 beats/min, blood pressure of 153/93 mm Hg and SpO₂ of 91% on room air. An arterial blood gas (ABG) analysis showed alkalaemia (pH 7.81, reference range: 7.35–7.45). The ABG analysis was repeated twice to rule out laboratory error, and the values remained consistent. We noted a respiratory alkalosis, a metabolic alkalosis and an anion gap metabolic acidosis by interpreting the laboratory results.⁵ We noted a secondary metabolic alkalosis with a bicarbonate of 22 mmol/L (expected bicarbonate 9 mmol/L) and an anion gap metabolic acidosis (anion gap=21, reference range: <12 mmol/L) due to lactic acidosis (serum lactate 8.7, reference range: 0.7–2.5 mmol/L). Serum electrolytes showed sodium 139 (reference range: 135–145 mmol/mL), potassium 3.6 (reference range 3.5–5.3 mmol/mL) and chloride 96 (reference range: 98–108 mmol/mL). The patient was transferred to the ICU for further management. We provided intravenous lorazepam for breakthrough episodes of anxiety to control his hyperventilation. We used intravenous ondansetron as needed to control his emesis. We initiated fluid resuscitation with normal saline to improve tissue perfusion. Imaging studies were performed to identify the source of his gastrointestinal bleeding.

Chest X-ray was unremarkable for pleural effusions, infiltrates or pneumothorax. CT of the brain without contrast did not show an acute process. CT of the chest, abdomen and pelvis with intravenous contrast showed no inflammatory changes in the gastrointestinal tract or overt signs of bleeding. A 6×7 mm pulmonary nodule was noted in the right lower lobe, suggesting a non-calcified granuloma; a 3.2×3.6 cm infrarenal abdominal aortic aneurysm was also noted. Vitals before esophagogastroduodenoscopy (EGD) were a respiratory rate of 18 breaths/min, a heart rate of 65 beats/min and a blood pressure of 141/72 mm Hg. Repeat ABG analysis showed a pH of 7.61 (reference range: 7.35–7.45) and partial pressure of carbon dioxide (pCO₂) of 23 (reference range: 35–45 mm Hg). EGD was performed and identified a 1 cm sliding hiatal hernia and antral/pyloric gastritis with superficial erosions, which were biopsied for Helicobacter pylori, resulting in negative results. Peptic duodenitis was noted in the duodenum bulb but without stigmata of recent bleeding. He then underwent a colonoscopy, identifying scattered diverticular changes in the sigmoid colon, multiple diminutive sessile colonic polyps and mild internal haemorrhoids. The patient received pantoprazole and started on a gastroprotective diet. He improved throughout the rest of his hospital stay, did not require tracheal intubation or blood transfusion, and was discharged medically stable with a respiratory rate of 16 breaths/min, heart rate of 82 beats/min and blood pressure of 114/72 mm Hg. The patient had good mentation, was functioning independently with tasks, and was breathing room air and oxygen on his own because of his significant improvement at the time of discharge; therefore, we did not repeat the ABG analysis. In a 2-month follow-up visit with his primary care provider, he had not had any additional episodes of emesis, melaena or bleeding since discharge, and the patient was working on seeing a psychiatrist for better control of his anxiety.

Treatment

Treatment focused on identifying and correcting acid–base derangements and controlling gastrointestinal bleeding. We provided intravenous fluid resuscitation to improve tissue perfusion and correct lactic acidosis. We provided breathing exercises to the patient to control his hyperventilation and administered intravenous lorazepam for breakthrough anxiety to address his respiratory alkalosis secondary to hyperventilation. Due to his recent gastrointestinal bleeding, we initiated pantoprazole, a proton pump inhibitor. We also administered ondansetron for his nausea and emesis. He underwent an EGD, which identified areas of recent bleeding but not active bleeding, indicating no further intervention was necessary. We started the patient on a gastroprotective diet and discharged him from the hospital in stable condition.

Discussion

Our patient had a remarkably high pH (alkalaemia) of 7.81 at presentation, confirmed upon repeated ABG. We suspected he had an underlying chronic respiratory alkalosis due to significant underlying anxiety. Research showed that there is rising morbidity and mortality with increasing alkalaemia. Alkalaemia impacts oxygen delivery through the Bohr effect, which involves increased affinity of haemoglobin for oxygen, leading to impairment in tissue oxygenation.^{6 7}

We also identified a mixed ABD that included respiratory alkalosis in our patient. Respiratory alkalosis develops when there is a significant increase in alveolar ventilation relative to carbon dioxide production. If carbon dioxide production remains stable, hyperventilation causes hypocapnia from the excess loss of carbon dioxide resulting in respiratory alkalosis.¹ The pH in the blood continues to rise (alkalaemia) due to a shift in the form of bicarbonate to its carbon dioxide form and the reduction of hydrogen ions in the blood. Respiratory alkalosis has many causes, including an enhanced respiratory drive from sepsis, hepatic failure, salicylate intoxication, subarachnoid haemorrhage, nicotine, iatrogenic interventions, pregnancy and anxiety due to hyperventilation. We suspected a chronic respiratory alkalosis likely secondary to hyperventilation syndrome and anxiety, which the patient struggled with at baseline. Unfortunately, we only had an ABG analysis available after his presentation; an ABG before admission is best to confirm this hypothesis. Hyperventilation syndrome will often present with dizziness or lightheadedness, paresthesia, palpitations, chest pain or tightness, and, less commonly, carpopedal spasms.^{2 3} Our patient reported some of these symptoms. An infectious workup was negative, and sepsis was ruled out as a cause for hyperventilation. The patient did not have a history of hepatic disease, and non-contrast CT of the head was negative for acute processes, including cerebrovascular accident.

We treated his hyperventilation by restarting duloxetine, initially tried closed-bag breathing, provided lorazepam intravenously for breakthrough anxiety and created a calming environment for the patient, thus helping to correct the respiratory alkalosis. We monitored the patient closely. He remained neurologically and haemodynamically stable. If the patient worsened (respiratory muscle fatigue, loss of consciousness), we could have used deep sedation with mechanical ventilation to adjust minute ventilation according to his acid–base status. Given that the patient had an acute gastrointestinal bleed, it is possible that low circulating blood volume could have led to a worsening of his hyperventilation through a baroreflex interaction.⁸

Next, we suspected a concomitant metabolic alkalosis because of the abnormal bicarbonate levels. In a chronic state, we anticipate the bicarbonate to decrease by 4–5 mEq/L for every 10 mm Hg change below the normal 40 mm Hg pCO₂.⁹ We estimated that the bicarbonate should be approximately 9 mEq/L, but the measured plasma bicarbonate was 22 mEq/L; thus, a diagnosis of mixed respiratory alkalosis with concomitant metabolic alkalosis was made. Metabolic alkalosis increases the plasma bicarbonate concentration above the expected plasma bicarbonate. There are different mechanisms by which metabolic alkalosis can develop. Commonly seen conditions include intracellular hydrogen ion shift (typical in patients with hypokalaemia),¹⁰ loss of hydrogen ions from the gastrointestinal (diarrhoea, vomiting or suctioning) or renal tract (mineralocorticoid excess, diuretic therapy or specific genetic syndromes), exogenous bicarbonate administration or intravascular fluid depletion when the extracellular bicarbonate amount remains stable (also called contraction alkalosis).¹¹ A low serum bicarbonate concentration is due to less filtration, more efficient reabsorption or less excretion by the kidneys. Otherwise, the excess bicarbonate would be lost in urine.⁶ Obtaining urine potassium and chloride can be helpful if the aetiology of metabolic alkalosis is not immediately evident. High urinary potassium in hypokalaemia suggests an intrinsic renal aetiology, whereas low urinary potassium levels indicate extrarenal losses. Low urinary chloride is usually seen with volume depletion and is responsive to normal saline infusion.¹¹ In the case of our patient, his serum potassium level was within normal limits, he was not undergoing gastric suctioning, and he had not been given diuretics or bicarbonate supplementation. Thus, we interpreted his metabolic alkalosis as secondary to hydrogen ion loss from his emesis, which was treated with adequate fluid resuscitation and controlled with ondansetron.

In addition, our patient had a significant arterial lactate elevation, which may explain the high anion gap. Lactic acidosis has a broad differential diagnosis and can depend on tissue hypoxia or aerobic glycolysis, with epinephrine-dependent stimulation of beta-2-adrenoceptor being an essential contributor to the latter.¹² Critically ill patients commonly have elevated lactic acid in extensive trauma, sepsis, hypovolaemic or cardiogenic shock, but multiple other culprits exist. Many studies reveal possible causes for an elevated anion gap in respiratory alkalosis. Alkalosis is known to diminish the clearance of lactic acid and increase plasma lactate levels in the serum; the latter effect is an intracellular physiological response known to occur with elevated blood pH.¹³ Current research on patients with panic disorder who hyperventilate showed that patients with severe hypocapnia had higher lactate concentrations in the serum than those with mild reductions in carbon dioxide levels.¹⁴ One possible explanation is a compensatory response of released hydrogen ions to combine with bicarbonate, decreasing the acidosis.¹³ Many patients will have mild chronic respiratory alkalosis due to hyperventilation and significant lactic acid production as they will have acute-on-chronic respiratory alkalosis.¹⁵ Given our patient’s clinical picture and history of diarrhoea and vomiting, we suspected lactic acidosis was more likely due to hypovolaemia and less likely due to septic or cardiogenic shock. The above compensatory response contributed to this scenario. Throughout treatment, fluid infusion and correction of alkalaemia improved lactic acid level to 0.9 mmol/L (reference range: 0.7–2.5).

Learning points.

• Consider acute anxiety and hyperventilation syndrome in the differential of respiratory alkalosis, particularly after ruling out common causes such as sepsis, fever and hypoxaemia.

• Suspect concurrent metabolic alkalosis based on elevation in the expected serum bicarbonate.

• Assess for occult acidosis using an anion gap which could be present in alkalaemia and mixed alkalosis.

• Adequate control of acute anxiety is an important management strategy in patients who develop anxiety-related respiratory alkalosis.

• Morbidity and mortality significantly increase as pH increases due to changes in oxygen–haemoglobin binding capacity.

Ethics statements

Patient consent for publication

Obtained.