ACETAZOLAMIDE, a carbonic anhydrase inhibitor, is used in patients with meningeal inflammation, mild intracranial hypertension, and basal skull fractures to decrease the formation of cerebrospinal fluid (CSF). It causes mild metabolic acidosis by inhibiting the reabsorption of bicarbonate (HCO−3) ions from renal tubules. This effect has been used successfully in the treatment of patients with chronic respiratory acidosis with superimposed metabolic alkalosis 1and central sleep apnea syndrome. 2Life-threatening metabolic acidosis during acetazolamide therapy has been observed only in patients with renal impairment or 3diabetes 4and in elderly patients. 5Severe metabolic acidosis, associated with acetazolamide, in the absence of other predisposing factors has not been reported in patients with central nervous system disease. We report three cases of severe metabolic acidosis and hyperventilation during acetazolamide therapy in normal doses in adult patients without renal impairment.
A 35-yr-old man with a head injury underwent craniotomy for evacuation of a traumatic left temporal extradural hematoma. Postoperatively, the patient underwent mechanical ventilation to maintain a partial pressure of arterial carbon dioxide (Paco2) of 30–35 mmHg. On the third postoperative day, 250 mg acetazolamide administered every 8 h through a nasogastric tube was started to treat a CSF leak from the operative wound. A T-piece trial of weaning was started on the fourth postoperative day. On the fifth postoperative day, patient respiratory rate increased to 40–44 breaths/min. Arterial blood gas analysis showed metabolic acidosis resulting in compensatory hypocapnia and a normal pH (table 1). The patient was sedated and underwent artificial ventilation for the next 6 days. Attempts at discontinuation of sedation resulted in resumption of hyperventilation. The patient was normovolemic and his blood urea, serum creatinine, and serum Na+and K+concentrations were normal. Acetazolamide administration was discontinued on the eleventh postoperative day. Three days after discontinuation of acetazolamide, patient respiratory rate and acid–base status returned to normal. He was weaned from mechanical ventilation and his trachea was extubated on the nineteenth postoperative day.
A 35-yr-old man underwent craniotomy for evacuation of a traumatic subdural hematoma and decompression of a right temporal contusion. Intraoperatively, moderate brain swelling developed, which necessitated postoperative mechanical ventilation. On the fifth postoperative day, 250 mg acetazolamide (administered every 8 h through the nasogastric tube) was started to treat a CSF leak from the operative wound. A T-piece trial of weaning from mechanical ventilation, commenced on the ninth postoperative day, was discontinued because of severe metabolic acidosis resulting in hypocapnia and a normal pH (table 1). Metabolic acidosis was corrected with administration of a 7.5% sodium bicarbonate infusion, 100 ml every 6 h for 48 h, with repeated analyses of blood gas. The patient was normovolemic and his blood urea, serum creatinine, and serum Na+and K+concentrations were normal. With the discontinuation of acetazolamide therapy on the tenth postoperative day, metabolic acidosis and hyperventilation improved gradually during the next 6 days, and the patient was weaned from mechanical ventilation.
A 42-yr-old woman with a diagnosis of neurocysticercosis presented with signs of increased intracranial pressure. She was treated with use of oral albendazole, intravenous dexamethasone, oral glycerol, and carbamazepine. Because her headache worsened despite continued therapy, oral acetazolamide therapy was started in a dosage of 250 mg every 8 h to decrease the intracranial pressure. During the next 2 days, she became febrile and delirious. Her temperature was 39.5°C and respiratory rate was 36 breaths/min. Empiric antibiotic therapy was started because of clinical suspicion of septicemia. Tachypnea and hyperventilation worsened during the ensuing 12 h. Blood gas analysis showed severe metabolic acidosis resulting in hypocapnia and a normal pH (table 1). After administration of 36 mg morphine and 20 mg intravenous diazepam, her lungs were mechanically ventilated in synchronized intermittent mandatory ventilation mode at a mandatory rate of 15 breaths/min, a pressure support of 8 cm H2O, a positive end-expiratory pressure of 4 cm H2O and an inspired oxygen fraction (Fio2) of 0.5. Ensuing moderate hypotension was treated with dobutamine infusion for 8 h. The patient was normovolemic and blood urea, serum creatinine, and serum Na+and K+concentrations were normal. Repeated doses of sodium bicarbonate were administered to control the metabolic acidosis during the next 48 h. Simultaneously, acetazolamide therapy was discontinued. Metabolic acidosis improved during the ensuing 48 h and the patient’s trachea was extubated on the third day of admission to the intensive care unit.
In this study, we presented three patients with central nervous system disease treated with 250 mg acetazolamide every 8 h. In each patient, more severe metabolic acidosis developed than is usually caused by acetazolamide. In each case, appropriate compensatory hyperventilation and hypocapnia partially or completely normalized arterial pH. All patients received acetazolamide in normal doses and none had any another predisposing factor.
The mechanism of this complication is difficult to explain. One possible explanation is that the central nervous system disease in these patients triggered hyperventilation, and acetazolamide may have had an enhanced ability to reduce plasma bicarbonate during these circumstances, resulting in dramatic metabolic acidosis. However, the partial pressure of carbon dioxide (Pco2) and pH responses to metabolic acidosis in our patients are those described as normally expected compensation. 6Alternatively, lactic acidosis can be speculated to have compounded the acetazolamide-induced acid shift. There are no specific causes for lactic acidosis in our patients, with the exception of hypocapnia, which has been suggested to increase the lactic acid levels in blood.
Acetazolamide therapy causes metabolic acidosis by inhibiting reabsorption of HCO−3 from renal tubules. In the brain, it causes CSF and extracellular fluid acidosis. The abrupt decrease of pH of the brain extracellular fluid is caused by a local increase of carbonic acid. In addition, acetazolamide also inhibits the conversion of carbon dioxide to HCO−3in erythrocytes. Thus, it increases brain Pco2and acidifies the brain extracellular fluid, which explains its ventilatory stimulus. 7,8
During systemic metabolic acidosis, carbonic anhydrase inhibition in the choroid plexus has been suggested to augment CSF acidosis. 9,10However, this hypothesis conflicts with the other evidence, which indicates that the decrease in pH is not caused by the choroid plexus but by the local increase of carbonic acid. 7
Hyperventilation during acetazolamide therapy seems to have a complex cause. Systemic or CSF acidosis caused by this drug seems to be an important mechanism. In human volunteers, short- and long-term acetazolamide administration has been shown to cause hyperventilation through central and peripheral chemoreceptor mechanisms. 11In cats, respiratory stimulation has been reported, which is mediated through peripheral chemoreceptor mechanisms and the effect of acetazolamide on cerebral blood flow. 12
Although the potential of acetazolamide to cause metabolic acidosis and hyperventilation has been shown in experimental animals and healthy volunteers, we are not aware of clinical reports of this complication in patients with cerebral disease without predisposing factors. Metabolic acidosis caused by acetazolamide usually is mild. Severe metabolic acidosis and respiratory alkalosis are seen only in patients with renal dysfunction 3or diabetes 4or in the elderly patients. 5None of our patients had these predisposing factors.
In all three patients, administration of acetazolamide had a temporal relation to the development of metabolic acidosis and hyperventilation. In patients 1 and 2, hyperventilation delayed weaning mechanical ventilation. In patient 3, severe hyperventilation necessitated administration of very high doses of sedatives and institution of mechanical ventilation. Cessation of acetazolamide therapy resulted in a trend toward correction of metabolic acidosis and hyperventilation; however, complete normalization occurred only after 2–6 days. Evidence from these cases also suggests that even normal doses of acetazolamide may precipitate severe metabolic acidosis in some patients.
Acetazolamide administration per se has been suggested to increase cerebral blood flow (CBF) via a mechanism that does not involve carbonic anhydrase. 13However, this evidence is not compelling because the authors did not account for carbonic acidosis in the brain. With long-term acetazolamide therapy, CBF returns to normal and the cerebral circulation has a normal response to carbon dioxide. 14This results from normal rates of excretion of metabolic H+and HCO−3at abnormally high concentration gradients in the steady state after long-term acetazolamide therapy. The extent of hyperventilation seen in our patients might be predicted to cause cerebral vasoconstriction. 15Experimental evidence, on the contrary, showed that brain tissue Po2increased, not decreased, with acetazolamide. 16This is understandable because cerebral acidosis caused by acetazolamide also dilates cerebral vessels.
In conclusion, in patients with central nervous system disease without other known predisposing factors, severe metabolic acidosis may occur with normal doses of acetazolamide. After discontinuation of the therapy, it may take more than 48 h for the metabolic acidosis and hyperventilation to be corrected. Therefore, acid–base status should be monitored closely in critically ill neurologic patients who are prescribed acetazolamide therapy. The chemistry in this field is complex and not understood well. More research is needed before the causes of this complication can be stated with certainty. No specific evidence implicates cerebral injury as the cause of hyperventilation in the patients described herein. It may be worthwhile to measure the lactate levels to further understand the problem. It would also be of interest to investigate subclinical renal or other mechanisms that might make some patients more susceptible than others to metabolic acidosis during acetazolamide therapy.