- Email: [email protected]
Alkalemia in diabetic ketoacidosis
Alkalemia in Diabetic Ketoacidosis
A patient with a history of diabetes mellitus and congestive heart failure was taking furosemide and metolazone as diuretics. Diabetic ketoacidosis developed, and the patient became lethargic and confused. Initial biochemical determinations showed an alkalemic pH, serum and urine ketones with an anion gap, and hyperventilation. The hyperventilation was appropriate for the degree of ketoacidosis but it was grossly inappropriate for the alkalemia. This could be explained by a direct effect of ketones on the respiratory center or a sudden increase in hydrogen ion concentration superimposed on previously chronic alkalemic pH due to the potent combination of furosemide and metolazone.
JON W. CRONIN, M.D. SUSAN F. KROOP, M.D. JONATHAN DIAMOND, M.D. ARTURO R. ROLLA, M.D. Boston, Massachusetts
The classic derangement of diabetic ketoacidosis is a metabolic acidosis with an increased anion gap secondary to the accumulation of ketoacids and hydrogen ions. Usually, this results in an acidemia (pH less than 7.35). Very rarely, alkalemia (pH greater than 7.45) occurs in this setting. When present, it signifies at least a dual acid-base disturbance with the alkalosis overriding the ketoacidosis. Several reports have shown protracted vomiting [l-6], alkali ingestion [ 1,4], and mineralocorticoid excess [7,8] as the underlying causes for the alkalosis. Review of the literature showed only one previous report
of alkalemia in diabetic ketoacidosis due to diuretic use [9]. We describe herein a patient with this mixed acid-base disturbance secondary to prolonged use of furosemide and metolazone. CASE REPORT
From the Department of Medicine, New England Deaconess Hospital and Harvard Medical School, Boston, Massachusetts. Requests for reprints should be addressed to Dr. Arturo R. Rolla. New England Deaconess Hospital, 185 Pilgrim Road, Boston, Massachusetts 02215. Manuscript accepted October 18, 1983.
192
July 1984
A 66-year-old white woman with diabetes mellitus and congestive heart failure presented to the New England Deaconess Hospital with a two-day history of polydypsia, polyuria, and rapid onset of lethargy and confusion. There was no history of vomiting, diarrhea, fever, or head trauma. Family members denied any recent ingestion of alcohol, alkali, salicylates, or other toxins. Current medications included 10 units of NPH insulin every morning, metolazone 2.5 mg twice daily, furosemide 40 mg daily, and digoxin 0.125 daily. There was a concern that the patient may have inadvertently increased her diuretic therapy for about two weeks prior to admission. On admission, the patient was minimally responsive to verbal stimuli, and acetone was noted on her breath. Blood pressure was 120170 mm Hg, pulse 110 per minute, respirations 26 per minute, and temperature 37% (99OF). There was no evidence of head trauma, and the neck was supple. The chest was clear to auscultation and percussion. The cardiac examination was noteworthy for a grade IV/VI holosystolic murmur at the apex radiating to the axilla. Neurologic examination revealed a confused and slowly responsive patient without localizing signs. Admission laboratory data included a hematocrit of 38 percent, leukocyte
The American Journal of Medicine
Volume 77
ALKALEMIA
IN DIABETIC KETOACIDOSIS-CRONIN
ET AL
Count Of 5,300/mm3, and the following chemical values: gkJCOSe1,047 mg/dl, sodium 127 meq/liter, potassium 4.1 rI’Nq/litet’, bicarbonate 30 meq/liter, chloride 78 meq/liter, calcium 10 mg/dl. The anion gap was calculated to be 19 rneqjliter. Arterial blood gas sample with the patient breathing room air showed a pH of 7.70, carbon dioxide tension of 11 mm Hg, and oxygen tension of 111 mm Hg. Ethanol and salicylates were not detected on toxic screening. Serum ketones (nitroprusside test) were found at a 1:8 dilution. Urinalysis revealed glycosuria and ketonuria. The electrocardiographic results were normal. Chest radiography showed moderate cardiomegaly and clear lung fields. Interpretation of the initial laboratory data revealed a triple acid-base disorder consisting of diabetic ketoacidosis with secondary respiratory alkalosis and metabolic alkalosis. The overall result was alkalemia. Treatment was begun with normal saline infusion intravenously and regular insulin (10 units were given as an initial intravenous bolus followed by 5 units of regular insulin per hour by intravenous pump infusion). Serial blood chemical and gas values are shown in Figure 1. After four hours of treatment, the pH had decreased to 7.60, and the serum bicarbonate had risen to 38 meq/liter. At this time, acetazolamide 250 mg intravenously every six hours was added. By the end of the first hospital day, the blood chemical values had returned to normal, and the patient’s mental status had improved.
COMMENTS Metabolic alkalosis and acidosis occasionally occur simultaneously. The resulting pH is the sum of these two opposing processes. Alkalemia complicating diabetic ketoacidosis is an uncommon metabolic aberration. It should be suspected when the serum chloride level is initially lower than normal, and the bicarbonate and pH are higher than expected for the observed anion gap. A thorough history reveals the cause of the alkalotic tendency in the majority of patients; prolonged vomiting [l-6], alkali ingestion [ 1,4], mineralocorticoid excess [7,8], and diuretic use [9] have been reported. In our patient, an alkalemic pH developed as the net effect of a triple acid-base disturbance. The metabolic acidosis was readily explained by the overproduction of ketoacids also leading to an increased anion gap. The metabolic alkalosis was secondary to diuretic administration. Furosemide effect is potentiated by the concomitant use of metolazone. The respiratory alkalosti; should be a natural consequence of diabetic ketoacidosis; however, in the presence of an elevated pH, it is
difficult to explain. The protective hypoventilation and consequent hypercapnia of metabolic alkalosis states were absent in our patient; on the contrary, she was hypocapnic (initial carbon dioxide tension was 11 mm Hg). The accepted cause for hyperventilation in diabetic ketoacidosis is the effect of the increase in hydrogen ion concentration on the chemoreceptors of the respiratory center of the medulla. This was not the case in our patient since she was alkalemic. Possible alterna-
tive explanations are a direct effect of ketoacids on the
Figure 1. Serum glucose, electrolytes, pH, and carbon dioxide tension during the first 10 hours of admission.
respiratory
center or a sudden change in pH and concentration of hydrogen ions on a previously chronically alkalemic respiratory center. The respiratory response to metabolic alkalosis has been known to be erratic and/or variable. It is possible that chronic alkalemia resets the respiratory center to a lower concentration of hydrogen ions. A sudden increase in their concentration, such as with the superimposition of diabetic ketoacidosis, could trigger a response of hyperventilation, even while still in the “alkalemic range.” This theory is supported in our patient by the fact that the hyperventilation and ketoacidosis disappeared at the same time, and her metabolic alkalosis persisted once the other acid-base disturbances were treated. This case underscores a number of important points in the management of diabetic ketoacidosis. First, that patients with diabetic ketoacidosis may have a mixed acid-base disorder, which can be severe enough to cause an alkalemic pH in rare circumstances. Second, arterial blood gases should be determined at least initially in patients with diabetic ketoacidosis to ascertain the presence and severity of other acid-base disorders. Third, protracted vomiting, mineralocorticoid excess, alkali ingestion, and diuretic administration in a patient with diabetic ketoacidosis suggest the possibility of a mixed disorder.
July 1984
The American Journal of Medlclne
Volume 77
193
ALKALEMIA IN DIABETIC KETOACIDOSIS-CRONIN ET AL
REFERENCES 1. 2.
3.
4. 5.
194
Bleicher S: Ketosis not always acidosis: “heartburn” can be relevant. Diabetic Outlook 1967; 2: 3-4. Rogging G, Moses D, Kautcher M, Wishner W, Shuman C: Ketosis and metabolic alkalosis in a patient with diabetes. JAMA 1970; 211: 296-296. Jimenez JA, Damiano A, Fernandez E, Rodriguez-Rubio F, Garrido M: Metabolic alkalosis in diabetic ketosis. JAMA 1975; 233: 1193-1194. Walsh CH, Lim KC: Diabetic ketoalkalosis, a readily misdiagnosed entity. Br Med J 1976; 2: 19. Melrose E, Morgan A, Harrower AD, Smith AW: Diabetic ke-
July 1994
The American Journal of Medicine
Volume 77
6. 7. 8.
9.
toalkalosis. Br Med J 1976; 2: 237. Sanders G, Boyle G, Hunter S, Poffenbarger P: Mixed acid base abnormalities in diabetes. Diabetes Care 1978; 1: 362. O’Reilly DJ, Delamere JP: Cause of alkalosis in “diabetic ketoalkalosis.” Clin Chem 1980; 26: 171-172. Pearson DW, Kennedy AC, Thomson JA, Toner PG, Ratcliffe JG: Diabetic ketoacidosis due to ectopic ACTH production from an oat cell carcinoma. Postgrad Med J 1981; 57: 455-456. Koett J, Howell J, Steinberg S, Wolf P: Diabetic ketoalkalosis. Clin Chem 1979; 25: 1328-1329.
A patient with a history of diabetes mellitus and congestive heart failure was taking furosemide and metolazone as diuretics. Diabetic ketoacidosis developed, and the patient became lethargic and confused. Initial biochemical determinations showed an alkalemic pH, serum and urine ketones with an anion gap, and hyperventilation. The hyperventilation was appropriate for the degree of ketoacidosis but it was grossly inappropriate for the alkalemia. This could be explained by a direct effect of ketones on the respiratory center or a sudden increase in hydrogen ion concentration superimposed on previously chronic alkalemic pH due to the potent combination of furosemide and metolazone.
JON W. CRONIN, M.D. SUSAN F. KROOP, M.D. JONATHAN DIAMOND, M.D. ARTURO R. ROLLA, M.D. Boston, Massachusetts
The classic derangement of diabetic ketoacidosis is a metabolic acidosis with an increased anion gap secondary to the accumulation of ketoacids and hydrogen ions. Usually, this results in an acidemia (pH less than 7.35). Very rarely, alkalemia (pH greater than 7.45) occurs in this setting. When present, it signifies at least a dual acid-base disturbance with the alkalosis overriding the ketoacidosis. Several reports have shown protracted vomiting [l-6], alkali ingestion [ 1,4], and mineralocorticoid excess [7,8] as the underlying causes for the alkalosis. Review of the literature showed only one previous report
of alkalemia in diabetic ketoacidosis due to diuretic use [9]. We describe herein a patient with this mixed acid-base disturbance secondary to prolonged use of furosemide and metolazone. CASE REPORT
From the Department of Medicine, New England Deaconess Hospital and Harvard Medical School, Boston, Massachusetts. Requests for reprints should be addressed to Dr. Arturo R. Rolla. New England Deaconess Hospital, 185 Pilgrim Road, Boston, Massachusetts 02215. Manuscript accepted October 18, 1983.
192
July 1984
A 66-year-old white woman with diabetes mellitus and congestive heart failure presented to the New England Deaconess Hospital with a two-day history of polydypsia, polyuria, and rapid onset of lethargy and confusion. There was no history of vomiting, diarrhea, fever, or head trauma. Family members denied any recent ingestion of alcohol, alkali, salicylates, or other toxins. Current medications included 10 units of NPH insulin every morning, metolazone 2.5 mg twice daily, furosemide 40 mg daily, and digoxin 0.125 daily. There was a concern that the patient may have inadvertently increased her diuretic therapy for about two weeks prior to admission. On admission, the patient was minimally responsive to verbal stimuli, and acetone was noted on her breath. Blood pressure was 120170 mm Hg, pulse 110 per minute, respirations 26 per minute, and temperature 37% (99OF). There was no evidence of head trauma, and the neck was supple. The chest was clear to auscultation and percussion. The cardiac examination was noteworthy for a grade IV/VI holosystolic murmur at the apex radiating to the axilla. Neurologic examination revealed a confused and slowly responsive patient without localizing signs. Admission laboratory data included a hematocrit of 38 percent, leukocyte
The American Journal of Medicine
Volume 77
ALKALEMIA
IN DIABETIC KETOACIDOSIS-CRONIN
ET AL
Count Of 5,300/mm3, and the following chemical values: gkJCOSe1,047 mg/dl, sodium 127 meq/liter, potassium 4.1 rI’Nq/litet’, bicarbonate 30 meq/liter, chloride 78 meq/liter, calcium 10 mg/dl. The anion gap was calculated to be 19 rneqjliter. Arterial blood gas sample with the patient breathing room air showed a pH of 7.70, carbon dioxide tension of 11 mm Hg, and oxygen tension of 111 mm Hg. Ethanol and salicylates were not detected on toxic screening. Serum ketones (nitroprusside test) were found at a 1:8 dilution. Urinalysis revealed glycosuria and ketonuria. The electrocardiographic results were normal. Chest radiography showed moderate cardiomegaly and clear lung fields. Interpretation of the initial laboratory data revealed a triple acid-base disorder consisting of diabetic ketoacidosis with secondary respiratory alkalosis and metabolic alkalosis. The overall result was alkalemia. Treatment was begun with normal saline infusion intravenously and regular insulin (10 units were given as an initial intravenous bolus followed by 5 units of regular insulin per hour by intravenous pump infusion). Serial blood chemical and gas values are shown in Figure 1. After four hours of treatment, the pH had decreased to 7.60, and the serum bicarbonate had risen to 38 meq/liter. At this time, acetazolamide 250 mg intravenously every six hours was added. By the end of the first hospital day, the blood chemical values had returned to normal, and the patient’s mental status had improved.
COMMENTS Metabolic alkalosis and acidosis occasionally occur simultaneously. The resulting pH is the sum of these two opposing processes. Alkalemia complicating diabetic ketoacidosis is an uncommon metabolic aberration. It should be suspected when the serum chloride level is initially lower than normal, and the bicarbonate and pH are higher than expected for the observed anion gap. A thorough history reveals the cause of the alkalotic tendency in the majority of patients; prolonged vomiting [l-6], alkali ingestion [ 1,4], mineralocorticoid excess [7,8], and diuretic use [9] have been reported. In our patient, an alkalemic pH developed as the net effect of a triple acid-base disturbance. The metabolic acidosis was readily explained by the overproduction of ketoacids also leading to an increased anion gap. The metabolic alkalosis was secondary to diuretic administration. Furosemide effect is potentiated by the concomitant use of metolazone. The respiratory alkalosti; should be a natural consequence of diabetic ketoacidosis; however, in the presence of an elevated pH, it is
difficult to explain. The protective hypoventilation and consequent hypercapnia of metabolic alkalosis states were absent in our patient; on the contrary, she was hypocapnic (initial carbon dioxide tension was 11 mm Hg). The accepted cause for hyperventilation in diabetic ketoacidosis is the effect of the increase in hydrogen ion concentration on the chemoreceptors of the respiratory center of the medulla. This was not the case in our patient since she was alkalemic. Possible alterna-
tive explanations are a direct effect of ketoacids on the
Figure 1. Serum glucose, electrolytes, pH, and carbon dioxide tension during the first 10 hours of admission.
respiratory
center or a sudden change in pH and concentration of hydrogen ions on a previously chronically alkalemic respiratory center. The respiratory response to metabolic alkalosis has been known to be erratic and/or variable. It is possible that chronic alkalemia resets the respiratory center to a lower concentration of hydrogen ions. A sudden increase in their concentration, such as with the superimposition of diabetic ketoacidosis, could trigger a response of hyperventilation, even while still in the “alkalemic range.” This theory is supported in our patient by the fact that the hyperventilation and ketoacidosis disappeared at the same time, and her metabolic alkalosis persisted once the other acid-base disturbances were treated. This case underscores a number of important points in the management of diabetic ketoacidosis. First, that patients with diabetic ketoacidosis may have a mixed acid-base disorder, which can be severe enough to cause an alkalemic pH in rare circumstances. Second, arterial blood gases should be determined at least initially in patients with diabetic ketoacidosis to ascertain the presence and severity of other acid-base disorders. Third, protracted vomiting, mineralocorticoid excess, alkali ingestion, and diuretic administration in a patient with diabetic ketoacidosis suggest the possibility of a mixed disorder.
July 1984
The American Journal of Medlclne
Volume 77
193
ALKALEMIA IN DIABETIC KETOACIDOSIS-CRONIN ET AL
REFERENCES 1. 2.
3.
4. 5.
194
Bleicher S: Ketosis not always acidosis: “heartburn” can be relevant. Diabetic Outlook 1967; 2: 3-4. Rogging G, Moses D, Kautcher M, Wishner W, Shuman C: Ketosis and metabolic alkalosis in a patient with diabetes. JAMA 1970; 211: 296-296. Jimenez JA, Damiano A, Fernandez E, Rodriguez-Rubio F, Garrido M: Metabolic alkalosis in diabetic ketosis. JAMA 1975; 233: 1193-1194. Walsh CH, Lim KC: Diabetic ketoalkalosis, a readily misdiagnosed entity. Br Med J 1976; 2: 19. Melrose E, Morgan A, Harrower AD, Smith AW: Diabetic ke-
July 1994
The American Journal of Medicine
Volume 77
6. 7. 8.
9.
toalkalosis. Br Med J 1976; 2: 237. Sanders G, Boyle G, Hunter S, Poffenbarger P: Mixed acid base abnormalities in diabetes. Diabetes Care 1978; 1: 362. O’Reilly DJ, Delamere JP: Cause of alkalosis in “diabetic ketoalkalosis.” Clin Chem 1980; 26: 171-172. Pearson DW, Kennedy AC, Thomson JA, Toner PG, Ratcliffe JG: Diabetic ketoacidosis due to ectopic ACTH production from an oat cell carcinoma. Postgrad Med J 1981; 57: 455-456. Koett J, Howell J, Steinberg S, Wolf P: Diabetic ketoalkalosis. Clin Chem 1979; 25: 1328-1329.