ANESTHESIA AND ANALGESIA . . . Current Researches VOL.56, No. 4, JULY-AUGUST, 1977
506
Hyperglycemic Hyperosmolar Nonketotic Coma Following Diazoxide, Anesthesia and Operation BAEKHYO SHIN, MD*
SAMUEL I. JOSEPH, PhD, MDt Baltimore, Maryland$
Unsuspected hyperglycemia caused prolonged postanesthetic coma in a nondiabetic patient treated with IV diazoxide and furosemide. The combined effect of these drugs with general anesthesia and surgical stress was postulated a s
the cause of the hyperglycemia. Blood glucose should be measured during and after general anesthesia in patients who are potential candidates for hyperglycemic hyperosmolar nonketotic coma.
H
border, enlarged prostate, and a pulsatile abdominal mass.
hyperosmolar nonketotic coma (HHNKC) has been reported in known diabetic and previously unknown diabetic patients, associated with general anesthesia.'.' HHNKC has also been reported in association with the use of diazoxide in the management of resistant hypertension,"J unassociated with anesthesia. This is a report of HHNKC following a single I V dose of diazoxide and subsequent surgical anesthesia. YPERGLYCEMIC
REPORT OF A CASE A 70-year-old man was admitted to the emergency room with a chief complaint of abdominal pain. He was known to have had gastric carcinoma 3 years before, at which time gastrectomy, distal pancreatectomy, and esophagojejunostomy were performed. Since then he had been in excellent health, taking no medications. Pertinent physical findings on admission were: blood pressure 220/110, irregular pulse with a rate of 90, grade 111-IV murmur at the left sternal
Soon after admission, blood pressure rose to 2601120. Diazoxide (300 mg) was injected IV. Furosemide (40 mg) was also injected, to minimize the hypernatremia and expansion of plasma volume often associated with diazoxide administration.5 Digoxin was given, apparently in anticipation of impending congestive heart failure. Lidocaine was given I V for frequent. premature ventricular contractions (PVCs) . Blood pressure promptly fell to lSO/SO and stabilized at that level. Routine laboratory data, including blood sugar, were within normal limits. Following an aortogram which revealed an abdominal aneurysm, resection was initiated under N,O-0,-halothane anesthesia supplemented with d-tubocurarine, 4 hours after administration of diazoxide. During the 51/2-hour procedure, blood pressure was maintained between 130170 and 150180,
*Assistant Professor. f Professor.
$Department of Anesthesiology, University of Maryland School of Medicine, Baltimore, Maryland 21201. Paper received: 2/9/76 Accepted for publication: 11/2/76
Hyperosmolar Nonketotic Coma . . . Shin and Joseph
TABLE Laboratory Findings in Recovery Room Arterial blood goses ( F I O ~ ~ . " )
PH Paoz
Blood chemistry
7.41
Na'
129mEq/L
108 t o r r
C1-
94mEq/L
Pacoa
34 torr
K'
Hb sab
99.8%
C02
3.6mEq/L 28mEq/L
Standard HCO, Base excess
21
BUN
17mg
2
507
DISCUSSION Diazoxide, which has been used successfully to treat patients with severe hypertension, has been reported to increase the blood glucose level mainly by inhibiting insulin release from the pancreas and stimulating catecholamine release.6-8 Hyperglycemia may be further compounded by the concomitant injection of furosemide or thiazides, which are frequently used with diazoxide and are structurally similar."." Diphenylhydantoin is also known to cause hyperglycemia; the mechanism is unknown.10
Elevated blood glucose is often found during and following general anesthesia in response to surgical and anesthetic stress,' particularly in light anesthesia and intraabdominal operations. Although the cause of hyperglycemia was believed to be the glycogenolytic effect of catecholamines released under anesthetic stress, it is also reported that glucose intolerance may play a role to increase blood glucose in halothane In the recovery room, the patient, not yet and methoxyflurane anesthesia.l3 It has responding, was placed on a ventilator. With been suggested that glucose intolerance due vital signs stable, he remained unresponsive to decreased response to insulin may be for several hours except for occasional associated with a low cardiac output state, twitching of the hands and feet. Block-AidcC as hyperglycemia has been observed fremonitor indicated no residual effect of d-tu- quently during cardiopulmonary bypass in bocurarine. ECG revealed occasional PVCs, both diabetic and nondiabetic patients.'" but no other abnormalities. Laboratory findings at this time were as shown in the table. Since a single dose of diazoxide rarely causes severe hyperglycemia, the magnitude A diagnosis of an idiopathic convulsive in this case may well have been the result disorder following anesthesia was made. of several of the factors mentioned. InadeTwo doses of diphenylhydantoin (100 mg) quate intraoperative replacement of fluid were given IV, although the EEG did not and excessive administration of diuretics show seizure activity. Twelve hours post- could contribute to hyperosmolality. operatively, HHNKC was first and finally suspected when attention was brought to Twitching of the hands and feet during the fact that the 2-hour postoperative blood the comatose state may have represented glucose and osmolality were 942 mg % and extrapyramidal symptoms, such as those 336 mOsm/L, respectively. However, blood observed by Neary and coworkers1.' in hyglucose at this time was 463 mg % and the pertensive patients treated with diazoxide, patient began to regain consciousness. Lac- although these usually develop during longtated Ringer's solution was infused at a term therapy. Such symptoms may lead to rate of 500 ml/hr. Twenty units of regular the erroneous diagnosis of cerebrovascular insulin were given subcutaneously. Five accident or convulsive disorder. hours later, the blood glucose was 44 mg % HHNKC is more likely to occur in older and osmolality was 288 mOsm/L. Five percent glucose was administered to correct the patients with mild diabetes. Dehydration, low glucose. The patient regained conscious- steroid therapy, burns, hyperalimentation, ness completely 18 hours postoperatively, sepsis, pancreatitis, uremia, and surgical without any residual neurologic deficit. He stress are other possible contributing factors. required no further insulin during his hos- Mortality in prolonged hyperosmolar coma pital stay and was discharged from the hos- may be as high as 40 to 60 percent. Early pital without further complication 13 days recognition, appropriate administration of insulin, and replacement of water, sodium, later. pulse 90-110, and CVP 7-14 cm H,O. Estimated blood loss was 3000 ml. Fluid replacement consisted of 500 ml of D,/W and 3000 ml of whole blood. Mannitol (100 gm) and furosemide (240 mg) were given in a vigorous attempt to overcome the oliguria of the first 3 hours of operation. Total urine output following diuretics was 1100 ml for the next 2 hours.
508
ANESTHESIA AND ANALGESIA. . . Current Researches VOL.56, No. 4, JULY-AUGUST, 1977
and potassium are essential to decrease mortality.15 Early recognition will be aided by glucose determinations during and after anesthesia in patients who are potential candidates for HHNKC, especially in patients treated with diazoxide. The foregoing report is not tendered as a model of suggested management but to warn of the posibjlities and dangers.
7. Senft G: Biochemical aspect of the hyperglycemic action of diazoxide. Ann NY Acad Sci 150:242-253, 1968 8. Porte D J r : Inhibition of insulin release by diazoxide and its relation to catecholamine effect in man. Ann NY Acad Sci 150:281-289, 1968 9. Floy JM, Furman BL: Effect of single dose administration of diuretics on the blood sugar of alloxan-diabetic mice or mice made hyperglycemic by acute administration of diazoxide. Br J PharmaCOI47:124-132, 1973
REFERENCES
10. Peters BH, Samaan NA: Hyperglycemia with relative hypo-insulinemia in diphenylhydantoin toxicity. N Engl J Med 281:91-92, 1969
1. Bedford R: Hyperosmolar hyperglycemic nonketotic coma following general anesthesia. Report of a case. Anesthesiology 35:652-655, 1971
11. Clarke RSJ: The hyperglycemic response to different types of surgery and anaesthesia. Br J Anaesth 42:45-53, 1970
2. Toker P: Hyperosmolar hyperglycemic nonketotic coma, a cause of delayed recovery from anesthesia. Anesthesiology 41:284-285, 1974
12. Merin RG, Samuelson PN, Schalch DS: Major inhalation anesthetics and carbohydrate metabolism. Anesth Analg 50: 625-632, 1971
3. Harrison BDW, Rutter TW, Taylor RT: Severe nonketotic hyperglycemic precoma in a hypertensive patient receiving diazoxide. Lancet 2:599-600, 1972
13. Mills NL, Beaudet RL, Isom OW, et al: Hyperglycemia during cardiopulmonary bypass. Ann Surg 177:203-205, 1973
4. Charles SJ, Harrington AH: Nonketoacidotic hyperglycemia and coma during intravenous diazoxide therapy in uremia. Diabetes 20:501-503, 1971 5. Koch-Weser J: Drug therapy. N Engl J Med 294: 1271-1274, 1976 6. Kozman ME: Evaluation of diazoxide. JAMA 224: 1422-1423, 1973
14. Neary D, Thurston H, Pohl JEF: Development of extrapyramidal symptoms in hypertensive patients treated with diazoxide. Br Med J 3:474475, 1973 15. Arieff A, Carroll HJ: Nonketotic hyperosmolar coma with hyperglycemia: clinical features, pathophysiology, renal function, acid-base balance, plasma-cerebrospinal equilibrium and the effects of therapy in 37 cases. Medicine 51:73-94, 1972
RENAL FUNCTION AND ENFLURANE. Fluoride ion kinetics after nitrous oxideenflurane anesthesia were determined in 16 healthy control patients, 18 anephric patients and 6 patients with creatinine clearance less than 5 ml/min and undergoing chronic hemodialysis. The dose of enflurane was similar in all 3 groups. There was no significant difference between groups with respect to maximum serum inorganic fluoride concentrations (control 13.2 pM/l, anephric 13.4 pM/l, low creatinine clearance 19.4 pM/1) o r time to reach these concentrations (about 4 hours a f t e r beginning enflurane). The greatest serum fluoride concentration was 46.3 pM/l in a low-clearance patient. Serum fluoride decreased rapidly after termination of enflurane even in anephric patients presumably reflecting ion uptake by bone. It was concluded t h a t uptake of fluoride by bone is the major determinant of serum concentrations a f t e r low-dose enflurane anesthesia. It seems unlikely t h a t toxic serum fluoride concentrations would occur with low-dose entlurane even i n the absence af renal function. (Carter R, Heerdt M , Acchi-
ardo S : Fluoride kinetics a fte r enflurane anesthesia in healthy and anephric patients and in patients with poor renal fzmction. Clin Pharrn Thor 20:565-570, 1976)
506
Hyperglycemic Hyperosmolar Nonketotic Coma Following Diazoxide, Anesthesia and Operation BAEKHYO SHIN, MD*
SAMUEL I. JOSEPH, PhD, MDt Baltimore, Maryland$
Unsuspected hyperglycemia caused prolonged postanesthetic coma in a nondiabetic patient treated with IV diazoxide and furosemide. The combined effect of these drugs with general anesthesia and surgical stress was postulated a s
the cause of the hyperglycemia. Blood glucose should be measured during and after general anesthesia in patients who are potential candidates for hyperglycemic hyperosmolar nonketotic coma.
H
border, enlarged prostate, and a pulsatile abdominal mass.
hyperosmolar nonketotic coma (HHNKC) has been reported in known diabetic and previously unknown diabetic patients, associated with general anesthesia.'.' HHNKC has also been reported in association with the use of diazoxide in the management of resistant hypertension,"J unassociated with anesthesia. This is a report of HHNKC following a single I V dose of diazoxide and subsequent surgical anesthesia. YPERGLYCEMIC
REPORT OF A CASE A 70-year-old man was admitted to the emergency room with a chief complaint of abdominal pain. He was known to have had gastric carcinoma 3 years before, at which time gastrectomy, distal pancreatectomy, and esophagojejunostomy were performed. Since then he had been in excellent health, taking no medications. Pertinent physical findings on admission were: blood pressure 220/110, irregular pulse with a rate of 90, grade 111-IV murmur at the left sternal
Soon after admission, blood pressure rose to 2601120. Diazoxide (300 mg) was injected IV. Furosemide (40 mg) was also injected, to minimize the hypernatremia and expansion of plasma volume often associated with diazoxide administration.5 Digoxin was given, apparently in anticipation of impending congestive heart failure. Lidocaine was given I V for frequent. premature ventricular contractions (PVCs) . Blood pressure promptly fell to lSO/SO and stabilized at that level. Routine laboratory data, including blood sugar, were within normal limits. Following an aortogram which revealed an abdominal aneurysm, resection was initiated under N,O-0,-halothane anesthesia supplemented with d-tubocurarine, 4 hours after administration of diazoxide. During the 51/2-hour procedure, blood pressure was maintained between 130170 and 150180,
*Assistant Professor. f Professor.
$Department of Anesthesiology, University of Maryland School of Medicine, Baltimore, Maryland 21201. Paper received: 2/9/76 Accepted for publication: 11/2/76
Hyperosmolar Nonketotic Coma . . . Shin and Joseph
TABLE Laboratory Findings in Recovery Room Arterial blood goses ( F I O ~ ~ . " )
PH Paoz
Blood chemistry
7.41
Na'
129mEq/L
108 t o r r
C1-
94mEq/L
Pacoa
34 torr
K'
Hb sab
99.8%
C02
3.6mEq/L 28mEq/L
Standard HCO, Base excess
21
BUN
17mg
2
507
DISCUSSION Diazoxide, which has been used successfully to treat patients with severe hypertension, has been reported to increase the blood glucose level mainly by inhibiting insulin release from the pancreas and stimulating catecholamine release.6-8 Hyperglycemia may be further compounded by the concomitant injection of furosemide or thiazides, which are frequently used with diazoxide and are structurally similar."." Diphenylhydantoin is also known to cause hyperglycemia; the mechanism is unknown.10
Elevated blood glucose is often found during and following general anesthesia in response to surgical and anesthetic stress,' particularly in light anesthesia and intraabdominal operations. Although the cause of hyperglycemia was believed to be the glycogenolytic effect of catecholamines released under anesthetic stress, it is also reported that glucose intolerance may play a role to increase blood glucose in halothane In the recovery room, the patient, not yet and methoxyflurane anesthesia.l3 It has responding, was placed on a ventilator. With been suggested that glucose intolerance due vital signs stable, he remained unresponsive to decreased response to insulin may be for several hours except for occasional associated with a low cardiac output state, twitching of the hands and feet. Block-AidcC as hyperglycemia has been observed fremonitor indicated no residual effect of d-tu- quently during cardiopulmonary bypass in bocurarine. ECG revealed occasional PVCs, both diabetic and nondiabetic patients.'" but no other abnormalities. Laboratory findings at this time were as shown in the table. Since a single dose of diazoxide rarely causes severe hyperglycemia, the magnitude A diagnosis of an idiopathic convulsive in this case may well have been the result disorder following anesthesia was made. of several of the factors mentioned. InadeTwo doses of diphenylhydantoin (100 mg) quate intraoperative replacement of fluid were given IV, although the EEG did not and excessive administration of diuretics show seizure activity. Twelve hours post- could contribute to hyperosmolality. operatively, HHNKC was first and finally suspected when attention was brought to Twitching of the hands and feet during the fact that the 2-hour postoperative blood the comatose state may have represented glucose and osmolality were 942 mg % and extrapyramidal symptoms, such as those 336 mOsm/L, respectively. However, blood observed by Neary and coworkers1.' in hyglucose at this time was 463 mg % and the pertensive patients treated with diazoxide, patient began to regain consciousness. Lac- although these usually develop during longtated Ringer's solution was infused at a term therapy. Such symptoms may lead to rate of 500 ml/hr. Twenty units of regular the erroneous diagnosis of cerebrovascular insulin were given subcutaneously. Five accident or convulsive disorder. hours later, the blood glucose was 44 mg % HHNKC is more likely to occur in older and osmolality was 288 mOsm/L. Five percent glucose was administered to correct the patients with mild diabetes. Dehydration, low glucose. The patient regained conscious- steroid therapy, burns, hyperalimentation, ness completely 18 hours postoperatively, sepsis, pancreatitis, uremia, and surgical without any residual neurologic deficit. He stress are other possible contributing factors. required no further insulin during his hos- Mortality in prolonged hyperosmolar coma pital stay and was discharged from the hos- may be as high as 40 to 60 percent. Early pital without further complication 13 days recognition, appropriate administration of insulin, and replacement of water, sodium, later. pulse 90-110, and CVP 7-14 cm H,O. Estimated blood loss was 3000 ml. Fluid replacement consisted of 500 ml of D,/W and 3000 ml of whole blood. Mannitol (100 gm) and furosemide (240 mg) were given in a vigorous attempt to overcome the oliguria of the first 3 hours of operation. Total urine output following diuretics was 1100 ml for the next 2 hours.
508
ANESTHESIA AND ANALGESIA. . . Current Researches VOL.56, No. 4, JULY-AUGUST, 1977
and potassium are essential to decrease mortality.15 Early recognition will be aided by glucose determinations during and after anesthesia in patients who are potential candidates for HHNKC, especially in patients treated with diazoxide. The foregoing report is not tendered as a model of suggested management but to warn of the posibjlities and dangers.
7. Senft G: Biochemical aspect of the hyperglycemic action of diazoxide. Ann NY Acad Sci 150:242-253, 1968 8. Porte D J r : Inhibition of insulin release by diazoxide and its relation to catecholamine effect in man. Ann NY Acad Sci 150:281-289, 1968 9. Floy JM, Furman BL: Effect of single dose administration of diuretics on the blood sugar of alloxan-diabetic mice or mice made hyperglycemic by acute administration of diazoxide. Br J PharmaCOI47:124-132, 1973
REFERENCES
10. Peters BH, Samaan NA: Hyperglycemia with relative hypo-insulinemia in diphenylhydantoin toxicity. N Engl J Med 281:91-92, 1969
1. Bedford R: Hyperosmolar hyperglycemic nonketotic coma following general anesthesia. Report of a case. Anesthesiology 35:652-655, 1971
11. Clarke RSJ: The hyperglycemic response to different types of surgery and anaesthesia. Br J Anaesth 42:45-53, 1970
2. Toker P: Hyperosmolar hyperglycemic nonketotic coma, a cause of delayed recovery from anesthesia. Anesthesiology 41:284-285, 1974
12. Merin RG, Samuelson PN, Schalch DS: Major inhalation anesthetics and carbohydrate metabolism. Anesth Analg 50: 625-632, 1971
3. Harrison BDW, Rutter TW, Taylor RT: Severe nonketotic hyperglycemic precoma in a hypertensive patient receiving diazoxide. Lancet 2:599-600, 1972
13. Mills NL, Beaudet RL, Isom OW, et al: Hyperglycemia during cardiopulmonary bypass. Ann Surg 177:203-205, 1973
4. Charles SJ, Harrington AH: Nonketoacidotic hyperglycemia and coma during intravenous diazoxide therapy in uremia. Diabetes 20:501-503, 1971 5. Koch-Weser J: Drug therapy. N Engl J Med 294: 1271-1274, 1976 6. Kozman ME: Evaluation of diazoxide. JAMA 224: 1422-1423, 1973
14. Neary D, Thurston H, Pohl JEF: Development of extrapyramidal symptoms in hypertensive patients treated with diazoxide. Br Med J 3:474475, 1973 15. Arieff A, Carroll HJ: Nonketotic hyperosmolar coma with hyperglycemia: clinical features, pathophysiology, renal function, acid-base balance, plasma-cerebrospinal equilibrium and the effects of therapy in 37 cases. Medicine 51:73-94, 1972
RENAL FUNCTION AND ENFLURANE. Fluoride ion kinetics after nitrous oxideenflurane anesthesia were determined in 16 healthy control patients, 18 anephric patients and 6 patients with creatinine clearance less than 5 ml/min and undergoing chronic hemodialysis. The dose of enflurane was similar in all 3 groups. There was no significant difference between groups with respect to maximum serum inorganic fluoride concentrations (control 13.2 pM/l, anephric 13.4 pM/l, low creatinine clearance 19.4 pM/1) o r time to reach these concentrations (about 4 hours a f t e r beginning enflurane). The greatest serum fluoride concentration was 46.3 pM/l in a low-clearance patient. Serum fluoride decreased rapidly after termination of enflurane even in anephric patients presumably reflecting ion uptake by bone. It was concluded t h a t uptake of fluoride by bone is the major determinant of serum concentrations a f t e r low-dose enflurane anesthesia. It seems unlikely t h a t toxic serum fluoride concentrations would occur with low-dose entlurane even i n the absence af renal function. (Carter R, Heerdt M , Acchi-
ardo S : Fluoride kinetics a fte r enflurane anesthesia in healthy and anephric patients and in patients with poor renal fzmction. Clin Pharrn Thor 20:565-570, 1976)
