1090 Short Communication: Diabetic Coma. Canad. Med. Ass. J. Dec. 7, 1968, vol. 99. SHORTCOMMUNICATION. Hyperosmolar, Non-Ketotic Diabetic Coma.
1090 Short Communication: Diabetic Coma
Canad. Med. Ass. J. Dec. 7, 1968, vol. 99
SHORT COMMUNICATION Hyperosmolar, Non-Ketotic Diabetic Coma W. A. MAHON, F.R.C.P.[C], M. B. UROWITZ,
fT\HE clinical syndrome of diabetic ketoacidosis -*¦ and its therapy have been long known and are familiar to all physicians. However, in 1957 Sament and Schwartz1 reported a case of "severe diabetic stupor without ketosis", and thus de¬ fined another form of diabetic coma. Subse¬ quent reports from many centres have estab¬ lished the major components of this syndrome
extreme hyperglycemia, serum hyperosmolality, dehydration, and the absence of ketosis.24 Early diagnosis of this syndrome and prompt initiation of appropriate therapy are prerequisites for recovery. Failure to recognize this form of diabetic coma probably accounts for the 46% mortality rate which was associated with the first 63 reported cases.3 Three cases illustrating some of the features of this entity have recently been treated at the Toronto General Hospital. Because it is likely that this condition occurs more frequently than is generally appreciated, it has been thought desirable to report these cases, emphasizing some of the aspects of clinical presentation and management, particularly where they differ from as
those of conventional diabetic ketoacidosis. Case Reports
Case 1..J.F., a 68-year-old man, had been a known diabetic for 14 years, during which time he had been taking 40 units of lente insulin daily. There were no previous episodes of hyperglycemia severe enough to require admission to hospital al¬ though he had had several hypoglycemic reactions. For more than 10 years he had exercised no dietary restrictions whatsoever. Four days before admission he developed slurring of his speech, polydipsia and polyuria. By the next morning, however, he felt well again. On the day before admission the slurred speech, polydipsia and polyuria recurred and he became increasingly con¬ fused. He was brought to the emergency depart¬ ment of the Toronto General Hospital. On examination, he was found to be profoundly dehydrated and in a stuporous state from which he could be roused. There was a prominent flapping From the Toronto General Hospital, Toronto, Ontario. Reprint requests to: Dr. W. A. Mahon, Department of Pharmacology, Toronto General Hospital, 101 College
Street, Toronto 2, Ontario.
J. HOLLAND, M.D. and M.D., Toronto table i.
Initial laboratory values
Case 1
Case 2
Case S
Glucose.
750 895 (mg. per 100 ml.) 1000 BUN. (mg. per 100 ml.) 49 82 39 Sodium. 145 144 127 (mEq. per litre) Potassium. 5.2 3.3 7.7 per litre) (mEq. Chloride. 106 115 92 (mEq. per litre) 15 22 18 CO2 content. (mEq. per litre) Serum ketones. Neg. Neg. Neg. Urine ketones. Neg. Neg. Neg. 364 359* 318* Osmolality. (mOsm. perkg.) Arterial pH. 7.35 7.31 7.32 Therapy first 24 hours Fluids (litres). 10.0 2.3 8.0 140. 250. 355. Regular insulin (units). Progress Time for osmolality to
become normal.
Time to become mentally normal.
*Calculation:
.
Less Less Less than 24 than 24 than 24 hours hours hours 4
.
days
2
days
osmolality + BUN + 2.8 (Normal serum osmolality 285- 295 mOsm. per kg.) 2 X sodium
serum
4
days
Blood
glucose 18
=
of the hands and the upper limbs. He was and there was no detectable odour of acetone. The results of the initial labora¬ tory investigations are set forth in Table I. During the first 24 hours he was given 10 litres of fluid intravenously, of which the first 4.0 litres were hypotonic (half normal) saline. Supplemental potas¬ sium chloride was given. Insulin administered during this period totalled 140 units. Within 12 hours the blood glucose, BUN and electrolytes were all essentially normal, but the patient's confused state persisted. Twenty-four hours after admission he be¬ came hallucinated and this state lasted for a further 12 hours. His intellectual function remained im¬ paired for a considerably longer period, and it was not until four days after admission that his usual mental state returned. The creatinine clearance was 80 ml. per min. 19 days after admission. This patient was discharged on the 31st hospital day on a regimen of 30 units of lente insulin daily. tremor not
hyperventilating,
Case 2..H.G., a 37-year-old woman, had been a diabetic since 3 years of age, requiring insulin for control of her hyperglycemia. Her most recent in¬ sulin therapy had been 28 units PZI and 10 units NPH insulin. Her control had always been quite erratic, and she had had numerous insulin reactions but no documented previous hyperglycemic coma. Four days before admission she began vomiting and took only fluids. She then became progressively more drowsy, and finally was brought to the hospital, where she was found to be stuporous, and at times
Canad. Med. Ass. J. Dec. 7, 1968, vol. 99
thrashing about. The ocular fundi showed early diabetic retinopathy. Her breathing was not laboured. The skin was dry and had a doughy consistence, and the eyeballs were soft and sunken. There were no focal neurologic signs. The initial laboratory values and calculated serum osmolality appear in Table I. During the first 24 hours she received 2.5 litres of normal saline intravenously, 2 litres of tap water by nasogastric tube, 3.5 litres of 5% glucose and water intravenously and a total of 250 units of regular insulin. Supplementary potassium chloride was given. On this therapy the patient's condition improved and in 24 hours her laboratory results had returned to normal. Her mental state remained somewhat clouded for an¬ other 24 hours, but after a further two days it was restored to normal. This patient was discharged on the 15th hospital day on 50 units of lente insulin. Case 3..T.E., a 50-year-old woman, had been a diabetic for three years, requiring 20 units of lente insulin daily. There was no history of hyperglycemic or insulin coma. This patient had many other problems including schizophrenia, diffuse bronchiectasis, alcoholism and an
undiagnosed anemia.
She was brought to the emergency department of the Toronto General Hospital, where she was thought to have taken an overdose of drugs, pos¬
sibly including ethchlorvynol (Placidyl), diazepam (Valium) and chlordiazepoxide (Librium). She was comatose and hypotensive (systolic blood pres¬ sure 50 to 60 mm. Hg). Laboratory studies revealed a blood glucose of 534 mg. per 100 ml., BUN of 36 mg. per 100 ml. and serum and urine negative for ketones. She was intubated, ventilated, hydrated and given 25 units of regular insulin. Following this therapy she woke up and became responsive. How¬ ever, approximately four to six hours later she again became stuporous. Laboratory studies and calcu¬ lated osmolality at this time are listed in Table I. At this point the patient's stupor was attributed primarily to the hyperosmolar non-ketotic diabetic state. A complicating feature was the renal failure and marked oliguria which ensued owing to the prolonged period of hypotension she had suffered before being brought to the hospital. Consequently, her fluid replacement was modified according to her urine output. She received a total of 2.3 litres of fluid intravenously and 355 units of insulin with¬ in a period of 24 hours. With this therapy her blood glucose and serum osmolality returned to within normal limits. However, her mental state remained somewhat clouded for approximately three more days, returning to normal after about 96 hours. This patient subsequently underwent peritoneal dialysis to manage the oliguric phase of her renal failure, and a tracheostomy to improve her pul¬ monary function. She tolerated all of these procedures extremely well. After 21 days her urinalysis and BUN were normal. She was discharged on the 44th hospital day on 30 units of lente insulin.
Short Communication: Diabetic Coma 1091 Discussion
The feature of hyperosmolar coma which disit from the diabetic keto-acidotic coma is the absence of ketones. This was evi¬ dent in the three patients described, as their serum and urine acetone tests were consistently
tinguishes
negative.
Clinically,
the striking features of this con¬ the extreme dehydration and the stuporous mental state from which the patients can be roused. Unlike keto-acidotic coma, there is no Kussmaul respiration and no acetone odour on the breath. Various neurologic abnormalities associated with this syndrome have been des¬ cribed.5 Some of the reported cases have suf¬ fered focal and grand mai seizures, and in some, various pareses have been noted, either follow¬ ing or independent of such seizures. It is of interest that in patient J.F. (Case 1) there was a very marked coarse flap of the upper limbs which resembled closely the asterixis associated with hepatic coma. Such a tremor has not been previously described in this condition. The relationship between these neurologic abnor¬ malities and the hyperosmolar state is not by any means clear, although it seems likely that cerebrospinal fluid hyperosmolality may be an important factor. Although studies on CSF electrolytes and osmolality were not carried out in these patients, such serial measurements would contribute to further understanding of the neu¬ rologic state. It is worth emphasizing that the clouded mental state of the patients may continue for a prolonged period after the blood biochemical abnormalities have been corrected. The three patients described here did not return to their normal mental state until two to four days after the onset of the stupor. In patient J.F. (Case 1) well-organized hallucinations de¬ veloped 24 hours after his admission to hospital, when most of the blood biochemical abnormali¬
dition
are
ties were no longer present. In many of the initial reports of this state
of diabetes mellitus
a
uncommon.2* 3 history Indeed, in 1965, in a review of the world litera¬ ture, Di Benedetto, Crocco and Soscia2 reported 24 such cases and found that only five had been known diabetics. The present three patients were all known diabetics, one of juvenile-onset type and two of maturity-onset type. All three had been on maintenance doses of insulin before admission and all required insulin at the time of
discharge.
was
A consistent laboratory finding in hyperosmo¬ coma is the marked elevation of blood glu¬ cose, ranging from 750 mg. to 1000 mg. per 100 ml. in these three cases. In previous reports, blood glucose levels have been as high as 2200
lar
1092
SHORT COMMUNICATION:
DIABETIC COMA
Canad. Med. Ass. J. 7, 1968, vol. 99
mg. per 100 mi.6 Although ketosis is conspicuously absent in this condition, arterial pH values have varied from normal to acidotic levels. The low p11 might be attributed to associated lactic acidosis or renal failure. As lactate levels have not been reported in detail in any of these case reports, it would be worth while making this measurement in any further studies. Patient T.E. (Case 3) was severely oliguric and azotemic at the time her arterial pH was recorded as 7.22. Patient J.F. (Case 1) did indeed have an anion gap of 26.0 mEq. per litre; however, lactic acid levels were not obtained. An elevated blood urea nitrogen is often found in these patients, and this was so in the three cases reported here. It is noteworthy, however, that following treatment renal function became normal. Because of the profound dehydration associated with the hyperosmolar state there is little doubt that initial therapy should consist of large quantities of fluid administered intravenously over a short period of time. There is some controversy as to whether the initial intravenous solution should be hypotonic or isotonic saline. Because of the excessive osmolality of the plasma, normal saline will be hyposmolar to the patient's plasma, but nonetheless there is reason to believe that the deficit of water is so marked that the initial therapy should be hypotonic (half normal) saline. Hypotonic fluids in the form of tap water can be administered by nasogastric tube, as in patient H.G. (Case 2). However, the value of this form of therapy must be weighed against the complications of an indwelling nasogastric tube in stuporous patients. Case 3 was complicated because of the onset of renal failure and marked oliguria which prevented the administration of large quantities of fluid. There seems to be no place in the initial therapy for the administration of intravenous
5% glucose and water. In reported cases the volume of fluid necessary to treat dehydration ranges from 6 to 16 litres in the first 24 hours. As in conventional ketoacidosis associated with diabetic coma, supplementary potassium should be administered. It is obvious from the cases reported here that there is considerable variation in the quantity of insulin required to restore the high blood glucose to normal. Patient J.F., who had the highest blood glucose of the three cases described, indeed required the smallest quantity of insulin and was given only 140 units over the first 24 hours. Previous reports have been at variance as to the amount of insulin required, some stating that large doses are necessary, and others that relatively small doses are sufficient. However, it is obvious that considerable attention must be paid to individual variations in the insulin requirements in this type of diabetic coma. It is believed that a greater awareness of hyperosmolar non-ketotic diabetic coma and its various manifestations will promote earlier diagnosis and treatment of this state, and consequently reduce the excessively high mortality reported from this condition. We wish to thank Dr. C. R. Burton for permission to report the clinical details of patients admitted to Ward I of The Toronto General Hospital. REFERENCES 1. SAMENT, S. AND SCHWARTZ, M. B.: S. Air. Med. J.,
31: 893. 1957. Arch. Intern. Med. (Chicago), 116: 74, 1965.
2. DI BENEDETTO, R. J., CR0000, J. A. AND SOsCIA, J. L.: 3. SCHWARTZ, T. B. AND APFELBAIJM, R. I.: Nonketotic diabetic coma. In: The year book of endocrinology (1965-1966 Year Book series), edited by T. B.
Schwartz, Year Book Medical Publishers Inc., Chicago, 1966, p. 165. 4. MARTIN, A.: Po8tgrad. Med. ,J., 44: 218, 1968. 3. MACCARlO, M., MEssIs, C. P. AND VASTOLA, E. F.: Neurology (Minneap.), 15: 195, 1965. 6. RosslER, P. H., RETJTTER, F. AND FRICK, P.: Deut8ch. Med. Wschr., 86: 2145, 1961.
