Transient Extreme Insulin Resistance in Shock During Diabetic Ketoacidosis

Endocrinol Japon 1992, 39 (6), 571-576

Transient Extreme Insulin Resistance in Shock during Diabetic Ketoacidosis

HIROKI YOKOYAMA, TARO WASADA, YUKO SHIMIZU, HIROKO YOSHINO, SUMIKO HASUMI AND YASUE OMORI Diabetes Center, TokyoWomen's Medical College, Tokyo162, Japan

Abstract. Transient extreme insulin resistance was encountered during an episode of diabetic ketoacidosis (DKA) in an insulin-treated diabetic patient. On admission, the plasma glucose level was 1241 mg dl-1 and arterial blood pH 6.895 with HCO3- 4.7mEql-1. An intravenous bolus injection of 20 units, followed by continuous infusion of 20 units h-1 of short-acting regular human insulin, was instituted. Ischemic myocardial changes were noted on the initial electrocardiogram, therefore fluid replacement was limited to 1,000ml of 0.9% saline solution in the first hour. As the plasma glucose level declined by only 203 mg dl-1 (41 mg dl-1 h-1) in the first 5h, the insulin dose was doubled every 2h. At hour 4, the patient developed circulatory shock which required vasopressor support and respiratory assistance. A plasma glucose level of 300 mg dl-1 was not achieved until the total dosage of insulin amounted to 91,580 units at hour 25. Insulin resistance was not observed from that point on. The patient had neither insulin antibodies nor anti-insulin receptor antibodies in serologic testing. The insulin binding characteristics of the patient's erythrocytes were similar to those from healthy controls

both with and without experimental acidosis and with a high level of ƒÀ-hydroxybutyrate. Among

multiple potential factors, the severe shock associated with DKA has been considered as a primary cause of the transient severe insulin resistance in this case.

Key words: Insulin resistance, Diabetic ketoacidosis, Shock. (Endocrinol Japon 39: 571-576, 1992)

INSULIN resistance has been said to exist in Case Report virtually all cases of diabetic ketoacidosis (DKA) [1, 2]. In the majority of cases, however, continuous A 65-year-old male patient with diabetes was infusion of low-dose insulin up to 10 units h-1 is admitted to our hospital because of a decreased described to be safe and effective in the treatment level of consciousness and urinary incontinence. of DKA [3, 4]. Severe insulin resistance requiring He has been diabetic since the age of 52. He more than 100 units h-1 is exceedingly rare [5, 6]. underwent partial pancreatectomy because of a We encountered a case of DKA with shock, which mucus-producing adenoma at the age of 62, at required an extremely large dose of insulin: which point insulin therapy was begun. 91,580 units over a 25-h period, for resolution. Eleven days prior to admission, the patient Several potential mechanisms for the severe in- noticed a mild sore throat, cough and low grade sulin resistance were evaluated. fever. Following these symptoms, he became anorectic and omitted his daily insulin injection of 14 units (0.31 units kg-1 of Humulin NR). One day prior to admission, he became drowsy and urinary Received: June 29, 1992 incontinence was noted. The patient was admitted Accepted: September 18, 1992 to our hospital on Nov. 13, 1988. On admission, he Correspondence to: Dr. Hiroki YOKOYAMA, Diabetes Cen- ter, Tokyo Women's Medical College, 8-1 Kawada-cho, Shin- was lethargic and appeared dehydrated. Vital juku-ku Tokyo 162, Japan. signs revealed a blood pressure of 92/60 mmHg, 572 YOKOYAMA et al.

heart rate of 76min-1 and body temperature of Table 1. Laboratory data on admission 35•Ž. Arterial blood gas analysis revealed a pH of

6.895, HCO3- of 4.7mEql-1. Urinary ketones were 2+ on the dipstick. The plasma glucose concentration was 1,241mgdl-1, CPK 8,095mU ml-1 (normal range: 8-80), myoglobin>500ng ml-1, blood urea nitrogen 68.0mgdl-1, creatinine

2.5mgdl-1, sodium 123mEql-1, and potassium 5.0mEql-1 (Table 1). Analysis of these laboratory data indicated diabetic ketoacidosis accompanied by rhabdomyolysis and acute renal failure. A chest X-ray showed an infiltrate in the upper and middle lobes of the right lung, suggesting acute

pneumonia. The electrocardiogram showed ischemic myocardial changes with decreased R waves in leads I, aVL and V1-5 and ST segment depression in II, III, aVF and V4-5. The clinical course for the first 2 days is depicted in Fig. 1. With the diagnosis of DKA, associated with acute renal failure secondary to rhabdomy-

Fig. 1. Clinical course (I) of a 65-y.o. male with diabetic ketoacidosis and shock, showing laboratory data on the changes in plasma glucose levels, arterial blood pH, serum ketone with ƒÀ-hydroxybutyrate

(B-OH), central venous pressure (CVP), volume replacement and insulin doses. EXTREME INSULIN RESISTANCE IN DKA 573

Fig. 2. Clinical course (II), showing clinical signs and data related to acute renal failure, pneumonia and disseminated intravascular coagulation (DIC).

olysis, a bolus injection of 20 units, followed by creased miosis, and the patient developed circula- continuous infusion of 20 units h-1 of short-acting tory shock at 4h. Dopamine hydrochloride at a human insulin (Humulin RR), was started intra- rate of 5ƒÊgkg-1min-1 was initiated to maintain venously. Since the central venous pressure (CVP) systolic blood pressure >90mmHg, and respira- of 8-11 cmH2O was within upper limits of normal tory assistance was eventually required with 100% and there was evidence of both acute renal failure oxygen by mechanical ventilation to avoid severe and ischemic heart disease, overhydration was hypoxia. carefully avoided. Intravenous fluids were limited Antibiotics and gabexate mesilate were adminis- to 1,000ml of 0.9% saline solution for the first h, tered for pneumonia and accompanying dissemi- 1,800ml for the next 2-h period and a total of nated intravascular coagulation (DIC), and the

3,200ml for the first 5h. This resulted in a pneumonia and state of shock resolved. Dopamine positive fluid balance of 2,800ml (urine volume hydrochloride and mechanical ventilation were 400ml 5h-1) for the first 5h. Because the blood discontinued on the sixth and tenth hospital day, glucose level declined by only 203mgdl-1 (41mg respectively. As shown in Fig. 2, the patient dl-1h-1) in the first 5h, the insulin dose was regained consciousness on the fifth hospital day. doubled every 2h and amounted to 20,000 units CT scan of the brain, performed on the tenth in a 2-h period. The blood glucose level declined hospital day, revealed only minimal cerebral atro- to 300mgdl-1 by 25h and the insulin infusion phy. Rhabdomyolysis with acute renal failure also was discontinued. The cumulative dose of insulin improved gradually. The patient was discharged 2 was 91,580 units. Restoration of acidosis was also months later, and has been stable with adequate retarded; arterial blood pH rose to 7.2 at hour 14 glycemic control on 22 units of daily insulin (0.49 and did not reach 7.35 until 8h. Over the units kg-1). following three days, the blood glucose level was maintained at 200-300mgdl-1 without insulin infusion. Methods and Results Despite the initiation of the intensive therapy, the level of consciousness deteriorated with in- Antibodies to insulin and its receptor were 574 YOKOYAMA et al.

considered as contributing to the insulin resistance. Insulin antibodies were measured by a standard method using polyethylene glycol [7] and

found to be 2.7% (normal range <5%) in this

patient. Anti-insulin receptor antibodies were ex- amined by the method described by Gambhir et al.

[8] using rabbit erythrocytes as insulin receptors. As shown in Fig. 3, the rabbit erythrocytes did not show any alteration in 125I-insulin binding characteristics after preincubation with patient's serum

or its ƒÁ-globulin fraction which was derived by Sephadex G-50 chromatography. It was therefore

felt that the patient's serum was negative for anti-insulin receptor antibodies.

The possibility that the patient's insulin receptor

could be altered by the lowered pH was then evaluated. As shown in Fig. 4, a pH ranging from

8.0 to 6.8 exerted almost identical effects on insulin binding by erythrocytes from both this

patient and healthy controls. ƒÀ-hydroxybutyrate (ƒÀ-OH), the major ketoacid of diabetic ketoacidosis, also did not influence the insulin binding to the

patient's erythrocytes even at the concentration Fig. 3. Insulin binding to rabbit erythrocytes, showing found in this patient (data not shown). 125I-insulin binding to rabbit erythrocytes pretre- Insulin degrading activity of the whole serum ated with patient's whole serum or its protein was evaluated by determining the % precipitation fraction after Sephadex G-50 column chromatogra- of 125I-insulin with 10% TCA after preincubation phy, and a known serum positive for anti-insulin receptor antibodies. Insulin binding assay was per- with serum. This was found to be 85.3% for the formed by a method previously described [8]. patient's serum drawn on Nov. 14 and 90.2% for a healthy control. This small difference observed seems unlikely to account for the severe insulin

resistance observed in this patient. Finally we evaluated the biologic activity of the

patient's serum insulin. An insulin fraction obtained by Sephadex G-50 chromatography of the patient's serum dated Nov. 14 was injected

intravenously into a rabbit. This contains an insulin dose approximating 0.15 units per kg of

body weight and lowered the blood glucose level from a baseline of 79mgdl-1 to 32mgdl-1 at 32 min following the injection. It was therefore felt

that the patient's insulin injected by us had true

biologic activity. Plasma levels of glucagon (212.4pgml-1), adrenalin (0.3ngml-1) and cortisol (92.4ƒÊgdl-1)

Fig. 4. Effect of pH on insulin binding to patient's erythro- were high but all remained within the range cytes. Patient's erythrocytes obtained on Jan. 20, previously reported in DKA [1]. 1989 and those from three control subjects were compared for 125I-insulin binding ability under decreasing from 8.0 to 6.8 pH of incubation media. Insulin binding assay was performed by a method previously described [8]. EXTREME INSULIN RESISTANCE IN DKA 575

deranged tissue microcirculation [14]. Under Discussion these conditions, circulating insulin might not We have presented a case which showed re- reach insulin-sensitive tissues such as skeletal markable insulin resistance, requiring 91,580 units muscle, and hence result in the profound unre- of insulin over a 25-h period for the treatment of sponsiveness to insulin. DKA. Through experimental studies, immune An additional potential cause of the insulin resistance was excluded by the absence of anti- resistance in this case is post-insulin receptor bodies to both insulin and its receptors. There was resistance of unknown mechanism. It was re- neither increased sensitivity of the patient's insulin ported that impairment in the post-binding level receptor to lowered blood pH nor rapid insulin of insulin receptor contributes to insulin resistance degradation in the circulation as reported pre- in the septic state [15]. This post-insulin receptor viously [9]. resistance does not respond to an increasing Several factors were considered to be causally concentration of plasma insulin [16]. A similar related to the resistance to massive dose of insulin response is observed to both low and high doses of observed in this patient. Prompt and adequate insulin because in both treatment schedules free fluid replacement constitutes a critical portion of insulin is enough to saturate the receptors. Despite therapy in DKA. It has been reported that approx- the lack of definitive evidence, the transient nature imately 3l should be replaced during the first 2h of the insulin resistance and the lack of significant of therapy in adults with severe dehydration response to maximally challenging doses of insulin [10-12]. In our patient only 1,000ml of 0.9% point to the involvement of a post-insulin receptor saline solution was replaced carefully in the first h resistance in this case. because of the presence of ischemic heart disease. In conclusion, the extreme insulin resistance Therefore fluid replacement at the beginning of observed in this case may be explained by multiple therapy may have been inadequate and resulted in factors. A severe state of shock may have been deranged peripheral microcirculation. mainly attributed to the insulin resistance. Other Secondly, circulatory shock which required both factors include inadequate initial fluid replace- dopamine to maintain blood pressure >90mmHg, ment, and the post-insulin receptor resistance to and mechanical ventilation to avoid severe hypoxia insulin action. The latter may be caused, at least in might have played a certain role in the insulin part, by the presence of severe pulmonary infec- resistance. The insulin dose was escalated after the tion [17, 18]. initiation of dopamine and amounted to 10,000 to 20,000 units in a 2-h period while receiving 100% oxygen by mechanical ventilation. It has been Acknowledgment described that the magnitude of hyperglycemia is related to the severity of the shock, and that serum The helpful advice of both Roger H. Unger, insulin in man is unresponsive to the hyper- M.D., Professor of the University of Texas, South- glycemia of shock [13]. Acidosis itself also lowers western Medical Center, and Yasuko Uchigata, the insulin binding to the receptor. In addition, M.D., Tokyo Women's Medical College, is greatly there was a severe pulmonary infection accompa- acknowledged. nied by DIC in this case, which may have further

References 1. Ginsberg HN (1977) Investigation of insulin resist- The efficacy of low-dose versus conventional ther- ance during diabetic ketoacidosis: role of counter- apy of insulin for treatment of diabetic ketoacido- regulatory substances and effect of insulin therapy. sis. Ann Intern Med 84: 633-638. Metabolism 26: 1135-1146. 4. Padilla AJ, Loeb JN (1977) Low-dose versus high- 2. Barret EJ, DeFronzo RA, Bevilacqua S, Ferrannini dose insulin regimens in the management of E (1982) Insulin resistance in diabetic ketoacidosis. uncontrolled diabetes, a survey. Am J Med 63: Diabetes 31: 923-958. 843-848. 3. Kitabuchi AE, Ayyagari V, Guerra SMO (1976) 5. Shade DS, Eaton RP, Alberti KGMM, Johnston DG 576 YOKOYAMA et al.

(1981) Insulin resistance. In: Diabetic Coma: New Hyde Park, NY: 109-121. Ketoacidotic and hyperosmolar. University of New 12. Foster DW, McGarry JD (1983) The metabolic Mexico Press, Albuquerque: 116-124. derangements and treatment of diabetic ketoacido- 6. Naghmi R, Kaboor M (1988) Transient severe sis. N Engl J Med 309: 159-169. insulin resistance in diabetic ketoacidosis. Horm 13. Carey LC, Lowery BD, Cloutier CT (1970) Blood Metabol Res 20: 584-586. sugar and insulin response of humans in shock. 7. Nakagawa S, Nakayama H, Sasaki T, Yoshino K, Ann Surg 172: 342-350. Ying Y, Shinozaki K, Aoki S, Mashimo K (1973) A 14. Lang CH, Dibrescu C (1989) In vivo insulin simple method for the determination of serum resistance during nonlethal hypermetabolic sepsis. free insulin levels in insulin-treated patients. Di- Circ Shock 25: 165-178. abetes 22: 590-600. 15. Igarashi M, Yamatani K, Fukase N, Daimon M, 8. Gambhir KK, Archer JA, Bradley CJ (1978) Ohnuma H, Takahashi H, Manaka H, Tominaga Characteristics of human erythrocyte insulin re- M, Sasaki H (1992) Sepsis inhibits insulin- ceptors. Diabetes 27: 701-708. stimulated glucose transport in isolated rat adipo- 9. Williams G, Pickup JC, Keen H (1987) Massive cytes. Diabetes Res Clinic Pract 15: 213-218. insulin resistance apparently due to rapid clear- 16. Kahn CR (1978) Insulin resistance, insulin insensi- ance of circulating insulin. Am J Med 82: tivity and insulin unresponsiveness: A necessary 1247-1252. distinction. Metabolism 27: 1893-1902. 10. Vignati L, Asmal AC, Black WL, Brink SJ, Hare 17. Raymond RM, Klein DM, Gibbons DA, Jacobs HK, JW (1985) Coma in diabetes. In: Marble A, Krall Emerson TE (1985) Skeletal muscle insulin unre- Lp, Bradley RF, Christlieb AR, Soeldner JS (eds) sponsiveness during chronic hyperdynamic sepsis Joslin's Diabetes Mellitus. Lea & Febiger, Phi- in the dog. J Trauma 24: 845-855. ladelphia: 526-552. 18. Drobny EC, Abramson EC, Baumann G (1984) 11. Matz R (1987) Uncontrolled diabetes mellitus: Insulin receptors in acute infection: A study of diabetic ketoacidosis and hyperosmolar coma. In: factors conferring insulin resistance. J Clin Endocri- Bergman M (ed) Principles of Diabetes Manage- nol Metab 58: 710-716. ment. Medical Examination Publishing Co. Inc