Search

Diabetic coma is the most severe form of hyperglycaemic metabolic disorders. The post-mortem diagnosis of this disorder of glucose metabolism can be difficult and vague due to a lack of characteristic morphological findings. Six death cases caused by diabetic coma are described with special focus on biochemical (and histological) findings. The possible glycaemia markers glucose, lactate, HbA1c, fructosamine, anhydroglucitol, and ketone bodies were measured and the usefulness of these parameters is evaluated and discussed. Estimations of glucose concentrations in vitreous humour or cerebrospinal fluid and of ketone bodies in blood or other matrices are obligatory while measurements of HbA1c, fructosamine, or anhydroglucitol can only provide additional information on the long-term adjustment of diabetes in the deceased. Lactate concentrations (addition of glucose and lactate levels to form the sum formula of Traub) do not give more information than the glucose concentration itself and can be therefore omitted., (Copyright © 2013 John Wiley & Sons, Ltd.)

In Germany, approximately 8 million people suffer from diabetes mellitus, of which over 90% have type 2 diabetes. Worldwide, there are around 246 million people with diabetes mellitus. As diabetes is often symptomless for a long time, around 2.5% of the population have diabetes without knowing it. Diabetic coma is a life-threatening complication, which affects up to 1.2% of all diabetics annually and is associated with a mortality rate of 20-25%. Since in 25% of patients the diabetic coma is the first manifestation of the disease, a valid post-mortem biochemical method for the diagnosis of the metabolic crisis is very important. The article presents cases from the authors' own autopsy material showing that the analysis of the vitreous humour is indispensable for the reconstruction of the events leading to death even in cases in which the cause of death is seemingly clear.

Objective: To describe a 24-year-old patient with immune-mediated primary adrenal insufficiency and type 1 diabetes mellitus (T1DM) receiving intensive diabetes management who was comatose at presentation attributable to severe hypoglycemia and had residual dysphasia after recovery and to summarize the related literature., Methods: We present a case report and the findings on systematic review of the pertinent literature to identify the cumulative incidence of severe hypoglycemia with use of intensive insulin therapy in patients with primary adrenal insufficiency and T1DM and to determine the incidence of dysphasia after severe hypoglycemia., Results: After 5 days of mechanical ventilation, our patient was revived. He had severe dysphasia after recovery of consciousness. Magnetic resonance imaging of the brain revealed encephalomalacia in the left temporal, frontal, and parietal lobes. After 6 years of follow-up, he continues to have residual deficits of expressive dysphasia and difficult-to-control seizures but no other neurologic disorders. Systematic review of the literature revealed that studies from the 1950s reported mortality due to hypoglycemia in such a cohort, but no recent studies have described the cumulative incidence of severe hypoglycemia in a cohort of patients with primary adrenal insufficiency and T1DM. To the best of our knowledge, we report the first findings on magnetic resonance imaging of the head in such a patient., Conclusion: Fortunately, residual dysphasia is an infrequent outcome after severe hypoglycemia.

Summary: We describe 2 cases of diffusion-weighted (DW) MR imaging in hypoglycemic coma. One patient, with diffuse cortical lesions, had a poor outcome, but the other, with transient white matter abnormalities, made a complete recovery. The distinctive patterns of DW MR imaging abnormalities in hypoglycemic patients should be recognized and may be a predictor of clinical outcome.

Insulin-induced hypoglycemic coma in animals with alloxan diabetes was observed at a higher basal glucose level in the blood compared to healthy animals. It was associated with inhibition of glycolysis and glycogenolysis and decreased activities of succinate dehydrogenase and glutamate dehydrogenase in the cerebral hemispheres and brainstem.

The biochemical test of the vitreous body (VB) may be used in post-mortem diagnosis of diabetes mellitus and diabetic coma. Concentrations of glucose, lactate, keton bodies in the VB of the eye do not depend on duration of post-mortem period. Methods of diagnosis of hyperglycemic, hypoglycemic and ketoacidotic comas in the postmortem period are proposed. VB glucose over 17 mmol/l is a specific marker indicating death due to diabetic coma with hyperglycemia. Blood lactate under 16 mmol/l and glucose absence in the VB specifically mark death of hypoglycemic coma. In death of diabetic coma with ketoacidosis, a sharp rise in the level of VB ketonic bodies was observed.

Hypoglycemia is a classic cause of coma and can result in irreversible neuronal loss. Until now, the main prognostic factors were depth of hypoglycemia and duration of coma. We report the case of a 55-Year-old woman who suffered severe hypoglycemic coma with abnormal cortico- subcortical diffusion weighted MR images. These MRI abnormalities preceded severe atrophy of these cerebral areas. This findings suggests that diffusion abnormalities in hypoglycemic coma may be related to neuronal loss and may thus have prognostic value.

Aims: The present study was initiated by a very recent histochemical observation of lipid accumulation in the renal cortex of a woman who died in a diabetic coma. Two older reports of lipid accumulation in the kidneys of patients who died, most likely in a state of non-regulated diabetes, supported this observation. We have examined whether lipid accumulation in the renal cortex is characteristic of diabetic coma and, if so, which type of lipid accumulates., Methods: Three groups were studied. Ten subjects who died in a diabetic coma, eight diabetics who died of known causes unrelated to diabetes, and seven normal control subjects without any diagnosed diabetes who died of known causes. All were subjected to histological examination for lipid accumulation in the renal cortex. Detailed analysis of cortex lipids was performed for two of the subjects who died in a diabetic coma and all diabetic controls as well as non-diabetic control subjects., Results: All subjects who died in a diabetic coma showed vacuolar lesions staining strongly for lipid in the proximal tubules. Neither normal controls nor non-coma diabetics showed these lesions. Compared with normal controls, renal cortex lipid was about tripled in the two analysed diabetic coma subjects due to 60-100-fold increases of triglycerides. The non-coma diabetics did not differ from the other controls with respect to triglycerides or other types of lipid, except that cholesteryl esters were elevated, though still a quantitatively minor component., Conclusion: Our findings strongly indicate that vacuolar lesions in the proximal tubules are characteristic of diabetic coma and that they are caused by accumulated triglycerides. Therefore, histological examination of renal cortex using a lipid stain may be a useful forensic tool in establishing diabetic coma as the cause of death.

Vacuolization of the renal tubular epithelial cells (the Armanni-Ebstein lesion) associated with diabetic hyperglycemia is usually regarded as an accumulation of glycogen. In a case of death of diabetic coma, the vacuoles were stained strongly for lipids. This observation may have both clinical and therapeutic consequences, and may increase our knowledge of the metabolism in diabetes.

In the period from January 1986 through April 1993, 47 cases with diabetes mellitus were autopsied at the Institute of Forensic Medicine, Odense University. In 26 cases, the diabetes had been treated with insulin, in 21 cases with oral medication or diet only. In eight insulin-dependent cases, tubular vacuolation was found in the kidneys, the so-called Armanni-Ebstein lesions. The circumstances of death and postmortem analyses of blood, urine and/or vitreous humor supported a presumed diabetic coma as the cause of death in these eight cases. Of the remaining 39 cases, six were too putrefied for histologic examination. In the remaining 33 cases and in a series of 20 non-diabetics, the cause of death was ascertained as illness (other than diabetes), traffic accidents, drowning or intoxication. In none of these cases was a diabetic coma suspected, and none of these cases showed tubular vacuolation in the kidneys. The authors conclude that tubular vacuolation of the kidneys strongly indicates death in diabetic coma.

Rhabdomyolysis is a rare but potentially fatal complication of hyperosmolar states. We report a case of severe hyperosmolar non-ketotic diabetic coma causing rhabdomyolysis in a young man. Despite very high levels of creatine kinase there was no detectable myoglobinuria. Creatine kinase estimation should be a standard investigation in all patients presenting with a hyperosmolar state.

Following stress such as heat shock or transient cerebral ischemia, global brain protein synthesis initiation is depressed through modulation of eucaryotic initiation factor (eIF) activities, and modification of ribosomal subunits. Concomitantly, expression of a certain class of mRNA, heat-shock protein (HSP) mRNA, is induced. Here we report that the activity of eucaryotic initiation factor-2 (eIF-2), a protein that participates in the regulation of a rate-limiting initiation step of protein synthesis, transiently decreases following insulin-induced severe hypoglycemia in the rat brain neocortex. Expression of HSP 72, a 72-kDa HSP, in surviving neurons was seen at 1-7 days of recovery following 30 min of hypoglycemic coma, but not at 1 h and 6 h of recovery. In the neocortex, HSP 72 was first seen in layer IV, and later also in surviving neurons in layer II. In the CA1 region and in the crest of dentate gyrus, HSP 72 expression was evident in cells adjacent to irreversibly damaged neurons. In the CA3 region and the hilus of dentate gyrus, HSP 72 was expressed in a few scattered neurons. In septal nucleus, HSP 72 was expressed in a lateral to medial fashion over a period of 1-3 days of recovery. We conclude that severe insulin-induced hypoglycemia induces a stress response in neurons in the recovery phase, including inhibition of protein synthesis initiation, depression of eIF-2 activity, and a delayed and prolonged expression of HSP 72 in surviving neurons. The HSP 72 expression may be a protective response to injurious stress.

The case history and the autopsy findings of a 44-year-old women who died shortly after her flight to Khartoum (Sudan) in a hospital is presented. The clinical diagnosis in Africa was "shock caused by gastroenteritis". The body was embalmed thoroughly and brought back to Germany. The autopsy was performed just prior to the cremation (according to "section 3 Abs. 2 Nr. 2 Feuerbestattungsgesetz"). Morphological findings (nodular glomerulosclerosis and glycogen nephrosis with Armanni-Ebstein-cells) and postmortem biochemical analyses of vitreous humour led to the diagnosis of a hyperglycaemic coma. The fatal course might have been prevented by sufficient health information to the patient (who suffered from type 1 diabetes); the diagnostic errors in the hospital could have easily been avoided by careful anamnesis and diagnostic procedures.

The pancreatic autopsy findings of 11 children with Type 1 (insulin-dependent) diabetes mellitus are presented. Nine children died within 24 h of initial presentation. In these 'recent-onset' diabetic patients many islets were shrunken and insulin-deficient. However, large islets containing B cells were present also. Insulitis was present in eight recent-onset diabetic subjects; in these, 18% of insulin-containing islets were inflamed, but only 1% of insulin-deficient islets were thus affected. This finding supports the concept of an immunologically mediated destruction of B cells in the pathogenesis of Type 1 diabetes. Severe acinar cell atrophy was present surrounding insulin-deficient islets, but acinar tissue around insulin-containing islets was normal. These exocrine changes are thought to be related to islet-exocrine vascular connections and the effects of the various islet hormones on pancreatic acini.

This study was performed to examine the relationship between postmortem biochemical values and cause of death. The follow samples were taken from 399 corpses: cerebrospinal fluid (CSF; n = 376, suboccipital), blood (n = 158, femoral vein), and urine (n = 101, at autopsy). (See Table 1 for causes of death) All samples were stored at -80 degrees C. A further 100 samples of blood were later taken and stored at +4 degrees C before testing. Biochemical determinations made were: glucose in CSF, blood, and urine (hexokinase method); lactate (LDH/GPT) and free acetone (HS-gas chromatography) in CSF; hemoglobin A1 in blood (microcolumn technique). In 34 cases fatal diabetic coma was considered verified by morphological and chemical findings. One hundred cases of sudden cardiac death were chosen as the main control group. In 32 of the 34 cases defined above, the value of the formula of Traub (glucose + lactate in CSF) exceeded 415 mg/dl. It is not influenced significantly by hyperglycemia or hyperlactatemia due to factors other than diabetes (i.e., carbon monoxide, asphyxia). After death the value rose till the 30th hpm, then remained stable for at least 1 week. Fatal coma was defined as the ketoacidotic form if free acetone in CSF ranged above 21 mg/l. In these cases, CSF glucose and free acetone correlated positively. Hemoglobin A1 remained stable after death. Its amount was independent from postmortem blood glucose, postmortem interval and total hemoglobin. Furthermore, the manner of storage (-80 degrees or +4 degrees C) had no significant influence on its values. In 29 of 34 cases of fatal coma, Hb A1 exceeded 12.1%. Analysis of urine glucose showed elevated levels (over 500 mg/dl) in diabetic comas. On conclusion, fatal diabetic coma seems indicated as the cause of death if measured values of postmortem biochemistry exceed the following limits: CSF-Traub 415 mg/dl, free acetone (CSF) 21 mg/l; Hb A1 12.1%; urine glucose 500 mg/dl. Most important are the Traub formula and hemoglobin A1. Usually, in fatal coma both values are elevated. If both of them are normal, diabetic coma can nearly be excluded. Combined evaluation of all values is absolutely necessary. Morphology must also always be taken into account. Consequently, a diagnosis of fatal coma can be obtained by a process of elimination.

Necrotizing fasciitis is a rapidly spreading infection of the subcutaneous tissue and fascia; diabetes mellitus appears to be the most frequent underlying disease. Early diagnosis and immediate aggressive surgical therapy are paramount to curtail morbidity and mortality, but diagnosis is often difficult and unnecessarily delayed. We describe a case of necrotizing fasciitis precipitating diabetic ketoacidotic coma where correct diagnosis was not made until the 14th hospital day. We stress the fact that physicians caring for critically ill patients should be keenly aware of the possibility of necrotizing fasciitis when tending diabetic patients with unexplained fever; failure to recognize the disease can have devastating results. Finally, we believe this to be the first reported case of diabetic ketoacidotic coma precipitated by necrotizing fasciitis.

The neuropathological findings in two cases of irreversible hypoglycemic brain injury are described. A 26-year-old diabetic man injected insulin without adequate food intake and died after 2 months in coma. An 84-year-old nondiabetic man accidentally received 10 mg of glibenclamide and died after 3 months in relatively superficial coma. In the first case, an extensive necrotizing injury with gliosis was present in the cerebral cortex with temporal preponderance, as well as in the amygdalae and hippocampus. Lesions were also present in the putamen and caudate nucleus whereas the globus pallidus and thalamus were less severely destroyed. The distribution of the lesions was therefore somewhat different from that commonly seen in hypoxic-ischemic brain injury, which, together with some previously published data, suggests some difference in the pathogenesis of hypoglycemic vs. hypoxic-ischemic brain injury. In the second case only a slight loss of cortical neurons with secondary gliosis could be attributed to the hypoglycemic insult. This case demonstrates the danger of accidental intake of sulfonylurea preparations, which can cause an irreversible brain injury due to their prolonged hypoglycemic effect.

Report of the typical clinical course and pathological findings in acute rhino-orbito-cerebral mucormycosis following diabetic coma. Invasion of orbital nerves by the fungus with neural and perineural changes may contribute to the neuroophthalmological symptoms in this disease besides the predilection for blood vessels with consequent thrombosis.

Three clinical cases of profound hypoglycemia are described with survival periods ranging from 2 1/2 months to 6 years. Although prolonged clinical coma was present in all cases, only subtle evidence of neuronal loss was found. The value of immunohistochemical examination of the grey and white matter of the telencephalon in detecting evidence of neuronal loss in such cases is demonstrated. Infarcts were uniformly absent, as would be expected from the lack of cerebral acidosis in hypoglycemia. In addition to the usual absence of infarction, the pathologic findings in hypoglycemia distinguish themselves from hypoxic/ischemic encephalopathy in that the cerebellum is regularly spared, and that there is sometimes a characteristic involvement of the dentate gyrus, a structure relatively resistant to hypoxic/ischemic damage.

Fifty-four cases of death from hyperosmolar comas were examined pathoanatomically together with the postmortem biochemical analysis of the CSF and blood. Brain of animals with disturbances of the blood and CSF osmolarity was studied electron cytochemically and autoradiographically. It was discovered that the hyperosmolar comas are manifested by excosis and brain collapse. The brain is reduced in volume due to deep cell-extracell dehydration and alteration of the hematoencephalic barrier with the irreversible depression of the neuronal and glial metabolism. In 7 patients dying with the purulent-necrotic changes of vessels resulting from the hyperosmolar effect of verografin used for the carotid angiography, numerous perivascular hemorrhages developed in the brain. Dysequilibrium syndrome in hemodialysis is manifested by an acute brain swelling or by a formation at a later period of symmetrical ischemic-hemorrhagic necrosis in the thalamus and occipital lobes of the large hemispheres.

Histopathological study of the liver has been undertaken on twenty-one diabetics with hyperlactatemia exceeding 5 mEq/1 of whom seven were treated with phenformin, six with metformin and eight not biguanide-treated. Hyperlactatemia occurred during the course or during resolution of severe ketoacidosis or of hyperosmolar coma. Hepatic lesions were invariably present. By light microscopy, massive steatosis, steatonecrosis or necrosis of variable extent were observed. Ultrastructural study showed constant mitochondrial abnormalities. These results support the hypothesis of a major role for mitochondrial changes in hepatic cells in provoking pathological hyperlactatemia. In diabetic patients, these mitochondrial lesions could be induced either by an anoxic process resulting from a variety of metabolic insults or by some as yet undefined toxic action of biguanides or by the combination of both of these factors.