Biguanides, Sulphonylureas and Insulin

Problem 1

Hypoglycaemic problem
Mrs AB died at home on the 5th of January 2000.

AB arrived in this country in 1997. Her medical history was not available, but she was first seen medically on 25/06/97 for a medical examination for an insurance policy and found to be in good health.

In early November 1999, AB first showed signs of ill health. She was dizzy, had slurred speech, fell down stairs and at one stage drove her car into a tree. She was given by Dr B (CB) a drink of red herbal tea. CB discussed AB’s condition with a colleague, telling him that AB had a shunt in her head after a head injury in South Africa. CB reported that Dr W confirmed that the shunt was functioning properly.

On 20/11/99 AB was admitted into hospital with hypoglycaemia. The hypoglycaemia appeared to require quite considerable efforts at reversal. A sulphonylurea screen for tolbutamide, glibenclamide, gliclazide, glipizide and chlorpropamide was performed and all were negative. The time given for this sample was 00:45h on 21/11/99, which is approximately 17 hours after admission.

Following this admission, further episodes of hypoglycaemia occurred on 27/11/99, 28/11/99, 29/11/99, 05/12/99 and 06/12/99, prompting a provisional diagnosis of insulinoma. C-peptide and insulin levels were raised in the presence of hypoglycaemic episodes. She underwent arterial stimulation tests of the pancreas using calcium gluconate to hepatic, superior mesenteric, gastro-duodenal and splenic arteries into the pancreas. During the test her insulin and C-peptide levels rose dramatically and this was associated with a subsequent drop in blood glucose to 1.6 mmol/L.

On the findings of this test a possible site of the insulinoma was localised to the tail of the pancreas. On 13/12/99 a surgical intervention was performed during which palpation and intraoperative ultrasound were negative. A 2.7 mm nodule was discovered in the ileum and histologically thought to be a carcinoid/insulinoma. A distal two-thirds pancreatectomy was performed. She was hypoglycaemic in hospital on two occasions, one pre-op and one post-op.

Following the surgery, the blood sugars remained relatively normal until 21/12/99, when over the next two days she experienced some hypoglycaemias with symptoms. AB was discharged on the 24th of December 1999 following a 36-hour fast without hypoglycaemia. She was discharged on pancreatic enzyme replacement and penicillin for post-splenectomy prophylaxis.

On 02/01/00 and 03/01/00, she again became hypoglycaemic corrected by diet. This persisted on 04/01/00, along with more marked symptoms. On 05/01/00 she was found dead by CB at 06:15.

1. While in hospital Mrs AB was on diazoxide from the 2nd of December, 1999 to the 13th of December. Would diazoxide override the administration of oral hypoglycaemics? On the 6th of December she had a hypoglycaemic episode, blood sugar level (BSL) 2.1 mmol/L. Would diazoxide prevent or counter a hypoglycaemic episode arising from an insulinoma?

2. At admission on the 29th of November 1999, C-peptides were high, viz 1390 picograms/mL – two conclusions presumably, has had oral hypoglycaemics or has an insulinoma. But AB was given a glucagon injection by Dr B when she went into coma at home before the ambulance took her to hospital. Would this not counteract or reverse the effect of the hypoglycaemics?

3. AB presented at admission on the 29th of November 1999 with symptoms of polydipsia, polyuria, polyphagia, and loss of weight and strength. These are symptoms of diabetes; are they symptoms of the administration of oral hypoglycaemics?

4. On the afternoon of the 1st of December 1999, during the arterial calcium stimulation AB suffered a hypoglycaemic episode, BSL dropped from 6 to 1.6 (gastro-duodenal artery). The last possible time oral hypoglycaemics could have been administered is the 28th of November 1999. Wouldn’t the half life of oral hypoglycaemics preclude their involvement (if present) in a hypoglycaemic incident on the 1st during the arteriography?

5. Are you able to comment on the effect of oral hypoglycaemics on BSL given that AB had readings as low as 0.8 mmol/L? Is there a cut off point for the effect of oral hypoglycaemics on BSLs?

Pathology
On 13 December 1999, surgical specimens were taken from Mrs.AB
The specimens were normal with no evidence of insulinoma.
Subsequent to the negative findings on the frozen specimens, the Histology Department received approximately the distal half of Mrs AB’s pancreas, the whole spleen and other tissue.
Routine pathology procedures were then used to divide the pancreas and other tissue into 59 tissue blocks consisting of 52 blocks of the pancreas and 7 of the non-pancreatic tissue.
After careful microscopic observation none of sections of these blocks showed an insulinoma.
During surgery a small nodule measuring 2.5 mm in diameter at the outer part of the lower small intestine (ileum) was found.
After microscopic observation, it was diagnosed as a carcinoid tumour. Immunohistochemistry and electron microscopy established that this carcinoid tumour was not producing insulin and therefore was not insulinoma.

Progress
Dr CB is arrested for the murder of his wife. The Police case is that Dr CB killed his wife by administering metformin, glibenclamide and glipizide which he obtained by writing false scripts. The defence case is that Dr B was stockpiling hypoglycaemics to commit suicide and the cause of death of AB is unknown.

Postmortem
Postmortem examination revealed a normal pancreas, an acute ischaemic myocardium, some early pneumonia in one lung but no evidence of a V-P shunt. There was no insulinoma found.

Toxicological analysis
Note that all specimens were post-mortem samples unless otherwise stated.

Alcohol
Blood (femoral)	Not detected
Glucose
Ante-mortem serum	1.6 millimoles per litre (mmol/L)
Glibenclamide
Blood (femoral)	0.78 milligrams per litre (mg/L)
Liver	Detected
Stomach contents	0.006 milligrams (mg)
Ante-mortem blood (EDTA tube)	0.2 mg/L
Glipizide
Blood (femoral)	0.15 mg/L
Liver	Detected
Stomach contents	0.004 mg
Ante-mortem blood (FDTA tube)	0.30 mg/L
Metformin
Blood (femoral)	16 mg/mL
Citalopram
Blood (femoral)	0.14 mg/L
Stomach contents	0.03 mg
Ante-mortem blood (EDTA tube)	0.06 mg/L
Clonazepam
Blood (femoral)	7.0 micrograms per litre
Blood (leg veins)	60 microgram/L
Ante-mortem blood (EDTA tube)	330 microgram/L
Serum (ante-mortem)	Not detected
7-Aminoclonazepam
Blood (femoral)	770 microgram/L
Blood (leg veins)	700 microgram/L
Ante-mortem blood (EDTA tube)	180 microgram/L
Serum (ante-mortem)	Not detected
Total clonazepam and 7-aminoclonazepam, as clonazepam equivalent
Blood (femoral)	860 microgram/L
Blood (leg veins)	830 microgram/L
Ante-mortem blood (EDTA tube)	530 microgram/L
Midazolam
Blood (femoral)	Not detected
Ante-mortem blood (EDTA tube)	Not detected
Fluoxetine
Blood (femoral)	Not detected

6. What is your interpretation of these post mortem analysis results?

7. What is the significance of AB only having a third of a pancreas at the time Dr CB is meant to be administering the hypoglycaemic drugs that supposedly killed her?

8. What is the effect, if any, of the delay of l7 hours post first hypoglycaemic coma before taking the blood sample for sulphonylurea screen? Would not the drug be detectable if hypoglycaemia was present? The time line is as follows: Mrs B was found comatose at 6.30 am. 10% dextrose was commenced at 13.00. BSL 2.1 at 18.30. BSL 3.1 at 2130. Blood taken 00.45 next morning.

9. What is the effect of hypoglycaemic drugs on a pancreatic venous sampling procedure? The notes state: ‘Her sampling results are somewhat unusual in that there is apparent response in the splenic artery but also possibly in the gastroduodenal plus or minus the superior mesenteric artery’. If hypoglycaemic drugs are non-selective in the pancreatic islet β-cells that they stimulate, then pancreas stimulation by calcium should not produce a different spike of insulin in the hepatic vein in respect of any of the three arteries into which the calcium was injected. Would not they result in the same amount of insulin release in response to calcium injected in each artery?

Further Progress
The Police now maintain that AB’s death was due to an injection of insulin obtained by Dr CB just prior to her death.
On the 4th of January, the day before AB’s death, Dr B picked up insulin (Humalog) at 3.00 pm. The children went in to say goodnight to their mother at about 11.30 pm. A (16) says her mother acknowledged her. G (18) says his mother was sleeping. They all then went to bed. AB was found dead at 6.00 am. Dr B did not sleep with her that night. He slept in the spare room next to his daughter and she could hear him snoring. She is quite clear she and her father went to bed at the same time. G says he read until 3:00 am. He is directly above his parent’s room where AB was and heard no movement in there.
So it would appear if Dr CB was to kill AB with an insulin injection he had to do it before 10.30 pm. Indeed any administration had to be before 10.10 pm because from 10.30 pm to 11.30 pm Dr B was walking on the beach with G and A, another friend and the dog.

10. How quickly would insulin kill, if it would kill?

The trial of Dr B for the murder of his wife is into its second week.
The Crown has just advised that further tests have shown that Mrs B’s pancreatic cells show hyperplasia. This is thought helpful to the defence, as it supports an opinion that Mrs B was suffering from endogenous pancreatogenic hyperinsulinaemic hypoglycaemia.

11. The defence needs to know urgently whether the administration of sulphonylurea drugs could or would cause the islet beta cells in the pancreas to show hyperplasia