Anaesthetic management of nesidioblastosis in a newborn.AA Soares, SA Karapurkar, SS Suresh
Anaesthesiology Department, GS Medical College & KEM Hospital, Parel, Mumbai.
This report details the management of a newborn with nesidioblastosis who underwent a 95% pancreatectomy under general anaesthesia. The baby presented with hypoglycemic convulsions, due to hyperinsulinism, and was treated with 12.5% dextrose infusions, glucagon and anticonvulsants. Intraoperatively and postoperatively the baby remained hyperglycemic. A postoperative osmotic diuresis necessitated the use of insulin for brief period. The infant remained euglycemic and convulsion free, following discontinuation of the dextrose infusions and starting of oral feeds. Recovery was uneventful.
Keywords: Anesthesia, General, methods,Case Report, Convulsions, drug therapy,etiology,Human, Hyperinsulinism, etiology,Hypoglycemia, classification,drug therapy,etiology,Infant, Newborn, Pancreatectomy, Pancreatic Diseases, complications,congenital,diagnosis,surgery,
Nesidioblastosis is a rare disorder characterised by an excess number of islet cells arising from the exocrine pancreatic ducts, which results in increased secretion of insulin. Nesidioblast means islet builder. This hyperinsulinism manifests as persistent hypoglycaemia in the first year of life, and if not treated early and effectively may lead to irreversible brain damage. Treatment of this condition entails hypertonic glucose infusions, drugs like glucagon, diazoxide and somatostatin and finally surgical resection of the pancreas.
Here, we report the diagnostic features and anaesthetic management of a new born with suspected nesidioblastosis who underwent 95% pancreatectomy for persistent hypoglycaemia under general anaesthesia.
A 3 kg, 24-day-old baby, born of a non-diabetic mother, was referred to the neonatology department for management of intractable seizures.
The baby had started getting convulsions from day two onwards and received treatment for the same at a private hospital. The baby was a 3.75 kg, full term normal delivery, with no history of birth asphyxia. Examination of the baby revealed no dysmorphic features, congenital anomalies or hepatomegaly. Investigations on admission, showed a blood sugar level (BSL) of 20 mg %. Subsequent BSLís done when the baby was not receiving glucose infusions or following a convulsion, revealed persistent hypoglycaemia.
To diagnose the cause of hypoglycaemia the following investigations were done: BSL: 19 mg % (Normal 40-80 mg %); Plasma insulin: 11.5 micro-units/ml (Normal 7-14 micro-units/ml); Insulin: Glucose ratio: 0.55 (Normal < 0.5 or no detectable level of insulin in presence of hypoglycaemia). Serum Ketones: Not detectable (should be raised in presence of hypoglycaemia): Serum Calcium .. 8.2 mg % (Normal 8.6-11 mg %). CSF - Sugar 25 mg % (Normal 20-40 mg %) - Chlorides 690 mg %. Blood lactate, amino acid and ammonia levels - normal. Urine : absent ketones and reducing substances. ABG: pH 7.37, 02 saturation 96%, base deficit - 2.4 meq/L. CT Scan Abdomen - minimally enlarged pancreas; normal abdominal scan. CT Scan Brain Changes suggestive of birth asphyxia.
These tests revealed a nonketotic hypoglycaemia accompanied by increased insulin levels and hence atentative diagnosis of nesidioblastosis was made.
This was further substantiated by the fact that
1. The infant required a high glucose infusion rate of 13 mg/kg/min of 12.5% Dextrose (normal 68 mg/kg/min) to maintain BSL above 40 mg %.
2. The baby showed a glycaemic response to Glucagon (0.1 mg/kg/IM).
The management also included calcium (2 cc/ kg IV 8 hrly) and antibiotics. Convulsions were controlled with IV Gardenal (5 mg/kg/day) and Phenytoin (3 mg/kg/day).
Despite this regime, hypoglycaemic convulsions still occurred especially when the infant was off the glucose drip during resisting of the infusion. Hence a near total pancreatectomy was decided upon. In between the convulsive episodes, the baby remained active with good muscle tone and reflexes. Preoperative examination revealed no systemic abnormalities. Investigations done the day prior to surgery were as follows: Hb 12.4 gm% RBS 180 mg % Na 128 meq/ L K+ 3.7 meq/L PCV 36% BUN 6 mg %.
The baby received 0.1mg/kg glucagon IM for treatment of a hypoglycaemic convulsion on the night prior to surgery. On the day of surgery, 7.5 mg Gardenal and 7.5 mg Phenytoin were given intravenously and the baby was shifted to the operation theatre with a 10% dextrose infusion on flow. The limbs and head of the infant were wrapped in warm gamgee and silver foil. The baby was placed on a heating mattress. The immediate preoperative BSI was 250 mg %. The infant was premedicated with atropine 0.01 mg/kg IV and hydrocortisone 5 mg/kg IV as it was a major procedure.
Anaesthesia was induced with N2O: O2: halothane and the baby was intubated with a No.3 portex plain endotracheal tube following 2 mg/kg suxamethonium IV. Anaesthesia was maintained with O2: N2O: intermittent halothane and muscle relaxant (pancuronium) using the Jackson-Rees circuit and IPPR. Intraoperatively, the baby was monitored using the cardioscope, pulse oximeter, rectal temperature probe and an oesophageal stethoscope. Blood glucose levels were monitored at 30 minutes intervals with a glucometer. IV fluids were adjusted to 10 ml/kg/hour with a glucose delivery rate of 13 mg/kg/min using 10% dextrose. Surgical exploration revealed no obvious adenomas in the pancreas, so a 95% pan createctomy was proceeded with. Following the incision, the first blood sugar level was 228 mg %. Hence, the 10% dextrose was replaced using a 5% dextrose electrolyte solution (Kidral). Despite this, the BSI remained between 250 to 300 mg% till resection of the pancreas. Post pancreatectomy, the BSI- rose to 392 mg %. Surgery lasted for 2 hours. The total blood loss was around 15 ml and no blood transfusion was given. At the end of the surgery, the baby remained apnoeic, despite one hour having elapsed since the last does of muscle relaxant. The core temperature at this stage was 35.6 deg. C. The baby was re-warmed to 37 deg. C and IPPR was continued till the baby made good spontaneous attempts to breathe. This took another one hour after the surgery. During the time, the BSL was constantly between 280-350 mg %.
The urinary bladder was not palpable and there was no urine output, despite adequate hydration. Hence 1 mg Lasix IV was given. ABG done at this time showed a pH of 7.43, O2 saturation 100% and base deficit - 0.3 meq/L. The serum K? was 4.2 meq/ L and serum Na? 134 meq/L.
After good spontaneous attempts to breathe, the muscle relaxant effect was reversed with atropine 0.02 mg/kg and neostigmine 0.05 mg/kg IV. The baby had good muscle tone but a weak cry and was shifted to the ICU and placed under an oxygen hood (3L/min) for further monitoring.
In the immediate postoperative period, the urine output of the baby was just adequate. PCV was around 25% and 30 ml blood was transfused. The blood sugar level was between 250-300 mg %. On the first postoperative day, the baby had persistent hyperglycaemia (BSL- 300-348 mg %) and osmotic diuresis, despite decreasing the dextrose delivery to 5 mg/kg/min. An insulin drip was therefore started at a rate of 0.06 U/kg/hour. Within 2 hours the BSI dropped to 40 mg %. The insulin infusion was gradually reduced to 0.013 U/kg/hour and finally stopped 2 hours later, as the BSL was maintained between 80-120 mg %. The baby had one convulsion during this period.
24 hours postoperatively the BSL was 189 mg %, with a normal dextrose infusion rate. Over the next 4 days, the BSL was maintained between 85-105 mg % without insulin. Oral feeds were started on the 5th postoperative day along with supplementary IV fluids. By the 8th postoperative day, the baby was receiving only breast milk. There were no further hypoglycaemic episodes. The convulsions, too, had stopped, but the anticonvulsants were continued. The baby was active and thriving well and the BSL were between 80-130 mg %.
The histopathology report showed normal pancreatic acini with increase in the number of Islets of Langerhans More Details. Between the acini were seen clusters of islet cells, suggestive of nesidioblastosis, associated with adenomatosis.
Hypoglycemia in babies is defined as a BSL less than 30 mg % in the first 3 days and 40 mg % thereafter, in the term infant. The incidence of hypoglycaemia is about 2 per 1000 live births in full term and 43 per 1000 births in preterm infants.
Cornblath and Schwartza have described 4 types of neonatal hypoglycaemia.
Type I - early transient neonatal hypoglycaemia - in infants of diabetic mothers and in fetal distress.
Type II - secondary neonatal hypoglycaemia in brain disorders and cardiac defects.
Type III - classical transient neonatal hypoglycaemia hypocalcemia, CNS anomalies and cardiac failure.
Type IV - persistent neonatal hypoglycaemia characterised by a duration of more than 7 days and of a severity requiring glucose at a rate of 12-14 mg/kg/ min.
1. Inborn errors of metabolism - Glucose, amino acid, fatty acid oxidation & ketogenesis.
2. Endocrine disorders - Hypopituitarism, adrenal insufficiency, hypothyroidism, glucagon deficiency.
3. Neonatal hyperinsulinism - comprise 1.4% to 5.6% of all hypoglycaemias. It is characterized by
- Nonketotic hypoglycaemia
- Birth weight more than gestational age (Birth)
- FFA level less than 0.40 mmol/L
- ? hydroxybutyrate less than 1.1 mmol/L.
- Insulin level more than 12 micro-units/ml
- Insulin glucose ratio more than 0.5 or a detectable insulin level when BSL less than 40 mg %.
- Glycaemic response to glucagon.
Aetiology of neonatal hyperinsulinism includes Beckwith Widemann syndrome ? cell adenoma localized by CT scan. Nesidioblastosis - diagnosed by histopathology.
Ansley-Green and other reported a study of 3 infants who had severe hypoglycaemia due to pancreatic nesidioblastosis. They underwent 75% pancreatectomy for the same. These babies were large for gestational age and presented with convulsions within 6 hours of birth. The mean BSL was 16.4 mg % (control 66 mg %) with a mean plasma insulin concentration of 28.1 micro-units/ml (control 6.8 micro-units/ml). All had low levels of ? -hydroxybutyrate.
In our case, a 3.75 kg baby large for gestational age, presented with a history of convulsion from the second day onwards. At the time of a convulsion, the infant BSL was 19 mg % and the simultaneous plasma insulin level was 11.5 micro-units/ml. The i/G ratio was 0.55 and serum ketones were absent. Other tests to rule out inborn errors of metabolism were negative.
The baby was treated with 12-5% dextrose infusions at a rate of 13 mg/kg/min. and received I.M. Glucagon (0.1 mg/kg) twice to control hypoglycaemic convulsions, which occurred despite the glucose infusions.
One study used concentrated glucose hydrocortisone 5 mg/kg/12 hrs, glucagon 0.1 mg/kg and diazoxide 10-25 mg/kg/day to treat the hypoglycaemia. They reported an increase in BSL to 128 mg % with diazoxide. Beneficial effects of this drug include suppression of insulin secretion and enhancement of glycogenolysis and catecholamine release. Sizonekis observed remission on 60% cases where diazoxide therapy was continued for 4-6 years.
The same study documented an increase in BSL from 19.8 to 109 mg % following glucagon. Only a minimal increase in BSL (from 32.4 to 41.4%) was observed with hydrocortisone. The baby did not receive diazoxide due to its non-availability. Failure to control the hypoglycaemia medically led to the decision to surgically resect 95% of the pancreas. Fonkalrud recommended early pancreatic resection to obviate the occurrence of mental retardation resulting from frequent hypoglycaemic convulsions.
Preoperatively our patient had a raised BSL 250 mg % as he had received a bolus of 10% dextrose just prior to transfer to the operation theatre. 30 minutes post induction and throughout the surgical procedure, the blood sugar remained high (250-300 mg %) despite switching the infusion from 10% dextrose to a maintenance solution containing 5% dextrose (Kidral).
This hyperglycemia could possibly be attributed to:
a) Catecholamine release due to stress of surgery.
b) Insulin inhibiting effect of diphenyihydantoin,.
c) Residual glycemic effect of I.M. glucagon administered the previous night.
d) I.V. hydrocortisone given at the time of induction.
Green used a continuous intraoperative infusion of 10% dextrose and recorded a mean BSL of 180 mg % prior to pancreatic resection in 3 infants with nesidioblastosis. Immediately following near total pancreatectomy we documented an increase in BSL to 392 mg %, which then decreased to between 300-350 mg % over the next few hours.
This was in concurrence with one study, which reported a postoperative BSL of 280 mg %. Fonkalrud too, found an increase in BSL to 250 mg % despite changing the I.V. infusion from 20% to 5% dextrose. 40 minutes before resection was complete. He recommended termination of the dextrose infusion 30 minutes prior to completing the pancreatic resection, in order to reduce the severity of the post surgical hyperglycaemic response. However, this may give rise to dangerous hypoglycaemia if insulin is released during handling of the pancreas. . In fact, Green 3 demonstrated an increase in insulin level to 87 ?U/ ?l immediately before resection, which decreased to 18 U/ml following 75% pancreatectomy. We could not monitor intraoperative insulin levels in our patient due to lack of facilities.
The baby in our study had delayed recovery from muscle relaxant effect, which could have been due to the reduced urine output. The preoperative, parenteral Gardenal could have had an additive effect on the anaesthetic agents. The baby's low core temperature (35.6 deg. C) at the end of surgery could also have contributed to the delayed recovery. A metabolic cause seems unlikely as ABG and serum K+ at this stage were normal. Lack of urine output was inexplicable considering the adequate hydration and the constant intraoperative hyperglycaemia, which should have caused an osmotic diuresis. Urine output improved following 1 mg Lasix 1.V. The baby developed osmotic diuresis 12 hours postoperatively when BSLís continued to remain between 300-350 mg %. At this stage, insulin was given for a period of 4 hours, following which BSLís gradually decreased to 80-105 mg %. The osmotic diuresis also reduced and the baby did not require any insulin thereafter.
In one study insulin was required to control hyperglycaemia and osmotic diuresis, for 3-48 hours postoperatively. Following this, there was persistent hypoglycaemia due to inadequate pancreatic resection, which necessitated a total pancreatectomy 4 weeks later.
Another study demonstrated a spontaneous reduction in hyperglycaemia after the first 3 postoperative days, to a BSL of 90-140 mg %. Following full oral feeds, the baby in this report, remained euglycemic and had no further recurrence of convulsions.
A long term follow up would be necessary to ensure that the baby remains euglycemic and to assess its neurological development in view of the changes on the Brain Scan. Two follow up studies, over a period of 4 years showed that most of the infants had normal neurological development except for one who had mild mental retardation.
In conclusion, it may be stated, that frequent, peri-operative monitoring of blood sugar levels in these babies is absolutely essential if irreversible brain damage, due to hypoglycaemia is to be avoided.