Hypothermia and sudden infant death syndrme?????

Disease in Childhood, 1988, 63

Discussion

Fungal infection in neutropenic patients is common.

In one postmortem study of young cancer patients,

23 out of 88 fatalities showed evidence of systemic

fungal infection.1

Our patient was at particular risk of such an

infection. The risk factors included prolonged neutropenia,

prolonged steroid treatment, diabetes, and

an indwelling Hickman catheter.2 In addition he was

on prolonged broad spectrum antibiotics, the most

important factor in the aetiology of serious and fatal

candidiasis. The stopping of antibiotics in those who

remain febrile and neutropenic, however, carries a

very high risk of serious bacterial infections.3

Empirical antifungal treatment is advocated for

neutropenic patients with an antibiotic resistant

fever.1 4 5 Amphotericin is the only antifungal with

proved efficacy in neutropenic patients and its use as

a single agent in this situation is therefore

recommended.4 5 Indeed in one study nearly complete

elimination of fungal infections was seen if

amphotericin was started in those who remained

febrile and neutropenic after a week of antibiotic

treatment.1 Starting antifungals only late in antibiotic

resistant neutropenic fever is known to result

in an increased mortality from fungal infection, but

the optimal time to start remains uncertain.1 S

In our patient amphotericin was started after six

days of fever resistant to antibiotics. Despite this,

symptomatic presentation of mycotic cerebral

abscesses occurred 12 days later and viable Candida

was isolated from these three weeks after the start of

amphotericin treatment. Doubt about the efficacy of

amphotericin alone as empirical treatment has been

expressed previously.6

The use of flucytosine combined with amphotericin

has been suggested, not least because of its

synergistic activity against Candida.6 This combination,

however, has been considered to have a

number of disadvantages-namely, the need to

lower the dose of amphotericin, the problems of

determining flucytosine concentrations, and myelotoxicity.

5 While using this combination therapeutically

in our patient none of these problems were seen.

Amphotericin dosage was reasonable (1 mg/kg on

alternate days), flucytosine concentrations were

readily determined by our laboratories, and myelotoxicity

was not an important problem. The addition

of itraconazole seemed justified by the extremely

high mortality previously documented for this

condition.

References

Pizzo PA, Robichaud KJ, Gill FA, Witebsky FG. Empiric

antibiotic and antifungal therapy for cancer patients with

prolonged fever and granulocytopenia. Am J Med 1982;72:

101-11.

2 Warnock DW, Richardson MD. Fungal infections in the

compromised patient. Chichester: Wiley, 1982.

Pizzo PA, Robichaud KJ, Gill FA, et al. Duration of empiric

antibody therapy in granulocytopenic patients with cancer. Am J

Med 1979;67:194-200.

4 Marcus RE, Goldman JM. Management of infection in the

neutropenic patient. Br Med J 1986;293:406-8.

Cohen J. Empirical antifungal therapy in neutropenic patients.

Journal of Antimicrobal Chemotherapy 1984;13:409-11.

6 Polak A. Combined therapy of systemic mycoses. Proceedings

of 13th International Congress of Chemotherapy. Vienna:

1983. Clinical and experimental studies in immunotherapy.

Amsterdam and Princeton: Excerpta Medica, 1984: Part 20:2-9.

Correspondence and requests for reprints to Dr MG Mott, Bristol

Royal Hospital for Sick Children, St Michael's Hill, Bristol BS2

8BJ.

Accepted 29 December 1987

Hypothermia and sudden infant death syndrome

K P DUNNE AND T G MATTHEWS

Department of Paediatrics, Rotunda Hospital, Dublin, Ireland

SUMMARY We recently reported an association between

recurrent episodes of severe apnoea requiring

vigorous resuscitation for which no cause could be

found and episodic hypothermia. Two similar cases

are now reported that give further evidence of a link

between hypothermia and acute life threatening

episodes of apnoea.

Our first report concerned a baby boy who at the

time of writing this paper was 4 years old.' He has

had no further apnoeic attacks, but does have

recurrent attacks of hypothermia (rectal temperature

32-35°C). The only new development is that he

now has occasional episodes of sweating and hypoglycaemia

that have been documented while he was

under observation in hospital. He is being monitored

at home with a skin temperature probe.

Downloaded from http://adc.bmj.com/ on January 18, 2015 - Published by group.bmj.com

Hypothermia and sudden infant death syndrome 439

We now report two similar cases that give further

evidence of an association between severe apnoea

for which no cause could be found and hypothermia.

Case reports

CASE 1 A normal baby boy was born at term

weighing 2270 g. There was no family history of

sudden infant death syndrome (SIDS) or hypothermia.

He was referred from another hospital at the

age of 2 months having had recurrent episodes of

apparent lifelessness during sleep when he became

pale, grey, cold, and limp, and had required

vigorous mechanical or mouth to mouth resuscitation

by his mother (a trained nurse) from the age of

4 weeks. Observations showed excessive periodic

breathing when compared with age matched controls,

and mild regurgitation was seen on barium

swallow examination. He was treated with oral

theophylline 5 mg/kgI24 hours and given a home

apnoea alarm (MR-10, Graseby Dynamics Ltd).

Between 1 and 12 months there was a gradual

improvement both in the frequency and the severity

of his attacks. The apnoeic episodes stopped when

he was 15 months old. From the age of 10 months he

had recurrent episodes of hypothermia with a rectal

temperature of less than 35°C recorded while he was

under observation in hospital. These episodes were

accompanied by extreme pallor and he was cold to

touch (described as 'snow white' by his mother) but

had no sweating. Rectal temperatures during the

day were normal, and he became feverish (38°C)

when he had an upper respiratory tract infection.

Physical examination yielded normal results, and

development at the time of writing was normal.

CASE 2-A baby girl of 38 weeks' gestation was

delivered by caesarian section (because of a bicornuate

uterus) weighing 2892 g. She had a brief

cyanotic attack at the age of 2 days, possibly

associated with secretion of mucus. She was seen

when she was 6 weeks old with a three week history

of multiple recurrent episodes of apnoea, limpness,

and cyanosis that had required vigorous resuscitation.

Observations and investigations showed excessive

periodic breathing, and a bradycardia of less

than 90 beats/minute was recorded on cardiorespiratory

tracing. She was given a home apnoea alarm

but theophylline 5 mg/kgl24 hours was stopped

because it made her irritable. She had recurrent

apnoeic attacks requiring vigorous resuscitation

every four or five weeks until she was 12 months old,

and recurrent episodes of hypothermia during which

she became pale and cold with a rectal temperature

of less than 35°C occurred during sleep throughout

the summer while she was under observation in

hospital. These started when she was 8 months old.

Physical examination yielded normal results, and

development at the time of writing was normal.

All three babies had the following measurements

made, all of which yielded normal results: full blood

count; erythrocyte sedimentation rate (mm in the

first hour); serum urea and electrolyte concentrations;

and calcium, glucose, and creatine phosphokinase

concentrations. Serum glutamic oxylacetic

transaminase and serum glutamic pyruvic transaminase

activities, were normal as were screening tests

for autoantibodies, serum concentrations of immunoglobulins,

chest radiographs, cultures of cerebrospinal

fluid, and Mantoux tests. Wasserman

reactions were negative, as were serological tests for

toxoplasma, rubella, cytomegalovirus, herpesvirus,

mumps, Q fever, and psittacosis. The electrocardiograms

and electroencephalograms were normal.

Computed tomography of the brain yielded normal

results in each case with normal copora callosa. The

patient previously reported had also had magnetic

resonance imaging of the brain, which was normal.

Estimations of the serum concentrations of thyroxine,

thyroid stimulating hormone, and prolactin

(samples taken at rest) were normal. Pronounced

hypoglycaemia induced by insulin produced normal

serum growth hormone and cortisol concentrations.

There were normal responses to thyroid releasing

hormone, and the gonadotrophin responses were

also normal. Cortisol concentrations showed normal

diurnal variations. Early morning urine specimens

concentrated normally.

Discussion

All three infants had recurrent unexplained life

threatening episodes of apnoea followed later by

episodes of hypothermia. This association may be

more important that has previously been recognised.

We have reviewed the presentation and management

of 73 cases of acute life threatening episodes of

apnoea diagnosed at this hospital from 1980-84.2

We studied all patients with episodes of lifelessness

that required vigorous mechanical or mouth to

mouth resuscitation to revive the infant, and in

whom investigations failed to show a cause. Twelve

parents (16%) stated as a presenting complaint that

the child was 'cold'. This could simply be explained

by the vasoconstriction that occurs in any ill child.

Six children (8%), however, had recurrent severe

apnoeic attacks, and three of these later developed

episodic hypothermia.

Although the pathophysiology of both apnoea

and hypothermia is uncertain, the most likely link

between severe apnoeic attacks and temperature

Downloaded from http://adc.bmj.com/ on January 18, 2015 - Published by group.bmj.com

440 Archives of Disease in Childhood, 1988, 63

disturbances is a defect in hypothalamic function.I

Episodic hypothermia is rare, and usually associated

with episodic hyperhidrosis, and absence of the

corpus callosum.3 This link with agenesis of the

corpus callosum is presumed not to be functional but

developmental, perhaps occurring in the hypothalamus.

One patient had gliosis and neuronal loss in the

premamillary area of the hypothalamus at necropsy.3

As the hypothalamus also controls regulation of

sleep, it is interesting to note the differences in sleep

patterns between normal controls and infants with

severe apnoeic attacks.4

Neurones sensitive to local temperature extend

from the hypothalamus far down into the spinal cord

in an interlinked system. It has been postulated that

the defect causing episodic hypothermia may be

below the hypothalamus. We have reported normal

brain stem potentials evoked by auditory stimuli in

one patient. 1 This, however, does not exclude a

brain stem lesion with possible effects on respiration.

Altered autonomic function may affect both

temperature and respiratory control, and play a part

in the pathogenesis of SIDS. We have previously

reported a child with poor short term variability of

heart rate who died of SIDS. There are, however,

few data on normal controls to compare with

variability in heart rate among these patients, and

we could not test their autonomic function fully

because they were too young.

The association between severe apnoeic attacks

and SIDS has not been proved, though some of

these infants have died from SIDS.2 Abnormally

high temperature has been postulated as a cause of

some deaths from SIDS, despite the fact that most

of the deaths occur in winter. Stress caused by cold

may be a sufficient trigger for the abnormalities in

fatty acid metabolism recently implicated in the

pathogenesis of SIDS. A defect in heat production

that is metabolically mediated and accentuated by a

cold environment may be a possible explanation for

the increase in such deaths during the winter

months. The metabolic state of these children is

currently being investigated.

The presence of brown adipose tissue may also be

important. There is evidence that brown adipose

tissue is present and functional in normal babies.5

Aherne and Hull, quoted by Blaza, however, found

that the brown adipose tissue was depleted of fat in

42 of 394 necropsies of infants who died before the

age of 4 weeks. Most of the 42 deaths were due to

starvation and the brown fat was only partially

depleted. In six infants with injury from cold,

however, they found total depletion of brown fat. A

similar depletion was found in two elderly adult

patients who had died of hypothermia. There were

no cases of SIDS in their series.

Contrary to earlier reports, Emery and Dinsdale

found less periadrenal brown fat in babies who died

of SIDS compared with controls aged less than 5

months.6 As the largest mass of brown adipose

tissue in the abdomen envelops the kidneys we

postulate that these infants may have been at greater

risk of hypothermia.

Abnormal temperature regulation should be

sought in infants with recurrent life threatening

apnoeic attacks, and should be investigated closely

in the search for the causes of SIDS.

References

Dunne KP, McKay M, Matthcws TG. Blowing hot and cold:

'near-miss" sudden death and episodic hypothermia. Br Med J

1986;293:856.

2 Dunne KP, Matthews TG. Clinical presentation and management

of "near-miss" SIDS. Pediatrics 1987;79:889-93.

3 Le Witt PA, Newman RP, Greengerg HS, et al. Episodic

hyperhidrosis, hypothermia, and agenesis of the corpus callosum.

Neurology 1983;33:1122-9.

4 Guilleminault C, Coons S. Sleep states and maturation of sleep:

a comparative study between full-term normal controls and

near-miss SIDS infants. In: Tildon JT, Roeder LM, Steinschneider

A, eds. Proceedings of the international conference on

SIDS, Baltimore, 1982. London: Academic Press, 1983;401-12.

5 Blaza S. Brown adipose tissue in man. A review. J Roy Soc Med

1983;76:213-6.

6 Emery J, Dinsdale F. Structure of periadrenal brown fat in

childhood in both expected and cot deaths. Arch Dis Child

1978;53:154-8.

Correspondence to Dr TG Matthews, Department of Paediatrics,

Rotunda Hospital, Dublin 1, Ireland.

Accepted 19 October