Journal of Internal Medicine 1996 : 239 : 451–456
CASE REPORT
Nitrogen dioxide pneumonitis in ice hockey players
C. K A R L S O N-S T I B E R, J. H O> J E R ,* AI . S J O> H O L M , ‡ G. B L U H M† & H. S A L M O N S O N
From the Swedish Poison Information Centre, the *Department of Internal Medicine, SoX der Hospital, and the †Department of Environmental
Health and Infectious Disease Control, Karolinska Hospital, Stockholm, ; and the ‡Department of Internal Medicine, LoX wenstroX mska Hospital,
Upplands VaX sby ; Sweden
Abstract. Karlson-Stiber C, Ho$ jer J, Sjo$ holm AI , Bluhm
G, Salmonson H (Swedish Poison Information Centre,
the Department of Internal Medicine, So$ der Hospital,
and the Department of Environmental Health and
Infectious Disease Control, Karolinska Hospital,
Stockholm ; and the Department of Internal Medicine,
Lo$ wenstro$ mska Hospital, Upplands Va$ sby ; Sweden).
Nitrogen dioxide poisoning with toxic pneumonitis
in ice hockey players (Case report). J Intern Med
1996 ; 239 : 451–6.
Exposure to the toxic gases carbon monoxide and
nitrogen dioxide (NO ) in indoor ice arenas occa#
sionally occurs and may result in severe symptoms.
The gases are produced by ice resurfacing machines
operating on hydrocarbons, and in certain conditions
toxic levels accumulate. The damage to lung tissues
caused by NO may not be evident until after a
#
latency time of "–2 days. The role of corticosteroids in
#
Introduction
The hazards of exposure to poisonous substances
whist playing ice hockey in indoor arenas are not
widely recognized, although they have been pointed
out previously [1–3]. The symptoms described have
been dizziness, headache, nausea, vomiting, cough,
haemoptysis, dyspnoea and even unconsciousness.
The toxic agents responsible have been identified as
carbon monoxide and}or nitrogen dioxide (NO )
#
# 1996 Blackwell Science Ltd
the treatment is controversial, but there are clinical
experiences as well as experimental data supporting
their use.
We report two cases of toxic pneumonitis, with
delayed onset, due to NO exposure during an ice
#
hockey game in an indoor arena. Signs and symptoms were cough, dyspnoea, haemoptysis, hypoxaemia and reduced peak expiratory flow. Chest
radiographs showed parenchymatous infiltrative
lesions and alveolar consolidation. Both patients
were treated with high doses of corticosteroids by
inhalation and orally or intravenously. Their condition rapidly improved and pulmonary function was
restored.
Keywords : corticosteroids, ice skating rinks,
irritating gas, nitrogen dioxides, poisoning, toxic
pneumonitis.
produced by ice resurfacing machines operating on
gasoline or liquefied petroleum gas. The aggravating
factors that have been reported to cause accumulation of these gases to toxic levels are inadequate
ventilation, high shields surrounding the rink and
malfunctioning resurfacing equipment. We report
two cases with delayed toxic pneumonitis and
pulmonary oedema that recently occurred in
Sweden.
During a 2-day period in December 1994, the
451
452
C. K A R L S O N-S T I B E R et al.
Swedish Poison Information Centre in Stockholm
received 10 inquiries from young men experiencing
irritation of the respiratory tract after having played
ice hockey in the same indoor arena in a suburb of
Stockholm. Five of these men were admitted to
hospital, two of them to intensive care units. Data
concerning pulmonary function were documented
throughout the hospital stay and at follow-up 4–6
weeks after the acute illness. The ice resurfacer
which had been used was powered by an internal
combustion engine using liquefied petroleum gas.
Monitoring of carbon monoxide and NO levels was
#
conducted by the Environment and Health Protection
Administration in Stockholm.
Case reports
Case 1
A 17-year-old male ice hockey player was admitted
to Lo$ wenstro$ mska Hospital at 11.00 hours on 8
December with difficulty in breathing. This nonsmoking patient had previously been in good health
and had no history of asthma or allergy. He had been
playing ice hockey with his school team in a
community-operated indoor ice arena between 14.00
and 15.00 hours on the previous day, when all the
players had experienced acute onset of slight but
irritating cough and diffuse chest pain. After the
game the cough vanished and until midnight, 9 h
later, the patient had only slight discomfort in his
chest. After midnight, however, his condition deteriorated, with intensified dyspnoea, dizziness, vertigo
and headache, and he consulted the hospital.
Upon physical examination the patient was afebrile
but exhibited shortness of breath and tachypnoea.
The heart rate was 95 b.p.m. and the blood pressure
130}60 mmHg. On auscultation of the lungs, diffuse
wet crackles were audible bilaterally and the peak
expiratory flow (PEF) rate was lowered to 350 L
min−" (reference 640 L min−"). Measurement of the
arterial blood gases disclosed severe hypoxaemia,
with PO 6.0 kPa, PCO 5.7 kPa and oxygen satura#
#
Fig. 1 (a) Chest radiographs of a 17-year-old ice hockey player (case 1) taken 22 h after exposure to toxic levels of nitrogen dioxide,
showing marked parenchymatous infiltrative lesions with a reticulo-nodular pattern and widened bronchi (most pronounced on the
right side), and (b) showing disappearance of these lesions 4 days later.
# 1996 Blackwell Science Ltd Journal of Internal Medicine 239 : 451–456
C A S E R E P O R T: N I T R O G E N D I O X I D E P N E U M O N I T I S
453
Fig. 2 (a) Chest radiographs of another player (case 2) taken 34 h after exposure to nitrogen dioxide, showing widespread alveolar
consolidation bilaterally, and (b) showing normalization 7 days later.
tion of haemoglobin 81 %. The carboxyhaemoglobin
value was in the normal range and methaemoglobin
was 0.8 % (reference ! 0.6 %). The white blood cell
count was elevated to 17.1¬10* L−" (reference
4–10¬10* L−") and the C-reactive protein (CRP)
was increased to 57 mg L−" (reference ! 10 mg L−").
All other routine laboratory blood tests were normal.
A chest radiograph disclosed marked parenchymatous infiltrative lesions with a reticulo-nodular
pattern (Fig. 1 a).
The patient was given 8 mg of betamethasone
intravenously and was transferred to the intensive
care unit on suspicion of toxic pneumonitis due to
exposure to some unknown irritating gas. There he
was treated with oxygen and 60 mg prednisolone
orally daily in conjunction with inhalation of 400 µg
budesonide four times per day. No antibiotics were
given. The patient remained at the intensive care
unit for 4 days, during which time his condition
gradually improved and the blood gases, PEF rate
and chest radiograph were normalized (Fig. 1b). After
a further 2 days of observation on an ordinary ward,
he was discharged in good condition and prescribed
6-day oral therapy with tapering doses of prednisolone. A spirometric assessment was performed 4
weeks later and showed completely normalized
pulmonary function.
Case 2
Another 17-year-old player in the school ice hockey
team was admitted to So$ der Hospital at 23.10 hours
on 8 December with cough and haemoptysis. He was
a nonsmoker and previously healthy. Like the other
players, he had had a slight cough during the game
in the previous afternoon, but during the rest of the
day he had felt quite well. In fact, he played hockey
again on an outdoor arena in the evening with his
club team. However, at 02.00 hours on 8 December
(11 h after the afternoon game) he deteriorated, with
irritating cough which worsened during the night,
and repeated haemoptysis. The following day he felt
better, but had headache and some fever, and in the
evening he experienced shortness of breath and
consulted the hospital.
Upon physical examination he exhibited dyspnoea.
The heart rate was 70 b.p.m. and the blood pressure
130}70 mmHg. Distinct rales were audible bilaterally on auscultation of the lungs, and the PEF rate
was only 270 L min−" (reference 640 L min−"). The
pulse-oximeter showed an oxygen saturation of 85 %.
All routine laboratory blood tests were normal except
for the white blood cell count, which was elevated to
13.8¬10* L−", and CRP, which was increased to
59 mg L−". The carboxyhaemoglobin level was
normal and methaemoglobin was 0.8 %. A chest
# 1996 Blackwell Science Ltd Journal of Internal Medicine 239 : 451–456
454
C. K A R L S O N-S T I B E R et al.
radiograph disclosed widespread alveolar consolidation bilaterally (Fig. 2 a).
The patient was given oxygen, 16 mg betamethasone intravenously and 2 mg budesonide through
an inhaler, and was transferred to the medical
intensive care unit. Toxic pneumonitis was suspected,
although bacterial pneumonia was initially considered as a differential diagnosis. Oral penicillin was
started, but was then discontinued together with the
systemic corticosteroid therapy after 2 days, when
the patient was discharged from hospital. During the
2 days at the intensive care unit his condition rapidly
improved and at follow-up on 15 December the
laboratory values, blood gases, PEF rate and chest
radiograph were normalized (Fig. 2b). The local
corticosteroid therapy with budesonide was continued, first in a dose of 2 mg twice daily up to 15
December and then 800 µg twice daily for 4 more
weeks. A full spirometric assessment 6 weeks later,
including measurement of the diffusion capacity,
showed completely normalized pulmonary function.
On clinical follow-up 1 year after the incident,
both patients were still active ice hockey players and
showed good health without any symptoms of the
respiratory tract.
In addition to these two cases, three other players
were admitted to hospital. However, as their radiographs were normal and they had no residual
symptoms, they were discharged after a few hours of
observation.
Discussion
Inhalation of nitrogen dioxide is a known cause of
toxic pneumonitis, pulmonary oedema and, although
rare, bronchiolitis obliterans [4]. The gas is insidious,
because it is relatively insoluble in water and not
particularly irritating to the mucous membranes.
Thus, it may exert harmful effects on the peripheral
parts of the lung, despite the presence of only mild
and transient initial symptoms, and more dramatic
symptoms may be delayed for a period of some hours
up to 2 days. The toxic mechanisms have not been
fully elucidated, but NO is a powerful oxidant and it
#
is assumed that the tissue damage is a result of
production of free radicals followed by lipid peroxidation and membrane destruction [5]. When
combined with moisture and water, it also slowly
forms nitric and nitrous acid, which may aggravate
the injury. It has been suggested that methaemoglobinaemia may occur as a result of absorption of
the gas by the blood [6].
The degree of toxic damage basically depends on
the concentration of the gas inhaled, as well as on
the length of the exposure period and volume of air
inhaled. Severe symptoms such as pulmonary
oedema are likely to occur after inhalation of
100 000 µg m−$ for 1 h [7]. However, during exercise, such as ice hockey, the respiratory minute
volume will considerably increase, resulting in
heavier exposure to NO and thus a risk of severe
#
reactions to much lower concentrations of the gas
[8]. Controlled exposure has shown that bronchoconstriction can occur in normal subjects at NO
#
levels of about 9000 µg m−$, and probably also at
4000–5000 µg m−$ [9]. In asthmatic individuals,
similar findings have been reported at concentrations
of about 500 µg m−$ [9]. Bronchial hyperactivity can
occur at concentrations above 2000 µg m−$ in
normal subjects and 200 µg m−$ in asthmatics, and
it has been suggested that repeated exposures may
cause persistent increased airway responsiveness [9].
Measurements conducted after operation of the ice
resurfacer in a similar way to that during the day of
the incident reported here, showed a maximal 1-h
NO level of 2358 µg m−$. The carbon monoxide
#
level was below the standard for outdoor air. Twenty
hours after the use of the resurfacing machine the
concentration of NO was still more than 2000 µg
#
m−$, which is the 8-h occupational exposure limit for
NO from motor combustion exhausts. Moreover, the
#
concentration of NO could have been higher during
#
the game. Air monitoring in indoor ice arenas in
Sweden has shown higher levels of NO (4000 µg
#
m−$) with a properly run ice resurfacing machine but
with poor ventilation [10]. Despite improvements of
the ice arena after the accident described here, with
replacement of the emission control device, which
was partly malfunctioning, elevation of the exhaust
pipe and improvement of the ventilation, the level of
NO after only half an hour’s use of the ice resurfacer
#
exceeded the national 1-h air quality standard of
110 µg m−$. Hence, we recommend that in indoor
ice arenas the ice resurfacing machines be replaced
by others driven by electricity.
The symptomatology of the patients reported here
clearly illustrates the characteristics of NO and is
#
consistent with that described by others [1, 2, 11].
The symptoms during the exposure were minor, with
# 1996 Blackwell Science Ltd Journal of Internal Medicine 239 : 451–456
C A S E R E P O R T: N I T R O G E N D I O X I D E P N E U M O N I T I S
only slight cough and diffuse chest pain. Moreover,
in both cases there was a latency time of several
hours (9 and 11 h) before the relapse to severe
symptoms including haemoptysis and hypoxaemia.
This emphasizes that under special conditions (e.g.
high physical activity, prolonged exposure and inadequate ventilation) even relatively low levels of
toxic gases can be deleterious. The patients were
both treated with high doses of corticosteroids locally
and systemically and their condition rapidly
improved. According to clinical assessment 1 year
after the incident the recovery was complete.
Concerning the medical therapy in cases of toxic
pneumonitis and pulmonary oedema, the general
conception is that administration of oxygen, Β #
stimulating agents if there are signs of airway
obstruction and, when required, mechanical ventilation with positive end-expiratory pressure is of
great importance. On the other hand, the use of
corticosteroids is controversial and there are no
conclusive studies on this issue in the literature.
Some experimental data have shown that the use of
steroids ameliorates the damage to lung tissues after
exposure to NO , but other experimental results have
#
failed to demonstrate any positive effect [12–14].
However, in studies of lung mechanics, gas exchange
and haemodynamics, early corticosteroid inhalation
reduced the impairment of respiratory function after
chlorine exposure [15, 16]. Furthermore, clinical
experience [17, 18] has indicated that administration
of corticosteroids through inhalation and}or intravenously could be of value in the acute management
of patients exposed to irritating gases. It has also
been suggested that later or prolonged treatment
with corticosteroids may prevent development of
permanent pulmonary dysfunction due to bronchiolitis obliterans, which in some cases follows
exposure to NO [19]. According to the recom#
mendations currently applied at the Swedish Poison
Information Centre, patients who have been exposed
to an irritating gas such as NO should be treated
#
locally with high doses of budesonide or beclomethasone. Initially a dose of 4 mg (10 inhalations of
0.4 mg) is given and this is followed by 2 mg one to
five times during the first 24 h. To avoid bronchoconstriction, pretreatment with locally administered
β -stimulating agents is advisable. Systemic treat#
ment with intravenous injections of 8–16 mg betamethasone is also recommended in cases of moderate
to heavy exposure.
455
Finally, when patients present with pulmonary
symptoms with onset while, or within 48 h after,
spending time in an indoor ice arena, physicians
should consider the possibility of exposure to NO
#
and be aware of the risk of delayed severe toxic
symptoms.
References
1 Hedberg K, Hedberg CW, Iber C, White KE, Osterholm MT,
Jones DBW et al. An outbreak of nitrogen dioxide-induced
respiratory illness among ice hockey players. JAMA 1989 ;
262 : 3014–17.
2 Smith W, Anderson T, Anderson HA, Remington PL. Nitrogen
dioxide and carbon monoxide intoxication in an indoor ice
arena – Wisconsin, 1992. MMWR 1992 ; 41 : 383–5.
3 Soparkar G, Mayers I, Edouard L, Hoeppner VH. Toxic effects
from nitrogen dioxide in ice-skating arenas. Can Med Assoc J
1993 ; 148 : 1181–2.
4 Anonymous. Ice hockey lung : NO poisoning (Editorial) .
#
Lancet 1990 ; 335 : 1191.
5 Meulenbelt J, Dormans JAMA, van Bree L, Rombout PJA,
Sangster B. Desferrioxamine treatment reduces histological
evidence of lung damage in rats after acute nitrogen dioxide
(NO ) intoxication. Hum Exp Toxicol 1993 ; 12 : 389–95.
#
6 Fleetham JA, Tunnicliffe BW, Munt PW. Methemoglobinemia
and the oxides of nitrogen. N Engl J Med 1978 ; 298 : 1150.
7 Hathaway GJ et al., eds. Nitrogen dioxide. In : Proctor and
Hughes’ Chemical Hazards & the Workplace, 3rd edn. New York :
Van Nostrand Reinhold, 1991 ; 432–4.
8 Anderson DE. Problems created for ice arenas by engine
exhaust. Am Ind Hyg Assoc J 1971 ; 32 : 790–801.
9 Bylin G. Health risk evaluation of nitrogen oxides. Controlled
studies on humans. Scand J Work Environ Health 1993 ;
2 (Suppl. 19) : 37–43.
10 Berglund M, Bra/ ba$ ck L, Bylin G, Jonson JO, Vather M.
Personal NO exposure among ice-skating schoolchildren.
#
Arch Environ Health 1994 ; 49 : 17–24.
11 Lee K, Yanagisawa Y, Spengler JD, Nakai S. Carbon monoxide
and nitrogen dioxide exposures in indoor ice skating rinks. J
Sport Sci 1994 ; 12 : 279–83.
12 Engelhardt G. Effect of corticosteroids on the toxic pulmonary
oedema induced by nitrogen dioxide inhalation in the rat.
Arzneimittelforschung 1987 ; 37 : 519–23.
13 Vassilyadi M, Michel R. Effect of methylprednisolone on
nitrogen dioxide (NO )-induced pulmonary oedema in guinea
#
pigs. Toxicol Appl Pharmacol 1989 ; 97 : 256–66.
14 Meulenbelt J. Experimental study to investigate the effects of
intervention in acute nitrogen dioxide intoxication to improve
human treatment (Dissertation). Utrecht : Universiteit Utrecht,
1994.
15 Gunnarsson M, Bloom G, Jansson I, Walther S, Lennquist S.
Effects of corticosteroid inhalation on respiratory function
after chlorine-gas exposure (Abstract). Prehospital and Disaster
Medicine 1991 ; 6 : 230.
16 Gunnarsson M, Walther S, Holmstro$ m A, Jansson I, Lennquist
S. Nebulized corticosteroid improves pulmonary function after
chlorine gas exposure in pigs (Abstract). Prehospital and
Disaster Medicine 1993 ; 8 : 95–6.
# 1996 Blackwell Science Ltd Journal of Internal Medicine 239 : 451–456
456
C. K A R L S O N-S T I B E R et al.
17 Chester EH, Kaimal PJ, Payne CB, Kohn PM. Pulmonary
injury following exposure to chlorine gas. Possible beneficial
effects of steroid treatment. Chest 1977 ; 72 : 247–50.
18 Lorin HG, Kulling PEJ. The Bhopal tragedy – what has Swedish
disaster medicine learned from it ? J Emerg Med 1986 ; 4 :
311–16.
19 Milne JEH. Nitrogen dioxide inhalation and bronchiolitis
obliterans. J Occup Med 1969 ; 11 : 538–47.
Received 1 November1995 ; accepted 28 December 1995.
Correspondence : Dr Christine Karlson-Stiber, The Swedish Poison
Information Centre, Karolinska Hospital, S-171 76 Stockholm,
Sweden.
# 1996 Blackwell Science Ltd Journal of Internal Medicine 239 : 451–456