The Pulmonary Sequelae Associated with Accidental Inhalation of ...

The Pulmonary Sequelae Associated with
Accidental Inhalation of Chlorine Gas*
David A. Schwartz, M.D. , MPH.; Dorsett D. Smith, M.D. , F.C.C.; and
Sambasiva Liikslinzinarayan, Al. D. , F. C. C. P
Twenty previously healthy individuals were accidentally
exposed to high concentrations of chlorine gas in 1975.
Pulmonary function tests were performed on these individuals
on several occasions over the next 12 years. On average,
each subject was followed up for 8.5 years and 13 of the 20
exposed persons were tested 12 years following the exposure.
Pulmonary function tests obtained one day following
the accident were most notable for the high prevalence of
airflow obstruction and air trapping. Over the ensuing
years, the airflow obstruction persisted; however, the high
prevalence ofair trapping resolved. Ofnote, the prevalence
of a low residual volume consistently increased during the
follow-up period (p15
percent decline in FEy) to inhaled methacholine. Those
individuals with reactive airways were older (p 0.004) and
had more marked airflow obstruction (p 0.03) and air
trapping (p = 0.03) immediately following the exposure.
These data suggest that exposure to high concentrations of
chlorine gas may result in long-term pulmonary complications
that are characterized by a reduced residual volume.
Unfortunately, these data preclude us from determining
whether the chlorine exposure led to the development of
airway reactivity or the presence of reactive airways
accounted for the air trapping that was observed following
the exposure to chlorine gas. (Chest 1990; 97:820-25)
C hlorine gas can cause immediate injury to) the
re5)iratory tract and, if inhaled in sufficient
quantity, can result in the rapid development of
interstitial 1)neumonia, ‘ .2 ptmhnonary edema,3#{176}’ and
death due to) progressive respiratory failure.I,16 These
changes have been documented to) occur in laboratory
animals experimentally exposed to chlorine gas,7 soldiers
exposed to chlorine gas during World War I,
For editorial comment see page 773
and among individuals accidentally exposed to this
noxious gas. ‘“ The acute cytotoxic injury induced by
chlorine is manifested by extensive epithelial cell
sloughing” and is thought to be mediated by highly
reactive oxygen radicals generated by the hydration
o)f chlorine.2
While the acute manifestations associated with this
single high-dose exposure have been well documnented,
the long-tern effects reniain unclear. Follow-
(11) studies o)fVok)rld \Var I chlorine gas victims indicate
that tiian’ o)fthe survivo)rs were subse(luentlv disabled
l’ asthnia, bronchitis, amid emnphvsemna.5.m2 Of the six
rt’ports’#{176}#{176}’’’’’t1iat filloved tip individuals for at least
one ‘ear after aCCi(lemital eXpO)stlre, 0)111%’t\x0)#{176}t’ 1)mo)\’ide
(lata to) stiggest that these in(lividuals may be at risk of
(leveloping ChrO)IliC m’t’s)iI’atory 1)r0)l)lt’I1s fromn their
exposures. Otherwise, results from these follow-up
studies provide a very optimistic prognosis for individual
cases of acute chlorine gassings. An earlier report
from our group2 documented the occurrence of airflow
obstruction and air trapping shortly after the inhalation
of chlorine gas. This, however, was followed by a
pro)gressive loss of residual volume with six of the 11
study subjects demonstrating a residual volume below
80 percent O)f their predicted value two years after the
exposure. This finding is particularly interesting since
low residual volumes have been reported in other
individuals exposed to high concentrations of chlorine
gas.3’6
The purpose o)f this investigation was to determine
the long-term consequences of chlorine gas exposure
by reevaluating the lung function in a cohort of workers
who had l)een exposed to chlorine gas 12 years earlier.
We were particularly interested in determining if the
abnormally low residual volume was a persistent
finding and also whether these individuals had a high
prevalemice O)f hyperreactive airxvays. In addition, we
investigated the association l)etWeen early (within 24
hours ofexposure) o)I)jective findings and adverse longterm
consequences to) determine ifa higher risk group
Cotil(l be ideuitified i n the i mnn)e(liate postexposure
setting. Vt) our knowledge, this case series represents
the longest reported folhw-up l)erio(l for persons
exposed to) high concentrations of chlorine gas.
*Fro)mll tho’ PImlm110)narv 1)isvaso I)ivision, University of loeva, Iosva
( itv (I)r. So’iosvartz); tiso’ l)ivisiomo of Pulmiomoarv Diseases amid
( ritio’al ( ,oi’(’. t1m,iersit oof \Lsilimogtoon (I)rs. Sm1litil an(I Lakshmuimooro’amo);
aml(1 Seattle Veteramis Adnlinistratioml I Ioosl)itdl (I)m’.
I.aksiumuinaravan). Seattle.
Nlamlmls(-ril)t roceiveol Jmmno 22; ro’visio)mI d(l’t’1)t0ol Auguist S.
Reprint re(ju(’sts: !)r. Selmmcartz, tnirersiti, of lomt’a C3.3 GH, Iomva
(:it,1 .52242
ME’I’I lOI)S
Stmulq Population 070(1 I’Xp0507’(’ Cloaraeteri.s’ties
On Decemuher 17, 197.5, 20 pres’ioumsly healthy comistrimction
woorkers were exposed to) high c(oncemltratio)ns o)f chlorine gas. The
accident toook I)iace at a I)m011) fllill Wilile liojuid chlorine was being
ooff-loaoledf’ronld railroad car otmtside the i)mmilding to a hoddimig tank
820 Pulmonary Sequelae from Inhalation of Chlorine Gas (Schwartz, Smith, Lakshmiroara yen)
Downloaded From: http://publications.chestnet.org/ on 03/29/2014

Table 1 -Clinical Characteristics* ofExposed Workers at
the Time ofExposure by Length of Follow-up
All Workers
Follow-up Time, yr
(N=20) 12(N=13)

Tinie
After
Table 3-Lung Volumes and Diffusing Capacity*
(Mean ± SD) During 12-Year Period of Observation
Expoosure N TLC RV FRC DCO)
Day 1 19 108.2±19.6 141.6±45.0 108.5±19.5 . .
I)ay 10 20 110.3±15.4 124.1±30.9 106.9±25.3 . .
Day 40 17 108.7 ± 16.6 108. 1 ± 25.8 99.4 ± 20.2 94.6 ± 18.3
1 yr 15 102.7±14.8 98.9±28.6 90.3±20.6 94.5±26.1
2vr 13 99.9±15.3 88.2±24.4 90.8±18.0 90.4±15.5
9 yr 6 101.8± 19.0 83.9±37.2 100.8±31.7 110.8± 16.9
12r 13 102.5±12.9 82.3±39.2 93.9±23.1 105.6±23.0
pvaluet 0.48 0.0001 0.12 0.82
* Percent predicted.
tCalculated tosing amialysis oof variance.
With the exception of an initially high mean residual
vo)lume that progressively diminishes throughout the
years of follow-up, few changes are apparent in either
the spirometry measures (Table 2) or lung volumes
(Table 3). This relationship is also depicted in Figure
I where the prevalence of airflow obstruction (FEy3!
vC ratio 120
percent predicted) diminishes from 61 percent to 8
percent and the prevalence of study subjects with a
low residual volume (

100
w z
1
Lu
(1)
80
>
LU
LI.
60
Baseline 5 25 50
CUMULATIVE DOSE (Breath Units)
FIo:oIOF: 2. Percentage of Loll fromml baseline FE\’
0 )i)so-rved soit Io tiot- didn’t-s iateol mnoiihe’l to uline
airway challenge iii tie’ 13 o.xisoseol u-orkers
1 50 tested 12 vo’.ors tI’ter oxposuro to (-illo)rimoo’ gos.
One in’eatio ommoitis coma1 too0)3W illll;llatio)IIod 1
mug/mill of mnttilao’holino
those with po)siti\’e airway responses to) methacholine
(N = 5) tended to) be older (p 0.004) and demonstrated
airflow obstructio)n (p = 0.03) with more air
trapping (p = 0.03) on standard pulmonary function
tests obtained within 24 hours of the exposure (Table
5). After controlling for age, we found that the degree
of air trapping on initial pulmonary function testing
was directly associated with airway reactivity only
among those who) were at least 35 years of age at the
time of the exposure to chlorine gas (data not shown).
The limited number of sUl)jects precluded a formal
multivariate
analysis.
D1s1sS1oN
Our findings suggest that short-term exposure to)
high concentrations of chlorine gas may lead to longterm
respiratory complications. We found that accidental
exposure to chlorine gas acutely maiuifests as
air trap)i11g. However, during the 12-year period of
observatio)n, we noted a progressive decline in residual
volume with 67 percent of those tested at year 12
having a residual volume below 80 percent of their
predicted value. This finding should be contrasted
with an expected increase in residual volume of 20
mi/yr given the age and height characteristics of our
study population. ‘ In addition, we found that five of
13 subjects tested at year 12 had an increased airway
reactivity to) inhaled methacholine and that the risk of
having reactive airways was directly related to) the
degree o)fairflow obstruction and air trapping that ‘as
observed immediately following the exposure.
The loss of residual volume iii our chlorine-exposed
cohort represents an important o)hservatio)n that helps
us to) piece together the findings from the other studies
that have addressed this type of ouma6mi.mS In
Table 6, we review the six other case series that have
evaluated the short- and long-term (at least one year
of follow-up) pulmonary changes in individuals acci-
dentally’ eXI)osed to high co)ncent rationis of chlori no’
gas. Of the four reports that actually prestiit ptilinonar
function results,l,10I5 studies56 report lower
than expected values for residual volume at tho’ emid
of the follow-up period. However, the most comprehensive
sttidy’5 to) date has found minimal changes in
residual \‘O)ltlfllC ( - 1 mllvr) during the six-ear folloxx’-
up peno)d. The inconsistency ofour findings with this
pre’is repo)rt is difficult to) explain l)LIt mnay reflect
the different exposure conditio )ns of on r respective
cohoorts. 1mm the only study2 that systemnatically revie\Ved
1)atients with isolated redtictioiis in I’csi(lual
volunie, toxic iruhalatiomus (ho’drocarbomus, amnmiio)ilia,
Table 4-Relationship Between Residual Volume at Year 12
and Clinical Charaiteristics* Within 24 hours of Exposure
to Chlorine Gas
o-ategorio.- variables are exI)resseol as tue frt’oiue’no.- (1o’rco-mot of total
smobjeets).
i10Si(iu1al\olimmoot it loam’ 12

Table 5-Relationship Between Airway Response to
Methacholine Challenge at Year 12 and Clinical
Characteristics* Within 24 Hours ofExposure to
Chlorine Gas
Niethacholin e Cllallemlge
Poositive
Negative
(N=5) (N=8) 1)dltme
Ageat exposure. yr 48.7±9.6 33.7±5.5 0.004
Smoking hishor; Noo. (%)
Never . . . 1 (13)
Ex 3 (61)) 6 (75) 0.42
CLmrre’mlt 2 (40) 1 (13)
Day 1 CXR, No. (%)
Infiltrates . . . 3 (:38) 0.23
Day 1 PFTst
FEy 75.8±29.5 86.6±24.2 0.48
FEV/VC ratio 59.8± 14.5 74.9±6.7 0.03
V(: 96.3±21.4 96.0±22.8 0.86
LIV 183.9±60.6 125.0±26.5 0.03
TIA: 126.4±20.7 99.2±17.6 0.03
FRC 128.3±5.6 99.9± 17.5 0.005
*(‘o,,Jti,)m300mos variai)les are (.xpresso(I as tue mucan ± SI) so bile
categoric variables are exl)ressed as tile freoiuencv (percent of total
subjects).
tPF”I’s: individual measures ni I)mmlmlloonarY function are tll expressed
as a I)t’rco’mot prt-olicteol eX(’(’pt for the FEVA’C ratio.
sulfur dioxide, and pesticides) were l)elieved to be
responsible for 16 percent of the cases. Although the
clinical significance of this physiologic abnormality
remains unclear, the reduced residual volume observed
in tir sul)jects is supported by other studies
and may represent a characteristic feature of lung
function following short-term exposure to this agent
as we11 as other no)XiouS gases. 20
The observed reduction in residual volume suggests
that the initial injury may result in chronic stiffening
of the airways. Since residual volume is determined
l)y the strength of the expirato)ry muscles and the
pressure-volume characteristics of the lung and chest
wall, reductions in residual volume result from an
increase in either muscle strength, chest wall compliance,
o)r ii) the elastic recoil of the lung. In healthy
persons older than 35 years of age, residual volume
has been found to be largely determined by airway
co)llapse.#{176}’ Presuming that chlorine gas has not affected
the mechanics O)f the chest wall, two possible explanations
exist for the observed reduction in residual
volume. First, diffuse parenchymal injury induced by
chlorine gas may result in mild roentgenographically
inapparent interstitial fibrosis that could increase the
elastic recoil O)f the lung. The increased elastic recoil
would result in a low residual volume by its tethering
effect on small airways, tending to) maintain their
patency at low lung volumes while permitting a nearly
normal expansion at total lung capacity. Alternatively,
acute airway damage may result in mild peribronchiolar
fibrosis that may stiffen the bronchioles and cause
them to be more resistant to closure during exhalation.
Maintaining airway patency at low lung volumes would
permit continued exhalation and result in a reduced
residual VO)luflle. Animal studies7 and po)stmorten)
sttidies in tw’() humans#{176} suggest that the latter hypothesis
is more likely: W7ithin hours ofexposure to chlorine
gas, patholo)gic specimens demonstrate extensive
sloughing of airsvay epithelia, svhich is followed by an
intense inflammatory response . Animals killed between
two weeks and six months following the exposure
were found to have persistent airway inflammatiO)fl
with extensive bn)nchitis, bronchiectasis, and
bronchiolitis n Thus, in some cases, acute
extensive airway injury is followed by persistent airway
inflammation and fibrosis. The peribronchial fibrosis
comld account for the persistent decline in residual
volume that sve observed among our expo)sed construetiO)I1
workers. However, without histologic confirmation,
the mechanism(s) accounting for this finding
remains speculative.
Although the prevalence of positive airway re-
SOn5C5 to) methacholine in our exposed cohort is
relatively high (38 percent of subjects tested), given
the high prevalence of smoking in this cohort, it is
Table 6-Reports ofAccidental Exposure to Chlorine Gas in Which the Subjects Were Followed upfor at Least One Year
Type
of
Noo. ool Subjects
- Folloow-up
Puinuomiary Fimmlction Tests
. ‘ .
Accident, Referemce Exposed Foollonved-up Time, yr Initial End of Fodioeo’-up
New Yoork (:it sui)wa\;’ 2()8 3.3 1.3 Not obtailled Not ooiotaimied
Loongshoretnan, 156 1 1 1.7 , FRC Noormnal
Lo)ngshooreman, 156 59 1 .6-2.9 Not oohtained Airway resistance
FEy amid IRV
Freight traimi,” 1(X) 12 7.0 Not o)htaineol J RV (5/7 tested)
FEy, (5/10 tested)
Filtration plant,#{176} 35 18 1.2 RV (7/18)
1 RV (3/12)
I FEy (5/18)
FE!I (3/12)
fecompatioonal,’’ 56 56 5.0 Noot rep.tortoi Noot reported
Freight traim,#{176} 145 61) 6.0 Nornlal spiroometr
and lumig volimnies
Normal PF’Ts with
boss oof 1 nol.5’r in RV
824 Pulmonary Sequelae from Inhalation of Chlorine Gas (Schwartz, Smith, Lakshminarayan)
Downloaded From: http://publications.chestnet.org/ on 03/29/2014

consistent with expected 27 However, after controlling
for age, we found that the degree of air
trapping on initial (within 24 hours of the exposure)
pulmonary function testing was directly associated
with airway hyperreactivity 12 years following the
exposure. These findings suggest that the early development
ofair trapping following chlorine gas exposure
may be predictive of those who go on to develop
bronchial hyperreactivity. Alternatively, those vorkers
who had more air trapping 0!) initial lung function
testing might also have underlying airflow obstruction
prior to the exposure and thus, they may be more
likely to) respond to methacholine 12 years after the
exposure. Unfortunately, these data preclude us from
determining whether the chlorine exposure led to) the
development ofhyperreactive airways or the presence
ofhyperreactive airways accounted for the air trapping
that was observed immediately following the exposure
to chlorine gas.
In the absence of preexposure pulmonary function
tests and individual measures of exposure to chlorine
gas, it is difficult to l)e certain that the changes we
observed were, in fact, caused by chlorine gas. However,
given the initial symptomatolog; pul n)onary
function al)normalities, and chest roentgenographic
findings, it is clear that these workers developed acute
respirato)ry problems associated with their exposure
to this no)xious gas. The diminished residual volume
observed years fillowing the exposure is unusual, and
given the nature of the injury, it appears to be related
to) the exposure. In the aI)sence of associated respiratory
symptoms, the exact clinical significance of this
physiologic abnormality remains unclear. The association
between the early development of air trapping
following the exposure and reactive airways may prove
to) be related to chlorine gas exposure; however, the
temporal sequence ofthese events needs to be further
clarified.
REFERENCES
1 Kaufman J, Btmrkoons D. Clinical, roentgemlooloogic, and physioloogic
effects of acute chlorine exposure. Arcil Environ ilealth 1971;
23:29-34
2 Charan NB, Lakshminaravan 5, Meyers CC, Smith DD. Effects
of accidental chlorine inhaiatiom in pulnloonary functioon. Vest J
Med 1985; 143:333-36
3 Kowitz TA, Reba RC, Parker RT, Spicer WS. Effects oof chlorine
gas upton respiratoory functioon. Arch Emoviron health 1967;
14:545-58
4 Adelsorn L, Kaufman J. Fatal chlorine pooisoning: report of twoo
cases with clinicopathooloogic coorrelatioon. Am J Clin Pathooi 1971;
56:430-42
5 Joovmier RE, I)urel E(;. Acciolo-motal li(lui(i chlorimie spill in d rural
conlmunit\ J Occump Med 1962; 4:152-54
6 Weill ii, (;eo)rge R, Schwarz NI, Ziskind NI. Late evaluationi of
pulmonary functioon after acmmte exposumre to) chloorine gas. Ammo
Rev Respir Dis 1969; 99:374-79
7 \Vinternits. Nl(:. Pathologv of war gas pooisonimig. New ilmven,
Conn: Yale University Press; 1920.
8 (;ilchnst IlL, Mats PB. The residual effects of warfare gases:
tile use oof cilloorimle gas, with report of cases. Nied Bull VoL
Adnlin 1933; 9:229-68
9 Sandall TE. The later effects oof gas poisooning. Lancet 1922;
2:857-59
10 Berghoff RS . The niore comumnon gases: tileir effect omi the
respiratory tract. Arch Intern Med 1919; 24:678-84
1 1 Ahbott ‘N . The incidemue of pulmonary disease folloaving
(-xpoosure too vesicant amlol asphyxiating gases. Br Meol J 193.3;
2:862-65
12 Hankimls JL, Klotz \VC. Permmoinent I)lmlnlonar effects of gas in
warfare. Aol Rev Tuioerc 1922; 6:57 1-74
13 Barret L, Famire J. Cilloorille pooisonin Iletterl. L.anct-t 1984;
1:561-62
14 Chasis ii, Z.aI)i) ,JA. Bannomi Jil, Vhittemlix-rger JL, ilo-Im J,
I)ohemlv JJ, et al. Cillowine acc.’iolemlt im Broooklvm. Occmmp Nleoi
1947; 4:152-76
15 Jones RN . llmmghes J, Clinolmever ii. \V-ill II. Lumug fmmnctiomu
after acumto o-hlorimie’ exposure. Am Rev Respir Dis 1986; 134:
1190-95
16 ATS Statenio’mot. Snowbird workshoop omo staml(lar(lizatiomo oof spiro)nletrs
Am Rev Respir I)is 1979; 119:8.31-38
17 olorris JF, Kooski A, Joiinson LC. Spiromnetric staml(lards for
healthy mlo)nsmnoking adults. Am Rev Respir Dis 197!; 10.3:57-67
18 (;oldmaml iii , Becklake NIR. Rospiratorv fimmictiomi tests: noormal
values at mileolian altitudes and the predictioon oof noornial results.
Anl Rev Tmmioerc 1959; 79:457-67
19 \on (;anso’ \\‘F, Ferris BC, Cotes JE. Cigarette snooking and
l)t1lmllmlarY difl’usiiug capacity (tramisfer factor). Aol Rev Respir
Dis 1972; 11)5:30-4 1
20 Chathamu NI, Blo-ecker ER, Normal P. Smith PL, Niasoomo P A
screemling test for airsvavs reactivity: .111 alobreviateol muetilacia)-
line inhalationo challenge. Chest 1982; 82:1.5-8
21 Coltom T. Statistics in medicine. Bostoml, Mass: Little Brown &
(:oo; 1974
22 Docker D\’S \Vare Jil, Ferris BC, Clicksherg DS, Fav ME,
Spiro A, et al. I)istriinmtioon oof forced expiratoory vo)lumne in oma
secoml(l amld forced vital capacity in health); white, adult nevem
smookers in six U.S. cities. Am Rev Respir Dis 198.5; 131:51 1-21)
23 Bumrroows B, I.eboowitz MD, Camilli AE, Knudson RJ. i.omgitudinal
cilanges iml forced expiratorv vo)lunle in oone secootul ill
adomlts. Am Rev Respir Dis 1986; 133:974-81)
24 Owens M\’u Kimiasewit-s CT, Andersoon WM . Clinical significamce
oof all isO)lated reductioon in residual voolionie. An Rev
Respir Dis 1987; 136:1377-80
25 Nlurrav JF. The no)rtnal lummig. Philadelphia, Pa: \VB Saumnders
Go; 1986
26 Leith DE, Mead J. Mechanisms determining residual volume
oftile lungs in mooormnal sui)jects. J AppI Physiool 1967; 2.3:221-27
27 Sparrow D, OComlmlor C, Colton T, Barry CL, Weiss ST. The
relatioonship oof mlOmlspeeific bronchial responsivemless too the
ooecurrence of respiratory sVnlpto)fllS amld do-creased levels of
pulmoonary fmoi.ction. Am Rev Respir Dis 1987; 13.5: 125.5-61)
CHEST I 97 I 4 I APRIL 1990 825
Downloaded From: http://publications.chestnet.org/ on 03/29/2014