.&r The poisoningThepathologyofthe lung inparaquatpoisoning paraquat poisoning as an experimental...

J. clin. Path., 28, Suppl. (Roy. Coll. Path.), 9, 81-93

.&r The pathology of the lung in paraquat poisoningPAUL SMITH AND DONALD HEATH

From the Department ofPathology, University ofLiverpool

There have been numerous fatalities followingaccidental or deliberate ingestion of the weedkillerparaquat and these have received great attentionboth in the popular and scientific press. Mostfatalities have occurred following the oral ingestionof Grammoxone, the concentrated form of thesubstance. Although this is not supposed to bereadily available to the general public, people

$ illicitly obtain samples of the weedkiller which theystore at home in lemonade or beer bottles. In thisway there is a great danger that the paraquat willbe accidentally drunk by children or even unwaryadults. Many fatalities have resulted from thiscareless disregard for precautions since only one

v mouthful of Grammoxone appears to be all that isrequired for a fatal outcome.The disease is confined almost exclusively to the

lungs and consists essentially of a fulminating

f~~

pulmonary fibrosis which causes death fromrespiratory failure. The development of pulmonarylesions can be divided into two stages: a destructivephase and a proliferative phase.

The Destructive Phase of Paraquat Poisoning

The primary target for the destructive action ofparaquat is the alveolar epithelium. In experimentalrats there is a swelling of the type I alveolar epithelialcells or membranous pneumocytes only four hoursafter an intraperitoneal injection of paraquat(Smith and Heath, 1974a). By eight hours afterinjection membranous pneumocytes show vacuola-tion and disruption of organelles (Smith and Heath,1974a) and by 18 hours these cells show hydropicdegeneration in the form of numerous large vacuo-lated swellings projecting into the alveolus

;. ,; > . 7tW-..4w.

;iv,4"N

AVI

2K

L

Fig 1 Granularpneumocyte from arat 24 hours afteran injection ofparaquat. The cellshows degenerativechanges consisting ofswelling ofmitochondria (M) andvacuolation oflamellar bodies (L).There is alsodisruption of therough endoplasmicreticulum. Electronmicrographx 25000.

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.s3-9.1.81 on 1 January 1975. Dow

nloaded from

http://jcp.bmj.com/

82 6,U*.. ....aS.....

.Sc:.b~~~~~~~~~~~~~~IJ.....p*...:m4 :.::'$a ;N

wb

4U .

#. /

Paul Smith and Donald Heath

a

Fig 2 Three days after an injection ofparaquat.Alveolar capillaries are congested with blood and thereis alveolar pulmonary oedema. Neutrophil polymorphsare scattered about the lung. The intraalveolarmononuclear cells may be the first few profibroblasts toinfiltrate the lung. Haematoxylin and eosin x 375.

Fig 3 Human case ofparaquat poisoning in which death.from renalfailure supervened while pulmonary lesionswere at an early stage. The figure shows the destructivephase ofparaquat lung in man comprising alveolarcapillary congestion, alveolar oedema, fibrinous exudate,and neutrophilpolymorphs. The respiratory bronchiolesare lined by hyaline membranes. Hand E x 150.

(Vijeyaratnam and Corrin, 1971; Smith and Heath,1974a). Also at this time after injection degenerativechanges occur in the type II alveolar epithelial cellsor granular pneumocytes. These take the form ofswelling of mitochondria, vacuolation of lamellarbodies and disruption of the endoplasmic reticulum(figure 1). Two days after the administration ofparaquat both granular and membranous pneumo-cytes start to disintegrate so that by three days thealveolar walls are denuded of their epithelial lining(Smith and Heath, 1974a). It is interesting that thiswidespread destruction does not involve the alveolarcapillaries.

In the rat, the sloughing of the alveolar epitheliumis associated with alveolar pulmonary oedema,capillary congestion and a mild acute inflammatoryreaction (figure 2). Eosinophilic hyaline membranesare also common (Manktelow, 1967; Robertson,Enh6rning, Ivemark, Malmqvist, and Modee, 1971;Smith, Heath, and Kay, 1974). Alveolar pulmonaryoedema during the early stage of paraquat poisoningin animals is often sufficiently extensive to causesevere dyspnoea, and many animals may die in thisacute phase of the disease (Clark, McElligott, andHurst, 1966; Smith et al, 1974).

In the destructive phase of paraquat poisoningin man, disintegration and sloughing of the alveolarepithelium also occur (Toner, Vetters, Spilg, andHarland, 1970). Although the sequence of changesleading to this lesion have not been determined it isprobable that they are similar to those in the rat.Alveolar pulmonary oedema, capillary congestion,hyaline membranes and an acute inflammatoryexudate are also found (figure 3). In man thesechanges are rarely fatal but may induce dyspnoea.Gardiner (1972) has demonstrated clinically thepresence of pulmonary oedema during the earlystages of paraquat poisoning.The reason for the liberation of oedema fluid into

the alveolar spaces is unclear since paraquat doesnot damage the pulmonary capillaries. Increasedvascular permeability may account for the oedema(Vijeyaratnam and Corrin, 1971) but Gardiner(1972) has suggested that loss of pulmonary surfact-ant may be responsible. It is commonly acceptedthat the granular pneumocytes secrete pulmonarysurfactant and, therefore, destruction of these cellsby paraquat would lead to loss of surfactant with acorresponding increase in surface tension of thealveolar fluid. This could then withdraw fluid fromthe alveolar capillaries to produce oedema. Such anincrease in surface tension has been demonstratedin animals (Manktelow, 1967; Robertson et al,1971). These authors also suggest that loss ofpulmonary surfactant leads to the formation ofhyaline membranes, and for this reason they propose

'4

4.

4

4

I

'411

4

4'-A

46



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

The pathology of the lung in paraquat poisoning

paraquat poisoning as an experimental model forthe idiopathic respiratory distress syndrome of thenewborn.During the destructive phase of paraquat poison-

ing, damage to organs in the systemic circulationmay occur. Thus, patients often show clinicalevidence of hepatic and renal failure which usuallyresolves, after a few days (Almog and Tal, 1967;Campbell, 1968; Matthew, Logan, Woodruff, andHeard, 1968; Fennelly, Gallagher and Carroll,1968). Occasionally renal failure may be the cause ofdeath (Oreopoulos, Sayannwo, Sinniah, Fenton,McGeown, and Bruce, 1968; Smith and Heath,1974b).

The Proliferative Phase in Animals

FOLLOWING A SINGLE DOSE OF PARAQUATApproximately three days after a single injection ofparaquat into rats scattered mononuclear cells arefound in the alveolar spaces (figure 2), and by sevendays alveolar spaces are filled with them (figure 4).These cells are irregular in shape and measurebetween 6 and 10 um in diameter. They present a'ragged' outline, with non-vacuolated, moderatelyeosinophilic cytoplasm. Their nuclei are large, ovoid,

Fig 4 Rat seven days after an injection ofparaquat.The lung is heavily infiltrated with profibroblasts whichfill the alveolar spaces. Alveolar walls are disrupted.Note the irregular 'ragged' outline of the cells and theirlarge, dark, eccentric nuclei. H and E x 600.

83

darkly staining and eccentrically placed. Althoughthese cells may superficially resemble macrophagesthey are not phagocytic but differentiate intofibroblasts. For this reason we have called themimmature or profibroblasts (Smith et al, 1974).The ultrastructure of a profibroblast is shown in

figure 5. Their most prominent feature is the posses-sion of numerous, long, filamentous pseudopodia attheir periphery (figs 5 and 6). Their oval nuclei aresmooth in outline withlarge quantities of dark chrom-atin. Mitochondria are small, rounded and dense withtightly packed cristae. At low magnifications mito-chondria can be confused with the few smalllysosomes also present in the cytoplasm (fig 5).Rough endoplasmic reticulum is scanty and consistsof undulating parallel pairs of membranes bearingribosomes. The pairs of membranes are unbranchedand the spaces between them (cisternal spaces) arevery narrow (fig 5). These profibroblasts are presentexclusively within the alveolar spaces (fig 6) (Smithet al, 1974; Smith and Heath, 1974a).

Profibroblasts rapidly undergo differentiation intomature fibroblasts and the sequence of changesinvolved in this maturation has been traced in therat (Smith et al, 1974; Smith and Heath, 1974a).Initially the cells become larger and more basophilicand then they elongate (fig 7). Mitotic figures aresometimes encountered. Later the cells undergofurther elongation and increase in basophilia untilthey are mature fibroblasts. During this maturationthe alveolar walls become disrupted (fig 7) untileventually they are unrecognizable. At an ultra-structural level the maturation of profibroblastscommences with the withdrawal of pseudopodia,enlargement of mitochondria and an increase inquantity of rough endoplasmic reticulum (fig 8).This is followed by branching of the rough endo-plasmic reticulum and widening of cisternal spaces.Progressive elongation of these cells produces maturefibroblasts (fig 9). It has been demonstrated thatthese fibroblasts are present exclusively within thealveolar spaces (fig 10) and that alveolar walls arenot infiltrated in any way (Smith et al, 1974). Onthe contrary alveolar walls are severely disruptedand often all that remains is the alveolar capillaries(figs 6 and 10). The histological appearance of thelung at this final stage of paraquat poisoning is thatof a dense mass of fibroblastic tissue with almosttotal obliteration of the lung architecture (fig 11).Death usually supervenes about 10 days after injec-tion before much collagen can be secreted. Electronmicroscopy shows small quantities of collagencontained within lacunae in the fibroblasts (fig 10).In this terminal stage of paraquat lung the originof the pulmonary fibrosis is difficult to determine, soextensive is the obliteration of normal lung structure.

'*4kA#



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


C.',P.. :, C

. b,'

!

9q

v g

AM;.'.: jA...

Fig 5 Electron micrograph from the same rat as in fig 4 showing details of a profibroblast. It has a large oval nucleus(N) and the periphery is thrown into numerous filamentous pseudopodia. Mitochondria (M) are inconspicuous and therough endoplasmic reticulum (R) consists of a few parallel membranes with narrow cisternae. A few lysosomes (L) arealso present. Electron micrograph x 18 750.

-4,~~~~~~~~~~~C

RA

0:... -11,

4.1



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


3 -..........

4~~~~~~~~~j~ ~ -

01~~~~~~~~~1

~~~~~....>.....;ork

Fig 7

Fig 6

It can sometimes resemble fibrosing alveolitis, andhas been described in the literature as a type ofinterstitial pulmonary fibrosis (Clark et al, 1966;Vijeyaratnam and Coffin, 1971). However, thesequential light and electron microscopic studiesdescribed above demonstrate that, in rats, the fibrosisis a diffuse, cellular, intraalveolar fibrosis of the lung.Apart from its intraalveolar origin, paraquat lungdiffers from fibrosing alveolitis in other respects.Thus, it is more cellular, more vascular, its growth

X' more rapid, and the quantity of collagen much less

Fig 6 Lung ofa rat treated with paraquatshowing a profibroblast (P) within thealveolar space (A). Note the filamentouspseudopodia at the periphery of theprofibroblast. Alveolar capillaries (C) can beseen and their endothelium (E) isundamaged. The alveolar epithelium is totallyabsent and all that remains is its basementmembrane (B). Electron micrographx 12500.

Fig 7 Lung ofa rat showing the alveolarspaces totally occluded by profibroblasts.Many of these cells are now elongated andresemble mature fibroblasts. The alveolarwalls are severely disrupted. H andE x 375.

than in fibrosing alveolitis. Neither does its patho-genesis involve a phase of proliferation and des-quamation of granular pneumocytes as occurs inearly fibrosing alveolitis (Scadding and Hinson,1967; McCann and Brewer, 1974).

FOLLOWING REPEATED DOSES OF PARAQUATWhen small doses of paraquat are administeredrepeatedly over a long period of time exudative andacute inflammatory changes are much less conspic-uous. If the paraquat is administered by repeated

85



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Fig 8 A cell intermediate inappearance between a profibroblasand a mature fibroblast. Notethat it lacks pseudopodia.Mitochondria (M) are larger andhave a paler matrix than infigure 5. The rough endoplasmicreticulum (R) is more extensiveand cisternae are wider. Electronmicrograph x 25 000.

Fig 9 Part ofa mature fibroblast fromwithin an alveolar space. The cell iselongated, has prominent mitochondriaand its rough endoplasmic reticulum is fextensive with dilated cisternal spaces.Electron micrograph x 50 000. *

4

I



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

** t KiiR N

IF~~~~~~~~~~~~

~~~~~~~~~~~4414

-_

elL,b;.#f l -a30^. 2> #olV + x @t

w81it, rvs 8 b - 7

---4zs 1 ~~Fig 10

Fig 10 Lung ofa rat 10 days after an injection ofparaquat. A fibroblast (F) can be seen within the alveolarspace. The alveolar wall consists of capillary endothelialnucleus (E) and capillary lumen (C) containingerythrocytes and a platelet. The fibroblastic reaction ispurely intraalveolar. Collagen fibrils (Col) can be seenin the vicinity of the fibroblast. Electron micrographx 18 750.

Fig 11 Lung ofa rat 10 days after an injection ofparaquat. The lung architecture is obliterated by a dense,mass offibroblasts and small quantities of collagen.H andE x 375.

Fig 11



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


injection (Smith et al, 1974) or in the diet (Clarket al, 1966) pulmonary fibrosis results after a variableperiod of time. The pathogenesis of this fibrosis isthe same as after a single dose (Smith et al, 1974)although the end result may be a looser fibrosis withfewer fibroblasts and more collagen (fig 12).

Occasionally one may encounter foci of interstitialfibrosis with patent alveoli containing foamy macro-phages. There may be an associated hyperplasia ofgranular pneumocytes on the alveolar walls. It isunclear why this lesion should develop although anidentical picture has been described in parts of thelungs of rats after prolonged intoxication withCrotalaria spectabilis seeds (Smith et al, 1970). Sincethese foci are uncommon it is unlikely that they playa significant role in paraquat lung. The proliferationof granular pneumocytes may represent areas of

:~~~~~~~j;S/j,AFig 12 Rat following repeated injections ofparaquat.There is a gland-like proliferation of the bronchiolarepithelium into surrounding alveolar ducts and alveoli.The fibrous tissue is much looser and more open inappearance than is seen after a single injection.HandE x 150.

epithelial regeneration where damage to the alveolararchitecture has been less severe. We have neverseen these changes in rats after single doses ofparaquat.Another lesion which is commonly found after

repeated doses of paraquat is a proliferation of thebronchiolar epithelium (fig 12) into surroundingalveolar ducts and alveoli (Clark et al, 1966; Smithet al, 1974). This may also be an attempt at regenera-tion. The lesion is, however, a common finding inhuman necropsy cases where a single dose ofparaquat has been taken.

V.~~~~

5 &,

IF~~~~"Fig 13 A human case ofparaquat poisoning showingobliteration of the lung architecture by a dense mass offibroblastic tissue. The large air spaces are dilatedrespiratory bronchioles which give the lung an earlyhoneycomb appearance. Lakes of haemorrhage can alsobe seen. HandE x 60.

400 ~j&4

~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~.

;'. s.~~~~~~~~~~~~~'4

Fig 14 Detail of the fibroblastic tissue shown infigure 13. The fibroblasts (F) are plump and basophilicand merge imperceptibly with fine collagen fibrils. Muchof the intercellular miatrix consists offibrin and groundsubstance. The fibroblastic tissue includes numerouscapillaries (C). There is an infiltrate of lymphocytes andplasma cells(H;and E x 600).

88



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


s . _a 4 ssj4 _..:s W. Bx .}. *.

alk ?¢N;

Fig 15 Human case ofparaquat poisoning showingearly intraalveolar fibrosis. Alveolar walls persist and canbe seen as chains of capillaries (arrowed). The alveolarspaces contain fibroblasts but the alveolar walls are notinfiltrated in any way. Hand E x 285.

AM

.iw4

tob N\tvW > t~~~~ 4

Fig 16 Same case as figure 15 showing a differentarea of lung. There is extensive intraalveolar fibrosisalthough the alveolar wall is more severely disruptedthan in the previous figure. Hand E x 375.

The Proliferative Phase in Man

Accidental or deliberate ingestion of paraquat byman commonly leads to the development of pul-monary fibrosis which causes death between oneand three weeks after ingestion. Elucidation of thepathogenesis of this fibrosis is hampered by the factthat one is usually examining the advanced terminalstage of the disease. The histological picture typicalof advanced paraquat lung in man is shown infigure 13. It consists essentially of a dense mass offibroblastic tissue causing obliteration of the normallung architecture. Surviving respiratory bronchiolesmay show dilatation to produce an early honeycombappearance (fig 13) (Matthew et al, 1968; Smith andHeath, 1974b). In a few cases there is also pulmonaryhaemorrhage but this is not an invariable part of thehistology. The fibroblastic tissue consists of numer-ous plump, basophilic fibroblasts with an acellularfibrillary matrix between them (fig 14). The matrixstains only faintly with Van Gieson's reagents andconsists largely of immature collagen and fibrin. Itis infiltrated by lymphocytes and plasma cells (fig14) (Smith and Heath, 1974b).Thepulmonaryfibrosisalso includes numerous capillaries (fig 14) which arepresumably derived from alveolar capillaries follow-ing breakdown of the alveolar walls. In some cases,surviving alveoli may be lined by a continuousepithelium of granular pneumocytes (Matthew et al,1968; Smith and Heath, 1974b) and there is com-monly a gland-like proliferation of the bronchiolarepithelium.

This sort of histological picture in the lung inparaquat poisoning is often described as being aninterstitial pulmonary fibrosis (Bullivant, 1966;Fennelly et al, 1968; Matthew et al, 1968), andparaquat lung is now held by many to be an exampleof diffuse fibrosing alveolitis as defined by Scaddingand Hinson (1967). Recently we described a case ofparaquat poisoning which suggested that this wasnot the case (Smith and Heath, 1974b). In this casethe fibrosis was less dense and had caused lessobliteration of the lung architecture than normal.Also, the alveolar walls, although damaged, hadpersisted and were easily identified as chains ofcapillaries containing red blood cells (fig 15). Theseacted as markers for the situation of the pulmonaryfibrosis which was present exclusively in the alveolarspaces (fig 15). The fibrosis consisted of elongated,basophilic strap-shaped fibroblasts with long cyto-plasmic prolongations which merged imperceptiblywith ground substance and fine collagen fibrils(fig 16). A few cells were seen which were round oroval with an irregular 'ragged' outline, unvacuolated,moderately eosinophilic cytoplasm and large, darklystaining nuclei. These cells were similar in appearance

89



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


to profibroblasts which have been described inconnexion with paraquat-induced pulmonary fibrosisin rats.We interpret these observations as representing an

earlier stage of the proliferative phase of paraquatpoisoning than one usually encounters. In this casethe origin of the pulmonary fibrosis was demon-strated as being intraalveolar and not interstitial.We suggest that the pathogenesis of human paraquatlung may parallel that in the rat and consist of anintraalveolar infiltration of profibroblasts followedby their maturation into mature fibroblasts. Cellsresembling profibroblasts were seen in the case justdescribed and similar cells were observed in a humanlung biopsy specimen (Toner et al, 1970), but wereinterpreted as a monocyte infiltration. At a sub-sequent necropsy of this case, these workers des-cribed an extensive pulmonary intraalveolar fibrosis.Thus the final stage of paraquat poisoning, likethat in the rat, may be reminiscent of fibrosing alveo-litis but obliteration of the alveolar walls belies itstrue intraalveolar origin. We think that the basicpathological change is a diffuse, cellular, intra-alveolar fibrosis.

The Influence of Oxygen Therapy

The severe dyspnoea experienced by patientspoisoned by paraquat requires treatment withartificial respiration of gas mixtures containingahighpartial pressure of oxygen (Almog and Tal, 1967;Campbell, 1968; Oreopoulos et al, 1968; Matthewet al, 1968; Toner et al, 1970; Smith and Heath,1974b). Prolonged respiration of gases containingmore than 40% oxygen carries a grave risk of oxygenpoisoning (Sevitt, 1974). In this condition there isan early exudative phase consisting of interstitialand alveolar oedema with the formation of eosino-philic hyaline membranes. This is followed by aproliferative phase after five days of oxygen therapyin which there is a hyperplasia of granular pneumo-cytes and the development of interstitial pulmonaryfibrosis (Nash, Blennerhassett, and Pontoppidan,1967; Sevitt, 1974). In some cases reviewed by Sevittthere was also an intraalveolar fibrosis. One should,therefore, consider the possibility that oxygentoxicity may be an important factor in the patho-genesis of pulmonary fibrosis in some cases ofparaquat poisoning. It is clear that paraquat iscapable of producing pulmonary damage on its ownsince intraalveolar fibrosis occurs when paraquat isgiven to rats breathing only atmospheric oxygen.Furthermore the histopathology of paraquat lungand oxygen poisoning is somewhat different in that,although intraalveolar fibrosis may occur in thelatter, there is always a prominent interstitial

component. Neither is there a profuse hyperplasiaof granular pneumocytes in paraquat lung sincethese cells disintegrate and most alveolar walls aredisrupted and become engulfed by fibroblastictissue. One cannot, however, completely excludeoxygen toxicity as a contributory factor in thepathogenesis of pulmonary fibrosis in some casesof paraquat poisoning in man.

The Mechanism of Toxicity of Paraquat

Paraquat in its concentrated form (Grammoxone)has a strong irritant action on various types ofepithelia. Thus it will cause blistering and sorenessof the skin (Swan, 1969), damage to the cornealepithelium when splashed into the eyes (Swan,1968), and damage to the nail-bed of the fingerscausing shedding of the nails (Samman and Johnston1969). Many patients following oral ingestion ofGrammoxone show ulceration of the epithelium ofthe buccal cavity and fauces. It may be that a similarcaustic action is responsible for the disintegrationof the alveolar epithelium during the destructivephase of paraquat poisoning. However, whenGrammoxone is diluted its dermal toxicity is verymuch reduced (Swan, 1969), and it seems unlikelythat its concentration in the blood followingingestion would be high enough for the alveolarepithelium to be damaged merely by the causticnature of the compound.

Paraquat appears to kill plants by entering thechloroplasts and then taking part in an oxidation-reduction cycle in which hydrogen peroxide isliberated (Conning, Fletcher, and Swan, 1969). Asimilar type of calalytic activity may occur in animaltissues in which a small concentration of paraquatcan lead to the synthesis of a high concentration oftoxic byproducts. The highly vulnerable alveolarepithelium may succumb to these toxins, perhaps byinhibition of the sodium pump with consequentwaterlogging of the cells.

Theories which may explain the destructive actionof paraquat on the lung do not adequately explain thesubsequent development of pulmonary fibrosis.Vijeyaratnam and Corrin (1971) suggested that thisarose simply from organization of acute inflam-matory exudate. However, chronic administrationof paraquat to rats induces an extensive fibrosisfollowing only a slight fibrinous exudate (Smithet al, 1974). Also, Gage (1968) had shown thatinhalation of paraquat aerosols by animals causesan extensive exudate of oedema and fibrin into thealveoli but that pulmonary fibrosis never results fromthis. It seems that the formation of a fluid exudateduring the destructive phase of paraquat poisoning

90



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


is not necessary for the initiation of the proliferativephase which is independent of it.

In an attempt to explain the fibrogenic nature ofparaquat, Conning et al (1969) studied organcultures of lung. They found that paraquat causesextensive necrosis of alveolar cells but not ofmesenchymal structures. Work with unicellularcultures showed that alveolar and peritonealmacrophages are more readily killed by paraquatthan are fibroblasts. Furthermore, the addition ofmacrophages treated with paraquat to a culture offibroblasts resulted in a more rapid proliferation ofthe latter. It appears therefore that paraquat activelystimulates proliferation of fibroblasts and that itachieves this through the medium of macrophages.The fact that paraquat will only produce pul-

monary fibrosis when administered systemicallysuggests that it is metabolized in another organ suchas the liver and that a metabolite is at least partlyresponsible for the proliferative changes. This wouldcertainly explain why pulmonary fibrosis in animalsoften develops after the paraquat concentration inthe body has dropped below detectable levels(Daniel and Gage, 1966). In man there is often asmall residue of paraquat in the body for up to 15days after ingestion (Carson, 1972). What thismetabolite may be, how it can stimulate an infiltra-tion of profibroblasts into the lung, and why onlythe lung is affected, are all questions which cannotbe answered in the present state of knowledge.

Histopathology of the Pulmonary Vasculaturein Paraquat Poisoning

The pulmonary fibrosis in paraquat poisoning isfrequently associated with pulmonary vascularlesions. The type of change in the pulmonary bloodvessels seems to depend upon the degree of honey-comb change in the lung (Smith and Heath, 1974b).Where there is a dense, cellular, obliterative

pulmonary fibrosis with little dilatation ofrespiratorybronchioles, the muscular pulmonary arteries showa moderate degree of medial hypertrophy (fig 17).There is also a pronounced muscularization ofpulmonary arterioles. In both these classes of vesselthere is a cellular intimal proliferation consisting offine elastic fibres and longitudinally orientatedsmooth muscle (fig 17). A minority of vessels mayshow focal atrophy of the media with splitting ofthe external elastic laminae. A slight degree ofmedial hypertrophy of muscular pulmonary arteries,muscularization of arterioles and longtitudinalmuscle in the intima have all been described in casesof hypertensive pulmonary vascular disease (HPVD)brought about by conditions of chronic hypoxia(Hasleton, Heath, and Brewer, 1968). Thus it is

91

.4.

*0~~~~~_

Fig 17 Human case ofparaquat poisoning showing amuscular pulmionary artery with hypertrophy of itsmedia and an intimal proliferation of longitudinallyorientated smooth muscle cells and fine elastic fibrils.Elastic Van Gieson x 375.

possible that pulmonary fibrosis in paraquatpoisoning produces alveolar hypoxia of sufficientseverity to induce the changes of hypoxic hyper-tensive pulmonary vascular disease, despite the factthat death occurs a mere three weeks after ingestion.Where there has been extensive honeycomb

change, the pulmonary vascular pathology is differ-ent. In the larger muscular pulmonary arteries themedia is atrophic so that individual smooth musclecells are widely separated (fig 18). External elasticlaminae commonly show splitting and fragmentation.The intimal proliferation in these vessels is acellularconsisting of anastomosing elastic fibres with finecollagen fibres in its interstices (fig 18). Longitudin-ally orientated smooth muscle cells are rare or absent.The medias of smaller muscular pulmonary arteriesare distinctly atrophic and fragmentation of bothelastic laminae is extensive (fig 19). In some pul-monary arteries medial atrophy and fragmentationof elastic laminae is so severe that they consist of anirregular ring of fibroelastic tissue with a diminishedlumen. Small pulmonary veins contain an extensiveintimal fibrosis of fine collagen fibrils which maytotally occlude the affected vessel. Similar changeshave been described in the pulmonary arteries ofhoneycomb lung complicating a variety of diseases(Heath, Gillund, Kay, and Hawkins, 1968).



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A

Fig 18 Fig 19Fig 18 Large muscular pulmonary artery showing atrophy of the smooth muscle cells in its media. There is an intimalproliferation consisting of elastic fibres andfine collagen fibres. EVG x 232.Fig 19 Muscular pulmonary artery showing extensive medial atrophy andfragmentation ofboth elastic laminae.EVG x 375.

Conclusions

The pathogenesis of paraquat lung involves adestructive phase and a proliferative phase, and thesetwo stages appear to be independent of one another.The destructive phase consists of swelling and frag-mentation of the alveolar epithelium followed byalveolar oedema and an acute inflammatory exudate.In rats the proliferative phase commences with aninfiltration into the alveolar spaces of profibroblastswhich then mature via a series of stages into maturefibroblasts to produce a diffuse intraalveolar fibrosis.There is evidence that the proliferative phase in manhas a similar pathogenesis. The pulmonary fibrosis inhuman paraquat lung is not therefore a type offibrosing alveolitis, but is instead a diffuse, cellularintraalveolar fibrosis.Although various theories have been postulated

to explain the destructive action of paraquat in thelung, there is no satisfactory explanation for thedevelopment of pulmonary fibrosis. This has made

treatment of the condition extremely difficult. Thecellular nature of the fibrosis plus its rapid develop-ment probably explains the lack of success of steroidtherapy and this treatment is no longer recom-mended. A degree of success has been claimed fromthe combined use of steroid and cyclophosphamide(Lancet, 1971). Treatment usually consists of attemp-ting to remove the paraquat from the body beforethe damage is done. Thus emetics, forced diuresis,and haemodialysis are common forms of treatment.Because paraquat is adsorbed on to fine particles,kaolin or fuller's earth is sometimes administered.However, once the fibrotic changes are initiated,they progress inexorably and there is little thatclinicians can do about it. Until the mechanism ofparaquat toxicity is more fully understood we canonly search blindly for an antidote. In the meantimethe onus is on the general public not to decantGrammoxone from the containers in which it issupplied and thereby reduce the chances of accidentalingestion of the compound.

92



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


References

Almog, C., and Tal, E. (1967). Death from paraquat after subcutan-eous injection. Brit. med. J., 3, 721.

Bullivant, C. M. (1966). Accidental poisoning by paraquat: report oftwo cases in man. Brit. med. J., 1, 1272-1273.

Campbell, S. (1968). Death from paraquat ina child. (Letter.) Lancet,1, 144.

Carson, E. D. (1972). Fatal paraquat poisoning in Northern Ireland.J. forens. Sci. Soc., 12, 437-443.

Clark, D. G., McElligott, T. F., and Hurst, E. W. (1966). The toxicityof paraquat. Brit. J. ind. Med., 23, 126-132.

Conning, D. M., Fletcher, K., and Swan, A. A. B. (1969). Paraquatand related bipyridyls. Brit. med. Bull., 25, 245-249.

Daniel, J. W., and Gage, J. C. (1966). Absorption and excretion ofdiquat and paraquat in rats. Brit. J. ind. Med., 23, 133-136.

Fennelly, J. J., Gallagher, J. T., and Carroll, R. J. (1968). Paraquatpoisoning in a pregnant woman. Brit. med. J., 3, 722-723.

Gage, J. C. (1968). Toxicity of paraquat and diquat aerosols generatedby a size-selective cyclone: effect of particle size distribution.Brit. J. ind. Med., 25, 304-314.

Gardiner, A. J. S. (1972). Pulmonary oedema in paraquat poisoning.Thorax, 27, 132-135.

Hasleton, P. S., Heath, D., and Brewer, D. B. (1968). Hypertensivepulmonary vascular disease in states of chronic hypoxia. J.Path. Bact., 95, 431-440.

Heath, D., Gillund, T. D., Kay, J. M., and Hawkins, C. F. (1968).Pulmonary vascular disease in honeycomb lung. J. Path. Bact.,95, 423-430.

Lancet (1971) Editorial. Paraquat poisoning. Lancet, 2, 1018-1019.McCann, B. G., and Brewer, D. B. (1974). A case of desquamative

interstitial pneumonia progressing to 'honeycomb lung'.J. Path., 112, 199-202.

Manktelow, B. W. (1967). The loss of pulmonary surfactant inparaquat poisoning: a model for the study of the respiratorydistress syndrome. Brit. J. exp. Path., 48, 366-369.

Matthew, H., Logan, A., Woodruff, M. F. A., and Heard, B. (1968).Paraquat poisoning-lung transplantation. Brit. med. J., 3,759-763.

93

Nash, G., Blennerhassett, J. B., and Pontoppidan, H. (1967). Pulmon-ary lesions associated with oxygen therapy and artificialventilation. New Engl. J. Med., 276, 368-374.

Oreopoulos, D. G., Soyannwo, M. A. O., Sinniah, R., Fenton, S. S. A.,McGeown, M. G., and Bruce, J. H. (1968). Acute renal failurein case of paraquat poisoning. Brit. med. J., 1, 749-750.

Robertson, B., Enhorning, G., Ivemark, B., Malmqvist, E., and Mod6e,J. (1971). Experimental respiratory distress induced byparaquat. J. Path., 103, 239-244.

Samman, P. D., and Johnston, E. N. M. (1969). Nail damage associa-ted with handling of paraquat and diquat. Brit. med. J., 1,818-819.

Scadding, J. G., and Hinson, K. F. W. (1967). Diffuse fibrosingalveolitis (diffuse interstitial fibrosis of the lungs). Thorax,22, 291-304.

Sevitt, S. (1974). Diffuse and focal oxygen pneumonitis. J. clin. Path.,27, 21-30.

Smith, P., and Heath, D. (1974a). The ultrastructure and timesequence of the early stages of paraquat lung in rats. J. Path.,114, 177-184

Smith, P., and Heath, D. (1974b). Paraquat lung: a reappraisal.Thorax, 29, 643-653.

Smith, P., Heath, D., and Kay, J. M. (1974). The pathogenesis andstructure of paraquat-induced pulmonary fibrosis in rats.J. Path., 114, 57-67.

Smith, P., Kay, J. M., and Heath, D. (1970). Hypertensive pulmonaryvascular disease in rats after prolonged feeding with Crotalariaspectabilis seeds. J. Path., 102, 97-106.

Swan, A. A. B. (1968). Ocular damage due to paraquat and diquat.(Letter). Brit. med. J., 2, 624.

Swan, A. A. B. (1969). Exposure of spray operators to paraquat.Brit. J. ind. Med., 26, 322-329.

Toner, P. G., Vetters, J. M., Spilg, W. G. S., and Harland, W. A.(1970). Fine structure of the lung lesion in a case of paraquatpoisoning. J. Path., 102, 182-185.

Vijeyaratnam, G. S., and Corrin, B. (1971). Experimental paraquatpoisoning: a histological and electron-optical study of thechanges in the lung. J. Path., 103, 123-129.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

.&r The poisoningThepathologyofthe lung inparaquatpoisoning paraquat poisoning as an experimental...

Documents

Transcript of .&r The poisoningThepathologyofthe lung inparaquatpoisoning paraquat poisoning as an experimental...