Document dkYLk6abRE4RVk9b89MEmwE0

Original Contributions JAMA The Journal of the American Medical Association Fa S. IM3 347. NO S Prevalence and Incidence of Benign Asbestos Pleural Effusion in a Working Population Gary R. Epfer, MO, MPH; Theresa C. McLoud, MO; Edward A. Gaensler, MO Benign asbestos effusion was defined by (1) exposure to asbestos, (2) confirmation by roentgenograms or thoracentesea, (3) no other disease related to pleural effusion, and (4) no malignant tumor within three years. There were 34 benign effusions among 1,135 exposed workers compared with no otherwise unexplained effusions among 717 control subiects. Prevalence was dose related with 7.0%, 3.7%, and 0.2% effusions with severe (III), Indirect (II), and peripheral (I) exposure, respectively. The latency period was shorter than for other asbestos*related disorders. Benign effusion was the most common asbestos-related abnormality during the first 20 years after exposure. Incidence studies showed B.2 effusions per 1,000 person-years for level ill exposure, 3.B for level II, and 0.7 for level I. Most effusions were small; 28.6% recurred, and 66% were aaymptomatic. There was one mesothelioma six years after effusion. Asbestos exposure should be carefully searched for in patients with "Idiopathic" pleural effusion. (JAMA 1982;247:617-622) AMONG the several asbestos-related pleural manifestations, benign effu sion is the most recent to come to our attention.' Though still considered a rare complication of asbestos expo sure, the clinical and histological fea tures of such effusions are now well described.111 However, these case re ports do not permit conclusions con cerning frequency in asbestos-ex posed populations. We studied serial roentgenograms and medical records of 1,135 employees in the asbestos industry and compared these with the records of 717 control subjects to determine (1) the prevalence and inci dence of asbestos effusions, (2) the duration and amount of exposure associated with effusion, (3) the rela tionship between effusion and diffuse pleural thickening, and (4) the fre quency of subsequent mesothelioma. From the Evant Memorial Department ol Clini cal Research. Dacanmanl of Medicine (Dr Epler) and the Thoracic Sarvtcaa, Department of Surgary (Dr GaanataO, Boaton University MadJcal Canfar. and tha Department of Radiology. Miaaacftuaatta Ganaral Hospital and Harvard Madleal School (Dr McLoud). Boston. Raprinl rsqusata 10 SO E Concord 3t. Boston. MA 031 IB (Or Gasnslar). MATERIALS AND METHODS Definition Benign asbestos effusion was defined by four criteria, as follows: (1) direct or indirect exposure to asbestos, (2) an effu sion confirmed by a transient pleural change in serial chest films or by thora centesis, (3) lack of evidence for any other disease related to pleural effusion, and (4) no malignant tumor detected within three years after the effusion. Study Populations Two groups were selected: a "survey group" of working exposed persons seen by us at yearly intervals at industrial sites and a "control group" of male employees from s large university. The survey group consisted of 1,135 employees, including 45 women, from six plants: 1. Shipyard A (new-ship construction): 93 directly exposed pipecoverers and 74 indirectly exposed shipfitters first seen in 1965. Asbestos had been used regularly since 1930. 2. Shipyard B (submarine new construc tion and refitting): 317 employees, includ ing directly exposed pipecoverers and sweepers and indirectly exposed welders, lead bonders, and others seen annually since 1976. Asbestos had been used from 1952 to 1975. 3. Fireproofing product manufacturing: 144 employees, some heavily exposed since the early 1930s, seen annually for the last ten years. 4. Mill A (specialty paper): 90 employees heavily exposed to erocidolite in the manu facture of filter paper during a special project between 1952 and 1956 and 126 employees with slight exposure, seen since 1971. 5. Mill B (specialty paper): 211 employ ees manufacturing filter paper and gas kets, who had slight but strictly monitored exposure since 1968, seen annually for the past six years. 6. Mill C (specialty paper): 80 employees exposed to "bonded asbestos" used for electrical insulation since 1930, seen for the last three years. The control group consisted of faculty and employees of a large university. Serial films, some dating back to 1940, were available because of a law requiring roent genograms of school employees before employment and every three years there after. We located an active "three-year recall list," which excluded persons who had been exposed to beryllium or asbestos at the university. We coded all films of the 717 male employees on this list according to methods used during our industrial surveys. For comparison of clinical features and follow-up data, we referred to a group of 178 asbestos exposed clinical patients (Ta ble 1). These included 26 with benign pleural effusion, 12 of whom have been described in detail.' Method* of Investlgotton The survey group answered a respirato ry questionnaire, had a physical examina tion, screening pulmonary function studies consisting of force vital capacity (FVC) and its time derivatives, including a forced expired volume in 1 s, and also a single breath diffusing capacity (Dab)." Occupa tional histories, obtained by specially trained physicians, included a listing of all past jobs and exposures, present job description, year first exposed, and total years exposed. Dust exposure had been JAMA. Feb 6. 1982--Vol 247, No. 5 Benign Aabeatoa E 001510 monitored with variable consistency and duration, while instrumentation, manner of counting, and sampling sites had also varied over the years. Therefore, for this study we assigned employees to one of three exposure levels that were baaed in part on job description, in part on dust or fiber counts, and in part on personal observations. Generally, employees with peripheral exposure, such as administra tors, clerks, and secretaries, were consid ered to have level. I exposure; indirectly exposed employees, such as certain electri cians, welders, mechanics, shipfitters, or machine operators, who worked in the plant were placed in the level II group; and the level III group consisted of directly exposed pipecoverers, asbestos mixers, or sweepers. Chest roentgenograms at the time of study and ail available previous films were read twice, first by a group of two to three chest physicians who knew the occupation al history, and later by a chest radiologist who was unaware of the history. Films were coded according to the ILO U/C 1971 international classification of radiographs of the pneumoconioses." This allowed for description of pleural disease, such as extent and thickness of plaques, calcifica tions, and diffuse thickening. Blunting of costophrenic angles was also recorded. A separate notation was made for "effusion" only if (1) there had been hospitalization with thoracentesis or (2) when effusions were obvious from transient roentgenographic changes or the sudden appearance of blunting of the lateral or posterior costophrenic angles with a miniscus typi cal of pleural effusion. RESULTS The survey and control groups were comparable in most respects. In the survey group, 96.0% were men, and for control subjects we selected only men. The mean age of the survey group at the first visit was 41.7 years and of the control group at the time of reading, 42.0 years. Although most of our control group were not profes sors but, rather, laboratory techni cians, maintenance personnel, and grounds workers, the university envi ronment admittedly is different from Table 1.--Pleural Effusions in Survey and Control Groups and Exposed Clinic Patients CtaeaMcatton Total Pleural Effusions Known diaoaoo rotated Moaothotloma Lung cancer Pnoumonto or ampyoma Cheat surgery Trauma Spontanaoua pnoumothorax Tuborcutoaia Conpoativa haart taHuro Bonign aabottoo offuaion Survey Oroup, n-1,135 54 Control Subjects, nT1T 7 40 20 62 22 4 20 0 0 36 0 Eipneeit CSnic Patients. -17B 45 16 2 O 1 0 Q 0 0 26 that of large shipyards or factories. Nevertheless, other than for an increased amount of trauma that may cause effusion, there seemed to be no important exposures except for asbes tos, which have been related to pleu ral disease. Initial roentgenograms > the oldest asbestos workers dated back to 1935; for the university group, to 1940. The entire survey group was presumed to have had asbestos expo sure, however slight; while among the control subjects, those who had been exposed to asbestos at the university were excluded. A few control subjects did have previous exposure, largely because of the many shipyards in this area. Indeed, 13 control subjects (1.8%) -had typical pleural plaques, four with calcifications. Occupational histories of these 13 showed that eight had previously worked in ship yards or as Navy machinists or steamfitters. Diffuse pleural thicken ing was seen in two controls, a former shipyard worker and a former ship's engineer. Prevalence of All Effualone Pleural effusions of all kinds were five times more common in the survey group (4.8%) than in the control subjects (1.0%), a highly significant difference (i><.001) (Table 1). This was in part because of a greater number of asbestos effusions and in part because of effusions related to mesothelioma and lung cancer in the survey group. Cancer-related effu sions were even more common in our 178 asbestos-exposed clinic patients. There were 19 (10.7%) with effusions Fig 1.--Chest roentgenogram ol 55-year-old asymptomatic shipyard pipecoverer aince 1969 Yearly films to January 1973 (left) were normal. During next 13 months asymptomatic left pleural effusion developed, resulting in diffusa pleural thickening in March 1974 (center). During next year asymptomatic right effusion with residua seen in March 1975 (right). No change occurred during next five years. 618 JAMA. Feb 5, 1982--Vol 247, No. 5 Benign Asbeatoa Effusion--Epler et al 001511 18-0000092 related to obvious disorders, 18 of which were pleural or pulmonary malignancy (Table 1). Prevalence of Benign Asbestos Effusion The overall occurrence of one or more asbestos effusions in the survey group was 3.1% (Fig 1 and 2), while there were no otherwise unexplained effusions in the control group (Table 1). " Oose Response In the first four plants listed in Table 2, where there were many employees at each exposure level, there was an obvious relationship between exposure and prevalence of asbestos effusions. These prevalences ranged from 7.2% to 14.3% at level III and ranged from 0% to 4.3% at level I. In the fifth plant, mill B, where exposure had been recent, slight, and carefully controlled, there were no documented effusions. In mill C, there were few level III exposed employees left at the time of our first survey. For the entire survey group, the prevalences of asbestos effusion were 7.0%, 3.7%, and 0.2% at expo sure levels III, II, and I, respectively. There was also a relationship to occupation (Table 3). Asbestos effu sions were most common among asbestos pipecoverers (7.3%) (Fig 1 and 2), less common in asbestos prod uct and paper machine operators (5.3%), and least common among shipfitters, maintenance personnel, and welders. The fact that effusions were seen in level II exposures and in an office executive (Table 3), and have been seen in the wife of an employee," suggests that the requisite exposure threshold may oe low. Litency The latent period, that is, the inter val between first exposure and clini cal evidence of disease, is shown for several asbestos-related disorders in Fig 3. Only persons exposed at level II or III were included, because only for these was onset of exposure precisely known. There were no cases of asbestosis, pleural plaques, or calcifica tions during the first ten years after initial exposure. These three manifes tations increased steadily over subse quent years. The latent period for benign asbestos effusion was shorter. Fig 2.--Chest roentgenograms of 51-year-old eftipyard pipecoverer had been normal tor 16 years (left). Three months later, pleuritic chest pain developed on right side, as well as pleural-based density (right). Exploratory thoracotomy for "mesothelioma" showed encapsulated bloody effusion. Patient has remained wall for throe years. Table 2. --Prevalence and Incidence of Asbeatoa Effusion Industry and Exposure Level Shipyard A m H Shipyard S 1 1 Asbestos products 1 Paper miN A M 1 Papar mill B I 1 Papar mid C i \ Survey Group 1 Total Total Control Subjects No. of fpployvM S3 74 136 79 112 27 71 46 70 20 126 s IS 187 4 30 40 320 296 S11 1.13S 717 Asbeatoa Effusion, No. (%) 0(6.2) 1(1.4) S(7.D 3(3.8) 0(0.0) 4(14.8) 6<8.S> 112-2) 4(8.7) 0(0.0) 0(0.0) 0(0.0) O(CLO) 0(0.0) 0(0.0) 1(2.8) 0(0.0) 23(7.0) 11(3.7) 1(02) 38(3.1) 0(0.0) ObssmsMun, ^MOa^VMft 672 484 1283 an 408 280 1010 230 322 79 288 80 aa 416 23 230 78 2,800 2.708 1.421 8.718 S.B33 Date per 1,000 Person-Years 10.8 2.0 7.2 3.3 ' 0.0 14.3 8.8 4.3 12.4 0.0 0.0 0.0 0.0 0.0 0.0 4.4 0.0 9.2 3.8 0.7 8.2 0.0 JAMA, Feb S. 1982--Vol 247, No. S Benign Asbestos Effusion--Epler at al 00151.2 ia-0000093 619 Table 3 --Job Descriptions of Employees* exposure Level III a Joe Title Ptp4COvf*rt. aftMttoa Asbestos mixer Asbestos sweepers - Msdime operators StUpStters Mssttanance Olliers Machm* operators* Pipecoverers. ibargless OlHee Others No. of 191 110 28 114 41 31 108 206 71 58 175 N-1.136. tOw ssctv welder. lesd bonder, bailor supply room worker. (Old not work with asbestos. 60 )- 50 U 2a 40 uEr 30 ui o c 20 u 10 Asbestos Effusion Pulmonary Fibrosis Plaural Plaques Plaural Calcification 166 112 effusion. No. (%) 14(7 3) 7 (6.4) 2(7 1) 8(5 3) 1 (2 4) 1 (3.2) 3t(2.8) 0(0.0) 0(0.0) 1(1.7) 0(0.0) 19 m 3-9 10-19 40-49 Years Since First Exposure Fig 3.--Asbestos-related manifestations smong employees exposed to asbestos at level II or III. Grouping is according to years since first exposure. Number of employees in each group indicated above columns. Benign asbestos effusion was observed earlier than other manifestations: it wss the only asbestos-related disorder observed during first tan years after exposure and was most common during first 20 years. It was the only manifestation seen within ten years, and it was the most common abnormality during the first 20 years (Fig 3). The possibility that latency is dose related could not be proved by these data. Although the latent period was shorter (13.3 years) for the 23 workers exposed at level III than for the 11 employees exposed at level II (15.2 years), this difference was not statistically significant Incidence Serial chest roentgenograms were svailable for up to 45 years. There fore, the number of new asbestos effusions per 1,000 person-years of observation could be calculated (Ta ble 2). This analysis indicated an annual occurrence of nine asbestos effusions per 1,000 employees exposed at level III, four for those at level II, and fewer than one per 1,000 for those at level I. Clinical Findings The mean age at the time of the first benign asbestos effusion was 46.1 years, and two workers were only 28 years old. Eight also had chronic bronchitis, one had asthma, and four had had childhood pneumonia. Con current medical problems included hypertension in three, coronary ar tery disease in three, and diabetes in one. In none was there any evidence of tuberculosis at the time or subse quently. Two thirds reported no symptoms during their effusion, even when told of their condition (Fig 1) (Table 4). In contrast, more than one half (53.8%) of the 26 clinic patients with asbestos effusions had pleuritic pain. Roentgsnographic Features Most asbestos effusions were small (Fig 1), and a few presented bilateral ly (Table 4). Plaques were seen in one fifth, calcifications in only one, and moderate to severe asbestosis in less than 10%. All of these three manifes tations were uncommon because they are usually late complications, while asbestos effusion often occurs rela tively early (Fig 3). Follow-up films showed blunted costophrenic angles in virtually all cases, and residual diffuse pleural thickening was Been in one half (Table 4). Follow-up and Prognosis The mean follow-up after initial asbestos effusion was 9.7 years, with a range of three to 27 years. Recur rent benign effusions developed in ten persons (28.6%), sometimes on the same side, more often on the opposite side (Fig 1). Physical findings con sisted of bilateral crackles in one third, and there were pleural friction rubs in two. The FVC was reduced in one half, and Dsb was reduced in two thirds. Few had evidence of airflow obstruction (Table 5). In'the survey group three persons have died, two from asbestosis, and one from a mesothelioma that devel oped six years after the firBt effusion. Mortality was greater among our 26 clinic patients with asbestos effusion. Five have died: two from infections, one from metastatic hypernephroma, and two from mesothelioma nine and 16 years after the first effusion. COMMENT Asbestos and other fibrous silicates are virtually unique among environ mental hazards as a cause of pleural manifestations. Hyaline plaques were readily associated with asbestos clini cally, because of the almost invariable history of exposure, however distant or brief, and epidemiologically, be cause of their exceedingly high preva- 620 JAMA, Fob 5. 1982--Vol 247, No. 5 Benign Asbestos Effusion--Epler et si 0 01c) 13 18-0000094 Table 4 -- Clinical and Roentgenograpftic Features of Benign Asbeatoa Effusion Table S.--Follow-up Oats Three or Mora Years After Aabeatos Effusions FMturt Ho. % Clearly, among these 35 effusions there may have been some that were related to other perplexing causes rather than to asbestos. However, the Feature MO. % Physical finding* absence of such cases among the I Presenting symptoms N O---n-g Pteuntte pan n 66.7 17.1 Clubbed fingers 4 11.4 Localized duffness 4 11.4 BSatsrsi fine crackles 1? 34.3 control subjects suggests that such problems must be relatively rare. Dyspnea 3 Pleural friction rub 2 6.7 The prevalence of 3.1% asbestos "Pneumonia" Coid symptoms Hmopiyits Cigarette use * Smokers Ex-emeker* MOftefltOfearS Roentgenograpnic Initial effusion Large. >500 mL Bilateral 2 5.7 1 2.9 0 0.0 18 51.4 14 40.0 3 5.8 4 11.4 3 8.6 Pulmonary function* PVC<90% predicted FVC, meantSO FEV,/FVC% <70 FEV,/PVC%. . meemSO Oab <80% predictad OeP. meantSO Recurrent iffunona, antra observation pohod 17 74.8t7S 13 72.0113.7 20 8.2*20.7 10 48.5 37.1 87.1 286 effusions undoubtedly represents an underestimate. Sometimes the first available roentgenograms were ob tained many years after initial expo sure, and then an initial finding of a blunted costophrenic angle or of dif fuse pleural thickening was not accepted as evidence of asbestos effu Signa oi aabeetoa xpoaura Plaques CeictScanone 7 20.0 1 2.9 PVC indicataa forced vital capacity; PEV,, fofcad axprad volume nil; Dad, braatti dtffuamg capacity. sion. Also, lateral films were some times missing, and oblique films were rarely available. Therefore, residua of Parenchymal 'tHroata" Slight {a.t.u 1/01/2)* Moderate to severs 15 42.9 localized effusions might have been and France.14'11 In most of the total 37 missed. Finally, some persons might (s.t.u 2/1-3/4)* RMKJual pleural findings Blunted anoie 3 9.6 aa 91.4 reported cases141 and in our 12 previ have had an effusion that disap ously reported,' there was physical peared without residua during the Diffuse thiefcerung 19 54.3 evidence of asbestos exposure from interval between films. This raised *ILO U/C elaanlcatton* biopsy specimens that showed pulmo the possibility of an underestimate of nary fibrosis,'-14'''1' asbestos bodies in the prevalence of effusions, in the lence among exposed workers. Meso the lung,'44'1" or from asbestos bod control group because there were few thelioma, another pleural disease, ies in the sputum.*1'0 Asbestos bodies er films, usually every third year, was firmly related to asbestos by a in the pleura were uncommon,14"11 while in the survey group films were single clinical study when this rare sometimes detected by x-ray diffrac obtained annually at least during tumor appeared with great frequency tion or electron microscopy,1" and recent years. However, most effu in certain locations." The significance they were not seen in the pleural sions, that is, 91.4% (Table 4), left a of pleural effusion with respect to fluid. residual blunted angle so that such a asbestos exposure was more difficult These case reports are uninforma finding is an important marker for to evaluate, because effusion, unlike tive concerning prevalence of asbestos past effusion. Among our 1,135 em plaques and mesotheliomas, is a com effusion. Also, it has been pointed out ployees in the survey group, there mon complication of a large number that conclusions concerning the rela were 92 who had blunted angles, 54 of of disorders. tionship of any two conditions must which were recognized as evidence of The first suspicion, in 1962, of a be made with caution, and that the effusions (Table 1), whereas in the relationship between asbestos expo more common they are, the greater control group there were eight per sure and effusion was based on histo the likelihood of error.'1 Clearly, both sons with blunted angles, seven of logical findings: an insulation worker pleural effusion and asbestos ex which were counted as evidence of with bilateral recurrent effusions was posure are extremely common. An effusions (Table 1). Therefore, if the found to have pulmonary fibrosis analogy may be drawn to the long- smaller number of films in the con with asbestos bodies as well as hya suspected relationship between effu trol group caused us to miss silent line plaques.1 Over the ensuing nine sion and rheumatoid arthritis-- effusions without roentgenographic years, additional observations of "ex another frequent condition. Here the residua, there certainly were not udative pleuritis" were described in answer came from comparison of many such eases. ten exposed persona.14 About that rheumatoid patients with a control Two distinct types of pleural reac time, we saw an otherwise healthy group with degenerative arthritis: tions are seen in the asbestos exposed: patient with recurrent bilateral pleural effusion was ten times more plaques and diffuse pleural thicken bloody effusions who eventually re prevalent in the former.'1 This sug ing. A relationship between diffuse quired decortication. Again, asbestos gested our comparison of the survey thickening and asbestos effusion has bodies and fibers were found in the group with a control group drawn not been suggested previously. How lung and this time also in the pleura. from the general population. It ever, it has been recognized that This prompted a review of clinical showed that, over a similar observa diffuse thickening differs from and histological material of 11 other tion period, effusions of any cause plaques in most respects. The costo patients with asbestos exposure and were five times more common in the phrenic angles are commonly in effusion.' Since then, 27 additional exposed group and that there were 35 volved, there is pleural symphysis, cases of asbestos effusion have been pleural effusions not related to other fibrosis is common, and there is usu reported from the United States,1 disease among the exposed but none ally loss of function.10 Indeed, by 1970, South Africa,1 Australia,1 Hungary,'1 in the control subjects (Table 1). diffuse thickness was so well recog- JAMA. Feb 5. 1982--Vol 247, No. S Benign Asbestos Effusion--Epler et si 821 nized that the International Classifi cation for the Pneumoconioses al lowed for a distinction from plaques.'0' Our longitudinal observa tions suggest that the sudden appear ance nf diffuse thickening frequently was caused by an effusion. Among our 35 workers with asbestos effusion, 54% had residua in the form of diffuse thickening (Fig 1). Further more, among the 1,135 employees, there were 44 with diffuse pleural thickening greater than 5 mm, and of these, almost one half had had a previous asbestos effusion. By con trast, there were 127 with typical plaques, but the development of the plaques never seemed to be related temporally to the effusion. Mesothelioma, the other well-rec ognized asbestos-related pleural com plication, generally is thought of as a rapidly growing and quickly fatal tumor. Therefore, we excluded from our count of benign asbestos effusions all persons who had a follow-up of less than three years. However, there has been a report of malignant meso thelioma of 17 years' duration,a and others have indicated that effusions sometimes occurred several years be fore the histological confirmation of a mesothelioma.0' In our "survey group," one person was recognized to have a mesothelioma six years after his initial "benign" effusion, and in our clinic group, two had such a tumor nine and 16 years after initial effusion. There are several possible explanations: inasmuch as both a benign effusion and mesothelioma are relatively common in the asbestos exposed, it may be that the two 1. Gaensler EA. Kaplan At Asbeitoa pleura) effusion. Ann Intern Med 1971,74.178-191. 2. EiMnsudt HR Pleural aabeeteeia. Am Pract 1962:18573-578 3. Eisenstadt HE' Benign aebeetoa pleurisy. JAMA 1965;192419-tZl. 4. Collini TFR Pleural reaction associated with eabeetoe expoeare. Br J Radiol 1968;41:656661. 5. Mattson S, Ringqviit T Pleural plaques and exposure to eabeetoe. Stand J Reepir Die 1970; 75isupplhl-41. 6. Smyth NPD, Goodman NG, Baiu AP, at al: Pulmonary aabestoeia Cheat 1971;60:270-271 7. Sluie-Cremer GK, Webeter I; Acute pleuriij in aabantoe expoead persona. Environ Ret 1972*380-392. 1 Elder Jit A etody of 16 eases of pleurisy with effusions in ex-miners from Wittanoon Gorge. And NZ J Med 19722328-329. 9. Chahinian P, Hirach A, Bignoo i, at ah Lea pleurisies asbestoeiques non tumoralea. Rev Ft Mai Reepir 1973;1:5-39. disorders occurred in the same pa tient in nonrelated fashion. It is also possible that the pleural drift of asbestos fibers caused mechanical irritation resulting first in effusion and eventually in mesothelioma. Fi nally, it is possible that the earlier effusion was the first manifestation of the tumor. Mesothelioma grows along interstitial planes and is de tected roentgenographically only late, and, therefore, its "doubling time" cannot be determined. It well may be that in some patients this lesion initially grows rather slowly. The spontaneous cessation of pleural exu dation, the disappearance of most roentgenographic residua, and the usually long-term stable course there after suggest that most of our "be nign asbestos effusions" were indeed benign. However, our follow-up was not long enough to determine the incidence of mesothelioma after as bestos effusion. Historically, idiopathic pleural ef fusion was generally attributed to tuberculosis, and, among younger persons, two thirds eventually showed development of active disease.11 Even now some textbooks and reviews on pulmonary disease deal at length with the concept of idiopathic effu sion presumably due to tuberculosis and make no mention of benign asbestos effusion. Our study suggests that in the general population, effu sions without immediately apparent cause have become extremely rare, while in the asbestos exposed, they are relatively frequent (Table 1). In our ambulatory consultation practice, asbestos exposure has become the References 10. Bootin C, Viallat J, Farisae P. et al: Plsurisies ssbestoaiques Meignet. Poumon Caeur 197*31:111-118. 11. Lemenager J. Rousselot P, LBouffant L, et al: let pleoresies btnignea da I'aminants. Rev Fr Mai Reepir 1975*325-342. 12. Nyiredy G. Benign aabeetoe-induced pleu risy. Pros Pneumai 197559:166-169. 13. Moignetcau Ch, Touxeao PY, GuiUement J: Calcifications pleurale* aebeatoeiqsea at patho logic sseociee. Poumon Coeur 197753:101-106. 14. Ferria BG: Epidemiology standardisation project. Am Rev Reepir Die 1978;118:7-88. 15. ILO U/C 1971 international classification of radiograph! of the pneqmoeonioeee. Med Radiogr Phatogr 1972:4867-76. 16. Epier GR, FitxGermld MX, Gaenaier EA, et al: Asbestos-related diaeaae from houaehold exposure. Reepiration 1980.39:229-240. 17. Wagner JC, SUgga CA, Marchand P Dif fuse pleural mesothelioma and asbestos expolure in the North Western Cape Province. Br J Induetr Med 1960,12260-271. most common cause of pleural effu sion.' Mesothelioma presents the principal differentia] diagnostic prob lem to benign asbestos effusion. Our experiences and those of others sug gest that benign effusion often occurs relatively soon after initial exposure, sometimes within ten years (Fig 1) and in two thirds within 20 years' 0 (Fig 2 and 3), while mesothelioma usually is seen more than 20 and often 30 to 40 years later. Pleuritic pain is the most frequent symptom in both conditions though benign effu sion goes entirely unnoticed by more than one half of patients (Table 4). Roentgenographic signs of other as bestos-related disease such as plaques or pulmonary fibrosis (asbestosis) may be absent both in benign asbes tos effusion and mesothelioma, and the presence of plaques or asbestosis is not helpful in differentiating between the two. It has been estimated that there are somewhere between 2 million and 6 million persons with significant asbestos exposure in the United States.11 A prevalence of 3.1% benign effusions in the asbestos exposed should alert physicians to the impor tance of this disorder in the differen tial diagnosis of "idiopathic" effu sion. This study was supported in part by program project gnat HL 19717 and career award HL lira from the National Heart, Luna nod Blood Institute, National Institutes of Health, Public Health Service. The authors wish to thank Melvin H. Rodman. MD, for permission to inspect leriai roentgeno grami of employees of Maeaeefauietts Institute of Technology, Boston. The Warren E. Collins Co of Braintree. Maas, donated equipment for nrvey studies. 18. Rheumaioie pieoritia? editorial. Lancet 1968;1:135-136. 19. Walker WC, Wright V: Rheumatoid pleuritil. Ana Rheum Die 196726:467-474. 20. Zigging A, Huutkonen MS, Aianko K, et aL Radiographic and physiologic findings is pa tients with asbestosis. Stand J Work Environ Health 1978;4:275-288 21. Bohlig H. Bristol LI, Cartier PH, et al: UICC/Cincinnati classification of the radio graphic appearances of pneumoconioses: A coop erative study by the U1CC committee. Cheat 1970,5857-67. 22. Helletrom PE, Freinaa C. Tcppo L Malig nant mesothelioma of 17 yean' duration with high pleural fluid concentrations of hyaluronate. &aitd J Reepir Die 19773897-102. 28 Roper WH, Waring JJ: Primary teroflbrinous pleural effusion in military personnel. Am Rev Tvberc 1965,-71.-616-684. 24. Leman RA, Dement JM, Wagoner JK: Epidemiology of asbestoe-related disease. Envi ron Health Penpect 1980*4:1-11. 022 JAMA, Feb S, 1982--Vo! 247, No 6 Benign Aabeatoa Effusion--Epier et al 001 18-0000096 IF